Dead Cas(t) light on new life: CRISPRa-mediated reprogramming of somatic cells into neurons

Abstract

Overexpression of exogenous lineage-specific transcription factors could directly induce terminally differentiated somatic cells into target cell types. However, the low conversion efficiency and the concern about introducing exogenous genes limit the clinical application. With the rapid progress in genome editing, the application of CRISPR/dCas9 has been expanding rapidly, including converting somatic cells into other types of cells in vivo and in vitro. Using the CRISPR/dCas9 system, direct neuronal reprogramming could be achieved by activating endogenous genes. Here, we will discuss the latest progress, new insights, and future challenges of the application of the dCas9 system in direct neuronal reprogramming.

Introduction

Central nervous system (CNS) diseases, including brain injury, spinal cord injury, and neurodegenerative diseases, such as Alzheimer's disease, amyotrophic lateral sclerosis, Parkinson's disease, and Huntington's disease, are usually accompanied by progressive loss and degeneration of central neurons [1–3]. In all these disorders, neuronal loss results in permanent damage, as neurogenesis does not occur in most adult mammalian CNS. At present, the main clinical treatments for nerve injury or neuronal loss include surgery, drug delivery, protection of residual neuronal function, and later rehabilitation treatment, which show the trivial effect on neurological degeneration due to the irreversible terminal differentiation of neurons and the complexity of the nervous system. So far, neuronal repair and functional reconstruction remain among the most challenging medical problems.

In recent years, stem cell-based cell replacement therapy has provided an alternative strategy for these refractory diseases by supplementing lost neurons, emerging as a promising approach to regenerative medicine [4]. However, the cell source still limits the clinical application of stem cells. In 2006, Yamanaka et al. reported the generation of induced pluripotent stem cells (iPSCs) from mouse embryonic and adult fibroblasts by introducing four pluripotency transcription factors Oct4, Sox2, Klf4, and c-Myc [5]. The generated iPSCs exhibited embryonic stem cell (ESC) morphology and growth properties and expressed ESCs marker genes [5]. Reprogramming terminally differentiated somatic cells into iPSC can produce patient-specific cells for autogenous transplantation, establish human disease models, and facilitate drug screening. Moreover, the production of iPSCs from patient-specific autogenous somatic cells eliminates the crucial issues associated with immune incompatibility and ethical concerns [6]. However, the application of iPSCs still faces significant challenges, such as the risk of tumor formation, inflammatory response, and differentiation efficiency [7].

Direct lineage reprogramming, or transdifferentiation, is a process of reprogramming a specific somatic cell type to another without passing through a pluripotent state, thus eliminating iPSC-derived cells' tumorigenic potential [8]. Neurons, for example, could be converted from various cell types by overexpression of defined transcription factors [9]. Nevertheless, this approach is mainly limited by the size limit of the inserted open reading frame (ORF) sequence in the virus expression vector [10]. Furthermore, it is difficult to precisely imitate the expression of noncoding elements and complex transcriptional subtype variation. In addition, the genetic delivery of exogenous genes has certain safety concerns, such as genetic mutation and gene insertion [11].

Currently, dead Cas9 (dCas9) mediated gene activation has been successfully archived in different cell lines and primary cultures [12–17]. Compared with traditional transgenic technology, one innovative advantage of dCas9-based activation systems is the flexibility to use a mixture of diverse single-guide RNAs (sgRNAs) to achieve accurate and efficient regulation of multiple loci. Another potential advantage of the dCas9 system is related to epigenetic modification around endogenous loci, which is assumed to simulate natural epigenetic markers better [18]. Moreover, dCas9 system activates endogenous genes, which may have an advantage over traditional ORF overexpression of exogenous genes [18]. With the help of the dCas9 system, direct neural reprogramming possesses immense potential for treating neurological degeneration diseases. This review summarizes recent progress in dCas9-mediated direct reprogramming of terminally differentiated somatic cells into neurons.

Historical perspective on CRISPR/dCas9

The type II clustered regularly interspaced short palindromic repeat and Cas9 nuclease (CRISPR/Cas9) system are found in prokaryotes such as bacteria and archaea and repurposed as a tool for genome editing in mammalian cells [19–21]. In prokaryotic organisms, new exogenous DNA fragments from the infection are integrated into the host CRISPR sequence and transcribed into CRISPR RNA. These RNA sequences are used as a guide for identifying exogenous genetic elements from previously infected prokaryotes. Cas9 protein is a nuclease that can cut DNA fragments to match CRISPR RNA [22]. dCas9, a deactivated form of Cas9, has been engineered to fuse with transactivation factor to form programmable synthetic transcription factor (TFs), which is termed the CRISPR activation (CRISPRa) system [12, 16, 23]. There are three major components of the CRISPRa system: the dCas9 nuclease protein, a single-stranded guide RNA (sgRNA), and transcriptional activators. TFs can be programmed to target any genomic locus of interest by simply exchanging 20 nt targeting sequence of the sgRNA, making the CRISPRa system a simple, robust, and highly scalable approach for controlling complex transcriptional networks [24]. This system reportedly can target constitutively silenced genes within compacted chromatin with high precision to initiate chromatin remodeling and promote downstream gene transcription [13, 14].

dCas9 and sgRNA

In 2012, Jinek et al. mutated the RuvC1 and HNH nuclease domains (D10A and H841A, respectively) of Streptococcus pyogenes Cas9 (SpCas9) to produce a nuclease-deficient dCas9 [19]. Although it lost its cleavage activity, dCas9 can still target and bind DNA with the same accuracy under the guidance of sgRNA without irreversibly altering the genome. Instead, it interferes with the transcription of the target site with the help of transcriptional activators or inhibitors fused to its termini, resulting in reversible gene silencing. In contrast, the standard CRISPR/Cas9 system relies on introducing double-strand breaks (DSBs) into DNA through the activity of Cas9 endonuclease and then manipulating the DNA repair mechanism for gene editing [25]. By using various transcriptional activators and inhibitors, the CRISPR/dCas9 system can be used for gene screening and regulating the expression of target genes.

Guided RNA (gRNA) is a single chimeric RNA formed by fusing tracrRNA with crRNA. There are two critical regions of gRNA, the scaffold, and spacer regions, which orient dCas9 to its target. The spacer region contains nucleotides complementary to those found in the target gene (usually in the promoter region). The scaffold region mainly forms a complex with dCas9. Together, they bind dCas9 and direct it to the target gene. The CRISPR system shows excellent flexibility because the gRNA spacer region can be modified for any potential sequences, as any gene or nucleotide whose sequence is complementary to the spacer can be chosen for targeting.

The recruitment of effector proteins

Transcriptional effectors or synthetic transcriptional regulators, including transcriptional activators or inhibitors, are proteins or protein domains fused with dCas9 or sgRNAs, which help recruit essential cofactors and RNA polymerase to transcribe genes to be manipulated and expressed. Table 1 lists CRISPRa-based transcriptional regulation tools. The dCas9-based targeting of transactivators, such as the herpes simplex viral protein 64 (VP64), VP16, and p65, a transcription factor NF-κB, recruits RNA polymerase II for specific and robust upregulation of target genes [26–28]. Similarly, the combination of VP64, p65, and Rta (a gammaherpesvirus transactivator), termed VPR, can also induce target gene expression and is more effective than VP64 or p65 alone [16]. On the other hand, dCas9 fusion proteins designed to inhibit specific sites have also been developed. These inhibitory dCas9 constructs recruit histone-modifying enzymes, such as Kruppel-associated Box (KRAB), to concentrate the chromatin environment to reduce gene expression.

Table 1.

CRISPR-based tools for transcriptional activation

Transcriptional regulation	CRISPR fusion protein	Characteristic	Efficiency	Reference
Activation	dCas9-VP64	C-terminal VP64 domain	Low	[13, 27]
Activation	VP64 dCas9 VP64	N-terminal and C-terminal VP64 transactivation domains	Tenfold improvement compared to dCas9-VP64	[31]
Activation	dCas9-VPR	VP64-p65-Rta tripartite activator	More efficient than VP64	[16]
Activation	dCas9-TV	The combination of 6TAL and VP128	More efficient than VP64	[29]
Activation	dCas9-SunTag-VP64	dCas9 fused to tandem GCN4 peptide repeats, and a single-chain variable fragment (scFv) GCN4 antibody fused to VP64	More efficient than VP64	[32, 74–76]
Activation	dCas9-SAM	sgRNA containing two protein-binding aptamers (MS2) can recruit transcription activators p65 and HSF1 to targeted promoters	More efficient than VP64	[12]
Activation	dCas9-SPH	Replacing VP64 in SunTag with p65-HSF1	The most potent dCas9 activator among VP64, VPR, SAM and SunTag	[33]
Activation	dCas9-VPH	Fusion of dCas9-VP192 and p65-HSF1 activator domains	More potent than VP64	[77]
Activation	dCas9-CAM	Replacing VP64 in SAM with VP64-p65-Rta	More potent than VPR	[36]

dCas9-based transcriptional activation provides a promising alternative gene activation approach by targeting specific genomic sites, usually endogenous gene promoters, using programmable DNA-binding modules [29]. The dCas9 transcriptional activation domain has unparalleled simplicity and repeatability compared with zinc-finger protein-transcriptional activation domain and transcriptional activator-like effector (TALE) because the synthetic guide RNA can be easily modified to obtain new targeting specificity, and dCas9 guided by multiple sgRNAs can bind to numerous target sites at the same time [30].

CRISPRa-mediated neuronal transdifferentiation

Transdifferentiation from one cell type to another has been achieved through ectopic expression of lineage-specific transcription factors. In recent years, alternative strategies have been utilized for transdifferentiation (Fig. 1). One strategy employs dCas9-mediated endogenous activation, not requiring the overexpression of exogenous TFs by traditional transgenic technology (Table 2). In 2016, Black and his colleagues demonstrated that activation of the endogenous Brn2, Ascl1, and Myt1l genes using engineered CRISPR/Cas9-based transcriptional activators could directly induce rapid reprogramming of primary mouse embryonic fibroblasts (MEFs) into induced neuronal cells (iNs) at a conversion efficiency of about 4% [31]. They used a transactivator with both N-terminal and C-terminal VP64 transactivation domains (^VP64dCas9^VP64) that generated a remarkable improvement in activation of target genes compared to the first-generation dCas9 transactivator with only one C-terminal VP64 domain. These iNs exhibit complex neuronal morphology, express neuronal markers, form functional synapses, and show electrophysiological activity. In addition, the removal of sgRNAs targeting the Brn2 locus reduced the efficiency by approximately fivefold compared with targeted activation of all three endogenous factors. Removing Myt1l sgRNAs slightly reduced the efficiency. Removal of Brn2 sgRNAs reduced the percentage of Syn⁺ cells (representing mature neurons) over twofold, but no significant change was detected after removing Myt1l sgRNAs. This study provides a direct neuronal conversion strategy by CRISPRa-mediated activating specific gene expression. Liu et al. demonstrated that activating only one endogenous factor, Ngn1, with the CRISPRa system could convert MEFs into Map2⁺ /Tuj1⁺ neurons 14 days after infection, but the conversion efficiency was extremely low (~ 1%) [32]. However, the combination of Brn2, Ezh2, and Foxo1 (Ngn1 + Brn2, Ngn1 + Ezh2, Ngn1 + Foxo1, with comparable efficiency) significantly enhanced the efficiency of Ngn1 for direct neuronal reprogramming up to 83%. Interestingly, the combination of Ezh2 + Brn2 or Ezh2 + Mecom also efficiently induced MEFs into neurons [32]. Meanwhile, the converted neurons by EZH2 + Brn2 combination highly expressed hindbrain markers Gbx2, Hoxa2, and Hoxa4. Moreover, these iNs induced by the Ezh2 + Ngn1, Ezh2 + Mecom, and Ngn1 + Brn2 combination could form functional synapses [32]. Zhou et al. established a platform known as SunTag-p65-HSF1 (SPH) to improve the efficiency of dCas9-fused activators by replacing VP64 in SunTag with p65-HSF1, which is used in the synergistic activation mediator (SAM). They showed that activating three endogenous neurogenic transcription factors, Ascl1, Neurog2, and Neurod1 (ANN), using this system directly converted mouse astrocytes into functional neurons in the midbrain with a yield of up to 35% being NeuN positive [33]. These studies pave the way for direct cellular reprogramming in vivo, providing a potential treatment for many neurodegenerative diseases [34]. Our group used the CRISPRa system to activate endogenous Ngn2 and Isl1 to convert primary spinal cord astrocytes and MEFs into functional motor neurons (MNs). A conversion efficiency of over 80% was achieved. Moreover, astrocytes can be converted into mature MNs in the adult spinal cord by CRISPRa-mediated activation of Ngn2 and Isl1 within 42 days at a conversion efficiency of 48%. More importantly, the converted MNs in the spinal cord could extend axonal projections into the sciatic nerve [35]. Recently, Russo et al. reported that CRISPRa-mediated induction of neuron-enriched mitochondrial proteins Prdx2 and Sod1 could improve the efficiency of Ascl1-induced neuronal reprogramming (Ascl1-Prdx2-Sod1). The presence of Prdx2 and Sod1 accelerated the conversion process of astrocytes into neurons and prolonged the survival time of these iNs [36]. Their study indicated that mitochondrial proteins might play an enabling and driving role in neuronal reprogramming.

Fig. 1.

Direct reprogramming of astrocytes or fibroblasts into iNs. Traditional approaches vs. CRISPRa reprogram astrocytes or fibroblasts in vitro. TFs transcription factors, iNs induced neurons

Table 2.

CRISPR-based transdifferentiation or reprogramming

Starting cell	CRISPR fusion protein	Transcription factors	Ending cell	Efficiency	References
MEFs, spinal cord astrocytes	dCas9-VP64	Ngn2 + Isl1	Motor neuron	High	[35]
MEFs	VP64 dCas9 VP64	Brn2 + Ascl1 + Myt1l	Neuron	Low	[31]
mouse astrocytes	dCas9-SPH	Ascl1 + Neurog2 + Neurod1	Neuron	Medium	[33]
MEFs	dCas9-SunTag	Ezh2 + Ngn1 Ezh2 + Mecom Ngn1 + Brn2	GABAergic neuron	High High High	[32]
MEFs	dCas9-SunTag	Oct4 or Sox2	iPSCs	High	[76]
Brain astrocytes	Cas13d	Ptbp1 knockdown	Dopamin neuron	High	[78]
Cortical astrocytes	dCas9-CAM	Ascl1 + Prdx2 + Sod1	Neuron	High	[36]
human fibroblasts	dCas9-SAM	Nr5a1 + Gata4 +	Leydig cells	?	[79]

MEFs mouse embryonic fibroblasts

In vivo application of CRISPRa in the nervous system

The use of CRISPRa in the nervous system has been widely studied recently. Zhou et al. used CRISPRa to activate multiple genes simultaneously. Adeno-associated virus (AAV) was used to deliver sgRNAs in the CRISPR–dCas9-activator transgenic mice. Astrocytes were induced into functional neurons in the mouse midbrain by activating Ascl1, Neurog2, and Neurod1 [33]. Several other studies further demonstrated the successful application of CRISPRa in animal models. One study showed that CRISPRa-mediated upregulation of the potassium channel gene Kcna1 in mouse hippocampal excitatory neurons reduced neuronal excitability, reducing spontaneous generalized tonic–clonic seizures and rescuing cognitive impairment associated with chronic epilepsy in a temporal lobe epilepsy model [37]. Another study reported that upregulation of the voltage-gated sodium channel gene Scn1a expression in the inhibitory neurons could significantly reduce seizures and improve behavioral phenotypes of Dravet syndrome model mice [38]. CRISPRa method was also used in animal models of drug addiction [39].

The delivery of CRISPR components into target cells is limited by the specific methods used. Traditionally, the components are delivered through viral vectors such as lentivirus or AAV, but other non-viral methods have also been evaluated. One of the main limitations of virus delivery is packaging size, which makes it challenging to package large gene sequences or even multiple sequences in one AAV vector. Strategies such as splitting dCas9 activators may solve some of these problems [40]. Böhm et al. used the split dCas9 approach for gene therapy for inherited blindness. The long-term efficient expression of cone photoreceptor-specific M-opsin was achieved using dual adeno-associated viral vectors expressing split dCas9-VPR, which improves retinal function and attenuates retinal degeneration in a mouse model of rhodopsin deficient retinitis pigmentosa [41]. The study also found that enhancing the AAV delivery efficiency could lead to a more remarkable improvement of retinal function.

Comparison between CRISPRa system and ORF overexpression in neuronal reprogramming

Transdifferentiation has been widely achieved by overexpression of lineage-specific TFs. Forced expression of specific TFs has been shown to produce specific types of neurons, including motor neurons [42], serotonergic neurons [43], dopaminergic neurons [44], glutamatergic and GABAergic neurons [45, 46]. These studies used ORF overexpression of exogenous copies of TFs. Still, with CRISPRa system, it is possible to activate endogenous genes, which may have more advantages than ORF overexpression considering the clinical safety issues. Moreover, CRISPRa system is related to epigenetic modification around endogenous loci, which is assumed to simulate natural epigenetic markers better [18]. However, it has been reported that Cas9 could induce host response in mice [40]. Generally speaking, direct ORF overexpression is preferred in vitro if the goal is to achieve the strong expression of target genes. For large-scale multiplexed screens, CRISPRa system may be more feasible as the design and synthesis of sgRNA could be more accessible. Considering the immune response, using ORF for in vivo application may have advantages over CRISPRa system as ORF from the same species is usually used [18]. The expression level of ORF (which can induce supraphysiological expression higher than CRISPRa), the species (human cells seem to be more challenging to reprogram than mouse cells), and the induced cell type should be considered when designing the reprogramming strategy. To activate target genes, CRISPRa system needs both CRISPRa protein and sgRNA, while the ORF strategy requires multiple ORF delivery. The main obstacles surrounding the application of CRISPRa in vivo include the delivery of components, the strength of activators, and CRISPR protein’s immunogenicity [18].

Current challenges and strategies

Glial cells constitute at least half of the cells in the CNS [47]. To reprogram glial cells into neurons could realize in situ regeneration of neurons. The regenerated cells supplement the lost non-renewable neurons to make the cell replacement therapy of neurodegenerative diseases feasible. Although multiple research groups have reported successful transdifferentiation of neurons from glial cells, the linage transition remains controversial. Matsuda et al. achieved mouse microglia-to-neuron conversion both in vitro and in vivo by overexpressing NeuroD1 [48]. However, a recent report challenged their results, suggesting that "NeuroD1 induced microglia-to-neuron reprogramming" comes from the virus nonspecific leakage. On the contrary, overexpression of NeuroD1 caused the massive death of microglia [49]. Wang et al. also claimed that NeuroD1 induced neurons are endogenous neurons, against the previous observation that overexpression of NeuroD1 could convert astrocytes into neurons [50]. These contradictions strongly demand a more rigorous experiment design, and the results need to be interpreted with caution.

Although iNs could be converted from somatic cells by CRISPRa-mediated activation in vitro and in vivo, challenges such as the efficiency of transgene expression and possible off-target effects could hinder its application. Therefore, this technology needs optimizations in three aspects: selecting a suitable vector, increasing transcriptional activation, and minimizing off-target effects.

Vector

At present, the main limitation of using CRISPRa in direct neuronal reprogramming is the efficiency of transgene expression [51]. Virus transduction is one of the most efficient methods to transfer CRISPR constructs to somatic cells [37, 38, 41, 52]. Depending on the application, various viruses can be chosen for gene delivery. Adeno-associated virus (AAV) vector can transfer genes to target cells safely and effectively. It has become the leading gene delivery platform for clinical gene therapy and vaccination [53–55]. Nevertheless, there are two major problems in AAV-mediated in vivo gene therapy. The first is the immune response; it has been reported that host immune response during rAAV administration limits long-term transgene expression in humans [56]. Several strategies have been developed to overcome this barrier, such as plasmapheresis and engineering rAAV capsid to prevent the binding of neutralizing antibodies with the virus. More details about current advances to minimize the immune response to AAV can be found in references [57, 58]. AAV-mediated CRISPRa system has been applied to gene therapy of animal model of inherited blindness, and no adverse immune responses were observed [41]. Another is the genome capacity of the virus vector. The coding region of SpCas9 alone is about 4.2 kb [21]. However, the capacity of AAV, commonly used in the nervous system and gene therapy trials in humans, is only about 4.7 kb. Compared with AAV, lentivirus has a larger transgenic capacity (~ 8 to 10 kb) [59], but is less diffusible in the CNS. Multi-point injection may be required if targeting in vivo. It is also challenging to achieve high virus titers with the large particle of the lentivirus.

To overcome the limitation imposed by the packaging capacity of the AAV vector, other species of Cas9, such as Staphylococcus aureus (SaCas9), which is almost 1 kb shorter than the commonly used SpCas9, has been proposed to reduce the package size [60]. However, the shortcoming of using SaCas9 is that the PAM sequence (5ʹ-NNGRRT-3ʹ) is more complex, restricting the range of accessible targets. Moreover, the CRISPR component can be decomposed into multiple vectors, and co-transduction can be performed. In addition, separation of dCas9 from its effector protein, such as SunTag [61], LACE [14], and the use of RNA aptamer [62, 63] can reduce the load of the virus vector. The disadvantage of this method is that cells must be co-transduced to receive all CRISPR components.

Strategies for increasing CRISPRa transcriptional activation

There are two main strategies employed to promote transcriptional activation by CRISPRa. The first strategy is to fuse various activators in tandem with dCas9. Taking VP64 as an example, when dCas9-VP64 is directed to the promoter sequence of the target gene through sgRNA, the complex can usually recruit and regulate the transcription of the target gene (Fig. 2). With extra activation domains, dCas9 VPR and dCas9 TV show higher transcription activation than dCas9-VP64. dCas9 TV is named for combining 6TAL (6 copies of the TALE transcription activation domain) and VP128 transcription activation domain [29]. Other transcriptional effectors, such as SAM and SunTag activators, together with VPR, are the most effective constructs by far. In human cell lines, SAM usually has up to a fivefold increase over other activators [30]. The second strategy uses modified sgRNA scaffolds to recruit more effectors [12, 64]. SAM sgRNA was generated by inserting two phage MS2 RNA hairpins into the nonessential region of sgRNA. Each MS2 hairpin binds two MCP molecules (MS2 coat protein), and each molecule is fused with a pair of transcription activator p65 and human heat shock factor HSF1 (Fig. 2). Guiding dCas9-VP64 fusion protein through this chimeric sgRNA leads to more efficient activation (2–10⁴ times) [30].

Fig. 2.

Schematic of six dCas9 activation systems. SunTag, dCas9 fused to tandem GCN4 peptide repeats, and a single-chain variable fragment (scFv) GCN4 antibody fused to VP64. VPR, a tandem fusion of VP64, p65, and Rta to dCas9. SAM (synergistic activation mediator), sgRNA containing two protein-binding aptamers (MS2) recruiting transcription activators p65 and HSF1 to targeted promoters. SPH, replacing VP64 in SunTag with p65-HSF1. CAM, replacing VP64 in SAM with VP64-p65-Rta. TSS transcriptional start site

Minimizing off-target effects

The main criticism of any gene-editing system is the off-target effect, including similar gene loci and overall cellular response. Measures are being taken to accurately identify and minimize any off-target binding. Although the CRISPRa system eliminate the risk of severe off-target DNA cleavage, gene expression may still be subject to off-target interference. The CRISPR system is guided by RNA, and the sgRNA binds to DNA in a sequence-specific manner. Therefore, a mismatch could happen in the target recognition, and unnecessary mutations may occur due to the similarity between sgRNA and unintended sites [65]. Predicting and avoiding off-target mutations are necessary for functional genomic analysis. Studies suggest that to minimize off-target effects, highly specific sgRNA must be designed to increase on-target efficacy, and the use of sgRNA containing homology with other genomic sites should be avoided [66–68]. To validate the specificity of sgRNA, a straightforward approach is to use dCas9 antibody (or fusion tag antibody) for chromatin immunoprecipitation (ChIP) sequencing to ascertain the genomic location of CRISPR complex binding [69, 70]. RNA-seq has also been used to characterize whether the dCas9 fusion protein is recruited to undesirable sites. One study analyzed the off-target effects of dCas9-p300 and dCas9-VP64 by RNA-seq and found no significant increase in nonspecific gene expression outside the sgRNA target of dCas9-VP64. Still, only two off-target genes increased significantly in dCas9-p300-transfected cells [71]. In practice, the GC content of the target sequence higher than 70% might increase the off-target effect [72]. However, using sgRNAs with GC content greater than 50% could effectively increase on-target mutagenesis due to high binding affinity to target sites [66].

Another strategy to reduce off-target effects is to lower sgRNA/Cas9 complex concentration. Minimizing the risk of off-target Cas9 can be achieved by titrating Cas9 [73] or directly introducing recombinant RNA/protein complexes. It is vital to limit the time of Cas9 expression in cells because once gene editing is completed, the whole system is no longer needed. In general, careful nuclease expression and engineering specific design may provide safe and effective gene editing for CRISPR even with minimal off-target effects.

Conclusion and perspectives

Direct reprogramming of somatic cells to neurons is an attractive alternative for cell replacement therapy in the CNS. Using lineage-specific transcription factors, multiple types of somatic cells have been converted into multiple subtypes of the neuron. CRISPR/dCas9 activation system has been proved to be an efficient and robust transcriptional activation tool. This paper has reviewed direct neuronal reprogramming using the CRISPR/dCas9 activation system in recent years. The key technical parameters of this strategy include gene expression level, sgRNA specificity, stable cell reprogramming, and component delivery. To take full advantage of the CRISPRa in cell reprogramming, careful analysis of the characteristics of target sites, the design of sgRNA, and the properties of the AAV vector are necessary. With the help of high-throughput cell phenotype analysis methods, such as single-cell technology, we are optimistic that the CRISPRa system will lead to the discovery of more neuronal reprogramming protocols. With the advance of this technology, CRISPRa systems will emerge as more powerful tools in exploring the fundamental mechanisms in neural cells fate and treating neurodegeneration diseases.

Author contributions

The manuscript was drafted by MZ. YC, MS, XS, FW, and BZ edited the manuscript. All authors have read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China 81471283 (B. Z.) and the National Natural Science Foundation of China 82072795 (F. W.).

Data availability

Not applicable.

Declarations

Conflict of interest

The authors declare no competing interests.

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Consent for publication was obtained from all participants.