The strains, plasmids, and primers used in this study are listed in Supplementary Table S1. S. cerevisiae BY4741 strain was preserved by the research group. The Top10 competent strains of E.coli were purchased from Bomed Genetic Technologies Ltd., and the plasmids pESC-Ura and pESC-His used for plasmid vector construction were from the storage of the research group. Yeast Extract Peptone Dextrose Medium (YPD) contained 20 g/L peptone, 10 g/L yeast powder and 20 g/L glucose. Luria-Bertani (LB) medium contained 10 g/L peptone, 5 g/L yeast powder and 10 g/L sodium chloride. SD-Ura medium (uracil trap medium) contained 6.7 g/L yeast nitrogen source (YNB) without amino acids, 1.3 g/L other essential amino acids, histidine (His) 100 mg/L, leucine (Leu) 100 mg/L, tryptophan (Trp) 100 mg/L, 20 g/L glucose. SD-His medium (histidine deficient medium) contained 6.7 g/L YNB without amino acids and 1.3 g/L other essential amino acids. Without amino acid YNB 6.7 g/L, uracil (Ura) 50 mg/L, leucine (Leu) 100 mg/L, tryptophan (Trp) 100 mg/L, 20 g/L glucose. SD-Ura and SD-His media and pH was adjusted to 6.2 by 2 mol/L NaOH. Solid medium was prepared by adding 20 g/L Agar powder.

Using plasmid PUC19-Cas9 as a template, primers were designed to split the SpCas9 gene into two fragments SpCas9(F) and SpCas9(R) and introduce mutations at the first and last ends to mutate the amino acid aspartic acid (D) at position 10 to alanine (A) and histidine (H) at position 841 to alanine (A), and the SpCas9(F) and SpCas9(R) fragments were ligated to obtain the SpdCas9 gene by OE-PCR, thus to realize SpdCas9 nuclease inactivation.

Using plasmid pY004-FnCpf1 as a template, primers were designed to split the FnCpf1 gene into two fragments FnCpf1(F) and FnCpf1(R) and mutations were introduced at the end to mutate the amino acid aspartate (D) at position 917 to alanine (A), and the FnCpf1(F) and FnCpf1(R) fragments were ligated by OE-PCR to obtain the FndCpf1 gene, thus to realize FndCpf1 nuclease inactivation.

The FBA1p/TEF1p promoter was amplified by PCR on the BY4741 genome. Taq DNA polymerase was purchased from TaKaRa (Dalian). Restriction endonuclease was purchased from Thermo Scientific. gRNA expression cassette was constructed on the basis of plasmid pRS423-gRNA. The target fragments SNR52p-20 bp and 20 bp-sgRN scaffold-SUP4t were obtained by PCR amplification. Plasmid pESC-Ura was cleaved with HindIII/KpnI to obtain fragment pESC-Ura, and the three fragments were Gibson assembled by Gibson ligation. Again, the target fragments ADH1p-MCP-linker-VP64-ADH1t and ADH1p-MCP-linker-KRAB-ADH1t were obtained by PCR amplification. Plasmid pESC-Ura was digested with EcoRI/BamHI to obtain fragment pESC-Ura, and the three fragments were subjected to Gibson assembly by Gibson ligation. Gibson assembly. The pESC-ura-PCP-VP64 (Ura-PV) and pESC-ura-PCP-KRAB (Ura-PK) expression vectors were obtained. The gRNA libraries were further transferred into E. coli Top10 and cultured on LB plates containing 100 μg/ml ampicillin. Extract the successfully transformed E. coli plasmid. The gRNA library was transferred to strain yDYW000 and cultured at 30°C in SC-Ura medium.

The fragments to be integrated on the S. cerevisiae genome in this study were transferred into S. cerevisiae cells by electrotransformation as follows. The yeast was transferred from the overnight culture to YPD liquid medium at a volume ratio of 1:10 (depending on the amount of transformation required) and incubated at 30°C for 4–5 h in a 200 rpm shaker. 1 ml of the bacterial solution was divided into 1.5 ml EP tubes and centrifuged at 5,000 rpm for 2 min. Resuspend the cells by adding 1 ml of distilled water to the EP tube, centrifuge at 5,000 rpm for 2 min, and discard the supernatant. Repeat. Add 40 μL of 1 M sorbitol solution to the EP tube, resuspend the cells by slow agitation with a gun tip, centrifuge at 5,000 rpm for 2 min, and discard the supernatant. Repeat. Resuspend with 50 μL of 1 M sorbitol solution. Add the fragments to be transformed to the EP tube at a molar ratio of 1:1 so that the amount of each fragment is not less than 1,000 ng. Mix well and add to a 2 mm electrotransformer cup, and after a 2,500 V shock, quickly add 1 ml of 1 M sorbitol medium to the electrotransformer cup, mix well by pipetting with a gun and aspirate into a clean EP tube. Incubate at 30°C for 1 h in a 200 rpm shaker. Centrifuge at 5,000 rpm for 2 min and discard the supernatant. Resuspend with 50 μL of 1 M sorbitol solution and spread evenly onto YPD solid medium with antibiotics and incubate at 30°C for 48 h.

Recombinant yeast strains were precultured in the corresponding selective medium for 2 days and then inoculated into the fresh synthetic media with an initial OD of 0.1. Mid-log phase yeast cells were diluted 5-fold in ddH2O and eGFP an mcherry fluorescence signals were measured at 485–528 nm and 587–610 nm, respectively, using a Tecan Infinite M1000 PRO multimode reader (Tecan Trading AG, Switzerland). The fluorescence intensity (relative fluorescence units; RFU) was normalized to cell density that was determined by measuring the absorbance at 600 nm using the same microplate reader.

β-Carotene producing strains with gRNAs were pre-cultured in SD-HIS-URA/G418 medium for approximately 2 days, inoculated into 20 mL of fresh medium with an initial OD600 of 0.1, and incubated for 5 days under aerobic conditions (30°C, 250 rpm). Y The yeast cells were collected, centrifuged at 13,000 g for 1 min, and the cell precipitate was resuspended in 1 ml of 3n HCl, boiled for 5 min and cooled in an ice bath for 5 min. The lysed cells were washed with ddH2O and resuspended in 400 μL of acetone to extract β-carotene. Cell debris was removed by centrifugation. The extraction step was repeated until the cell precipitate was white. Analyze the absorbance of the supernatant containing β-carotene at 454 nm.

gRNAs and crRNAs for CRISPRa/i were designed using the CRISPOR tool (http://crispor.tefor.net/crispor.pybatchId=OZeMIBqmOOy55UwylOOJ) and those with high targeting and off-target scores were selected. Since there was no targeting score and off-target score for Cpf1, the following criteria were considered. gC content between 35% and 65%, no polyT, no secondary structure.

Mid-log phase yeast cells were collected and used to determine the relative expression levels via qPCR. A total of 1 µg RNA samples were reverse transcribed into cDNA using a Transcriptor First Strand cDNA Synthesis Kit using an oligonucleotide dT primer (Roche, Indianapolis, Indiana, United States).

qRT PCR was performed using the 217 Applied Biosystems StepOnePlus Real-Time PCR System and the detaSYBR^® kit under the following reaction conditions: 95°C for 1 min, 45 cycles of 95°C for 15 s, 60°C for 15 s, 72°C 45 s, 95°C 10 s, 65°C 60 s, 97°C 1 s, 37°C 30 s. The standard internal reference gene is ACT1.

The statistical analysis was carried out using one-way analysis of variance followed by Duncan’s multiple comparison tests. p values < 0.05 were considered statistically significant. The results are presented as the mean ± the standard deviation (SD) for a replication of n = 3.

To investigate the quantified effects of promoter regulatory strength in the CRISPR/dCas9 system within S. cerevisiae, we selected two robust yeast promoters, FBA1 and TEF1, and employed mcherry as a reporter gene. Transcriptional regulator proteins VP64 (activator) and KRAB (inhibitor) were synthesized to create complexes with Sp-dCas9, serving as fusion proteins in dCas9-CRISPRa/i systems.

In total, we designed 16 gRNAs targeting distinct regions of the FBA1 and TEF1 promoters, with gRNA1-10 specific to FBA1 and gRNA11-16 dedicated to TEF1 (Figure 1A). The integration of gRNA scaffold MS2 with the activating effector VP64 yielded a 1 to 6.27-fold enhancement of mcherry gene expression (Figure 1B). Conversely, combining the MS2 gRNA scaffold with the repressor effector KRAB resulted in a 1 to 4.5-fold reduction of mcherry gene expression. In both scenarios, the most pronounced regulatory effects were observed in proximity to the TATA-box and CAAT-box elements (Figure 1C).

Conversely, combining the MS2 gRNA scaffold with the repressor effector KRAB resulted in a 1 to 4.5-fold reduction of mcherry gene expression. In both scenarios, the most pronounced regulatory effects were observed in proximity to the TATA-box and CAAT-box elements TATA-box, CAAT-box, and transcription start site (TSS). Notably, gRNA7 (-194) targeting TATA-box (TATA) of the FBA1 promoter, and gRNA16 (-138) targeting TATA-box (TAAAGATT) and TSS of TEF1 promoter display significant stronger regulation in mcherry expression. In addition, the higher regulation rates caused by gRNA4 (-426) targeting CAAT-box (CAAT) of FBA1 promoter and gRNA14 (-304) targeting CAAT-box (TGAT) of TEF1promoter, were attributed to their closer proximity to the cis-acting elements within the promoter core region (Figures 1B, C).

The crucial component within the CRISPR/dCas9 system, responsible for identifying target genes and regulating dCas9 protein targeting, is known as the guide RNA, or gRNA. In efforts to achieve precise transcriptional regulation, certain researchers have integrated RNA-binding proteins (RBPs) fused with various effector proteins into the guide RNA sequence (Zalatan et al., 2015). The establishment of RNA-protein complexes followed a systematic procedure. Initially, RNA scaffolds, either MS2 RNA scaffold or PP7 RNA scaffold, were incorporated into the four-loop and stem-loop regions of selected guide RNAs (Figure 2A). Subsequently, MCP (MS2 capsid protein) or PCP (PP7 capsid protein) was fused with different effector proteins, namely, VP64 for activation (CRISPRa) and KRAB for inhibition (CRISPRi), forming the dCas9-gRNA complex. The results indicated that the gRNA/MS2-MCP/VP64 complex demonstrated higher activation rates in transcriptional regulation compared to the gRNA/PP7-PCP/VP64 complex (as shown in gRNA4/MS2, gRNA7/MS2, gRNA12/MS2, gRNA13/MS2, gRNA14/MS2) (Figure 2B). Conversely, stronger inhibition rates were observed in the gRNA/PP7-PCP/KRAB complex compared to the gRNA/MS2-MCP/KRAB complex when considering the gRNA scaffold and effector protein combinations (shown in gRNA1/PP7, gRNA3/PP7, gRNA4/PP7, gRNA16/PP7) (Figure 2C).

To further enhance the regulatory potential of the CRISPR/dCas9 system, multiple effector proteins VP64-p65-Rta (VPR) and KRAB-MeCP2 (KM), were introduced. Both combinations exhibited more potent regulation rates compared to single effector protein. The multiple effector complex of gRNA/MS2-MCP/VPR significantly increased transcriptional activation up to 5.5-fold (shown in gRNA7) in contrast to the single effector complex of gRNA/MS2-MCP/VP64 (Figure 3A). Similarly, the complex of gRNA/PP7-PCP/KM increased transcriptional inhibition up to 2.6-fold (shown in gRNA7/PP7) compared tothe single effector complex of gRNA/PP7-PCP/KRAB (Figure 3B). Overall, multiplexed RNA-protein complexes in the CRISPR/dCas9 system yielded more efficient regulatory rates for both CRISPRa and CRISPRi. Interestingly, concerning the compatibility of gRNA scaffolds and effector proteins within the CRISPR/dCas9 system, we observed that the MS2-MCP pair exhibited better compatibility with activator proteins (VP64, p65, Rta), while the PP7-PCP pair demonstrated better compatibility with repressor protein (KRAB, MeCP2) in transcriptional regulations.

To investigate the regulatory effect of CRISPR/dCpf1 system, we designed four crRNAs targeting the yeast ERG9 promoter region based on an analysis of its transcription start site (TSS), TATA-box, and CAAT-box using CRISPR-direct software and RNA-fold software (Figure 4A). RT-qPCR results demonstrates efficient inhibition rates of ERG9 gene ranged from 54.2% (crRNA2) to 72.5% (crRNA4) (Figure 4B).

In the CRISPR/dCpf1 system, the short (∼20 bp) crRNA secondary structures often show only a single loop region, which is not conducive to constructing complicated RNA scaffold. As shown in Figure 4C, the DR scaffold loop region of the four selected crRNAs targeting the ERG9 promoter was mutated from UGUU to UAUG (ecrRNA). However, these mutations of the DR loops of crRNA1, crRNA2 and crRNA3 did not enhanced the inhibition rate in CRISPR/dCpf1 system. In fact, the inhibition rate of ecrRNA4 reduced from 72.5% to 47.8% when compared to that of crRNA4 (Figure 4D).

Therefore, several crRNA arrays were designed to further quantify the regulatory effect of ERG9 promoter (Figure 4E). The CRISPR/dCpf1 system can directly employ the Fn-dCpf1 protein to cleave crRNA arrays and subsequently utilize multiple crRNAs to simultaneously target multiple sites, enabling multiplex genome editing. All three multiplex crRNAs (crRNA1-4, crRNA1-3-4, and crRNA1-2-3-4) have displayed significantly higher transcriptional inhibitory rates (60%, 240%, and 530%) comparing with single crRNAs (Figure 4F).

The development of multifunctional CRISPRa/i systems has enabled versatile regulation using orthogonal dCas proteins. Orthogonal Cas proteins, sourced from different bacterial species, exclusively recognize their corresponding complementary RNAs and nd necessitate distinct PAM sequences for targeting. The dCas9 protein can only be guided by its complementary gRNAs and dCpf1 proteins can only respond to its complementary gRNAs, while dCpf1 proteins are responsive solely to their complementary crRNAs due to the recognition of distinct PAM sequences, with the dCas9 system recognizing cytosine-rich PAM sequences and the dCpf1 system recognizing thymine-rich PAM sequences. To assess the orthogonality of Sp-dCas9 and Fn-dCpf1 proteins, experiments were conducted in two separate yeast strains, namely, dCas9-HO and dCpf1-NT. These strains had the dCas9 gene integrated into the HO site of the genome and the dCpf1 gene integrated into the NTR1 site within the S. cerevisiae genome. Upon introducing complementary gRNAs and crRNAs targeting the FBA1 promoter into these yeast strains, the dCas9-bound gRNAs effectively induced transcriptional inhibition of the reporter gene eGFP, with a 54.6% repression rate for gRNA2 observed exclusively in the dCas9-HO strain, while the dCpf1-bound crRNAs efficiently mediated transcriptional inhibition of the reporter gene eGFP, yielding a 72.5% repression rate for crRNA2, exclusively in the dCpf1-NT strain (Figures 5A, B).

Furthermore, to establish a CRISPR/dCas9-dCpf1 orthogonal system, an engineered yeast strain named dCas9-dCpf1-HN was created. This strain incorporated genes encoding dCas9 and mcherry, and dCpf1 and eGFP, inserted into either the HO or NTR1 site of the genome, respectively (Figure 5A). Plasmids containing both gRNA and crRNA sequences targeting the FBA1 promoter were introduced into the aforementioned three engineered yeast strains: dCas9-dCpf1-HN, dCas9-HO, and dCpf1-NT. The fluorescence results for the reporter genes, mcherry and eGFP (both driven by FBA1 promoters), exhibited clear orthogonality in regulatory patterns. Specifically, the dCas9-HO strain demonstrated 55.6% effective inhibition exclusively in mcherry (corresponding to gRNAm) and not in eGFP (corresponding to crRNAe), while the dCpf1-HO strain displayed 73.5% effective inhibition exclusively in eGFP (corresponding to crRNAe) and not in mcherry (corresponding to gRNAm). In the case of the dCas9-dCpf1-HN strain, it exhibited effective inhibition in both mcherry and eGFP, without crosstalk between the two dCas systems. Notably, the inhibition rates of the dCas9-dCpf1 co-expressing system (60.3% for eGFP, 78.6% for mcherry) closely mirrored those of the single Cas protein expressing systems (55.6%, 73.5%) (see Figure 5C). Thus, the CRISPR/dCas9-dCpf1 dual system has been successfully validated for gene regulation in S. cerevisiae.

To study the application of CRISPR multifunctional system in terpenoid yeast cell factories, the β-carotene pathway is commonly employed for validation (Naseri et al., 2019). The construction of the β-carotene metabolic pathway in the engineered dCas9-dCpf1-HN yeast strain necessitates the integration of three heterologous genes—CrtE, CrtI, and CrtYB—under the control of FBA1 and TEF1 promoters into the YPRC3 single-copy locus within the genome. This integration forms the basis for constructing a CRISPR/dCas9-dCpf1 bifunctional system. The newly developed β-carotene-producing yeast strain, named dCas9-dCpf1-Carotene, comprises two integral modules: the CRISPRa/dCas9 module and the CRISPRi/dCpf1 module (Figures 6A, B). The CRISPRa/dCas9 module operates in conjunction with a library containing 136 plasmids (CP1 to CP136), each harboring distinct combinations of gRNA-protein complexes designed to finely regulate the expression of crtE, crtI, and crtYB genes. The multiplexed combinations within the plasmid library are detailed in Table 1, featuring the previously mentioned 16 gRNAs, each possessing varying regulatory intensities and targeting the FBA1 and TEF1 promoters. These gRNAs are customized with either MS2 or PP7 RNA scaffolds, and the RNA-binding proteins MCP and PCP are fused with activating effector proteins such as VP64, VP64-p65-Rta (VPR), or repressing effector proteins such as KRAB, KRAB-MeCP2 (KM). In the CRISPRi/dCpf1 module, a small library of crRNA arrays was constructed to downregulate the ERG9 gene expression in the competing pathway. Each module can be activated through the application of the above stated plasmid libraries.

The highest β-carotene yield, quantified via light absorbance at 454 nm, was observed in the complex of gRNA6-gRNA14/MS2-MCP/VP64-P65-Rta, targeting the heterologous pathway within the CRISPRa/dCas9 module, and the complex of crRNA1-2-3-4 targeting the endogenous pathway within the CRISPRi/dCpf1 module. This resulted in a production yield 85% higher than that of the control. In comparison to individual monofunctional systems, the CRISPR/dCas9-dCpf1 dual system exhibited a 45% increase in efficiency over the dCas9 system alone and a 65% increase in efficiency over the dCpf1 system alone. Thus, we have successfully established a CRISPR/dCas9-dCpf1 bifunctional system for precise regulation of the heterologous β-carotene pathway and the competing endogenous ERG9 pathway (Figure 6C). In contrast to monofunctional CRISPR/dCas9 systems, the CRISPR/dCas9-dCpf1 dual system not only reduces crosstalk resulting from simultaneous activation and inhibition by gRNAs but also enables modular and quantitative metabolic engineering within S. cerevisiae.