Design and characterization of allantoin-inducible expression systems in budding yeast

Abstract

Background

Saccharomyces cerevisiae has been extensively employed as a host for the production of various biochemicals and recombinant proteins. The expression systems employed in S. cerevisiae typically rely on constitutive or galactose-regulated promoters, and the limited repertoire of gene expression regulations imposes constraints on the productivity of microbial cell factories based on budding yeast.

Results

In this study, we designed and characterized a series of allantoin-inducible expression systems based on the endogenous allantoin catabolic system (DAL-related genes) in S. cerevisiae. We first characterized the expression profile of a set of DAL promoters induced by allantoin, and further combined with the galactose-inducible (GAL) system to create a highly responsive genetic switch that efficiently amplifies the output signals. The resulting allantoin–GAL system could give a ON/OFF ratio of 68.6, with 6.8-fold higher signal output over that of direct P_DAL2-controlled gene expression. Additionally, when a centromeric plasmid was used for EGFP expression, the ON/OFF ratio was increased to > 67.2, surpassing the EGFP expression levels driven by the DAL2 promoter. Subsequently, we successfully demonstrated that allantoin–GAL system can be used to effectively regulate carotenoid production and cell flocculation in S. cerevisiae.

Conclusions

In summary, we characterized several allantoin-inducible DAL promoters from budding yeast and further developed a layered allantoin–GAL system that utilizes the DAL2 promoter to regulate the galactose regulon in budding yeast. The resulting allantoin–GAL system could give an impressive ON/OFF ratio that surpassed the traditional P_DAL2-controlled gene expression. It is anticipated that utilizing our allantoin-inducible system in budding yeast with allantoin as the alternative nitrogen source might favor the low-cost production of biochemicals and pharmaceuticals.

Background

The budding yeast S. cerevisiae is the first eukaryotic organism to have its genome fully sequenced [1]. Due to the extensive research on its physiology and genetics, including the transcriptome analysis [2], proteome analysis [3], and metabolome analysis [4], S. cerevisiae is considered one of the organisms with the most comprehensive experimental datasets available. As a generally recognized as safe (GRAS) microorganism, S. cerevisiae combines the advantages of single-cell organisms such as ease of genetic manipulation and the ability for post-translational modifications found in higher eukaryotes [5, 6]. Furthermore, due to its exceptional characteristics including robust growth properties, tolerance to environmental stressors such as low oxygen and acidity, as well as low susceptibility to contamination by bacteriophages and bacteria, S. cerevisiae is frequently employed as a eukaryotic microbial cell factory for the production of biochemicals and recombinant proteins, e.g., naringenin [7], vanillin, [8] and insulin [9]. The utilization of microorganisms for the synthesis of biochemicals often necessitates the introduction of multiple genes encoding enzymes involved in relevant metabolic pathways [10, 11].

To construct efficient yeast cell factories, the selection of an appropriate expression system for target genes is of paramount importance in attaining this objective. Promoter systems serve as a crucial tool to drive effective transcription of the genes-of-interest (GOI) and ensuring optimal expression efficiency of target proteins. Typically, there are two main options for transcribing target genes: constitutive or inducible promoters. Strong constitutive promoters such as P_PGK1, P_TEF1 and P_ADH1 [12, 13] from S. cerevisiae typically resulted in high-level enzyme expression, which may have detrimental effects during the early growth stage and might lead to strain instability. Therefore, the inducible expression systems capable of enabling conditional regulation of gene expression offer a broader spectrum of choices for practical industrial applications. Inducible promoters initiate gene expressions by responding to physical or chemical signals in the environment, thus the selection of the inducible expression system also critically depends on the choice of inducer and induction mechanism. The carbon source-dependent inducible regulatory systems are widely employed as a controllable induction system, wherein sucrose [14], ethanol [15], galactose [16], etc., serve as alternative inducers for regulating gene expression. Nevertheless, due to the preference of yeast cells for glucose as the carbon source, gene expression controlled by these alternative carbon sources is commonly inhibited under glucose repression [17,18,19].

Physical stimuli such as light and temperature can efficiently and conveniently regulate the expression of GOI without significantly perturbing the composition of culture medium. Currently, a number of researchers have devised physical induction systems in budding yeast. Zhao et al. developed OptoEXP (the light-inducible gene circuit) and OptoINVRT (the light-repressible gene circuit), through modifications to the GAL system [20]. Despite the ease of controlling light as a signal, challenges arise in terms of inadequate penetration and uneven light intensity in high-density cultivation, further hindering its industrial implementation. Recently, temperature-responsive expression systems have drawn increasing attention. By employing directed evolution techniques, a temperature-sensitive Gal4 mutant, named Gal4M9, was generated [21]. This variant was subsequently utilized for engineering the GAL system in S. cerevisiae, resulting in the development of a temperature-responsive GAL system. Taking lycopene as an example, the accumulation of lycopene was 177% higher in the strain controlled by Gal4M9 than that in the Gal4 control strain [21]. More recently, intein-mediated Gal4 transcriptional factor [22] has also been explored for temperature control for complete biosynthesis of sanguinarine in budding yeast [23].

Chemical inducers exhibit the capacity to regulate a wide dynamic range of gene expression and possess broad applicability, rendering them the preferred choice for industrial applications. S. cerevisiae possesses a set of DAL genes for allantoin utilization [24, 25], which are repressed under normal nitrogen source such as NH₄⁺. Despite the well-known characteristics of allantoin catabolic pathways in S. cerevisiae, no attempts have been made thus far to genetically modify the components of the DAL gene cluster for biochemical productions. In this study, we aim to develop the allantoin-dependent gene expression system in S. cerevisiae for bioproduction applications. We first characterized allantoin-inducible promoters, namely, P_DAL2 and P_DAL3 from S. cerevisiae, by using enhanced green fluorescent protein (EGFP) as a reporter. We found that DAL2 promoter exhibited the good performance with less leakiness and impressive ON/OFF ratio (14.4-fold) after replacing the nitrogen source with allantoin. To further improve the signal output magnitude, we subsequently constructed a layered genetic network by refactoring the yeast galactose regulon by deleting Gal80 repressor and introducing P_DAL2-controlled Gal4 activator. As a result, the engineered allantoin–GAL system was tightly restricted by normal nitrogen source, and the EGFP expression was induced by 68.6-time when using allantoin as alternative nitrogen source. When a centromeric plasmid was used for EGFP expression, we found that the ON/OFF ratio was increased to > 60. We further explored the allantoin–GAL system for metabolic engineering applications. Conditional carotenoid production in S. cerevisiae was achieved by introducing the allantoin–GAL genetic switch, resulting in a titer up to ~ 80.0 mg/L. Furthermore, the allantoin–GAL switch was also demonstrated for conditional cell flocculation. Taken together, we have constructed an effective allantoin-inducible system in budding yeast, which might have a broad utility for metabolic engineering applications.

Materials and methods

Strains and reagents

Escherichia coli strain DH5α or TOP10 was used for general plasmid constructions and the strain was maintained at 37 °C in Luria–Bertani (LB) medium (10 g/L tryptone, 5 g/L yeast extract, and 10 g/L NaCl). Strain BY4741 from EUROSCRAF and JS-BE5-PEST (laboratory stock) was used as the starting yeast strain for further genetic modifications. The engineered yeast strains were cultured at 30 °C in the yeast-peptone-dextrose (YPD) medium (20 g/L tryptone, 10 g/L yeast extract, 20 g/L glucose), the yeast-nitrogen-base-glucose (YNBD) medium with appropriate dropouts or the yeast-allantoin-base-glucose (YABD) medium (2 μg/L biotin, 400 μg/L calcium pantothenate, 2 μg/L folic acid, 2000 μg/L inositol, 400 μg/L nicotinic acid, 200 μg/L 4-aminobenzoic acid, 400 μg/L pyridoxine hydrochloride, 200 μg/L riboflavin, 400 μg/L thiamine hydrochloride, 500 μg/L H₃BO₃, 40 μg/L CuSO₄, 100 μg/L KI, 200 μg/L FeCl₃, 400 μg/L MnSO₄, 200 μg/L Na₂MoO₄, 400 μg/L ZnSO₄, 1 g/L KH₂PO₄, 0.5 g/L MgSO₄, 1 g/L NaCl, 0.1 g/L CaCl₂, 20 g/L glucose, and 1.7 g/L allantoin) with appropriate dropouts. Enzymes (High-Fidelity Phusion polymerase, SacI-HF, BamHI-HF, T4 DNA ligase and rSAP) were purchased from New England Biolabs (Ipswich, MA, USA). PCR purification kit, gel extraction kit, and plasmid DNA extraction kit were all purchased from BioFlux (Shanghai, China). All chemicals used in this study were obtained from Sigma-Aldrich or otherwise stated.

Construction of plasmids and strains

The oligonucleotides used in this study are listed in Supplementary Table S1. All the DNA fragments were PCR amplified using High-Fidelity Phusion DNA polymerase. The plasmid pRS425Gal1-EGFP [26] was used as the backbone for the construction of pRS425DAL2/3-EGFP. The DAL2/DAL3 promoter sequences were amplified from the genomic DNA of S. cerevisiae using the primer pairs listed in Supplementary Table S1 and digested with SacI-HF and BamHI-HF, then inserted into pRS425Gal1-EGFP digested with the same enzymes to yield pRS425DAL2/3-EGFP. The gRNA expression plasmids were all constructed using the Golden Gate assembly method as previously reported [27], using the oligonucleotides listed in Supplementary Table S1. All the details of plasmids are provided in Supplementary Table S2.

The CRISPR/Cas9-mediated genome editing was conducted using the previously described electroporation method [28]. Briefly, yeast cells (50 μL) were mixed with approximately 2 μg of the genome editing cassette and subjected to electroporation in a 0.2 cm cuvette at 1.6 kV. Immediately after electroporation, the cells were recovered by a mixture containing 1 M sorbitol and YPD medium in a ratio of 1:1 (1 mL total volume) on a rotary shaker for 1 h. Subsequently, the cells were collected by centrifugation at 5000 rpm for 1 min, washed, and resuspended in ddH₂O (50 μL). The resuspended cells were then plated onto YNBD-URA-TRP plates. Randomly selected colonies underwent diagnostic PCR validation to confirm genome editing events. Both gene modifications including Δgal80 and P_DAL2:Gal4 were achieved utilizing CRISPR/Cas9 technology. Detailed information regarding all strains can be found in Supplementary Table S3.

Measurement of OD₆₀₀ and green fluorescence intensity

Cell cultures were cultivated in YNBD and YABD, respectively, at 30 °C and 250 rpm in a rotating shaking incubator. Samples was periodically collected for the measurement of OD₆₀₀ and EGFP intensity using the microplate reader (Biotek, Synergy H1). The EGFP intensity was monitored with excitation wavelength set at 480 nm and emission wavelength set at 520 nm. The induction ratios were calculated as the ratio of EGFP/OD₆₀₀ with allantoin treatment to that without allantoin treatment. To eliminate interference from complex components in the culture medium, all samples were centrifuged and resuspended in ddH₂O prior to measuring the OD₆₀₀ and EGFP values.

Carotenoid production

For small-scale carotenoid production, the yeast culture was inoculated at a seeding rate of 1% into 12-mL shake tubes containing 2 mL YNBD or YABD medium, respectively. These cultures were incubated at 30 °C and 250 rpm, the sample of the cell culture (300 μL) was collected after 96 h and centrifuged to remove the supernatant. Subsequently, the harvested cells were then suspended in screw-cap tubes with 300 μL acetone and disrupted using a bead beater (OMNI, USA). After centrifugation, the cellular residue was discarded while the supernatant containing carotenoids was transferred to new brown tubes for protection against light-induced degradation. The acetone extract underwent analysis using high-performance liquid chromatography (HPLC) on an LC-20A model Shimadzu system equipped with a C18 column (250 mm × 4.6 mm, 5 μm). The mobile phase consisted of methanol and acetonitrile in a ratio of 1:1. During HPLC analysis, the flow rate remained constant at 1.0 mL/min while maintaining the column temperature at 40 °C. Carotenoids were detected at a wavelength of 450 nm as employed as previously reported [29].

Results and discussion

Characterization of DAL promoters induced by allantoin utilization

Under normal physiological conditions, allantoin is sequestered within the vacuole of S. cerevisiae as an intermediate product of purine degradation [30]. In scenarios where other nitrogen sources are restricted, it can be ultimately catabolized by a series of allantoin degradation enzymes into ammonia [31], which is then assimilated into various nitrogen-containing compounds to support cell growth and metabolism. The metabolic and regulatory pathways of allantoin metabolism, as depicted in Fig. 1, encompass one of the largest metabolic gene clusters known as DAL cluster in S. cerevisiae. The DAL cluster comprises contiguous six out of eight proteins involved in allantoin degradation pathway [24], namely DAL1 (allantoinase), DAL2 (allantoicase), DAL3 (ureidoglycolate hydrolase), DAL4 (allantoin permease), DAL7 (an isoform of malate synthase) and DCG1 (a probable allantoin racemase). DAL5 (encodes for allantoate permease) and DUR1,2 (encodes for urea amidolyase) are the two other genes which are not located in the DAL cluster [32]. Normally, when allantoin or its metabolic intermediates are used as the main nitrogen source, the transcriptional repressions of enzymes are rapidly de-activated [33, 34]. Therefore, allantoin can function as a dual-functional component in the culture medium: both as an inducer and a nitrogen source. Considering the direct condensation of allantoin from glyoxylic acid and urea [35], this method offers advantages such as facile access to raw materials, low production costs, a straightforward production process, high yield, and convenient post-processing in industrial-scale allantoin production. The application of allantoin as both a nitrogen source and an inducer obviates the need of the supplementation of conventional chemical inducers or physical inducing devices to control the expression of targeted genes. The establishment of an allantoin-induction system in budding yeast holds significant potential for industrial applications. Among the DAL-related genes, DAL2 and DAL3 are the two genes located on the DAL cluster, encoding enzymes that decompose precursors to generate urea, a more accessible nitrogen source. Their expression activities are directly modulated by nitrogen metabolism. We selected DAL2 and DAL3 promoters to regulate the expression of GOI, and compared the relative strength of EGFP expression controlled by different DAL promoters when allantoin is utilized as the nitrogen source.

Fig. 1

Schematic diagram of metabolic and regulatory pathways of allantoin metabolism in budding yeast. S. cerevisiae possesses a series of DAL genes for allantoin utilization, which are repressed under normal nitrogen source. Among the DAL-encoded enzymes, there are six proteins primarily involved in allantoin utilization, namely DAL1 (allantoinase), DAL2 (allantoicase), DAL3 (ureidoglycolate hydrolase), DAL4 (allantoin permease), DAL7 (an isoform of malate synthase), and DUR1,2 (urea amidolyase). Additionally, DAL81 and DAL82 function as transcriptional activators of the DAL genes, whereas DAL80 and GZF3 act as inhibitors of their transcription. The green arrows indicate activation, and the red dashed arrows represent inhibition

After replacing the P_Gal1 of pRS425Gal1-EGFP with DAL2/3 promoters, we inoculated the BY4741 strains carrying these plasmids into YNBD and YABD medium (Fig. 2A). As shown in Fig. 2B, we found that the EGFP expressions were successful activated by both DAL2 and DAL3 promoters when allantoin was used as alternative nitrogen source. Notably, the DAL2 promoter exhibited significantly higher expression levels of EGFP, and approximately 14.4-time stronger EGFP expression after 48 h of cultivation was observed when compared to that of the control group. In comparison, the DAL3 promoter can also be triggered by allantoin and the ON/OFF was slightly lower than that of DAL2 promoter, indicating that following the induction by allantoin or its metabolic intermediates, the DAL3 promoter exhibits a modest increase in activating downstream gene expression. As depicted in Fig. 2C, we also monitored the growth profile of yeast cells on different nitrogen sources. In contrast to the direct utilization of NH₄⁺ as a nitrogen source, BY4741 yeast cells exhibit a shorter lag phase when utilizing allantoin. Moreover, after 48 h of cultivation, they could ultimately attain a relatively high cell density, suggesting that allantoin can serve as a viable alternative nitrogen source to support cell proliferation.

Fig. 2

Characterization of DAL promoters induced by allantoin. A Schematic diagram of plasmid composition for characterization of the strength of DAL promoters. The engineered yeast cells were cultivated in normal YNBD medium (repression state) or YABD medium (activation state). B Comparison of EGFP expressions under the control of DAL2/DAL3 promoter in normal YNBD medium (repression state) or YABD medium (activation state). C The cell growth profile of BY4741-derived strains in normal YNBD medium (repression state) or YABD medium (activation state). All experiments were performed in triplicate, and data represent the mean values with standard deviations

Certainly, there is still a scope for further optimization of our allantoin-induction system. DAL2 is a known substrate for nonsense-mediated decay (NMD) [36,37,38], and its RNA half-life is significantly reduced in cells with functional NMD capabilities. Similar to the degradation regulation observed in the arginine metabolism gene CPA1 transcription products through the arginine-sensitive arginine inhibitory peptides encoded 5′ upstream of the open reading frame (uORF) on CPA1 [39], this phenomenon may be attributed to the presence of the 5′ uORF of DAL2 [36]. The impact of environmental pressure on NMD of DAL2 in industrial settings has not been thoroughly investigated yet, and it might lead to aberrant activity within our allantoin-induction system that relies on NMD dependency, which requires further investigation on allantoin-induction system.

A layered-genetic design of coupling to the GAL regulon to improve the performance of allantoin switch

Although we have identified that the DAL2 promoter has a relatively good ON/OFF ratio, the promoter strength of DAL promoters is still not as strong as the commonly used P_PGK1, P_TEF1 and P_GAL1. Therefore, direct control of gene expression using the allantoin-induced DAL promoters could only yield suboptimal protein expressions with limited applications, and further modifications are needed to enhance signal output of the allantoin-inducible system. Recently, we have successfully constructed a number of layered-genetic circuits for creating various expression systems in eukaryotic and prokaryotic microorganisms. For instance, by substituting the Gal4 promoter with the copper-inducible P_Cup1 and replacing the Gal80 promoter with the copper-repressible P_Ctr1, the GAL system can exhibit an enhanced signal output response to copper ions [26]. Additionally, we have also engineered a cyanamide-inducible GAL system in budding yeast by coupling the cyanamide-inducible promoter P_Ddi2 with the GAL system [40]. The induced expression level of EGFP by this cyanamide-induced GAL system is 13.5 times that of the direct control of EGFP expression using the Ddi2 promoter. Moreover, through the integration of the GPCR signaling pathway with the GAL system using a synthetic transcription factor (sTF) composed of pheromone responsive domain (PRD) from Ste12 and DNA binding domain (DBD) of Gal4 [41], we can significantly augment the sensitivity and amplitude of signal output in the engineered yeast sensors by employing layered-genetic design for signal cascade amplification [42]. Similarly, the thermoresponsive quorum-sensing (ThermoQS) circuit in E. coli enabled the detection and conversion of transient heat stimuli into accumulation of quorum-sensing molecules such as acyl-homoserine lactone (AHL), thereby establishing a genetic circuit capable of recording transient physical stimuli [43, 44].

As shown in Fig. 3A, we also employed the layered design to reprogram the allantoin-inducible system and created an allantoin–GAL switch. Firstly, we deleted the Gal80 gene in BY4741 strain, which acts as a negative regulator of the Gal4 activator. Subsequently, we substituted the promoter of the Gal4 gene with the DAL2 promoter, resulting in a modified strain designated as JS-Allan. Following transformation of JS-Allan with pRS425Gal1-EGFP plasmid, we cultivated it in YNBD and YABD medium to monitor the changes in EGFP signal values. As depicted in Fig. 3B, while maintaining a minimal leaky expression from the GAL system, our allantoin–GAL switch exhibited an elevated EGFP signal output value that was 6.8-fold higher compared to that of direct control of gene expression using the DAL2 promoter after 96 h cultivation, confirming the successful amplification of allantoin input signals through the activation by the Gal4 activator. To further evaluate the robustness of allantoin–GAL switch, we also constructed a single copy plasmid using the centromeric replication origin for EGFP expression. Although the absolute EGFP/OD₆₀₀ was slightly reduced, we found that the ON/OFF ratio was further increased to > 60 because of the reduced basal expression level (Fig. 3B), which is expected to greatly favor metabolic engineering applications. Overall, the performance of our engineering allantoin–GAL switch was comparable to our previously constructed layered-genetic circuits [26, 40, 41].

Fig. 3

Characterization of allantoin–GAL expression system with signal amplification effect. A Schematic design of the allantoin–GAL system. By deleting the Gal80 gene (encoding the negative regulator of Gal4 transcriptional activator) and replacing the Gal4 native promoter with the DAL2 promoter in BY4741 strain, the input signal from allantoin inducer can be amplified by Gal4, which binds to the upstream activation sequences (UAS_G) of GAL genes and would eventually result in a stronger signal output of GOI under the GAL promoter. Once the Gal4 transcriptional activator is expressed, it can persistently drive the GOI expression under the GAL promoters. GOI, genes-of-interest. B The EGFP expression of engineered yeast cells in normal YNBD medium (repression state) or YABD medium (activation state). Strain BY4741 transformed with plasmid pRS425DAL2-EGFP and strain JS-BE5-Allan transformed with plasmid pRS415Gal1-EGFP and pRS425Gal1-EGFP was used to collect the data. All experiments were performed in triplicate, and data represent the mean values with standard deviations

Harnessing the allantoin–GAL switch for regulating the carotenoid synthesis

β-Carotene (C₄₀H₅₆) is a crucial type of carotenoid, playing a significant role in various aspects such as maintaining normal visual health, enhancing immune function, and contributing to antioxidant activity, and cardiovascular disease prevention [45, 46]. Recently, our group employed a multidimensional approach to enhance the carotenoid production in S. cerevisiae. These strategies included introducing the non-oxidative glycolysis (NOG) pathway to augment acetyl-CoA supply, restricting the ergosterol pathway to reduce farnesyl pyrophosphate (FPP) consumption, and incorporating human lipid binding/transfer protein saposin B (hSapB) as a compensatory space for carotenoid storage. Consequently, the engineered strain JS-BE5-PEST achieved a titer of β-carotene at 112.9 mg/L in YNBD medium under shaking flask conditions [29]. In this study, we introduced the allantoin–GAL switch into this strain to achieve conditional β-carotene synthesis by varying nitrogen sources (Fig. 4A). As depicted in Fig. 4B, following 48 h cultivation period on YABD agar plate, the resulting JS-BE5-Allan strain exhibited a distinct orange color, whereas no obvious orange color was observed when this strain was cultivated on YNBD agar plate. These findings confirmed that the allantoin–GAL switch can be tightly regulated by allantoin and amplify signals within this particular strain. Subsequent shake tube fermentation results demonstrated that after 96 h of fermentation, the JS-BE5-Allan strain produced β-carotene at a level of approximately 80.0 mg/L in YABD medium (Fig. 4C). Although the growth of engineered strain JS-BE5-Allan was decelerated in allantoin-containing medium than that of JS-BE5-PEST in the medium with the normal nitrogen source NH₄⁺ [29], the β-carotene titer is still comparable to the predecessor of JS-BE5-PEST without Pgi1 restriction in rich YNBD medium with yeast dropouts. We also noticed that the engineered carotenoid-producing yeast could not grow well in both YNBD and YABD without the supplementation of amino acids (Fig. 4D), whereas its parental strains could grow normally as shown in Fig. 2C. It is likely that that some genetic modifications in the carotenoid-producing yeast [29] are unfavorable for the yeast cell growth in amino acids-free medium. To fully release the potential of carotenoid production using the allantoin–GAL switch, it might require the reconstruction of the carotenoid biosynthesis module at more neutral genomic locations.

Fig. 4

Carotenoid synthesis by allantoin-induced switch. A Schematic design of allantoin-induced carotenoid synthesis. After deleting the Gal80 gene and substituting the Gal4 promoter with the DAL2 promoter, the JS-BE5-Allan strain exhibits responsiveness to allantoin signal for β-carotene production. IPP, isopentenyl pyrophosphate; DMAPP, dimethylallyl diphosphate; FPP, farnesyl pyrophosphate. B The color differences of carotenoid-producing strains on the normal YNBD plate (repression state) or YABD plate (activation state). Image was captured after 48 h cultivation. C The carotenoid levels of engineered yeast cells in normal YNBD medium (repression state) or YABD medium (activation state) after 96 h cultivation. D The cell growth profile of engineered yeast cells in normal YNBD medium (repression state) or YABD medium (activation state). All experiments were performed in triplicate, and data represent the mean values with standard deviations

Exploring the allantoin–GAL switch for conditional cell flocculation

During the product separation process, it is necessary to remove the suspended yeast cells prior to further processing of the fermented product. This may require time-consuming and costly techniques such as centrifugation or filtration. As a convenient and cost-effective approach, natural sedimentation can be considered as a viable option in industrial processes for the separation of biomass from fermentation media through cell aggregation (also known as flocculation). Flocculation is defined as the calcium-dependent, reversible, and asexual process of yeast cell aggregation, resulting in the formation of flocs containing a high number of cells that can rapidly settle at the bottom of liquid culture medium [47]. S. cerevisiae possesses five flocculation genes namely Flo1, Flo5, Flo9, Flo10, and Flo11 [48]. All Flo proteins are glycosylphosphatidylinositol (GPI)-anchored glycoproteins with an identical three-domain structure [49].

Among these proteins, Flo1 acts as a sugar-sensitive cell surface flocculin that utilizes its C-terminal GPI anchor sequence to directly bind to mannose residues on adjacent cells, leading to yeast cell aggregation and subsequent sedimentation [50]. In this study, the allantoin–GAL switch was employed for conditional regulation of flocculation in budding yeast. The promoter region of the Flo1 gene was replaced with GAL1 promoter allowing control over its expression through allantoin (Fig. 5A). As shown in Fig. 5B, after 24 h of cultivation in YABD medium, the JS-Allan P_GAL1:Flo1 strain formed aggregates and settled at the bottom of the shake tube, while under identical conditions, the yeast cells cultured in YNBD medium remained well suspended. This observation suggested that the allantoin–GAL switch could be utilized to control and regulate the flocculation behavior of yeast strains to meet specific industrial requirements.

Fig. 5

Conditional yeast cell flocculation by allantoin-induced switch. A Schematic design of allantoin-induced cell aggregation in yeast. The native FLO1 promoter in JS-Allan strain is replaced by the GAL1 promoter. The allantoin-induced signal triggers the activation of Flo1 gene expression under the control of GAL1 promoter, resulting in cell aggregation and subsequent sedimentation. B The morphological disparities between the engineered yeast cells following 24 h of cultivation in the normal YNBD medium (repression state) or YABD medium (activation state). All experiments were performed in triplicate and only a representative image was provided

Conclusion

Compared to the carbon source-dependent inducible regulatory systems that are commonly inhibited under glucose repression [17,18,19], allantoin can serve as both a nitrogen source and an inducer, which greatly simplify the fermentation process. We optimized the allantoin-inducible system by developing a layered allantoin–GAL system that utilizes the DAL2 promoter to regulate the galactose regulon in budding yeast. The resulting allantoin–GAL system could give an impressive ON/OFF ratio that surpassed the traditional P_DAL2-controlled gene expression. Subsequently, we applied this layered design of allantoin–GAL system for the β-carotene production, and the titer of β-carotene in JS-BE5-Allan reached ~ 80.0 mg/L when growing in allantoin-containing YABD medium. Furthermore, conditional flocculation experiments on yeast cells induced by allantoin was demonstrated to expand the range of industrial applicability of our system.