Prime editing in neuropsychiatric disorders: From mutation‐specific target selection to clinical translation

Abstract

Prime editing, a novel clustered regularly interspaced short palindromic repeats (CRISPR)‐based technology, fuses a reverse transcriptase (RT) to an engineered CRISPR‐associated protein 9 (Cas9) and uses a prime editing guide RNA (pegRNA)‐encoded template. It enables precise base substitutions, small insertions, and deletions without introducing double‐strand breaks, thereby expanding the range of correctable mutations while reducing undesired repair outcomes. This technology offers a promising strategy for genomic correction in the nervous system. Here, we review the development of prime editing, its mechanistic rationale, and emerging preclinical evidence that supports its application in neuropsychiatric disorders. We discuss key biological and technological barriers, including limited editing efficiency in post‐mitotic neurons, complex pegRNA design, reverse transcription‐related errors, vector payload limitations, and blood‐brain barrier (BBB) penetration. Nevertheless, in vitro and in vivo studies have demonstrated proof‐of‐concept correction and functional rescue in several monogenic neurodevelopmental disorders. Advances such as split‐adeno‐associated virus (AAV) systems, lipid nanoparticles, engineered peptides, and compact Cas variants are actively expanding their therapeutic potential. Further clinical translation will rely on improved editors with guide engineering, BBB‐penetrant and neuron‐targeted delivery platforms, transient or cell‐type‐specific expression strategies, and comprehensive genome‐wide safety evaluations.

Highlights

• Prime editing enables precise genomic correction without inducing double‐strand breaks, offering a safer alternative to traditional clustered regularly interspaced short palindromic repeats (CRISPR)‐CRISPR‐associated protein (Cas) 9 editing.

• The technology can correct diverse mutation types, including single‐nucleotide variants and small insertions/deletions, relevant to neuropsychiatric disorders.

• Current preclinical studies demonstrate functional rescue in monogenic neurodevelopmental disease models.

• Key translational challenges include delivery to post‐mitotic neurons, cellular heterogeneity, and off‐target effects.

• Advances in delivery systems, pegRNA engineering, and editor optimization are paving the way for clinical applications in central nervous system disorders.

1. INTRODUCTION

1.1. Prime editing

Conventional clustered regularly interspaced short palindromic repeats (CRISPR)‐CRISPR‐associated protein 9 (Cas9) introduces DNA double‐strand breaks (DSBs) and relies on endogenous DNA repair mechanisms to achieve targeted modification. However, DSB‐dependent editing often results in genomic instability, including undesired insertions or deletions (indels) and chromosomal rearrangements. ¹ , ² Primary endogenous DSB repair pathways, such as non‐homologous end joining (NHEJ) and homology‐directed repair (HDR), each have limitations. NHEJ is intrinsically error‐prone and frequently generates random indels at the junction site. ³ HDR demonstrates higher fidelity and precision but offers variable and relatively low efficiency across cell types, especially in nondividing cells, which limits its therapeutic applicability. ¹ , ²

To overcome these limitations, base editing systems were developed to introduce precise single‐nucleotide conversions without generating DSBs. ⁴ , ⁵ Although effective, base editors are constrained to single‐base edits and can result in substantial off‐target and bystander editing. ⁶ , ⁷ , ⁸ Motivated by the need for a more precise and reliable gene editing tool, prime editing was introduced in 2019. ⁹

Prime editing employs a Cas9 nickase (H840A variant, a single‐strand cutting form of Cas9) and a fused reverse transcriptase (RT), guided by a prime editing guide RNA (pegRNA). ¹⁰ , ¹¹ The pegRNA specifies the genomic target and encodes an RT template that directs precise sequence changes at the locus. After the nick is introduced, the RT synthesizes the desired sequence directly at the target site, allowing replacement, insertion, or deletion of base pairs without generating DSBs or relying on additional exogenous donor templates. ¹² , ¹³ , ¹⁴ Since its introduction, the system has undergone substantial refinement, including the development of enhanced editor variants (such as PE2, PE3, and PEmax), engineered pegRNAs, dual prime editing strategies, and intein‐mediated protein trans‐splicing to overcome size constraints. ⁵ These innovations are collectively aimed at improving efficiency, precision, and cell‐type compatibility.

Technically, the advantage of prime editing lies in its versatility and precision. Compared with traditional base editors, prime editors can perform a broad spectrum of small DNA edits, including single‐base transitions, short insertions or deletions, and multiplex editing. ¹⁴ By avoiding DSBs, prime editing reduces the risk of chromosome breakage or translocation, large deletions, and unintended indels. ¹² , ¹³ The performance of prime editing systems can be further optimized by refining pegRNA design, enhancing reverse transcriptase activity, and exploring alternative Cas variants. Recent research has shown that engineered prime editors demonstrate improved editing efficiency in plants, holding significant promise for optimization of these systems. ¹⁵ Additionally, machine learning has been employed to predict editing efficiency across diverse chromatin contexts, enabling more precise and effective gene editing. ¹⁶ Delivery methods also play a crucial role; for example, pseudoviral NanoScribe particles have proven effective in delivering prime editing components into human stem cells, offering a novel approach to enhance gene editing in human cells. ¹⁷ Ribonucleoproteins (RNPs), when delivered via optimized lipid nanoparticle formulations, can improve both the safety and efficiency of base and prime editing. ¹⁸ Moreover, in vivo applications, such as mRNA delivery using acid‐degradable lipid nanoparticles, have achieved widespread gene editing in the brain, underscoring the potential of lipid nanoparticles for therapeutic gene editing. ¹⁹ These strategies can limit exogenous protein exposure, thereby reducing immunogenicity and off‐target effects.

Despite these advantages, there are still several limitations. Its efficacy in post‐mitotic cells, including mature neurons, remains suboptimal. ²⁰ The design of pegRNAs is complex and highly variable. Vector payload limits in vivo delivery options, especially for applications requiring adeno‐associated virus (AAV). To address these challenges, enzyme engineering, delivery innovation, and computational designs that enhance pegRNA performance are required (Figure 1). ²¹ , ²²

Figure 1.

Mechanism, optimization and applications of prime editing. Prime editing facilitates precise genome editing through direct DNA synthesis at the target site. Its versatility supports diverse genome modifications, offering therapeutic potential for correcting pathogenic mutations underlying neurological and psychiatric disorders. CRISPR, clustered regularly interspaced short palindromic repeats; Cas, CRISPR‐associated protein; MT, mutant; PBS, primer binding site; PE, prime editing; RT, reverse transcriptase; WT, wild type.

1.2. Neuropsychiatric disorders and challenges in current treatments

Neuropsychiatric disorders encompass a broad and heterogeneous group of conditions, including neurodevelopmental disorders, emotional and mental illnesses, and late‐onset neurodegenerative diseases. These disorders can be classified as single‐gene neurodevelopmental syndromes such as Rett syndrome and fragile X syndrome; multifactorial conditions such as autism spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD); and neurodegenerative diseases such as Alzheimer's and Parkinson's disease. The etiology of these disorders varies. Some are attributed to rare monogenic mutations, while others result from complex interactions between genetic susceptibility and environmental factors.

From a genome editing perspective, neuropsychiatric disorders can also be categorized based on the type of underlying genetic variation and their theoretical amenability to precise sequence correction. Pathogenic variants include single‐nucleotide substitutions, indels, splice‐site alterations, repeat expansions, and large copy number variations (CNVs). Among these, single‐nucleotide variants and short indels are particularly well‐suited for template‐directed correction strategies such as prime editing, whereas large structural rearrangements and repeat expansions remain technically more challenging. Therefore, understanding the mutational architecture of each disorder is critical for rational target selection in genome‐editing‐based interventions.

The pathogenetic mechanisms of these disorders are multifactorial and exhibit common themes across categories (Figure 2). For example, Alzheimer's disease and Parkinson's disease exemplify progressive neuronal dysfunction driven by synaptic abnormality, circuit disruption, mitochondrial dysfunction, and chronic neuroinflammation. ²³ , ²⁴ Environmental factors, including pesticide exposure, further increase dopaminergic vulnerability in Parkinson's disease, while alterations in defined neuronal subtypes can exacerbate disease‐associated phenotypes. ²⁵ , ²⁶ Mitochondrial dysfunction and bioenergetic decline represent central, cross‐disease features. ²⁷ Calcium dyshomeostasis constitutes an additional mechanistic axis, supported by studies employing genetically encoded sensors such as Yellow Cameleon 3.6 in Alzheimer's disease. ²⁸ Increasing evidence also implicates nonneuronal elements, including satellite glial cells, alongside disrupted detoxification and metabolic stress pathways in age‐related neurodegeneration. ²⁹

Figure 2.

Multifactorial etiology of neuropsychiatric disorders. Genetic susceptibility, including monogenic mutations, interacts with diverse risk factors such as neurotransmitter alterations, infections, somatic conditions, abnormal neurodevelopment, and epigenetic modifications to shape disease pathogenesis.

In contrast, the genetic landscape of amyotrophic lateral sclerosis (ALS) has become clearer, with findings pointing to mutations in genes like SOD1, TDP‐43, FUS, and C9ORF72 as key contributors to disease onset. ³⁰ , ³¹ , ³² , ³³ Regarding ASD, while it involves polygenic risk and environmental modulation, a subset of cases is driven by highly penetrant single‐gene mutations, which may be more suitable for mutation‐specific correction strategies. This distinction is crucial when evaluating the translational feasibility of prime editing for neurodevelopmental disorders. ³⁴ , ³⁵ These observations highlight substantial genetic and mechanistic heterogeneity, which complicates the identification of uniformly effective therapeutic targets.

Current standard therapies for neuropsychiatric conditions include pharmacotherapy, behavioral therapy, and rehabilitation. Current interventions predominantly provide symptomatic relief without modifying the underlying genetic or molecular pathology. Psychotropic agents often result in variable treatment outcomes and notable side effects. ³⁶ , ³⁷ For developmental disorders caused by mutations, conventional interventions cannot reverse established neural circuit defects. ³⁸ Genetic and molecular approaches to treating central nervous system (CNS) disorders face several significant biological and technical challenges. The blood‐brain barrier (BBB) remains a major obstacle, limiting the delivery of therapeutic agents to the brain and driving recent advancements in imaging technologies designed to overcome it. ³⁹ Moreover, the limited regenerative capacity of mature neurons complicates efforts to repair neural damage. While innovative approaches such as bio‐nanotechnology are being explored for precision gene therapy in Alzheimer's disease, these methods still struggle with targeting efficiency and long‐term stability. ⁴⁰ Additionally, the need for long‐term safety and stability of in vivo therapeutic agents remains a critical issue, as addressed by studies on various gene therapy modalities and their translational challenges. ⁴¹ These challenges emphasize the complexity of CNS gene therapy and the need for continued research to develop strategies that ensure both effective delivery and safety.

2. CLINICAL APPLICATIONS OF TRADITIONAL GENE EDITING IN NEUROPSYCHIATRIC DISORDERS AND THEIR LIMITATIONS

Gene editing technologies have rapidly progressed from experimental tools to clinically validated therapies. Ex vivo CRISPR‐based treatments such as exa‐cel (Casgevy) demonstrate durable therapeutic benefit in hematologic diseases through targeted disruption of BCL11A. ⁴² In vivo systems, including VERVE‐101 (a base editing therapy for familial hypercholesterolemia) and EDIT‐101 (a CRISPR nuclease therapy for inherited retinal disease), have established the feasibility of delivering gene editors directly into human tissues. ⁴³ , ⁴⁴ , ⁴⁵ These successes illustrate the transformative potential of genome editing for correcting pathogenic variants or modulating gene expression in a durable manner (Table 1).

Table 1.

The clinical landscape of in vivo gene‐editing therapies.

Disease indication	Therapeutic approach	Phase	Clinical trial
Acute lymphoblastic leukemia	CAR‐T cell therapy	Approved	NCT02435849
Adenosine deaminase severe combined immunodeficiency	Lentiviral gene therapy	Approved	NCT01852071
Beta‐thalassemia	Ex vivo CRISPR/Cas9 edited autologous CD34+ cells and lentiviral gene therapy	Approved	NCT03655678; NCT02906202
Cerebral adrenoleukodystrophy	Lentiviral gene therapy	Approved	NCT01896102
Choroideremia	AAV gene therapy	III	NCT03496012
Cystic fibrosis	CFTR with a deletion in the regulatory domain	I/II	NCT05248230
Diffuse large B‐cell lymphoma	CAR‐T cell therapy	Approved	NCT02348216
Duchenne muscular dystrophy	Mini‐dystrophin	I/II; III; approved	NCT03368742; NCT04281485; NCT05096221
Geographic atrophy	AAV gene therapy	II	NCT04514445
HIV	SaCas9 with guide RNAs targeting the HIV genome	I	NCT05144386
Hemophilia A	AAV gene therapy and FVIII replacement	Approved/I/II; approved	NCT03392974; NCT03033533; NCT04323098
Hemophilia B	FIX	III; approved	NCT03861273; NCT03569891
Hereditary angioedema	SpCas9 with guide RNA targeting the gene encoding kallikrein B1	I/II	NCT05120830
Heterozygous familial hypercholesterolaemia	Adenine base editor with guide RNA targeting the gene encoding PCSK9	I	NCT05398029
Leber congenital amaurosis 10	SaCas9 with guide RNAs targeting the CEP290 mutation	I/II	NCT03872479
Leber congenital amaurosis type 2	RPE65	Approved	NCT00999609
Mantle cell lymphoma	CAR‐T cell therapy	Approved	NCT02601313
Metachromatic leukodystrophy	Lentiviral gene therapy	Approved	NCT01560182
Multiple myeloma	CAR‐T cell therapy	Approved	NCT03361748; NCT03548207
Primary immunodeficiency with Wiskott‐Aldrich syndrome	Lentiviral gene therapy	III	NCT03837483
Retinitis pigmentosa	AAV gene therapy	I/II	NCT03328130
Sickle cell disease	Ex‐vivo CRISPR/Cas9 edited autologous CD34+ cells	Approved	NCT03745287
Spinal muscular atrophy type 1	SMN	Approved	NCT03306277
Stargardt disease	AAV gene therapy	I/II	NCT01367444
Transthyretin amyloidosis	CRISPR gene editing and RNAi targeting transthyretin	Approved/I	NCT01960348; NCT04601051
Usher syndrome type 1B	AAV gene therapy	I/II	NCT02065011
Vaccination	SARS‐CoV‐2 spike protein	Approved	NCT04368728; NCT04470427
Wet age‐related macular degeneration	Aflibercept and an anti‐VEGFC RNAi	I/II	NCT05197270
Wiskott‐Aldrich syndrome	Lentiviral gene therapy	I/II	NCT02333760
X‐linked severe combined immunodeficiency	Lentiviral gene therapy	Approved	NCT01410019

Abbreviations: AAV, adeno‐associated virus; CAR‐T, chimeric antigen receptor T cells; CD34+, hematopoietic stem/progenitor cells; FVIII, coagulation factor VIII; RPE65, retinal pigment epithelium‐specific 65 kDa protein; SaCas9, Staphylococcus aureus Cas9; SpCas9, Streptococcus pyogenes Cas9; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2; SMN, survival motor neuron protein; VEGFC, vascular endothelial growth factor C.

However, applying traditional gene editing strategies to neuropsychiatric disorders presents biological and technical challenges. ⁴⁶ , ⁴⁷ Most gene editing tools can trigger undesired indels, chromosomal rearrangements, and activation of DNA damage responses, raising significant safety concerns for the DNA damage‐sensitive, nonregenerating neurons. ⁴⁸

Base editors provide an alternative by enabling DSB‐free nucleotide conversions, but they present notable constraints for their therapeutic use in the CNS (Figure 3). Operating at the single‐nucleotide level, they are incapable of correcting larger pathogenic lesions, including frameshifts, multi‐base insertions or deletions, exon losses, and complex structural variants. ⁴⁹ Delivery also remains a central bottleneck for base editors due to their large size. Furthermore, their clinical translation is hampered by pre‐existing immunity to viral or bacterial components and risks of sustained intracellular expression. ⁴⁰ The aforementioned technical and biological limitations preclude base editing from being a turnkey solution for many neuropsychiatric indications.

Figure 3.

CRISPR‐based genome editing platforms for neuropsychiatric disorders. CRISPR‐Cas nucleases (Cas9/Cas12, CRISPR‐associated nucleases) mediate gene disruption or precise insertion via nonhomologous end joining (NHEJ) or homology‐directed repair (HDR). Base editors enable targeted point mutations without double‐strand breaks (DSBs). Prime editors support precise installation of point mutations and small indels through template‐directed reverse transcription, offering versatile tools for correcting pathogenic variants. Cas, CRISPR‐associated protein; CRISPR, clustered regularly interspaced short palindromic repeats; crRNA, CRISPR RNA; PAM, protospacer adjacent motif; pegRNA, prime editing guide RNA; RT, reverse transcriptase; sgRNA, single guide RNA.

3. CURRENT RESEARCH STATUS OF PRIME EDITING IN NEUROPSYCHIATRIC DISORDERS

Prime editing, with its capacity to correct diverse mutation types without generating DSBs, represents a particularly promising tool for the treatment of neuropsychiatric disorders. Evidence from in vitro and animal model studies confirms the capacity of prime editing to mediate precise base substitutions and small insertions in various cell types. Optimized systems such as PEmax and engineered pegRNAs have demonstrated robust editing efficiency in the CNS (Figure 3).

The suitability of prime editing for neuropsychiatric disorders depends critically on the underlying mutational architecture. Disorders driven by single‐nucleotide substitutions or short indels are theoretically optimal candidates for template‐directed correction, whereas large repeat expansions or copy number variations may require alternative approaches. Consequently, current research efforts prioritize monogenic, high‐penetrance variants in which partial restoration of gene function could provide meaningful therapeutic benefit.

Recent progress on multiple fronts is helping to overcome the main obstacles in translating prime editing for CNS applications. The landscape of delivery strategies of gene therapy has significantly diversified, incorporating both viral and nonviral platforms that cater to specific therapeutic needs. Advanced AAV systems, which employ techniques such as split‐intein reconstitution, the use of smaller Cas proteins, and BBB‐penetrant serotypes, have shown promise in achieving targeted delivery to the CNS. ⁵⁰ However, these approaches face challenges related to immune responses and production scalability. In contrast, nonviral approaches, including lipid nanoparticle‐mediated delivery of mRNA and ribonucleoproteins, enable transient expression and may reduce long‐term genomic risk. Emerging materials, such as glutathione‐responsive silica nanocapsules, further expand strategies for blood‐brain barrier penetration and cell‐type targeting. ⁵¹ , ⁵² Parallel advances in editor engineering and chemical modification continue to improve precision while reducing off‐target activity.

Prime editing has demonstrated satisfactory preclinical effects across various models, supporting its potential for precise gene correction and complex genomic engineering. Efficient editing of post‐mitotic neurons has been achieved in adult mouse brains following local delivery of engineered Cas9 ribonucleoprotein complexes, establishing the feasibility of in vivo CNS editing. ⁵³ Large‐animal models further enhance this potential, such as a nonhuman primate model of inherited retinal disease caused by a PDE6C mutation, which provides valuable systems for optimizing gene editing strategies in the macula. ⁵⁴ Mechanistic insights also inform therapeutic targeting; for example, studies of Neurofibromatosis type 1 indicate that neuronal hyperexcitability contributes to tumor progression, underscoring the rationale for precise neuronal genome intervention. ⁵⁵ Expanding toolkits such as Cas9‐based prime editing platforms capable of multiplexed precise edits further broaden the scope of programmable genomic modification. ⁵⁶ Together, these advances demonstrate both the feasibility and versatility of prime editing in CNS disease models, while highlighting the need for continued optimization of efficiency, scalability, and clinical applicability.

Nonetheless, prime editors are not without risk. Potential issues such as pegRNA‐derived byproducts, low‐frequency off‐target edits, reverse transcription‐associated events, and immune responses to editor components necessitate comprehensive genome‐wide and transcriptome‐wide safety profiling. ²⁰ , ⁵⁷ Before clinical translation, several hurdles must be overcome: comprehensive validation in physiologically relevant CNS models, rigorous long‐term safety and efficacy studies, and the advent of scalable, cell‐type‐specific delivery strategies. ⁴⁹ , ⁵⁸ , ⁵⁹

4. TRANSLATIONAL CHALLENGES OF PRIME EDITING IN THE TREATMENT OF NEUROPSYCHIATRIC DISORDERS

Prime editing represents a groundbreaking advancement in genetic modification, but its clinical application in the treatment of neuropsychiatric disorders currently faces several challenges. These challenges stem from reduced editing efficiency in postmitotic neurons, the cellular heterogeneity within the nervous system, the limitations of viral and nonviral delivery platforms, and safety concerns regarding off‐target effects and immune responses.

A primary limitation is the diminished editing efficiency observed in mature neurons. Unlike proliferating cells, post‐mitotic neurons lack replication‐coupled DNA repair pathways that facilitate efficient integration of engineered edits. In dividing cells, S/G2‐phase repair processes support incorporation of template‐directed modifications. By contrast, neurons predominantly rely on maintenance pathways such as base excision repair, mismatch repair, and flap processing, which are less conducive to efficient integration of exogenous sequences. ⁶⁰ , ⁶¹ , ⁶² In addition, the relatively compact chromatin landscape of mature neurons can restrict access of pegRNAs and reverse transcriptase to target loci, further reducing editing efficiency. ¹⁶ , ⁶³

Cellular heterogeneity represents an additional obstacle. The brain comprises diverse neuronal and glial populations with distinct transcriptional programs, chromatin states, and DNA repair capacities. ⁶⁴ , ⁶⁵ These differences substantially influence prime editing performance, such that optimization in one cell type may not translate to others. Disease‐associated molecular changes, including neuroinflammatory remodeling and epigenetic reprogramming, can further alter editing outcomes. ⁶⁶ , ⁶⁷ Achieving consistent and therapeutically meaningful correction, therefore, requires cell‐type‐specific and context‐aware optimization strategies.

Delivery remains a central translational bottleneck. Viral vectors such as AAV enable efficient neuronal transduction but are constrained by limited payload capacity and potential immunogenicity, particularly in the setting of repeat administration. ⁶⁸ , ⁶⁹ Nonviral systems, including lipid nanoparticles and engineered extracellular vesicles, allow transient expression and larger cargo delivery but face challenges in blood‐brain barrier penetration and cell‐type specificity. ⁷⁰ , ⁷¹ Balancing efficiency, payload size, targeting precision, and safety remain a key determinant of clinical feasibility.

Safety considerations are equally critical. Although prime editing reduces double‐strand break‐associated risks, low‐frequency off‐target edits, pegRNA‐derived byproducts, reverse transcription‐associated events, and immune responses to editor components remain concerns. ⁷² Comprehensive genome‐wide and transcriptome‐wide profiling, along with long‐term functional assessment in relevant CNS models, will be essential before clinical application.

In conclusion, while prime editing offers a powerful framework for precision correction of pathogenic variants, overcoming barriers related to neuronal biology, cellular heterogeneity, delivery logistics, and safety profiling is essential to enable its translation into effective therapies for neuropsychiatric disorders.

5. FUTURE PROSPECTS OF PRIME EDITING IN THE TREATMENT OF NEUROPSYCHIATRIC DISORDERS

Despite current limitations, prime editing remains a highly promising strategy for the treatment of neuropsychiatric disorders, particularly those driven by defined genetic variants. Realizing its therapeutic potential will require coordinated advances in molecular engineering, neuronal biology, delivery science, and safety assessment.

Improving editing efficiency in post‐mitotic neurons is a central priority. Rational modulation of endogenous DNA repair pathways, such as through transient small‐molecule enhancers or auxiliary repair factors, may increase template incorporation without disrupting neuronal homeostasis. ⁷³ Parallel efforts to engineer next‐generation editors, including optimized reverse transcriptases, enhanced pegRNA architectures (e.g., epegRNAs), could further improve efficiency and fidelity in the CNS environment. ⁷⁴ , ⁷⁵

Specificity across heterogeneous neural cell populations must also be refined. Cell type‐restricted promoters may enable more controlled and context‐sensitive editing. ⁴⁹ Such strategies are particularly important given the dynamic transcriptional and chromatin landscapes observed in neurodegenerative and psychiatric conditions. ⁶⁴

Delivery innovation will be equally decisive. Next‐generation viral systems with compact editor configurations and engineered serotypes optimized for CNS tropism may expand the feasibility of in vivo correction. ⁶⁸ , ⁷⁵ In parallel, nonviral platforms, including lipid nanoparticles, engineered extracellular vesicles, and receptor‐targeted nanocarriers, offer opportunities for transient, repeatable, and potentially less immunogenic administration. ¹⁸ , ⁷⁰ Approaches that enhance blood‐brain barrier penetration or enable region‐specific targeting may further increase therapeutic precision. ⁷⁶ , ⁷⁷

Comprehensive safety evaluation will underpin clinical translation. Advanced genome‐wide off‐target detection methods, long‐read sequencing to capture structural variants, transcriptome‐wide analyses, and long‐term functional assessment in relevant animal models will be essential to define acceptable therapeutic windows and risk thresholds. ⁷²

Looking forward, the application of prime editing in neuropsychiatric disorders will revolutionize the treatment of genetic conditions such as autism, schizophrenia, and neurodegenerative diseases. With continued optimization in enzyme, guide RNA, and delivery systems, prime editing could enable precise, site‐specific genetic corrections in vivo, offering long‐term therapeutic solutions. The ongoing refinement of these technologies, combined with a deeper understanding of neuronal biology, will pave the way for successful clinical applications of prime editing, providing new treatments for patients with genetic neuropsychiatric conditions (Figure 4).

Figure 4.

Application of prime editing in neuropsychiatric disease. Prime editing (PE) systems enable precise genomic correction and can be delivered via viral (adeno‐associated virus [AAV]) or nonviral (lipid nanoparticles [LNPs]) vectors. This approach facilitates targeted repair of pathogenic mutations and functional rescue in affected cells, supporting the potential treatment of genetic neuropsychiatric disorders. BBB, blood‐brain barrier; MMLV, moloney murine leukemia virus; PBS, primer binding site; RTT, reverse transcriptase template.

6. CONCLUSION

Prime editing represents a major conceptual and technical advance for precise and versatile genomic correction. It is especially relevant for neuropsychiatric diseases caused by single‐nucleotide changes or small indels, which often lead to severe and early‐onset phenotypes. Preclinical studies in cells and animal models have provided proof‐of‐concept for correcting pathogenic alleles and have informed the development of delivery strategies that can access the CNS. However, its low efficiencies in mature and post‐mitotic neurons and the need for scalable and safe in vivo delivery pose challenges to its clinical translation.

To realize the therapeutic potential of prime editing, coordinated progress on multiple fronts is required. Continued enzyme and pegRNA engineering can be utilized to increase activity and fidelity in neurons. To limit immunogenicity and prolonged exposure, compact or split delivery systems can be developed, and transient, cell‐type‐restricted expression paradigms can be implemented. Equally important are rigorous, genome‐wide safety and off‐target assays, comprehensive toxicology in relevant large‐animal models, and pragmatic clinical development strategies that focus on high‐penetrance, monogenic disorders with clear unmet need.

Successful translation will depend on aligned advances in molecular engineering, delivery science, preclinical modeling, and clinical trial design, supported by appropriate regulatory pathways and long‐term monitoring approaches. With these joint efforts, prime editing has the potential to move from a powerful research tool to a viable therapeutic modality for a subset of neuropsychiatric diseases. The outlook is optimistic: the approach is well‐suited to precise correction in the nervous system, but its clinical realization requires addressing delivery, efficacy, and safety.

AUTHOR CONTRIBUTIONS

Tianshan Ji: Conceptualization (equal); data curation (equal); formal analysis (equal); methodology (equal); validation (equal); visualization (equal); writing—original draft (equal). Yuan Zhang: Conceptualization (equal); data curation (equal); formal analysis (equal); methodology (equal); resources (equal); validation (equal); visualization (equal); writing—original draft (equal); writing—review and editing (equal). Jinyi Zhao: Conceptualization (equal); formal analysis (equal); methodology (equal); writing—original draft (equal). Yiping Lu: Conceptualization (equal); data curation (equal); methodology (equal). Chengkun Wang: Funding acquisition (lead); investigation (lead); project administration (lead); supervision (lead); writing—original draft (equal); writing—review and editing (equal).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHIC STATEMENT

Not applicable.

Ji T, Zhang Y, Zhao J, Lu Y, Wang C. Prime editing in neuropsychiatric disorders: from mutation‐specific target selection to clinical translation. Neuroprotection. 2026;1‐11. 10.1002/nep3.70041

DATA AVAILABILITY STATEMENT

All relevant data are within the paper. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.