고효율 유전자적중 및 초정밀 유전자교정 전략 개발 = Development of novel strategies for an enhanced gene knockout and a high-fidelity gene editing

다국어 초록 (Multilingual Abstract)

The CRISPR–Cas system has emerged as one of the most powerful and versatile genome-editing platforms, enabling precise modification of virtually any genomic locus. Among various CRISPR endonucleases, Cas9 and Cpf1 (also known as Cas12a) are the most extensively utilized systems. Despite their broad applicability, conventional CRISPR-based knockout approaches remain largely dependent on cell proliferation, displaying high efficiency in immortalized or rapidly dividing cells but limited success in primary or slow-growing cells. This constraint has restricted the use of CRISPR technology for functional studies that require physiologically relevant cell types. Chapter I of this study addresses this limitation by establishing a multiplex-Cpf1–based single-gene knockout system optimized for efficiency and accuracy. A major challenge in CRISPR-mediated gene disruption is that roughly one-third of induced mutations are in-frame, yielding truncated but partially functional proteins. To overcome this, we exploited the intrinsic pre-crRNA processing capability of Acidaminococcus sp. Cpf1 (AsCpf1) to construct a multi- target system in which multiple crRNAs are transcribed as a single array and autonomously processed into mature crRNAs. This design allows simultaneous cleavage of several target sites within one gene using a single construct, significantly increasing the likelihood of complete functional knockout. Multiple crRNAs targeting distinct exons were first screened for activity, and the most efficient candidates were assembled into a four-crRNA array (4CR). The 4CR system was validated in both hepatocellular carcinoma (HCC) cell lines and primary fibroblasts, achieving efficient gene disruption without clonal isolation. Western blot analysis and targeted deep sequencing confirmed biallelic knockout and complete loss of protein expression in mixed populations. Moreover, the 4CR platform enabled rapid functional studies of gene–drug interactions. Knockout of autophagy-related 5 (ATG5) and EI24 autophagy associated transmembrane protein (EI24) genes, two key autophagy regulators, altered sorafenib sensitivity in HCC models both in vitro and in vivo, demonstrating that this system can effectively generate physiologically relevant knockout phenotypes. Collectively, these findings establish the multiplex-Cpf1 system as a robust, stable, and scalable platform for gene knockout in diverse cell types, including those refractories to conventional Cas9-mediated editing. Chapter II focuses on enhancing genome-editing precision by developing an intentional mismatch–based Cas9 system designed for single-nucleotide discrimination. Off-target cleavage remains a critical obstacle for therapeutic genome editing. Although high-fidelity Cas9 variants such as eSpCas9(1.1), SpCas9-HF1, and evoCas9 have been engineered to improve target specificity, these variants often exhibit reduced on-target activity due to attenuated DNA binding. To address this trade-off without altering the Cas9 protein itself, we rationally introduced one or two deliberate mismatches within the 20-bp guide RNA (gRNA) sequence. This subtle destabilization weakens imperfect pairing at off-target loci while preserving robust hybridization at the intended target site, thereby reducing off-target activity without compromising efficiency. The C–C motif chemokine receptor 5 (CCR5) and its homolog C-C motif chemokine receptor 2 (CCR2) were chosen as a well-established model to assess Cas9 specificity, since sequence similarity between these loci frequently causes off-target editing. Experimental results showed that gRNAs harboring intentional mismatches effectively eliminated cleavage at CCR2 while maintaining high activity at CCR5. Structural analysis suggested that mismatches near residues interacting with R661 in the Cas9 REC1 domain play a key role in modulating specificity. Systematic comparison further revealed that dual mismatches at positions 10 and 18 of the spacer sequence most effectively suppressed off-target cleavage. Importantly, this strategy proved compatible with both wild-type Cas9 and high-fidelity variants, underscoring its broad applicability. To explore the therapeutic potential of this approach, we applied the intentional-mismatch gRNA design to the KRAS proto-oncogene (KRAS), a critical driver of human cancers. KRAS encodes a small GTPase that alternates between active (GTP-bound) and inactive (GDP- bound) states to control the RAS/MAPK signaling cascade. Single-base substitutions at codons 12, 13, and 61 (e.g., G12D, G12V, G13D) abolish intrinsic GTPase activity, leading to constitutive activation and oncogenic transformation. Using pancreatic cancer cell lines harboring wild-type or mutant (G12D, G12V) KRAS alleles, we demonstrated that an intentionally mismatched sgRNA (10A variant) selectively cleaved the mutant KRASG12D allele while sparing the wild type. Deep sequencing and Western blot analysis with a mutation- specific antibody confirmed allele-specific editing with single-nucleotide precision. These results collectively demonstrate that intentional incorporation of mismatches into the gRNA sequence provides a simple yet powerful method for enhancing Cas9 specificity. Unlike protein engineering approaches, this strategy preserves the catalytic and structural integrity of the enzyme while achieving high-fidelity editing across multiple Cas9 platforms. In summary, this study introduces two complementary genome-editing strategies that address both efficiency and precision, the two key bottlenecks in CRISPR-based gene manipulation. The multiplex-Cpf1 platform enables efficient, clonal-free knockout generation even in primary cells, facilitating functional genomics and disease modeling. Meanwhile, the intentional mismatch Cas9 system offers a practical route to allele-selective editing and precision therapeutics without modifying Cas9 itself. Together, these technologies establish an integrated framework for next-generation genome editing that balances accuracy, versatility, and broad biological applicability.

목차 (Table of Contents)

• Abstract i
• List of figures viii
• List of tables xi
• Background 1
• 1. Evolution of genome editing tools: ZFNs, TALENs, and the CRISPR-Cas systems 1
• 2. Overview of hepatocellular carcinoma: pathogenesis, treatment modalities, and targeted therapy with sorafenib 4
• CHAPTER I. Establishment of a multiplex-Cpf1-based knockout model for drug sensitivity evaluation 5
• Introduction 6
• Materials and methods 10
• 1. Cell culture 10
• 2. Vector construction 10
• 3. Virus production and infection 12
• 4. Detection of indel mutations 12
• 5. Western blot analysis 13
• 6. Immunofluorescence cytochemistry 13
• 7. Cell viability assay 14
• 8. Xenograft mouse model 14
• Results 16
• 1. Design of a single-gene-directed multiplex CRISPR-Cpf1 system 16
• 2. Validation of the multiplex CRISPR-Cpf1 system by establishing EI24-deficient cancer cells 17
• 3. Validation of ATG5 knockout using the multiplex CRISPR-Cpf1 system 18
• 4. Establishment and validation of double-gene knockout cell lines using the multiplex CRISPR-Cpf1 system 19
• 5. Validation of the multiplex CRISPR-Cpf1 system in primary cells 19
• 6. Evaluation of sorafenib drug responsiveness in ATG5 or EI24 knockout cell lines generated via the multiplex CRISPR-Cpf1 system 21
• 7. In vivo evaluation of sorafenib responsiveness in ATG5- and EI24-deficient HepG2 cell line xenograft models generated using the multiplex CRISPR-Cpf 1 system 22
• 8. Prognostic interaction between ATG5 and EI24 expression in human hepatocellular carcinoma 23
• Discussion 24
• Figures 28
• Tables 41
• CHAPTER II. Development of an intentional mismatch-based Cas9 system for precision genome editing 44
• Introduction 45
• Materials and methods 47
• 1. Cell culture 47
• 2. Vector construction 47
• 3. Virus production and infection 47
• 4. Detection of indel mutations 48
• 5. Western blot analysis 48
• Results 50
• 1. Design of the intentional mismatch-based Cas9 system 50
• 2. Selection of CCR5 sgRNAs for validation of the intentional mismatch strategy utilizing the sequence similarity between CCR5 and CCR2 51
• 3. Validation of the intentional mismatch strategy based on transition substitution 51
• 4. Evaluation of intentional mismatches across all 20-bp spacer positions in CCR5- RG12 sgRNA 52
• 5. Validation of Cas9 target specificity modulation through intentional mismatches at positions 10 and 18 53
• 6. Validation of the intentional mismatch sgRNA strategy in high-fidelity Cas 9 variants 54
• 7. Validation of the intentional mismatch strategy in KRAS point mutation models (G12D and G12V) 55
• Discussion 58
• Figures 62
• Tables 73
• Reference 81
• 국문요약 90