Skip to main content
NCBI home page
Frontiers in Microbiology logoLink to Frontiers in Microbiology
. 2026 Sep 14;17:1934502. doi: 10.3389/fmicb.2026.1934502

An unfolded protein response-inducible protein (ULI1) mediates aromatic alcohols biosynthesis through reinforcing the Ehrlich pathway in Saccharomyces cerevisiae

Xiaowei Gong 1,†, Yong Shen 2,†, Shanzhai Shang 1, Jianbo Zhan 1, Guanghui Ma 3, Huajun Luo 4, Shiwei Li 1, Chunxia Song 4, Donglai Zhu 1,*, Yiyong Luo 2,*
PMCID: PMC13617065  PMID: 42807480

Abstract

Aromatic alcohols such as 2-phenylethanol (2-PE) and tryptophol (TOL) possess diverse biological functions. Here, an unfolded protein response-inducible gene ULI1 and an aldehyde dehydrogenase gene ALD3 in Saccharomyces cerevisiae were disrupted using the CRISPR-Cas9 system. Fermentation assays revealed that both the ULI1 frameshift (Kmuli1) and knockout (KΔuli1) mutants exhibited significantly reduced 2-PE and TOL yields relative to the wild-type strain KMLY1-6, retaining 46.32 and 64.54% for 2-PE, and 34.61 and 77.73% for TOL, respectively. In contrast, ALD3 knockout moderately elevated 2-PE and TOL production by 4.98 and 20.57%, respectively. These results indicate that both ULI1 and ALD3 mediate the biosynthesis of 2-PE and TOL in S. cerevisiae. To further confirm the role of ULI1, three chromosomal integration expression (CIE) strains (KXuli1, KXIIuli1, KΔald3uli1) and one plasmid-based expression (PBE) strain (BY4741/pY26-uli1) were constructed. Compared with the control KMLY1-6, the CIE strains increased 2-PE and TOL production by 10.13–12.91% and 25.76–29.28%, respectively. Meanwhile, the PBE strain produced significantly higher levels of 2-PE and TOL than the empty vector control (BY4741/pY26), with maximum increases of 38.80% and 30.86% detected at 60 h of cultivation. To further elucidate the molecular mechanism of ULI1 in 2-PE and TOL biosynthesis, comparative transcriptomic analysis for Kmuli1 vs. KMLY1-6, together with transcriptomic and proteomic analyses for BY4741/pY26-uli1 vs. BY4741/pY26, was performed. Multi-omics results revealed that ULI1 reinforces the Ehrlich pathway, modulates metabolic flux, maintains mitochondrial structure and cellular protein homeostasis, and coordinates stress responses with metabolic redistribution. Collectively, ULI1 acts as a novel regulatory hub linking stress tolerance and cellular homeostasis, and facilitates aromatic alcohol biosynthesis by strengthening the Ehrlich pathway in S. cerevisiae.

Keywords: 2-phenylethanol, gene editing, proteomics, Saccharomyces cerevisiae, transcriptomics, tryptophol, unfolded protein response

1. Introduction

Aromatic alcohols, a distinct class of aromatic compounds, are predominantly biosynthesized in natural systems via the metabolic transformation of aromatic amino acids. These compounds possess great economic value and research potential due to their versatile biological functions. For instance, 2-phenylethanol (2-PE), a metabolite with a pleasant rose-like fragrance, is widely used in the food, cosmetics, and tobacco industries (Hua and Xu, 2011; Bernardino et al., 2024). Additionally, it acts as a quorum-sensing molecule in yeast (Chen and Fink, 2006) and exhibits broad-spectrum antibacterial and antifungal activities (Fraud et al., 2003; Mo and Sung, 2007; Wang et al., 2020), thus garnering considerable research attention. Tryptophol (TOL), another representative aromatic alcohol, also boasts multiple biological functions, including the inhibition of cytokine production, the induction of cell apoptosis (Inagaki et al., 2007; Schirmer et al., 2016), the promotion of plant growth (Elleuch et al., 2010), and the exertion of antiviral and antipathogenic properties (Zhu and Jin, 2015), underscoring its research potential and broad application prospects.

Nowadays, aromatic alcohols are primarily obtained either through chemical synthesis or extraction from plant materials. However, the chemical approach is environmentally unfriendly and often yields lower-grade products, whereas the extraction method suffers from high costs and unstable supply (Xu et al., 2017; Dubhashe et al., 2018; Yan et al., 2021; Wang et al., 2025). In response to these challenges, biosynthesis has attracted growing interest due to its eco-friendly nature and promising industrialization potential. Saccharomyces cerevisiae, a generally recognized as safe (GRAS) food-grade microorganism, is an ideal chassis cell for synthetic biology. Aromatic alcohols are biosynthesized in this strain via two metabolic pathways. The first is the Ehrlich pathway, which initiates from aromatic amino acids and proceeds through three enzymatic steps: (1) transamination to 2-oxoacids catalyzed by ARO8 and ARO9, (2) decarboxylation to aromatic aldehydes by ARO10, PDC1, PDC5 and PDC6, and (3) reduction to aromatic alcohols by ADH1–ADH5 and SFA1 (Dickinson et al., 2003; Gu et al., 2020; Liu et al., 2024). The second route is the de novo pathway, which converts glucose to aromatic alcohols through the following steps: (1) phosphoenolpyruvate (from glycolysis) and erythrose 4-phosphate (from pentose phosphate pathway) are condensed into 3-deoxy-D-arabino-heptulosonate-7-phosphate (DAHP) under the catalysis of ARO3 and ARO4; (2) DAHP is sequentially converted to chorismate through reactions catalyzed by the pentafunctional arom protein ARO1 and chorismate synthase ARO2; (3) for 2-PE biosynthesis, chorismate is converted to prephenate by ARO7, then to phenylpyruvate by PHA2, and finally enters the Ehrlich pathway for decarboxylation and reduction; for TOL biosynthesis, chorismate is transformed into tryptophan by enzymes TRP1–TRP5, before proceeding through the Ehrlich pathway (Hassing et al., 2019; Guo et al., 2020; Wu et al., 2024).

In S. cerevisiae, besides the core catalytic enzymes, the aromatic alcohol biosynthesis is also regulated by pathway-related regulatory proteins. For instance, transcriptional activator ARO80 and GATA factors (e.g., GLN3, GAT1) modulate aromatic alcohol biosynthesis by upregulating ARO9 and ARO10 gene expression (Cooper, 2002; Dai et al., 2021). The decarboxylation reaction in the Ehrlich pathway relies on thiamine pyrophosphate (TPP) as a coenzyme (Dai et al., 2021), meaning thiamine regulon genes (THIs) indirectly control aromatic alcohol production. This is supported by our previous work, where overexpression of the THI4 gene significantly increased yields of 2-PE and TOL in S. cerevisiae (Gong et al., 2022). In addition, the reduction step in the Ehrlich pathway requires NADH. Lipoamide dehydrogenase (LPD1) catalyzes the oxidative regeneration of dihydrolipoamide coupled with NAD+ reduction to NADH. Thus, reduced 2-PE and TOL yields in the LPD1 mutant may be linked to NADH availability (Dickinson et al., 2003).

The unfolded protein response (UPR) can be induced by endoplasmic reticulum (ER) stress (Read and Schröder, 2021; Hrach et al., 2023; Hong et al., 2024). ER stress arises from stimuli such as Ca2+ depletion, hypoxia, altered glycosylation, and viral infection, and ultimately results in the accumulation of unfolded or misfolded proteins in the ER lumen. In S. cerevisiae, the UPR is mediated exclusively through the IRE1 (Inositol-requiring enzyme 1) pathway (Read and Schröder, 2021). IRE1, an ER-localized type I transmembrane protein, functions as a key sensor for both ER stress and ER bilayer stress (Adams et al., 2019; Ho et al., 2020). Upon sensing either stress type, IRE1 is activated and acquires endonuclease activity, which enables it to specifically splice HAC1 mRNA and generate mature HAC1 protein (Cox et al., 2011). Mature HAC1 then binds to unfolded protein response elements in the promoters of UPR target genes, driving their expression to support processes including protein folding, degradation, and secretion (Halbleib et al., 2017; Read and Schröder, 2021). Within cells, a successful UPR restores normal protein folding homeostasis, while failure triggers cellular apoptosis (Read and Schröder, 2021). To date, no studies have reported an association between UPR and aromatic alcohol biosynthesis.

In this study, an unfolded protein response-inducible gene ULI1 was disrupted and heterologously expressed in S. cerevisiae, revealing its involvement in the biosynthesis of 2-PE and TOL. Using transcriptomics and proteomics, we further elucidated the molecular mechanisms by which ULI1 regulates aromatic alcohol biosynthesis, aiming to provide theoretical support and genetic resources for efficient production of aromatic alcohols in S. cerevisiae. ULI1 overexpression may enhance aromatic alcohol biosynthesis in industrial fermentation. Nevertheless, critical factors governing strain stability and scaled production require ongoing investigation.

2. Materials and methods

2.1. Strains and culture conditions

The strains used in this study are listed in Supplementary Table S1. The wild-type S. cerevisiae strain KMLY1-2 (isolated from homemade sourdough; Gong et al., 2022) and its MATa haploid derivative KMLY1-6 (obtained by ascospore selection) were cultured in YPD medium (1% yeast extract, 2% peptone, 2% dextrose). S. cerevisiae strain BY4741 was maintained in synthetic complete dropout medium without uracil (SC-URA; Coolaber, Beijing, China). For 2-PE and TOL production, all yeast strains were cultured in TM medium (30 g/L glucose, 5 g/L KH2PO4, 0.5 g/L MgSO4, 1 g/L NaCl) supplemented with 1.75 g/L L-phenylalanine (TM-1.75P) or 0.6 g/L L-tryptophan (TM-0.6T), respectively. Escherichia coli strain DH5α, used for plasmids construction, was grown in LB medium (1% tryptone, 0.5% yeast extract, 1% NaCl). Solid media were prepared by adding 2% agar to the corresponding broth. When required, antibiotics were supplemented as follows: 100 μg/mL ampicillin for E. coli, and 200 μg/mL geneticin plus 100 μg/mL nourseothricin for S. cerevisiae.

2.2. Gene sequence and phylogenetic analyses

Based on our previous RNA-seq data (Gong et al., 2022), two differentially expressed genes (DEGs; Supplementary Table S2), namely the unfolded protein response-inducible gene ULI1 and the aldehyde dehydrogenase gene ALD3, were selected to investigate their effects on 2-PE and TOL production. The nucleotide sequences of ULI1 and ALD3 were obtained from the KMLY1-2 genome (Gong et al., 2022). Their deduced amino acid sequences, theoretical isoelectric points (pI), and molecular weights were predicted using EditSeq (DNASTAR, Madison, United States). Conserved functional domains in ULI1 and ALD3 proteins were identified using InterPro.1 SignalP 5.0, TargetP 2.0 and TMHMM 2.02 were used to detect signal peptides, subcellular localization, and transmembrane helices in ULI1 and ALD3, respectively. Homologous proteins of ULI1 and ALD3 from different yeast species were downloaded from NCBI3 and aligned using Clustal X. A neighbor-joining tree was constructed using the MEGA 6.0 software package (Mega Limited, Auckland, New Zealand).

2.3. Genes disruption by CRISPR-Cas9 system

The oligonucleotides used in this study are listed in Supplementary Table S3. All gene disruption mutants were generated using the CRISPR-Cas9 gene editing system (Jessop-Fabre et al., 2016). To disrupt the ULI1 gene, a single-guide RNA (sgRNA) targeting ULI1 was designed using CHOPCHOP,4 and two complementary oligonucleotides (Sangon Biotech, Shanghai, China) containing a 20-bp spacer flanked by 15-bp homologous arms were annealed to form the dsDNA fragment. This fragment was cloned into the linearized pCfB3052 vector, amplified using primers 3052-F and 3052-R, via In-Fusion assembly with the ClonExpress MultiS One Step Cloning Kit (Vazyme Biotech, Nanjing, China), generating plasmid pCfB3052-uli1 (Supplementary Figure S1). The donor DNA for homologous recombination was constructed by overlap PCR using the ULI1 upstream and downstream arms amplified from KMLY1-6 genomic DNA with primer pairs uli1-uF/uli1-uR and uli1-dF/uli1-dR. The Cas9-expressing plasmid pCfB2312 was co-transformed into KMLY1-6 with either pCfB3052-uli1 alone or pCfB3052-uli1 plus donor DNA. Positive transformants selected on geneticin and nourseothricin were verified by PCR and sequencing, ultimately obtaining the ULI1 frameshift mutant Kmuli1 and knockout mutant KΔuli1. For ALD3 disruption, two sgRNAs targeting ALD3 were designed, and three cassettes containing two 20-bp spacers (Cassette I, amplified from pCfB3052-uli1 with F1/ald3-R1; Cassette II amplified from pCfB3052 with ald3-F2/ald3-R2; Cassette III, amplified from pCfB3052-uli1 with ald3-F3/R3) were In-Fusion cloned into the linearized pCfB3052 backbone, which was prepared using primers 3052c-F/3052c-R, generating plasmid pCfB3052-ald3 (Supplementary Figure S1). Donor DNA was constructed by combining the upstream and downstream arms of the ALD3 gene, amplified from KMLY1-6 genomic DNA using primer sets ald3-uF/ald3-kuR and ald3-kdF/ald3-dR, respectively. Following co-transformation of KMLY1-6 with pCfB2312, pCfB3052-ald3, and the donor DNA, candidate colonies were isolated and verified by PCR, generating the ALD3 knockout strain KΔald3.

2.4. Construction of ULI1 chromosomal integration expression (CIE) strains

Jessop-Fabre et al. (2016) identified multiple genomic loci suitable for gene integration and expression in S. cerevisiae strain CEN.PK113-7D. The corresponding sites (X-3 and XII-3) in KMLY1-6, along with the ALD3 gene, were selected as integration targets for ULI1. Plasmids pCfB3052-X-3 and pCfB3052-XII-3 (targeting X-3 and XII-3 loci, respectively) were constructed analogously to pCfB3052-ald3 but using different primers (F1/X-3-R1 and F1/XII-3-R1 for cassette I; X-3-F2/X-3-R2 and XII-3-F2/XII-3-R2 for cassette II; X-3-F3/R3 and XII-3-F3/R3 for cassette III) as listed in Supplementary Table S3. To assemble the ULI1 expression cassette, the ULI1 open reading frame (ORF) was PCR-amplified, digested with EcoRI and XhoI, and ligated into similarly digested pY26TEF-GPD, yielding pY26-uli1, from which the cassette was amplified using primers Uec-F and Uec-R. The upstream and downstream arms of X-3, XII-3, and ALD3 were amplified using the primer pairs X-3-uF/X-3-uR and X-3-dF/X-3-dR, XII-3-uF/XII-3-uR and XII-3-dF/XII-3-dR, and ald3-uF/ald3-uR and ald3-dF/ald3-dR, respectively. Three donor DNAs were generated by overlap PCR, each fusing the ULI1 cassette to flanking arms of either X-3, XII-3, or ald3. Finally, three ULI1 CIE strains comprising KXuli1 (uli1 integrated at X-3), KXIIuli1 (ULI1 integrated at XII-3), and KΔald3uli1 (uli1 replacing ald3 ORF) were constructed following transformation, selection, and verification procedures equivalent to those used for KΔald3.

2.5. Construction of ULI1 plasmid-based expression (PBE) strains

The plasmid pY26-uli1 was introduced into the yeast strain BY4741 by electroporation, and the ULI1 PBE strain (BY4741/pY26-uli1) was selected on SC-URA plates. Strain BY4741 harboring the empty plasmid pY26TEF-GPD (BY4741/pY26) served as the control.

2.6. Determination and statistical analysis of 2-PE and TOL production

Frozen stock cultures of KMLY1-6 and its engineered derivatives were revived by incubation in 5 mL YPD medium at 30 °C and 150 rpm for 12 h, followed by inoculation (1% v/v) into 50 mL of fresh YPD medium for further cultivation under identical conditions. In parallel, strains BY4741/pY26 and BY4741/pY26-uli1 were pre-cultured and subsequently scaled up in SC-URA medium using the same growth parameters as KMLY1-6. All cells were harvested by centrifugation at 10,000 rpm for 5 min, washed twice with sterile ddH2O, and resuspended. The resulting suspension was inoculated into TM-1.75P and TM-0.6T media to an OD600 of 0.15, then cultivated for 48 h with 150 rpm shaking at 30 °C. Cell-free supernatant (CFS) was prepared by centrifugation (12,000 rpm, 10 min, 4 °C) followed by sterile filtration through a 0.22-μm filter (Millipore, Darmstadt, Germany), and was used to determine 2-PE and TOL content using high-performance liquid chromatography (HPLC; Agilent, Santa Clara, United States) according to Gong et al. (2022). To analyze the time course of 2-PE and TOL accumulation, the engineered strains Kmuli1 and BY4741/pY26-uli1, along with their controls, were cultured for 60 h. The concentrations of 2-PE and TOL in the CFSs were measured at 12-h intervals. Each experiment was conducted in at least triplicate and the data were expressed as mean ± standard deviation. Student’s t-test was performed using GraphPad Prism 10.1 (San Diego, United State), and significance was set at p < 0.05.

2.7. RNA isolation and RNA-seq analysis

Strains KMLY1-6, Kmuli1, BY4741/pY26 and BY4741/pY26-uli1 were cultured in TM-1.75P and TM-0.6T media for 48 h as described above, generating the following samples: K1.75P and K0.6T (for KMLY1-6), M1.75P and M0.6T (for Kmuli1), B1.75P and B0.6T (for BY4741/pY26), and U1.75P and U0.6T (for BY4741/pY26-uli1). Cells from these samples were harvested by centrifugation, and total RNA was immediately extracted using the Trizol reagent kit (Invitrogen, Carlsbad, United State). With three biological replicates per condition, a total of 24 high-quality RNAs were obtained, which were then used to construct RNA-seq libraries following the manufacturer’s protocol of the TruSeq RNA sample preparation kit (Illumina, San Diego, United State). The libraries were sequenced on an Illumina HiSeq 4000 system to generate 100 nt paired-end raw reads. Clean reads were obtained using a Perl program following the criteria described by Wu et al. (2020) and subsequently mapped to the KMLY1-2 genome with HISAT2 (Kim et al., 2015) under default parameters. Aligned reads were quantified using the fragment per kilobase of transcript per million mapped reads (FPKM) method (Pertea et al., 2016). DEGs between two groups were identified using DESeq2 (Love et al., 2014), with filtering thresholds set as | log2(fold change)| ≥ 1 and adjusted p-value < 0.05. The DEGs were classified and enriched by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotation. Heatmaps of gene and protein expression profiles were generated using R package.

2.8. Proteomic analysis

Total proteins from samples B1.75P, B0.6T, U1.75P, and U0.6T were extracted and trypsinized following the method described by He et al. (2020). Then, the peptides were desalted and analyzed using liquid chromatography-tandem mass spectrometry (LC-MS/MS; ThermoFisher Scientific, Waltham, United State) at Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China). Briefly, the peptides were separated using a VanquishNeo instrument equipped with a uPAC High Throughput column (75 μm × 5.5 cm). A gradient elution was performed with mobile phases A (2% acetonitrile and 0.1% formic acid) and B (80% acetonitrile and 0.1% formic acid), using the program described in Supplementary Table S4. Data-independent acquisition (DIA) data were acquired using an Orbitrap Astral mass spectrometer operated in DIA mode. The detection was carried out over a mass range of 380–980 m/z (MS1), and 150–2000 m/z (MS2). The MS raw data were searched using Xcalibur 4.7 (ThermoFisher Scientific) against the UniProt-Saccharomyces cerevisiae database with the following parameters: protein false discovery rate (FDR) ≤ 0.01, peptide FDR ≤ 0.01, peptide confidence ≥ 99%, and extracted ion chromatogram (XIC) width ≤ 75 ppm. Differentially expressed proteins (DEPs) between groups were identified using the R package “t-test” with thresholds of fold change ≥ 1.2 or ≤ 0.83 and p-value < 0.05. These DEPs were further subjected to GO and KEGG enrichment analyses.

3. Results and discussion

3.1. Bioinformatic characterization of the ULI1 and ALD3 genes

The predicted ORFs of the ULI1 and ALD3 genes are 510 bp and 1521 bp in length, encoding polypeptides of 169 and 506 amino acid residues, respectively. The theoretical molecular weight and pI of ULI1 are 18.81 kDa and 6.35, while those of ALD3 are 55.42 kDa and 5.54 (Table 1). ULI1 contains a signal peptide and transmembrane (SP-TM) domain, with residues 1–19 comprising the SP sequence and residues 5–27 forming a TM helix (Figure 1A). This indicates that ULI1 is a membrane-bound protein, which is consistent with the previous finding that ULI1 may localize to the ER (Yofe et al., 2016). In contrast, ALD3 contains an ALDH-like domain (Figure 1A) and is predicted to be a cytosolic protein. Phylogenetic analysis showed that the homologous ULI1 and ALD3 from different yeasts were separated into two clades (Figure 1B). Notably, ULI1 homologs are found exclusively within the genus Saccharomyces, and exhibit significantly greater genetic divergence among themselves than do ALD3 homologs. This suggests that ULI1 is a genus-specific protein, and its evolutionary mechanism requires further investigation.

TABLE 1.

Characteristics of ULI1 and ALD3 proteins.

Proteins Molecular weight (kDa) pI
ULI1 18.81 6.35
ALD3 55.42 5.54
Open in a new tab

FIGURE 1.

Two-panel figure showing Saccharomyces gene homology and phylogeny. Panel A diagrams domain structure for ULI1 and ALD3 proteins with color-coded boxes labeled SP_TM and ALDH-like, including residue numbers. Panel B features a phylogenetic tree of ULI1 and ALD3 homologs from various Saccharomyces and closely related yeasts, with bootstrap values indicated at major nodes and selected strain names in red text.

Open in a new tab

Domain organization and evolutionary analysis of ULI1 and ALD3 in yeasts. (A) Domain structure of ULI1 and ALD3 from KMLY1-2. Domain positions and counts are schematically shown (not to scale). (B) Neighbor-joining phylogenetic tree of ULI1 and ALD3 homologs. The numbers adjacent to the branches indicate the percentages at which given branches were supported in 1,000 bootstrap replications. GenBank accession numbers are displayed after the species name. KMLY1-2 ULI1 and ALD3 are highlighted in red.

3.2. Involvement of ULI1 in 2-PE and TOL biosynthesis revealed by gene disruption assays

To test whether ULI1 and ALD3 are involved in aromatic alcohol biosynthesis, ULI1 and ALD3 of KMLY1-6 were disrupted by using CRISPR-Cas9 gene editing system. Sequence alignment analysis identified a single thymine (T) insertion in the ULI1 gene between the 100th and 101st base pairs, which corresponds to the fifth nucleotide upstream of the PAM site (Supplementary Figure S2A). This insertion caused a frameshift mutation, resulting in a truncated 42-amino-acid peptide where 34 residues are identical to those encoded by the ULI1 gene in the wild-type strain. This mutation stems from the repair of a Cas9-induced genomic double-strand break (DSB) via non-homologous end joining, an error-prone mechanism characterized by its propensity to introduce small indels (Chen et al., 2019), which directly resulted in the thymine insertion observed in the Kmuli1 strain. Additionally, ULI1- and ALD3- knockout mutants (KΔuli1 and KΔald3) were constructed using homology-directed recombination (HDR) with donor DNAs containing upstream and downstream homologous arms (Supplementary Figures S2B–D,G). One strain of each mutant was analyzed for aromatic alcohol production (Figures 2a,b). Disruption of ULI1 significantly (p < 0.01) reduced 2-PE and TOL production. Kmuli1 and KΔuli1 produced 46.32 and 64.54% of the control (KMLY1-6) 2-PE levels, and 34.61 and 77.73% of its TOL levels, respectively. In contrast, ALD3 knockout increased 2-PE and TOL production by 4.98 and 20.57%, respectively. These results indicate that ULI1 and ALD3 mediate the biosynthesis of 2-PE and TOL in S. cerevisiae. Current knowledge regarding ULI1 function is limited. Existing literature indicates that ULI1 localizes to the ER and is involved in and induced by the ER unfolded protein response (UPRER) (Metzger and Michaelis, 2009; Yofe et al., 2016). Given that TM-1.75P and TM-0.6T are nutrient-poor media, S. cerevisiae cultured in these media inevitably encounters starvation stress, which is likely to trigger the UPRER. Thus, ULI1 may be a UPRER component that influences the biosynthesis of 2-PE and TOL by aiding in maintaining cellular homeostasis and regulating cellular metabolism. The lower production of 2-PE and TOL in Kmuli1 compared to KΔuli1 (Figures 2a,b) may be related to the extent of ULI1 disruption. Complete deletion in KΔuli1 might activate intracellular proteins functionally redundant to ULI1, conferring better cellular homeostasis than in Kmuli1, which retains a 34-amino-acid peptide. Regarding ALD3, it is proposed to function in a competing branch pathway by oxidizing the key aldehydic precursors (phenylacetaldehyde/indoleacetaldehyde) to their corresponding acids, thereby diverting flux away from the ADH-mediated production of 2-PE and TOL. This model explains the observed increase in these aromatic alcohols upon ALD3 knockout. Since the ULI1 frameshift exerts the most pronounced effect, Kmuli1 was selected for production dynamics analysis. As shown in Figures 2C,D, 2-PE and TOL levels increased gradually with time, but KMLY1-6 consistently had much higher yields than Kmuli1, with differences of 378.04–697.78 mg/L (2-PE) and 180.31–299.58 mg/L (TOL). These results further support ULI1’s involvement in 2-PE and TOL biosynthesis.

FIGURE 2.

Figure containing four panels compares 2-phenylethanol (2-PE) and TOL concentrations among four yeast strains and over time. Panels A and B are bar graphs showing strain comparisons for 2-PE (green) and TOL (blue) concentrations, respectively, with statistical significance indicated. Panels C and D are line graphs displaying 2-PE and TOL concentrations over time for strains KMLY1-6 and Kmuli1, showing higher production in KMLY1-6.

Open in a new tab

Effects of gene disruption on 2-phenylethanol (2-PE) and tryptophol (TOL) production. (A) 2-PE concentration of ULI1 and ALD3 disruption strains in TM-1.75P medium. (B) TOL concentration of ULI1 and ALD3 disruption strains in TM-0.6T medium. (C) 2-PE production kinetics curves of KMLY1-6 and ULI1 mutant in TM-1.75P medium. (D) TOL production kinetics curves of KMLY1-6 and ULI1 mutant in TM-0.6T medium. Error bars denote the standard deviation from three biological replicates. p-values were calculated using Student’s t-test (**p < 0.01, ***p < 0.001).

3.3. Involvement of ULI1 in 2-PE and TOL biosynthesis confirmed by gene expression analysis

Given its stronger impact on 2-PE and TOL production than ALD3 (Figures 2a,b), ULI1 was selected for expression analysis. Using HDR, donor DNAs containing the ULI1 cassette flanked by homologous arms were integrated into the X-3, XII-3 and ALD3 loci of the KMLY1-6 chromosome. This resulted in size shifts of the corresponding PCR products from 2,782 to 3,262, 2,383 to 2,644, and 2,822 to 2,750 bp, respectively (Supplementary Figures S2D–G), confirming the successful construction of three ULI1 CIE strains (KXuli1, KXIIuli1, KΔald3uli1). Yield analysis showed that ULI1 overexpression at the X-3, XII-3 and ALD3 loci increased 2-PE production by 12.91, 11.83, and 10.13%, and TOL production by 29.28, 28.44, and 25.76%, respectively (Figures 3a,b). These results indicate ULI1 is involved in 2-PE and TOL biosynthesis in a ULI1 abundance-dependent manner. The greater yield improvement of KΔald3uli1 compared to KMLY1-6 relative to KΔald3 is attributed to the combined effect of ULI1 overexpression and ALD3 knockout in KΔald3uli1. To verify whether ULI1 plays the same role in the model S. cerevisiae strain, it was expressed in BY4741 using a plasmid-based system. As shown in Figures 3C,D, the yields of 2-PE and TOL increased with cultivation time. After 48 h of cultivation, the yields of 2-PE and TOL in BY4741/pY26-uli1 were significantly higher (p < 0.01) than those in the control strain BY4741/pY26, with respective increases of 38.80 and 30.86% observed at 60 h. These results further confirm that the involvement of ULI1 in 2-PE and TOL biosynthesis in an abundance-dependent manner is applicable across different S. cerevisiae strains.

FIGURE 3.

Panel A shows a bar graph comparing 2-PE concentrations among four strains, with significant differences indicated by asterisks. Panel B displays a bar graph of TOL concentrations for the same strains, also highlighting significant differences. Panel C presents a line graph of 2-PE concentration over time for two groups, showing higher levels in BY4741/pY26-uli1. Panel D contains a line graph of TOL concentration over time for the same groups, with BY4741/pY26-uli1 exhibiting higher concentrations. Data points and error bars are included throughout.

Open in a new tab

Effects of gene expression on 2-phenylethanol (2-PE) and tryptophol (TOL) production. (A) 2-PE concentration of ULI1 chromosomal integration expression strains in TM-1.75P medium. (B) TOL concentration of ULI1 chromosomal integration expression strains in TM-0.6T medium. (C) 2-PE production kinetics curves of BY4741/pY26 and BY4741/pY26-uli1 in TM-1.75P medium. (D) TOL production kinetics curves of BY4741/pY26 and BY4741/pY26-uli1 in TM-0.6T medium. Error bars denote the standard deviation from three biological replicates. p-values were calculated using Student’s t-test (**p < 0.01, ***p < 0.001).

3.4. Transcriptome sequencing and DEGs analysis of KMLY1-6 ULI1 disruption

Given that the ULI1 frameshift mutation resulted in the most significant decrease in 2-PE and TOL production, Kmuli1 and the wild-type strain KMLY1-6 were selected for comparative transcriptomics analysis. Principal component analysis (PCA) revealed that all samples clearly formed four distinct groups, with tight clustering of the three intra-group replicates, indicating high reproducibility (Figure 4A). Notably, groups K0.6T and M0.6T were proximal, while both were distant from K1.75P and M1.75P (which also separated from each other), suggesting that medium TM-1.75P had a more pronounced effect on the transcriptome than TM-0.6T. Following the criteria described in the methods, 3078 (486 upregulated, 2592 downregulated) and 212 (97 upregulated, 115 downregulated) DEGs were identified in the comparisons M1.75P vs. K1.75P and M0.6T vs. K0.6T, respectively (Figures 4B,C). Among these, 112 DEGs were shared between the two comparisons (Figure 4C), of which 75 displayed consistent up- or down-regulation patterns in both comparisons (Supplementary Table S5). These shared and consistently regulated DEGs may be closely associated with the observed reduction in both 2-PE and TOL production due to the ULI1 frameshift mutation, and were selected for GO and KEGG enrichment analyses. As shown in Figure 4D, eight of the top 10 enriched GO terms were categorized as biological processes, among which the thiamine-containing compound metabolic process was the most enriched. Correspondingly, thiamine metabolism was identified as the most significantly enriched KEGG pathway (Figure 4E and Supplementary Table S5). In our previous study, we proved that thiamine regulon genes (e.g., THI4) were involved in TOL production (Gong et al., 2022). These findings collectively indicate that ULI1 likely regulates 2-PE and TOL biosynthesis by modulating thiamine metabolism.

FIGURE 4.

Panel A shows a PCA scatter plot visualizing group clustering by PC1 and PC2. Panel B presents a bar chart of differentially expressed gene counts, separated by up and down regulation across two comparisons. Panel C features a Venn diagram quantifying unique and shared DEGs between the two conditions. Panel D displays a horizontal bar chart of gene ontology (GO) terms with color intensity representing significance. Panel E provides a bubble chart summarizing enriched pathways, with bubble size indicating gene number and color denoting statistical significance. All panels collectively illustrate statistical differences in gene expression and pathway enrichment among experimental groups.

Open in a new tab

Statistics and enrichment analysis of differentially expressed genes (DEGs) between Kmuli1 and KMLY1-6 from transcriptome data. (A) Principal component analysis (PCA) derived score plot. (B) Number of DEGs. (C) Venn diagram of DEGs. (D) Top 10 Gene Ontology (GO) terms of shared DEGs with consistent up- or down-regulation in M1.75P vs. K1.75P and M0.6T vs. K0.6T. Pink terms denote biological processes and green those represent molecular functions. (E) Top 10 Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment pathways of shared DEGs with consistent up- or down-regulation in M1.75P vs. K1.75P and M0.6T vs. K0.6T. Three biological replicates were set for every sample (n = 3).

3.5. Transcriptome sequencing and DEGs analysis of BY4741 ULI1 expression

To further characterize the function of the ULI1 gene in yeast, comparative transcriptomic analysis was performed using strains BY4741/pY26 and BY4741/pY26-uli1. PCA revealed four clearly separated clusters, with tight grouping among three biological replicates, demonstrating high reproducibility (Figure 5A). The distinct separation between groups suggested that both culture media (TM-0.6T and TM-1.75P) and ULI1 over-expression significantly influenced the transcriptomic profile. According to the thresholds described in the methods, 917 DEGs (167 upregulated, 750 downregulated) and 1002 DEGs (88 upregulated, 914 downregulated) were identified in the U1.75P vs. B1.75P and U0.6T vs. B0.6T comparisons, respectively (Figure 5B). Among these DEGs, 540 were shared between the two comparisons (Figure 5C), and all exhibited consistent expression trends (Supplementary Table S6). These common and stably regulated DEGs were likely closely related to the increased 2-PE and TOL production caused by the ULI1 over-expression, and were therefore used for subsequent GO and KEGG enrichment analyses. As shown in Figure 5D, among the top 10 enriched GO terms, mitochondrion organization was the most significantly enriched term. The remaining nine terms included cellular macromolecule catabolic process, proteolysis involved in cellular protein catabolic process, macromolecule catabolic process, cellular protein catabolic process, modification-dependent protein catabolic process, organelle organization (biological processes, BP), as well as basal transcription machinery binding, basal RNA polymerase II transcription machinery binding, and protein binding (molecular functions, MF). Collectively, seven of the top 10 GO terms belonged to the BP category, while the other three were assigned to the MF category. In the KEGG pathway enrichment analysis, the top 10 pathways were proteasome, mitophagy-yeast, endocytosis, peroxisome, SNARE interactions in vesicular transport, basal transcription factors, non-homologous end-joining, ABC transporters, other glycan degradation, and longevity regulating pathway-multiple species. Functional classification of these KEGG pathways showed that four were involved in genetic information processing and three in cellular processes, with the remaining pathways distributed across metabolism and environmental information processing (Figure 5E and Supplementary Table S6). The convergent enrichment of mitochondrion organization (top GO term) and proteasome/mitophagy (top KEGG pathways) strongly indicates that ULI1 overexpression triggers a coordinated cellular response centered on organelle remodeling and protein homeostasis in BY4741. These data demonstrate that ULI1 functions as a regulator of a pleiotropic network encompassing mitochondrial structure, protein degradation, organelle crosstalk, and transcriptional regulation, which likely underlies its reported effects on metabolic engineering and stress tolerance in yeast.

FIGURE 5.

Panel A shows a PCA scatter plot with four color-coded sample groups. Panel B is a bar graph with upregulated and downregulated DEGs counts for two pairwise comparisons. Panel C is a Venn diagram showing the overlap of DEGs between the two comparisons. Panel D is a horizontal bar chart of GO terms for gene enrichment, with color intensity representing significance. Panel E is a dot plot of pathway enrichment, where dot size indicates gene number and color intensity represents significance.

Open in a new tab

Statistics and enrichment analysis of differentially expressed genes (DEGs) between BY4741/pY26-uli1 and BY4741/pY26 from transcriptome data. (A) Principal component analysis (PCA) derived score plot. (B) Number of DEGs. (C) Venn diagram of DEGs. (D) Top 10 Gene Ontology (GO) terms of shared DEGs with consistent up- or down-regulation in U1.75P vs. B1.75P and U0.6T vs. B0.6T. (E) Top 10 Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment pathways of shared DEGs with consistent up- or down-regulation in U1.75P vs. B1.75P and U0.6T vs. B0.6T. Three biological replicates were set for every sample (n = 3).

3.6. Proteome sequencing and DEPs analysis of BY4741 ULI1 expression

PCA showed that the sample replicates were closely clustered and ULI1-expressing samples (U1.75P and U0.6T) were clearly separated from their respective controls (Figure 6A), suggesting that the proteomic data were highly reproducible and ULI1 expression significantly altered the cellular proteomic profile. A total of 4,472 proteins were identified, of which 2,258 and 2,345 were DEPs in the comparisons of U1.75P vs. B1.75P and U0.6T vs. B0.6T, respectively (Supplementary Table S7). In both comparisons, upregulated DEPs (1,353 and 1,313) outnumbered downregulated ones (905 and 1,032) significantly (Figures 6B,C). Given that ULI1 co-regulates the biosynthesis of both 2-PE and TOL (Figures 2, 3), we hypothesized that its expression would induce common proteomic changes. Therefore, 1,644 shared DEPs in the U1.75P vs. B1.75P and U0.6T vs. B0.6T comparisons were identified, of which 1,620 (984 up- and 636 down-regulated) exhibited a consistent expression trend across both comparisons (Figure 6D and Supplementary Table S7). These 1,620 DEPs were enriched into 522 GO terms (Supplementary Table S7). The top-enriched terms included biological process (e.g., peptide metabolic process, amide biosynthetic process), cellular component (e.g., mitochondrion, mitochondrial protein-containing complex), and molecular function (e.g., structural constituent of ribosome, oxidoreductase activity) (Figure 6E). The up- and down-regulated shared DEPs were subjected to KEGG enrichment analysis (Supplementary Table S7). As shown in Figures 6F,G, the up-regulated DEPs were significantly enriched in catabolic pathways including carbon metabolism, tryptophan metabolism, and citrate cycle; in contrast, the down-regulated DEPs were associated with anabolic pathways such as amino acid biosynthesis, other types of O-glycan biosynthesis, and monobactam biosynthesis. These findings suggest that ULI1 expression enhances the utilization of exogenous nutrients while reducing non-essential biosynthesis to conserve energy. Since phenylalanine and tryptophan are key nutrients in TM-1.75P and TM-0.6T media, respectively, their extensive consumption will inevitably facilitate the production of metabolites like 2-PE and TOL.

FIGURE 6.

Panel A displays a principal component analysis scatterplot comparing four sample groups; panels B and C show volcano plots of differential gene expression; panel D is a Venn diagram indicating gene overlap between two comparisons; panel E is a bar chart of enriched gene ontology terms; panels F and G are dot plots representing KEGG pathway enrichment analyses with pathway names, rich factor, p-adjust values, and gene counts.

Open in a new tab

Statistics and enrichment analysis of differentially expressed proteins (DEPs) between BY4741/pY26-uli1 and BY4741/pY26 from proteome data. (A) Principal component analysis (PCA) derived score plot. (B) Volcano plot of DEPs from the comparison of U1.75P vs. B1.75P. (C) Volcano plot of DEPs from the comparison of U0.6T vs. B0.6T. (D) Venn diagram of DEPs from both comparisons. (E) Top 20 Gene Ontology (GO) terms of shared DEPs with consistent up- or down-regulation in both comparisons. (F) Top 20 Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment pathways of up-regulated shared DEPs in both comparisons. (G) Top 20 KEGG enrichment pathways of down-regulated shared DEPs in both comparisons. Three biological replicates were set for every sample (n = 3).

3.7. Heatmap cluster analysis

The genes and proteins associated with 2-PE and TOL de novo biosynthesis, regulation of Ehrlich pathway, and the unfolded protein response were subjected to heatmap cluster analysis. The transcriptomes of KMLY1-6 ULI1 disruption and BY4741 ULI1 overexpression exhibited no significant differential pattern between control samples (K0.6T, K1.75P, B0.6T, B1.75P) and experimental samples (M0.6T, M1.75P, U0.6T, U1.75P), respectively (Figures 7a,b). In contrast, the proteome of BY4741 ULI1 overexpression showed an evident differential pattern between the corresponding control samples (B0.6T, B1.75P) and experimental samples (U0.6T, U1.75P) (Figure 7C). The results indicate that ULI1 mainly functions at the post-transcriptional translation stage or the post-translational protein folding stage, which supports that ULI1 was involved in the UPR (Metzger and Michaelis, 2009). Benisch et al. (2023) found that UPR regulation could optimize protein production. In our study, compared to the control samples B0.6T and B1.75P, many aromatic alcohols de novo synthesis enzymes (ARO2, ARO7, ARO9, ARO10, PDC5, ADH2, ADH4, and SFA1), plenty of thiamine regulon proteins (THI3, THI4, THI6, THI13, THI20, THI21, THI72, THI73, and THI80), ARO80 (the key positive regulation factor of ARO9 and ARO10), as well as LPD1 were up-regulated in the experimental samples U0.6T and U1.75P (Figure 7C). In experimental samples, numerous proteins with large changes compared to the control samples, such as ARO9 (8.32- to 16.84-fold), ARO10 (18.29- to 40.32-fold), PDC5 (21.71- to 41.90-fold), ADH2 (5.84- to 7.65-fold), ADH4 (69.44- to 196.82-fold), THI4 (31.83- to 77.91-fold), THI6 (8.80- to 14.12-fold), THI13 (47.54- to 118.40-fold), THI20 (14.10- to 18.89-fold), THI21 (14.19- to 18.43-fold), may contribute substantially to the improved biosynthesis of 2-PE and TOL. Transamination serves as the rate-limiting reaction of the Ehrlich pathway (Iraqui et al., 1999; Dai et al., 2021). The upregulation of ARO9 in the ULI1 overexpression groups U0.6T and U1.75P may enhance the Ehrlich pathway for aromatic alcohol biosynthesis. Meanwhile, the increased abundances of decarboxylase ARO10 and PDC5, as well as alcohol dehydrogenase ADH2, ADH4 and SFA1, further facilitate the progression of the Ehrlich pathway. Decarboxylation reaction in this pathway is strictly TPP-dependent and rate-limiting for 2-PE and TOL biosynthesis, and TPP acts as an essential coenzyme to stabilize catalytic intermediates and sustain decarboxylase activity (Romagnoli et al., 2012; Dai et al., 2021; Gong et al., 2022). Accordingly, the upregulation of thiamine regulatory proteins in experimental groups also indirectly promotes the synthesis of aromatic alcohols. The transcription factor ARO80 can activate the transcription of ARO9 and ARO10 during aromatic alcohol biosynthesis (Iraqui et al., 1999). Therefore, the upregulated expression of ARO80 in experimental samples may promote the transcription of ARO9 and ARO10, thereby indirectly enhancing the Ehrlich pathway. The LPD1, which is involved in NADH generation, was also elevated in ULI1-overexpressing strains. Given that the reduction reaction of the Ehrlich pathway is NADH-dependent (Dickinson et al., 2003), the upregulation of LPD1 could further promote this metabolic pathway. In conclusion, ULI1 promotes the biosynthesis of aromatic alcohols (2-PE and TOL) by reinforcing the Ehrlich pathway.

FIGURE 7.

Three clustered heatmaps labeled A, B, and C display gene expression data for multiple genes and sample types, with color gradients from blue to red indicating relative expression levels from low to high. Each gene name is listed along the y-axis, and sample names are listed on the x-axis. Dendrograms show gene clustering patterns for each heatmap.

Open in a new tab

Heatmap profiling of genes involved in de novo biosynthesis of 2-phenylethanol (2-PE) and tryptophol (TOL), Ehrlich pathway regulation, and the unfolded protein response in uli1-disrupted KMLY1-6 (A) and ULI1-overexpressing BY4741 (B). Heatmap profiling of proteins associated with the aforementioned biosynthetic and regulatory pathways in the ULI1-overexpressing BY4741 (C). Three biological replicates were set for every sample (n = 3).

3.8. Omics insight into the regulatory role of ULI1

Transcriptomic analysis of the ULI1 frameshift mutant showed that DEGs were highly enriched in thiamine metabolism, suggesting that ULI1 modulates 2-PE and TOL biosynthesis by regulating thiamine-related pathways. In ULI1-overexpressing cells, transcriptomic profiling revealed prominent enrichment in mitochondrial organization, proteasome function, mitophagy, and vesicular transport, indicating that ULI1 may maintain organelle structure and protein homeostasis. Proteomic analysis further showed that ULI1 overexpression tended to elevate the abundance of proteins associated with carbon metabolism, tryptophan metabolism, and the TCA cycle, whereas an overall decline was observed for proteins participating in anabolism including amino acid biosynthesis. This metabolic rearrangement may promote efficient utilization of exogenous nutrients, reduces energy consumption on unnecessary biosynthesis, and redirects metabolic flux toward aromatic alcohol production. Under nutritional starvation and ER stress, we propose that the ER-localized ULI1 orchestrates a global regulatory network by modulating thiamine metabolism, preserving mitochondrial architecture and cellular protein homeostasis, promoting catabolic pathways while restraining anabolic processes, and coordinating stress responses with metabolic flux redistribution, thereby enhancing the biosynthesis of 2-PE and TOL. This study identified the regulatory role of ULI1 in aromatic alcohol biosynthesis, while its molecular interactions remain unclear. ULI1 may act as a regulatory mediator to modulate yeast metabolism. Future studies will be carried out to identify its direct targets, further elucidating its regulatory mechanism for targeted metabolic engineering.

4. Conclusion

In this study, we characterized the bioinformatic features and regulatory function of ULI1, a genus-specific ER membrane protein in Saccharomyces. Gene knockout and overexpression experiments demonstrated that ULI1 positively regulates the biosynthesis of 2-PE and TOL in a dosage-dependent manner, whereas ALD3 negatively affects aromatic alcohol production by competing for key aldehyde precursors. Multi-omics analyses revealed that ULI1 does not directly catalyze aromatic alcohol synthesis but may function as a pleiotropic regulator connecting ER stress responses, organelle homeostasis, and central metabolism. ULI1 may coordinate thiamine and mitochondrial metabolism, enhance catabolic flux, and repress non-essential anabolic processes, thereby optimizing energy utilization and redirecting metabolic precursors toward 2-PE and TOL production. Our results suggest that ULI1 may serve as a novel regulatory hub linking stress tolerance, cellular homeostasis, and aromatic alcohol biosynthesis in yeast. The proposed regulatory role of ULI1 in 2-PE and TOL biosynthesis remains to be fully verified. This work provides mechanistic insights and a promising target for future metabolic engineering strategies aimed at improving the industrial production of high-value aromatic alcohols and other valuable alcohols in S. cerevisiae.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the Key Scientific Research Project of China Tobacco Yunnan Industrial Co., Ltd. (2025XY01).

Edited by: Li Wen, Changsha University of Science and Technology, China

Reviewed by: Guihu Zhang, Beijing Technology and Business University, China

Simiat Olanike Jimoh, Federal University of Health Sciences, Ila-Orangun, Nigeria

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material. Sequence data of this study have been deposited in the Short Read Archive under accession numbers SRR38241256–SRR38241279.

Author contributions

XG: Methodology, Visualization, Writing – original draft. YS: Investigation, Visualization, Writing – original draft. SS: Formal Analysis, Writing – review & editing. JZ: Funding acquisition, Methodology, Writing – review & editing. GM: Formal Analysis, Visualization, Writing – original draft. HL: Data curation, Methodology, Resources, Writing – original draft. SL: Data curation, Writing – review & editing. CS: Data curation, Investigation, Validation, Writing – original draft. DZ: Conceptualization, Supervision, Writing – review & editing. YL: Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing.

Conflict of interest

XG, SS, JZ, SL, and DZ were employed by the R&D Center of China Tobacco Yunnan Industrial Co., Ltd.

The remaining author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb.2026.1934502/full#supplementary-material

Data_Sheet_1.zip (4.2MB, zip)
Supplementary Figure S1

Schematic diagram of gene-editing plasmids assembly.

Data_Sheet_1.zip (4.2MB, zip)
Supplementary Figure S2

Verification of gene engineering strains. (A) Sequence alignment of the ULI1 gene in KMLY1-6 and Kmuli1. (B) Schematic diagram of ULI1 gene knockout and primers used for PCR verification. (C) Agarose gel electrophoresis of PCR products for verifying gene deletion events. M: DNA marker; Lanes 1–4: KΔuli1 strains; Lane 5: KMLY1-6. The PCR products from KΔuli1s and KMLY1-6 were 1,436 and 1,946 bp in length, respectively. (D) Schematic diagram of ALD3 gene knockout, ULI1 expression cassette integration, and primers used for PCR verification. (E) Schematic diagram of ULI1 expression cassette integration at the X-3 site and primers used for PCR verification. (F) Schematic diagram of ULI1 expression cassette integration at the XII-3 site and primers used for PCR verification. (G) Agarose gel electrophoresis of PCR products for verifying genomic integration events, amplified using the primers shown in (D-F). M: DNA marker; Lanes 1: KΔald3 (1,301 bp); Lane 2: KΔald3uli1 (2,750 bp); Lane 4: KΔuli1 (3,262 bp); Lane 3, 5, and 6: KMLY1-6 (2,822 bp, 2,782 bp, 2,383 bp); Lane 7: KXIIuli1 (2,644 bp).

Data_Sheet_1.zip (4.2MB, zip)

References

  1. Adams C. J., Kopp M. C., Larburu N., Nowak P. R., Ali M. M. U. (2019). Structure and molecular mechanism of ER stress signaling by the unfolded protein response signal activator IRE1. Front. Mol. Biosci. 6:11. 10.3389/fmolb.2019.00011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Benisch M., Benzinger D., Kumar S., Hu H., Khammash M. (2023). Optogenetic closed-loop feedback control of the unfolded protein response optimizes protein production. Metab. Eng. 77, 32–40. 10.1016/j.ymben.2023.03.001 [DOI] [PubMed] [Google Scholar]
  3. Bernardino A. R. S., Torres C. A. V., Crespo J. G., Reis M. A. M. (2024). Biotechnological 2-phenylethanol production: Recent developments. Molecules 29:5761. 10.3390/molecules29235761 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen H., Fink G. R. (2006). Feedback control of morphogenesis in fungi by aromatic alcohols. Genes Dev. 20, 1150–1161. 10.1101/gad.1411806 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen W., McKenna A., Schreiber J., Haeussler M., Yin Y., Agarwal V., et al. (2019). Massively parallel profiling and predictive modeling of the outcomes of CRISPR/Cas9-mediated double-strand break repair. Nucleic Acids Res. 47, 7989–8003. 10.1093/nar/gkz487 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cooper T. G. (2002). Transmitting the signal of excess nitrogen in Saccharomyces cerevisiae from the Tor proteins to the GATA factors: Connecting the dots. FEMS Microbiol. Rev. 26, 223–238. 10.1111/j.1574-6976.2002.tb00612.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cox D. J., Strudwick N., Ali A. A., Paton A. W., Paton J. C., Schroder M. (2011). Measuring signaling by the unfolded protein response. Met. Enzymol. 491, 261–292. 10.1016/B978-0-12-385928-0.00015-8 [DOI] [PubMed] [Google Scholar]
  8. Dai J., Xia H., Yang C., Chen X. (2021). Sensing, uptake and catabolism of L-phenylalanine during 2-phenylethanol biosynthesis via the Ehrlich pathway in Saccharomyces cerevisiae. Front. Microbiol. 12:601963. 10.3389/fmicb.2021.601963 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Dickinson J. R., Salgado L. E. J., Hewlins M. J. E. (2003). The catabolism of amino acids to long chain and complex alcohols in Saccharomyces cerevisiae. J. Biol. Chem. 278, 8028–8034. 10.1074/jbc.M211914200 [DOI] [PubMed] [Google Scholar]
  10. Dubhashe Y. R., Sawant V. M., Gaikarm V. G. (2018). Process intensification of continuous flow synthesis of tryptophol. Ind. Eng. Chem. Res. 57, 2787–2796. 10.1021/acs.iecr.7b04483 [DOI] [Google Scholar]
  11. Elleuch L., Shaaban M., Smaoui S., Mellouli L., Karray-Rebai I., Fourati-Ben Fguira L., et al. (2010). Bioactive secondary metabolites from a new terrestrial Streptomyces sp. TN262. Appl. Biochem. Biotechnol. 162, 579–593. 10.1007/s12010-009-8808-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fraud S., Rees E. L., Mahenthiralingam E., Russell A. D., Maillard J. (2003). Aromatic alcohols and their effect on gram-negative bacteria, cocci and mycobacteria. J. Antimicrob. Chemoth. 51, 1435–1436. 10.1093/JAC/DKG246 [DOI] [PubMed] [Google Scholar]
  13. Gong X., Luo H., Hong L., Wu J., Wu H., Song C., et al. (2022). Effects of tryptophan and phenylalanine on tryptophol production in Saccharomyces cerevisiae revealed by transcriptomic and metabolomics analyses. J. Microbiol. 60, 832–842. 10.1007/s12275-022-2059-2 [DOI] [PubMed] [Google Scholar]
  14. Gu Y., Ma J., Zhu Y., Xu P. (2020). Refactoring Ehrlich pathway for high-yield 2-phenylethanol production in Yarrowia lipolytica. ACS synth. Biol. 9, 623–633. 10.1021/acssynbio.9b00468 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Guo W., Huang Q., Feng Y., Tan T., Niu S., Hou S., et al. (2020). Rewiring central carbon metabolism for tyrosol and salidroside production in Saccharomyces cerevisiae. Biotechnol. Bioeng. 117, 2410–2419. 10.1002/bit.27370 [DOI] [PubMed] [Google Scholar]
  16. Halbleib K., Pesek K., Covino R., Hofbauer H. F., Wunnicke D., Hänelt I., et al. (2017). Activation of the unfolded protein response by lipid bilayer stress. Mol. Cell 67, 673–684.e8. 10.1016/j.molcel.2017.06.012 [DOI] [PubMed] [Google Scholar]
  17. Hassing E., de Groot P. A., Marquenie V. R., Pronk J. T., Daran J. G. (2019). Connecting central carbon and aromatic amino acid metabolisms to improve de novo 2-phenylethanol production in Saccharomyces cerevisiae. Metab. Eng. 56, 165–180. 10.1016/j.ymben.2019.09.011 [DOI] [PubMed] [Google Scholar]
  18. He Y., Yu S., Liu S., Tian H., Yu C., Tan W., et al. (2020). Data-independent acquisition proteomics unravels the effects of iron ions on coronatine synthesis in Pseudomonas syringae pv. tomato DC3000. Front. Microbiol. 11:1362. 10.3389/fmicb.2020.01362 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Ho N., Yap W. S., Xu J., Wu H., Koh J. H., Goh W. W. B., et al. (2020). Stress sensor Ire1 deploys a divergent transcriptional program in response to lipid bilayer stress. J. Cell Biol. 219:e201909165. 10.1083/jcb.201909165 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Hong S., Lee H. G., Huh W. K. (2024). ARV1 deficiency induces lipid bilayer stress and enhances rDNA stability by activating the unfolded protein response in Saccharomyces cerevisiae. J. Biol. Chem. 300:107273. 10.1016/j.jbc.2024.107273 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hrach V. L., King W. R., Nelson L. D., Conklinm S., Pollockm J. A., Patton-Vogt J. (2023). The acyltransferase Gpc1 is both a target and an effector of the unfolded protein response in Saccharomyces cerevisiae. J. Biol. Chem. 299:104884. 10.1016/j.jbc.2023.104884 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hua D., Xu P. (2011). Recent advances in biotechnological production of 2-phenylethanol. Biotechnol. Adv. 29, 654–660. 10.1016/j.biotechadv.2011.05.001 [DOI] [PubMed] [Google Scholar]
  23. Inagaki S., Morimura S., Gondo K., Tang Y., Akutagawa H., Kida K. (2007). Isolation of tryptophol as an apoptosis-inducing component of vinegar produced from boiled extract of black soybean in human monoblastic leukemia U937 cells. Biosci. Biotechnol. Biochem. 71, 371–379. 10.1271/bbb.60336 [DOI] [PubMed] [Google Scholar]
  24. Iraqui I., Vissers S., André B., Urrestarazu A. (1999). Transcriptional induction by aromatic amino acids in Saccharomyces cerevisiae. Mol. Cell Biol. 19, 3360–3371. 10.1128/MCB.19.5.3360 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jessop-Fabre M. M., Jakočiunas T., Stovicek V., Dai Z., Jensen M. K., Keasling J. D., et al. (2016). EasyClone-MarkerFree: A vector toolkit for marker-less integration of genes into Saccharomyces cerevisiae via CRISPR-Cas9. Biotechnol. J. 11, 1110–1117. 10.1002/biot.201600147 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kim D., Langmead B., Salzberg S. L. (2015). HISAT: A fast spliced aligner with low memory requirements. Nat. Methods 12, 357–360. 10.1038/nmeth.3317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Liu Y., Fu C., Zhang X., Gu B., Hu H., Yang R., et al. (2024). Recent advances in metabolic engineering of microorganisms for production of tyrosol and its derivatives. Chin. J. Biotechnol. 40, 2604–2625. 10.13345/j.cjb.240174 [DOI] [PubMed] [Google Scholar]
  28. Love M. I., Huber W., Anders S. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15:550. 10.1186/s13059-014-0550-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Metzger M. B., Michaelis S. (2009). Analysis of quality control substrates in distinct cellular compartments reveals a unique role for Rpn4p in tolerating misfolded membrane proteins. Mol. Biol. Cell. 20, 1006–1019. 10.1091/mbc.e08-02-0140 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mo E. K., Sung C. K. (2007). Phenylethyl alcohol (PEA) application slows fungal growth and maintains aroma in strawberry. Postharvest Biol. Tec. 45, 234–239. 10.1016/J.POSTHARVBIO.2007.02.005 [DOI] [Google Scholar]
  31. Pertea M., Kim D., Pertea G. M., Leek J. T., Salzberg S. L. (2016). Transcript-level expression analysis of RNA-seq experiments with HISAT, stringtie and ballgown. Nat. Protoc. 11, 1650–1667. 10.1038/nprot.2016.095 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Read A., Schröder M. (2021). The unfolded protein response: An overview. Biology 10:384. 10.3390/biology10050384 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Romagnoli G., Luttik M. A., Kötter P., Pronk J. T., Daran J. M. (2012). Substrate specificity of thiamine pyrophosphate-dependent 2-oxo-acid decarboxylases in Saccharomyces cerevisiae. Appl. Environ. Microbiol. 78, 7538–7548. 10.1128/AEM.01675-12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Schirmer M., Smeekens S. P., Vlamakis H., Jaeger M., Oosting M., Franzosa E. A., et al. (2016). Linking the human gut microbiome to inflammatory cytokine production capacity. Cell 167, 1125–1136. 10.1016/j.cell.2016.10.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Wang W., Du L., Wei Q., Lu M., Xu D., Li Y. (2025). Synthesis and health effects of phenolic compounds: A focus on tyrosol, hydroxytyrosol, and 3,4-dihydroxyacetophenone. Antioxidants 14:476. 10.3390/antiox14040476 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wang Y. Q., Zhang Z. Y., Lu X. Y., Zong H., Zhuge B. (2020). Transcription factor Hap5 induces gsh2 expression to enhance 2-phenylethanol tolerance and production in an industrial yeast Candida glycerinogenes. Appl. Microbiol. Biot. 104, 4093–4107. 10.1007/s00253-020-10509-y [DOI] [PubMed] [Google Scholar]
  37. Wu H., Zhao W., Zhao Y., Xiao Y., Yang L., Qin Y., et al. (2024). Research progress on the biosynthesis pathways and molecular regulatory mechanism of tryptophol in Saccharomyces cerevisiae. J. Kunming Univ. Sci. Technol. 49, 137–146. 10.16112/j.cnki.53-1223/n.2024.01.263 [DOI] [Google Scholar]
  38. Wu W., Deng G., Liu C., Gong X., Ma G., Yuan Q., et al. (2020). Optimization and multiomic basis of phenyllactic acid overproduction by Lactobacillus plantarum. J. Agric. Food Chem. 68, 1741–1749. 10.1021/acs.jafc.9b07136 [DOI] [PubMed] [Google Scholar]
  39. Xu J., Yu J., Jin Y., Li J., Yu Z., Lv Y. (2017). A continuous flow microwave-assisted fischer indole synthesis of 7-ethyltryptophol. Chem. Eng. Process. 121, 144–148. 10.1016/J.CEP.2017.09.001 [DOI] [Google Scholar]
  40. Yan W., Gao H., Jiang Y., Qian X., Zhou J., Dong W., et al. (2021). Research progress in 2-phenylethanol production through biological processes. Synth. Biol. J. 2, 1030–1045. [Google Scholar]
  41. Yofe I., Weill U., Meurer M., Chuartzman S., Zalckvar E., Goldman O., et al. (2016). One library to make them all: Streamlining yeast library creation by a SWAp-Tag (SWAT) strategy. Nat. Methods 13, 371–378. 10.1038/nmeth.3795 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zhu F., Jin M. (2015). The effects of a thermophile metabolite, tryptophol, upon protecting shrimp against white spot syndrome virus. Fish Shellfish Immun. 47, 777–781. 10.1016/j.fsi.2015.10.015 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data_Sheet_1.zip (4.2MB, zip)
Supplementary Figure S1

Schematic diagram of gene-editing plasmids assembly.

Data_Sheet_1.zip (4.2MB, zip)
Supplementary Figure S2

Verification of gene engineering strains. (A) Sequence alignment of the ULI1 gene in KMLY1-6 and Kmuli1. (B) Schematic diagram of ULI1 gene knockout and primers used for PCR verification. (C) Agarose gel electrophoresis of PCR products for verifying gene deletion events. M: DNA marker; Lanes 1–4: KΔuli1 strains; Lane 5: KMLY1-6. The PCR products from KΔuli1s and KMLY1-6 were 1,436 and 1,946 bp in length, respectively. (D) Schematic diagram of ALD3 gene knockout, ULI1 expression cassette integration, and primers used for PCR verification. (E) Schematic diagram of ULI1 expression cassette integration at the X-3 site and primers used for PCR verification. (F) Schematic diagram of ULI1 expression cassette integration at the XII-3 site and primers used for PCR verification. (G) Agarose gel electrophoresis of PCR products for verifying genomic integration events, amplified using the primers shown in (D-F). M: DNA marker; Lanes 1: KΔald3 (1,301 bp); Lane 2: KΔald3uli1 (2,750 bp); Lane 4: KΔuli1 (3,262 bp); Lane 3, 5, and 6: KMLY1-6 (2,822 bp, 2,782 bp, 2,383 bp); Lane 7: KXIIuli1 (2,644 bp).

Data_Sheet_1.zip (4.2MB, zip)

Data Availability Statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material. Sequence data of this study have been deposited in the Short Read Archive under accession numbers SRR38241256–SRR38241279.

Articles from Frontiers in Microbiology are provided here courtesy of Frontiers Media SA

RESOURCES