Abstract Background Non-conventional yeasts hold significant potential as biorefinery cell factories for microbial bioproduction. Currently, gene editing systems used for these yeasts rely on antibiotic and auxotrophic selection mechanisms. However, the drawbacks of antibiotics, including high costs, environmental concerns, and the dissemination of resistance genes, make them unsuitable for large-scale industrial fermentation. For auxotrophic selection system, the engineered strains harboring auxotrophic marker genes are typically supplemented with complex nutrient-rich components instead of precisely defined synthetic media in large-scale industrial fermentations, thus lack selection pressure to ensure the stability of heterologous metabolic pathways. Therefore, it is a critical to explore alternative selection systems that can be adapted for large-scale industrial fermentation. Results Here, a novel glucose-dependent selection system was developed in a high pullulan-producing non-conventional strain A. melanogenum P16. The system comprised a glucose-deficient chassis cell Δpfk obtained through the knockout of the phosphofructokinase gene (PFK) and a series of chromosomal integration plasmids carrying a selection marker PFK controlled by different strength promoters. Utilizing the green fluorescent protein gene (GFP) as a reporter gene, this system achieved a 100% positive rate of transformation, and the chromosomal integration numbers of GFP showed an inverse relationship with promoter strength, with a customizable copy number ranging from 2 to 54. More importantly, the chromosomal integration numbers of target genes remained stable during successive inoculation and fermentation process, facilitated simply by using glucose as a cost-effective and environmental-friendly selectable molecule to maintain a constant and rigorous screening pressure. Moreover, this glucose-dependent selection system exhibited no significant effect on cell growth and product synthesis, and the glucose-deficient related selectable marker PFK has universal application potential in non-conventional yeasts. Conclusion Here, we have developed a novel glucose-dependent selection system to achieve customizable and stable multilocus chromosomal integration of target genes. Therefore, this study presents a promising new tool for genetic manipulation and strain enhancement in non-conventional yeasts, particularly tailored for industrial fermentation applications. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-024-02531-3. Keywords: Non-conventional yeasts, Phosphofructokinase, Glucose-deficiency, Chromosomal integration, Selection system Background Non-conventional yeasts represent a diverse group of yeasts beyond well-established model organisms such as Saccharomyces cerevisiae, Candida albicans, and Schizosaccharomyces pombe [[37]1]. They hold significant potential as biorefinery cell factories for microbial bioproduction owing to their inherent capacity to utilize a wide range of carbon sources and produce desired chemicals and proteins [[38]2]. Additionally, non-conventional yeasts exhibit high tolerance to various stressful environments in biological processes, including osmotolerance, thermotolerance, and acid resistance [[39]3]. However, the application of advanced metabolic engineering efforts in non-conventional yeasts remains limited due to the scarcity of well-characterized synthetic biology components and optimized genome engineering tools, impeding further progress in this field. Aureobasidium spp., belonging to the subphylum Pezizomycotina within the class Dothideomycetes, are polyextremotolerant, melanin-producing, and dimorphism non-conventional yeasts with high biotechnological and biological importance [[40]4]. They demonstrate a high adaptability to complex and varied environments, leading to their wide ecological distribution in marine, soil, beehive, polar marine sediments, desert, limestone, and mangrove ecosystems [[41]5]. Moreover, Aureobasidium spp. possess the ability to synthesize various economically valuable primary and secondary metabolites such as fumaric acid [[42]6], polymalate [[43]7, [44]8], pullulan [[45]9], intracellular lipids [[46]10], liamocins [[47]11], siderophore [[48]12], Aureobasidin A [[49]13], β-glycan [[50]5], etc. Altogether, Aureobasidium spp. have garnered significant attention as promising eukaryotic hosts for industrial microbial bioproduction. Genome editing plays a crucial role in enhancing the performance of non-conventional yeasts as cell factories by precisely engineering metabolic pathways to redirect carbon flow towards specific products [[51]2]. Currently, the majority of metabolic engineering endeavors in Aureobasidium spp. rely on a Cre/LoxP site-specific recombination system, which allows for sequential gene deletion and expression, and has proven successful in investigating biosynthesis and regulation mechanism and constructing new cell factories [[52]14]. However, it is noteworthy that the selectable markers of this system are limited to a few antibiotic resistance genes towards nourseothricin, hygromycin, and bleomycin [[53]6, [54]11, [55]14]. Consequently, the maintenance of heterologous genes in engineered strains relies on the presence of antibiotics. However, the high cost and serious environmental issues associated with problems make them unsuitable for use in large-scale industrial fermentation, ultimately leading to a decline in the performance of engineered strains. Moreover, the excessive utilization of antibiotic resistance genes in metabolic engineering carries substantial risks, including their wide dissemination across multiple microbial taxa and their increasing accumulation in environments, thereby posing significant threats to public health and food safety [[56]15, [57]16]. Therefore, the development of non-antibiotic selectable markers in Aureobasidium spp. is essential. Auxotrophic marker genes, which do not rely on antibiotics, are extensively employed in genetic manipulation for microbial cell factory construction. In the case of non-conventional yeasts, a variety of auxotrophic strains have been developed as chassis cells, including those with amino acids, nucleotides, and vitamin auxotrophs [[58]17–[59]19]. These strains require specific nutritional formula that can be provided in defined synthetic media, facilitating control over growth and metabolism, and simplifying genetic manipulation. Moreover, due to their limited survival ability in the natural environment, they pose a lower risk of biological contamination [[60]20]. However, the growth and product biosynthesis of auxotrophic strains are inherently constrained by the insufficient supplementation of essential growth factors such as amino acids and nucleotides [[61]21]. This limitation becomes particularly significant in industrial fermentation, where there is a desire to streamline the process and reduce costs. In large-scale industrial fermentations, nutritional complementation of auxotrophs is typically accomplished using complex medium components that are rich in the required growth factors, rather than using precisely defined synthetic media [[62]3, [63]6, [64]11]. In such cases, engineered strains harboring auxotrophic marker genes are not subject to selection pressure to maintain the stability of heterologous metabolic pathways. Consequently, it becomes crucial to explore novel selection systems that can be adapted for large-scale industrial fermentations. Glucose, a simple sugar molecule, plays a crucial role as both an energy source and a carbon building block for cellular processes. It is widely recognized as the preferred carbon source for the majority of microorganisms in natural environments and is extensively utilized as the primary carbon source in industrial fermentation processes. Therefore, exploiting glucose deficiency as a selectable marker offers several advantages, including its low cost, widespread availability, robust screening capabilities, and the ability to exert continuous selection pressure during industrial fermentation. However, these possibilities remain to be explored. In this study, we developed a novel glucose-dependent selection system in the high pullulan-producing strain A. melanogenum P16. This system involved generating of a glucose-deficient chassis cell obtained through the knockout of the phosphofructokinase gene (PFK). Additionally, we constructed a series of chromosomal integration plasmids carrying the selection marker PFK, controlled by various promoters. To assess the impact of different promoter strength of PFK on transformation efficiency, as well as on the number and stability of chromosomal integrations of target genes, we employed the green fluorescent protein gene (GFP) as a reporter gene. The results of our study demonstrate the feasibility of the glucose-dependent selection system in Aureobasidium spp., specifically enabling customizable and stable multilocus chromosomal integration of target genes under typical fermentation conditions. This advancement presents a promising new tool for genetic manipulation and strain improvement in non-conventional yeasts. Results and discussion The phosphofructokinase gene can serve as a glucose-deficient related selectable marker Glucose plays a fundamental role in the metabolism of microorganisms and serves as the primary carbon source for industrial fermentation processes. Glycolysis initiates glucose metabolism by breaking down one glucose molecule into two pyruvate molecules while providing energy for cell growth. The phosphofructokinase enzyme (pfk) catalyzes the ATP-dependent phosphorylation of fructose 6-phosphate to form fructose 1,6-bisphosphate and ADP, representing a critical rate-limiting step in glycolysis [[65]22]. Disruption of two copies of the PFK gene in S. cerevisiae abolished growth on glucose [[66]23], indicating that PFK can serve as a viable glucose-deficient related selectable marker. In this study, we targeted the PFK gene in A. melanogenum P16, a remarkable non-conventional yeast known for its excellent pullulan-producing capability [[67]24]. Unlike S. cerevisiae, A. melanogenum P16 possessed a single copy of PFK gene. After disrupting the PFK gene in A. melanogenum P16, the obtained mutant Δpfk-N was further to eliminate nourseothricin resistance gene by transforming a plasmid of pAMCRE-1 with Cre recombinase gene, resulting in the final mutant Δpfk (Table. S1) All the strains were confirmed by genomic PCR (Fig. [68]1a). As shown in Fig. [69]1b, the PFK mutant Δpfk showed minimal growth impairment when utilizing glycerol and lactate as carbon sources compared to the wild-type strain P16. However, Δpfk exhibited complete growth impairment in glucose (Fig. [70]1b), indicating the critical role of PFK in glucose metabolism and its significance as a stringent selectable marker. Growth curves further demonstrated that Δpfk was unable to utilize glucose but exhibited similar growth levels in glycerol and lactate after 36 h, albeit with an initial growth delay, compared to P16 (Fig. [71]1c-d). Furthermore, this growth deficiency on glucose of Δpfk was restored by the complementation of PFK in the strain epfk (Table S1 Fig. [72]1c). These results highlight the complete switch in growth phenotype on glucose in Aureobasidium spp. achieved through PFK knockout and complementation, thereby establishing PFK as a selectable marker associated with glucose deficiency and enabling the development of a novel glucose-dependent selection system. Fig. 1. [73]Fig. 1 [74]Open in a new tab Growth phenotypes and growth curve of A. melanogenum P16, Δpfk (P16 with PFK knocked out), epfk (Δpfk with PFK complemented). a The genotypes verified by genomic PCR revealed the following: the fragment of the PFK knockout cassette was detectable in Δpfk-N using the primes PFK-5F/PFK-3R, the fragment of the PFK knockout cassette without nourseothricin resistance gene was detectable in Δpfk using the primes PFK-5F/PFK-3R, and the fragment of the PFK expressed cassette was detectable in epfk using the primes PFK-F/PFK-R. b The growth phenotype of P16, Δpfk, and epfk on YPD (glucose) and YPGL (glycerol and lactate) media. c Time course of cell growth of P16, Δpfk, and epfk in YPD medium within 33 h. d Time course of cell growth of P16, Δpfk, and epfk in YPGL medium within 60 h. e The survival rate analysis of P16 and Δpfk in YPD medium. Data are given as means ± SD, n = 3 To ensure the maintenance of all engineered strains harboring exogenous genes during fermentation processes, it is important to assess the impact of selectable marker gene loss on cell survival during fermentation. To simulate this scenario, Δpfk cells pre-cultured on glycerol and lactate were transferred to a glucose medium, and their survival rate was analyzed. As shown in Fig. [75]1e, the analysis revealed a rapid decline in the survival rate of Δpfk cells, with over 90% of yeast cells dying within 4 h. In contrast, the survival rate of P16 yeast cells continued to increase. These findings highlight the timely responsiveness and sensitivity of the glucose-dependent selection system, as the loss of the PFK gene results in the rapid death of yeast cells. In addition, it is important to note that in this system, glucose, a readily available and cost-effective carbon source, can be utilized as a reliable and environmental-friendly selectable molecule, providing continuous selectable pressure in industrial fermentation. On the contrary, the high cost of antibiotics, particularly antifungal antibiotic, presents a significant challenge to their application in industrial production. For example, hygromycin B is priced at $104.21 per gram, while the cost of nourseothricin sulfate and bleomycin sulfate are $1528.53 and $1444.5 per gram, respectively. Furthermore, proper treatment of antibiotic fermentation waste is necessary to prevent irreversible ecological damage, incurring a significant economic cost of approximately $300 per ton [[76]25]. The ecological risk of antibiotic resistance genes is another concern in the disposal of fermentation waste. Antibiotic resistance genes have the potential to transfer to humans through the food chain, posing a threat to human health. Reports indicate that the number of deaths caused by antibiotic-resistant infections worldwide is projected to rise from 700,000 in 2014 to a staggering 10 million annually by 2050, with the cumulative cost in terms of reduced healthcare and productivity reaching up to $100 trillion [[77]26]. In contrast, the phosphofructokinase gene (PFK), being a widely distributed gene in organisms, does not pose a risk of gene contamination through horizontal gene transfer, making it a safe and environmentally friendly selectable marker. Based on the above results, the PFK gene was chosen as the selectable marker for glucose-dependent selection system, and the strain Δpfk was used as the chassis cell with glucose deficiency. Screening promoters of varying strength for the construction of the glucose-dependent selection system The strength of selection pressure is influenced by the expression levels of the selectable marker genes [[78]27]. Promoters play a crucial role in regulating the initiation and intensity of transcription, thereby influencing gene expression at various levels [[79]28]. Currently, the exploration of promoters in non-conventional yeasts is limited, mainly focusing on a few species such as Yarrowia lipolytica and Pichia pastoris [[80]29–[81]31]. To explore the specific promoters in Aureobasidium spp. that are adapted to the glucose-dependent selection system in this study, we conducted a temporal transcriptome analysis at five-time intervals during pullulan fermentation of A. melanogenum P16 (Fig. S1). After temporal variations analysis using the Short Time-series Expression Miner (STEM) program, a total of 17,915 genes were classified into 19 significantly different (P < 0.05) gene expression model profiles (Fig. [82]2a). To ensure stable and consistent selectable marker gene expression throughout the fermentation process, 1122 promoters of Profile 1 (Table S3) were selected as the candidate promoter database for further investigation. KEGG enrichment analysis revealed that these genes were enriched in glycometabolism, amino acid metabolism, and lipid metabolism (Fig. [83]2b), further indicating their constitutive expression patterns. From this set of candidates, six genes with distinct expression levels were selected, including genes encoding phosphoglycerate kinase (PGK), pyruvate kinase (PK), general substrate transporter (GST), nascent polypeptide-associated complex α-subunit (NPA) and alcohol dehydrogenase (ADH). The selection of these genes was corroborated by RT-qPCR. As shown in Fig. [84]3c, all tested genes demonstrated relatively stable transcription levels during the fermentation process. Furthermore, we categorized their transcription strength into three groups: PGK exhibited a transcription level approximately 200% of that of the reference gene GAPDH, while the analogs of PK and GST, and NPA and ADH were approximately 80% and 5%, respectively. Finally, we chose a diverse set of promoter modules based on these three groups, referred to P[PGK], P[GST] and P[ADH], to drive the expression of the selectable marker gene PFK in our glucose-dependent selection system. Fig. 2. [85]Fig. 2 [86]Open in a new tab Transcriptome analysis of P16 and RT-qPCR analysis of genes. a Nineteen model profiles recognized by STEM, the number in the upper left-hand corner of the rectangle was the significance level P-value value, and the number in the upper right-hand corner was the number of genes. b KEGG pathway enrichment analysis of 1122 stably expressed genes. The horizontal axis represents the gene ratio, while the vertical axis represents the enriched pathway name. The color scale indicates different thresholds of the P-value, and the size of the dot indicates the number of genes corresponding to each pathway. c The relative transcriptional levels of PGK, GST, ADH, PK, NPA and GAPDH in P16. The values are the mean of three biological replicates, and error bars represent the standard deviations. Data are given as means ± SD, n = 3 Fig. 3. [87]Fig. 3 [88]Open in a new tab a Map of the pP[ADH]/P[GST]/P[PGK]-PFK-rDNA. b The gene copy number analysis of P16:P[TEF]-NAT:GFP and Δpfk containing diverse promoters. c The relative fluorescence intensity and the relative transcriptional level analysis of Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, Δpfk:P[PGK]-PFK:GFP and P16:P[TEF]-NAT:GFP. d Bright-field and fluorescence images of the Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, Δpfk:P[PGK]-PFK:GFP, P16:P[TEF]-NAT:GFP and Δpfk. Different letters in each group represent significant differences, Tukey HSD, P < 0.05. Data are given as means ± SD, n = 3 The glucose-dependent selection system achieves a 100% positive rate of transformation Based on the obtained promoter elements, a glucose-dependent selection system was constructed as shown in Fig. [89]3a. The new selectable marker gene PFK gene driven by the promoter of P[PGK], P[GST], or P[ADH]replaced the nourseothricin selection element (P[PGK]-NAT) of the original expression plasmid pNAT-LoxP-rDNA for Aureobasidium spp. (Fig. S2), resulting in the plasmids of pP[ADH]-PFK-rDNA, pP[GST]-PFK-rDNA, or pP[PGK]-PFK-rDNA, respectively (Fig. [90]3a, Table S1). To assess the functionality of this glucose-dependent selection system, the reporter gene GFP was inserted into the multiple cloning sites (MCS) of these plasmids, resulting in the plasmids of pP[ADH] / pP[GST] / pP[PGK]-PFK-rDNA-GFP (Table S1). Subsequently, these plasmids were linearized using SmaI and then transformed into the chassis cell Δpfk. Transformation efficiency and positive rate were determined by screening the transformants on YPD plates and verifying them through genomic PCR. As a control, the original expression plasmid carrying GFP was transformed into the strain P16 and screened on YPD plates with nourseothricin. As shown in Table [91]1, the positive rate of the transformants carrying GFP can reach 100% when using the glucose-dependent selection system with all three promoters. This result was consistent with the growth phenotype of Δpfk on glucose, suggesting that the glucose-dependent selection system exhibited strong screening pressure when using glucose as sole carbon source, leading to a 100% positive rate. In contrast, the positive rate with the nourseothricin-dependent selection system was only 75% (Table [92]1), which might be attributed to variations in the ability of transformants to tolerate antibiotics. Hence, when utilizing a genetic engineering system that relies on antibiotic resistance, including our original nourseothricin-dependent selection system for Aureobasidium spp., an additional screening round should be conducted with higher antibiotic concentrations [[93]9, [94]14]. In addition, the result in Table [95]1 revealed that the transformation efficiency of the glucose-dependent selection system with the promoters P[PGK], P[GST] and P[ADH] were 18.43, 12.67, and 6.67 positive cfu per 1.0 μg of added DNA fragments, respectively, which were lower than that of the nourseothricin-dependent selection system. This means that the glucose-dependent selection system possesses a significantly higher positivity rate but sacrificing the transformation efficiency. Table 1. Transformation efficiency and positive rate of the glucose-dependent selection system Strains cfu/μg DNA Positive rate % P16: P[TEF]-NAT:GFP 30.78 ± 4.53^a (75 ± 5.5) Δpfk: P[ADH]-PFK:GFP 6.67 ± 1.27^d 100 Δpfk: P[GST]- PFK:GFP 12.67 ± 2.88^c 100 Δpfk: P[PGK]-PFK:GFP 18.43 ± 3.46^b 100 [96]Open in a new tab Data are given as means ± SD, n = 3. Each plate had 10^4 competent cells. Different letters in each group represent significant differences, Tukey HSD, P < 0.05 Achieving high positive rate is an essential and desirable criterion for genetic engineering systems, because it can significantly reduce the burden of subsequent transformant screening and improve overall efficiency. In a previous study, using the CRISPR-Cas9 system for multigene expression in Issatchenkia orientalis, the integration efficiency was only 80% [[97]32]. In another study, utilizing the amino acid nutrient auxotroph for genetic manipulation in Pichia pastoris, the deletion efficiency of individual genes ranged from 4 to 88% [[98]33]. In contrast, the glucose-dependent selection system in this study, which could eliminate false positives through rigorous screening, is a promising tool for genetic manipulation. The strength of the PFK gene promoters can modulate the chromosomal integration numbers of target genes To investigate the relationship between the chromosomal integration number of target genes and the strength of the PFK gene promoters, twenty transformants from each transformation with a specific PFK promoter were selected, and the copy numbers of the GFP gene were analyzed. As shown in Fig. [99]3b, strains containing P[ADH]-PFK exhibited a copy number range of 30–54 copies for the GFP gene, with a median of 39. For strains containing P[GST]- PFK, the copy number ranged from 8 to 20 copies, with a median of 14. In the case of PFK driven by the strongest promoter P[PGK], the copy number only varied between 2 and11 copies, with a median of 6. These results indicate an inverse relationship between promoter strength and chromosomal integration numbers of target genes, while weaker promoter strength correlates with higher copy numbers. In addition, it is noteworthy that when using our original expression system with the nourseothricin selection element (P[PGK]-NAT), the GFP copy number exhibited range of 15–31 copies, which was significantly lower than that of the glucose-dependent selection system with P[ADH]-PFK (Fig. [100]3b). Generally, a high copy number integration is crucial for achieving high-level expression of target genes. In a previous study, a multicopy plasmid integration system based on Leu2 auxotrophic selection was constructed in Pichia pastoris, and the strains containing 20 integrated copies of the vector were successfully isolated [[101]34]. In comparison, the glucose-dependent selection system with P[ADH]-PFK in this study can achieve higher copy numbers (54 copies, Fig. [102]3b), suggesting the potential application of this system for efficient expression of target genes. To further confirm this relationship, each strain with a specific PFK gene promoter, named Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, and Δpfk:P[PGK]-PFK:GFP with GFP chromosomal integration numbers of 39, 14, and 6, respectively (Table. S1), was subjected to analysis of fluorescence intensity, GFP transcriptional level and fluorescence microscopy. The control strain transformed with the original expression plasmid carrying GFP and nourseothricin resistance gene, named P16:P[TEF]-NAT:GFP with a GFP chromosomal integration number of 19, was also included. As shown in Fig. [103]3c and d, the strain Δpfk:P[ADH]-PFK:GFP exhibited the highest fluorescence intensity and GFP transcriptional level, followed by the strains Δpfk:P[GST]-PFK:GFP and Δpfk:P[PGK]-PFK:GFP. This sequential decrease in GFP expression was consistent with the trend of their GFP gene chromosomal integration numbers, further confirming that the strength of the PFK gene promoters were negatively correlated with the chromosomal integration numbers and expression levels of target genes. Therefore, based on the candidate promoter database in this study, the expression level of PFK can be finely tuned by selecting appropriate promoters, thereby realizing the customization of the chromosomal integration number and expression level of target genes in Aureobasidium spp. The chromosomal integration of heterologous biosynthetic pathways is indispensable for constructing robust and high-performing cell factories. To achieve commercially viable production of target compounds, the expression levels of the metabolic genes must be fine-tuned to allow achieve a balanced metabolic flux, ensuring efficient conversion of substrates into the desired product [[104]35]. Therefore, it is important to develop effective synthetic biology tools capable of customizing the gene chromosomal integration number. To date, the methods of customizing gene chromosomal integration number have primarily involved increasing screening pressure or modifying the sequence length of the selectable marker promoter [[105]36, [106]37]. For instance, a modified version of post transformational vector amplification, known as Liquid PTVA, involves gradually increasing the concentration of antibiotics in the liquid medium to obtain strains with different copy numbers ranging from 1 to 18 [[107]36]. However, this method still relies on the use of antibiotics, and overall time cost is 12 days. Another approach involves using the truncated URA3 promoter (P[URA]) to regulate the expression of resistance genes in S. cerevisiae, resulting in the copy number of target gene within the range of 5 to 35 [[108]37]. In this study, through utilizing P[PGK], P[GST] or P[ADH] as the PFK promoter, the glucose-dependent selection system was able to achieve target gene chromosomal integration numbers ranging from 2 to 54, resulting in an 8.3-fold difference in expression levels (Fig. [109]3). Moreover, the candidate database consisting of 1122 promoters will provide additional opportunities to finely tune the expression levels of target genes within the system, facilitating the desired gene copy number range and providing the necessary range of expression levels for synthetic biology applications. The chromosomal integration numbers of target genes remain stable during successive inoculation and fermentation process The rDNA gene, responsible for encoding the RNA components of ribosomes, is highly repetitive with typically 100–1000 copies. These gene are arranged in large stretches of tandem repeats, forming loci that are highly susceptible to copy loss due to intrachromatid homologous recombination between copies [[110]38]. Moreover, since only a subset of rDNA copies is essential for normal cellular function, the silencing of unnecessary rDNA copies can result in DNA damage and copy number variation [[111]38]. Previous studies have reported a significant decrease in the number of target genes integrated in rDNA after 48–72 h of incubation without screening pressure [[112]39]. Therefore, ensuring the stability of the glucose-dependent selection system is an essential requirement for industrial fermentation application. Successive inoculation is a common practice in industrial fermentation processes, such as strain activation and preservation, preparation of seed liquid, and the amplification of multi-stage fermentation. To evaluate the passaging stability of target gene in glucose-dependent selection system, the strains Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP and Δpfk:P[PGK]-PFK:GFP (Table S1) were subjected to five successive inoculations in YPD medium. The strains P16:P[TEF]-NAT:GFP cultured in YPD medium with and without 100 μg/mL nourseothricin were used as controls. As indicated in Fig. [113]4a, the GFP copy numbers of all the strains constructed using glucose-dependent selection system showed excellent stability, which was consist with the observation about the strain P16:P[TEF]-NAT:GFP cultured with nourseothricin. In contrast, the GFP integrated into chromosomes in the strain P16:P[TEF]-NAT:GFP was almost lost after five successive inoculations without nourseothricin (Fig. [114]4a). These results suggest that using the glucose-dependent selection system for strains construction, maintaining stable chromosomal integration numbers of target genes during successive inoculation can be accomplished by simply cultivating the strains on the most commonly used medium containing glucose. Furthermore, this antibiotic-free system will simplify the process of successive inoculation reduce, the costs associated with antibiotics, and avoid complex post-treatment process of wastewater containing antibiotics, particularly in large-scale multi-stage fermentations. Fig. 4. [115]Fig. 4 [116]Open in a new tab Stability of the system during successive inoculation and fermentation process. a The gene copy number analysis of GFP in Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, Δpfk:P[PGK]-PFK:GFP, and P16:P[TEF]-NAT:GFP during five successive inoculations. b The gene copy number analysis of GFP in Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, Δpfk:P[PGK]-PFK:GFP, and P16:P[TEF]-NAT:GFP during fermentation process. c The relative fluorescence intensity analysis of Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, Δpfk:P[PGK]-PFK:GFP, and P16:P[TEF]-NAT:GFP during fermentation. d The PFK relative transcript levels of Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, Δpfk:P[PGK]-PFK:GFP, and P16:P[TEF]-NAT:GFP. Different letters in each group represent significant differences, Tukey HSD, P < 0.05. Data are given as means ± SD, n = 3 In addition to the stability of successive inoculation, the consistent performance of fermentation throughout the fermentation process is a vital criterion for engineered strains utilized in industrial fermentation. As shown in Fig. [117]4b, the copy numbers of the GFP gene remained stable at 40–45 in Δpfk:P[ADH]-PFK:GFP, 14–17 in Δpfk:P[GST]-PFK:GFP, and 6–9 in Δpfk:P[PGK]-PFK:GFP throughout the fermentation process. However, in the case of P16:P[TEF]-NAT:GFP, there was a gradual decrease, leading to complete loss at 120 h. These results were consistent with the relative fluorescence intensity that the relative fluorescence intensities of Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, and Δpfk:P[PGK]-PFK:GFP remained stable throughout the fermentation process, while P16:P[TEF]-NAT:GFP continuously decreased (Fig. [118]4c). These results suggest that the glucose-dependent selection system was capable of maintaining stable copy numbers of target genes during fermentation. This may be attributed to the need for maintaining high transcriptional levels of PFK, as it plays a crucial role in glucose metabolism, as well as pullulan synthesis. As expected in Fig. [119]4d, though driven by different promoters with different strength, the expression of PFK in P16, Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, and Δpfk:P[PGK]-PFK:GFP exhibited comparable transcriptional levels. This highlights the advantage of using key genes involved in glucose metabolism as selection markers, as the significant metabolic flux requirements render strains devoid of copy numbers for these genes less competitive during the fermentation process. Industrial fermentation aims to achieve efficient and cost-effective production processes. In order to fulfill nutritional requirements in fermentation media, complex nitrogen sources like yeast extract, as well as inexpensive crude nitrogen sources such as corn steep liquor, soybean meal, and cottonseed flour, are commonly utilized [[120]40]. However, the various nutritional components of these sources in fermentation media, including nucleic acids, amino acids, vitamins, and other cofactors, do not provide effective selection pressure on the engineered strains obtained through auxotrophic selection, leading a risk of destabilizing of chromosomal integration numbers of target genes during fermentation process. In contrast, based on the system constructed in this study, glucose, one of the most used carbon sources in industrial fermentation, can maintain a constant and rigorous screening pressure during successive inoculation and fermentation process (Fig. [121]4). Therefore, the glucose-dependent selection system is a promising tool to construct engineered strains for industrial fermentation. Glucose-dependent selection system has no significant influence on pullulan fermentation As mentioned above, the glucose-dependent selection system could successfully restore the glucose metabolism in strain Δpfk (Fig. [122]1b and c). However, considering that pullulan synthesis necessitates a significant provision of glucose precursors and energy [[123]41], the pullulan production of Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, and Δpfk:P[PGK]-PFK:GFP still required evaluation. As indicated in Fig. [124]5a, the pullulan titer and cell dry weight of these strains showed no significant differences compared to the wild-type strain P16. Furthermore, the two key genes involved in pullulan synthesis, including AGS2 responsible for pullulan synthesis and UGT1 responsible for UDP-glucose synthesis [[125]41], showed comparable transcriptional levels among three engineered strains and P16 (Fig. [126]5b). These results indicate that the glucose-dependent selection system successfully satisfied the carbon flux required for cell growth and pullulan synthesis, which consistent with the observation that the expression of PFK in engineered strains showed comparable transcriptional levels to that of the wild-type P16 (Fig. [127]4d). For auxotrophic selection system, it is reported that the most amino acid auxotrophs of S. cerevisiae showed reduced cell growth and fatty acid production when supplemented with commonly used concentrations of amino acids, and these observed side effects could be remedied by the application of higher supplement [[128]21]. Nevertheless, employing elevated supplementation levels is economically impractical for industrial fermentation processes. In addition, a leucine-auxotroph Y. lipolytica strain was unable to restore the biomass equivalent to that of a prototrophic strain even after complementing the marker gene [[129]42]. In contrast, the glucose-dependent selection system has no significant influence on growth and product biosynthesis of engineered strains, showcasing potential for application in industrial fermentation, especially in the fermentation polysaccharide, organic acids, and lipids that rely heavily on glucose. Fig. 5. [130]Fig. 5 [131]Open in a new tab a Pullulan yield and biomass of Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, Δpfk:P[PGK]-PFK:GFP, and P16 was determinate after fermentation. b The relative transcript levels of AGS2 and UGT1 in Δpfk:P[ADH]-PFK:GFP, Δpfk:P[GST]-PFK:GFP, Δpfk:P[PGK]-PFK:GFP, and P16. Different letters in each group represent significant differences, Tukey HSD, P < 0.05. Data are given as means ± SD, n = 3 Universality of the PFK gene as a glucose-deficient related selectable marker Based on the above results, the PFK gene, serving as a glucose-deficient related selectable marker show significantly advantages over antibiotics resistance genes and auxotrophic markers, including rigorous and rapid screening with 100% positive rate, stability across successive inoculations and fermentation processes, and no adverse effects on cell growth and metabolism. To provide references for utilizing this gene as a selection marker in other