Mechanisms of formic acid stress response in evolved yeast for microbial-electrocatalytic systems

Abstract

Integrated electrocatalytic microbial systems (iEMS) offers a sustainable route for carbon cycling. However, formic acid by-product stress severely hinders microbial conversion. Herein, we reveal the adaptive mechanisms of Saccharomyces cerevisiae under formic acid stress during ethanol cultivation through adaptive laboratory evolution and transcriptomic analysis. Formic acid exposure systematically downregulates central carbon metabolism while upregulating the acetaldehyde branch to replenish carbon skeletons. Concurrently, cell cycle genes are activated via MAPK signaling, and ribosomal biogenesis is enhanced to bolster stress-response translation. By improving the formic acid tolerance of Saccharomyces cerevisiae and investigating the underlying mechanisms, we engineered yeast to produce 3-hydroxypropionic acid (3-HP) from ethanol. Finally, we integrated an upstream electrocatalytic system with a downstream engineered yeast producing 3-HP, achieving a titer of 95.8 mg/L, which is 4-fold higher than that of the initial strain. This work elucidates the molecular basis of formic acid tolerance and demonstrates a viable route for iEMS.

1. Introduction

The continuous rise in atmospheric CO₂ concentrations has become a global environmental challenge, giving rise to issues such as global warming and the greenhouse effect [1], and poses a serious threat to ecological security and sustainable development [2,3]. Against this background, technologies for converting CO₂ into high value added chemicals have attracted widespread attention. Among these, the CO₂ reduction reaction (CO₂RR) shows great potential for application due to its mild reaction conditions and the ability to utilize renewable electricity as a power source [4]. However, the single CO₂RR process often faces bottlenecks such as low product selectivity and intense competition from hydrogen evolution reactions. To overcome these limitations, researchers have proposed the concept of integrated electrocatalytic microbial systems (iEMS) [5]. This system combines the high reaction rates of electrocatalytic reduction with the high selectivity of microbial synthesis. First, CO₂ is reduced via electrocatalysis to simple intermediates such as methanol and ethanol [6], which are then further converted by engineered microorganisms into complex long-chain chemicals, thereby achieving a synergistic improvement in both energy efficiency and product specificity [7,8].

Significant progress has been made in the field of CO₂RR through strategies such as the rational design of copper-based catalysts, and efficient conversion to multi-carbon products such as ethanol has now been achieved [9]. However, during the electroreduction of CO₂ to ethanol, the electrons transferred to CO₂ are sometimes insufficient to drive carbon–carbon coupling, leading to the inevitable formation of formic acid as a one-carbon by-product [10]. In the iEMS framework, formic acid may accumulate in the electrolyte, resulting in carbon source waste and imposing toxic stress on microbial metabolism, which may in turn compromise the subsequent utilization of ethanol.

Saccharomyces cerevisiae (S. cerevisiae), a model host for genetic editing, possesses a well-defined genetic background and mature metabolic engineering tools [11]. Formic acid is a weak acid. When weak acids are present in yeast culture medium, they may diffuse into the cytoplasm where they dissociate, lowering the internal pH. Simultaneously, this dissociation releases H⁺ and HCOOH⁻, thereby disrupting intracellular pH equilibrium [12,13]. Consequently, cells expend additional ATP energy to expel protons [14]. At the same time, anions accumulate significantly in weakly acidic environments, increasing intracellular osmotic pressure and heightening water demand. This leads to excessive intracellular pressure, causing irreversible damage to cells [13]. Severely impairs metabolism and growth within yeast cells, resulting in slowed cell growth, increased cell death, and reduced levels of metabolic products [15]. Although glucose is the commonly used substrate for S. cerevisiae, ethanol serves as an excellent alternative substrate for this yeast. Compared to other strains, S. cerevisiae exhibits remarkably high ethanol tolerance [16,17]. When S. cerevisiae utilizes ethanol as a carbon source, ethanol can be directly converted into the crucial intermediate acetyl-CoA without consuming additional carbon [18]. This process, unlike the release of CO₂ when yeast uses glucose, aligns more closely with our strategic requirements for carbon conservation.

In this study, we focus on the transcriptional response of S. cerevisiae under extreme formic acid stress conditions using ethanol as the sole carbon source. By integrating scanning electron microscopy (SEM) and RNA-Seq transcriptomic analysis, we systematically investigate the metabolic pathway alterations and molecular response mechanisms of yeast during formic acid stress in an ethanol-containing environment. Our study aims to provide a theoretical basis for understanding the mechanisms underlying formic acid tolerance in S. cerevisiae when utilizing ethanol as a carbon source. Furthermore, we constructed a CO₂ electrocatalytic reduction system that successfully reduces CO₂ to ethanol and trace amounts of formic acid. Concurrently, we engineered S. cerevisiae adapted to formic acid to enable growth and metabolic production of 3-hydroxypropionic acid (3-HP) using ethanol as a substrate, thereby achieving the complete conversion from CO₂ to 3-HP within the iEMS framework.

2. Materials and methods

2.1. Strain activation and culture

In this study, all background strains used were S. cerevisiae CEN.PK102-5B (MATa; ura3-52, his3Δ1, leu2-3112). The strains were provided by the research group of Professor Hou at Shandong University. The activation of the S. cerevisiae strain is performed in Yeast Extract Peptone Dextrose Medium (YPD) liquid medium. (The composition of 1 L YPD liquid medium is 2% trypticase peptone, 1% yeast extract, and 2% glucose). The normal growth of yeast strains occurs in ethanol-containing medium. (The composition of 1 L ethanol-containing medium is 2% trypticase peptone, 1% yeast extract, and 2.5% ethanol). Synthetic complete dropout (SC) medium was used for the screening of genetically engineered yeast strains. Each 1 L of SC medium contained 1.7 g yeast nitrogen base, 5 g ammonium sulfate, 0.77 g corresponding dropout powder, and 2.5% ethanol. The standard culture conditions for S. cerevisiae are 30°C and 220 rpm in a shaking incubator. The growth of S. cerevisiae was monitored by measuring the OD₆₀₀ absorbance of the culture using a microplate reader. In subsequent experiments, varying concentrations of formic acid were added to observe the yeast's growth response. All experiments were performed in triplicate.

2.2. Adaptive laboratory evolution ALE enhances the formic acid tolerance of S. cerevisiae

The initial strain underwent gradient adaptation to formic acid in ethanol-containing medium under formic acid effect conditions. Over a period of 150 days, the S. cerevisiae was gradually adapted to grow in the presence of formic acid in ethanol-containing medium. Initially, add 0.1% formic acid by volume to the ethanol-containing medium. When growth reaches an OD₆₀₀ value of 1.5, inoculate 1% of the yeast culture into fresh medium. Increase the formic acid concentration by 0.01% with each passage. If yeast fails to grow normally, continue culturing at the original formic acid concentration for two additional passages to adapt to the formic acid environment, then resume cultivation with increasing formic acid concentrations. Three parallel groups were cultured for each treatment.

2.3. Scanning Electron Microscope (SEM) analysis

Inoculate S. cerevisiae into ethanol-containing medium and shake-culture until the logarithmic growth phase. Add 0.25% formic acid to the experimental group and an equal volume of sterile water to the control group. Continue shaking for 3 h. Subsequently, take 1 mL of bacterial culture for enrichment centrifugation. Wash three times with sterile 1× PBS solution. Finally, take the centrifuged yeast pellet, resuspend it in 2.5% glutaraldehyde solution, and fix overnight. Take the fixed bacterial suspension and dehydrate it sequentially with 30%, 50%, 70%, 90%, and 100% ethanol solutions, each dehydration step lasting 10 min. Finally, the yeast sediment was applied to a silicon wafer, dried, gold-sputtered, and then observed and photographed using a Fehn Phenom SEM.

2.4. Transcriptome sample preparation and analysis

Take the propagated S. cerevisiae strain and inoculate it into the same ethanol-containing medium. Once it reaches the logarithmic growth phase, add 0.25% formic acid to three experimental groups and an equal volume of sterile water to three control groups. Continue shaking the culture for 3 h, then immediately transfer 1 mL of the culture to cryovials. Rapidly freeze them in liquid nitrogen and store at −80°C. We employ standard extraction methods to isolate RNA from S. cerevisiae cells, followed by precise assessment of RNA integrity using the Agilent 2100 Bioanalyzer to ensure rigorous quality control of RNA samples. After performing quality control on the entire RNA sample, mRNA was enriched from total RNA using Oligo dT magnetic beads. Following fragmentation and a series of modifications, the library construction was completed. The effective concentration of the library was quantified using qRT-PCR. After library quality control, libraries were pooled according to effective concentration and target read depth requirements before undergoing Illumina sequencing. Transcriptome sequencing was performed by Novogene Co., Ltd.

2.5. Quantitative real-time qPCR analysis

RNA was extracted from preserved samples using the FastPure Universal Plant Total RNA Isolation Kit (Vazyme Biotech Co., Ltd), and subsequently, RNA was reverse transcribed into cDNA using the ABScript Neo RT Master Mix for qPCR with gDNA Remover (ABclonal Technology Co., Ltd.). qPCR was performed using the 2X Universal SYBR Green Fast qPCR Mix (RK21203).We randomly selected six genes that were either upregulated or downregulated from the transcriptomic analysis results. Using the highly conserved ATC1 gene as the internal reference, we measured the Cq values via qPCR to calculate gene expression levels. Each experiment was performed in triplicate.

2.6. Electrocatalytic reduction of CO₂ and gene-edited metabolic pathways in S. cerevisiae

By reviewing the relevant literature [9], we successfully synthesized CuO_x copper oxide nanocluster materials using acetylacetonate copper and sodium nitrate, in accordance with the experimental methods described. Under electrocatalytic conditions, this material can reduce CO₂ to ethanol, but this process is accompanied by the formation of formic acid as a by-product. At the same time, we genetically edited the S. cerevisiae strain cultured via serial passage. We constructed gene expression cassettes for the MCRca (Chloroflexus aurantiacus), ADH2 (S. cerevisiae) and ALD6 (S. cerevisiae) genes and integrated them into plasmids, which were then transformed into the yeast. Positive clones were selected using nutrient-deficient media and subsequently cultured in liquid medium in shaking flasks. After collecting the bacterial culture, the concentration of 3-HP was determined using high-performance liquid chromatography (HPLC) with a C18 column (Inertsil ODS-3).

2.7. Data analysis

The results are presented as mean ± standard deviation (SD), and statistical analysis was performed using SPSS 20.0 statistical software (SPSS Inc, Chicago, IL, USA). Statistical significance was determined using one-way analysis of variance (ANOVA) and Duncan's test. p < 0.05 indicates statistical significance.

3. Results and discussion

3.1. Adaptive evolution of formic acid-tolerant S. cerevisiae and analysis of cell morphology under stress

In this study, S. cerevisiae strain CEN.PK 102-5B was used as the starting strain for adaptive laboratory evolution (ALE). As shown in Fig. S1a, the growth of the parent strain was almost completely arrested in culture medium supplemented with 0.1% (v/v) formic acid. The strain achieved robust growth without formic acid supplementation, which demonstrated that formic acid exerted remarkable growth-inhibitory effects on the parental strain [19]. To select tolerant mutants, serial subcultivation was performed for successive generations under gradually elevated formic acid stress, and dynamic growth curves of strains harvested at distinct evolutionary passages (1st, 13th, 19th and 25th generations) are summarized in Fig. S1b. A clear stepwise improvement in formic acid adaptability was observed alongside prolonged ALE domestication: the 1st generation barely proliferated under 0.15% formic acid, while the 13th generation gained moderate tolerance and showed detectable biomass accumulation. After 19 rounds of passaging, the isolate survived 0.2% formic acid, and the 25th passage strain exhibited the strongest stress endurance. We isolated this stable mutant and named it evolved strain 5B-2 (CEN.PK 102-5B-2), whose maximum tolerable formic acid concentration was successfully elevated to 0.25% (v/v). We further characterized the full growth performance of evolved strain 5B-2 across a continuous formic acid concentration gradient (0, 0.05%, 0.1%, 0.15%, 0.2%, 0.25% v/v), with OD₆₀₀ monitored over a 72 h cultivation period (Fig. 1a). In the absence of formic acid or under low-concentration formic acid stress (0.05%-0.15%), the strain exhibited fast exponential proliferation and reached peak biomass within 36 h, with final OD₆₀₀ values approaching 1.7. Growth retardation emerged markedly at 0.2% formic acid: the lag phase was greatly extended, and the maximum OD₆₀₀ dropped to approximately 1.4 after 72 h incubation. Severe growth suppression was observed at the limiting tolerance concentration of 0.25% formic acid. The strain displayed an extremely long lag stage, with negligible biomass increase within the first 36 h; although slow proliferation persisted after 60 h, the OD₆₀₀ still failed to reach the peak level observed in low-stress groups even at the 72 h endpoint. This gradient growth assay verified that 0.25% (v/v) formic acid represented the critical threshold stress that triggered the full-scale adaptive response of strain 5B-2. Accordingly, we set this concentration as the treatment condition for subsequent transcriptomic profiling, established formic acid-treated evolved cells as the experimental group (Group D) and untreated wild-type parent cells as the control group (Group B). We further characterized the overall growth performance of the evolved strain 5B-2 by monitoring the OD₆₀₀ values over a 72-h cultivation period at different formic acid concentrations (Fig. 1a). In the absence of formic acid or under low-concentration formic acid stress (0.05%–0.15% v/v), the strain exhibited rapid exponential proliferation and reached peak biomass within 36 h, with the final OD₆₀₀ value approaching 1.7. At a formic acid concentration of 0.2% (v/v), the growth of the strain was significantly inhibited: the lag phase was remarkably prolonged, and after 72 h of cultivation, the maximum OD₆₀₀ value dropped to approximately 1.4. Under the condition of 0.25% (v/v) formic acid, the strain suffered severe growth suppression. It displayed an extremely long lag phase, with negligible biomass increase within the first 36 h; although slow proliferation persisted after 60 h, the OD₆₀₀ value still failed to reach the peak level observed in the aforementioned groups, even at the 72-h endpoint. This gradient growth assay confirmed that 0.25% (v/v) formic acid represented the critical threshold stress that triggered the full-scale adaptive response of strain 5B-2. Based on this, we set this concentration as the treatment condition for subsequent transcriptomic analysis, and established three independent biological replicates for both the formic acid-treated evolved cells (Group D) and the untreated wild-type parental cells (Group B) as the control.

Fig. 1.

The macroscopic effects of formic acid stress on S. cerevisiae. (a) Growth curves of S. cerevisiae in ethanol medium at different concentrations of formic acid. (b-c) Scanning electron microscopy observation of S. cerevisiae control group and formic acid stress group; (d) Boxplot of Sample Gene Expression Levels. The horizontal axis represents sample names, while the vertical axis represents log₂ (FPKM+1). (e) Sample Correlation Heatmap. The horizontal and vertical axes represent the squared correlation coefficients of each sample.

The surface morphology of S. cerevisiae in the normal conditions and under formic acid stress is shown in Fig. 1b–c. When ethanol serves as the substrate and formic acid is absent, the yeast cells exhibit smooth, rounded ellipsoidal spheres. In the experimental group subjected to formic acid effect, the smooth surface of the yeast cells began to wrinkle. This phenomenon can be attributed to the dissociation of formic acid into hydrogen ions after entering ∗Saccharomyces cerevisiae∗ cells, which in turn leads to a sharp decrease in intracellular pH. To maintain intracellular pH homeostasis, yeast cells need to consume a large amount of ATP, which ultimately hinders the proliferation of wild-type strains [20], and this result is consistent with the above-mentioned research conclusions. This change in the intracellular environment imposes significant survival pressure on yeast cells, further triggering osmotic imbalance and cellular dehydration, resulting in a reduction in cell volume, and ultimately causing the surface of yeast cells to wrinkle.

3.2. Transcriptomic data of S. cerevisiae under formic acid effect

To investigate the mechanism underlying the enhanced tolerance of the Saccharomyces cerevisiae strain 5B-2 under formic acid stress, this study employed transcriptomic RNA-seq to analyze the gene expression profile of this strain under formic acid stress conditions. Transcriptome analysis yielded gene expression values (Fragments Per Kilobase of exon model per Million mapped fragments, FPKM) for six samples, namely the control group BD (B1D, B2D and B3D) and the experimental group BS (B1, B2 and B3). These values were corrected for sequencing depth and gene length, and box plots statistically illustrated the distribution of gene expression levels across different samples. As shown in Fig. 1d, this indicates that systematic technical biases such as sequencing depth and gene length across different samples have been effectively corrected, laying the foundation for reliable inter-sample comparisons and differential gene analysis. The correlation of gene expression levels between samples is a crucial indicator for assessing experimental reliability and the validity of sample selection. A correlation coefficient of at least 0.8 between biological replicates is required to demonstrate sample validity. As shown in Fig. 1e, intra-group correlations for both the experimental and control groups exceed 0.97, while inter-group correlations exceed 0.866. This heatdmap clearly demonstrates that inter-sample variations are minimal, and expression patterns within each group are highly consistent. Differentially expressed genes (DEGs) were identified by calculating log₂FoldChange based on gene expression profiles of S. cerevisiae under formic acid stress. As shown in Fig. 2a, Figs. S2 and S3, volcano plots intuitively displayed the distribution of DEGs in each comparison group. Using the KEGG database, we performed functional enrichment analysis to examine transcriptional changes and compared the top 20 significantly enriched pathways as shown in Fig. 2b, Figs. S4 and S5. The results revealed complex pathways and gene enrichment patterns associated with formic acid tolerance in S. cerevisiae.

Fig. 2.

Transcriptome analysis of S. cerevisiae under formic acid stress (a) Volcano plot of differentially expressed genes screening in S. cerevisiae B1and B1D under formic acid stress. The green points represent downregulated genes, the red points represent upregulated genes, and the gray points represent genes with no significant difference. (b) Scatter plot of upregulated genes in the KEGG pathway analysis of S. cerevisiae B1and B1D under formic acid stress. The size and color of the dots represent the number of genes and the confidence level, respectively.

3.3. Formic acid stress leads to downregulation of carbon metabolism and upregulation of the glyoxylate cycle pathway in S. cerevisiae

Key genes associated with carbon metabolic pathways are listed in Table S1. Analysis of differentially expressed genes (DEGs) under formate stress revealed that the vast majority of genes related to carbohydrate metabolism—including those involved in glycolysis, the pentose phosphate pathway (PPP cycle), and glycosylation (as shown in Fig. 3) were significantly downregulated. In contrast, upregulated DEGs were mainly enriched in the glyoxylate cycle, indicating that S. cerevisiae initiates a pathway-specific adaptive metabolic strategy to resist formate stress [21]. In glycolysis, the HK gene responsible for converting glucose to glucose-6-phosphate [22], the GAPDH gene converting glyceraldehyde-3-phosphate to glyceraldehyde-1,3-bisphosphate, and the ENO gene converting glyceraldehyde-3-phosphate to phosphoenolpyruvate were all significantly downregulated [23]. The transcriptional repression of these genes blocks the cascade reactions of glycolysis, thereby restraining glucose catabolism and impairing the generation of intracellular energy substrates [24]. The PPP pathway was also transcriptionally suppressed by formate stress. The PGLS and PGD genes, which encode pivotal enzymes responsible for the sequential generation of 6-phosphogluconate and 5-phosphoribulose in the oxidative PPP branch, were significantly downregulated [25]. Meanwhile, E2.2.1.1 and TALDO1, which dominate the interconversion of tetrose, hexose, and heptose sugar intermediates in the non-oxidative PPP branch, also exhibited a pronounced declining expression trend. The downregulation of these PPP-related genes not only blocks the biosynthesis of pentose precursors for nucleic acid synthesis but also impairs intracellular NADPH regeneration, thereby disrupting cellular redox balance and further weakening the stress resistance of S. cerevisiae [26]. In starch and sucrose metabolism, the trehalose biosynthesis pathway was severely inhibited under formate stress. The key trehalose synthesis genes, including TPS1(otsA), TSL1, and TPS, which collaboratively catalyze the two-step transformation of UDP-glucose into intracellular trehalose [[27], [28], [29]], displayed consistent downregulated expression. As a crucial compatible solute and stress-protective metabolite in S. cerevisiae, the reduction of trehalose synthesis further impairs cellular osmotic homeostasis and aggravates metabolic perturbation under formate stress. Concurrently, the persistent deficiency of energy substrates caused by suppressed glycolysis and PPP pathways led to a widespread decline in the expression of glycogen metabolism-related genes [30]. Genes responsible for glycogen synthesis and cyclic degradation, such as GYS, GBE1, AGL, PYG, and IMA, were significantly downregulated, indicating that yeast cells actively suppress energy-consuming glycogen cycling to adapt to energy shortage under stress [31]. Overall, the significantly downregulated DEGs were predominantly enriched in pathways governing carbon source metabolism and energy production, indicating that formate stress may disrupt normal carbon source metabolism in S. cerevisiae. This disruption leads to insufficient energy supply, markedly slowed glycolysis and the PPP cycle, and reduced polysaccharide synthesis.

Fig. 3.

Schematic Diagram of Carbon Metabolism. Genes highlighted in red are upregulated. Genes highlighted in green are downregulated.

3.4. Effects of formic acid stress on MAPK signaling, cell cycle and Meiosis in S. cerevisiae

A schematic diagram of MAPK signaling pathway transduction is shown in Fig. 4a. Furthermore, KEGG pathway analysis (Table S2) indicates that following formic acid effect, the expression levels of MAPK pathways induced by pheromone pathways in S. cerevisiae significantly increased, while some genes associated with cell wall, high osmolarity, or starvation stress-related MAPK signaling pathways exhibited downregulation. MAPK pathway activation within this network is promoted by exogenous formic acid's impact on pheromone changes, exemplified by the activation of ERK genes and the FAR1 gene [32,33], or by regulating subsequent cell cycle processes through the CLB1/2 genes that control mitosis-specific cyclin expression [34].

Fig. 4.

Schematic Diagram of MAPK Signaling Pathway and Cell Cycle. (a) Signal Transduction in the MAPK Signaling Pathway. The genes highlighted in red are upregulated genes. (b) Gene Heat Map Related to the Cell Cycle.

Fig. 4b and Table S3 list the core regulatory genes of the yeast cell cycle and meiosis pathways. Analysis of differentially expressed genes indicates that a large number of genes associated with mitotic cell cycle processes show a significant upregulation trend. Concurrently, a subset of genes linked to meiosis-specific events also exhibit expression changes, collectively pointing to significant alterations in cell proliferation and development. At the onset of the mitotic cycle, the G1/S-specific cyclin CLN1 gene [35], which propels the cell cycle through the START checkpoint; the CDC6 gene [36], responsible for replication origin licensing and assembling the pre-replication complex; and the homologous box protein YOX1 gene [37], which provides transcriptional fine-tuning for G1 progression, all exhibit an upward trend. Together, they ensure the correct initiation of the cell division program. Regarding DNA replication and genomic stability maintenance, the CDC45 gene [38], a core component of the replication fork; the MRC1 gene, a mediator protein that monitors replication stress and activates checkpoint signaling; and the FOB1 gene [39], which prevents replication fork collisions in rDNA regions, are all significantly upregulated. Together, they safeguard the integrity and accuracy of genetic material replication. Following entry into the G2/M phase, the G2/M-specific cyclin genes CLB1 and CLB2, which drive progression into mitosis; the CDC5 gene encoding the Polo kinase [40], a core coordinating kinase regulating spindle assembly and chromosome separation; the YCG1 sub-unit of the condensin complex involved in prophase chromosome condensation [41]; and the SCC1 subunit of the cohesin complex, which maintains sister chromatid cohesion until late separation, all exhibit a marked upward trend. Together, they constitute the entire process for completing mitosis. Furthermore, the upregulation of the ERK and FAR1 genes indicates that the MAPK signaling pathway promotes the cell cycle process. During meiosis, multiple core regulatory elements similar to those in the cell cycle are equally active: the CLB1 gene driving meiotic spindle assembly and homologous chromosome separation, the CDC6 gene ensuring DNA replication licensing in meiotic S phase, and the CDC5 kinase gene acting throughout meiosis all play crucial roles. Specifically, the upregulation of the SGO1 and SPS4 genes during meiosis guarantees the successful progression of the entire process [42]. Overall, this series of upregulated genes related to the cell cycle and meiosis essentially covers the entire division process. This indicates that under specific conditions, yeast cells may precisely regulate their proliferation cycle or initiate the meiosis developmental program by coordinating the upregulation of these core regulatory factors. Simultaneously, they maintain connections with the MAPK signaling pathway, making adaptive adjustments at the metabolic and signal reception levels to address challenges posed by changes in the external environment.

3.5. Formic acid stress leads to upregulation of ribosome biogenesis

In S. cerevisiae, formic acid effect induces the upregulation of a series of ribosomal biogenesis-related genes (Fig. 5 and Table S4). These genes include members of the UTP-A complex involved in the early assembly of the ribosomal small subunit (e.g., UTP21, UTP4, IMP4, BMS1, and UTP9), members of the UTP-B complex (e.g., PWP2, UTP18, and UTP14), and members of the UTP-C complex (e.g., UTP6, UTP22, and UTP15), which collectively promote 18S rRNA processing and maturation of the 40S subunit [43,44]. Simultaneously, genes associated with the assembly and export of the large subunit (such as RIX7, LSG1, NUG2, and NOP4) are also upregulated to ensure the proper assembly and transport of the 60S large subunit [45,46]. Furthermore, the upregulation of rRNA modification enzyme genes (such as EMG1 and NAT10) facilitates post-transcriptional modification of rRNA [47], while SNU13 participates in the rRNA processing pathway. To match this enhanced assembly capacity and guarantee the structural integrity of newly synthesized ribosomes, a series of ribosomal structural protein genes are simultaneously upregulated. These include the large subunit structural proteins RP-L22e, RP-L6e, RP-L35Ae, and RP-L36e, which directly participate in the assembly and structural stabilization of the 60S large subunit. They assist in the correct folding of rRNA and maintain the conformation integrity during the translation elongation phase. Concurrently, the upregulation of small subunit structural proteins RP-S28e, RP-S27e, and RP-S9e ensure the assembly quality of the 40S small subunit. These proteins play a crucial role in translation initiation, influencing mRNA binding, scanning, and recognition of the start codon. The synergistic overexpression of these ribosomal proteins indicates that yeast cells enhance translation efficiency and accuracy under formate stress by strengthening ribosomal structural integrity and functional activity. This supports the rapid synthesis of proteins required for stress responses, thereby improving cellular adaptability and survival capacity in challenging environments.

Fig. 5.

Schematic Diagram of the Synthetic Components of Ribosomes and Eukaryotic Ribosomes. The genes highlighted in red are upregulated genes.

3.6. Formic acid stress leads S. cerevisiae to discard non-essential secondary metabolite synthesis pathways

Table S5 presents genes upregulated in the biosynthesis pathways of secondary metabolites in S. cerevisiae under formate stress. Specifically, the ELO3 gene involved in elongating fatty acid carbon chains was upregulated [48], indicating that yeast enhances membrane integrity by adjusting cell membrane lipid composition to resist formate-induced proton leakage and membrane fluidity disruption. This process directly impacts the synthetic flux of long-chain fatty acid-derived secondary metabolites. Concurrently, the expression changes in the key enzyme genes ERG5 (C-22 sterol desaturase) and ERG2 (C-8 sterol isomerase) within the sterol biosynthetic pathway further indicate that yeast dynamically adjusts the composition of sterol secondary metabolites to maintain membrane stability and potentially reduce reactive oxygen species damage [49,50]. This is because sterols are not only crucial structural components of the cell membrane, but their derivatives are also closely associated with stress signaling and antioxidant defense. On the other hand, changes in the expression of the SER2(serB) gene encoding phosphoserinase suggest adjustments in serine synthesis and one-carbon metabolism pathways [51]. This not only affects the supply of amino acid-derived secondary metabolites such as glycine and serine but also indirectly supports the synthesis of antioxidants like glutathione by regulating one-carbon unit cycling, thereby counteracting oxidative stress potentially induced by formate. Crucially, the expression response of the key enzyme gene DAL7(aceB) (malate synthase) in the glyoxylate cycle indicates that yeast may have enhanced acetyl-CoA/glyoxylate metabolic shunting under formate stress [52]. This not only reduces dependence on the glycolytic pathway for carbon but also supports amino acid synthesis, gluconeogenesis, and other secondary metabolic processes by generating intermediates such as oxaloacetic acid. glyoxylate metabolic diversion under formate stress. This approach reduces reliance on glycolysis for carbon supply while simultaneously supporting amino acid, gluconeogenic, and other secondary metabolic precursor synthesis through intermediates like oxaloacetic acid. Consequently, it achieves a rebalancing of energy supply and anabolic processes during carbon metabolic reprogramming. Collectively, the expression responses of these genes reveal that S. cerevisiae, under formate stress, reshapes its secondary metabolic networks through fine-tuned regulation of cell membrane component synthesis, antioxidant production, and carbon metabolic flux. This enables the yeast to maintain basic survival while enhancing its adaptability and tolerance to acidic environments.

Additionally, Table S6 displays genes downregulated in relation to the biosynthesis of secondary metabolites in S. cerevisiae. This is primarily manifested in a global deceleration of carbon metabolic flux, a weakening of the antioxidant defense system, and the coordinated downregulation of key metabolic nodes. Collectively, these changes reflect the cell's survival strategy adjustments under energy crisis and toxic stress.

Specifically, formic acid intervention severely disrupted core carbon metabolism while further extending to the synthesis and mobilization of storage polysaccharides, consistent with the preceding discussion. Additionally, cellular antioxidant and detoxification systems exhibit diminished function. Downregulation of genes encoding catalase CTA1(katE) [53], aldehyde dehydrogenase ALDH [54], and proline dehydrogenase PRODH [55] significantly weakens the cell's capacity to clear reactive oxygen species (ROS) and toxic aldehydes potentially induced by formate stress. Simultaneously, it may affect the metabolic turnover of compatible solutes such as proline, rendering cells more susceptible to oxidative damage. Furthermore, the downregulation of IDH1 [56] (isocitrate dehydrogenase) associated with mitochondrial metabolism and E4.1. 1.15 (malate synthase) further restricts the efficiency of the tricarboxylic acid cycle and the recycling capacity of C2 carbon sources (e.g., acetate), severing another critical supply line providing carbon skeletons and reducing power for secondary metabolism (e.g., sterols, fatty acid derivatives). Importantly, the downregulation of fatty acid β-oxidation-related genes ACAA1 (acyl-CoA acyltransferase) and ACOT1_2_4 (acyl-CoA thioesterase) suggests that cells may have reduced the process of utilizing fatty acids as alternative carbon sources to generate acetyl-CoA [57]. This directly impacts the supply of precursors for secondary metabolites such as polyketides and volatile esters. Meanwhile, the downregulation of BDH (butanediol dehydrogenase) may disrupt the balance of pyruvate metabolism and flavor compounds [58].

Overall, under formate stress, S. cerevisiae exhibits a significant systemic metabolic response. Cells enhance membrane lipid and sterol synthesis by upregulating genes such as ELO3, ERG2, and ERG5 to maintain membrane integrity. Concurrently, they reroute carbon flux via genes like DAL7(aceB) to support the supply of key precursors and antioxidant defenses. Concurrently, genes associated with glycolysis, the PPP cycle, polysaccharide metabolism, and antioxidant systems were broadly downregulated. This reflects the cells' active reduction of anabolic synthesis and energy expenditure, indicating that under stress conditions, yeast prioritizes resource allocation toward survival-related membrane structures and basal metabolism while significantly suppressing nonessential synthetic pathways, including those producing various secondary metabolites.

3.7. RT-qPCR validation of RNA-seq results and electrocatalytic CO₂ reduction coupled with S. cerevisiae for 3-HP production

To validate the authenticity of RNA-Seq results from transcriptome analysis, we randomly selected six genes from metabolic pathways: four upregulated genes (YHP1, CLB2, RPS9A, UTR2) and two downregulated genes (SPG1, FMP45). Primers are listed in Table S7. Using the ATC1 gene as an internal control, we performed RT-qPCR on each to verify gene expression levels. As shown in Fig. S6, the experimental group exhibited significant differences in gene expression levels. Notably, the high consistency in expression trends between RT-qPCR and transcriptome sequencing data demonstrates the stability and accuracy of the transcriptomic profiling in response to formic acid stress in S. cerevisiae. The differentially expressed genes verified in this study are closely associated with yeast cell cycle regulation, ribosomal synthesis, and stress resistance metabolism. Collectively, these consistent molecular expression patterns not only confirm the credibility of the transcriptome data but also provide reliable molecular evidence for revealing the adaptive response and regulatory mechanism of laboratory-evolved yeast strains under formic acid stress. On this basis, we have successfully established a preliminary process for the electrocatalytic reduction of CO₂ to ethanol, followed by the utilization of the ethanol by S. cerevisiae to produce 3-HP, as illustrated in Fig. 6a. This opens up a new pathway for the direct conversion of CO₂ into high-value chemicals. First, we carried out the reaction using a CuO_x catalyst in a flow cell we had constructed, and successfully reduced CO₂ to ethanol at a potential of −1.2 V (vs. RHE) with a faradaic efficiency of 74.7%. We measured the concentration of the by-product formic acid after 30 min of electrolysis, and the concentration of formic acid in the electrolyte was 1.57 ± 0.13 mM. The results of the gradient formic acid stress experiment on bacterial growth (Fig. S1) indicated that when the formic acid concentration was ≥ 0.1% (26.5 mM), the growth of the primary yeast almost completely stopped. Based on the formic acid accumulation rate, after continuous electrolysis for approximately 9 h, the formic acid concentration could reach this inhibitory critical value. This suggests that the continuously accumulated formic acid during CO₂ electroreduction can cause significant growth inhibition to the un-evolved primary yeast. During this period, the evolved strain has certain tolerance. At the same time, as shown in Table S8, we constructed the Pjfe3-MCRca, PRS425-ADH2 and PRS423-ALD6 plasmids and transformed them into S. cerevisiae to obtain the strains. The MCRca gene encodes malonyl-CoA reductase, which catalyzes the reduction of malonyl-CoA to 3-HP [59]. Meanwhile, ADH2 and ALD6 respectively encode alcohol dehydrogenase and aldehyde dehydrogenase to facilitate ethanol utilization [60]. As shown in Fig. 6b, to establish the 3-hydroxypropionic acid metabolic pathway, this study used strain 5B-2 as the parental strain to prepare competent cells. A series of recombinant strains were constructed using the lithium acetate transformation method, and positive clones were screened using nutrient-deficient media: strain 5B-3, which expresses only MCRca, was obtained; Based on 5B-3, ADH2 and ALD6 were overexpressed individually to yield strains 5B-4 and 5B-5, respectively; ALD6 was further overexpressed in 5B-4 to construct strain 5B-6. When cultured for 96 h in a sugar-free nutrient-deficient medium with 0.43 M ethanol as the sole carbon source, the 3-HP yields of the aforementioned strains were 18.44, 50.83, 47.31 and 98.5 mg/L, respectively, as shown in Fig. 6c. This represents a five-fold increase in 3-HP yield compared to the 5B-2 strain.

Fig. 6.

Electrocatalytic reduction of CO2 coupled with S. cerevisiae metabolism. (a) Coupling process schematic diagram. (b) S. cerevisiae 3-HP metabolic pathway. (c) The yield of 3-HP from S. cerevisiae metabolism.

4. Conclusion

In this study, RNA-seq transcriptomic analysis was employed to investigate the yeast response mechanism of S. cerevisiae strain CEN.PK 102-5B under formic acid effect with ethanol as substrate. Results indicate that in core carbon metabolism, glycolysis, the PPP cycle, and polysaccharide synthesis are extensively downregulated, impairing energy supply. Meanwhile, the glyoxylate cycle—a key pathway in ethanol metabolism—is compensatorily upregulated to replenish the carbon skeleton. Simultaneously, key regulatory genes of the cell cycle and meiosis (such as CLN1, CDC6, and CLB1/2) are significantly upregulated and interact with the MAPK signaling pathway to drive the division process. Ribosomal biogenesis is comprehensively activated to enhance the translation capacity of stress response proteins. Additionally, membrane lipid and sterol synthesis (ELO3, ERG2/5) were upregulated to maintain membrane integrity, while antioxidant and fatty acid oxidation pathways were downregulated. This indicates that yeast enhances adaptation to acidic environments by redistributing resources to prioritize membrane stability and ethanol metabolic flux, while suppressing non-essential synthesis. Finally, we preliminarily established an iEMS that couples the electrocatalytic reduction of CO₂ to ethanol with the subsequent conversion of ethanol to 3-HP by S. cerevisiae. This approach opens a new avenue for the direct conversion of CO₂ into value-added chemicals and holds dual significance for promoting green carbon emission reduction and sustainable energy storage.

CRediT authorship contribution statement

Jin Liu: Writing – original draft, Data curation, Methodology. Weilin Bi: Formal analysis, Conceptualization. Miao Liu: Formal analysis, Data curation. Shang Wang: Formal analysis. Xiangrui Wu: Formal analysis. Yuanhong Xu and Jin Hou: Resources, Supervision. All authors edited the manuscript before submission. All authors have read and agreed to the published version of the manuscript.

Data availability statement

The RNA-seq data reported in the current study are available in the NCBI SRA database (Bioproject: PRJNA1441775).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: