Inducible promoters are activated to initiate gene transcription or repressed to decrease target expression by the addition of stimuli to the medium. They have been developed to be applied for the optimization of heterologous metabolic pathways and regulatory networks for accurate gene mediation. Glycolytic promoters, the most powerful promoters of S. cerevisiae obtained from genes encoding glycolytic enzymes, are the first promoters applied in protein production, and can be induced by glucose addition to the medium (Rajkumar et al., 2019). In contrast, galactose-regulated promoters, which are strongly repressed by glucose and generated from genes involved in metabolizing galactose (such as GAL1, GAL2, GAL7, and GAL10) (Romanos et al., 2001; Peng et al., 2017), are the most powerful and tightly regulated promoters in S. cerevisiae. Those GAL promoters have been used in metabolic engineering approaches for the reconstruction of complicated heterologous metabolic pathways and in genetic engineering for stable genetic manipulations. The glucose-repressible promoters ADH2 (alcohol dehydrogenase 2) (Peng et al., 2015), invertase gene SUC1 (Rodriguez et al., 1981), and invertase gene SUC2 are strongly repressed by the addition of glucose while they are activated by reducing glucose concentration in the medium (Sharma et al., 2022). The powerful and tightly-regulated promoter ADH2 has been used for the expression of heterologous genes (especially genes encoding toxic proteins, e.g., insulin-like growth factor I) owing to its trait of being heavily repressed by glucose (Shuster, 1989; Price et al., 1990). Promoters that are not induced by the nutrients in the medium can also be used for heterologous gene expression; these include temperature-regulated promoters, steroid-regulated promoters, foreign promoters, and the copper-induced promoter CUP1. CUP1 is responsible for moderate transcriptional strength under a certain copper concentration, which is typically used to avoid metabolic imbalance during the log growth phase or post-log phase without extra copper addition. Although copper-induced expression systems show a high copy number and stable transcriptional strength, the repeatedly used CUP1 gene tends to result in a loss of introduced GOI by homologous recombination. Synthetic regulatory modules for gene transcription are expected to reduce genetic instability caused by a reduplicated induction system for metabolic modification and regulatory introduction of heterologously metabolic pathways (Peng et al., 2017).

While inducible promoters have the capability to maintain the yeast culture at a non-expression status during the cell growth phase, which results in the minimization of the selection for non-expression mutants (Mattanovich et al., 2012), constitutive promoters seem to be preferred. Compared to regulated inducible promoters, constitutive promoters are unregulated, which means that the transcription efficiency of a heterologous gene controlled by constitutive promoters is almost unaffected by both internal and external factors. Moreover, inducible promoters are generally activated under specific circumstances (e.g., addition of biotic or abiotic stimuli), and commonly negatively affect the purity and concentration of target proteins, even requiring a more rigorous isolation strategy. Notably, constitutive promoters are responsible for the application of the expression of both intracellular and extracellular proteins by S. cerevisiae, yet overexpression of a target protein often specifically depends on the target protein being expressed in S. cerevisiae (Da Silva and Srikrishnan, 2012). For example, the promoter ADH1, which is one of the most extensively utilized constitutive promoters for the production of heterologous proteins in yeast, naturally controls the expression of alcohol dehydrogenase 1 in S. cerevisiae. Another constitutive promoter (pTEF1) controls the expression of the translation elongation factor EF1 alpha, showing higher expression intensity than promoter alcohol dehydrogenase 1 and being unrepressed by ethanol (Partow et al., 2010). pTEF1 has been used to enhance the production of r-proteins and homologous enzymes by S. cerevisiae (Steinborn et al., 2007; Zhang et al., 2017; Xiao et al., 2023). Other constitutive promoters have also been applied to the construction of expression vectors (Partow et al., 2010; Lee et al., 2013). For example, Pgpd (which encodes glyceraldehyde 3-phosphate dehydrogenase) is used to regulate heterologous or recombinant gene expression processes (Knudsen et al., 2015); pPGK1 (which encodes phosphoglycerate kinase 1) is used for accurate metabolic and transcriptional regulation (Lamour et al., 2019; Jin et al., 2020; Shi et al., 2022); pTPI1 (which encodes triosephosphate isomerase) is used to precisely regulate protein expression and markedly increase industrial product production by S. cerevisiae (Chen et al., 2019; Azizoglu et al., 2021; Olzhausen et al., 2021; Paramasivan et al., 2022); pHXT7 (which encodes a hexose transporter) is used to improve the sugar transportation efficiency in the metabolic pathways of S. cerevisiae (Vasylyshyn et al., 2020; Zhu et al., 2021); and pPYK1 (which encodes pyruvate kinase 1) is used to rewire certain metabolic networks (Kim et al., 2016; Zhang et al., 2019). Although constitutive promoters are the most used promoters for foreign gene expression because of their unimpaired transcriptional strength, the effects of various culture conditions on heterologous metabolic pathways are unpredictable.

Although varieties of powerfully inducible and constitutive promoters that strengthen the transcriptional intensity have been used for the production of heterologous proteins and r-proteins, developing a double-promoter expression system (DPES) with higher efficiency is still desirable (Öztürk et al., 2017). DPESs that control the transcription of the same gene through the combinatorial use of two promoters tend to result in higher extracellular protein production and more stable co-transfection. DPESs can also be used for heterologous protein expression by both prokaryotes and eukaryotes via ligation-independently cloning of the foreign gene into the dual expression vectors; this process markedly simplifies the cloning procedures, functionality expression, and protein purification process (Sinah et al., 2012). Except for typically streamlined DPESs, bidirectional promoter expression vectors featuring bidirectional initiation of the transcription of each GOI by collectively using identically constructed transcription regulator modules represent competent or higher r-protein production (Figure 1A) (Öztürk et al., 2017). For instance, a bidirectional vector containing dual cytomegalovirus promoters for the individual initiation of the expression of each GOI in a single plasmid was developed to produce recombinant antibodies; this enabled the simultaneous expression of the heavy chain and light chain for monoclonal antibody production; further, it represents competent antibody production, superior functionalities, as well as simplified manipulations compared to traditional two-plasmid co-transfection (Carrara et al., 2021). Promoter trap systems have been used to screen active promoters by cloning random DNA fragments into the multiple cloning sites of a marker gene upstream. This technique is utilized to identify active promoter elements and simultaneously obtain powerful promoters by shuffling and recombining cloned unidentified DNA fragments (Yang et al., 2013). A double or multiple promoter expression system that strongly initiates the transcription of the same target gene by using two or more promoters is expected to achieve a remarkable enhancement of heterologous protein production. Regarding different operational mechanisms, double promoter expression systems that improve foreign gene expression by elevating the transcription intensity can mainly be classified into consecutively and simultaneously operating double promoter expression systems (Öztürk et al., 2017). Consecutively operating double promoter expression systems have been generally used to prolong the transcription length through the differently phased activation of each promoter under different conditions (Figure 1B); the simultaneously operating double expression systems efficiently enhance the transcriptional strength by activating both promoters under the same circumstances (Figure 1C) (Öztürk et al., 2017).

In addition to the construction of a common bidirectional or double promoter expression vector, a promising approach to overexpress heterologous and r-proteins is the use of a synthetic promoter. Synthetic promoters can be defined by inserting functional cis-elements into the distal promoter region to either positively or negatively recruit the corresponding transcription factors, specific regulation of transcription efficiency and strength can be achieved without being constrained by endogenous factors (Machens et al., 2017; Yasmeen et al., 2023). For example, the α-amylase titer increased by 95% in S. cerevisiae by using an unfolded protein response synthetic responsive promoter, which is constructed by inserting the unfolded protein response element in front of the upstream enhancer region of yeast native promoters (Xiao et al., 2023). Artificially synthesized promoters commonly feature shorter sequences, which are more accurately regulated and more stably condition-induced for heterologous metabolic transcription in dynamic metabolic networks (Wichmann et al., 2023).

An effective SP serves as the leader in orienting the target protein into the ER secretory pathway that is commonly attached to the N-terminus of the secretory protein (Bae et al., 2022). Furthermore, the efficiency and capability of translocation that greatly impacts the secretion of target proteins in both prokaryotes and eukaryotes mainly depend on the selection of SPs and secretion mechanisms (Owji et al., 2018). In addition, only if a specific SP is conferred to the target protein (von Heijne, 1990), the newly synthesized polypeptide can enter the ER secretory pathway for subsequent processes. Despite the low specificity of endogenous SPs in yeasts, the use of foreign SPs usually obtains a poorer secretion efficiency than that of SPs originating from yeast (Kaiser et al., 1987). Moreover, whether foreign signals will function in yeast remains unknown and unpredictable, and secreted signal sequences that are homologous to S. cerevisiae (e.g., those from acid phosphatase, α-factor, and invertase) are extensively used in heterologous protein secretion. Among them, the most widely used signal sequence originating from the pre-pro region of mating factor α (MFα) has been demonstrated to be generally applicable and functional in the expression of various heterologous proteins by S. cerevisiae (Ernst, 1986).

Unfortunately, the construction of a universal SP for all heterologous protein production seems to be impossible. Although a foreign SP can be expressed in a host strain (Kober et al., 2013), compared to the endogenous SP, target protein secretion is always inefficient, and the SP cannot be recognized by native translocons and cleaved off. For instance, the use of E. coli OmpA SP for heterologous protein production represents both a higher expression level and more efficient secretion of Bacillus subtilis chitosanase than Bacillus SP by the E. coli expression system (Pechsrichuang et al., 2016). Appropriate selection of SPs for heterologous proteins is often conducive to maintaining a balance between translocating efficiency and ER folding capability (Owji et al., 2018) rather than triggering a series of degradation mechanisms that lead to holistic proteolysis of both properly folded proteins and aberrantly aggregative proteins.

Foreign gene expression mainly hinges on the folding rate of the ER, while the secretion of heterologous proteins that are overexpressed via the ER secretory pathway always reaches saturation regardless of the posttranscriptional mechanism of S. cerevisiae (Schröder, 2008; Young and Robinson, 2014). Rational selection of promoters and designated modification of secretion signals are capable of enhancing heterologous protein production to a certain extent (Rakestraw et al., 2009); however, these approaches are generally not applicable to all heterologous proteins. Most studies mainly concentrated on host strain engineering, including overexpression of target genes, modulation of factors related to protein secretion, optimization of expression vector systems, and perfection of the fermentation process (Idiris et al., 2010). Compared to host strain engineering, translational fusion (in which the target gene is fused with a specific TFP) can improve target gene expression without impairing the structure of the target protein or imposing an additional burden on the ER (Bae et al., 2015).

Translational fusion is a promising expression system for enhancing the expression of both heterologous and r-proteins, in addition to using the designed fusion partners. The commonly utilized fusion system, which is a fusion tag called small ubiquitin-like modifier (SUMO), is the S. cerevisiae SUMO protein Smt3 and SUMO protease 1. Analogous to the removal of TFPs, the fusion tag SUMO can be removed by digestion of a highly specific SUMO protease (Lau et al., 2018), e.g., Smt3 can be deconjugated from the target protein by SUMO protease 1 (Mossessova and Lima, 2000). Similar to the functions of translational fusion, SUMO fusion is responsible for facilitating expression, increasing solubility, and simplifying purification procedures of target proteins (Malakhov et al., 2004; Marblestone et al., 2006). A stably constructed and highly expressed functional TFP that involves heterologous metabolic pathways and biosynthesis of a target protein always exhibits a remarkable upregulation of target gene expression.

The pre-pro peptide of S. cerevisiae MFα has been identified as a powerful fusion partner for the production of various heterologous proteins, yet MFα is not an all-purpose secretion leader for all target proteins. Therefore, the optimal selection of a fusion partner and its appropriate modifications are deemed to directly impact secretion efficiency. The fusion partners function well for enhancing heterologous protein production, which can be mainly attributed to the increase of solubility of fusion proteins in the ER and the facilitation of transport of fusion proteins to the Golgi apparatus (Dälken et al., 2010; Xu et al., 2018). As Kex2p is a Ca²⁺-dependent serine protease that specifically recognizes dibasic residues (Lys-Arg, Arg-Arg) (Fuller et al., 1989), fusion partners can be removed by cleavage after dibasic residues in the late Golgi network through the artificial introduction of the Kex2p cleavage site. The efficient endoprotease Kex2p that was discovered in S. cerevisiae is a processing protease that is responsible for the proteolytic maturation of killer toxins and MFα (Kim et al., 2022). Kex2p has been used for the production of r-proteins; for example, when expressed with the MFα signal peptide that contains a Kex2p cleavage site, overexpression of KEX2 has been shown to enhance the production of the r-protein in Pichia pastoris (Sreenivas et al., 2015). Even though the fusion system retains the intact construction of the target protein, no fusion partner is omnipotent for all protein varieties. Fortunately, the origination of fusion partners can be generated from open reading frames that encode proteins carrying secretion signals. In addition, fusion partners can be generated from different-sized fragments of a single open reading frame, and the secretion efficiencies of different translational fusion systems always depend on the secretory capabilities of target proteins. Powerful fusion partners have been generated and screened from the known open reading frames encoding a segment of signal peptide or fusion peptides that highly secreted proteins in yeast (Bae et al., 2015).

The ER is one of the largest membrane organelles responsible for the accurate folding, posttranslational modification, and final sorting of various proteins. Newly-synthesized peptides must be properly folded before being assembled into their target organelles (Gardner et al., 2013). When secretory demands are high, protein synthesis errors may occur, resulting in the accumulation of aberrantly folded proteins. Consequently, abnormal accumulation in the ER induces an unfolded protein response (UPR), which is an adaptive signaling pathway to combat both misfolded and unfolded proteins by elevating the expression of genes that enhance the ability and efficiency of ER protein folding (Krishnan and Askew, 2014). Inevitably, a vast fraction of polypeptides will still fail to complete the appropriate conformation even under the impact of UPR (Hartl and Hayer-Hartl, 2009). Ultimately they are eliminated by ER-associated degradation, which is a signaling pathway that retro-translocates misfolded proteins back into the cytosol and then degrades them via the ubiquitin-proteasome system (Smith et al., 2011; Ruggiano et al., 2014). Additionally, unfolded proteins that cannot be transported from the ER to the Golgi apparatus accumulate in the lumen of the ER, which represents one of the rate-limiting steps in protein secretion of eukaryotic microorganisms (Lodish et al., 1983; Shuster, 1991). The ER lumen contains amounts of ER-resident protein folding chaperones and foldases that play a role in sustaining ER homeostasis by correcting both unfolded and misfolded peptides. Thus, the increase in the expression of these protein-folding chaperones and foldases is bound to exert a positive impact on the folding efficiency of heterologous proteins expressed by S. cerevisiae.

The karyogamy gene Kar2, which encodes the S. cerevisiae chaperone binding immunoglobulin protein (Bip), is homologous to mammalian BiP and shows the highest match to it. Kar2 includes a functional secretory signal sequence that facilitates ER translocation at the N-terminus, lacks N-linked glycosylation sites, and contains an “ER retention” signal HDEL (His-Asp-Glu-Leu)/(-GCTTGACGAACT-) at the C-terminus (Normington et al., 1989; Rose et al., 1989; Kasuya et al., 1999). Saccharomyces cerevisiae Bip (Kar2) is the sole member of the heat shock 70 kDa protein family that is segregated into the ER for protein folding (Plemper et al., 1997; Brodsky et al., 1999). Kar2 also functions in translocating newly synthesized polypeptides across the ER membrane and recycling aberrantly folded proteins back into the ER for degradation (Normington et al., 1989; Simons et al., 1995). Furthermore, Bip transiently interacts with a variety of newly synthesized extracellular proteins but interacts more permanently with both aberrantly folded and unassembled proteins (Gething and Sambrook, 1992). Regardless of whether it is glycosylated or nonglycosylated processed, BiP binds to these aberrant proteins in the form of a stable complex (Pobre et al., 2019), leading to a reduction of the concentration of BiP in the ER lumen.

The decrease in the concentration of free BiP in the ER as well as the interaction between BiP and these aberrant proteins are intimately linked to UPR initiation. As Bip can be induced by the UPR to increase the protein folding capability of the ER, thus alleviating the high secretory demand the ER encounters; the resulting increase of Bip expression is expected to elevate the production of heterologous proteins. Bip has been demonstrated to play a positive role in heterologous protein production and the secretion of heterologous protein is proportional to the expression level of Bip (Tang et al., 2015; Zinkevičiūtė et al., 2015; Jiao et al., 2018; Jiang et al., 2023). However, overexpression of Bip in the S. cerevisiae expression system is not always adaptable for the secretion of heterologous protein, even negatively impairing the r-protein production. For example, no or neglectable increase was observed in the secretion of several different r-proteins with overexpression of chaperone Bip in S. cerevisiae (Robinson et al., 1996; Kauffman et al., 2002; van der Heide et al., 2002). A possible explanation is that Bip might more frequently bind to the unoptimized heterologous proteins, directing expressed proteins to degradation rather than to the secretory pathway (Gasser et al., 2008).

When the ER encounters a high secretory demand or aggregation of newly synthesized polypeptides, the UPR will be triggered to buffer the folding intensity, featuring an increase in the folding capability of the ER (Menzel et al., 1997). The constitutive induction of UPR aids protein folding in the ER not only for native protein production but also for the production of heterologous proteins, which is mainly associated with the upregulated expression of foldases and ER-resident chaperones in the ER. Moreover, the disruption of the gene HAC1, which encodes the UPR regulator that upregulates the expression of folding factors in the ER, leads to a remarkable decrease in heterologous protein production (Valkonen et al., 2003).

The type-Ⅰ transmembrane protein kinase IRE1, which is also called endoplasmic reticulum to nucleus-1 protein, resides in the ER lumen where it acts as a stress sensor that transmits stress signals to the luminal domain of IRE1 in response to protein folding fluctuation (Kaufman, 1999; Patil and Walter, 2001). Another type-Ⅰ transmembrane protein kinase (PER-like ER kinase, PERK) can also be activated to combat environmental or native stresses by phosphorylating the α-subunit of eukaryotic translation initiation factor 2 (Shi et al., 1998), which directly inhibits the synthesis of nascent proteins (Harding et al., 1999). When secretory stress occurs (Figure 2), i) unfolded proteins trigger oligomerization and trans-autophosphorylation of both IRE1 and PER-like ER kinase and activate their RNase domains (Rubio et al., 2011; Gardner et al., 2013; Kopp et al., 2019); ii) Bip commonly binds to the luminal domain of IRE1 and under unstressed conditions, PER-like ER kinase is released to assist protein folding or to degrade aberrant polypeptides (Bertolotti et al., 2000); iii) UPR causes the reduction of free Bip in ER lumen. Activated IRE1 is responsible for the cleavage of unconventional introns that are absent from the HAC1 mRNA to maintain ER homeostasis (Sidrauski and Walter, 1997); it also encodes a functional transcription factor that belongs to the basic leucine zipper family (Nikawa et al., 1996; Sidrauski et al., 1996). This unconventional mRNA splicing increased the expression of the transcription factor HAC1 that induces UPR target genes to combat secretory demand by increasing the folding capability of ER and the expression of both folding factors and ER-resident chaperones (Kimata and Kohno, 2011; Walter and Ron, 2011).

However, forced induction or overexpression of UPR can only increase a fraction of heterologous protein production (Valkonen et al., 2003); thus, overexpression of foldases or ER-resident chaperones is not omnipotent for all foreign gene expressions. Once the translational level of a foreign gene reaches that of endogenous expression, proper folding for nascent proteins in the ER probably becomes one of the main limitations of the production of heterologous proteins. Because the efficiency of protein folding and assembly is regarded as the most critical process that is directly related to the final secretion level, modulation of the UPR pathway is expected to overcome the secretory bottleneck of heterologous proteins.

Attributed to the simplicity of type II precise genome editing technology CRISPR/Cas systems—in which the CRISPR RNA-effector (crRNA-effector) functions by a single protein Cas9, which differs from that of type I and type III(Makarova et al., 2015; Rainha et al., 2020)—CRISPR/Cas9 systems have been extensively used for single- or multiple-gene disruptions, single nucleotide mutations, and gene targeting insertions. For example, the production of α-amylase is increased by nearly 1.8-fold in S. cerevisiae by CRISPR/Cas9 mediated multiplex point mutation and gene deletion (Wang et al., 2022). Also, CRISPR/Cas9-mediated heterologous expression of enzymes for metabolic regulation is widely used to enhance the production of chemicals, natural products, and therapeutics. For instance, the secretion of taxadiene synthase is improved by using an optimized genetic editing toolkit, which leads to a 25-fold increase in anti-tumor chemical taxadiene production in S. cerevisiae (Reider Apel et al., 2017). The CRISPR/Cas9 system that has been most widely used for heterologous gene editing is composed of: i) the Cas9 protein that is oriented to the target DNA site for the recognition of the given target DNA sequence, contains a protospacer-adjacent motif (PAM) sequence, and achieves cleavage of the target genomic DNA site (Rainha et al., 2020); ii) CRISPR RNA (crRNA) that is transcribed from the CRISPR sequence as precrRNA; iii) trans-activating crRNA (tracrRNA) for the maturation of crRNA via the activation of the conserved RNase III(Deltcheva et al., 2011); iv) a donor DNA template containing homology arms in 5′ and 3’ as well as carrying a GOI. Briefly, the type II genome editing technology functions in the direction of the Cas protein to interact with tracrRNA-crRNA duplexes; then, a double-stranded break (DSB) is created that is repaired via nonhomologous end-joining (NHEJ) and/or homologous recombination (HR) at the target gene site for the precise deletion or insertion of genes (Figure 3).

The most extensive utilization of a single guide RNA (sgRNA) in S. cerevisiae increases the efficiency of gene editing by the CRISPR/Cas9 system. For example, Ryan et al. developed multiplex CRISPR, utilizing a self-cleaving hepatitis delta virus ribozyme that is fused to sgRNA for the removal of redundant RNA sequence, achieves nearly 100% efficiency of the CRISPR/Cas9 mediated homologous direct targeting by a single sgRNA (Ryan et al., 2014). The guide RNA that is composed of a scaffolding RNA (which is required for forming the Cas9/gRNA complex) and a 20 bp gRNA spacer sequence (which is complementary to the target DNA site that directs the Cas protein to the target DNA PAM site) consolidates the functions of both crRNA and tracrRNA. Additionally, the induction of DSBs, which is generally regarded to cause genome lesions that threaten genome stability and cell survival (Gnügge and Symington, 2020), is also applicable for genomic integration resulting from the highly efficient HR repair of S. cerevisiae; this repair maintains high integrity of the host genome and achieves high-fidelity repair of DSBs by using donor DNA (Lisby et al., 2001; Wright et al., 2018). Because of the specific PAM sequence that is recognized by the Cas protein, the efficiency and accuracy of both-end design for gRNA always constrain the genome editing capability and efficiency when using the CRISPR/Cas9 system, especially for multiplex genome editing (Adiego-Pérez et al., 2019). Therefore, computer-aided design or other auxiliary tools for the desired gRNA that are appropriate for a variety of Cas proteins are required for efficient genome editing and multiplex gene integration (Bourgeois et al., 2018; Li et al., 2020). For example, the programmable red light switch PhiReX 2.0 has been developed to rewire metabolic fluxes by targeting endogenous promoter sequences through sgRNAs, which leads to 6-fold upregulation of a native promoter gene (Machens et al., 2023).

Biosynthetic techniques based on the CRISPR/Cas9 system are expected to achieve multiplex genetic manipulations and genome integration without the need for additional repeated gene edits and selective markers. The single RNA-guided Cpf1 endonuclease that was defined as a class 2/type V CRISPR/Cas system shows specificity to several bacteria and an exceptional archaeon (Candidatus Methanomethylophilus alvus) (Vestergaard et al., 2014; Makarova et al., 2015); it was classified as the effector protein Cfp1 that is responsible for target cleavage of the PAM sequence and crRNA processing (Makarova et al., 2017). The following presents developments and features that differ from those of the type II CRISPR/Cas system: i) Cpf1 protein is guided only by crRNA without the requirement for a tracrRNA compared to a gRNA that is required for Cas9(Zetsche et al., 2015); ii) the T-rich PAM sequence of the Cpf1 protein is located at the 5′ end of the target DNA, which contrasts with the location of Cas protein at the 3’ end of target DNA (Verwaal et al., 2018); iii) Cpf1 protein is responsible for processing pre-crRNA to the mature state and can be expected to simplify the multiplex genome editing procedures as well as facilitate the development of the CRISPR/Cas gene editing system.

Based on a self-cloning CRISPR/Cas9 system (scCRISPR/Cas9), which contains a self-cleaving palindromic sgRNA plasmid that is repaired by HR for CRISPR/Cas9-mediated genomic mutation and site-specific knock-in transgene creation (Arbab et al., 2015), the scCRISPR/Cpf1 system facilitates the genome editing process by circumventing any cloning step; this circumvention leads to highly efficient singleplex and tripleplex genomic integration (Li et al., 2018).

Microorganisms have evolved to combat environmental stresses with the goal to maintain metabolic homeostasis, which hinders the realization of high titer, rate, and yield for target proteins, and S. cerevisiae is no exception. Hence, rewiring the metabolism of S. cerevisiae is needed to extend substrate ranges, improve cellular properties, obtain new capabilities, enhance endogenous or heterologous protein production, and tolerate internal or external stresses (Du et al., 2011; Hong and Nielsen, 2012; Jens and Keasling, 2016; Lian et al., 2018). Genomic editing technologies are derived from CRISPR/Cas9 systems, including multifunctional CRISPR systems with truncated gRNAs without creating DSBs(Figure 4): i) tri-functional CRISPR-AID combines transcriptional activation (CRISPRa), transcriptional interference (CRISPRi), and gene deletion (CRISPRd). It achieves the phenotype desired by combinatorial metabolic engineering by generating possible gRNA combinations in a single system that uses three orthogonal nuclease-deficient CRISPR proteins to prevent crosstalk between gRNAs(Lian et al., 2017; Schultz et al., 2018); ii) tri-functional CRISPR-ARE combining CRISPRa, transcriptional repression (CRISPRr), and genome editing (CRISPRe) achieve the combinatorial metabolic engineering of non-model organisms by using a single nuclease-active Cas9-VPR fusion protein (compared to three independent PAM-recognizing CRISPR proteins of CRISPR-AID) (Dong et al., 2020). Moreover, EvolvR, which combines the target specificity of CRISPR/Cas systems, enables random or target insertion by an error-prone DNA polymerase in a single-stranded break. This break is created with a nicking variant of Cas9 (nCas9) when native homology repair is initiated (Halperin et al., 2018). EvolvR-mediated mutagenesis is expected to realize continuous target nucleotide diversification and multiplex gene mutations in higher eukaryotes, e.g., yEvolvR applied in S. cerevisiae (Tou et al., 2020). Another CRISPR/Cas9-based multiplexed integration system features in recruiting Rad51 recombinase to the proximity of the DSBs by fusing a protein Brex27 to Cas protein, which enhances homologous recombination and achieves 78% quadruple integration efficiency in S. cerevisiae (Meng et al., 2023).

Remarkably, because of the complicated cellular metabolism and intimate interaction between gene expression and metabolites, the changes that various modifications impose on metabolic pathways are unpredictable.

Numerous specifically developed promoter expression systems and pathway modification procedures have been used to enhance the gene transcription level for metabolic engineering; however, the transcriptional regulation efficiency for the complex metabolic pathway remains relatively low. The direct fusion of functional units (activator or repressor) and nuclease-deficient CRISPR proteins that have been used in CRISPRa/i exhibit highly precise and efficient transcriptional regulation. For example, the fusion of location functional units that can orient metabolites to target metabolic pathways to the deactivated Cas9 (dCas9) is expected to reduce the residence of bioproducts and facilitate the elucidation of unclear mechanisms (Xie et al., 2020), leading to efficient secretion and discovery. However, single-effector fusion and finite fusion partners are always unsuitable and incompetent for the efficient regulation of complicated metabolic pathways and multi-gene regulation. A CRISPR-mediated regulator recruiting system has been developed to amplify transcriptional activation and repression. This system has been developed by fusing protein scaffolds (SPY and Sun Tag systems) with nuclease-deficient CRISPR proteins so that several effectors are recruited for accurate and efficient multi-regulation of target gene transcription (Zhai et al., 2022).

Even though protein fusion strategies can achieve a substantial improvement of genetic manipulation efficiency, nucleotide-level fusion strategies can also achieve efficient genome modulation while leaving out the translation. The synthetic sRNA-mediated gene regulation system achieves the repression of target gene expression at the post-transcriptional level with the regulator Hfq protein (Yeom et al., 2022), which involves the binding of sRNA and target mRNA. Moreover, the synthetic sRNA assembled with a specifically modified mRNA-binding module sequence can be expected to bind to any target mRNA via base-paring, resulting in specific repression and precise regulation of metabolic pathways. To simultaneously regulate multiple target genes, the cloning of several synthetic sRNAs to a vector or the use of several sRNA vectors are conventional processes. However, the repeated use of the same scaffold sequences always leads to low regulatory efficiency, which can be attributed to unintentional HR. Therefore, the fusing of several mRNA binding modules that are specific for target genes to a single sRNA scaffold can achieve multiplex gene regulation, thus simplifying genetic procedures while imposing no additional burden on the cell (Yeom et al., 2022).