Proteomic analysis revealed the roles of YRR1 deletion in enhancing the vanillin resistance of Saccharomyces cerevisiae

Background Vanillin is one of the important phenolic inhibitors in Saccharomyces cerevisiae for bioconversion of lignocellulosic materials and has been reported to inhibit the translation process in cells. In our previous studies, it was confirmed that the deletion of the transcription factor gene YRR1 enhanced vanillin resistance by promoting some translation-related processes at the transcription level. In this work, we investigated the effects of proteomic changes upon induction of vanillin stress and deletion of YRR1 to provide unique perspectives from a transcriptome analysis for comprehending the mechanisms of YRR1 deletion in the protective response of yeast to vanillin. Results In wild-type cells, vanillin reduced two dozens of ribosomal proteins contents while upregulated proteins involved in glycolysis, oxidative phosphorylation, and the pentose phosphate pathway in cells. The ratios of NADPH/NADP+ and NADH/NAD+ were increased when cells responded to vanillin stress. The differentially expressed proteins perturbed by YRR1 deletion were much more abundant than and showed no overlaps with transcriptome changes, indicating that Yrr1 affects the synthesis of certain proteins. Forty-eight of 112 upregulated proteins were involved in the stress response, translational and transcriptional regulation. YRR1 deletion increased the expression of HAA1-encoding transcriptional activator, TMA17-encoding proteasome assembly chaperone and MBF1-encoding coactivator at the protein level, as confirmed by ELISA. Cultivation data showed that the overexpression of HAA1 and TMA17 enhanced resistance to vanillin in S. cerevisiae. Conclusions Cells conserve energy by decreasing the content of ribosomal proteins, producing more energy and NAD(P)H for survival in response to vanillin stress. Yrr1 improved vanillin resistance by increasing the protein quantities of Haa1, Tma17 and Mbf1. These results showed the response of S. cerevisiae to vanillin and how YRR1 deletion increases vanillin resistance at the protein level. These findings may advance our knowledge of how YRR1 deletion protects yeast from vanillin stress and offer novel targets for genetic engineering of designing inhibitor-resistant ethanologenic yeast strains. Supplementary Information The online version contains supplementary material available at 10.1186/s12934-021-01633-z.

they are available in abundance, are renewable, and do not compete with food resources [3]. However, the pretreatment of lignocellulose is inevitably generates many inhibitors, such as organic acids, furans, and phenolic compounds, which hamper microorganism growth and fermentation [4]. Vanllin, a guaiacyl phenol, has been considered as an important inhibitor of lignocellulosic hydrolysates, as it inhibits the viability of many microorganisms at very low concentration [5]. The concentrations of vanillin ranges from 1 to 26 mM according to the types of biomass materials and the method of pretreatment [6,7]. In addition, as one of the most widely useful flavoring agents in the food and cosmetics industry, vanillin is produced by microorganisms from glucose or ferulic acid, which is other components of lignin [8][9][10][11]. Thus, it is essential to understand the physiological mechanism of how vanillin inhibits S. cerevisiae growth and fermentation during the generation of vanillin-resistant S. cerevisiae cell factories to improve bioethanol and vanillin production.
In our previous work, we reported that the deletion of the multidrug resistance-related transcription factor gene YRR1 remarkably enhanced S. cerevisiae vanillin resistance. The rate of growth and vanillin transformation to vanillyl alcohol in the YRR1-deleted strain under vanillin stress was faster than that in the wild-type yeast. This phenomenon is generally evident in different background strains (BY4741 and CEN.PK2-1C) [12]. Transcriptome analysis showed that the genes related to ribosome biogenesis and rRNA processing are upregulated in the YRR1-deleted strain compared to the wild-type strain under vanillin stress [12]. Among the upregulated genes, the overexpression of DBP2, encoding RNA helicase and participating in ribosome biogenesis, was confirmed to enhance vanillin resistance [12]. Ribosomes and rRNA are vital for translation and protein synthesis [12]. Thus, YRR1 was thought to impact translation and protein synthesis. In addition, Iwaki et al. [13] have revealed that vanillin represses translational initiation by blocking polysome assembly, which led to the accumulation of cytoplasmic processing bodies (P-bodies) and stress granules (SGs). Furfural, 5-hydroxymethyl furfural (HMF), and acidic stress also exhibited the same phenomenon [14,15].
Therefore, global mRNA profiles alone are not sufficient to represent the changes in the biological system because translational and post-translational regulation mechanisms, protein degradation, and other processes have not been taken into account. Transcript levels frequently do mismatch the corresponding protein levels [16,17].
In general, to gain a complete insight into how YRR1 deletion protects yeast from vanillin stress, in this study, we conducted a quantitative proteomic analysis of yeast treated with specific vanillin stress and deletion of YRR1, respectively. The proteomic data were analyzed in combination with previous transcriptome data obtained under the same conditions. We found that YRR1 deletion increased Haa1, Tma17 and Mbf1 expression at the protein level, and overexpression of HAA1 and TMA17 enhanced the resistance of yeast to vanillin. This work enriches the knowledge of the protective response conferred by YRR1 deletion to vanillin stress and offered novel strategies for genetic engineering manipulation to improve microbe resistance.

Quantitative proteomic analysis reveals remarkable alterations in the protein abundance in response to vanillin
To investigate the changes in protein abundance perturbed by vanillin stress, the proteomic differences were examined with the laboratory strain BY4741 cultivated with 5 mM vanillin for 15 h or without vanillin for 6 h when OD 600 reached to 1.6. We used 1.2-, 1.3-, 1.5-and 2-fold thresholds to analyze the differentially expressed proteins (DEPs). The numbers of 1.2-, 1.3-, 1.5-and 2-fold DEPs was 1002, 218, 85 and 19, respectively. In addition, the corresponding numbers of DEPs between strains BY4741 and BY4741(yrr1Δ) without vanillin treatment was 198, 121, 32 and 0. According to the advice offered by Jingjie PTM Biolab, a proteomics company, and many published articles on proteomics [18,19], we chose 1.3 as the threshold. According to our stringent data-mining criteria, 135 proteins were upregulated and 83 proteins were downregulated in response to vanillin (P-value < 0.05). Functional annotation of all the identified proteins was conducted based on the Gene Ontology (GO) database [20] and the Kyoto Encyclopedia of Genes and Genomes (KEGG) database [21]. The KEGG pathway enrichment of DEPs suggests that the upregulated protein genes are mainly involved in the functional categories of oxidative phosphorylation and glycolysis (Fig. 1a). In contrast, downregulated protein genes are largely involved with the function of ribosome (Fig. 1b).

Proteomes exhibited visible differences in the presence of vanillin according to the transcriptome analysis
Considering of the repression of the translation initiated by vanillin, mRNA most likely cannot represent protein expression. Thus, a proteomic dataset was compared to data obtained from a previous transcriptomic outcome (GEO accession number: GSE89854) to investigate the correlation between mRNA and protein levels [12]. Overlaps (i.e., homodirectional) between DEPs and differentially expressed genes (DEGs) are shown in the Venn diagrams (Fig. 2a). Only 23.8% of these DEPs exhibited a homodirectional change at the level of their transcript counterparts. A scatter plot was used to compare high-quality protein with the transcript expression ratios for all 218 DEPs (Fig. 2b). The positive Spearman rank correlation coefficient (Sr) of 0.47 that was calculated for these 218 DEPs revealed a weak correlation between the transcriptome and proteome. This poor correlation between changes in protein and the corresponding transcripts indicated that the changes in protein expression in response to vanillin stress are mainly regulated at the posttranscriptional level.
The DEPs and DEGs enriched (P < 0.05) in KEGG pathways are listed in Fig. 2c. There were total nineteen glycolysis-related genes and ten ABC transporters in S288C S. cerevisiae according to Saccharomyces Genome Database (SGD). At both the mRNA and protein levels, four proteins (Tdh1, Tdh2, Hxk2 and Eno1) in glycolysis-related pathways and three ABC transporters (Pdr5, Snq2 and Pdr12) were upregulated, which indicates that vanillin regulates the expression of these genes at the transcriptional level. On the other hand, at the protein level, the ribosomal proteins and RNA polymerases were downregulated significantly, which indicated that vanillin inhibited the translation of proteins in these two pathways. However, this inhibition was gene-specific. Nguyen et al. [22] found that both of ADH6 and ADH7, which encode mediumchain alcohol dehydrogenases, were transcriptionally upregulated by vanillin stress. However, only the protein of ADH7 was increased, while the protein of ADH6 decreased under vanillin stress. The mechanism behind this finding is not clear, but it is probably related to the transcriptome and proteome inconsistency.

Ribosomal proteins and rRNA processing-related proteins were significantly decreased in response to vanillin stress
Although the transcriptome showed that 13 genes related to ribosome biogenesis were notably downregulated (Table 1), no changes were exhibited at the corresponding protein level in response to vanillin stress [12]. However, twenty ribosomal proteins showed decreased expression under vanillin stress, including 5 small ribosomal subunits (40S) and 15 large ribosomal subunits (60S) ( Table 1). Moreover, one ribosome biogenesisrelated protein (Rlp24) [23], three proteins involved in rRNA processing (Esf1, Nsa2, Pop8), two RNA polymerase I subunits (Rpa12 and Rpa190), one RNA polymerase III subunit (Rpc11), tRNA 4-demethylwyosine synthase Tyw1, and an ATPase for tRNA processing of Nbp35 also displayed a visible decrease in protein abundance. These downregulated proteins are constituents of the cellular translational machine, which indicates that vanillin blocks translation efficiency and protein synthesis by reducing the quantity of ribosomal proteins available. This finding coincides with the fact that vanillin inhibits translation initiation, resulting in cells not needing many more ribosomes.
Under same conditions, a proteomic analysis were performed with another haploid laboratory strain CEN. PK2-1C, which has a different genetic background than BY4741. After treatment of CEN.PK2-1C with or without vanillin, the proteome data showed that vanillin treatment decreased the content abundance of certain proteins related to ribosomes, as indicated by a pathway enrichment analyses (Additional file 1: Fig. S1). Therefore, these results demonstrated that the reduction in ribosome biogenesis is likely representative of the cellular Fig. 1 Statistics of KEGG pathway enrichment of DEPs of BY4741 response to 5 mM vanillin. BY4741 cultivated with 5 mM vanillin for 15 h or without vanillin for 6 h when OD 600 reached to 1.6. a Upregulated proteins; b down-regulated proteins. The log 2 fold enrichment is the log 2 converted value of the ratio of DEP number annotated in this pathway term to all protein numbers annotated in this pathway term. Greater log 2 fold enrichment indicates a greater effect of vanillin on the analyzed pathway conservation of energy needed for survival under vanillin stress, since ribosome biogenesis demands a great supply of energy [24,25].

Vanillin stress increased the quantity of proteins involved in energy generation
As shown in Table 2  energy generation pathways, such as glycolysis, oxidative phosphorylation, and the pentose phosphate pathway (PPP), were dramatically upregulated. Six glycolytic enzymes (Pyc1, Tdh1, Tdh2, Hxk2, Eno1, and Gpm2) associated with glycolysis were remarkably increased. In particular, Hxk2, the predominant enzyme that catalyzes the first irreversible step of glycolysis [26], was upregulated more than 1.5-fold. Hxk2 is one of two rate-limiting enzymes in glycolysis [27]. Moreover, the proteins related to adenosine triphosphatase (ATP) synthesis, including the five ATP synthase subunits (Atp20, Atp7, Atp15, Atp14, and Tim11), three cytochrome c oxidase subunits (Cox4, Cox12, and Cox8), four ubiquinol cytochrome c reductase subunits (Qcr9, Qcr2, Cor1, and Qcr6), and one ATPase (Pma2), were also increased significantly. Upregulated glycolysis and oxidative phosphorylation pathways tend to supply more ATP for cellular survival as the detoxification of vanillin consumes energy. For example, when facing vanillin stress, some ATP binding cassette (ABC) transporters are upregulated, such as Pdr5 and Snq2 (upregulated 1.68 and 3.78 times, respectively, at the protein level), to extrude vanillin from the cells, which requires a large amount of energy. These two transporters were upregulated at the mRNA level, and their overexpression was confirmed to shorten the lag phase under vanillin stress [12]. A similar result was also found in the proteome of CEN.PK2-1C treated with vanillin, in which a visible amount of upregulated protein was related to glycolysis (Additional file 1: Fig. S2). Two enzymes (Gnd1 and Sol3) of the PPP increased more than 1.5-fold, particularly the enzyme Gnd1, which generates nicotinamide adenine dinucleotide phosphate (NADPH) during its catalysis [28]. PPP is the major source of NADPH, a vital cofactor for vanillin conversion b log2 ratio (BY4741with/without 5 mM vanillin stress) which is less than 0 means down-regulation at mRNA level. False discovery rate (FDR) ≤ 0.001  to vanillyl alcohol [29][30][31]. This finding agrees with the finding confirmed in our previous work showing a higher NADPH/NADP + cell ratio can enhance vanillin resistance [32].
Our findings related to the involvement of the energy metabolism in the cellular response to vanillin. The levels of NAD(P) + , NAD(P)H and ATP/ADP were measured in this study (Fig. 3). The intracellular content of NADH and NADPH increased by at least 1.5-and 1.8-fold, respectively, while NAD + was not obviously changed, and NADP + decreased by 1.7-fold, leading to a moderately increased NADH/NAD + ratio and a significantly increased NADPH/NADP + ratio in vanillin-treated BY4741 cells compared with the control. These results comport with the upregulation of certain enzymes in carbon metabolism that produce more NADH and NADPH in response to vanillin stress. However, the level of ATP/ ADP ratio showed a significant decrease in the presence of vanillin. This outcome was probably because the urgent requirement of vanillin detoxication causes a severe shortage of ATP, leading to the upregulation of certain enzymes involved in the glycolysis to produce as much ATP as possible to defend yeast cells from the deleterious effects of vanillin. Thus, the lower number of ribosomal proteins and ribosome biogenesis proteins may reflect the need to conserve energy, and the high activation of the glycolysis and oxidative phosphorylation pathways to produce more energy is thought to be a defense mechanism of yeast for survival in response to vanillin stress.

Comparative proteomics analysis after the deletion of YRR1
Our previous study found that vanillin repressed the ribosome biogenesis pathway. In addition, the deletion of YRR1 can attenuate this repression and lead to enhanced resistance of S. cerevisiae to vanillin [12]. Thus, YRR1 is thought to be related to ribosome biogenesis or translation pathways. However, the known YRR1 targets are mostly related to ABC transporters and permeases, not the two aforementioned pathways. Further, the deletion of YRR1 triggered the differential expression of only eight genes (YBR230W-A, CAR2, FMP45, YCL048W-A, SCS3, UTH1, PMP3, and YIL002W-A) at the mRNA level, and none has an effect on vanillin resistance, as shown in our previous work [12]. To investigate the protein quantity changes caused by deleting YRR1, a proteomic analysis was conducted with BY4741(yrr1Δ) and BY4741. As mentioned before, we also used 1.3-fold change as the threshold for analysis. The outcome revealed that there The effect of vanillin on intracellular coenzymes and ATP. NAD + , NADH, NADP + and NADPH content. Cells were cultivated at 30 °C, 200 rpm in SC-URA liquid medium with 5 mM vanillin for 15 h or without vanillin for 6 h when OD 600 reached to 1.6 (a), the ratio of NADH to NAD + , NADPH to NADP + (b), the ratio of ATP to ADP (c). CDW, cell dry weight. V-BY4741 represents BY4741 under vanillin stress. Three independent replicates were performed for each sample. Data are the mean ± standard deviation of independent triplicate experiments. *P value < 0.05 and **P value < 0.01 in significance analysis using two-tailed Student t test were 121 DEPs, including 112 upregulated and 9 downregulated proteins that showed no overlap with the levels of their corresponding mRNA. The upregulated DEPs were classified into functional categories according to the MIPS Functional Catalog and mainly fell into transcription, translation, and energy generation categories ( Table 3). The functions of these differentially expressed proteins are listed in Table 4.

Deletion of YRR1 stimulated changes in the quantity changes of translational proteins
According to the proteomic data, a series of rRNA processing proteins were upregulated after deleting YRR1. Among these proteins, Nop6, Nhp2, Naf1, Lsm5, Nsr1, and Fyv7 participate in the synthesis of the 40S (small) ribosomal subunit. Cgr1 is involved in processing rRNA for the 60S ribosomal subunit [33]. The factors involved in rRNA synthesis also are increased, including the subunit of RNA polymerase I Rpa34, the RNA polymerase III assembly protein Rbs1, and two transcription factors (Hmo1 and Abf1) involved in activating rRNA and ribosomal protein transcription under the control of the target of rapamycin (TOR) pathway [34][35][36]. Moreover, several regulators of translation, including three translation initiation factors (Tif11, Hcr1, and Gis2), were also upregulated. Considering its ability to repress ribosome genesis and translation by vanillin stress, the deletion of YRR1 might strengthen translational elements to promote vanillin resistance in yeast. Interestingly, the expression of three translational suppressors also increased, including that of two 4E-binding proteins, Caf20 and Eap1, which inhibit translational initiation by competing with translation initiation factor eIF4G to bind with eIF4E [37]. The other translational repressor with upregulated expression was Stm1, which was previously demonstrated to repress translational elongation by limiting the interaction of elongation factor eEF3 with ribosomes and stalling ribosomes [38]. Further, one recent study indicated that Stm1 also functions as a ribosome preservation factor in response to nutrient stress and facilitates ribosomal protein synthesis rates when nutrients levels are restored [39]. These three proteins facilitate the formation of P-bodies [40,41]. P-bodies are viewed as reservoirs of mRNA after translation inhibition, and they are later released to reenter translation [42,43]. It is likely that the absence of YRR1 protects yeast cells against vanillin stress by promoting the formation of P-bodies to store translation elements and saving energy for the de novo synthesis of translation machinery.

The deletion of YRR1 promotes RNA polymerase II-directed transcription
Arbitrary extracellular and intracellular stimulation can trigger a global adaptive transcriptional regulatory program in yeast cells. The deletion of YRR1 upregulated the expression of several basal transcription factors (Taf10, Mbf1, Nhp6b, and Pgd1). These factors facilitate recognition of promoters that contain a TATA box and assembly of the transcription preinitiation complex, thereby promoting transcription of TATA-containing genes [44][45][46][47]. TATA-containing genes tend to be responsive to stress [48,49]. Moreover, the levels of four transcription elongation factors (Gim5, Yke2, Spn1, and Npl3) were also increased under conditions of protein abundance. Npl3, a mRNA-binding protein, is a positive transcription elongation factor shown to interact with the carboxy terminal domain (CTD) of RNA polymerase, exerting a direct stimulatory effect on the elongation activity of the polymerase [50]. These results indicate that the deletion of YRR1 promotes TATA-containing gene transcription through the promotion of basal transcription factors to improve survival under vanillin stress. YRR1 deletion increased the protein content of Tma17, Haa1 and Mbf1, whose overexpression enhanced S. cerevisiae resistance to vanillin To search for possible novel genes/proteins for enhancing vanillin resistance, the DEPs were investigated for their functions in vanillin resistance. First, two proteins attracted our attention since they were both downregulated by vanillin stress and were upregulated upon YRR1 deletion: Rtc3, which is involved in RNA metabolism, and plasma membrane protein Hsp12, which plays a role in maintaining membrane organization. These two proteins may be the critical for promoting the resistance of  yeast to vanillin stress upon YRR1 deletion. However, the overexpression of these proteins did not improve vanillin resistance (Additional file 1: Fig. S3). We investigated several upregulated proteins that were associated with transcription and the stress response in a YRR1-deleted strain as candidates. These candidates included five stress response-related proteins (Crz1, Dsk2, Haa1, Def1, and Tma17) and two transcriptionrelated proteins (Taf10 and Mbf1). To determine the functions of these proteins/genes in a vanillin-resistant phenotype, their genes were overexpressed in BY4741 cells. With 5 mM vanillin treatment, the growth superiority of the recombinant strains was not as apparent as it was with 6 mM vanillin treatment. It was assumed that the contribution of a single gene was limited. An obvious resistant phenotype was observed only under severe vanillin stress. The maximum specific growth rates of the cells transformed with HAA1, TMA17, and MBF1 were 18%, 33%, and 15% faster than the rate of the control, strain BY4741(pJFE3), in the presence of 6 mM vanillin, respectively ( Fig. 4; Table 5). However, their maximum specific growth rates were more than 40% slower than those of BY4741(yrr1Δ, pJFE3). The specific consumption rates for vanillin in the strains overexpressing HAA1, TMA17, and MBF1 were 0.041, 0.041, and 0.029 g g −1 h −1 , respectively. Compared to the 0.034 g g −1 h −1 rate of the control strain, BY4741(pJFE3) ( Table 5), only HAA1 and TMA17 significantly increased the specific vanillin consumption rates. MBF1 accelerated growth to a greater extent than vanillin reduction. Compared to that contributed by HAA1 and TMA17, the improvement conferred by MBF1 was marginal. Other genes related to translation did not significantly enhance vanillin resistance (Additional file 1: Fig. S4). Then HAA1, TMA17 and MBF1 were overexpressed simultaneously in BY4741. Under 6 mM vanillin stress, the maximum specific growth rates of BY4741(HAA1, TMA17, MBF1) were 0.110, which was 26% higher than that of the control strain BY4741(pJFE3) but a bit lower than that of BY4741(TMA17). The specific consumption rates for vanillin was 32% higher than that of BY4741(pJFE3). The overexpression of multiple genes did not exhibit superimposed effect. It was probably because the overexpression of multiple genes bring burden to the cells as the result showed that the maximum specific growth rates of BY4741(HAA1, TMA17, MBF1) were 26% lower than that of BY4741(pJFE3) without stress. Furthermore, an ELISA was conducted with strains BY4741(yrr1Δ, pJFE1), BY4741(yrr1Δ, pJFE1-YRR1) and BY4741(pJFE1) in SC-URA medium without vanillin stress to determine the protein quantities of Tma17, Haa1 and Mbf1, which were replaced in situ by the corresponding genes tagged with 6×His at the C-terminal. The results showed that the protein abundance of Mbf1 (1.47-fold change in the proteome), Tma17 (1.59-fold change in the proteome) and Haa1 (1.56-fold change in the proteome) increased 1.75-, 1.55-and 1.36-fold in the strain BY4741(yrr1Δ, pJFE1) compared to the control strain, BY4741(pJFE1). YRR1 was inserted into the centromere plasmid pJFE1 under the control of the YRR1 promoter and the PGK1 terminator to get pJFE1-YRR1. The pJFE1-YRR1 was transformed into the BY4741(yrr1Δ) resulted in the YRR1-compensated strain BY4741(yrr1Δ, pJFE1-YRR1). The protein abundance of Mbf1, Tma17 and Haa1 in BY4741(yrr1Δ, pJFE1-YRR1) was decreased 35%, 28 and 22% compared to BY4741(yrr1Δ, pJFE1) (Fig. 5). These results indicated that the protein quantities of Haa1, Tma17 and Mbf1 are affected by YRR1. Besides, the growth rate of the YRR1compensated strain under 6 mM vanillin stress was decreased obviously compared to according to our previous work [12]. Thus, the deletion of YRR1 could upregulates HAA1 and TMA17 at the protein level to enhance vanillin resistance.
The proteome analysis and comparison of CEN. PK2-1 C(yrr1Δ) and CEN.PK2-1 C showed that only Mbf1 exhibited an increase in protein quantity (1.32fold change). There were 83 genes and approximately 22,000 point mutations identified as absent in CEN.PK family relative to S288C, which is derived from strain BY4741 [51,52]. This result indicated that the different genetic background strains probably influenced the gene regulatory network.
Haa1 is a critical transcription factor involved in the acid stress adaption of yeast cells by activating the expression of TPO2 and TPO3 encoding two drug/ H + -antiporters that presumably mediate the active   [53,54]. The overexpression of HAA1 also contributes to vanillin resistance. But the mechanism behind it is still unknown. Mbf1 is a transcriptional coactivator that mediates transcriptional activation by bridging the DNA-binding region of an activator and a TBP (TATA-binding protein) [55]. However, recently, Liang et al. have confirmed that the TBP binding function of Mbf1 was not the reason for the enhancement of vanillin resistance mediated by Mbf1p [56]. Under stress conditions, Tma17 (also named as Adc17) is the most important chaperone for proteasome biogenesis and assembly in yeast cells. Hanssum et al. [57] found that overexpression of TMA17 rescues 26 S proteasome defects, while deletion of TMA17 compromises proteasome integrity and cell fitness. The maintenance of adequately functional proteasomes is vital for cells and organisms to adapt to changes and survive in their environment [57]. The overexpression of TMA17 may increase proteasome levels or activity, and the exact mechanism needs to be analyzed in the future. Interestingly, the deletion of YRR1 only enhanced resistance to vanillin, not to furfural, or acetic acid, and not even to another kind of phenolic aldehyde, syringaldehyde (Additional file 1: Fig. S5). This result indicates that the enhanced resistance caused by YRR1 deletion is vanillin-specific.

Conclusions
Here, the changes in general protein composition perturbed by vanillin stress and YRR1 deletion were investigated via quantitative proteome analysis. The results indicated that S. cerevisiae adapted to vanillin stress by repressing ribosomal protein abundance and enhancing energy production. YRR1 deletion led to the upregulation of proteins related to the stress response and transcriptional and translational regulation, which provide unique perspectives from transcriptome analysis for comprehending the mechanisms behind the protective response of YRR1 deletion in yeast. The finding that two proteins (Haa1 and Tma17) can enhance vanillin resistance upon their overexpression suggests novel strategies for designing inhibitor-resistant ethanologenic yeast strains for bioethanol production.

Strains, plasmids, and medium
Saccharomyces cerevisiae strain CEN.PK2-1 C (MATa; ura3-52; trp1-289; leu2-3,112; his311; MAL2-8 C; and SUC2) and BY4741 (MATa, his3Δ1, leu2Δ0, met15Δ0 and ura3Δ0) were stored in our laboratory. The YRR1deleted strain CEN.PK2-1 C(yrr1Δ) (CEN.PK2-1 C derivative; yrr1::loxP) and BY4741(yrr1Δ) (BY4741 derivative; yrr1::loxP) [12] that were previously stored and constructed in our laboratory, respectively. The laboratory strain BY4741 was engineered for gene overexpression in this study. The genes CRZ1, HAA1, TMA17, DSK2, DEF1, MBF1, TAF10, RTC3, HSP12, CAF20 and TIF11 were amplified by polymerase chain reaction (PCR) using the genomic DNA of S. cerevisiae BY4741 as the template. The PCR products were gel-purified, and DEF1, DSK2 and CRZ1 were digested by BamHI and SalI, and the other genes were digested by BamHI and PstI. Then, the genes were ligated into plasmid pJFE3 predigested with BamHI and SalI or PstI. Plasmid pJFE3 is a 2-μ plasmid with the TEF1 promoter, the PGK1 terminator, and URA3 as a selection marker [58]. To overexpress of multiple genes (HAA1, TMA17 and MBF1) in one strain, the MBF1 was cloned into the 2-μ plasmid pIYC04 [59] under the control of the TEF1 promoter and the ADH1 terminator with NotI and SacI. The TMA17 was cloned into plasmid pIYC04 under the control of the PGK1 promoter and the CYC1 terminator with BamHI and SacI. Then the expression cassettes of TMA17 and MBF1 were cloned in to the plasmid pJFE3-HAA1 (HAA1 was cloned into pJFE3) using Gibson assembly. Next, the resulting recombinant plasmids were transformed into BY4741. The open reading frames of HAA1, MBF1, TMA17 were tagged at their natural chromosomal locations with six-His epitopes encoding proteins with C-terminally tags. Tagging was performed using homologous recombination as described in our previous report [12]. The expression cassette containing LoxP-KanMX4-LoxP cassettes and 6×His-tagged homologous sequence with targetgenes were transformed into the strains BY4741 and BY4741(yrr1Δ), and the mutants were screened in YPD or SC-URA medium with 800 mg L −1 G418. All strains and primers employed in this study are listed in Additional file 1: Tables S1 and S2, respectively. The YRR1 was cloned into the centromere plasmid pJFE1 and under the control of the YRR1 promoter and the PGK1 terminator. The TEF1 promoter of pJFE1 was replaced by the YRR1 promoter, producing the plasmid pJFE1-YRR1. Then the plasmid pJFE1-YRR1 was transformed into BY4741(yrr1Δ) to get the YRR1-compensated strain BY4741(yrr1Δ, pJFE1-YRR1). Synthetic complete medium (SD) or SC-URA medium (5 g L −1 ammonium sulfate, Sangon, China, 1.7 g L −1 yeast nitrogen base without amino acids, Sangon, China, CSM or CSM-URA, MP Biomedicals, Solon, OH, USA) was used for activation, batch fermentation, and protein sample preparation of recombinant strains upon the addition of 20 g L −1 glucose. Vanillin was added to the medium as indicated. All of the cultures were incubated at 30 °C.

Protein sample preparation for proteomic analysis
Yeast cells of strains BY4741, BY4741(yrr1Δ), CEN.PK2-1C and CEN.PK2-1C(yrr1Δ) were harvested when the OD 600 reached approximately 1.6 in logarithmic phase after different times in culture with SD medium with or without of 5 mM vanillin. All of the proteomic analyses were carried out as biological triplicates. The sample was ground by liquid nitrogen into cell powder and then transferred to a 5-mL centrifuge tube. Four volumes of lysis buffer (8 M urea, 1% Triton X-100, 10 mM dithiothreitol, 2 mM EDTA and 1% protease inhibitor cocktail) were added to the cell powder, sonicated three times on ice using a high-intensity ultrasonic processor. The debris was removed by centrifugation at 12,000×g for 10 min at 4 °C. Finally, the supernatant was collected, and the protein concentration was determined using a BCA kit (Beyotime, P0010S, Shanghai, China) according to the manufacturer's instructions. For digestion, the protein solution was reduced with 5 mM dithiothreitol for 30 min at 56 °C and alkylated with 11 mM iodoacetamide for 15 min at room temperature in darkness. The protein sample was then diluted by adding 100 mM triethylammonium bicarbonate (TEAB) to reduce the urea concentration to < 2 M. Finally, trypsin was added at a 1:50 trypsin-to-protein mass ratio for the first digestion overnight; the second 4-h digestion process was based on a 1:100 trypsin-to-protein mass ratio. A trypsin digestion peptide was labeled with a tandem mass tag (TMT), according to the TMT kit manufacturer's protocol (Thermo Fisher Scientific, 90068, Waltham, USA).

High-performance liquid chromatography fractionation and mass spectrometric analysis
The labeled peptides were fractionated by high pH reverse-phase high-performance liquid chromatography (HPLC) using an Agilent 300 Extend C18 column (5-µm particles, 4.6-mm ID, and 250-mm length). Briefly, the peptides were first separated into 60 fractions with a gradient ranging from 8 to 32% acetonitrile (pH 9.0) for more than 60 min. Then, the peptides were combined into 18 fractions and dried by vacuum centrifugation. The tryptic peptides were dissolved in solvent A (0.1% formic acid and 2% acetonitrile); the gradient was composed of an increase in solvent B (0.1% formic acid in 90% acetonitrile) from 6 to 20% over 20 min, from 20 to 33% in 11 min, to 80% in 4 min and then holding at 80% for the last 3 min, all at a constant flow rate of 320 nL min −1 on an EASY-nLC 1000 UPLC system (Thermo Fisher Scientific, EASY-nLC 1000, Waltham, USA). The peptides were subjected to a nano-electrospray ionization (NSI) source followed by tandem mass spectrometry (MS/MS) in a Q Exactive ™ Plus (Thermo) coupled online to a UPLC. The applied electrospray voltage applied was 2.0 kV. The m/z scan range was 350-1800 for a full scan, and intact peptides were detected in an Orbitrap at a resolution of 70,000. The secondary mass spectrometry scan range was set as a fixed first point of 100 m/z, and the resolution was 17,500. The data acquisition mode was based on a data-dependent scanning (DDA) program; the first 20 peptides with the highest signal intensity were selected to enter the high energy collision dissociation (HCD) collision cell, and then 30% of the fragmentation energy was used for fragmentation after the first-stage scanning; the same step was performed again during the secondary mass spectrometry analysis. To improve the effective utilization of the mass spectrometer, the automatic gain control (AGC) was set as 5E4, the signal threshold was set as 1E4, and the maximum injection time was set as 200 ms with 30 s of dynamic exclusion time.

Database searching and data analysis
The resulting MS/MS data were processed by using the MaxQuant search engine (v.1.5.2.8), a platform for mass spectrometry (MS)-based proteomic data analysis, to match the secondary mass spectrum with the theoretical spectrum in the protein database. The search parameters were as follows: trypsin/P was specified as the cleavage enzyme and to two missing cleavages were allowed; the mass resistance for the precursor ions was set as 20 ppm for the first search and 5 ppm for the main search, and the mass resistance for fragment ions was set as 0.02 Da. Carbamidomethyl on Cys was specified asa fixed modification, and oxidation of Met was specified as a variable modification. The method used to obtain relative protein expression ratios was set as TMT-6plex, since relative quantification of peptides/proteins was carried out based on the signal intensity of the reporter ions in the TMT tag fragment, and the false discovery rate (FDR) for protein identification and peptide spectrum match (PSM) identification was 1% eliminate false positive peptide matching. Proteins with an average increase or decrease ratio > 1.3-fold were classified into functional categories by GO annotation based on the Uniprot-GOA database (http:// www. ebi. ac. uk/ GOA/). Some undetermined proteins were identified using protein sequence-based algorithm software (InterProScan) to predict the GO function of the protein. Pathway enrichment analyses were carried out using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, and the subcellular location of the proteins was determined using software WoLF PSORT software. The function of representative proteins was estimated using the MIPS Functional Catalog database (http:// funsp ec. med. utoro nto. ca/, P-value < 0.01, Bonferroni correction). To explore the relationship between the proteome and the transcriptome, the relationship of them was contrasted and verified by utilizing the Statistical Package for the Social Sciences (SPSS) software (Spearman's correlation analysis), and their overlap is displayed in a Venn diagram generated using a bioinformatic tool (http:// bioin forma tics. psb. ugent. be/ webto ols/ Venn/).

Fermentation
A single colony was precultured in 3 mL of SC-URA at 30 °C for 24 h. Then, each culture was transferred into 10 mL of fresh medium with an initial OD 600 of 0.2 and cultured overnight. Next, the cells were inoculated into 100-mL flasks containing 40 mL of fermentation medium in the presence or absence of 6 mM vanillin with an initial OD 600 of 0.2. The fermentation was carried out at 30 °C and 200 rpm.

Analysis of extracellular vanillin
The concentrations of vanillin and vanillyl alcohol were determined by a HPLC Waters system e2695 (Waters, USA) equipped with an Xbridge TM -C18 column (Waters, USA). The peaks were detected at room temperature using ultraviolet detector (PDA-2998) at 210 nm with a mobile phase containing 40% absolute methanol (Chromatographic grade, Fisher Chemical, USA) supplied at a flow rate of 0.6 mL min −1 [12].

Growth and vanillin consumption assays
The maximum growth rates were regarded as the linear regression coefficients of the ln (OD 600 ) versus time during the exponential growth phase [60]. The specific consumption rates of vanillin were calculated using the following equation: where r is the specific consumption rate during the phase from sampling point m to sampling point n; A, B, and t are the metabolite concentration, biomass concentration, and time, respectively, at sampling points n, i, and m, as previously described [12,61]. The biomass concentrates were investigated to determine their correlation with OD 600 and the dry cell weight (DCW). Different strains exhibited different coefficients between OD 600 and DCW.

NAD(P) + , NAD(P)H and ATP/ADP assays
Cells were cultured in 40 mL of SC-URA medium with or without 6 mM vanillin in a 100-mL flask starting at an initial OD 600 of 0.2. Rapid sampling, quenching, and metabolite extraction of biomass (in approximately 10 mL of culture at exponential phase) for the NAD(P) + and NAD(P)H assays were performed as described previously [61]. Subsequently, the concentration of the resultant extract was quantified using a sensitive enzymatic cycling assay as previously reported [28,62].
Samples of 1 mL of culture were taken for intracellular ATP/ADP ratio analysis and quenched by liquid nitrogen. The yeast cells were suspended in 0.5 mL of 0.6% (w/w) HClO 4 , after which cell extraction was vigorously mixied in a freezing grinder (Jin Xing, China) for 10 min. The cellular debris was removed by centrifugation at 13,000 rpm for 10 min and subsequent extracts were neutralized with 3 M KOH (pH 7.0). The ratio of ATP/ADP in the extracts was determined using an ADP/ATP ratio assay kit (Sigma, USA) according to the manufacturer's instructions. In summary, first, ATP immediately reacted with the luciferase substrate d-luciferin to produce light, and the luminescence was measured (relative light units, RLU) on a luminometer for an ATP assay (RLUA). Then, after 10 min of incubation, luminescence was measured for ATP (RLUB) in a background. In the second step, an ADP reagent was added to each well, ADP was converted to ATP through an enzymatic reaction and then the resulting product was allowed to react with the d-luciferin as in the first step, which was measured as RLUC. The ATP/ADP ratio was calculated using the following equation: ,