A
A hybrid system enables plasmid copy number control in yeastA
A
A
Anni
Li
1
Qingyang
Zhao
1
Zhunyi
Yang
1
Jiesheng
Cheng
1
Dong
Xu
1
Yueyao
Zhang
1
Ting
He
1
Bingzhao
Zhuo
1
Xing
Zhao
1
Xiaolin
Rao
1
Hui
Wang
1
Lizhu
Chen
1
Zhouqing
Luo
1,2✉
Emailluozq@xmu.eu.cn
1
State Key Laboratory of Cellular Stress Biology, State-Province Joint Engineering Research Center of Targeted Drugs from Natural Products, School of Life Sciences, Faculty of Medicine and Life Sciences
Xiamen University
361102
Xiamen
Fujian
China
2
Xiang’an Hospital of Xiamen University, Xiamen University
Xiamen
Fujian
China
Anni Li1,#, Qingyang Zhao1,#, Zhunyi Yang1,#, Jiesheng Cheng1, Dong Xu1,Yueyao Zhang1, Ting He1, Bingzhao Zhuo1, Xing Zhao1, Xiaolin Rao1, Hui Wang1, Lizhu Chen1, Zhouqing Luo1,2,*.
1State Key Laboratory of Cellular Stress Biology, State-Province Joint Engineering Research Center of Targeted Drugs from Natural Products, School of Life Sciences, Faculty of Medicine and Life Sciences, Xiamen University, Xiamen, Fujian, 361102, China.
2Xiang'an Hospital of Xiamen University, Xiamen University, Xiamen, Fujian, China.
*Correspondence: luozq@xmu.eu.cn (Zhouqing Luo)
A
Anni Li, Qingyang Zhao and Zhunyi Yang Co-first author.A
Abstract
A
Synthetic biology requires plasmid systems that offer not only stable, fixed copy numbers but also tunable copy numbers to enable multi-level regulation of gene expression. While dynamic plasmid copy number (PCN) control has been engineered in prokaryotes such as E. coli, a parallel capability has been lacking for eukaryotic systems. Here, we bridge this gap by developing a programmable PCN platform for S. cerevisiae based on its endogenous 2µ plasmid. First, we engineered a p2µ-Cir0 system that exhibits a superior combination of high copy number (up to 20 per cell), enhanced population homogeneity, and improved segregation stability compared to conventional yeast episomal plasmids (YEps). This system supports protein expression levels up to 60-fold higher than a single chromosomal integrant. Introduction of a CEN element into p2µ enabled the construction of programmable YTp-C and YTp-I systems, which allow temporal PCN control with enhanced stability. These switchable vectors enable efficient PCN transition from 1 to 38 through time-dependent induction. Further incorporation of Leu2d-mediated metabolic selection in the YTp-CL and YTp-IL elevated the PCN to nearly 70 copies and boosted expression capacity to approximately 110-fold relative to chromosomal integration. We demonstrated the versatility of this platform through diverse applications, demonstrating that PCN elevation facilitated phenotyping of tRNA overexpression, enhancing the production of several compounds, including the therapeutic peptide GLP-1 precursor, 2-phenylethanol, β-carotenoid, and ergothioneine. These results establish the first quantitative and multi-dimensional PCN regulation toolkit for yeast, addressing the long-standing issue of instability arising from multiple copies and providing critical insights for synthetic biology that integrates gene dosage control across DNA, RNA, and protein levels.A
Introduction
Saccharomyces cerevisiae (S. cerevisiae) and Escherichia coli (E. coli) are predominant chassis organisms in fundamental research and industrial biotechnology. Plasmids are indispensable in both systems for DNA manipulation, genome engineering, and controlling gene expression levels1–3. While E. coli benefits from a versatile plasmid toolkit with replicons covering a wide copy number range4,5, the options for S. cerevisiae remain limited. Conventional yeast episomal plasmids are primarily of two types6: Yeast Centromere plasmids (YCps), which contain a centromere (CEN) and an autonomously replicating sequence (ARS) and maintain low copy numbers (1–5 copies/cell) with high segregation stability but often provide insufficient gene dosage for high-level expression7,8; and Yeast Episomal plasmids (YEps) harbor a 2µ origin, which replicate at higher copy numbers (~ 20 copies/cell) but suffer from greater instability during long-term cultivation9. Plasmid copy number (PCN) and stability are thus critical yet under-optimized parameters for advanced yeast engineering10,11. Expanding the repertoire of yeast plasmids with tunable copy numbers and enhanced stability is therefore essential to unlock more complex and rigorous synthetic biology applications4.
Dynamic control of PCN, which is highly compatible with regulatory layers at transcriptional, translational and post-translational stages12,13, overcome the limitation of canonical plasmids with fixed PCN, constituting a powerful and modular strategy for flexible regulation of protein expression, metabolic engineering and synthetic gene circuits4,14,15. So far, dynamic PCN control has only been achieved by modulating the RNA-p/RNA-i balance for ColE1 plasmid or Rep proteins for pSC101 plasmid, which are both designed for E. coli 16,17. By aTc mediated full induction of RNA-p, the PCN of ColE1 could increase from 1.4 to roughly 50 copies per cell16. A profound understanding of the replication regulation mechanisms of ColE1 and pSC101 plasmids is essential for the success of the aforementioned dynamic regulatory work. Developing dynamic PCN control strategies in microorganisms beyond E. coli, especially in eukaryotic organisms like S. cerevisiae, has become the next frontier subject in synthetic biology.
Although the Flp-based PCN amplification mechanism of the endogenous yeast 2µ plasmid was well characterized18–20, no work on dynamic PCN regulation based on such a mechanism has been reported thus far. Most existing efforts have focused on modifying the 2µ derived YEps. For instance, truncating the marker gene promoter on YEps or increasing antibiotic concentrations enables a ~ 20-fold increase in PCN, but careful engineering of promoter length and antibiotic concentrations is needed to achieve the best performance21. However, YEps rely on endogenous 2µ for replication and segregation but compete with it for intracellular resources, leading to instability akin to plasmid incompatibility observed in prokaryotes that will lead to loss of these engineered YEps as few as 20 generations22,23. Meanwhile, the large-scale industrial use of antibiotics is neither permissible nor cost-effective24. This underscores the absence of an effective, inducible regulatory switch for yeast plasmids comparable to those available in prokaryotic systems4.
To bridge this technological gap, we developed a novel switchable-copy-number plasmid system with enhanced stability and elevated PCN. We first engineered a high-copy-number vector with improved stability based on the endogenous 2µ plasmid of S. cerevisiae. The incorporation of a centromeric (CEN) element then created a foundation for inducible copy number switching. Further optimization via marker promoter truncation enabled the generation of the YTp-L variant, which achieved a higher PCN. Using fluorescence quantified by flow cytometry, we demonstrated that our system enables programmable transitions across a wide copy number range (1–90 copies/cell), while exhibiting significantly greater population stability than conventional YEp vectors. As illustrated, we present a switchable genetic tool for S.cerevisiae that integrates high copy number with robust stability, which facilitates fundamental investigations into plasmid maintenance and provides a quantitative means to correlate gene dosage with transcriptional, translational, and metabolic outputs for synthetic biology.
RESULTS
Engineering the p2µ-
Cir
0
system to enhance PCN and stability
A
A
A uniform plasmid background could be conducive to the PCN and stability engineering5. To achieve this, it is necessary to eliminate the endogenous 2µ plasmid which is rarely lost naturally (loss frequency < 0.01% per generation25). Thus, we employed CRISPR-Cas9 to excise the endogenous 2µ ori, generating a 2µ-free host strain (Cir0) (Fig. 1a and Fig. S1). The resulting Cir0 strains of both mating types (a and α) exhibited growth profiles comparable to the wild type (Fig. 1b), consistent with previous reports26.Next, we constructed a shuttle vector, p2µ, by engineering the endogenous 2µ sequence. The engineered plasmid carries elements for propagation in E. coli and a HIS3 selectable marker for yeast (Fig. S2). Then p2µ was transformed into both Cir0 and Cir⁺ strains to assess its PCN and segregation stability. Interestingly, the transformation efficiency of p2µ was approximately threefold higher in Cir+ than in Cir0 strains (Fig. S3), though absolute efficiency in Cir0 remained robust. We hypothesize that this difference stems from the time required for p2µ to establish its self-encoded replication and segregation machinery (Flp, Rep1, Rep2, and Raf1) in Cir0 cells, whereas these proteins are already present in Cir⁺ cells27.
Furthermore, we employed the mCherry fluorescence to characterize p2µ, in comparison to YCp type plasmid pRS413 and the YEp type plasmid pRS423. The distinct color contrast of mCherry under blue light illumination enabled direct visual differentiation of strains harboring various plasmids, with p2µ-Cir0 exhibiting the highest intensity (Fig. 1c). To quantitatively assess PCN, we employed both quantitative PCR (qPCR) and flow cytometric measurement, normalizing all values to a genomically integrated single-copy reference. Our analysis revealed a proportional relationship between the flow cytometry and the qPCR measurements, with a consistent ratio of approximately 3:1 (Fig. 1d). This discrepancy suggests a higher transcriptional output per DNA copy for episomal plasmids compared to the chromosomally integrated reference. Notably, the PCN of the pRS423-mCherry dropped significantly from the reported about 20 to just 8.8 (Fig. S4), highlighting the substantial metabolic burden imposed by heterologous gene expression on the PCN. In contrast, our p2µ vector maintained a higher copy number. This was particularly enhanced in the Cir0 strain (≈ 17 copies), likely due to the absence of competition with endogenous 2µ plasmid.
Flow cytometric analysis of population heterogeneity demonstrated that the p2µ-Cir0 system exhibited a markedly tighter fluorescence distribution (Coefficient of Variation, CV = 15%) than both p2µ-Cir+ (CV = 18%) and the pRS423 control (CV = 23%) (Fig. 1e and S5). To systematically evaluate plasmid stability, all strains were subjected to serial passaging in non-selective medium, with plasmid retention rates quantified at intervals (Fig. S6). Even under initial selection, p2µ-Cir0 and pRS413 exhibited a similar proportion of His+ cells (approximately 80%), while the proportion for p2µ-Cir+ was slightly lower than this value, but pRS423 had a His+ proportion of less than 50% (Fig. 1f). This was consistent with the observation that a population of non-fluorescent cells was detected in the pRS423-mCherry group shown in Fig. 1e, indicating that even in the selective medium, a substantial number of cells had lost their plasmids. After 4 rounds of serial transfers, p2µ in Cir0 and Cir+ demonstrated significantly higher retention frequencies of 14% and 10%, respectively, markedly outperforming pRS423, which declined to merely 1.3%. These results are consistent with the conclusion on plasmid stability determined by flow cytometry (Fig. S7).
Collectively, our p2µ plasmid, particularly in the Cir0 background, combines higher copy number (up to ~ 17 copies per cell), enhanced population homogeneity, and improved segregation stability, surpassing conventional YEps. It supports protein expression levels up to 56-fold higher than a single chromosomal integrant, providing a robust genetic tool for S. cerevisiae.
Fig. 1
The p2µ-Cir0 system with enhanced PCN and stability.
a.
a. Strategy for generating 2µ-free (Cir0) strains. A CRISPR/Cas9 system expressing two sgRNAs targeting sequences flanking the 2µ origin was used to excise the endogenous plasmid from wild-type S. cerevisiae (Cir⁺), yielding isogenic Cir0 strains.
b.
b. Growth phenotypes of Cir+ and Cir0 strains. Serial dilutions of strains with different mating types (a or α) and 2µ status (Cir+ or Cir0) were spotted on YPD agar. All strains showed comparable growth after 36 h at 30°C.
c.
c. Visual assessment of plasmid copy number (PCN) via mCherry fluorescence. Cell suspensions of equal density, harboring the indicated mCherry expression constructs, were imaged under 480 nm blue light. From left: wild-type (no plasmid), chromosomal single-copy integrant, YCp-pRS413, YEp-pRS423, p2µ in Cir⁺, and p2µ in Cir0. Color intensity correlates with PCN.
d.
d. Quantitative correlation between PCN and fluorescence. PCN determined by qPCR strongly correlated (R² = 0.987) with normalized fluorescence intensity measured by flow cytometry. Data are mean ± SD (n = 3 biological replicates).
e.
e. Single-cell expression homogeneity. Flow cytometry analysis of the plasmid-bearing subpopulation revealed that p2µ in the Cir0 background produced the most uniform fluorescence distribution (lowest coefficient of variation), indicating superior population homogeneity compared to other constructs.
f.
f. Plasmid segregation stability. Plasmid retention rates were monitored by serially passaging strains in non-selective medium and quantifying the fraction of His⁺ cells. p2µ maintained significantly higher retention than the conventional YEp-pRS423 after four transfers. Data are mean ± SD (n = 3).
A hybrid plasmid system enabling switchable copy number control
The superior stability of p2µ-Cir0 provides an ideal foundation for developing dynamic PCN control in yeast. An optimal system should maintain a true single-copy state before induction, then enable robust amplification via a specific molecular switch.
We first questioned whether a dynamic regulatory system could be established using YEps. We engineered a derivative of the YEp pRS425 by integrating a CEN/ARS element to enforce low-copy propagation, as centromeric function dominantly suppresses 2µ-mediated replication28. The transition to multicopy propagation was designed by inactivating the centromere through inducible transcriptional interference mediated by the Pgal1 promoter, a method previously used for aneuploidy construction29. However, this construct (Pgal1-CEN-2µ ori) failed to function as intended. It maintained a basal copy number of ~ 3 before induction and showed no significant amplification thereafter (Fig. S8). It is reasoned that the 2µ system retains partial functionality before induction, leaving the plasmid in a mixed mode of replication and segregation, but then the endogenous 2µ plasmid lacks sufficient ability to fully amplify the copy number of YEps (with the 2µ origin only) when the centromere was transcriptionally inactivated.
To engineer a dynamically switchable system, we therefore turned our focus to the p2µ-Cir0 system. The amplification of the 2µ plasmid was mediated by Flp recombinase through a site-specific recombination mechanism, which is critical for maintaining high-copy maintenance30. To establish a stringent single-copy control, a CEN sequence flanked by I-SceI recognition sites was inserted into the FLP coding sequence, thereby disrupting the amplification machinery (Fig. S9a). Galactose-induced expression of I-SceI (integrated at the HO locus) was designed to excise the CEN cassette and restore a functional FLP gene, as verified by PCR (Fig. S9b). However, this construct failed to trigger PCN amplification upon induction.
Rep1 and Rep2, two transcriptional regulators encoded by and essential for the endogenous 2µ plasmid, form a complex that suppresses Flp31. We speculated that the prior expression of Rep1 and Rep2 during the single copy state could continually suppresses Flp expression, preventing the initial amplification of plasmid even after the formation of an intact FLP sequence. To overcome this, we replaced the native FLP promoter with the constitutive, medium-strength Pcyc1. This design served three purposes: to minimize transcriptional interference with the centromere before induction; to ensure robust expression of full-length FLP after induction; and to mitigate repression by the Rep1/Rep2 complex. Consequently, we called this system Yeast PCN-switchable plasmid (YTp-C, where C denotes constitutive FLP expression) (Fig. 2a). The YTp-C plasmid was maintained at a single copy in glucose medium and could be induced to ~ 38 copies per cell by galactose (Fig. 2c-d). We also tested the effect of promoter strength on system performance (Fig. S10). Replacement of the native promoter with either the minimal Pleu2d or the strong constitutive Ptdh3 resulted in suboptimal amplification profiles. Notably, the promoter TDH3-driven construct deregulated PCN even without induction, failing to maintain a tight single-copy baseline, likely due to interference with centromeric function.
To uncouple CEN excision from FLP expression but maintain a high Flp level, we integrated another copy of FLP driven by the inducible Pgals into the genome, which was named the YTp-I system (I denotes inducible expression of FLP) (Fig. 2b). Here, galactose induction simultaneously triggers I-SceI-mediated CEN excision and Flp expression, leading to replication derepression and an increase in PCN to ~ 22 copies (Fig. 2d). Similarly, we utilized mCherry fluorescence to measure the PCN and population heterogeneity by flow cytometry (Fig. 2d and 2e). Both YTp-C and YTp-I showed more centralized populations after induction than the conventional YEp pRS423, with PCN ranging from 1 to 105 and 67, respectively (Fig. 2d). The induced cultures showed a marked increase in red fluorescence under blue light (Fig. 2e).
We next assessed segregation stability by serial passaging in non-selective medium (Fig. 2f). Initially, YTp-C and YTp-I showed higher retention (84% and 73%, respectively) than pRS423 (45%). Despite higher copy numbers, the engineered YTp-C and YTp-I demonstrated greater stability, especially YTp-C, which retained 60% of its plasmids after four consecutive passages without selection (Fig. 2f). For YTp-C, nearly all cells were fluorescence-positive even after 8 consecutive passages without selection (Fig. S12). This stability surpassed that of the CEN/ARS plasmids, underscoring the system's capabilities in plasmid maintenance and its significant advantages for producing proteins and metabolite products.
The YTp-C and YTp-I systems present the programmable plasmid systems for yeast featuring temporally controlled PCN and superior stability. These switchable systems enable precise transition from single-copy to high-copy through time-dependent induction.
Fig. 2
Engineering and characterization of a switchable-copy-number plasmid system.
a.
Mechanism of the PCN switch of YTp-C system. Galactose-induced expression of I-SceI nuclease excises the CEN element. Concurrent constitutive Flp expression then triggers the transition from a single-copy to a multi-copy state.
b.
Genomic FLP dependency for robust amplification in YTp-I system. In this variant, which lacks constitutive FLP, the native FLP promoter is insufficient to drive amplification after CEN excision. Effective switching requires galactose-induced co-expression of a genomically integrated FLP gene.
c.
Experimental regimen for PCN amplification. A single transformant was subjected to a sequential induction protocol: initial growth in glucose-containing medium, a shift to raffinose medium to deplete residual glucose, followed by induction in galactose-containing medium to express I-SceI endonuclease and Flp.
d.
PCN quantification of induced PCN amplification. PCN was determined by both quantitative PCR and flow cytometry. Data points in the lower left quadrant represent the pre-induction copy numbers for both systems. Values are shown as the mean ± SD (n = 3).
e.
Single-cell homogeneity and visual phenotype. Flow cytometric analysis show the population-wide shift in fluorescence intensity (reflecting PCN) upon induction. Corresponding cultures of equal cell density were imaged under 480 nm blue light.
f.
Segregation stability under non-selective growth. Plasmid retention rates for the YTp systems were monitored over four serial passages in YPD medium by scoring auxotrophy (His⁺/Ura⁺). Data are mean ± SD (n = 3).
PCN can be super-elevated by Leu2d-mediated metabolic selection
To further increase PCN, we attempted to leverage extra driving forces for its improvement. Inspired by reports that truncating selection marker promoters can elevate YEps copy number21, we engineered a truncated LEU2 allele (designated Leu2d), shortening its promoter from 405 bp to 29 bp (Fig. 3a). This Leu2d allele was introduced into our basal “Pgal-CEN-2µ ori” system, whose PCN is not switchable (Fig. S13). The Leu2d significantly reduced the expression of Leu2p and thus the efficiency of leucine biosynthesis. Consequently, when cultured in leucine-deficient medium, strains with low plasmid copy number showed severely impaired growth, while high-copy-number variants proliferated robustly, enforcing plasmid maintenance through metabolic selection (Fig. 3b). This leucine auxotrophy-driven PCN amplification was a multi-generational process, gradually reaching a maximum of ~ 150 copies/cell (Fig. S13d). Notably, even at this high copy number, plasmid maintenance imposed no measurable burden on host growth under selective conditions (Fig. S13c). However, upon removal of selective pressure, PCN declined (Fig. S13d), indicating that the high-copy state is metabolically sustained.
Based on this metabolic selection principle, we engineered two advanced plasmid vectors, YTp-CL and YTp-IL (L is for Leu2d), by incorporating the Leu2d cassette into the YTp-C and YTp-I backbones, respectively (Fig. S14). These constructs enable programmable transitions across three distinct replication states (Fig. 3c): (i) a single-copy state maintained by the CEN element; (ii) a low-copy state achieved upon galactose-induced excision of CEN by I-SceI; and (iii) a high-copy state triggered by leucine deprivation. Quantitative analysis confirmed this programmable control (Fig. 3d). Under uninduced conditions, both plasmids maintained single-copy levels (1.3 and 1.0 copies/cell for YTp-CL and YTp-IL, respectively). Leucine deprivation drove robust amplification to 68 (YTp-CL) and 50 (YTp-IL) copies per cell. Correspondingly, revealed a dramatic increase in mCherry expression, exceeding chromosomally integrated controls by 109-fold (YTp-CL) and 101-fold (YTp-IL). Notably, the ratio of fluorescence intensity (flow cytometry) to gene copy number (qPCR) for YTp-CL decreased to approximately 1.6, suggesting a declining marginal contribution of each additional plasmid copy to the total protein output at very high copy numbers.
On the other hand, the high-copy-number populations exhibited markedly improved homogeneity, with coefficients of variation (CV) as low as 11% and 12% (Fig. S15). Quantitative stability assessment during serial passaging revealed distinct profiles: YTp-IL retention dropped sharply from 81.7% to 15.3%, whereas YTp-CL declined from 92.2% to 34.4%, showing relatively better maintenance (Fig. 3f). This reduced stability of both Leu2d-bearing constructs compared to their parental vectors (YTp-C and YTp-I) is likely attributable to the metabolic burden of sustained high-copy replication. Once selective pressure is alleviated, plasmid-free cells gain a pronounced growth advantage32, accelerating plasmid loss (Fig S16). However, both YTp-CL and YTp-IL still maintained significantly higher stability than conventional traditional YEps. Notably, under non-selective, single-copy conditions, the initial retention of YTp-CL and YTp-IL was comparable to that of their parents, confirming that the stability difference emerges specifically under high-copy, selective pressure.
The YTp-CL and YTp-IL system further elevate the PCN in yeast to nearly 70 and recombinant protein expression levels ~ 110-fold above genomic integration, providing robust dynamic regulatory tools for protein expression and other applications.
Fig. 3
PCN can be super-elevated by Leu2d-mediated metabolic selection.
a.
Mechanism of Leu2d-driven copy number amplification. Truncation of the native LEU2 promoter (409 bp to 29 bp) generates the Leu2d allele, drastically reducing leucine biosynthesis. This creates a conditional auxotrophy that ties plasmid maintenance to cellular viability.
b.
PCN-dependent growth under leucine selection. In SC-Leu medium, strains harboring single-copy plasmids undergo growth arrest due to insufficient leucine production, whereas high-copy-number variants outcompete through gene dosage compensation.
c.
Three-stage induction regimen for PCN switching. Transformants are maintained as single copy in SC-His medium. Transfer to galactose-containing medium triggers the shift to a low-copy state. Subsequent leucine deprivation induces amplification to the high-copy state.
d.
Quantification of copy number before and after induction. Plasmid copy numbers (PCN) for YTp-CL (green) and YTp-IL (purple) were determined by flow cytometry and qPCR. Values are shown as mean ± SD (n = 3 biological replicates).
e.
Characterization of mCherry fluorescence before and after induction via flow cytometry. The right panel shows corresponding fluorescent phenotypes under blue light irradiation, similar to that of Fig. 2e.
f.
Plasmid stability under non-selective growth. Retention of the YTp-L plasmid was monitored over four serial passages in YPD medium by scoring histidine prototrophy (His⁺). Data are mean ± sSD (n = 3 biological replicates).
Dynamic expression of RNA and protein
We sought to characterize the potential of our PCN-switchable system for phenotyping of multi-copy genes. tRNA genes represent a diverse and multicopy gene family, with their promoters encoded by internal tRNA sequences33. This characteristic renders it highly challenging to modulate specific tRNA levels through conventional promoter engineering34. Thus, we exploited the YTp-CL system to overexpress all valine-coding tRNAs by amplifying their gene dosage. The valine in yeast is decoded by three tRNA genes: tV(CAC), tV(TAC), and tV(AAC), with native genomic copy numbers of 2, 2, and 14, respectively35. As outlined in Fig. 4a, each gene was cloned into YTp-CL vector and transformed into a Pgal1-I-SceI host. Under uninduced conditions, the PCNs of tRNA-bearing plasmids remained comparable to those in the empty vector control (Fig. S17). Following the induction procedure in Fig. 3c, elevated PCN led to markedly increased tRNAVal transcript levels. The PCNs increased by 27-, 37-, and 64-fold for tV(CAC), tV(TAC), and tV(AAC), respectively. The corresponding transcript levels rose by 32-, 10-, and 11-fold compared to the uninduced state (Fig. 4b). This discrepancy between DNA copy number amplification and transcript accumulation suggests potential differences in transcriptional efficiency, post-transcriptional regulation, or cellular tolerance among these tRNA isoacceptors, although they decode the same amino acid. Notably, such pronounced increases did not impair growth under optimal conditions (Fig. 4c, lower panel). However, under nocodazole stress, strains overexpressing different tRNAVal isoforms exhibited divergent phenotypes: the tV(CAC)-overexpressing strain slightly outperformed the vector control, whereas tV(TAC) was severely impaired (Fig. 4c, upper panel). These results demonstrate that our PCN-switchable system is a powerful tool for revealing functional specialization within tRNA families.
We further extended the application of our system to produce the therapeutically relevant peptide glucagon-like peptide-1 (9–37) [GLP-1(9–37)], a direct precursor to the long-acting antidiabetic drug semaglutide36. To facilitate detection and purification, we constructed a fusion protein consisting of an N-terminal hexahistidine (6X His) tag, mCherry fluorescent protein, and the target peptide, with these domains separated by an enterokinase-cleavable DDDDK site (Fig. 4d). This expression cassette was introduced into a protease-deficient yeast strain (yps1Δ pep4Δ)37, and the intact peptide product was verified by LC-MS analysis (Fig. S18). Induction of the YTp-CL system substantially enhanced both PCN and peptide yield, leading to 22-fold and 42-fold increases, respectively (Fig. 4e). To further limit proteolytic degradation, we engineered a host strain by additionally deleting the protease gene PRB138. This modification boosted peptide accumulation to 119-fold (Fig. 4e). Absolute quantification showed that the recombinant peptide constituted 6.8% and 11.2% of the total cellular protein in the respective host strains (Fig. 4f). Following 48-hour cultivation in a 5-L fermenter and subsequent nickel-affinity purification, the final target protein concentration was determined to be 373 mg/L by BCA assay (Fig. S19a). This yield surpasses those reported for previous yeast expression systems even after normalization to cell mass or culture volume39,40.
Together, this work demonstrates a dual-purpose tunable plasmid copy number system for functional genomics and bioproduction. By combining dynamic gene dosage control with tailored host engineering, the platform effectively addresses potential transcriptional/translational limitations, proving its value for both fundamental genetic studies and applied metabolic engineering.
Fig. 4
The YTp system accelerates the design and optimizing production of genetic modules.
A
a༎
Valine-tRNA expression module in the YTp-CL vector. The module harbors the three valine-coding tRNA genes—tV(CAC), tV(TAC), and tV(AAC)—with native copy numbers of 2, 2, and 14, respectively. This design enables programmable switching between distinct overexpression states within a single strain.
b༎
Induction-fold changes in tRNA gene copy number in plasmid and transcript abundance. Changes were quantified by qPCR before and after induction. Data are mean ± SD of two biological replicates, each with three technical replicates.
c༎
Growth phenotypes of tRNA-overexpressing strains. Strains were assayed by serial dilution spot assays on SC-His (low-copy state) or SC-Leu (high-copy state) medium. Differential growth under 5 µg/mL nocodazole stress at 30°C for 48 hours is shown relative to the vector control.
d༎
Design of the GLP-1(9–37) expression construct in the YTp-CL vector. The target peptide is fused to an N-terminal 6×His tag and mCherry, separated by an enterokinase-cleavable DDDDK site to enable tag removal and subsequent chemical modification.
e༎
Induction-fold changes in plasmid copy number and target protein yield. Data for the yps1Δ pep4Δ (left) and yps1Δ pep4Δ prb1Δ (right) strains. Protein yield fold change (YFC) was determined by western blot, normalized to β-tubulin. Data are mean ± SD of three biological replicates.
f༎
Target protein proportion in total cellular protein. The fraction of mCherry-GLP-1(9–37) fusion protein after induction in the yps1Δ pep4Δ and yps1Δ pep4Δ prb1Δ strains. Data are mean ± SD of three biological replicates.
YTp system enhances the production of diverse metabolites
To demonstrate the broad utility of the YTp platform in synthetic metabolism, we applied it to produce three high-value compounds with distinct biochemical origins: 2-phenylethanol (a rose-scented monoterpenoid), β-carotene (a tetraterpenoid pigment), and ergothioneine (a sulfur-containing amino acid derivative). The engineered biosynthetic pathways for these compounds are schematically summarized in Fig. 5a.
2-Phenylethanol (2-PE) is a valuable rose-scented aroma compound widely used in fragrance and cosmetics, its global market was valued at USD 33.58 billion in 202541. In yeast, 2-PE is biosynthesized from L-phenylalanine via the Ehrlich pathway, a three step conversion catalyzed by phenylalanine transaminase (encoded by ARO8 and ARO9), phenylpyruvate decarboxylase (ARO10), and alcohol dehydrogenases (such as ADH2-5) (Fig. 5a)42. To engineer 2-PE production, we constructed the YTp-CL vector harboring ARO9 and ARO10 driven by PGK1 and TDH3 promoters. In uninduced conditions without leucine selection, the system maintained a baseline PCN of 0.4 copies/cell, producing 40.2 mg/L 2-PE in shake-flask culture (Fig. 5b). Galactose induction coupled with leucine selection pressure dramatically increased the PCN to 27.5 copies/cell (68.75-fold increase), which drove a concomitant rise in 2-PE titer to 840.8 mg/L─a 20.9-fold enhancement over the baseline (Fig. 5b, S20).
Next, we investigated the biosynthesis of the complex tetraterpenoid β-carotene. Codon-optimized variants of yeast CrtE, CrtYB, and CrtI were integrated into the pRS425 vector and transformed into the host strain ANY05143,44. The resulting transformants exhibited pronounced colony color heterogeneity, only a subset showed the yellow pigmentation characteristic of β-carotene accumulation (Fig. S21a), This result indicates that conventional YEps are inefficient for maintaining such a metabolically burdensome heterologous pathway. To address this, we deployed the same biosynthetic cassette using the YTp-CL system, generating strain ANY693. This strain showed markedly improved phenotypic homogeneity compared to the pRS425-based control (Fig. S21b). In the YTp-CL system, the plasmid maintained a basal copy number of 0.34 prior to induction. Upon induction, copy number increased 32-fold to 10.7, which drove a 9-fold enhancement in β-carotene titer, reaching 17.1 mg/L (Fig. 5c). This comparatively modest enhancement in both plasmid copy number and product titer, when compared with the 2-PE pathway, is likely attributable to the greater metabolic burden imposed by the β-carotene biosynthetic pathway on the host cell. This comparatively modest enhancement in both PCN and titers than 2-PE biosynthesis, is likely attributable to a greater metabolic burden imposed by β-carotene biosynthetic pathway on the host cell45.
Further, we engineered the biosynthesis of ergothioneine (EGT), a high-value thiol-histidine derivative with notable antioxidant and cytoprotective activities46. The NcEGT1 and CpEGT2 genes were assembled into the YTp-CL system47, yielding strain ANY694, which produced 9.2 mg/L EGT under uninduced conditions. Upon induction, the PCN increased by 35-fold to 15 copies, yet EGT titers rose only 1.9-fold to 17.5 mg/L (Fig. 5d). This disproportionate response implies that pathway flux is limited primarily by precursor availability, notably L-histidine and S-adenosylmethionine (SAM)48. Combinatorial engineering of precursor supply modules may therefore be necessary to fully exploit the production capacity of the YTp system.
Collectively, these results demonstrate that the YTp system enables a dynamically tunable strategy to enhance metabolic flux across diverse biochemical pathways through programmable copy number control. The differential yield responses reveal the complex, non-linear coupling between gene dosage, enzyme abundance, and precursor availability within reconstituted pathways. Therefore, the YTp platform serves not only as a tunable expression tool but also as a probe for flux distribution. By mapping production outputs to specific gene dosage inputs, it helps identify and rank rate-limiting steps in complex biosynthetic networks, thereby advancing the rational design of yeast cell factories for natural product synthesis.
Fig. 5
Enhanced diverse metabolites production via dynamic gene dosage amplification.
a. Biosynthetic pathways for 2-phenylethanol (2-PE), β-carotene, and ergothioneine (EGT) in S. cerevisiae. The representative products are (b) 2-PE (a monoterpenoid), (c) β-carotene (a tetraterpenoid), and (d) EGT (an amino acid derivative). The plasmid copy number (PCN, red circles) and product titer (mg/L, solid columns) are shown on the left and right y-axes, respectively. Strains were cultured in SC-His medium (pre-inductio) or SC-Leu medium (post-induction) for 48 h (2-PE, EGT) or 72 h (β-carotene). L-phe, L-phenylalanine; PPY, Phenylpyruvate; PAA, Phenylacetaldehyde; FPP, farnesyl diphosphate; GGPP, geranylgeranyl diphosphate; SAM, S-adenosylmethionine. Data are represented as mean ± SD (n = 3).
Discussion
The endogenous 2µ plasmid of S. cerevisiae is a natural high-copy genetic element. In this study, we engineered it into the p2µ-Cir0 backbone and its switchable derivatives (YTp-C, YTp-I, YTp-CL, YTp-IL). These systems provide a convenient and powerful platform for bioproduction, enabling over 100-fold enhancements in recombinant protein expression and up to several dozenfold increases in the titers of high-value metabolites. Notably, the protein output per plasmid copy in our systems significantly exceeds that of a single genomic integrant. When combined with the enhanced stability demonstrated by YTp-C, plasmid-based expression offers superior efficiency, convenience, and scalability. Thus, our work establishes a modular platform for high-level gene expression with broad utility, poised to assist in functional phenotyping, improve therapeutic protein production, and boost metabolic yields in industrial settings.
Dynamic control of gene expression is a fundamental challenge in eukaryotic synthetic biology, as current strategies inherently limited by competing constraints. Inducible promoters (e.g., Pgal1) couple expression control to specific carbon sources that compromise host fitness49; RNA interference provides reversible regulation but lacks genetic stability50; and genomic multi-copy integrations offer stability at the cost of dynamic tunability51. To address these limitations, we present YTp (Yeast switchable Plasmid), the first programmable plasmid combining galactose-inducible PCN amplification with the native stability of endogeneous 2µ backbone, ensuring robust and homogeneous plasmid maintenance across populations. This system provides a versatile DNA-copy-level tool that is capable of integrating regulatory inputs from transcription, translation, and post-translational modification levels.
Our work demonstrates that introducing metabolic pathways triggers dynamic oscillations in PCN within yeast cells, which ultimately stabilizes at a threshold level dictated by the host’s metabolic capacity. This self-regulatory behavior represents a cellular trade-off between heterologous protein overexpression and basal metabolic demands: once recombinant protein synthesis exceeds a critical threshold, the host downregulates plasmid replication through native homeostatic networks. These findings facilitates the identification of the threshold of wild-type cells and promotes the engineering of adaptive networks to support higher yields. This principle can be leveraged to engineer intelligent bio-manufacturing platforms with built-in homeostatic regulation, enabling self-tuning of gene dosage in response to real-time metabolic states.
In this study, galactose-induced formation of the Flp recombinase effectively amplified the copy number of the 2µ plasmid. The two engineered systems, YTp-C and YTp-I, differ solely in Flp levels yet exhibit markedly distinct PCNs and genetic stabilities, indicating that Flp level is a critical determinant for regulating both PCN and stability of the 2µ plasmid. Notably, the native 2µ plasmid maintains its high copy number and robust stability not only through Flp-mediated site-specific recombination but also via a highly optimized regulatory network involving Rep1, Rep2, and Raf126. Therefore, systematic engineering of these four proteins holds promise for programmable and precise control over 2µ PCN and stability, thereby introducing a new layer of regulation for synthetic biology chassis platforms based on the 2µ backbone.
Although the YTp system developed in this study has been successfully implemented only in S. cerevisiae, its core engineering principles hold broader applicability. The system utilizes complete replication and regulatory elements of the endogenous 2µ plasmid, which exhibit high functional conservation across the yeast kingdom. This offers a particularly promising solution for non-model yeasts—such as the probiotic Saccharomyces boulardii, the industrial host Pichia pastoris, or the pathogen Candida albicans—which often suffer from inefficient genetic tools, challenging genome editing, and a lack of stable, high-copy expression vectors52–54. Our work offers a novel strategy for these genetically recalcitrant yeasts: establishing a high-expression heterologous gene platform via straightforward transformation rather than complex genomic integration. Furthermore, with the chromosomal tethering of Rep1-Rep2 complexes55 and the FLP/FRT system's broad eukaryotic utility56, the 2µ plasmid system shows significant promise across eukaryotes.
Method
Strains and growth condition
The yeast strain BY4741 and BY4742, which are derived from S288C, were used as the host in this study. Based on them, a series of strains was constructed, which are listed in Supplementary Table 1.
For routine growth, yeast were cultured at 30°C in YPD medium (20 g/L glucose, 20 g/L peptone, 10 g/L yeast extract) for routine growth. Transformant selection and PCN induction, synthetic complete (SC) medium was used, which was deficient in uracil, methionine, lysine, leucine, histidine, adenine and trytophan (SC-Ura-Met-Lys-Leu-His-Ade-Trp) with appropriate supplements added as required by specific selection markers. The carbon sources in SC medium was adjusted based on experimental needs: 20 g/L glucose for standard cultivation, 20 g/L galactose to induce GAL1 and GALs promoter-driven expression, and 20 g/L raffinose to prevent the repression of galactose-inducible promoters by glucose. All liquid yeast cultures were incubated at 30°C with orbital shaking at 220 rpm.
Escherichia coli strains (DH5α and XL-10) used for plasmid construction and propagation were grown in LB medium (20 g/L bacto-peptone, 10 g/L yeast extract, 20 g/L NaCl) at 37°C with shaking at 220 rpm. Transformed E. coli was selected using LB medium supplemented with 100 mg/L carbenicillin. All plasmids used in this study are listed in Supplementary Table 2.
Construction of Cir0 strains using CRISPR/Cas9
The Cas9 plasmid was construct based on the pRS416 backbone, withCas9 expression driven by the GAL10 promoter and terminated by the CYC1 terminator. Two sgRNAs (with N20 sequences 5’-GGAAGACAATGTATGTATTT-3’ and 5’-GAAAATCACGTAATACTTCT-3’) were designed to target the sites flanking the 2µ ori. This plasmid was transformed into BY4741 and BY4742. Single colonies from the transformation plates were first pre-cultured in SC-Ura medium containing raffinose, then induced in SC-Ura medium containing galactose. Cells were subsequently plated onto selective medium (SC-Ura). Single colonies were screened by colony PCR, and those confirmed to have lost the endogenous 2µ plasmid were designated as Cir⁰ strains (Fig. S1).
Flow cytometry protocol
Overnight yeast cultures were diluted to an initial OD₆₀₀ of 0.1 in 3 mL of fresh medium and incubated at 30°C with shaking at 220 rpm for 8 hours. Cells were harvested by centrifugation at 4,200 rpm for 1 minute, washed twice with 800 µL of 50 mM sodium citrate buffer (pH 4.5), and finally resuspended in the same buffer to maintain cellular integrity during analysis.
Fluorescence intensity measurement was performed by Flow cytometry using a Quanteon cell analyzer (ACEA Bioscience) equipped with an automatic sampler. The mCherry fluorescence was excited with a 561 nm laser and detected using a 610/20 nm bandpass filter.. For each sample, ≥ 10,000 events were recorded. Data were analyzed using FlowJo software.
Plasmid copy number was calculated as the ratio of mCherry MFI in engineered strains to that of a reference strain harboring a single chromosomal copy of mCherry (denoted as the "one-copy reference"). This normalization accounts for variations in cellular physiology and instrument settings between experiments. All measurements were performed in biological triplicates.
qPCR assay for the plasmid copy number
Plasmid copy numbers were quantified by quantitative PCR (qPCR) using the total yeast DNA extracts based on the phenol-chloroform method. The yeast total DNA was diluted to 100 µg/mL and used for qPCR analysis. Genomic and reporter plasmid sequences were amplified using primers targeting the genomic ALG9 region (for: 5’-CACGGATAGTGGCTTTGGTGAACAATTAC-3’ and rev: 5’-TATGATTATCTGGCAGCAGGAAAGAACTTGGG-3’), and plasmid AMP region (for: 5’-TGACTTGGTTGAGTACTCACC-3’ and rev: 5’-TGTCACGCTCGTCGTTTGGTA-3’), respectively. Copy numbers were quantified using the 2–ΔΔCT method.
Plasmid stability experiment
The stability of plasmid expression was evaluated by plate count. Specifically, the strain was transformed with the corresponding plasmids, a single colony was picked and cultured overnight in selective medium, marking today as the day 0. Cells were harvested at the mid-log phase (OD600 0.6 ~ 0.8), 8 hours after subculture into fresh medium. Approximately 200 cells were plated onto YPD medium, and single colonies were allowed to grow for 48 hours at 30°C. This plate was then replica plated to selective medium plates. The plasmid retention rate at day 0 was calculated as the ratio of colony-forming units (CFUs) on the selective medium to those on the YPD medium. To assess stability over time, the Day 0 liquid culture was subcultured into non-selective YPD medium at an OD600 of 0.1. This culture was then subjected to the same plating, replica-plating, and CFU counting procedure described above. This sequential subculturing and assay process was repeated daily, and plasmid retention rates were determined for Days 1, 2, 3, and 4.
tRNA extraction and quantification
Total RNA was isolated from yeast cells using the TRIzol-based phenol-chloroform method. The cell pellet was washed twice and resuspended in 150 µL of lysis buffer (containing 0.1 M lithium acetate and 0.5% SDS), followed by incubation at 70°C for 5 minutes. Subsequently, 450 µL of TRIzol LS reagent was added and vortexed for 15 seconds, followed by the addition of 150 µL chloroform and inversion mixing for 15 seconds. The mixture was incubated at room temperature for 5 minutes. The aqueous phase was collected by centrifugation at 12,000 rpm for 30 minutes, and RNA was precipitated by 450 µL isopropanol. The RNA pellet was collected by centrifugation at 10,000 rpm for 20 minutes and washed twice with 75% ethanol. Finally, the RNA was dissolved in 25 µL of RNAase-free water and immediately used for cDNA synthesis. RNA concentration and purity were determined by spectrophotometric measurement, with all samples exhibiting A260/A230 ratios > 2 and A260/A280 ratios > 2. cDNA synthesis was performed using a commercial kit (Novoprotein, Suzhou, China) according to the manufacturer's instructions.To ensure the isolated RNA was DNA-free, about 2 µg samples underwent DNaseI treatment at 37°C for 30 minutes. The DNase reaction was halted by adding 50 mM EDTA and heating the sample at 65°C for 10 minutes. Subsequently, RNA was reverse-transcribed using the RevertAid First Strand cDNA Synthesis Kit (Novoprotein).
Individual tRNAs were quantified by qPCR using 1 µL of 1:10 diluted cDNA template with gene-specific primers (Supplemental Table 3). Expression values were normalized to the geometric mean of four reference genes (ALG9, TAF10, TFC1, and UBC6)57. Data were derived from two biological replicates, each with three technical replicates.
Extraction, purification and analysis of GLP-1
Each transformant was inoculated at OD₆₀₀ 0.1 into 50 mL of the corresponding medium in a shake flask and grown overnight, and 5 mL aliquot was used for protein sample preparation. For protein extraction, cultures were centrifuged and washed twice with 1 mL double-distilled water (ddw). The pellet was resuspended in 100 µL of 0.2 M NaOH and 100 µL ddw, incubated at room temperature for 5 minutes, then neutralized with 100 µL of 0.2 M HCl. The sample was centrifuged at 10,000 rpm for 5 minutes at 4°C. The supernatant was aspirated, and the pellet was resuspended in Simple Buffer (10 µL per OD600 unit). Then, 100 µL of glass beads was added to each tube, and cells were lysed using a bead beater (30 seconds ON, 60 seconds OFF, 6 cycles at 3950 rpm). The lysates were boiled at 100°C for 10 minutes, centrifuged at 14,800 rpm for 10 minutes, and the supernatant was loaded for analysis.
For Western blot analysis, protein samples were separated by 10% SDS-PAGE gel and transferred onto a polyvinylidene difluoride (PVDF) membrane (Millipore). The membrane was blocked with 5% skim milk in TBST and incubated overnight at 4°C with primary antibodies: Anti-β-Tubulin (AbMART, Shanghai) and Anti-His-tag (Yeasen, Shanghai). After washing, the membrane was incubated at room temperature for 2 hour with HRP-conjugated secondary antibodies—anti-rabbit for His-tag and anti-mouse for β-tubulin—both diluted 1:10,000 in TBST containing 5% skim milk. Enhanced chemiluminescence signals were detected using Immobilon ECL Ultra Western HRP Substrate (Millipore), followed by chemiluminescent imaging.
To purify target proteins, cells (150–200 OD600 units) were harvested and lysed in PBS using 200 µL glass beads and 500 µL lysis buffer, followed by mechanical disruption in a pre-chilled bead beater (6 cycles of 20 s at 5.5 m/s, with 1 min intervals at 4°C). The supernatant was incubated with pre-equilibrated His-Pur Ni-NTA resin (LABLEAD) at 4°C for 2 h with gentle agitation. The resin was then washed with wash buffer (PBS with 100 mM imidazole). The targe proteins were eluted with 250 mM imidazole. The eluate was pooled, desalted, and concentrated using a 10 kDa MWCO Pierce™ protein concentrator (Millipore) by centrifugation at 10,000 × g. Protein purity and concentration were assessed by SDS-PAGE and the Bradford method, respectively (Fig. S19). Subsequently, the target protein sample was sequenced using timsTOF HT (Bruker, Beijing). Figure S18 illustrates that the final purified GLP-1 analogue maintained an identical sequence to GLP-1 (E G T F T S D V S S Y L E G Q A A K E E F I A W L V R G R G).
Extraction and analysis of 2-PE
The induced strains were transferred into liquid medium for overnight culture, then the seed culture was transferred into 250 mL flasks containing 50 mL of medium with an initial OD600 of 0.2. The two-phase shake-flask fermentation was initiated by adding 5 mL of isopropyl myristate. The cultures were incubated in a shaker at 30°C, 220 rpm for 72 hours. After cultivation, the yeast broth was centrifuged and the supernatant was analyzed for aromatic compound concentrations.
For 2-PE analysis, the organic phase was centrifuged and collected at 4200 rpm for 10 minutes for Gas Chromatography–Mass Spectrometry (GC − MS) analysis. An Agilent Technologies 7820A GC system equipped with a HP-5 column (30 m × 0.25 mm × 0.25 µm). The sample (1 µL) was injected in split mode (1:20), and the GC oven temperature program was applied with 3 mL/min nitrogen as carrier gas: 100°C for 1 min, 10°C/min to 270°C, an increase of 30°C/min to 320°C, hold for 15 min. The injector temperature was 250°C. The structure of 2-PE was further analyzed using a GC/ MS-QP2010 Plus insturment of Shimadzu (Japan) with a quadrupole mass analyzer (QMA) and a HP-5MS column (30 m × 0.25 mm × 0.25 µm). The mass scan range was 50 − 800 m/z. A 2-PE standard (GC grade) was purchased from Macklin (Shanghai, China). All of the GC analysis were quantified using a six-point calibration curve.
Fermentation, extraction and HPLC quantification of carotene
The yeast strains were cultured in 50 mL SC-His or SC-Leu medium at 30°C (220 rpm) for 72 h. A 10% aliquot of the total culture was harvested by centrifugation (4200 rpm, 2 min) and washed twice with sterile deionized water. Cell pellets were resuspended in 1 mL of methanol:dimethyl sulfoxide (1:1, v/v) with 100 µL 0.5 mm glass beads. Mechanical lysis was performed using a Bioprep-24R homogenizer (Hangzhou, Allsheng) with 10 cycles of 45 s shaking followed by 20 s pausing. The lysate was incubated at 55°C for 10 min with intermittent vortexing and centrifuged at 14,800 rpm for 10 min at 4°C. The supernatant was collected as the crude β-carotene extract and stored at -20°C until analysis. β-carotene quantification was performed using an Agilent 1220 HPLC system equipped with a diode-array detector (DAD) and a InfinityLab Poroshell 120 EC-C18 column (4.6 × 250 mm, 4 µm particle size; Agilent Technologies). For each sample, 20 µL of extract was injected onto the column, and separation was achieved using an isocratic mobile phase of methanol:acetonitrile:dichloromethane (42:42:16, v/v/v) at a flow rate of 1 mL/min for 25 min. The column temperature was maintained at 40°C, and absorbance was monitored at 450 nm. β-carotene was identified by comparing retention times with an authentic standard and quantified by integrating peak areas using Agilent ChemStation software.
Fermentation, extraction and HPLC quantification of ergothioneine
Overnight-cultured yeast was transferred into a shake flask containing 50 mL of the corresponding medium. After 48 hours of cultivation, an additional 5 mL of 20% glucose was supplemented. Cells were harvested at 24 hours post-glucose supplementation. 1/10 cells was collected with 150 µL of 0.5 mm glass beads and cell disruption was performed (5 cycles of 45s shaking followed by 20s pausing). 1 mL ddH2O was added for thorough dissolution, followed by centrifugation at 12,000 rpm for 5 minutes. Ergothioneine quantification was performed using an Agilent 1220 HPLC system equipped with a diode-array detector (DAD) and an InfinityLab Poroshell 120 Hilic column (3.0 × 100 mm, 2.7 µm particle size; Agilent Technologies). The mobile phase consisted of 5 mM lithium acetate and acetonitrile in a ratio of 15:85. Separation was performed at a flow rate of 0.5 mL/min for 25 minutes. The column temperature was maintained at 30°C, and absorbance was monitored at 254 nm by using the same machine that analyses carotene. Ergothioneine was identified by comparing retention times with an authentic standard.
Data availability
All data generated or analysed during this study are included in the main text or the supplementary table.
A
Acknowledgements
This work was supported by the National Key Research and Development Program of China (2024YFA0916503), National Natural Science Foundation of China (32122050 and 32370074) and Fujian Provincial Natural Science Foundation of China (2025J011005). We are grateful to National Demonstration Center for Experimental Life Sciences Education, Xiamen University for providing HPLC instrument and for help in performing the GC-MS.
A
Author contributions
Z.Q.L. conceived and supervised the project. Z.Q.L. and A.N.L. designed the experiments, analyzed the data, and wrote the original manuscript. Q.Y.Z. performed the HPLC analysis. Z.Y.Y. developed and conducted all protein assay experiments. J.S.C. performed all stability testing experiments. All authors reviewed, edited, and approved the final manuscript. Z.Q.L. acquired the funding.
A
Competing interestsPortions of the content in this article have been applied for a Chinese invention patent (Patent Application Number: CN202411825285.2), which does not affect the openness of the data or academic discussions. All authors declare no competing interests.
Electronic Supplementary Material
Below is the link to the electronic supplementary material
Reference
1.
Garcillán-Barcia MP, de la Cruz F, Rocha E (2025) P. C. The extended mobility of plasmids. Nucleic Acids Res 53:gkaf652
2.
Castañeda-Barba S, Top EM, Stalder T (2024) Plasmids, a molecular cornerstone of antimicrobial resistance in the One Health era. Nat Rev Microbiol 22:18–32
3.
Wein T, Dagan T (2020) Plasmid evolution. Curr Biol 30:R1158–R1163
4.
Wang G, Wang Q, Qi Q, Wang Q (2023) Dynamic plasmid copy number control for synthetic biology. Trends Biotechnol 42:147–150
5.
Kumar S, Lezia A, Hasty J (2024) Engineering plasmid copy number heterogeneity for dynamic microbial adaptation. Nat Microbiol 9:2173–2184
6.
Karim AS, Curran KA, Alper HS (2013) Characterization of plasmid burden and copy number in Saccharomyces cerevisiae for optimization of metabolic engineering applications. FEMS Yeast Res 13:107–116
7.
Clarke L, Carbon J (1980) Isolation of a yeast centromere and construction of functional small circular chromosomes. Nature 287:504–509
8.
Tschumper G, Carbon J (1983) Copy number control by a yeast centromere. Gene 23:221–232
9.
Gnügge R, Rudolf F (2017) Saccharomyces cerevisiae Shuttle vectors. Yeast 34:205–221
10.
Kang CW et al (2018) Synthetic auxotrophs for stable and tunable maintenance of plasmid copy number. Metab Eng 48:121–128
11.
Ramiro-Martínez P, De Quinto I, Lanza VF, Gama JA, Rodríguez-Beltrán J (2024) Universal rules govern plasmid copy number. Nat Commun 16:6022–6035
12.
Dwidar M, Yokobayashi Y (2019) Riboswitch Signal Amplification by Controlling Plasmid Copy Number. ACS Synth Biol 8:245–250
13.
Baumgart L, Mather W, Hasty J (2017) Synchronized DNA cycling across a bacterial population. Nat Genet 49:1282–1285
14.
Li C, Zou Y, Jiang T, Zhang J, Yan Y (2022) Harnessing plasmid replication mechanism to enable dynamic control of gene copy in bacteria. Metab Eng 70:67–78
15.
Li J, Cui M, Zhao J, Wang J, Fang X (2023) A self-amplifying plasmid based ultrasensitive biosensor for the detection of As(Ⅲ) in water. Biosens Bioelectron 221:114937
16.
Rouches MV, Xu Y, Cortes LBG (2022) Lambert, G. A plasmid system with tunable copy number. Nat Commun 13:3908–3920
17.
Joshi SH-N, Yong C, Gyorgy A (2022) Inducible plasmid copy number control for synthetic biology in commonly used E. coli strains. Nat Commun 13:6691
18.
Chan K-M, Liu Y-T, Ma C-H, Jayaram M, Sau S (2013) The 2 micron plasmid of Saccharomyces cerevisiae: A miniaturized selfish genome with optimized functional competence. Plasmid 70:2–17
19.
Ma C-H et al (2019) A Flp-SUMO hybrid recombinase reveals multi-layered copy number control of a selfish DNA element through post-translational modification. PLoS Genet 15:e1008193
20.
Zakian VA, Brewer BJ, Fangman WL (1979) Replication of each copy of the yeast 2 micron DNA plasmid occurs during the S phase. Cell 17:923–934
21.
Lian J, Jin R, Zhao H (2016) Construction of plasmids with tunable copy numbers in Saccharomyces cerevisiae and their applications in pathway optimization and multiplex genome integration. Biotech Bioeng 113:2462–2473
22.
Hohnholz R, Achstetter T (2019) Recombinant multicopy plasmids in yeast – interactions with the endogenous 2 µm. FEMS Yeast Res 19:foz001
23.
Vieira Gomes AM, Carmo S, Silva Carvalho T, Mendonça L, Bahia F, Parachin NS (2018) Comparison of Yeasts as Hosts for Recombinant Protein Production. Microorganisms 6:38
24.
Van Boeckel TP et al (2017) Reducing antimicrobial use in food animals. Science 357:1350–1352
25.
Futcher AB, Cox BS (1983) Maintenance of the 2 microns circle plasmid in populations of Saccharomyces cerevisiae. J Bacteriol 154:612–622
26.
Rizvi A, Prajapati SM, Nag HK, P., Ghosh SK (2017) The 2-µm plasmid encoded protein Raf1 regulates both stability and copy number of the plasmid by blocking the formation of the Rep1–Rep2 repressor complex. Nucleic Acids Res 45:7167–7179
27.
Som T, Armstrong KA, Broach F, Volkert C (1988) Autoregulation of 2µm circle gene expression provides a model for maintenance of stable plasmid copy levels. Cell 52:27–37
28.
Ghosh SK, Hajra S, Jayaram M (2007) Faithful segregation of the multicopy yeast plasmid through cohesin-mediated recognition of sisters. Proc. Natl. Acad. Sci. U.S.A. 104, 13034–13039
29.
Shen Y et al (2023) Dissecting aneuploidy phenotypes by constructing Sc2.0 chromosome VII and SCRaMbLEing synthetic disomic yeast. Cell Genom 3:100364
30.
Volkert FC, Broach JR (1986) Site-specific recombination promotes plasmid amplification in yeast. Cell 46:541–550
31.
McQuaid ME et al (2017) The yeast 2-µm plasmid Raf protein contributes to plasmid inheritance by stabilizing the Rep1 and Rep2 partitioning proteins. Nucleic Acids Res 45:10518–10533
32.
Liu Z, Zhao qiuyun, Xu C, Song H (2024) Compensatory evolution of chromosomes and plasmids counteracts the plasmid fitness cost. Ecol Evol. https://doi.org/10.1002/ece3.70121
33.
Schimmel P (2018) The emerging complexity of the tRNA world: mammalian tRNAs beyond protein synthesis. Nat Rev Mol Cell Biol 19:45–58
34.
Pinzaru AM, Tavazoie SF (2023) Transfer RNAs as dynamic and critical regulators of cancer progression. Nat Rev Cancer 23:746–761
35.
Chan PP, Lowe TM (2016) GtRNAdb 2.0: an expanded database of transfer RNA genes identified in complete and draft genomes. Nucleic Acids Res 44:D184–189
36.
Won H, Cho J-Y, Lee S (2025) Clinical development of oral semaglutide for the treatment of type 2 diabetes mellitus: focusing on early phase clinical trials. Transl Clin Pharmacol 33:1
37.
Luo G et al (2022) Enhancement of HSA-pFSHβ production by disrupting YPS1 and supplementing N-acetyl-L-cysteine in Pichia pastoris. Front Microbiol 13
38.
Lin Y, Feng Y, Zheng L, Zhao M, Huang M (2023) Improved protein production in yeast using cell engineering with genes related to a key factor in the unfolded protein response. Metab Eng 77:152–161
39.
Hedin KA et al (2023) Cold Exposure and Oral Delivery of GLP-1R Agonists by an Engineered Probiotic Yeast Strain Have Antiobesity Effects in Mice. ACS Synth Biol 12:3433–3442
40.
Jiang P, Ma C, Hao J, Han Y, Li M (2019) Construction of a dietary-cure Saccharomyces cerevisiae expressing long-acting glucagon-like peptide-1 and investigation of its hypoglycemic activity in type 2 diabetes mellitus mouse model. J Biosci 44:9
41.
Region & Segment Forecasts (2025) Flavors And Fragrances Market Size | Industry Report. https://www.grandviewresearch.com/industry-analysis/flavors-fragrances-market
42.
Kim B, Cho B, Hahn J (2014) Metabolic engineering of Saccharomyces cerevisiae for the production of 2-phenylethanol via Ehrlich pathway. Biotech Bioeng 111:115–124
A
43.Mo Q, Song W, Xue Z, Yuan J (2022) Multi-level engineering of Saccharomyces cerevisiae for the synthesis and accumulation of retinal. Green Chem 24:8259–8263
A
44.Guo Y et al (2015) YeastFab: the design and construction of standard biological parts for metabolic engineering in Saccharomyces cerevisiae. Nucleic Acids Res 43:e88
45.
Ma Y, Zu Y, Huang S, Stephanopoulos G (2023) Engineering a universal and efficient platform for terpenoid synthesis in yeast. Proc. Natl. Acad. Sci. U.S.A. 120, e2207680120
46.
Han Y, Tang X, Zhang Y, Hu X, Ren L-J (2021) The current status of biotechnological production and the application of a novel antioxidant ergothioneine. Crit Rev Biotechnol 41:580–593
47.
Van Der Hoek SA et al (2019) Engineering the Yeast Saccharomyces cerevisiae for the Production of L-(+)-Ergothioneine. Front Bioeng Biotechnol 7:262
48.
Van Hoek D (2022) Engineering precursor supply for the high-level production of ergothioneine in Saccharomyces cerevisiae. Metab Eng 70:129–142
49.
Weinhandl K, Winkler M, Glieder A, Camattari A (2014) Carbon source dependent promoters in yeasts. Microb Cell Fact 13:5
50.
Malakondaiah S et al (2024) Gene silencing by RNA interference: a review. GENOME INSTAB DIS 5:225–241
51.
Du F et al (2024) Optimizing multicopy chromosomal integration for stable high-performing strains. Nat Chem Biol 20:1670–1679
52.
Carvalho JP, Moreno DS, Domingues L (2025) Genetic engineering of Saccharomyces boulardii: Tools, strategies and advances for enhanced probiotic and therapeutic applications. Biotechnol Adv 84:108663
53.
Gassler T, Heistinger L, Mattanovich D, Gasser B, Prielhofer R (1923) CRISPR/Cas9-Mediated Homology-Directed Genome Editing in Pichia pastoris. Methods Mol Biol 211–225 (2019)
54.
Ng H, Dean N (2017) Dramatic Improvement of CRISPR/Cas9 Editing in Candida albicans by Increased Single Guide RNA Expression. mSphere 2, e00385-16
55.
Liu Y-T, Chang K-M, Ma C-H, Jayaram M (2016) Replication-dependent and independent mechanisms for the chromosome-coupled persistence of a selfish genome. Nucleic Acids Res 44:8302–8323
56.
Rodríguez CI et al (2000) High-efficiency deleter mice show that FLPe is an alternative to Cre-loxP. Nat Genet 25:139–140
57.
Torrent M, Chalancon G, De Groot NS, Wuster A (2018) Madan Babu, M. Cells alter their tRNA abundance to selectively regulate protein synthesis during stress conditions. Sci Signal 11:eaat6409