Descubrimiento de un nuevo y potente péptido dirigido a KRAS(G12V) con actividad antiproliferativa contra células de cáncer colorrectal.

Colorectal cancer (CRC) remains one of the leading causes of cancer-related morbidity and mortality worldwide, imposing a substantial public health burden^[1],^[2]. As a pivotal proto-oncogene, Kirsten rat sarcoma viral oncogene homolog (KRAS) drives CRC tumorigenesis and progression when mutated^[3]. Codon 12 of KRAS is a mutation hotspot that is critically implicated in CRC tumorigenesis. Among KRAS codon 12 mutations, G12V (30%), G12C (11–13%), and G12D (35%) are the most prevalent variants in human cancers^[4],^[5]. Multiple small-molecule and peptide inhibitors targeting KRASG12D have been developed[6–8]. However, as the second most frequent oncogenic KRAS mutant, KRASG12V remains insufficiently explored as a therapeutic target^[9]. Therefore, the development of KRASG12V-targeted agents remains an important unmet need.

As a small GTPase belonging to the RAS superfamily, KRAS modulates fundamental cellular processes including proliferation, survival, and migration^[10]. However, KRASG12V differs substantially from KRASG12D in both structure and oncogenic function. Substitution with valine at codon 12 (G12V) generates marked steric hindrance and increased hydrophobicity, stabilising a rigid, compact protein conformation which may contribute to impaired GTP hydrolysis and reduced ligandability by currently available inhibitors^[11]. Conversely, the polar, negatively charged aspartic acid in G12D provides a favourable interface for inhibitor binding, supporting the extensive development of both noncovalent inhibitors and emerging covalent agents^[12],^[13]. KRASG12V is an oncogenic KRAS mutation that has been implicated in dysregulation of extracellular signal-regulated kinase (ERK) signalling, altered p21/CDKN1A (cyclin-dependent kinase inhibitor 1 A)-related cell-cycle control, and colorectal cancer progression^[11],[14–16]. Thus, the unique structural and functional features of KRASG12V make the development of novel and potent inhibitors against this mutant particularly challenging.

Currently, investigational therapies for patients with KRASG12V-mutant cancers include pan-KRAS (ON) inhibitors, pan-KRAS degraders, and KRASG12V-selective RNAi agents. Among these, RMC-5127 exerts inhibitory activity through noncovalent binding and a ternary complex mechanism, with preclinical efficacy superior to that of other clinical-stage agents^[17]. The small-molecule ACBI3 degrades KRAS proteins harbouring 13 different mutations through the recruitment of E3 ubiquitin ligases^[18]. siRNA-EFTX-G12V selectively silences mutant KRASG12V, with no effect on wild-type KRAS or other isoforms^[19]. Furthermore, LY4066434 and BBO-11818 both function as pan-KRAS inhibitors^[20],^[21]. However, most existing inhibitors exhibit poor binding stability towards KRASG12V and show modest inhibitory effects in cellular models. To date, no inhibitor specifically targeting KRASG12V has received clinical approval^[9].

Many small molecules face challenges in achieving high-affinity binding towards KRASG12V, which creates substantial obstacles in KRASG12V-targeted drug development^[22]. Furthermore, the binding interface of KRASG12V is flat and extensive, which hinders effective binding by small-molecule compounds^[23]. In contrast, peptides exhibit greater structural flexibility and can engage more extensively with the KRASG12V binding interface, thereby potentially enabling broader and more stable interactions with KRASG12V. Studies have shown that the H-REV107 peptide interacts with KRASG12V and inhibits tumour growth^[24]. Accordingly, constructing and screening a peptide library based on the H-REV107 sequence represents a rational strategy for identifying KRASG12V-targeting inhibitors. In this study, the H-REV107 peptide was used as a control.

Molecular docking-based virtual screening can markedly improve the efficiency and accuracy of identifying bioactive molecules from compound libraries. Docking scores can be used to prioritise compounds with favourable predicted binding profiles^[25],^[26]. In our previous work, we performed molecular docking-based virtual screening to identify high-affinity compounds targeting specific proteins from compound libraries^[27],^[28]. Furthermore, we previously identified the cyclic D-peptide NKTP-3 as a dual-target inhibitor of NRP1 and KRASG12D in lung cancer. Our prior investigations on KRASG12D have provided a sound experimental basis for this study^[7]. In this study, we identified novel peptide inhibitors targeting KRASG12V, among which Peptide-1 exhibited the best binding affinity and inhibitory potency.

Materials and methods

Reagents

Peptides 1–4, H-REV107, and Scrambled peptide were synthesised by GL Biochem (Shanghai, China). Recombinant human KRAS mutant proteins, including KRASG12V (cat. no. ab268713), KRASG12D (cat. no. ab314422), KRASG12C (cat. no. ab314426), KRASG13D (cat. no. ab315077), and KRASQ61H (cat. no. ab315078), were purchased from Abcam (Cambridge, UK). In addition, recombinant human KRAS mutant proteins, including KRASG12S (Cat. No. KRS-H5144) and KRASG12A (Cat. No. KRS-H51H1), were purchased from ACROBiosystems (Newark, DE, USA). The Protein Labelling Kit RED-NHS 2nd Generation (cat. no. MO-L011) and Monolith NT.115 capillaries (cat. no. MO-K022) were purchased from NanoTemper Technologies (Munich, Germany). shControl and shKRASG12V were obtained from Sigma-Aldrich (St. Louis, MO, USA). The primer sequences for ERK1 and p21 were synthesised by Sangon Biotech (Shanghai, China). TRIzol Reagent (cat. no. 15596026) was purchased from Invitrogen (Carlsbad, CA, USA). Thermo Fisher High-Capacity cDNA Reverse Transcription Kit (cat. no. 4374966) was purchased from Thermo Fisher Scientific (Waltham, MA, USA). FxCycle™ PI/RNase Staining Solution (cat. no. F10797) was obtained from Thermo Fisher Scientific (Waltham, MA, USA).### Cell culture

SW480 (cat. no. CCL-228, KRASG12V), SW620 (cat. no. CCL-227, KRASG12V), and HT-29 (cat. no. HTB-38, KRASWT) cell lines were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). NCM460 (cat. no. iCell-h373) cells were obtained from Cellverse (Shanghai, China). SNU-C4 cells (KRASG12V) were acquired from the Korean Cell Line Bank (KCLB, Seoul, Republic of Korea). RPMI 1640 medium and DMEM medium were obtained from Nanjing Senbeijia Biotechnology Co., Ltd. (Nanjing, China). Foetal bovine serum (FBS) was purchased from Gibco (Thermo Fisher Scientific, Waltham, MA, USA; Cat. No. 10099141). Opti-MEM™ was obtained from Gibco (Grand Island, NY, USA). SW480, SW620, and SNU-C4 cells were cultured in RPMI 1640 medium, whereas HT-29 and NCM460 cells were cultured in DMEM medium. All media were supplemented with 10% foetal bovine serum (FBS), 100 U/mL penicillin, and 100 μg/mL streptomycin. Cells were maintained in a humidified incubator at 37 °C with 5% CO2.### Peptide library construction

The peptide library was constructed according to previously reported methods^[7]. A combinatorial peptide library was constructed using the QuaSAR-CombiGen module in the Molecular Operating Environment (MOE) software. The library, consisting of 59,319 peptides, was generated by assembling two tripeptide fragments and one tetrapeptide fragment. The virtual library was exported as a database containing 59,319 peptides. Subsequently, energy minimisation in MOE was applied to convert the two-dimensional (2D) peptide structures in the database into optimised three-dimensional (3D) conformations.### Virtual screening

The 3D database containing 59,319 peptides was constructed for molecular docking. The crystal structure of the KRASG12V–H-REV107 peptide complex (PDB ID: 7C41, resolution: 2.28 Å) was retrieved from the Protein Data Bank (PDB) database. The complex was structurally refined using the Structure Preparation module, followed by energy minimisation. The peptide database was then docked into the active pocket of KRASG12V using the Dock module. Rigid receptor-flexible ligand docking was performed against KRASG12V. Molecular docking was performed in two stages: initial placement via the Triangle Matcher algorithm to generate peptide binding poses, followed by Rigid Receptor refinement to optimise ligand conformations. London dG and GBVI/WSA dG scoring functions were used to rapidly prioritise poses and calculate accurate binding free energies, respectively. Finally, the top four peptides ranked by docking score were selected for in vitro evaluation, and the docking score of the reference peptide H-REV107 was also recorded.### Microscale thermophoresis assay

Microscale thermophoresis (MST) was performed to determine the dissociation constant (Kd) of peptide binding to GppNHp-loaded KRASG12V, thereby evaluating peptide affinity under an active-like nucleotide-bound state. Recombinant human KRASG12V protein was loaded with 200 μM GppNHp using 5 mM EDTA-mediated nucleotide exchange procedure to approximate the active-like nucleotide-bound state^[29],^[30]. After nucleotide exchange, 20 mM MgCl2 was added to terminate the reaction and stabilise nucleotide binding. Excess unbound nucleotide was removed by ultrafiltration, and the 10 μM GppNHp-loaded recombinant human KRASG12V protein was subsequently fluorescently labelled using the Protein Labelling Kit RED-NHS 2nd Generation. After purification, the labelled protein stock was 2 μM. Peptides 1–4, H-REV107, and Scrambled peptide were prepared in PBST (Phosphate-buffered saline with Tween-20) buffer. DMSO was added to improve peptide solubility. All peptides were subjected to a two-fold serial dilution to generate 16 concentration gradients. The diluted peptides were incubated with the fluorescently labelled KRASG12V protein in the dark at room temperature for 5 min. After incubation, samples were loaded into standard capillaries, and MST measurements were carried out on a Monolith NT.115 instrument using a 1:1 binding model for KRASG12V and peptide. Measurements were carried out at 26 °C with 100% LED power, 20% MST power, and a data acquisition time of 30 s. Data processing and binding affinity analysis were performed using MO. Affinity Analysis software (NanoTemper Technologies).### Molecular dynamics simulations

Molecular dynamics (MD) simulations were performed to investigate the stability and dynamic behaviour of the KRASG12V–Peptide-1 complex under physiological conditions. The complex was parameterised using the AMBER99SB-ILDN force field, placed in a cubic water box with a minimum distance of 1.0 nm from the solute to the box edge, and solvated. Na+ and Cl- counterions were added to neutralise the system. Subsequently, the system was subjected to 5000 steps of steepest descent energy minimisation, followed by sequential NVT and NPT equilibration. Root mean square deviation (RMSD), root mean square fluctuation (RMSF), secondary structure, radius of gyration (Rg), and the number of hydrogen bonds were analysed, and data processing was performed using GraphPad Prism 9.5.### MM/PBSA calculation

MM/PBSA was employed to calculate the binding free energy from the equilibrated MD simulation trajectories. The MM/PBSA method was applied as previously described^[31]. Energy decomposition analysis was performed on the binding pocket residues of the KRASG12V–Peptide-1 complex, quantifying the contributions from van der Waals interactions, electrostatic interactions, polar solvation free energy, and nonpolar solvation energy.### Free energy landscape

Free energy landscape (FEL) analysis was conducted as previously described^[32]. Based on the equilibrated MD simulation trajectories, RMSD and Rg were selected as two-dimensional collective variables to construct the two-dimensional FEL of the KRASG12V–Peptide-1 complex, thereby characterising its conformational distribution and energetic stability. Furthermore, 2D and 3D FEL diagrams were generated to identify the dominant conformations, energy barriers, and lowest-energy states of the complex.### Serum stability assay

The serum stability of Peptide-1 was evaluated in 25% (v/v) human serum. This diluted-serum condition was used to maintain serum-derived proteolytic activity while minimising sample viscosity and matrix interference during subsequent HPLC analysis^[7]. Briefly, human serum obtained from Sigma-Aldrich (St. Louis, MO, USA; catalogue number H4522) was used in this study. The serum was collected by the supplier using standard venipuncture into serum-separating tubes, allowed to clot at room temperature for 30 min, and then centrifuged at 3,000 × g for 15 min at 4 °C to separate serum from blood cells. The resulting serum supernatant was filtered, aliquoted, and stored at −80 °C until use. Prior to the stability assay, the frozen serum was thawed, diluted to 25% (v/v) with pre-warmed phosphate-buffered saline (PBS, pH 7.4), and centrifuged at 15,000 rpm for 10 min at 4 °C to remove insoluble components. Peptide-1 was then added to the collected supernatant and mixed thoroughly to achieve a final concentration of 5 μM. The mixture was incubated at 37 °C under continuous gentle agitation to mimic physiological conditions. At predetermined time points (0, 30, 90, 180, and 360 min), 200 μL aliquots were withdrawn and immediately mixed with an equal volume of ice-cold acetonitrile, followed by incubation at 4 °C for 30 min to ensure complete protein precipitation. The samples were then centrifuged at 15,000 rpm for 10 min at 4 °C, and the clarified supernatants were collected and analysed by reversed-phase HPLC as previously described³³. The remaining amount of Peptide-1 at each time point was quantified based on the integrated peak area corresponding to the intact peptide. The percentage of remaining Peptide-1 was calculated by normalising the peak area at each time point to that at 0.0 min (defined as 100% intact peptide).### MTT assay

The MTT assay was performed as previously described^[34]. SW480, SW620, SNU-C4, HT-29, and NCM460 cells were selected to investigate the antiproliferative activity of the peptides. Cells in the logarithmic growth phase were trypsinized and seeded into 96-well plates at a density of 5 × 10³ cells per well in their corresponding complete media. The edge wells were filled with PBS to avoid the edge effect. After incubation at 37 °C with 5% CO2 for 24 h to reach 70–80% confluence, cells were treated with 16 serially diluted concentrations of each peptide for 72 h. Then, 10 μL of 5 mg/mL MTT was added to each well and incubated for 4 h. The supernatant was removed, and 100 μL DMSO was added to dissolve formazan crystals. The absorbance at 490 nm was measured using a microplate reader, and the IC50 values were calculated using GraphPad Prism 9.5.### Cellular NanoBRET assay for KRASG12V target engagement

The competitive NanoBRET target engagement assay was performed to assess the cellular engagement of Peptide-1 with KRASG12V in SW480 cells. Briefly, SW480 cells were seeded in white opaque 96-well plates at 1.5 × 104–2.0 × 104 cells per well and cultured overnight. Cells were then transiently transfected with a plasmid encoding NanoLuc-tagged KRASG12V and incubated for 18–24 h. For the competition assay, cells were treated with a fluorescently labelled Peptide-1 tracer at a fixed concentration of 0.3 μM and serial dilutions of unlabelled Peptide-1 ranging from 0.01 to 30 μM. After incubation at 37 °C with 5% CO2 for 2 h, NanoBRET substrate was added according to the manufacturer’s instructions, and donor and acceptor emission signals were recorded using a dual-wavelength luminescence plate reader. The BRET ratio was calculated as the acceptor-to-donor emission ratio, followed by background subtraction and normalization^[35]. Dose–response curves were fitted using a four-parameter logistic model to determine the apparent IC50 value.### Lentiviral shRNA-mediated KRASG12V knockdown

Transfection of shKRASG12V was performed as previously described^[7]. Polybrene was added to Opti-MEMTM medium to prepare a solution at a final concentration of 6 μg/mL. Lentiviral particles encoding KRAS-targeting shRNA or non-targeting control shRNA were separately diluted in Opti-MEM medium containing 6 μg/mL polybrene. The diluted lentiviral suspension was added to SW480 cells. After 3 days of incubation, stably transfected cell lines were selected using puromycin at a final concentration of 5 μg/mL. Additionally, knockdown efficiency was verified before subsequent MTT assays. The selected cells were then seeded into 96-well plates, and their IC50 values were determined by the MTT assay.### ERK1/2 phosphorylation assay by ELISA

Total and phosphorylated ERK1/2 levels in SW480 cells were measured using the Multispecies ERK1/ERK2 (Total/Phospho) InstantOne ELISA Kit (85–86012, Thermo Fisher Scientific). Briefly, SW480 cells were seeded in 96-well plates and cultured to approximately 70–80% confluence. Cells were then treated with Peptide-1 (1 μM) or H-REV107 (10 μM) for 12 h, while the control group received the corresponding vehicle treatment. Before cell lysis, cells were stimulated with EGF (10 ng/mL) for 10 min. After treatment, the culture medium was removed, and the cells were gently washed with PBS and lysed with freshly prepared 1× Cell Lysis Mix for approximately 10 min at room temperature with gentle shaking. Cell lysates and antibody reagent were then added to the InstantOne assay plate according to the manufacturer’s instructions and incubated for 1 h at room temperature. After washing, Detection Reagent was added for 10–30 min, followed by Stop Solution to terminate the reaction. Absorbance was measured at 450 nm with correction at 650 nm. Phospho-ERK1/2 and total ERK1/2 signals were determined separately, and the relative ERK1/2 phosphorylation level was expressed as the p-ERK1/2/total ERK1/2 ratio and normalised to the control group^[36].### Reverse transcription quantitative polymerase chain reaction

RT-qPCR was performed as previously described^[37],^[38]. After SW480 cells were treated with Peptide-1 at 1 and 5 μM for 24 h, total RNA was extracted using the TRIzol reagent, and its concentration and purity were determined. Subsequently, cDNA was synthesised from the total RNA using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher). Using cDNA as the template, qPCR was performed with gene-specific primers under standard denaturation, annealing, and extension conditions. The primer sequences were as follows: ERK1 (Forward, 5′-CCCATACCTGGAGCAGTATTATG-3′; Reverse, 5′-CTATTGTGTGCTCCCTCTCATC-3′), p21 (Forward, 5′-TTGGCCCAAGAGGTCGGCTCC-3′; Reverse, 5′-GCGGATTAGGGCTTCCTCTTG-3′), and GAPDH (Forward, 5′-CAGGAGGCATTGCTGATGAT-3′; Reverse, 5′-GAAGGCTGGGGCTCATTT-3′). The mRNA levels of ERK1 and p21 were detected using the Applied Biosystems 7500 Real-Time PCR System. GAPDH was used as the internal reference gene, and the relative mRNA expression levels of target genes were calculated using the 2-ΔΔCt method.### Flow cytometric assay of cell-cycle distribution

Cell-cycle distribution was analysed by flow cytometry after Peptide-1 treatment. SW480 cells were seeded in 6-well plates and treated with Peptide-1 at different concentrations (0.5, 2.5, and 12.5 μM) for 24 h. After treatment, cells were collected by centrifugation at 1,000 rpm for 5 min, washed twice with cold PBS, and prepared as single-cell suspensions. The cells were then fixed with 70% ethanol at −20 °C overnight. After fixation, the cells were washed with PBS and stained with PI/RNase staining solution for 30 min at room temperature in the dark^[39]. Finally, cellular DNA content was measured by flow cytometry, and the percentages of cells in the G0/G1, S, and G2/M phases were calculated using cell-cycle analysis software.### Statistical analysis

Data processing and visualisation were performed using GraphPad Prism software (version 9.5; GraphPad Software Inc., San Diego, CA, USA). All experiments were performed in biological triplicate, and data are presented as the mean ± standard deviation (SD). Statistical comparisons among groups were analysed by one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test. Differences with P 10 μM. In comparison, Peptides 1–4 all showed stronger antiproliferative activity. Among them, Peptide-1 displayed the most potent inhibitory effect, with an IC50 value of 0.92 ± 0.04 μM, indicating superior antiproliferative activity. By contrast, Scrambled peptide failed to achieve 50% inhibition even at the highest tested concentration (10 μM), further demonstrating that the biological activity of Peptide-1 depends on its specific amino acid sequence rather than non-specific effects. In addition, integrated analysis of the binding free energy, binding affinity, and antiproliferative activity of Peptides 1–4 (Figure 8) revealed a strong overall consistency among these parameters, further supporting the correlation between the computational predictions and experimental results.

To further assess the cellular target engagement of Peptide-1 with KRASG12V, a cellular NanoBRET target engagement assay was performed. As shown in Figure 9, increasing concentrations of unlabelled Peptide-1 induced a concentration-dependent decrease in the normalised BRET signal, yielding an apparent IC50 value of 1.91 ± 0.20 μM. These results support the cellular engagement of Peptide-1 with KRASG12V in SW480 cells.

To further evaluate the antiproliferative activity of Peptide-1, its effects on the proliferation of additional colorectal cancer cell lines (SW620, SNU-C4, and HT-29) and normal colonic epithelial cells (NCM460) were assessed by the MTT assay, as summarised in Table 4. Peptide-1 exhibited pronounced antiproliferative activity against KRASG12V-mutant SW620 and SNU-C4 cells, with IC50 values of 1.06 ± 0.07 μM and 1.27 ± 0.09 μM, respectively. In contrast, Peptide-1 showed no appreciable antiproliferative activity in KRASWT HT-29 cells, with an IC50 value >10 μM. Similarly, Scrambled peptide showed no obvious activity in this cell line. Compared with its activity in KRASG12V-mutant SW480 cells, the markedly weaker effect of Peptide-1 in HT-29 cells further supports that its cellular activity is preferentially associated with the KRASG12V-mutant context rather than broad or non-specific cytotoxicity.

In addition, its inhibitory effect on NCM460 cells was markedly weaker (IC50 >10 μM), suggesting lower cytotoxicity towards normal colonic epithelial cells while maintaining potent activity against tumour cells. Moreover, Scrambled peptide showed no obvious activity in any of the tested cell lines, further demonstrating that the antitumor effect of Peptide-1 depends on its specific amino acid sequence and structural features rather than non-specific effects. Collectively, these results indicate that Peptide-1 exerts potent and consistent antiproliferative activity against multiple colorectal cancer cell lines while showing weaker inhibitory effects on normal colonic epithelial cells, further supporting its potential as an anti-colorectal cancer peptide.### KRASG12V-dependent antiproliferative activity of peptide-1

To further validate the target dependence of the antiproliferative activity of Peptide-1, a KRASG12V knockdown SW480 cell model (shKRASG12V) was established using shRNA and compared with control cells (shControl). Antiproliferative activity was assessed by the MTT assay (Table 5). Peptide-1 retained potent inhibitory activity in shControl cells (IC50 = 0.92 ± 0.04 μM), whereas its activity was markedly reduced in shKRASG12V cells, indicating that its antiproliferative effect is closely associated with KRASG12V expression. In contrast, Scrambled peptide showed no detectable activity in either cell model, excluding non-specific cytotoxicity. Collectively, these results suggest that the antiproliferative activity of Peptide-1 in SW480 cells is closely associated with KRASG12V expression and is unlikely to result from non-specific cytotoxicity.### Peptide-1-associated changes in ERK signalling and cell-cycle distribution

To preliminarily explore the downstream cellular responses associated with Peptide-1 treatment, ERK1/2 phosphorylation, ERK1/p21 mRNA expression, and cell-cycle distribution were examined. As shown in Figure 10(A), EGF stimulation markedly increased the p-ERK1/2/total ERK1/2 level. Compared with EGF treatment alone, co-treatment with EGF and Peptide-1 reduced ERK1/2 phosphorylation. This reduction showed a trend similar to that observed with the positive reference peptide H-REV107, but was more pronounced under the present experimental conditions. These results suggest that Peptide-1 treatment is associated with reduced EGF-induced ERK-related signalling activity in KRASG12V-mutant SW480 cells.

Further RT-qPCR analysis showed that Peptide-1 treatment downregulated ERK1 mRNA expression and upregulated p21 mRNA expression, with a concentration-related trend (Figure 10B). In parallel, cell-cycle analysis revealed a gradual increase in the proportion of cells in the G0/G1 phase with increasing concentrations of Peptide-1, accompanied by corresponding decreases in the S and G2/M phase populations (Figure 10C). Collectively, these findings suggest that Peptide-1 treatment is associated with ERK-related signalling changes, p21 upregulation, and G0/G1 cell-cycle accumulation, providing preliminary evidence for downstream responses related to its antiproliferative activity.

Docking-based selection of KRASG12V-targeting peptides

Using an integrated workflow that combined computational screening with experimental validation, we identified KRASG12V-targeting peptides for subsequent experimental evaluation. This workflow included molecular docking, binding affinity assays, structure–activity relationship analysis, molecular dynamics simulations, and in vitro cellular activity assessment. The overall workflow is shown in Figure 1. First, preliminary virtual screening of the constructed peptide library containing 59,319 sequences was performed against KRASG12V using the Dock module in MOE, and docking scores were calculated for each KRASG12V–Peptide complex. In general, lower docking scores indicate more favourable KRASG12V–Peptide interactions and stronger predicted binding potential. Subsequently, the top four peptides ranked by docking score (Peptides 1–4), together with the co-crystallized reference peptide (H-REV107) from the KRASG12V structure (PDB ID: 7C41) were selected for further evaluation. In addition, the amino acid sequences of Peptides 1–4 are listed in Table 1. As shown in Figure 2, H-REV107 exhibited a docking score of −7.34 ± 0.39 kcal/mol, whereas Peptides 1–4 consistently showed lower docking scores. Notably, Peptide-1 displayed the lowest docking score (−13.65 ± 0.77 kcal/mol), indicating a more favourable interaction with KRASG12V and greater predicted binding potential.

Binding validation and correlation with docking predictions

To further assess the binding affinities of docking-selected Peptides 1–4 for GppNHp-loaded KRASG12V, Kd values were measured by MST under an active-like nucleotide-bound state, with the results summarised in Table 1. Lower Kd values indicate higher binding affinity between the peptide and recombinant human KRASG12V protein. H-REV107 exhibited a Kd of 2.57 ± 0.14 μM, whereas Peptides 1–4 showed consistently lower Kd values ranging from 0.85 to 1.69 μM. Among them, Peptide-1 displayed the lowest Kd value (0.85 ± 0.03 μM), indicating the highest binding affinity for recombinant human KRASG12V protein. To further assess the sequence specificity of Peptide-1 binding to KRASG12V, Scrambled peptide was generated by randomising the amino acid sequence of Peptide-1. Scrambled peptide displayed a Kd value of >10 μM, indicating that the interaction of Peptide-1 with KRASG12V depends on its specific residue arrangement rather than simple amino acid composition, arguing against non-specific binding.

To evaluate the relationship between docking predictions and experimentally determined binding affinity, linear regression was performed between the mean docking scores and Kd values of Peptides 1–4 and H-REV107 (Figure 3). A strong positive correlation was observed (R2 = 0.9611; Y = 0.2812X + 4.6699), indicating good agreement between predicted binding free energy and experimental affinity. Notably, more favourable docking scores (lower values) were associated with lower Kd values, reflecting higher binding affinity. Among the tested peptides, Peptide-1 exhibited both the lowest docking score and the lowest Kd, supporting Peptide-1 as the strongest binder among the tested peptides.

To further assess the binding selectivity of Peptide-1 towards KRASG12V, its binding affinities for several representative KRAS mutants, including KRASG12D, KRASG12C, KRASG12A, KRASG12S, KRASG13D, and KRASQ61H, were also evaluated by MST. As shown in Table S1, Peptide-1 showed no appreciable binding to these KRAS mutants under the same experimental conditions, with Kd values greater than 10 μM. These results indicate that Peptide-1 exhibited weak or undetectable binding to the tested KRAS mutants within the assay range. Together with its binding affinity for KRASG12V, these data suggest that Peptide-1 may display a preferential binding profile towards KRASG12V rather than broadly engaging multiple KRAS mutants.

Structure–activity relationship analysis

Given the measurable binding affinities of Peptides 1–4 for recombinant human KRASG12V, their interaction patterns in the optimal docking poses were systematically analysed. The predicted docking poses and binding surface views of Peptides 1–4 at the binding site of KRASG12V are presented in Figure 4. The interaction analysis primarily focused on hydrogen bonding and hydrophobic interactions. To more clearly illustrate the binding features, Table 2 summarises the key KRASG12V residues involved in hydrogen-bond formation, the corresponding number of hydrogen bonds, and the residues participating in hydrophobic interactions. Detailed interaction analysis showed that all Peptides 1–4 formed hydrogen bonds with the KRASG12V active-pocket residues Gly13, Ala59, Gln61, Asn85, Asn86, Asp92, and Lys117. Notably, Peptide-1 and Peptide-4 each formed additional hydrogen bonds with Tyr32 and Tyr96, whereas these residues mainly participated in hydrophobic interactions with Peptides 2–3.

The distinct structural features of Peptides 1–4 led to differences in their interaction patterns, which may underlie their varying binding affinities for KRASG12V. Further analysis showed that, compared with Peptide-4, Peptide-2 formed three hydrogen bonds with Asn85, indicating stronger binding interactions, consistent with its lower Kd value and higher affinity. Relative to Peptide-2, Peptide-3 formed one additional hydrogen bond with Ala59, suggesting a more stable binding conformation, in agreement with its experimentally observed higher affinity. Among Peptides 1–4, Peptide-1 formed the greatest number of hydrogen bonds with KRASG12V, supporting the strongest overall interaction. Collectively, the interaction patterns closely correlated with the observed binding affinity trends.

MD Analysis of KRASG12V–Peptide-1 interaction

Given its high binding affinity and stable docking pose towards KRASG12V, Peptide-1 was therefore subjected to 50 ns molecular dynamics simulations to further evaluate the stability of the KRASG12V–Peptide-1 complex under physiological conditions. RMSD analysis (Figure 5(A)) showed that the RMSD of apo KRASG12V continued to increase throughout the simulation, without reaching a clear equilibrium. In contrast, the KRASG12V–Peptide-1 complex stabilised after approximately 5 ns, with the RMSD remaining around 0.2 nm. These results indicate that Peptide-1 binds stably to KRASG12V and contributes to the conformational stability of the complex throughout the simulation. Rg analysis (Figure 5(B)) showed that the KRASG12V–Peptide-1 complex maintained a stable radius of gyration throughout the simulation, with an Rg value of approximately 1.56 nm, indicating a compact and structurally stable conformation. Secondary-structure analysis (Figure 5(C)) further revealed that KRASG12V was predominantly composed of α-helices, coils, and β-sheets during the simulation, with no obvious abnormal transitions in secondary-structure elements, suggesting that the overall secondary structure of the protein remained stable. RMSF analysis (Figure 5(D)) showed that most KRASG12V residues in the complex exhibited only limited fluctuations, with RMSF values generally below 0.2 nm, indicating that Peptide-1 binding enhanced the conformational stability of the protein. Higher flexibility was mainly confined to the terminal regions distal to the binding site. Notably, the key interfacial residues all displayed RMSF values below 0.1 nm, further suggesting that Peptide-1 effectively restricted local conformational fluctuations around the binding site.

To further elucidate the binding mechanism of Peptide-1 with KRASG12V, per-residue energy decomposition and hydrogen-bond analyses were performed for the active-pocket residues of the complex. As shown in Figure 5(E), key residues including Asp92, Asn85, Asn86, Gly13, Lys117, and Tyr32 exhibited significant contributions to the binding free energy, with most values below −1 kcal/mol, indicating their critical roles in stabilising the complex. Figure 5(F) further revealed that the binding of Peptide-1 to KRASG12V was governed by a combination of interactions, primarily van der Waals and electrostatic contributions, where van der Waals interactions provided the main stabilising force, while electrostatic interactions enhanced binding specificity at key sites. Consistently, Figure 5(G) showed that the KRASG12V–Peptide-1 complex maintained a stable hydrogen-bonding pattern with high occupancy throughout the simulation, further supporting the essential role of the hydrogen-bonding network in maintaining conformational stability of the complex. These findings further confirm the consistency between molecular docking, binding affinity data and molecular dynamics results, revealing that the predicted binding interactions remained stable throughout the simulation and could be sustained under simulated physiological conditions.

FEL analysis of KRASG12V–Peptide-1 complex

To further evaluate the conformational stability and dynamic behaviour of the complex from an energetic perspective, FEL analysis was performed on the molecular dynamics trajectories, as shown in Figure 6. Both the two- and three-dimensional free energy distributions constructed using RMSD and Rg exhibited a well-defined low-energy region, mainly clustered at RMSD values of approximately 0.15–0.20 nm and Rg values of approximately 1.56–1.57 nm, indicating the presence of a stable dominant conformation during the simulation. Notably, the RMSD and Rg values corresponding to the global minimum were highly consistent with the preceding conformational analyses. Collectively, the FEL results further corroborate the dynamic findings from a thermodynamic perspective, supporting the favourable conformational stability and energetically preferred state of the complex.

Human serum stability of peptide-1

Given the stable KRASG12V–Peptide-1 complex observed during the 50 ns molecular dynamics simulation, the stability of Peptide-1 in human serum was further assessed. As shown in Figure 7, Peptide-1 exhibited a modest, time-dependent decrease in the proportion of remaining intact peptide. More than 90% of Peptide-1 was retained during the first 90 min, and approximately 80% remained after 360 min of incubation. Notably, no pronounced rapid degradation phase was observed throughout the tested period. These results suggest that Peptide-1 possesses favourable in vitro stability in human serum, thereby supporting its further evaluation in cell-based assays.

In vitro antiproliferative effects of KRASG12V-targeting peptides

To evaluate the in vitro anti-colorectal cancer activity of the selected peptides, their inhibitory effects on SW480 cell proliferation were measured by the MTT assay, and the corresponding IC50 values are summarised in Table 3. H-REV107 was used as the reference peptide and exhibited an IC50 value of >10 μM. In comparison, Peptides 1–4 all showed stronger antiproliferative activity. Among them, Peptide-1 displayed the most potent inhibitory effect, with an IC50 value of 0.92 ± 0.04 μM, indicating superior antiproliferative activity. By contrast, Scrambled peptide failed to achieve 50% inhibition even at the highest tested concentration (10 μM), further demonstrating that the biological activity of Peptide-1 depends on its specific amino acid sequence rather than non-specific effects. In addition, integrated analysis of the binding free energy, binding affinity, and antiproliferative activity of Peptides 1–4 (Figure 8) revealed a strong overall consistency among these parameters, further supporting the correlation between the computational predictions and experimental results.

To further assess the cellular target engagement of Peptide-1 with KRASG12V, a cellular NanoBRET target engagement assay was performed. As shown in Figure 9, increasing concentrations of unlabelled Peptide-1 induced a concentration-dependent decrease in the normalised BRET signal, yielding an apparent IC50 value of 1.91 ± 0.20 μM. These results support the cellular engagement of Peptide-1 with KRASG12V in SW480 cells.

To further evaluate the antiproliferative activity of Peptide-1, its effects on the proliferation of additional colorectal cancer cell lines (SW620, SNU-C4, and HT-29) and normal colonic epithelial cells (NCM460) were assessed by the MTT assay, as summarised in Table 4. Peptide-1 exhibited pronounced antiproliferative activity against KRASG12V-mutant SW620 and SNU-C4 cells, with IC50 values of 1.06 ± 0.07 μM and 1.27 ± 0.09 μM, respectively. In contrast, Peptide-1 showed no appreciable antiproliferative activity in KRASWT HT-29 cells, with an IC50 value >10 μM. Similarly, Scrambled peptide showed no obvious activity in this cell line. Compared with its activity in KRASG12V-mutant SW480 cells, the markedly weaker effect of Peptide-1 in HT-29 cells further supports that its cellular activity is preferentially associated with the KRASG12V-mutant context rather than broad or non-specific cytotoxicity.

In addition, its inhibitory effect on NCM460 cells was markedly weaker (IC50 >10 μM), suggesting lower cytotoxicity towards normal colonic epithelial cells while maintaining potent activity against tumour cells. Moreover, Scrambled peptide showed no obvious activity in any of the tested cell lines, further demonstrating that the antitumor effect of Peptide-1 depends on its specific amino acid sequence and structural features rather than non-specific effects. Collectively, these results indicate that Peptide-1 exerts potent and consistent antiproliferative activity against multiple colorectal cancer cell lines while showing weaker inhibitory effects on normal colonic epithelial cells, further supporting its potential as an anti-colorectal cancer peptide.

KRASG12V-dependent antiproliferative activity of peptide-1

To further validate the target dependence of the antiproliferative activity of Peptide-1, a KRASG12V knockdown SW480 cell model (shKRASG12V) was established using shRNA and compared with control cells (shControl). Antiproliferative activity was assessed by the MTT assay (Table 5). Peptide-1 retained potent inhibitory activity in shControl cells (IC50 = 0.92 ± 0.04 μM), whereas its activity was markedly reduced in shKRASG12V cells, indicating that its antiproliferative effect is closely associated with KRASG12V expression. In contrast, Scrambled peptide showed no detectable activity in either cell model, excluding non-specific cytotoxicity. Collectively, these results suggest that the antiproliferative activity of Peptide-1 in SW480 cells is closely associated with KRASG12V expression and is unlikely to result from non-specific cytotoxicity.

Peptide-1-associated changes in ERK signalling and cell-cycle distribution

To preliminarily explore the downstream cellular responses associated with Peptide-1 treatment, ERK1/2 phosphorylation, ERK1/p21 mRNA expression, and cell-cycle distribution were examined. As shown in Figure 10(A), EGF stimulation markedly increased the p-ERK1/2/total ERK1/2 level. Compared with EGF treatment alone, co-treatment with EGF and Peptide-1 reduced ERK1/2 phosphorylation. This reduction showed a trend similar to that observed with the positive reference peptide H-REV107, but was more pronounced under the present experimental conditions. These results suggest that Peptide-1 treatment is associated with reduced EGF-induced ERK-related signalling activity in KRASG12V-mutant SW480 cells.

Further RT-qPCR analysis showed that Peptide-1 treatment downregulated ERK1 mRNA expression and upregulated p21 mRNA expression, with a concentration-related trend (Figure 10B). In parallel, cell-cycle analysis revealed a gradual increase in the proportion of cells in the G0/G1 phase with increasing concentrations of Peptide-1, accompanied by corresponding decreases in the S and G2/M phase populations (Figure 10C). Collectively, these findings suggest that Peptide-1 treatment is associated with ERK-related signalling changes, p21 upregulation, and G0/G1 cell-cycle accumulation, providing preliminary evidence for downstream responses related to its antiproliferative activity.

Conclusions

KRASG12V is a clinically relevant oncogenic driver in colorectal cancer, contributing substantially to malignant progression through its unique structural and functional properties. Despite its high prevalence and biological importance, the development of effective KRASG12V-targeted inhibitors remains limited, highlighting a critical unmet need in current anticancer drug discovery. Accordingly, the development of peptide inhibitors based on H-REV107 scaffold represents a rational strategy for KRASG12V targeting. This study implemented a structure-based virtual screening strategy to identify four peptides targeting KRASG12V from the peptide library containing 59,319 sequences. Among them, Peptide-1 exhibited the most favourable docking score, suggesting superior predicted binding. MST assays validated enhanced binding affinities of all four peptides towards active-like KRASG12V relative to the reference peptide H-REV107, with Peptide-1 exhibiting the strongest interaction. The observed agreement between computational docking scores and MST-derived Kd values supports the utility of this screening workflow for prioritising KRASG12V-binding peptides, while MST selectivity profiling further suggests that Peptide-1 may preferentially engage KRASG12V over other representative KRAS mutants under the same assay conditions.

Structural and molecular dynamics analyses revealed that differential binding affinities among Peptides 1–4 arose primarily from distinct hydrogen-bonding networks and hydrophobic interactions within the KRASG12V active pocket. Peptide-1 formed the most extensive and stable interaction interface with critical pocket residues, and molecular dynamics simulations supported conformational stability of the KRASG12V–Peptide-1 complex under physiological conditions, characterised by restrained residue fluctuations, preserved structural compactness, and a well-defined low-energy free-energy landscape relative to apo KRASG12V. Energy decomposition underscored the cooperative contributions of van der Waals forces, electrostatic interactions, and persistent hydrogen bonds to stable and high-affinity binding.

Peptide-1 demonstrated favourable in vitro stability in human serum, retaining more than 90% of the intact peptide within 90 min and approximately 80% after 360 min, supporting its suitability for cell-based evaluation. In cellular assays, Peptide-1 showed the most potent antiproliferative activity among Peptides 1–4 in SW480 cells and maintained consistent activity across additional colorectal cancer cell lines, while exhibiting a weaker inhibitory effect on normal colonic epithelial cells. Cellular NanoBRET analysis supported the intracellular engagement of Peptide-1 with KRASG12V in SW480 cells. In addition, Peptide-1 exhibited KRASG12V-associated antiproliferative activity in SW480 cells, with no apparent non-specific cytotoxicity. Peptide-1 treatment was associated with reduced EGF-induced ERK1/2 phosphorylation, ERK1 downregulation, p21 upregulation, and increased G0/G1 cell-cycle accumulation. Collectively, Peptide-1 exhibited appreciable antiproliferative activity in colorectal cancer cells, potentially in association with KRASG12V intracellular target engagement, ERK-related signalling changes, and altered cell-cycle distribution.