Abstract
Background:
Genome editing mediated by clustered regularly interspaced short palindromic repeat -CRISPR-Cas- was first reported using cell lines in 2012, and since then, its applications in other systems have steadily increased. However, this approach still needs improvement and optimization. Prime Editing (PE) was originally proposed by Anzalone and colleagues in 2019, who showed significant lower rates of "off target" effects compared to conventional CRISPR-Cas9.
Objectives:
In this review, we describe and compare the "conventional" CRISPR-Cas9 system with a novel homology-directed repair-independent editing approach called Prime Editing (PE).
Methods:
Numerous recent publications involving the PE molecular mechanism for targeted-mutagenesis, advantages, limitations, and applications of the approach in different biomedicine fields were reviewed.
Results:
From the literature, it appears that PE neither relies on DSB nor needs HDR with exogenous donor DNA templates, which shows the clinical potential of PEs in correcting a broad range of mutations for genetic diseases. Four generations of PE mechanisms have been developed, reaching higher efficiency from one generation to the next one. However, reviewed studies support that PE also introduces new challenges such as unwanted mutations mediated by the double nicking strategy required for the PE3's efficiency, the selection of the optimal combination of PBS and RT templates, and the limitation of large DNA insertions that conventional CRISPR-Cas9 is capable of.
Conclusion:
The reviewed literature demonstrate that genome editing-based PE is a promising technology aiming at decreasing the common undesirable effects associated with conventional genome editing approaches such as CRISPR-Cas9 or BE. Even though progress has been made to improve the efficiency of the genome editing by the PE ribonucleo-protein complex, further research is required to optimize PE tools and maximize its efficiency.
Introduction
Prime Editing (PE) is a site-specific mutagenesis approach developed 2 years ago for highly precise and efficient genome editing. This novel approach evolved from the base editing (BE) and clustered regularly interspaced short palindromic repeats (CRISPRs)-Cas9 technique opening new perspectives for the study of gene function. PE is also known as homology-directed repair (HDR)-independent CRISPR-Cas9 in contrast to HDR-dependent CRISPR-Cas9. For a better understanding of the topic, the terms PE and conventional CRISPR-Cas9 will be used throughout this review.
Among several advantages, critically, PE allows site-specific insertion of 12 single-base mutations on target sequences as well as deletions. These targeted mutations do not require double-stranded breaks (DSBs) or HDR donor molecules, given Cas9-nickase processes the strand to be edited. The absence of DSB highly reduces the chances of unexpected mutations and/or off-target mutant sites. PE directly introduces exogenous DNA molecules sequences into the target site using a primer editing gRNA (pegRNA) that includes the sequence to be inserted combined with a catalytically impaired Cas9-nickase and an engineered reverse transcriptase (RT).1
Jennifer Doudna and Emmanuelle Charpentier, 2020 Nobel Prize in Chemistry awardees, with Francisco Juan Martínez Mojica, had been studying the system for many years in bacteria. In 2011, Doudna, Charpentier, and their colleagues began experimenting with Streptococcus pyogenes and discovered a noncoding RNA molecule called trans-activating CRISPR RNA (tracrRNA). The tracrRNA is a component of an adaptive immune system in most archaea and bacteria known as CRISPR-Cas9, which neutralizes bacteriophages by specifically cleaving their DNA.2 This immune mechanism employed by prokaryotes has been cleverly adapted as a genome editing tool for eukaryotic cells. It allows targeted mutagenesis and site-specific insertion of transgenes. However, it has been shown that conventional CRISPR-Cas9 may induce an expected high number of off-targeted DSB and mutations that are then repaired by the HDR mechanism. The latter mechanism works in the presence of an exogenous DNA template flanked by homology arms, which leads to precise sequence replacement3. However, HDR is less efficient in nondividing cells and inevitably accompanied by unwanted indels; in addition, DSB themselves can produce highly heterogeneous mixtures of off-target mutations, such as indels, translocations, and duplications.4 The high frequency of off-target mutations (≥50%) introduced by conventional CRISPR-Cas9 is one of the major concerns of this technology, in particular when applied to gene therapy.5
PE does not require any of those two mechanisms to edit the DNA, increasing the efficiency considerably.6 One of the key advantages of PE is the higher specificity of edition compared to the conventional Cas9 (Table1).
The main differences between both gene editing techniques, CRISPR-Cas9 and Prime Editing (PE)
The need to improve treatments of disorders associated with gene mutations and genomic rearrangements has boosted the search for novel and more efficient genome editing technologies. PE represents a promising approach for gene therapy. More than 75,000 disease-associated with DNA variants are currently described,7 and PE could correct up to 89% of genetic variants associated with human diseases.8
In addition to gene therapy, PE may have other biomedical applications on diverse fields, which are discussed in the “Applications” section of this Review.
Advantages of Prime Editing Over Conventional CRISPR-Cas9
PE involves a complex formed by an engineered Cas9 nickase with an RT, this is a recently developed Cas9 endonuclease with wider range of protospacer adjacent motif (PAM) sequences and higher specificity, that enables specific DNA regions to be replaced.9 Those are paired with a pegRNA that encodes the desired modification to be introduced into the target site (Fig.1A).10 On the other hand, conventional CRISPR-Cas9 guide RNAs (gRNA) tend to show high mismatch tolerance, increasing the likelihood of cleaving off-target sites.11 This may lead to the introduction of large deletions and complex genome rearrangements into edited cells.11 Moreover, other controversies have been raised, including the carcinogenic potential of conventional CRISPR components12 and the immunogenicity of Cas9 nucleases.11,13

Overview of Prime Editing mechanism. (A) Prime Editing Complex (PE Complex) consists of a reverse transcriptase (RT) domain, a Prime Editing guide RNA (pegRNA), and a Cas9 nickase domain, which enables precise DNA editions in the target sequence near the protospacer adjacent motif (PAM) region. (B) PE Complex binds target DNA and nicks the PAM strand. The 3’ end hybridizes to the primer-binding site (PBS) and primes reverse transcription of DNA with RT template with an edit. (C) Flap equilibration takes place between edited 3’ flap with edit and 5’ flap without edit, then there is 5’ flap cleavage and ligation and finally DNA repair of edited DNA.
Also, PE has been shown to introduce distant point mutations (>30 bp) to the site of nicking, offering more versatility than the nuclease-mediated HDR approach, which is unable to introduce mutations beyond 10 bp from the predicted cut site.14
Prime Editing might be the right tool to edit the genome
The higher specificity of PE compared with conventional CRISPR-Cas9 might be driven by the three consecutive hybridizations occurring in the former compared to the latter approach. During PE, hybridizations occur (1) between the target DNA and spacer in the pegRNA molecule, (2) between target DNA and primer binding site (PBS) in the pegRNA molecule, and (3) between the target DNA and edited DNA flap. A single hybridization step between the target DNA and protospacer from single guide RNA (sgRNA) takes place during the conventional CRISPR-Cas9 system (Fig.1A and B).15
In addition, PE has shown to be more effective to precisely introduce all 12 combinations for base-to-base conversions, small insertions, and deletions. In contrast, other editing systems such as conventional CRISPR-Cas9 can only induce sequence replacements by providing an exogenous DNA template6 or BE, which only allows for a subset of possible edits C/T, G/A, A/G, and T/C limited to preexisting windows within the genomic target sequence.3
Remarkably, the lack of DSB and HDR mechanisms in PE has expanded the versatility of genome editing, being able to correct almost 90% of all mutations known to be associated with human diseases.1 Moreover, PE avoids both off-target effects, normally caused by conventional CRISPR-Cas9 mechanism. Off-target effects occur when conventional CRISPR-Cas9 edits unintended sites, whereas on-target edits represent changes in site-specific localization but with unintended consequences. This would lead to changes not only in the target sequence but also the surrounding DNA, which may affect other genes.16
Finally, the PE efficiency has increased along with the designing of better and more precise prime editors. Anzalone etal.1 designed up to four types of Primer Editors (PE1, PE2, PE3, and PE3b); each one displays new features that incrementally improve the PE efficiency. Optimization of the prime editors is key to develop a fully functional and high-precision genome editing tool.1 , 6
Mechanism By Which PE Introduced Site-Specific Mutations
The PE Complex
The PE mechanism is triggered by a ribonucleoprotein complex called PE Complex constituted by an RT Domain, a pegRNA, and a Cas9 nickase Domain (Fig.1).1
The pegRNA guides the Cas9 endonuclease into the target sequence adjacent to the PAM region—a short DNA sequence located next to the targeted region that CRISPR cleaves and that the Cas9 system requires to bind and cut (Fig.1B).17 The Cas9 (H840A) nickase targets the DNA by using a pegRNA containing a spacer sequence that hybridizes to the target site. The Cas9 nicks the strand containing the target sequence creating single-stranded with complementary sequences to the 3’ and 5’ pegRNA’s strands.15
The pegRNA is complementary to the target DNA strand, which also works as a primer-binding site (PBS) region and contains the sequence that will be introduced to the targeted gene. The PBS region, complementary to the second DNA single-strand, acts as a primer for the RT. The RT is an RNA-dependent DNA polymerase that uses the sequence from the pegRNA to start the reverse transcription. The information is copied from the pegRNA into the target DNA sequence, therefore, altering the preselected target sequence in a customized manner (Fig.1B and C).15
The RT domain uses a nicked genomic DNA strand as a primer for the synthesis of an edited DNA flap templated by an extension on the pegRNA. Subsequent DNA repair incorporates the edited flap by incorporating in a site-specific manner the edited sequence (Fig.1B).10,18
Flap equilibration
The flap equilibration is the process that occurs after the RT has edited the target DNA sequence, and two single-stranded DNA flaps are left—the 5’ and 3’ flaps, unedited and edited DNA strands, respectively.1 Cellular DNA repair processes excise the 5’ flap, allowing the 3’ flap to be introduced into the target site generating a heteroduplex DNA, containing the edited and the nonedited strands. Finally, the nonedited strand is resynthesized using the edited strand as a template, resulting on a fully edited duplex.14 The probability to repair the desired strand is then 50%. One outcome is the hybridization of the 5’ flap to the unmodified strand, followed by the DNA repair process and introduction of the mutation in the second DNA strand. On the contrary, if the flap equilibration does not occur, the 5’ flap is excised, and the target sequence will remain unchanged (Fig.1B).19
Prime Editing Generations
Accelerated progress on PE techniques led to major improvements of its components and mechanisms. We discuss the different generations of PE strategies in the following section.
PE1
The first generation uses an engineered Moloney murine leukemia virus Reverse Transcriptase enzyme (M-MLV RT) fused to the C terminus of the Cas9 (H840A) Nickase, both complexed with a pegRNA. The pegRNA specifies both the genomic target and edit sequence. PE1 shows an efficiency of 0.7–5.5% introducing point transversions (Fig.2).3 In this complex, the PE chimeric protein (PEFP) nicks one of the strands and the RT domain generates complementary DNA (cDNA) by copying the pegRNA to reinstall a segment of the nicked DNA strand.20

Prime Editing generations | PE1 and PE2. Equilibration between the 3’ edited and 5’ unedited flap takes place. PE3: twelve transition and transversion mutations were generated; PE3b separates the two nicking steps and reduces the generation indels.
PE2
Compared to PE1, PE2 shows significantly higher editing efficiencies at tested genomic loci1 by using different variants of M-MLV RT, which affect thermostability, processivity, DNA–RNA substrate affinity, and RNase H activity.3
The first change was the improvement of the thermostability, which was achieved through the implementation of three mutations (D200N, L603W, and T330P) in the M-MLV RT to increase the RT activity at higher temperatures, to improve the editing efficiency up to 1.3-3.0-fold. In addition, two additional mutations (T306K and W313F) were introduced to enhance the binding of RT to the pegRNA complex. As a result, the PE2 was described as a pentamutant RT linked to the nickase [Cas9(H840A)-M-MLV RT (D200N/L603W/T330P/T306K/W313F)], which showed an improvement of efficiency of 1.6 to 5.1-fold in introducing point mutations when compared with PE1.15
PE3
Two factors limited the efficiency of the DNA edits driven by PE2, the uncertainty of whether it will be the edited 5’ or the unedited 5’ flap that is paired with the unmodified DNA strand (Fig.2).1 To improve the efficiency, the PE is cotransfected with a second sgRNA to nick the unedited DNA strand, taking advantage of the endogenous mismatch repair pathway to retain the information of the edited strand.6 The additional sgRNA is a standard gRNA that directs the Cas9 H840A nickase element to the PEFP to nick the opposite strand as the original nick. This approach was developed given the edition with one strand might be repressed due to a mismatch between both strands.20
Through this approach, all 12 base-to-base combinations mentioned earlier can be generated with a 33% efficiency, which is similar to the efficiency level of the existing cytidine and adenine BE. However, the number of off-target effects, such as undesired mutations, was lower compared with conventional CRISPR-Cas9.1
PE3b
To suppress the remaining nonhomologous end joining and unwanted indels induced by the PE3, the nicking sgRNA was redesigned in the PE3b version to recognize the edited DNA sequence and ensure that the second cut occurs after the resolution of the edited strand flap. Thus, the PE3b temporally separates the two nicking steps, reducing the generation of undesirable indels (Fig.2).6
PE Limitations
PE studies have shown some discrepancies in its efficiency. PE3 was used to introduce mutations in the Dnmt1, Chd2, and Tyr genes in the mouse genome invivo, showing that even though PE3’s efficiently introduces targeted mutations in mouse embryos, indels and unwanted mutations are introduced by the double nicking, hampering its utility for gene therapy in its current form.21
A proper pegRNA design is vital to ensure a high efficiency. The combination of PBS and RT template must be selected to support an optimal PE efficiency. Generally, efficient PBSs are in a range between 8 and 15 nt, while RT templates are between 10 and 20 nt in length; although this combinatorial matrix of possible PBS and RT template is currently determined empirically, there are some factors to select the optimal pegRNA design, such as GC content, primary sequence motifs, and secondary structures within pegRNA 3’ extensions.14
Also, PE does not allow to introduce large insertions or deletions, due that such length must be proportional to the pegRNA, which the length must be within the mentioned range; otherwise, it could be degraded or eliminated.22
Though PE is confronting an ever-increasing interest, there remain challenges that require further research. One challenge is off-target editing; PE shows an average ≤0.6% off-target changes compared to an average 32% off-target changes by conventional CRISPR-Cas9 at the same loci. However, further investigations are required to better understand the PE off-target mechanism. Another challenge is the sequence specification and editing window; PE requires further in-depth study of edit range.23
Applications
PE technologies are being explored in many fields of biological science looking to expand the scope and capabilities of the existent CRISPR-Cas9-based tools. Some of the areas PE have been (or plan to be) tested are gene therapy, genetic manipulation of parasites, crop improvement and metabolic engineering.23
Genome edition for therapeutics
PE has been used to prove the correction of several pathogenic mutations by enabling precise edits at individual nucleotides, such as Duchenne muscular dystrophy,24 Sickle Cell disease, Tay-Sachs disease, and prion disease in humans.20 PE presents the potential for safer editing, thus eventually will become a therapeutically important approach to human gene therapy. However, obstacles remain. For instance, for clinical applications, the PE complex and pegRNA molecule have to be delivered together into the target cell; however, the size of the constructs that express these molecules is too large as a cargo for adeno-associated viral vectors.25 “pegFinder” is an online tool that assists with the design of the pegRNA molecule (http://pegfinder.sidichenlab.org); this is a promising approach for on-target and off-target scoring predictions.26
Prime editing in cancer research
All the cells have the necessary mechanisms to repair mutations
Although cancer is considered as a multifactorial malady, the genetic component is one of the most relevant. As part of the understanding of cancer, it was published the life history and evolution of mutational processes in the 38 most common types of cancer showing driver mutations. The main conclusions of the study were that in early stages of cancer, mutations are in a constrained set of driver genes although specific copy number gains. In the late stages, the genetic instability increases by four times as the number of driving genes also increases. This will ultimately lead to new mutations.29
One of the areas of interest in cancer research is precisely finding the ability to reverse malignant mutations. Conventional CRISPR-Cas9 provided an extraordinary opportunity to achieve this goal; but the limitations mentioned above, as off-target effects, have hampered the expectation.
PE has been hard applied in organoid production for both mutation induction and mutation corrections. Considering the differences in safety and efficiency of PE among the different organoid types it is efficient and specific to introduce desired mutations with undetectable off-target effects in hepatocytes and colon tumor organoids.30,31
Application of PE in human-induced pluripotent stem cells (hiPSC) and HEK293t cells showed differences in efficiency using the same pegRNAs, generally being better in hiPSC.32 PE has also been tested in human adult stem cells and cancer cell lines, demonstrating that PE can generate insertions, deletions, and various point mutations functionally reverting disease phenotype.33
An example of the high impact that PE can have on cancer research is the ability to repair one or more single mutation in a single gen such as TP53, which is the most frequently mutated gene with more than 5,000 described mutations including 75% of missense or nonsense mutations.32,34 The possibility of repairing DNA mutations with higher efficiency by PE brings us closer to the aim at repairing, errors in the DNA without generate extra errors.
Genome edition for parasitic worms
Approximately one-fifth of humans are infected with at least one species of parasitic worms (helminths). Helminths comprise platyhelminths (“flatworms”) and nematodes (“roundworms’), and the infections they cause are characterized by long term, chronic illness that impairs mental and physical developments, thus impeding socioeconomic development and driving a vicious cycle of poverty in endemic areas.35 Critically, the reliance on only a few drugs combined with their mass-administration is likely to drive drug resistance.36 Therefore, novel targets to develop control strategies are desperately needed. Large-scale genomic and transcriptomic datasets and their functional analysis to translate computational outcomes into biological insights become vital to achieve this objective.
Approaches to drive targeted-mutations and site-specific integration of transgenes will boost the development of transgenic lines of parasitic worms and will ultimately lead to high-throughput screening platforms for gene functional analysis. CRISPR-Cas9-based genome editing was demonstrated for few species of both nematode and platyhelminth phyla; however, the efficiency of this platform is still extremely low, making the approach unsuitable for high-throughput functional analysis to discover new drug or vaccine targets.37,38 The significantly lower off-target effects and higher efficiency for targeted-mutagenesis place PE as a promising alternative to explore in these parasites.
Conclusions
Genome editing-based PE is a promising technology aiming at decreasing the common undesirable effects associated with conventional genome editing approaches such as CRISPR-Cas9 or BE. PE neither relies on DSB nor needs HDR with exogenous donor DNA templates.
Currently, progress has been achieved in improving the efficiency of genome editing by the PE ribonucleoprotein complex. Four generations of PE mechanisms have been developed, reaching higher efficiency from one generation to the next one. Recent data support the clinical potential of PEs in correcting a broad range of mutations for genetic diseases, suggesting that screening of pegRNAs in vitro is essential before conducting invivo studies.
However, PE also introduces new challenges such as unwanted mutations mediated by the double nicking strategy required for the PE3’s efficiency, the selection of the optimal combination of PBS and RT templates, and the limitation of large DNA insertions that conventional CRISPR-Cas9 is capable of. Therefore, further research is required to optimize PE tools and maximize its efficiency.
