Targeted CRISPR/Cas9 Lipid Nanoparticles Elicits Therapeutic Genome Editing in Head and Neck Cancer

Abstract

Squamous cell carcinomas of the head and neck (HNSCC) originate in the upper aerodigestive tract, including the oral cavity, pharynx, and larynx. Current treatments of locally advanced HNSCC often lead to high treatment failure, and disease recurrence, resulting in poor survival rates. Advances in mRNA technologies and lipid nanoparticle (LNP) delivery systems led to several clinical trials involving LNP-CRISPR-Cas9 mRNA-based therapeutics. Despite these advances, achieving cell-type-specific extrahepatic mRNA delivery is still challenging. This study introduces a safe and effective intratumoral EGFR-targeted CRISPR-LNP delivery strategy for knocking out SOX2, which is a cancer-specific gene. To assess their therapeutic potential, it is shown that LNPs made from ionizable lipids with helper lipids co-encapsulating Cas9 mRNA and sgRNA targeting SOX2 (sgSOX2), lead to a ≈60% reduction in HNSCC cell viability in vitro. Next, using a xenograft HNSCC mouse model, targeted delivery of 𝜶EGFR- CRISPR-sgSOX2-LNPs to HNSCC cells resulted in a 90% inhibition of tumor growth and a 90% increase in survival for > 84 days, with tumor disappearance observed in 50% of the mice. These findings emphasize the potential of targeted mRNA-Cas9-LNPs in clinically accessible solid tumors, specifically in reaching tumor cells and inducing persistent therapeutic responses in tumors with high-recurrence rates like HNSCC.

Keywords: CRISPR‐Cas9; Genetic Medicines; Head and neck cancer; Lipid nanoparticles; SOX2; mRNA.

Figure 1

sgRNA screening and viability assays. A) Percentage of gene editing events upon RNP complex transfection with three guides for SOX2 in FADU and UMSCC‐104 HNSCC cell lines. B) Percentage of gene editing of SOX2 and YAP1 (positive control for editing) in UMSCC‐104. Data in A, and B are a representation of three independent experiments. C,D) XTT viability assay of UMSCC‐104 cells and Fadu treated with Mock, sgTurboGFP, sgYAP1, or sgSOX2 for 96 h. Bar chart representing % of cell viability normalized to sgTurboGFP‐treated cells. Data are means ±SD of three independent experiments. An unpaired T‐test was used to assess the significance. *p < 0.05.

Figure 2

The screen of ionizable cationic lipids for the transfection of human HNSCC cells. A) Schematic illustration of LNP preparation. B) Chemical structures of selected ionizable cationic lipids from the lipid library. C) LNPs mean diameter (nm). D) polydispersity index (PDI). E) ζ potential (mV), as measured by Zeta Sizer. F) Percentage of encapsulation efficiency as measured by a RiboGreen assay. Data are means ±SD of three independent experiments.

Figure 3

Efficacy of CRISPR‐ LNPs. A) The percentage of UMSCC cell viability 72 h post‐transfection with different concentrations of sgSOX2‐cLNP (1–10 µg mL⁻¹ of total RNA) as measured by DAPI‐Annexin assay, the viability percentages of each cell treated with sgSOX2‐cLNP were normalized to those treated with the same concentration (1 to 10 µg mL⁻¹ of total RNA) of sgYAP‐cLNPs. Data are means ±SD of three independent experiments. One‐way analysis of variance (ANOVA) with Tukey multiple comparison test was used to assess the significance. *p < 0.05, **p < 0.01, ***p < 0.001 B) SOX2 gene editing percentage of each formulation with the highest concentration. Data are means ±SD of three independent experiments.

Figure 4

Targeted cLNP characterization and genome editing efficiency. A) Representative transmission electron microscopy of naked and tLNPs. The experiment was repeated three times independently (Scale bar = 100 µm). B) LNPs mean diameter (nm). C) polydispersity index (PDI). D) ζ potential (mV), as measured by Zeta Sizer. E) SOX2 gene editing percentage of each cLNP formulation with 10 µg mL⁻¹ of total RNA, 72 h post‐transfection. Data in B‐E are means ±SD of three independent experiments. F) Indel contribution in the edited population and their edited proportions. The cut site is presented by a vertical dotted line, WT signal marked by orange +.

Figure 5

Uptake of targeted LNPs by tumor cell in vivo. A) Expression of luciferase at the tumor site, liver, spleen, and kidney at 6 h post IV injections of L31 α‐EGFR‐targeted LNPs (t‐LNPs), Isotype LNPs (Iso‐LNPs), and naked LNPs (LNPs) (n = 3 mice/group). B) Bars represent each formulation relative to the un‐injected HNSCC‐bearing mouse. Data are presented as mean ±SD; one‐way ANOVA with Tukey multiple comparison test was used to assess the significance. *p < 0.05, **p < 0.01, ***p < 0.001 C) Percentages of GFP‐positive HNSCC cells in the tumor bed, 24 h after injection of naked, isoLNPs, and α‐EGFR‐LNPs as analyzed by flow cytometry, n = 3/group. One‐way ANOVA with Tukey multiple comparison test was used to assess the significance. *p < 0.05.

Figure 6

αEGFR‐SOX2‐LNP mediates therapeutic gene editing in a xenograft HNSCC mouse model. A) Schematic illustration of targeted cLNPs preparation and IT injection with Debakey tweezers and Kel‐F‐needle inserted to Hamilton syringe. B) SOX2 gene editing percentage events 5 days after injection of Iso or 𝜶EGFR‐T‐sgSOX2‐cLNPs or sgNC (scramble control guide), (n = 3 mice/group). C) Indel's contribution to the edited population and their edited proportions. D) Experimental design. UMSCC‐104 cells were subcutaneously inoculated in FoxN1 nude mice. After tumors had reached ≈50 mm³ (day 10), mice were IT injected with 𝜶EGFR‐LNPs or iso‐LNPs encapsulating sgSOX2 or sgNT on days 10,17,24, n = 10 mice/group and mock group injected with BPS (n = 5). E and F) Representative fluorescence imaging of UMSCC‐bearing mice for 6 weeks for all groups and long‐follow of 12 weeks for 𝜶EGFR‐LNPs encapsulating sgSOX2. G) Tumor growth inhibition by triple‐dose treatment with cLNPs, data are presented as means ± SEM; H) Tumor volume of each treatment group at week 6, data presented as means ±SD; one‐way ANOVA with Tukey multiple comparison test was used to assess the significance. *p < 0.05, **p < 0.01, ***p < 0.001. I) Survival curves of UMSCC‐bearing mice. n = 10 animals per treatment group, showing 3 representative mice per cage. ****p < 0.0001. Log‐rank (Mantel‐Cox) test was used for curve comparison.