To validate the feasibility of our modular system for enabling dynamic, precise stress adaptation in yeast cell factories, we separately characterized the performance of three core functional modules—stress sensing & resistance, degradation degron, and repressor circuit—and confirmed their ability to fulfill key technical roles in the overall stress-responsive framework.
Each module targets a critical need: The stress sensing & resistance module provides specific "stress recognition-execution" genetic elements. The degradation degron module achieves tight temporal control of protein expression to reduce metabolic burden. And the repressor circuit module constructs the rhythmic core for oscillatory regulation. Collectively, the characterization results of these three modules lay a solid foundation for integrating the complete system and realizing adaptive stress regulation in industrial yeast strains.
I. Theory and Mechanism
The core theory of this experiment lies in using synthetic biology approaches to enhance the tolerance of cell factories towards environmental stresses, thereby stabilizing or increasing the yield of target products. The intrinsic mechanism is as follows: when cells encounter specific stresses (such as high temperature, hyperosmotic pressure, oxidative stress, etc.), the corresponding intracellular signaling pathways are activated. We clone promoters that respond to these specific stresses and splice them with functional genes that can enhance cellular stress resistance (such as genes for synthesizing compatible solutes, molecular chaperones, or antioxidant enzymes), thereby constructing a "stress-sensing and resistance-enhancing" genetic circuit. The design logic of this circuit is to maintain low-level expression under normal conditions to conserve cellular resources; once stress is encountered, the promoter is activated, driving the high-efficiency expression of the resistance genes to help the cell counteract the stress, maintain normal metabolic flux, and ultimately ensure the efficient synthesis of the target product. Additionally, a degradation tag is appended to the gene to help the cell promptly downregulate the circuit's expression after the stress subsides.
II. Experimental Design and Methods
)
Fig. 1 High-temperature promoter
Fig. 2 High osmotic promoter
Fig. 3 ROS promoter fluorescence
Validation of Resistance Genes: The test conditions were set up similarly to the method described above, except that the plasmids used constitutive promoters driving the resistance genes. The OD600 value of the culture was monitored during the process to characterize yeast growth. Partial results are shown below:
Fig. 4 Anti-high temperature gene
Fig. 5 Anti-high osmotic gene
Fig. 6 Anti-ROS gene
Finally, we screened the most effective components: the anti-high temperature promoter – pUAS-CYC1, the anti-high temperature gene – SHSP; the anti-hyperosmotic promoter – pGPD1, the anti-hyperosmotic gene – HOG1; the anti-ROS promoter – pTRX, the anti-ROS gene – SOD1.
2. Integration of the Optimal Circuit:
The best-performing promoter-gene pairs screened in the previous step were assembled with the optimal degradation tag using homologous recombination technology and integrated into the genome of the carotenoid-producing yeast cell factory, constructing stable engineered strains.
3. Performance Verification:
In shake flasks, parallel fermentation was conducted comparing the engineered strains harboring the integrated optimal circuit with control strains (the original strain without the integrated resistance circuit). The integrated yeast strains were fermented at 37°C. During the fermentation process, carotenoid concentration (represented by OD450) and cell density (represented by OD600) were measured daily.
III. Results and Analysis
Analysis of the experimental data yielded the following results: In the preliminary screening experiments, yeast strains carrying specific promoter-gene combinations showed significantly better growth under the corresponding stress conditions compared to the control group. The optimal combinations identified were: for high-temperature resistance: pUAS-CYC1---SHSP; for hyperosmotic resistance: pGPD1---HOG1; for ROS resistance: pTRX---SOD1.
Fig. 7 Fermentation result of β-carotene producing control strain
Fig. 8 Fermentation result of β-carotene producing, G9
Fig. 9 Fermentation result of β-carotene producing, G6
In the final cell factory validation, the engineered strain with the integrated optimal circuit showed an increased yield of the target product compared to the control strain after being subjected to stress. The introduction of this resistance circuit effectively mitigated the impact of stress on cell growth and production, enhancing the stability of yeast cell production under stressful conditions. These results confirm that the constructed genetic circuit functions as designed, significantly improving the robustness and production capacity of the cell factory under adverse conditions.
IV. Discussion
The successful validation of our engineered yeast strains underscores the efficacy of a synthetic biology approach in bolstering the resilience of cell factories. The core of our strategy was the design of a dynamic, stress-responsive genetic circuit that remains largely silent under optimal conditions but is rapidly activated upon stress perception. The data presented confirm that this design logic functions as intended.
Validation of Degradation Tags
To achieve precise temporal control of protein pulses while conserving cellular resources, we implemented a proteolytic control strategy based on degrons. This degradation mechanism can not only rapidly degrade useless proteins and recycle resources when stress is eliminated or switched, but also reduce interference between the regulatory processes of different stresses during stress switching. In this way, it achieves the goal of intelligent stress-resistant regulation desired in the yeast SMART system.Through a literature survey, we selected six ubiquitin-proteasome-dependent degradation tags for testing. These degrons had previously been used as components of fusion proteins. We chose ubiquitin-proteasome-dependent degrons primarily for three reasons: first, the ubiquitin-proteasome degradation mechanism is relatively well-characterized. With the cooperation of ubiquitin-activating enzyme E1, ubiquitin-conjugating enzyme E2, and ubiquitin ligase E3, the small protein ubiquitin is added to specific lysine (Lys) residues of unstable proteins. Subsequent polyubiquitin chains are recognized as "degradation signals" and targeted for degradation by the proteasome; second, these degrons are straightforward to apply—simply appending them to the N-terminus of the target protein is sufficient to trigger degradation; third, their half-lives can be easily regulated—altering just a few amino acids can change the structural stability of the degron, thereby adjusting the half-life of the target protein, a feature that offers significant advantages for controlling oscillation periods.
Fig. 1 Working principle of ubiquitin degradation
> Yu H, Matouschek A. Recognition of Client Proteins by the Proteasome. Annu Rev Biophys. 2017 May 22;46:149-173. doi: 10.1146/annurev-biophys-070816-033719. Epub 2017 Mar 9. PMID: 28301771.
For degron performance testing, we used the inducible DDI2 promoter to drive the expression of fluorescent reporter proteins. Different N-terminal degrons were fused to the reporter proteins. We set up a negative control group (4741) without fluorescent protein, a positive control group (J0) with fluorescent protein but no degradation tag, and experimental groups (J1[1] J2[2] J3[3] J4[3] J5[3] J6[4]) with fluorescent proteins tagged with six different degrons. All groups were given the same nutritional conditions and inducer concentration. At 30°C, we continuously measured fluorescence intensity and optical density at 600 nm (OD600) to generate fluorescence intensity changes and growth curves, thereby evaluating the degradation efficiency of each degron.
Fig. 2 The ratio of fluorescence intensity to OD600 over time at 30°C
Preliminary experiments were conducted at 30°C, the optimal growth temperature for Saccharomyces cerevisiae. Under these conditions, all fusion proteins exhibited extremely high stability with excessively long half-lives (t½), making reliable discrimination between different degrons impossible. Inspired by reports that UPS-mediated degradation is temperature-sensitive, we next subjected the strains to a temperature gradient ranging from 30°C to 40°C. A clear temperature dependence was observed: increasing the culture temperature shortened the half-life (t½) of all fusion proteins, with 37°C providing the optimal balance between degron activity and cell fitness. Among the six candidate degrons, degron 6 resulted in the shortest half-life (t½) for the fusion protein. Based on this, we obtained an efficient degradation tag in S. cerevisiae, which was subsequently applied in the ternary oscillator.
Fig. 3 The ratio of fluorescence intensity to OD600 over time at 37°C
Reference
[1] Meyn MA 3rd, Melloy PG, Li J, Holloway SL. The destruction box of the cyclin Clb2 binds the anaphase-promoting complex/cyclosome subunit Cdc23. Arch Biochem Biophys. 2002 Nov 15;407(2):189-95. doi: 10.1016/s0003-9861(02)00467-8. PMID: 12413490.
Mechanism
There is a class of DNA-binding proteins that can recognize and bind to specific operator sequences, block the binding of RNA polymerase to promoters, and thereby inhibit the transcription of downstream genes—these proteins are referred to as repressors in this study. Repressors enable transcriptional regulation of yeast gene expression, acting as "on/off switches" for gene expression. In the expression of stress-resistant genes, repressors allow for precise regulation of stress-resistant gene expression, saving intracellular resources and achieving "intelligent stress resistance" in yeast, i.e., yeast SMART.
Additionally, yeast is a eukaryote, and its genome is localized in the nucleus. Since repressors used in this study are derived from prokaryotic systems, they need to enter the nucleus to bind to operator sequences. Therefore, a nuclear localization signal (NLS) was fused to the C-terminus or middle region of all repressors and anti-repressors in the experiment. The NLS guides proteins into the nucleus via nuclear transport signals, ensuring effective binding between repressors and DNA operator sequences—this is a critical prerequisite for prokaryotic repressors to function in yeast.
Fig. 1 Addition of nuclear localization signal (NLS) to the C-terminus of repressors and anti-repressors (taking LacI as an example)
Experiments
To construct the tripartite oscillator in yeast, the experiment was carried out in three phases: "Monopartite Validation of Oscillator Components → Design and Testing of Closed-Loop Tripartite Oscillator → Construction and Detection of Tripartite Oscillator with Yeast Stress-Resistant Genes".
Fig. 2 Mechanism of monopartite validation for oscillator components
Step 1: Plasmid Construction
For each set of "repressor-operator sequence-anti-repressor", 3 recombinant plasmids were constructed: ① Plasmid containing the "melamine-inducible promoter - repressor (with NLS)" cassette; ② Plasmid containing the "constitutive promoter inserted with repressor operator sequence - mScarlet red fluorescent protein gene" cassette; ③ Plasmid containing the "galactose-inducible promoter - anti-repressor" cassette. A total of 12 plasmids were constructed using this method.
Step 2: Yeast Transformation
The 12 plasmids were separately transformed into competent Saccharomyces cerevisiae cells.
Step 3: Induction Treatment and Detection
All strains were inoculated in the same YPD medium and cultured for 18–20 hours to reach the logarithmic phase. Then, melamine was added to a final concentration of 0.5 mM (the same dosage for all experimental groups). The OD600 (cell density) and mScarlet fluorescence intensity (excitation wavelength: 579 nm; emission wavelength: 616 nm) were continuously monitored, with measurements recorded every 15 minutes for 24 hours.
Experimental results are as follows:
Fig. 3 Results of monopartite validation for oscillator components
As shown in Fig. 3, when the OD600 curves (indicating cell density) of all groups overlapped (ruling out the influence of cell growth differences on fluorescence intensity), the induced expression groups of LacI^ts, TetR, and λCI exhibited significantly inhibited fluorescence expression compared to the control groups. In contrast, LacI and CI857 showed unsatisfactory inhibitory effects.
These results indicate that LacI^ts, TetR, and λCI can be normally expressed under melamine induction. The fused NLS can effectively guide these repressors into the nucleus. These repressors specifically bind to their respective operator sequences (LacO, TetO, OR) and significantly inhibit mScarlet transcription. Thus, LacI^ts, TetR, and λCI are compatible with the yeast expression system and can be used as core components of the tripartite oscillator.
[Phase 2: Design and Testing of Closed-Loop Tripartite Oscillator]
This phase aimed to verify whether LacI^ts, λCI, and TetR can form a closed-loop negative feedback circuit ("Repressor A → Repressor B → Repressor C → Repressor A") and achieve rhythmic transcriptional regulation.
After successful monopartite validation, the team constructed these three repressors into a closed-loop negative feedback circuit. To detect its function via fluorescence oscillation, the operator sequence of the previous repressor was inserted into the promoter upstream of each repressor (for LacI^ts/λCI/TetR). Additionally, mScarlet expression was coupled with TetR, allowing the observation of mScarlet oscillation over time (fluorescence intensity reflects TetR expression, indirectly indicating the oscillation state of the circuit).
Fig. 4 Schematic design of the closed-loop tripartite oscillator
Four recombinant plasmids were constructed to form the closed-loop circuit:
① Plasmid with LacO inserted into the promoter of λCI (LacI^ts binds to LacO to inhibit λCI transcription); ② Plasmid with OR inserted into the promoter of TetR (λCI binds to OR to inhibit TetR transcription); ③ Plasmid with TetO inserted into the promoter of LacI^ts (TetR binds to TetO to inhibit LacI^ts transcription); ④ Plasmid with TetO inserted into the promoter of mScarlet (fluorescence intensity reflects TetR expression, indirectly indicating circuit oscillation).
Step 2: Yeast Transformation
The four plasmids were separately transformed into competent Saccharomyces cerevisiae cells, resulting in a total of 7 strains.
Step 3: Fluorescence Detection
The 7 strains were cultured at a constant temperature of 30°C (no inducer added). Changes in mScarlet fluorescence intensity were continuously monitored for 136 hours, with measurements recorded every 15 minutes.
The results are as follows:
Fig. 5 Results of closed-loop testing for the tripartite oscillator
As shown in Fig. 5:
All 7 strains exhibited periodic changes in mScarlet fluorescence intensity, confirming that the closed-loop circuit was functional (Fig. 5a). After superimposing and averaging the results of the 7 strains, the system’s oscillation became more intuitive (Fig. 5b). Further denoising processing made the oscillation pattern even clearer (Fig. 5c).
The oscillation period of the tripartite oscillator was approximately 21 hours. These results verify the establishment of the "LacI^ts → λCI → TetR → LacI^ts" closed-loop negative feedback logic: The three repressors restrict each other to form rhythmic expression, indicating the successful construction of the tripartite oscillator.
To further explore the application potential of biological oscillators, the team also reproduced an oscillator design reported in a literature study.
Fig. 6 Tripartite oscillator proposed by Li et al.
> Li, Z., Qiao, G., Wang, X. et al. De novo designed protein guiding targeted protein degradation. Nat Commun 16, 6598 (2025). https://doi.org/10.1038/s41467-025-62050-z
The core design of the literature-derived oscillator relies on a negative feedback loop between three components to achieve periodic regulation: GP-McsB (targets and degrades LacI); LacI (inhibits antiMcsB expression); antiMcsB (inhibits GP-McsB function).
Specifically, GP-McsB targets and degrades LacI, leading to a decrease in LacI levels; Reduced LacI levels relieve its inhibition of antiMcsB expression; Expressed antiMcsB inhibits GP-McsB function by targeted binding, thereby reducing LacI degradation; LacI levels rise again, re-inhibiting antiMcsB expression—forming a periodic negative feedback cycle that ultimately regulates the oscillation of mKate2 abundance.
The team replaced mKate2 with mScarlet and constructed 9 plasmids according to the literature design, which were then transformed into yeast. The 9 plasmids included: ① Plasmid for constitutive expression of LacI; ② Plasmid with LacO inserted into the promoter of mScarlet (fluorescent protein); ③ Plasmid with LacO inserted into the promoter of antiMcsB; ④ Plasmid for constitutive expression of GP-McsB; ⑤ Plasmid for constitutive expression of the degradation protein ClpP; ⑥ Plasmid for constitutive expression of the chaperone protein ClpC (ClpC provides "substrate recognition, unfolding, and delivery" services for ClpP and activates ClpP’s degradation activity); ⑦ Plasmid for constitutive expression of McsA (a protein that activates McsB); ⑧–⑨ Plasmids containing selection markers.
Fig. 7 Design and working mechanism of the literature-derived oscillator in yeast
The strains were cultured at a constant temperature of 30°C (no inducer added). Changes in mScarlet fluorescence intensity were continuously monitored for 48 hours, with measurements recorded every 15 minutes.
However, no fluorescence fluctuations were detected. The core issue is likely the poor compatibility between prokaryotic components and the yeast eukaryotic system. The literature-derived oscillator relies on prokaryotic degradation enzymes (ClpP/ClpC) to degrade repressors. However, yeast lacks a Clp protease system homologous to that of prokaryotes, leading to the accumulation of repressor proteins. Accumulated repressors continuously inhibit downstream genes, disrupting the oscillation balance.
[Phase 3: Construction and Detection of Tripartite Oscillator with Yeast Stress-Resistant Genes]
To achieve dynamic, periodic stress resistance in industrial yeast while minimizing metabolic burden, we designed an oscillatory anti-stress control scheme—coupling the validated three-node oscillator (LacI^ts → λCI → TetR) to a set of yeast endogenous stress-resistance genes. This design enables coordinated, periodic expression of protective functions: instead of constitutively overexpressing stress-resistance genes (which wastes cellular resources), the oscillator drives these genes to “turn on/off” rhythmically, matching the timing of stress challenges in industrial fermentation processes.
A core advantage of this scheme is the ability to artificially program the response threshold to complex stress: by tuning the oscillator’s period and the repressors’ DNA-binding affinities, we can precisely control when and how strongly stress-resistance genes are expressed, allowing adaptation to diverse fermentation environments (e.g., fluctuating temperature, osmolarity, or oxidative stress).
As a proof-of-concept, we targeted a β-carotene producing Saccharomyces cerevisiae strain: we introduced both the three-node oscillator and three key stress-tolerance genes into this host. Each stress-tolerance gene is regulated by one of the oscillator’s repressors, and their functions are:
shsp (Small Heat Shock Protein, enhancing resistance to heat stress.)
SOD1 (Superoxide Dismutase 1, mitigating oxidative stress by scavenging reactive oxygen species.)
HOG1 (Mitogen-Activated Protein Kinase Hog1, mediating osmotic stress response and promotes cell survival under high-osmolarity conditions.)
By rhythmically expressing these genes, we hypothesize that yeast can achieve “timely defense” against multiple stresses while reducing steady-state metabolic burden—ultimately boosting β-carotene yield.
Step 1: Plasmid Construction for Stress-Resistant Oscillator
Based on the three-node oscillator and stress-resistance gene regulation logic, we constructed three 3 recombinant plasmids: ① Plasmid with LacO inserted into the promoter of shsp; ② Plasmid with OR inserted into the promoter of SOD1; ③ Plasmid with TetO inserted into the promoter of HOG1.
Step 2: Yeast Transformation and Strain Construction
A β-carotene producing Saccharomyces cerevisiae strain was used as the host. We sequentially introduced genetic modules into this strain.
Step 3: Fermentation and Continuous Detection
Engineered strains (with oscillatory stress-resistance modules) and control strains (β-carotene producer without the oscillator or oscillatory stress-resistance gene expression) were inoculated into YPD medium and cultured at both 30°C and 37℃. Over 72 hours (3 days), we continuously monitored two key parameters, cell density (reflected by OD600) and β-carotene concentration (reflected by OD450). Measurements were recorded every 24 hours to capture both cell growth dynamics and effects of oscillatory stress protection on β-carotene productivity.
The results are as follows:
Fig. 8 Comparison of β-Carotene Production in Engineered vs. Control Strains
Analysis: we assessed β-carotene production in engineered and control strains at 30°C and 37°C over 3 days via quantitative OD₄₅₀/OD₆₀₀ ratio measurement (reflecting β-carotene yield relative to cell density).
For the engineered strain, the OD₄₅₀/OD₆₀₀ ratio at 37°C increased over time and even surpassed the 30°C group on Day 2 and Day 3. Conversely, the control strain exhibited a consistently lower OD₄₅₀/OD₆₀₀ ratio at 37°C compared to 30°C across all time points, with no upward trend throughout fermentation.
In summary, the engineered strain not only tolerates heat stress but also maintains or enhances β-carotene productivity—while the control strain lacks such adaptability and suffers reduced production under heat stress. This direct comparison validates that our stress-resistant oscillator effectively improve both thermal tolerance and β-carotene yield in industrial-relevant conditions.