In the model engineering strain, S. cerevisiae, CRISPR systems have been researched more deeply compared with those in other types of yeast. Using CRISPR strategies for gene editing in S. cerevisiae typically involves the following steps, including selecting a target gene, designing sgRNA sequences, constructing plasmids containing CRISPR elements, marker gene, and homology arms, transforming the plasmids or linearized DNA fragments along with donor DNA, and conducting selection and verification processes. From the initial research on the tolerance of Cas9 and guide RNA (gRNA) mismatches in vivo and in vitro to today’s study [46], many related genetic tools are becoming mature, such as some vector toolboxes: MoClo-based yeast chromosome modification kit [47], unmarked and commercialized EasyClone-MarkerFree [48], and a controllable copy number plasmid [49] constructed with dominant markers. Regarding the effects of CRISPR-relevant tools, researchers can knock out fragments over 30 kb in S. cerevisiae [50]. By using the delta site as a targeting site, the Di-CRISPR platform can integrate up to 18 copies of the gene in one step [51]. In addition, the CRISPR Cas9 system can construct auxotrophic mutants of industrial tetraploid yeast strains [52]. In the research of S. cerevisiae genomics, mCRISPR combined CRISPR-Cas9 and transformation-associated recombination (TAR) to enable single-marker multiplexed promoter engineering of large gene clusters, which enabled efficient multiplexed engineering of natural product biosynthetic gene clusters [53].

Efficient screening of ideal industrial yeast strains depends on the establishment of related libraries. Taking advantage of Cas9 mutants with confirmed genetic variations (deletion, substitution, and insertion) in yeast, a genome-wide library established in S. cerevisiae-short, trackable, integrated cellular barcodes (MAGESTIC) achieved saturation editing of the essential gene SEC14, identifying amino acids highly related to inhibiting lipid signaling [54]. Guo and colleagues integrated next-generation sequencing technology (NGS) to construct a genome library targeting 315 small open reading frames with unclear characteristics in the yeast genome, which evaluated genes crucial for growth in various environments [55]. S. cerevisiae is an industrial strain commonly used for the production of ethanol. Liu and colleagues have constructed a Global Regulatory Network (MINR) with 43,020 specific mutations of 25 regulatory genes and identified yeast strains with improved ethanol tolerance [16].

Common transcription regulation systems in eukaryotic cells can be divided into three generations [56], all of which use Cas proteins lacking endonuclease activity. For example, dCas9 can be obtained by mutating specific amino acids in the RuvC and HNH functional domains [57]. Similarly, other types of Cas proteins can also be utilized for transcriptional regulation [56]. The principle of the first-generation transcription control system is that when Cas protein is co-expressed with sgRNA, a DNA recognition complex is produced, and then it becomes a roadblock in the transcription process. Furthermore, fusion expression of transcription activator or transcription repressor with dCas9 can achieve gene activation or inhibition. Differently, compared with the additional modification of dCas9 in the first generation, the second generation used a special scaffold RNA to recruit specific effectors. The formed RNA aptamer–RBP pair could regulate the transcription of target genes [58]. The third-generation system is a dimerization system controlled by chemistry and light, further realizing the temporal and spatial control of gene functions.

In recent years, dCas9 has been widely applied in the regulation of the yeast metabolic network. In the glycerol synthesis pathway, by changing the target position of sgRNA, dCas9 can be used to express genes in stages (activation or inhibition) and test systematically enzyme perturbation sensitivity (STEPS) [59], in order to identify the rate-limiting steps in the metabolic pathway. dCas9 can not only fulfill individual inhibition or activation, but also achieved orthogonal transcriptional activation, transcriptional interference, and gene deletion (CRISPRAID). The three-function system containing Cas9 proteins from different sources could modularly, parallelly regulate and interfere with the metabolic network in high-throughput ways. After optimizing different combinations of inactivated nucleases, activators and inhibitors, the production of β-carotene could be increased by three times [17]. A more powerful synthetic biology tool, multi-functional genome-wide CRISPR system (MAGIC), can precisely control the defined genes in the genome. It has great application prospects in the genome-scale engineering of advanced eukaryotes [60].

Beyond the spatial level, synthetic biological elements can realize real-time gene regulation. Researchers have constructed and compared two dCas9-mediated systems, constitutive and inducible, with different gRNA design strategies. In this way, high production of isoprenoids was achieved and production of triglycerides was greatly improved [18].

As dCas9 and Cas9 nickase (D10A) will not cause DNA double-strand breaks, fusion expression of their functional domains with adenine/cytidine deaminase can change the corresponding DNA bases and achieve single-base editing [61]. Earlier, through the fusion of CRISPR/Cas9 lacking nuclease activity and activation-induced cytidine deaminase (AID) orthologs, the artificial synthetic complex (Target-AID) [62] can work efficiently in yeast, while in mammalian cells, Uracil DNA glycosidase inhibitors (UGI) are needed to inhibit insertions and deletions.

In addition to the developed adenine base editor ABEs [63] without causing double-strand DNA breaks, Liu and colleagues have recently built a single-base editing system [64,65], which brought a milestone breakthrough in the field of gene editing. The simplest system only contained engineered reverse transcriptase, Cas9 nickase and the guide RNA. It directly wrote new genetic information into the designated position of DNA in eukaryotic cells. Since many human diseases are related to single-base mutations, the CRISPR system has been demonstrated the potential to correct known pathogenic human genetic variants in the medical field.

By means of connecting genetic components, a genetic circuit with complex logic functions can be constructed. In order to solve the problem of the signal degradation when layered in dCas9 system, the NOR gate developed in S. cerevisiae based on dCas9-Mix1 has been established [66]. In a single-gene NOR logic gate, the input and output signals of the gate are programmable signal molecules-gRNAs of the same molecular type. The input end is composed of RNA Pol II pGRR promoter, and the output end is the gRNA transcript flanking with self-splicing ribozyme (RGR). On the basis of this, the NOR gate is integrated into a single yeast genome, and the fluorescent protein intensity at the output end was the signal comparing the effects of different logic circuits. It was found that digital logic circuits with up to seven gRNAs, five NOR gates, and up to seven ladders could be constructed.

In addition, in S. cerevisiae, an AND logic gate was developed with a transcription activation system composed of MCP-VP64, scRNA, and dCas9 [67]. The AND logic gate consisted of two switches, and activating two independent switches made the reporter gene express. The endogenous Gal10 regulated by galactose and the heterologous transcription factor LexA-ER-AD regulated by β-estradiol controlled the constitutively expressed RNA scaffold. It is worth mentioning that the system utilized ribosome skipping T2A sequence to bind target gene and fluorescent reporter gene. This peptide could make a transcript translate multiple proteins due to ribosome jumping, so as to adapt to any target gene and ensure the reporting function simultaneously.