Genetic engineering allows precise manipulation of the genome. The available systems make it possible to perform site-specific modification of the genome, which can be used to analyze the functions of genes, create cellular and animal models of diseases, and develop new methods of treatment. Zinc finger nuclease (ZFN) ^[1][2][1,2] and transcriptional activator-like effector nuclease (TALEN) [3] technologies employ sequence-specific DNA-binding modules to induce DNA damage and increase gene targeting efficiency, whereas a method based on clustered regularly interspaced palindromic repeats (CRISPR) with CRISPR-associated protein 9 (Cas9) [4] utilizes RNA guides.

As a result of the action of ZFNs, TALENs, or Cas9, a DNA double-stranded break (DSB) occurs, which can be repaired by one of three major mechanisms: microhomology-mediated end joining (MMEJ), non-homologous end joining (NHEJ) or homology-directed repair (HDR) [5]. The NHEJ pathway has a significant drawback: in the process of ligation of DNA ends, insertions or deletions (indels) can be introduced into the DNA sequence in the DSB region [6]. The HDR pathway requires a donor molecule (usually from a sister chromatid) to recombine to restore the correct DNA sequence [7]. Such precise modifications are desirable for targeted genome engineering. It should be noted that the term ‘HDR’ is often used to refer to both homology-directed repair (HDR), which is activated when a single-stranded oligodeoxyribonucleotide (ssODN) is used as a template and is regulated by the BRCA1–RAD52 axis, and homologous recombination (HR), which is activated when a double-stranded donor is used as a template and is regulated by the BRCA2–RAD51 axis. ReIn thisearchers article, we use the term ‘HDR’ for both cases. MMEJ is a variant of the alternative NHEJ and is based on the occurrence of microhomology of sequences ranging in length from 5 to 25 base pairs. This DSB repair pathway is classified as highly error-prone [5].

The cell cycle consists of several phases: the synthetic (S-phase), mitotic (M-phase), and growth (G1 and G2) phases. The transition between these phases is regulated by specific factors, the main factors of which are cyclin-dependent kinases (CDKs). Their activity is influenced by various external (inhibitors) and internal (for example, cell size or DNA damage) factors [8]. The key points of cell cycle advancement and arrest and the possibility of influencing these processes are discussed below.

In somatic mammalian cells, G1 is the longest phase of the cell cycle because, during this phase, all regular activities of the cell and its organelles take place. Among them, there is a global increase in histone acetylation and transcriptional activity ^[9][31], potentially exposing large regions of the genome to unwarranted programmable nuclease-induced NHEJ during G1. The anaphase-promoting complex (APC) with the activator protein Cdh1 (APC-Cdh1) forms the E3 ubiquitin ligase complex, which is active in the late M and G1 phases of the cell cycle, timely triggering ubiquitination and ensuing proteasomal degradation of the target cell cycle proteins, including geminin ^[10][11][32,33]. Gutschner et al. (2016) ^[12][34] proposed an elegant solution for cell cycle-dependent expression of Cas9—fusion of Cas9 with geminin—resulting in lower expression of Cas9 in G1 and higher expression in S/G2/M phases because of the activity of APC-Cdh1. Cas9-geminin fusion resulted in a 1.9-fold increase in knock-in at the MYH7 locus in porcine fibroblast cultures ^[13][35] and a growth in the HDR efficiency at the MALAT1 locus from 9.7% to 13.8% in HEK293T cells. The combination of this approach with nocodazole treatment in the case of the MALAT1 locus led to an increase in the HDR rate to 16.2% ^[14][36]. It was shown that the NHEJ/HDR ratio significantly decreased independently of the chromatin structure when using geminin ^[15][37].

RThesrache authors of the cited studies did not note a cell apoptosis increase or geminin toxicity; moreover, Yang et al. noted that the Cas9-geminin fusion shortened the lifespan of Cas9 in the cell and thereby reduced its toxicity to mice neurons in vivo ^[16][38]. Many studies using geminin have shown that a fragment of the protein, 110 amino acids (aa) for human geminin ^[12][15][34,37] or even 40 aa for mouse geminin ^[16][38], is sufficient to control Cas9 expression in the S/G2/M phases.

A similar approach to the regulation of Cas9 expression in a cell cycle-dependent manner was proposed by Matsumoto et al. ^[17][39]. The researchers used AcrIIA4, a natural inhibitor of Cas9, which was fused with the human chromatin licensing and DNA replication factor 1 (hCdt1). The AcrIIA4-hCdt1 complex inhibits Cas9 in the G1 phase; in the S/G2 phases, the complex undergoes proteolysis by the SCF-Skp2 complex and releases Cas9 activity. Indeed, this approach helped increase the HDR efficiency of CRISPR-Cas9 by 1.7–4.5 times, depending on the target locus.

Despite the fairly good results obtained with cell cycle-dependent degradation of molecules (geminin or Cdt1) in genome editing experiments, there is still little scientific research in this area, which is inexplicable.

Cyclin-dependent kinases (CDKs) are heterodimeric serine/threonine protein kinases that regulate cell cycle progression. Among them, the Cdk1-cyclin B complex controls the cell transition from the G2 to the M phase, while the Cdk2-cyclin E and Cdk2-cyclin A complexes regulate the G1/S and S/G2 transitions. In addition to CDKs, many CDK inhibitors (CDKIs), including members of the CIP/KIP family (p21, p27, and p57), are also involved in the regulation of the cell cycle. p21 interacts with a number of transcription factors, and the overexpression of p21 induces cell cycle arrest in various phases of the cell cycle ^[18][40].

Several cyclin-dependent kinases can be inhibited by various molecules to achieve cell cycle synchronization in the G1/S or G2/M phases, for example, by indirubins ^[19][41]. In GFP knock-in experiments using CRISPR-Cas9, treatment of porcine fetal fibroblasts with indirubin-3′-monoxime (4 µg/mL), an inhibitor of cyclin-dependent kinase 1 (CDK1), increased HDR by 1.9 times (up to 19.7%) ^[20][24]. Similar results were obtained for HeLa, HT-1080, and U-2 OS cells: an increase in the HDR rate by 2–5 times using transfection with expression vectors, coding meganuclease I-SceI, or ZFNs. In addition, in mesenchymal stem cells, indirubin-3′-monoxime also led to a 10-fold increase in HDR [21]. However, indirubins were found to increase apoptosis at almost all tested concentrations ^[22][42].

CDC7, a factor necessary to enter the S phase, can be indirectly included in this section. The CDC7 protein is involved in DNA replication; therefore, its inhibition leads to cell cycle arrest at the G1/S point ^[23][24][43,44]. The inhibition of CDC7 by XL413 led to a 2.1-fold increase in HDR in the K562 cell line, while its inhibition by siRNA increased HDR by 1.4 times ^[25][45]. Similar results were obtained in iPSCs: XL413 increased the HDR rate by 1.7 times (33.7% vs. 19.4%) during the integration of a BFP gene fragment into the AAVS1-EGFP locus. The combination of XL413 with the NHEJ inhibitors NU7441 and SCR7 further increased HDR by 2.7 times (45.7% vs. 17%) in iPSCs ^[24][44]. Importantly, XL413 must be added 24 h after CRISPR-Cas9 injection in order to achieve cell synchronization in the S phase ^[25][45].

As can be seen from the results of the studies presented above, both strategies (synchronizing cells in the G1/S or G2/M phases) ultimately lead to an increase in the efficiency of HDR/HR in genome editing experiments. This is probably due to the fact that the DSB repair pathway regulated by BRCA1/BRCA2, i.e., HDR/HR, is active in the S and G2 phases ^[26][12].

As stated above, CDK1 inhibition increases HDR; however, the opposite approach also enhances it. It has been shown that CDK1 promotes efficient end resection by phosphorylating the DSB resection nuclease, so its activation also contributes to an increase in HDR efficiency. CDK1 activation by CRISPRa increased HDR in HEK293, HEK293T, and HeLa cell cultures by 2.0–4.4 times (up to 7.58%), depending on the culture, locus of integration, and transgene. This activation had a synergistic effect (up to 15.3%) with the inhibition of KU80, a key factor in NHEJ, by CRISPRi ^[27][46].

Nonetheless, for unknown reasons, CDK inhibitors are rarely used to increase the efficiency of HDR in genome editing experiments despite the fact that several dozens of them have already been described, and many of them are commercially available ^[28][47].

Nuclease-mediated DNA double-stranded breaks by themselves can also cause cell cycle arrest. If DNA double-stranded breaks occur, then ATM kinase is activated, which in turn activates Chk2 (Figure 2). Then, the ATM-Chk2 complex phosphorylates p53, which promotes p21 expression. The latter binds and inhibits cyclin and cyclin-dependent kinase complexes, leading to cell cycle arrest in the G1/S phase as well as in the M/G1 and G2/M phases ^[29][48]. Genome editing methods involve the production of DSBs and thereby can activate the ATM–Chk2 pathway, which may result in cell cycle arrest ^[30][49]. Even single nuclease-induced DSBs in hematopoietic stem and progenitor cells can activate the p53 pathway, although this phenomenon is reversible ^[31][50]. G2/M arrest is indicative of DNA damage, likely caused by the combined on- and off-target activities of the nucleases. However, when using highly specific methods, DSBs should not be generated in large quantities (their significant increase can be caused only by a high nonspecific activity); therefore, cell cycle arrest cannot be expected to occur in all cells. Nevertheless, pweople must be aware of its possibility.

The activation of p53 is accompanied by cell cycle arrest in the G1 phase ^[32][33][51,52]. As noted earlier, HDR does not occur in the G1 phase; therefore, the inhibition of p53 may help to increase HDR. Indeed, the efficiency of GFP restoration using CRISPR-Cas9 in p53-deficient RPE1 cells was significantly higher than in the cells with wild-type p53 ^[32][51]. Subsequently, this observation was confirmed by several independent scientific groups in human pluripotent ^[34][53] and hematopoietic stem cells ^[31][50], as well as in ductal and hepatocyte organoids ^[35][54]. The main explanation for this phenomenon is the activation of p53 by nuclease-induced double-stranded breaks, which leads to p53-dependent arrest ^[32][51]. Nevertheless, the genome-scale CRISPR screen in different cell lines showed that p53 activation does not occur in all cell cultures, which should be taken into account in experimental work ^[36][55].

It is known that p53 is a key factor causing cell apoptosis in cases of abnormalities ^[37][56]; therefore, temporary inhibition of p53 may be unsafe and may lead to the accumulation of DNA damage in the cell, which must be taken into account when using this approach. It has been proven that p53 activates p21, which in turn inhibits all CDKs and arrests the cell cycle in any phase (M/G1, G1/S, and G2/M) ^[29][48]. Another concern is the possible clonal expansion of cells with mutations in the TP53 gene under conditions of in vivo genome editing, which could potentially lead to the development of tumors ^[38][57]. Recently, the selective advantage of cells with p53-inactivating mutations was experimentally confirmed in CRISPR-Cas9 studies ^[39][58].

Cell cycle synchronization can be achieved not only by the arrest of all cells in the culture at a certain point in the cell cycle but also by the simultaneous progression of the cycle for all available cells. This can be achieved with the help of mitogens, molecules that signal cells to enter the S phase ^[40][59].

Mitogens are typically small proteins that act as a signal to start cell division. Some growth factors are mitogens, such as epidermal growth factor (EGF) ^[41][60] and platelet-derived growth factor (PDGF) ^[42][61], but others, such as vascular endothelial growth factor (VEGF), are not ^[43][62]. They act via mitogen-activated protein kinases (MAPKs) and lead to the induction of mitosis (MAPK signaling pathway). Once the cells pass the G1 checkpoint, which is controlled by mitogens, the latter are no longer needed to continue the progress through the cell cycle. The MAPK signaling pathway is implicated in many cancers because dysregulation of this pathway leads to uncontrolled growth ^[44][45][63,64]. Mammalian cells require mitogens to proliferate; mammalian cell membranes have mitogen receptors that are typically receptor-associated tyrosine kinases, such as the epidermal growth factor receptor (EGFR). The binding of a mitogen to its receptor induces the activation of the Ras (rat sarcoma) protein, which leads to the activation of gene expression through transcription factors, such as c-Myc, serum response factor, etc. ^[46][65]. They activate the expression of cyclin D, which forms a complex with Cdk4 or Cdk6 called cyclin D-Cdk complex. This complex phosphorylates the retinoblastoma protein (Rb). Phosphorylated Rb interacts with the transcription factor E2F, which controls the expression of a number of genes required for DNA replication and mitosis, such as cyclin A and cyclin E ^[47][48][66,67]. This is not the only control exerted by mitogens; they can inhibit the glycogen synthase kinase (GSK3β) via phosphoinositide 3-kinase. GSK3β is a kinase that phosphorylates cyclin D at Thr286, keeping the cyclin D-CDK complex present in the G0 phase inhibited ^[49][68]. Thus, mitogens can be used to force cells to simultaneously enter the S phase.

The use of mitogens to increase HDR in genome editing experiments is often limited only to phytohemagglutinin for editing T-lymphocytes. Phytohemagglutinin (PHA) is a lectin derived from red kidney bean extract. It has strong agglutinating and mitogenic activities. PHA has been used for T-cell activation since 1960 ^[50][69]. It is also actively used in protocols for karyotyping T-lymphocytes. Kuo et al. showed that the efficiency of integration of the GFP cDNA into the 5′-UTR of the CD40LG gene in Jurkat T-cells using TALEN increased with the addition of PHA in a dose-dependent manner up to 20.7% (with 3 μg/mL PHA) ^[51][70]. However, no other similar studies, either with PHA or with other mitogens, have been published to date.

The lack of studies using mitogens may indicate that researchers understand the likely negative consequences. Hypophosphorylated (inactive) Rb is one of the key factors that prevent cells with damaged DNA from proliferating. Thus, overactivation (due to phosphorylation) of Rb would result in the appearance of cells with damaged DNA, which is not acceptable in clinical applications of this approach. Because of this, the use of mitogens to increase genome editing efficiency can be a double-edged sword.