Enhancing Specificity, Precision, Accessibility, Flexibility, and Safety to Overcome Traditional CRISPR/Cas Editing Challenges and Shape Future Innovations

Abstract

Derived from the bacterial immune system, CRISPR/Cas9 induces DSBs at specific DNA sequences, which are repaired by the cell's endogenous mechanisms, leading to gene insertions, deletions, or substitutions. Despite its transformative potential, several challenges remain in optimizing of CRISPR/Cas systems, including off-target effects, delivery methods, PAM restrictions, and the limitations of traditional editing approaches. This review focuses on the interplay between these challenges and their contributions to gene editing precision, specificity, accessibility, flexibility, and safety. How reducing off-target effects enhances specificity and safety is explored, while discussing the role of HDR-based editing in achieving precise gene modifications, alongside alternative methods such as base editing and prime editing. Improved delivery mechanisms are examined for their impact on accessibility and efficiency, while the reduction of PAM restrictions is highlighted for its contributions to flexibility. Lastly, emerging cleavage-free editing technologies are evaluated as they relate to safety and accessibility. This focused review aims to clarify the connections among these aspects and outline future research directions for advancing CRISPR-based applications.

Keywords: CRISPR/Cas; HDR; delivery systems; double‐stranded breaks; genome editing; off‐target.

Figure 1

Impact of genetic variation (GV) on CRISPR‐based targeting. A) GVs cause loss of gRNA efficiency at on‐target sites. a) GVs within the protospacer (near the seed region) result in loss of gRNA activity, failure in gRNA recognition and pairing that impair nuclease binding and DNA cleavage. b) GVs in the PAM site lead to a complete loss of gRNA activity and failure in DNA cleavage. B) GVs cause gain of gRNA efficiency at off‐target sites. a) GVs at potential off‐target sites can create novel off‐targets and PAM sites, resulting in increasing the potency of gRNA recognition and pairing, as well as nuclease binding and DNA cleavage. b) GVs at the PAM sequence of potential off‐target sites can create novel off‐targets, resulting in increased potency of gRNA recognition and pairing, as well as nuclease binding and DNA cleavage. GVs are highlighted in red. Bases in colored squares indicate mismatches between the on‐target and off‐target, which are predicted based on the reference genome. Vertical lines present the base pairing between the gRNA and the corresponding matching sequence at the on‐target.

Figure 2

Repair‐specific decision‐making choice toward HDR. A) A schematic illustration of DNA repair pathways. The Cas9 or Cas12 nucleases, guided by a specific sgRNA, cause a DSB at a specific site within the genome. The NHEJ repair pathway is initiated when Ku70/80 proteins recognize and bind to the break ends. Following end recognition, end protection and processing begin when Ku recruits DNA‐dependent protein kinase catalytic subunit (DNA‐PKcs), which has a tight affinity to the DNA ends, forming a stable complex that protects the DNA ends from resection and further damage. This recruitment enhances the affinity of the subsequent enzymatic components, forming a highly stable complex that phosphorylates downstream NHEJ proteins, including artemis, XRCC4, and XLF. The latter nucleases process the incompatible and chemically modified ends, making them ligatable and facilitating end ligation by the DNA ligase IV complex (Lig4). These reactions promote the alignment of DNA ends in an error‐prone pattern, resulting in small indels that can generate a loss‐of‐function mutation. This forms the basis of NHEJ‐based error‐prone editing. The HDR pathway involves sophisticated repair processing mechanisms led by 5′‐3′ DNA end resection to form 3′ single‐stranded DNA overhangs and the presence of undamaged sister chromatid or donor DNA. End resection begins when the MRN (MRE11‐RAD50‐NBS1) complex recruits CtBP‐interacting protein (CtIP), leading to the generation of short single‐stranded tails. Exo1 and the DNA2/BLM complex then perform long‐range resection, resulting in 3′ ssDNA tails. RPA stabilizes the ssDNA, which RAD51 subsequently replaces with the assistance of recombination mediators such as BRCA1, BRCA2, and PALB2. RAD51 forms nucleoprotein filaments on the ssDNA, facilitating homology search and strand invasion of a homologous DNA template. Finally, resolvases process the resolution junction, completing the repair and restoring the chromosome to its original state. B) Strategies to redirect cell's repair‐specific decision‐making choice toward HDR. The process of pathway decision‐making between NHEJ and HDR is influenced by several repair factors. As NHEJ is the default repair pathway, researchers aim to manipulate the factors that govern the transition between NHEJ and HDR to selectively enhance HDR over NHEJ. a) Exploiting NHEJ inhibition is a main strategy that favors the HDR pathway. This includes targeting 53BP1, a key promoter of the NHEJ pathway, through methods such as using ubiquitin variants (i53) or expressing factors that displace 53BP1 like (e18). It also involves using small molecule inhibitors against Ku70/80, DNA‐PKcs and LIG4, which are critical components of the NHEJ machinery. b) Manipulating the cell cycle to favor HDR that is active during the S/G2/M phases, in contrast to NHEJ, which operates throughout the cell cycle. This strategy involves the use of compounds to block cells in HDR‐permissive phases. c) Developing a Cas9‐CtIP fusion protein, d) dual molecules of CtIP (2x CtIP) to Cas9 using MS2 tagging, e) or fusing Cas9 with a small motif of BRCA2 (Brex27), forming MiCas9, enables end‐resection. f) Rational design of the ssODN donor template to be complementary to the non‐target strand can also improve the efficiency of precise editing.

Figure 3

A schematic illustration of CRISPR base editors. An illustration depicting various base editors. A) Base editing (BE) includes cytosine, adenine, guanine, and thymine base editors. A) The cytosine base editing involves nCas9 protein fused to a) a cytidine deaminase enzyme, rAPOBEC1. This complex targets the DNA and converts cytosine (C) to uracil (U), which is read as thymine (T) during DNA replication, resulting in C‐to‐T substitutions. b) Combining an additional uracil glycosylase inhibitor (UGI) blocks the initial step of the base excision repair process, thereby favoring the retention of deaminated C as U. c) Conversely, when incorporating a uracil DNA N‐glycosylase (UNG) facilitates C‐to‐G transitions at the abasic sites. d) The CDG is directly fused to the nCas9‐mediated C‐to‐G conversion process, involving the specific binding of the CDG‐nCas9 complex to target genomic DNA loci. This complex directly excises cytosine or thymine, creating an apurinic site and enabling base editing. B) The adenine base editing utilizes nCas9 fused to a) an adenine deaminase enzyme derived from the tRNA deaminase TadA7.10. The ABE complex binds to the target DNA and converts adenine (A) to guanine (G), resulting in an A‐to‐G substitution. b) TadA8e is an enhanced version of TadA7.10 derived from phage‐assisted non‐continuous and continuous evolution, showing a preference for A‐to‐G conversion. c) Fusing the TadA8e to an alkyladenine DNA glycosylase (AAG) or d) an N‐methylpurine DNA glycosylase (MPG) permits A‐to‐C or A‐to‐T conversions. C) Thymine deaminase‐free base editor derived from engineered UNG attached directly to nCas9 without the need of any deaminase facilitates the substitution of T‐to‐G or T‐to‐C. D) Guanine deaminase‐free base editor derived from mutagenized MPG attached directly to nCas9 without the need of any deaminase induces the substitution of G‐to‐T or G‐to‐C.

Figure 4

A schematic illustration of CRISPR prime editing. A) Prime editing uses a prime editing guide RNA (pegRNA) to direct the nCas9 protein that nicks one DNA strand and a reverse transcriptase enzyme that synthesizes the desired edit. This allows for a wide range of precise genetic modifications, including all base‐to‐base conversions, small insertions, and deletions. B) TwinPE system targets genomic DNA sequences with two protospacer sequences located on opposite strands. PE2–pegRNA complexes engage each protospacer, creating a single‐stranded nick and reverse transcribing the pegRNA‐encoded template with the desired insertion sequence. This process leads to the formation of 3′ DNA flaps, resulting in a hypothetical intermediate that has annealed 3′ flaps containing the edited sequence alongside 5′ flaps with the original DNA. The excision of the original sequence from the 5′ flaps, followed by the ligation of the 3′ flaps at the corresponding excision sites, produces the intended edited product. Abbreviations: PBS, prime binding site; RT, reverse transcriptase; RTT, reverse transcriptase template.

Figure 5

Overview of CRISPR/Cas combines transposases/recombinases for long sequence insertion. Type I‐F CRISPR‐associated transposase (CAST) and type V‐K CAST can facilitate the insertion of cargos. A) Type I‐F CAST binds to a target in an RNA‐guided manner, aided by bacterial proteins TniQ and TnsA‐C, along with the expression of CASCADE (Cas) proteins; Cas6, Cas7, and Cas8‐5; which help disassemble the post‐transposition complex. B) In contrast, type V‐K CAST also binds to a target in an RNA‐guided manner but requires the coexpression of TniQ and a bacterial S15 protein. Both CAST systems introduce target site duplications and leave scars from left end (LE) and right end (RE) of the cargo. This prime editing‐based insertion strategy relies on integrase or recombinase‐catalyzed donor insertion. C) Similar to the TwinPE system, attB and/or attP sites can be integrated at the edited site to enable subsequent insertion of large DNA sequences. This is achieved by delivering a plasmid containing the large DNA sequence flanked by attP sites, along with a coding sequence for the serine recombinase Bxb1, which catalyzes the integration of the DNA cargo. D) Programmable addition through site‐specific targeting element (PASTE) is designed for targeted insertion. An attachment site‐containing guide RNA (atgRNA) directs a PASTE fusion complex; comprising Cas9 nickase (nCas9), reverse transcriptase (RT), and integrase; to a specific genomic locus, facilitating the integration of an attB site. The integrase then recognizes the attB site and integrates a large DNA fragment flanked by attP sites without creating a double‐stranded break (DSB).

Figure 6

CRISPR‐mediated transcriptional modifications. (A) CRISPR‐mediated transcriptional activation (CRISPRa). a) The CRISPRa‐VP64 system combines dead Cas9 (dCas9) and a guide RNA (gRNA) scaffold with the VP64 transcriptional activator to recruit the transcriptional activation domain to the target gene. b) For robust transcriptional activation, three strong activation domains—VP64, p65, and Rta—are fused to dCas9, resulting in a higher transcriptional output compared to VP64 alone. c) The SunTag system uses a repeating peptide array (SunTag) fused to dCas9, which recruits multiple copies of an antibody‐fused activation domain (such as VP64). Each peptide in the array can bind an activator, thus amplifying the recruitment of transcriptional activators to the target gene. d) SAM (Synergistic Activation Mediator) combines dCas9 with a modified gRNA scaffold that recruits additional activation domains, MS2, p65, and HSF1. e) CRISPRon involves the fusion of dCas9 with a modified gRNA scaffold that recruits Rta, p65, and MS2, which together bind to TET1. The dCas9‐TET1 complex binds to the target gene, where TET1 induces DNA demethylation, leading to transcriptional activation. f) CRISPRon involves the induction of acetylation and gene expression through the fusion of dCas9 with the catalytic core of the human acetyltransferase p300. This fusion facilitates the acetylation of histone H3 at lysine 27 in its target regions, resulting in strong transcriptional activation of target genes from their respective promoters. B) CRISPR‐mediated transcriptional inhibition (CRISPRi). a) The dCas9‐KRAB complex binds to the target gene, recruiting the repressive chromatin modifier (Krüppel‐associated box) to silence gene expression. b) The dCas9‐MeCP2 complex binds to the target gene, utilizing MeCP2's repressive function to inhibit gene transcription. c) The dCas9‐SID4X complex recruits additional repressive complexes to the target gene. d) CRISPRoff involves the binding of the dCas9‐KRAB fusion to the DNA methyltransferases Dnmt3A and Dnmt3L. DNA methyltransferase domains facilitate the methylation of CpG sites in the target genomic region, resulting in stable transcriptional repression. e) The CHARM system utilizes a coupled histone tail (H3‐tail) and the Dnmt3L protein to modulate the autoinhibition of methyltransferases. When the H3‐tail binds to Dnmt3L. It triggers a conformational change that releases the autoinhibitory domain of the methyltransferase, leading to its activation that facilitates endogenous DNA methylation and gene silencing.

Figure 7

CRISPR‐mediated RNA modifications. This figure illustrates the process of RNA modifications using CRISPR technology. A) RNA Knockdown via Cas13, Cas7‐11, and Csm. a) The Cas13 system is a CRISPR‐associated, RNA‐guided nuclease that targets and cleaves single‐stranded RNA. A distinctive feature of Cas13 is collateral cleavage, which results in the degradation of non‐target RNAs. b) In contrast, Cas7–11 is a large single‐protein effector that enables RNA‐guided RNA degradation with reduced cellular toxicity and no collateral activity. c) CRISPR–Csm complex is a multisubunit structure that also utilizes a programmable RNA‐guided mechanism to degrade target RNAs without causing collateral damage. B) Precise nucleotide editing of RNA is achieved through deaminase‐mediated conversion of adenosine to inosine (A‐to‐I) and cytidine to uridine (C‐to‐U). a,b) RNA A‐to‐I base editing relies on ADAR (adenosine deaminases acting on RNA) enzymes to target double‐stranded RNA (dsRNA) substrates. ADAR deaminases can be directed to specific sites using catalytically inactive Cas13 (dCas13). Alternatively, guide RNAs can be fused to the bacteriophage MS2 coat protein (MCP). The MCP attaches to the MS2 stem‐loop RNA, and the MS2 loop region subsequently recruits the MCP‐ADAR2 to the editing site. c,d) RNA C‐to‐U base editing involves two main approaches: RESCUE and CURE. In the RESCUE approach, an artificially evolved ADAR2 is used to deaminate cytosine, recruited to target sites via dCas13. In the CURE approach, dCas13 mediates the recruitment of the APOBEC3A mutant to deaminate RNA within a small loop substrate.

Figure 8

Schematic illustration of strategies for enhancing CRISPR/Cas technology in genome editing, focusing on specificity, precision, accessibility, flexibility, and safety. To enhance CRISPR specificity, strategies include the identification and characterization of genetic variants (GVs), the use of advanced computational tools for off‐target analysis, high‐fidelity Cas9 variants, and optimized sgRNA design can result in enhancing GV sensitivity, and reducing off‐target edits that increase targeting specificity, ensure safety, and avoid unintended modifications. To improve precision, employing engineered Cas9 variants, alternative nucleases (e.g., Cpf1), and advanced tools like base and prime editors can enhance targeting fidelity, allow genetic corrections, increase editing accuracy, and predict editing outcomes. For accessibility, we highlight approaches including optimizing delivery methods, managing packaging size, enhancing cellular uptake, modulating repair pathway choices, and fine‐tuning editing factors. These strategies are designed to improve delivery efficiency, bias the cellular repair pathway toward desired outcomes, and enhance the efficiency of Homology Directed Repair (HDR). Additionally, the CRISPR complexes are illustrated alongside strategies that utilize Cas9 variants with relaxed or PAM‐free options to expand targeting capabilities. Key approaches include using diverse Cas variants (e.g., Cas12b, Cas3) to broaden the targeting scope and enhance adaptability across different biological contexts and organisms. The safety aspect emphasizes the use of non‐immunogenic delivery vehicles, comprehensive off‐target characterization, and the development of therapeutic applications while addressing potential risks. This comprehensive approach aims to enhance the efficacy and safety of CRISPR‐based interventions.