High-fidelity Cas9-mediated targeting of KRAS driver mutations restrains lung cancer in preclinical models

Abstract

Missense mutations in the 12^th codon of KRAS are key drivers of lung cancer, with glycine-to-cysteine (G12C) and glycine-to-aspartic acid (G12D) substitutions being among the most prevalent. These mutations are strongly associated with poor survival outcomes. Given the critical role of KRAS in lung cancer and other cancers, it remains as a major target for the development of new and complementary treatments. We have developed a CRISPR-High Fidelity (HiFi)-Cas9-based therapy strategy that can effectively and specifically target KRAS^G12C and KRAS^G12D mutants, avoiding KRAS^WT off-targeting and affecting KRAS downstream pathways, thereby significantly reducing tumorgenicity. The delivery of HiFiCas9 components via ribonucleoprotein particles (RNPs) and adenovirus (AdV) effectively abrogates cell viability in KRAS-mutant Non-Small Cell Lung Cancer (NSCLC) preclinical models, including 2D and 3D cell cultures, cell-derived xenografts (CDX), and patient-derived xenograft organoids (PDXO). Our in vitro studies demonstrate that HiFiCas9-based therapy achieves superior KRAS inhibition compared to Sotorasib and effectively circumvents certain resistance mechanisms associated with Sotorasib treatment. Moreover, in vivo delivery using adenoviral particles significantly suppresses tumor growth in preclinical NSCLC models. Collectively, our findings establish HiFiCas9 as an effective therapeutic strategy with promising clinical applications, especially if in vivo delivery methods are further optimized.

Fig. 1. Efficiency and specificity of KRAS^mut-specific sgRNAs.

A Representation of KRAS^WT and KRAS^mut alleles at genomic DNA level, with the mutated nucleotide highlighted in red. B Graphical representation of the therapeutic approach: KRAS oncogene-addicted cells die off when the KRAS^mut-specific KO is induced. No effect on KRAS^WT/non-tumor cells. C sgRNA designs utilizing different protospacer adjacent motifs (red frames for PAM1 and green for PAM2) to target KRAS^mut alleles. Mutant nucleotides are displayed in red, artificially introduced mismatches in blue, and PAM sequences in orange. D T7-endonuclease assay in KRAS^WT and KRAS^mut cell lines. Efficiency calculation of KRAS^mut-specific sgRNAs on digitized agarose gels after T7-endonuclease assay (shown as percentage below the lane with edition) using the formula: $\mathrm{Editing}\mathrm{efficiency}\left(\%\right)=100\times \left(1-\sqrt{1-\frac{\left(F1+F2\right)}{\left(F1+F2+FL\right)}}\right)$. F1 and F2 refer to the relative pixel density of fragments 1 and 2; FL refers to the full-length undigested amplicon. Gel images are representative of n = 3. E T7-endonuclease assay in KRAS^WT and KRAS^G12C cell lines using sgRNAs with single mismatches. C-: negative control (100% complementary dsDNA). C + : positive control (heteroduplex with indels). The image shown is representative of three independent experiments. F KRAS allele edition frequency by sgRNA. Proportion of unedited and edited targeted high-throughput sequencing reads from either WT, G12C and G12D alleles in heterozygous (H358, A427) and WT homozygous cell lines (H838). Bar graphs represent mean edited read percentages ± SD from three biological replicates for each cell line. Statistical analysis was performed using two-tailed unpaired t-tests with FDR-corrected p values. Source data for panels D, E, and F are provided in the Source Data file.

Fig. 2. Genome editing of KRAS^G12C and KRAS^G12D impairs viability of tumor cells in vitro.

A Cell viability of KRAS^mut NSCLC cell lines targeted with mutation-specific RNPs, 7 days after transfection. Data represent the mean ± SD of N independent biological replicates per cell line, analyzed using a two-tailed one-sample t-test (H358, n = 3, p = 0.016; H1792, n = 5, p = 0.0034; A427, n = 5, p = 0.011; SKLU1, n = 6, p = 0.0002; H838, n = 4, p = 0.27). Source data are provided as a Source Data file. B 2D cell viability assay of NSCLC cell lines 10 days after AdV transduction. Data represent the mean ± SEM of N independent biological replicates per cell line, analyzed using a two-tailed one-sample t-test (H358, n = 7, p = 0.0001; A427, n = 5, p = 0.65; H838, n = 7, p = 0.31). C Left: 3D cell viability assay of NSCLC cell lines 10 days after AdV transduction. Right: representative 3D spheroid culture of H358 cells. Data represent the mean ± SEM of N independent biological replicates per cell line, analyzed using a two-tailed one-sample t-test (H358, n = 7, p = 0.0001; A427, n = 6, p = 0.0014; H838, n = 6, p = 0.12). D Western Blots displaying KRAS-dependent signaling 72 h after RNP transfection. Representative Western blot images (left) and corresponding densitometric quantification (right). Data represent mean ± SEM from three independent biological replicates, analyzed using a two-tailed one-sample t-test (H358: KRAS p = 0.016, ph-p70 p = 0.025, ph-ERK p = 0.045; A427: KRAS^G12D p = 0.0044, ph-p70 p = 0.016, ph-ERK p = 0.0003). E Western Blot displaying KRAS-dependent signaling in H358 cells 72 h after AdV transduction. (Representative image of two biological replicates). Treatment was administered once at time point t = 0 in all the experiments. Source data for panels A-E are provided in the Source Data file.

Fig. 3. Edition of mutant KRAS induces tumor growth inhibition in PDX and CDXs models.

A Schematic representation of experimental design with representative ex vivo images of tumors harvested 60 days after implantation. 10⁶ pretreated cells were transplanted subcutaneously and tumor growth was monitored for two months. Representative images of n = 9. B Quantification of ex vivo tumor volumes of H358 (Adv-Control, n = 9; Adv-HiFiCas9, n = 7) and A427 (Adv-Control, n = 8; Adv-HiFiCas9, n = 8) CDXs. 63% reduction in tumor volume for H358 (left; p-value: 0.00017) and a ~ 42% reduction for A427 (right; p-value: 0.015). Boxplots display the median (50th percentile, center line), the 25th and 75th percentiles (lower and upper box bounds, respectively), and the whiskers extend to the smallest and largest values within 1.5 times the interquartile range (IQR) from the lower and upper quartiles. C Schematic representation of PDX experimental design. D Tumor volumes of LU5245 (G12C; Adv-Control, n = 6; Adv-HiFiCas9, n = 6) and LU5162 (G12D, Adv-Control, n = 4; Adv-HiFiCas9, n = 4) PDXs. Mean tumor volume (mm³) normalized to day 1, accompanied by standard deviation. Tumor growth was modeled using a linear mixed-effects model with random intercepts. E Ex vivo pictures from KRAS^G12C tumors extracted 28 days post-treatment. Scale bar=10 mm. Diameters (mm): PDX1: Control=14, HiFiCas9 = 11; PDX2: Control=13.6, HiFiCas9 = 11.8. Volumes (mm³): PDX1: Control=1084, HiFi-Cas9 = 548; PDX2: Control=948, HiFi-Cas9 = 658. Source data for panels B and D are provided in the Source Data file.

Fig. 4. Comparison of Sotorasib and HiFiCas9 therapy.

A Cell viability assay combining HiFiCas9 therapy and Sotorasib in parental H358 cells (n = 3 independent biological replicates). Control/HiFiCas9 infection was performed 24 h before Sotorasib treatment. Data represent the mean ± SD of three independent biological replicates, analyzed using a two-way ANOVA. (AdV-Control vs AdV-HiFiCas9; 0 nM Sotorasib p = 0.0005, 3 nM Sotorasib p = 0.0008, 6 nM Sotorasib p = 0.0056) B Colony formation assay combining HiFiCas9 therapy and Sotorasib in parental H358 cells. C Cell viability of PDXOs treated with HiFiCas9 therapy and/or Sotorasib. Data represent the mean ± SD of three independent biological replicates derived from three different PDXs, analyzed using a two-tailed unpaired t-test. TP79: AdV-HiFiCas9, p = 0,0133, Combo, p = 0.003; TP60: AdV-HiFiCas9, p = 0.0004, Sotorasib, p = 0.0058, Combo, p = 0.04. D Representative images of organoid cultures five days post-treatment with adenovirus control or HiFiCas9. Scale bar = 100 μm. E Cell viability of Sotorasib-resistant cells treated with HiFiCas9 therapy in 2D cultures. Data represent mean ± SEM from N independent biological replicates, analyzed using a two-tailed one-sample t-test (H358R1-R2, n = 4; H23, n = 3). F Cell viability of Sotorasib-resistant cells treated with HiFiCas9 therapy in 3D cultures. Data represent mean ± SEM from N independent biological replicates, analyzed using a two-tailed one-sample t-test (H358-R1, n = 3; H358-R2, n = 4; H23, n = 5, p = 0.0005). G Representative images of 3D cultures of cells ten days post-treatment with Adenovirus control or HiFiCas9. Scale bar = 50 μm. Source data for panels A, C, E and F are provided in the Source Data file.