AmXlnR, a transcription factor involved in xylan degradation and pentose catabolism, enhances pullulan production from xylose in Aureobasidium melanogenum

Abstract

The abundant presence of xylan in low-cost lignocellulosic biomasses establishes its monomeric sugar, xylose, as a pivotal substrate for sustainable bioconversion processes. Although significant advances have been devoted to rewiring the metabolic pathways for xylose assimilation, the effort on regulatory dimensions remains scare. Aureobasidium spp., yeast-forming fungi known for their broad xylanolytic activities and high-yield production of value-added metabolites, represent promising microbial cell factories. However, the regulatory mechanism of xylan degradation and pentose catabolism in these yeasts has not been elucidated. In this study, we identified AmXlnR as the key transcriptional regulator governing xylan degradation and pentose catabolism (d-xylose and l-arabinose) in Aureobasidium melanogenum. Integrated transcriptomic, biochemical, and genetic analyses demonstrate that AmXlnR functions as an activator of xylanase and β-xylosidase genes, while functions as a repressor of α-l-arabinofuranosidase genes, through recognition of specific promoter motifs. Furthermore, AmXlnR positively influences pullulan biosynthesis from pentoses by significantly upregulating the pentose catabolic pathway. Notably, overexpression of AmXLNR alone enhanced pullulan production from xylose, achieving a titer of 56.75 ± 1.51 g/L, which is comparable to the yield obtained with glucose as the carbon source. These findings establish AmXlnR as a key regulator of xylanolytic gene expression in A. melanogenum and provide a new strategic approach to improve bioconversion of lignocellulosic biomass-derived pentose for high-value product synthesis.

Graphical abstract

1. Introduction

Xylan, a predominant hemicellulose component in hardwoods and cereals, features a β-1,4-linked d-xylopyranosyl backbone decorated with diverse branching residues, including l-arabinosyl, 4-O-methyl-glucuronosyl, d-galactosyl, feruloyl, and acetyl groups [1]. This structural complexity necessitates synergistic action of multiple xylanolytic enzymes for complete hydrolysis, targeting both backbone linkages and side-chain substituents [2]. Core enzymes involved in this process include xylanase (EC 3.2.1.8), β-xylosidase (EC 3.2.1.37), α-l-arabinofuranosidase (EC 3.1.1.72), and β-glucuronidase (EC 3.2.1.139) [2]. As the most abundant hemicellulose after cellulose, xylan represents a substantial renewable carbon source and has attracted growing interest for biotechnological utilization [3]. The main monosaccharide released from xylan hydrolysis is xylose, which is the second most abundant sugar in nature after glucose [4]. It serves as a key substrate for the microbial production of biofuels, biochemicals, and other value-added products [5,6]. Although significant advances have been made in optimizing the metabolic pathways for xylose assimilation in industrial microorganisms [4,7,8], the effort on regulatory dimensions remains limited.

To date, the transcriptional regulation of xylanolytic enzymes and the xylose catabolism pathway has been extensively studied in filamentous fungi, particularly Aspergillus and Trichoderma species [9,10]. Expression of genes encoding xylanolytic enzymes is intricately regulated by transcription factors such as AraR, GalR, XlnR, and CreA [[11], [12], [13], [14], [15], [16]]. XlnR, first identified as the master activator of hemicellulose degradation in Aspergillus niger, triggers the expression of xylanolytic genes by binding specifically to upstream promoter regions [11,15]. Beyond regulating extracellular enzymes, XlnR also modulates intracellular xylose catabolism by regulating the pentose catabolic pathway (PCP) and the pentose phosphate pathway (PPP) [17]. Orthologous XlnR regulators have also been studied in other filamentous fungi, including Aspergillus nidulans, Aspergillus oryzae, Trichoderma reesei, Fusarium graminearum, Magnaporthe oryzae, and Neurospora crassa, revealing species-specific variations exist in target genes and DNA-binding motifs [11,[18], [19], [20]]. However, the regulatory role of XlnR in yeasts have been largely overlooked, despite their phylogenetically diverse xylanolytic representatives and ecological prevalence in lignocellulosic-rich environments [21].

Aureobasidium spp., a kind of yeast-forming fungi ubiquitous in diverse habitats, exhibit a remarkable capacity to produce extracellular enzymes that collaboratively degrade plant-derived polymers [22,23]. Unlike most xylan-degrading yeasts with restricted enzymatic profiles, Aureobasidium strains display a broad spectrum of xylanolytic activities, including xylanase, β-xylosidase, α-l-arabinofuranosidase, α-glucuronidase, lytic polysaccharide monooxygenases, ferulic acid esterase [24]. Furthermore, Aureobasidium spp. Possess robust xylose-utilizing capabilities facilitated by multiple catabolic pathways: the isomerase pathway, oxidoreductase pathway, non-oxidative pentose phosphate pathway, and phosphoketolase pathway [8]. Recently, Aureobasidium spp. Have gained recognition as high-yield producers of value-added metabolites including pullulan, β-glucan, siderophore, aureobasidin A, polymalate, organic acids, liamocin, and melanin [[24], [25], [26]]. Therefore, Aureobasidium strains emerge as promising candidates for serving as cell factories in consolidated bioprocessing (CBP) utilizing xylose, xylan and xylan raw materials. However, despite the biotechnological and biological importance, the transcriptional regulatory mechanisms underlying xylan degradation and xylose utilization in Aureobasidium remain poorly understood, hindering metabolic engineering efforts.

We previously isolated a marine yeast, Aureobasidium melanogenum TN2-1-2, which efficiently co-utilizes glucose and xylose for pullulan production [27]. In this study, we elucidated the transcriptional regulatory roles of XlnR in the strain TN2-1-2, termed AmXlnR, focusing on its function in governing xylan degradation and pentose utilization. Furthermore, the effects of overexpressing AmXLNR on the pentose catabolic pathway and pullulan production were investigated. Our findings provide fundamental insights into the function of XlnR in yeasts and offers a new strategic approach for engineering Aureobasidium to effectively convert lignocellulosic biomass-derived pentose into valuable biochemicals.

2. Materials and methods

2.1. Strains, medium, and plasmids

The strains and plasmids used in this study were listed in Table S1 and Table S2, respectively. Escherichia coli DH5 was cultivated at 37 °C in Luria Bertani (LB) medium supplemented with 100 μg/mL ampicillin. A. melanogenum TN2-1-2 and its derivatives were cultured at 28 °C in YPD medium (20.0 g/L glucose, 20.0 g/L peptone, 10.0 g/L yeast extract). When required, the media were supplemented with either 100 μg/mL hygromycin or nourseothricin for the selection of strains carrying the hygromycin phosphotransferase (HPT) or nourseothricin acetyltransferase (NAT) resistance genes, respectively. Growth tests were performed on minimal medium (1.7 g/L Yeast Nitrogen Base without amino acids and 5.0 g/L (NH₄)₂SO₄ appropriately supplemented with 1 g/L d-glucose, d-xylose, l-arabinose, corn cob xylan, sugarcane xylan, or birchwood xylan (Sigma–Aldrich). The extracellular enzymes medium was composed of 5.0 g/L xylose, 5.0 g/L (NH₄)₂SO₄, 2.0 g/L yeast extract, 1.0 g/L KH₂PO₄, 0.5 g/L MgSO₄. The pullulan production medium was composed of 0.2 g/L yeast extract, 0.2 g/L (NH₄)₂SO₄, 7.0 g/L KH₂PO₄, 2.5 g/L Na₂HPO₄·12H₂O, and 1.5 g/L MgSO₄·7H₂O, pH 6.5, supplemented with 110.0 g/L glucose, d-xylose, or l-arabinose [27].

2.2. Cloning and bioinformatics analysis of AmXLNR

The AmXLNR gene was PCR amplified from A. melanogenum TN2-1-2 genomic DNA using the primers XR-F and XR-R (Table S3). XlnR homolog sequences were obtained via BLAST search across Ascomycota and Basidiomyceta. Due to the abundance of sequences available in Pezizomycotina subphylum, a representative subset from each class was selected for analysis. Multiple alignments were performed with Clustal X and visualized using Jalview 2.10.3 (http://www.jalview.org/). A maximum likelihood (ML) phylogenetic tree was constructed in MEGA 11 under the JTT + G substitution model. The putative transcription factor binding motifs in the AmXLNR promoter were identified using the YEASTRACT database [9].

2.3. Inactivation, complementation and overexpression of AmXLNR

Gene knock-in and knock-out in A. melanogenum were performed as previously described [28]. Briefly, to inactivate the AmXLNR gene, a 540 bp 5′ arm and a 420 bp 3′ arm were amplified from the genomic DNA of A. melanogenum TN2-1-2 using the primers 5a-F/5a-R and 3a-F/3a-R (Table S3), respectively. The resulting PCR fragments were subsequently ligated into the PstI/SalI and BamHI/EcoRI sites of the knock-out plasmid pFL4a-NAT-loxp, generating pFL4a-NAT-loxp-ΔAmXLNR. A linear DNA fragment (3′-arm_XlnR-NAT cassette-5′-arm_XlnR) obtained by PstI and EcoRI digestion was transformed into A. melanogenum TN2-1-2 competent cells to disrupt the AmXLNR gene. Subsequently, the plasmid pAMCRE1, harboring the Cre recombinase gene and HPT gene, was transformed into the AmXLNR mutant to excise the NAT gene flanked by loxP sites. The plasmid pAMCRE1 was eventually lost after successive subculturing of the mutant without antibiotic pressure. The resulting nourseothricin and hygromycin sensitive strain, designated TNΔXR, was obtained and used for subsequent studies.

For complementation and overexpression, the AmXLNR gene was amplified from the cDNA of A. melanogenum TN2-1-2 using the primers XR-F2 and XR-R2 (Table S3) and ligated into the SpeI/KpnI site of the expression plasmid pNATX13-NS-loxp, yielding pNATX13-AmXLNR. The linear DNA fragment (rDNA-AmXLNR-NAT cassette-rDNA) obtained by PstI and EcoRI digestion was transformed into TNΔXR (complementation strain AX8) and TN2-1-2 (overexpression strain EX5). Linear DNA fragments for knock-out of the AmXLNR gene and for overexpression of the AmXLNR gene are depicted in Fig. S2. Transformants were selected on nourseothricin, and genomic integration was confirmed by PCR (Fig. S2). The AmXLNR expression levels in TN2-1-2, TNΔXR, AX8, and EX5 were confirmed by qRT-PCR using the primers qAmXLNR-F and qAmXLNR-R (Table S4).

2.4. Activity analysis of xylanolytic enzymes

The yeast strains were cultivated in the enzyme fermentation medium at 28 °C and 180 rpm for 48h. Culture supernatants were obtained by centrifugation (12,000×g, 10 min, 4 °C) and used for activity assays as described [29]. Briefly, for xylanase activity, the reaction system contained 50 μL of the suitably diluted enzyme and 100 μL of 1% (w/v) corn cob xylan in sodium acetate buffer at pH 4.8. After incubation at 50 °C for 30 min, the production of reducing sugars was measured using the DNS method. For activities of β-xylosidase and α-l-arabinofuranosidase activity, the reaction system contained 10 μL of the suitably diluted enzyme, 80 μL of sodium acetate buffer at pH 5.0, and 10 μL 0.01% (w/v) pNP-β-d-xylopyranoside and pNP-α-l-arabinofuranoside (Sigma-Aldrich) respectively. After incubation at 37 °C for 30 min, the release of p-nitrophenol was quantified spectrophotometrically at 420 nm.

2.5. Transcriptome assays

Yeast strains were cultured in enzyme-producing fermentation medium at 28 °C and 180 rpm for 48 h. Cell pellets were harvested by centrifugation, flash-frozen in liquid nitrogen, and stored at −80 °C. RNA was extracted using TRIzol reagent according to the manufacturer's instructions. Libraries were prepared and sequenced on the Illumina NovaSeq platform generating approximately 5 Gb of 150-bp paired-end reads per sample. Raw reads were processed using Trimmomatic v0.36 to remove adapters and low-quality sequences. Clean reads were mapped to the A. melanogenum TN2-1-2 genome using HISAT2. Differential gene expression analysis was performed using DESeq2. Data are available under NCBI BioProject accession PRJNA971762.

2.6. RT-PCR

Total RNA extraction and cDNA synthesis were performed as protocols that had been previously described [30]. qRT-PCR was conducted using ChamQ Universal SYBR qPCR Master Mix (Vazyme, China) on a Rotor-Gene Q Real-time PCR Cycler (QIAGEN Hilden, Germany). Gene-specific primers are listed in Table S4. The actin gene served as the reference. Relative expression levels were calculated using the 2^−ΔΔCt method.

2.7. Promoter analysis using GFP reporter system

The function of putative AmXlnR-binding motifs was tested using an GFP reporter system [31]. The native promoters of AmXLN2, AmBXL2, and AmABF1 genes were PCR amplified from the genomic DNA using the primers pXLN2-F/R, pBXL2-F/R and pABF1-F/R (Table S3), respectively. The obtained promoter fragments were ligated into a GFP expression plasmid of pNATX13-GFP using ClonExpress II One Step Cloning Kit C112-02 (Vazyme), yielding pNATX13-GFP-AmXLN2, pNATX13-GFP-AmBXL2, and pNATX13-GFP-AmABF1. Subsequently, the variant promoters of P_XLN2Δ (deletion of GGCTGA site), P_BXL2Δ (deletion of GGTTAA) and P_ABF1Δ (deletion of GGCTAT) genes were constructed using a Site-Directed Mutagenesis Kit (NEB, USA) using the primers BXL2WD-F/R, BXLWD2-F/R and ABF1WD-F/R (Table S3) and cloned similarly, yielding pNATX13-GFP-ΔAmXLN2, pNATX13-GFP-ΔAmBXL2, and pNATX13-GFP-ΔAmABF1. Finally, the obtained expression plasmids carrying GFP driven by the native promoters and their binding sites variants were linearized by PstI and EcoRI digestion and transformed into the competent cells of A. melanogenum TN2-1-2 or TNΔXR. GFP fluorescence intensity was visualized using an Olympus U-LH100HG fluorescent microscope and quantified using a BioTeK Synergy H1 Hybrid Reader (BioTek Instruments Inc., USA) (485 nm excitation and 520 nm emission). The relative transcription levels of EGFP were determined by qRT-PCR using the primers qGFPS and qGFPA (Table S4).

2.8. AmXlnR-binding motif analysis

The 700 bp upstream regions of differentially expressed xylanolytic genes (identified via transcriptome profiling) were analyzed for conserved motifs using the MEME suite (Multiple EM for Motif Elicitation-https://meme-suite.org). This tool integrates multiple algorithms to identify conserved TF binding motifs and validate their biological relevance [32].

2.9. Pullulan production and quantification

The purification and quantification of pullulan were conducted according to the previously described method [33]. Briefly, yeast strains were pre-cultured in 5 mL YPD tubes at 28 °C for 36 h. Subsequently, 5.0 mL of seed culture was inoculated into 50 mL of pullulan production medium in 250-mL flasks. Flasks were incubated at 28 °C with agitation at 180 rpm for 5 days. Cultures were centrifuged at 14,000×g and 4 °C for 10 min. Pullulan in the supernatant was precipitated by adding three volumes of ice-cold ethanol, washed twice with 70% ethanol, lyophilized and weighed.

2.10. Statistical analysis

All experiments were performed in triplicate. Data are presented as mean ± standard deviation (SD). Statistical significance was assessed using Student's t-test in GraphPad Prism 8.01. Relative growth rate (k) was determined by nonlinear regression of exponential growth using the model Y Y₀ × exp(k × x).

3. Results and discussion

3.1. AmXlnR is involved in pentose utilization and xylanolytic gene regulation

An XlnR ortholog-encoding gene, AmXLNR (accession number ON505002.1) was identified in A. melanogenum TN2-1-2 according to genome annotation. Sequence analysis revealed that AmXLNR contains a 3096-bp open reading frame interrupted by two introns (77 bp and 49 bp), encoding a 989-amino acid protein (Fig. S1). For AmXlnR, sequence alignment indicates its DB domain characterized by a GAL4 transcription factor superfamily-specific motif, Cys-X₆-Cys-X₅-Cys-X₂-Cys-X₆-Cys, located at positions 94–114 amino acids (Fig. S1). Two conserved sequences, 5′-SYGGRG-3′ at −196 bp and −173 bp, were identified as putative binding sites for the glucose repressor CreA [34], suggesting potential glucose-mediated repression of AmXLNR.

To investigate the role of AmXlnR, the AmXLNR gene was disrupted in A. melanogenum TN2-1-2, yielding the mutant TNΔXR (Fig. S2). The complementation strain AX8 and the overexpression strain EX5 were then constructed by expressing additional AmXLNR driven by a constitutive TEF promoter in TNΔXR and TN2-1-2, respectively (Fig. S2). qRT-PCR analysis confirmed the expected AmXLNR expression levels in TN2-1-2, TNΔXR, AX8, and EX5 (Fig. S3). Subsequently, these strains were grown on minimal medium supplemented with d-glucose, d-xylose, l-arabinose, or xylan. With glucose as the carbon source, all strains exhibited similar growth profiles (Fig. 1a). However, when grown on d-xylose or l-arabinose, TNΔXR exhibited significantly delayed growth and reduced relative growth rates (k) compared to the wild-type strain (Fig. 1b and c). This growth defect on pentoses was rescued in the complementation strain AX8. Furthermore, the overexpression strain EX5 demonstrated enhanced growth rates on these sugars relative to TN2-1-2, suggesting the critical role of AmXlnR in pentose utilization in A. melanogenum. Notably, TNΔXR still retained the ability to grow on d-xylose and l-arabinose, albeit at a reduced rate, suggesting the potential existence of additional transcription activators supporting pentose utilization in this yeast.

Fig. 1.

Growth phenotypes and xylanolytic enzyme activity of the wild-type strain TN2-1-2 and its derivatives including TNΔXR (TN2-1-2 with AmXLNR knocked out), AX8 (TNΔXR complemented with AmXLNR), and EX5 (TN2-1-2 with AmXLNR overexpression). (a–c) Time courses of cell growth of TN2-1-2 and its derivatives on minimal medium supplemented with 1 g/L d-glucose, 1 g/L d-xylose, or 1 g/L l-arabinose, respectively. (d) The growth phenotype of TN2-1-2 and TNΔXR on xylan derived from corn cob, sugarcane, and birch. (e–g) Cell biomass and xylanolytic enzyme activity of TN2-1-2 and its derivatives after culture on enzyme fermentation medium for 48 h. Three independent replicates were performed for the statistical analysis. ∗P < 0.05,∗∗P < 0.01.

The xylanolytic capability of the strains was determined by detecting xylan hydrolysis halos on plates containing corn cob, sugarcane, or birch xylan, along with quantitative assays of extracellular xylanolytic activity. As shown in Fig. 1d and e, TNΔXR showed a significantly reduced ability to hydrolyze xylan, consistent with its significantly lower xylanase activity compared to the wild-type strain. Furthermore, the expression of AmXLNR in TNΔXR and TN2-1-2 led to restored and even stimulated xylanase activity in AX8 and EX5, respectively, as expected. A similar trend was observed for β-xylosidase activity (Fig. 1f), supporting the positively regulatory role AmXlnR plays in xylanase and β-xylosidase gene expression. In contrast, α-l-arabinofuranosidase activity increased 4.6-fold in the TNΔXR mutant relative to the wild type (Fig. 1g), indicating that AmXlnR acts as a negative regulator of genes involved in cleaving xylan side chains. This may explain why TNΔXR, despite lacking clear xylan hydrolysis halos, still exhibited growth on xylan, possibly through the utilization of arabinose from xylan side chain.

3.2. AmXlnR is a bidirectional-function transcriptional regulator for xylanolytic gene expression

To further investigate the regulatory role of AmXlnR, we performed comparative transcriptomic profiling of the TNΔXR mutant and the wild-type strain TN2-1-2 after culturing on xylan for 48 h (Transcriptome data accession number: PRJNA971762). Notably, the aligned reads from transcriptome sequencing confirmed the presence of two introns in AmXlnR in the wild-type and the deletion of AmXlnR in TNΔXR (Fig. 2a). As shown in Fig. 2b, compared to the wild-type strain, eighteen of twenty-three xylanase genes distributed across GH5, GH10, GH11, GH30, and GH51 families, and nine of twelve GH43 family β-xylosidase genes showed reduced expression in TNΔXR. In terms of α-l-arabinofuranosidase genes distributed across the GH43, GH3, and GH62 families, the expression of five out of seven increased in TNΔXR. qRT-PCR analysis was further conducted to confirm the results (Fig. 2c). Deletion of AmXLNR led to significantly reduced transcription of two xylanase genes AmXLN1 (A6146) and AmXLN2 (A2296) to 52.74% and 10.16%, respectively. AmXLNR complementation or overexpression could restore or enhance the gene expression. A similar regulatory pattern was observed for β-xylosidase genes AmBXL1 (A0566) and AmBXL2 (A3053). In contrast, two α-l-arabinofuranosidase genes AmABF1 (A0995) and AmABF2 (A3053) exhibited increased transcriptional levels (345.66% and 177.9%, respectively) in TNΔXR, which were restored upon AmXLNR complementation.

Fig. 2.

Transcriptomic and qRT-PCR analysis of xylanolytic genes in the wild-type strain TN2-1-2 and the AmXLNR mutant TNΔXR. (a) The aligned reads of AmXLNR from transcriptome sequencing of TN2-1-2 and TNΔXR. (b) Phylogenetic analysis and family distribution of xylan-degrading genes, along with their transcriptional profiles in TN2-1-2 and TNΔXR. The heatmap shows the log₂-scaled gene transcriptional levels normalized to those in TN2-1-2. Blue stars mark genes encoding xylanases and β-xylosidases that exhibited the most significant downregulation, while red stars highlights the most significantly upregulated α-l-arabinofuranosidase-encoding gene. (c) The relative transcriptional levels of xylanase genes (AmXLN1 and AmXLN2), β-xylosidase genes (AmBXL1 and AmBXL2) and α-l-arabinofuranosidase genes (AmABF1 and AmABF2) of TN2-1-2 and its derivatives. Three independent replicates were performed for the statistical analysis. ∗P < 0.05,∗∗P < 0.01.

Together with the enzymatic activity data shown in Fig. 1e–g, our results demonstrated that AmXlnR serves as a bidirectional-function regulator in A. melanogenum, activating transcription of xylanase and β-xylosidase genes while repressing α-l-arabinofuranosidase genes. XlnR homologs are known to perform conserved activating functions in regulating xylanase and β-xylosidase genes across numerous filamentous fungi [19,35,36], which is consistent with our observation. In contrast, their regulatory effects on α-l-arabinofuranosidase genes exhibit considerable species-specific diversity. For instance, XlnR acts as an activator in A. nidulans, M. oryzae, and T. reesei, but as a repressor in A. niger [19,36], a pattern that aligns with our findings in A. melanogenum (Fig. 1g). AraR, an l-arabinose-responsive regulator, recognizes promoter sequences highly similar to those of XlnR [37]. In A. niger, the deletion of XlnR leads to the upregulation of AraR-target genes, including those encoding α-L-arabinofuranosidases, due to the antagonistic effect between the two regulators [38,39]. Similarly, a homologous protein of AmAraR was also identified in A. melanogenum, implying the existence of a comparable two-regulator system. Transcriptomic data revealed that the expression level of AmARAR in TNΔXR remains at 82.3% of that in the wild-type strain (Fig. S4), suggesting that AmAraR may contribute to the elevated expression of α-l-arabinofuranosidase genes upon the deletion of the repressor AmXlnR.

In addition, AmXLN2 (accession number: OP839390), AmBXL2 (accession number: OP801682), and AmABF1 (accession number: OP839391) were identified as the genes exhibiting the most significant expression changes among the xylanase, β-xylosidase, and α-l-arabinofuranosidase encoding genes, respectively, in response to AmXlnR regulation in A. melanogenum (Fig. 2b–c). These genes were selected for further analysis to explore the regulatory mechanism of AmXlnR.

3.3. AmXlnR exerts its bidirectional regulatory function through specific promoter binding sites

XlnR regulates the expression of xylanolytic genes through binding specific promoter motifs, such as 5′-GGNTAA-3′ in A. niger, 5′-GGC(A/T)₃-3′ in T. reesei, and 5′-GGCT(A/G)A-3′ in A. oryzae [10,32]. Accordingly, putative XlnR-binding motifs of 5′-GGCTGA-3′, 5′-GGTTAA-3′, and 5′-GGCTAT-3′ were identified in the promoters of AmXLN2, AmBXL2, and AmABF1 genes, respectively, in A. melanogenum TN2-1-2 (Fig. 3a). Green fluorescent protein (GFP) reporter constructs driven by either the intact promoters (P_XLN2, P_BXL2, and P_ABF1) or their motif-deletion variants (P_XLN2Δ, P_BXL2Δ, and P_ABF1Δ) were expressed in both wild-type (TN2-1-2) and AmXLNR-deletion (TNΔXR) strains (Fig. 3a and Table S2). As indicated in Fig. 3, the TN2-1-2 transformants carrying P_XLN2-GFP and P_BXL2-GFP exhibited strong intracellular green fluorescence, while those carrying motif-deletion variants (TN2-1-2:P_XLN2Δ and TN2-1-2:P_BXL2Δ) showed no fluorescence. Furthermore, TNΔXR transformants with both intact and variant promoters (TNΔXR:P_XLN2 and TNΔXR:P_BXL2) showed slight fluorescence (Fig. 3b–3d) and GFP expression (Fig. 3c). These results indicate that the presence of upstream 5′-GGCTGA-3′ and 5′-GGTTAA-3′ motifs is indispensable for gene activation by AmXlnR.

Fig. 3.

In vivo GFP expression analysis control of the native and mutated promoters of AmXLN2, AmBXL2 and AmABF1. (a) Schematic representation of the native promoters and their variants with deleted putative XlnR-binding sites: P_XLN2Δ:P_XLN2 with 5′-GGCTGA-3′ deleted; P_BXL2Δ:P_BXL2 with 5′-GGTTAA-3′ deleted; P_ABF1Δ:P_ABF1 with 5′-GGCTAT-3′ deleted. (b) Relative fluorescence intensity, (c) relative GFP transcriptional level, and (d) bright-field and corresponding fluorescence images of the reporter strains. Three independent replicates were performed for the statistical analysis. ∗P < 0.05,∗∗P < 0.01, ns: no significance.

Compared to the TN2-1-2 transformants harboring P_ABF1-GFP, TN2-1-2:P_ABF1Δ showed significantly higher relative fluorescence intensity (Fig. 3b–3d) and GFP transcription (Fig. 3c). Furthermore, both TNΔXR:P_ABF1 and TNΔXR:P_ABF1Δ exhibited significantly higher relative fluorescence intensity and GFP expression level compared to TN2-1-2:P_ABF1, regardless of ABF1 promoter integrity. These results indicated that AmXlnR represses AmABF1 expression by binding the 5′-GGCTAT-3′ motif. Notably, no significant difference in fluorescence intensity or GFP expression was observed between TN2-1-2:P_XLN2Δ and TNΔXR:P_XLN2, TN2-1-2:P_BXL2Δ and TNΔXR:P_BXL2, and TN2-1-2:P_ABF1Δ and TNΔXR:P_ABF1. This further confirm that the identified binding motifs (5′-GGCTGA-3′ for AmXLN2, 5′-GGTTAA-3′ for AmBXL2, and 5′-GGCTAT-3′ for AmABF1) are essential for AmXlnR binding and its activation and repression function on target promoters.

To further investigate the conserved DNA motifs recognized by AmXlnR in A. melanogenum, we performed Multiple EM for Motif Elicitation (MEME) analysis on the promoter regions of xylanolytic genes exhibiting differential expression in the wild-type strain versus the TNΔXR mutant (Table S5). The analysis revealed a significant correlation between conserved DNA motifs and gene expression patterns (Fig. S5). Specifically, the motif associated with transcriptional activation was defined as 5′-GGNT(A/t/g)A-3′ (Fig. S5a), while the repression motif was identified as 5′-GGC(T/a)AT-3′ (Fig. S5b). Accordingly, a putative AmXlnR activating sequence of 5′-GGCTAA-3′ was identified at position −412 bp within the AmXLNR promoter, suggesting a potential autoregulatory mode. Furthermore, analysis of promoter sequences for xylanolytic genes with distinctive expression patterns in the transcriptome data, such as A6695 (encoding a xylanse) and A1899 (encoding an α-l-arabinofuranosidase) (Fig. 2b), revealed the presence of a repressive sequence (5′-GGCAAT-3′) in A6695 and an activating sequence (5′-GGCTAA-3′) in A1899 (Table S5), consistent with the respective motifs defined above. These findings suggest that the bidirectional regulatory function of AmXlnR is likely mediated by position-specific nucleotide conservation within its core binding motif.

The discovery and characterization of such bidirectional regulatory elements are critical to expanding the operational diversity and metabolic flexibility of industrial bio-production strains [40]. Various methods exist for realizing bidirectional transcriptional regulation, such as integrating artificial upstream activating and repressing sequences through high-throughput screening and rational modification of endogenous promoter motifs [41], and fusing the activation and repression domains of CRISPRa and CRISPRi to two orthologous nuclease-deficient Cas proteins [42]. Herein, we confirmed AmXlnR as a natural and robust bidirectional transcriptional regulator, which has considerable potential in synthetic biology and xylose-based biomanufacturing applications.

3.4. AmXlnR enhances pullulan production from pentoses though positively regulation the pentose catabolic pathway in A. melanogenum

In addition to its role in regulating xylanolytic genes, AmXlnR is involved in pentose utilization in A. melanogenum (Fig. 2a–c). Given that the wild-type strain TN2-1-2 can produce pullulan from xylose [27], we further evaluated the pullulan titer of TN2-1-2 and its AmXlnR-deleting or overexpressing derivatives (TNΔXR, AX8, EX5), using d-glucose, d-xylose, or l-arabinose as the sole carbon source. As indicated in Fig. 4a, the pullulan titers of TNΔXR, AX8, and EX5 were comparable to that of TN2-1-2 (∼56.80 g/L) when grown on d-glucose. In contrast, when d-xylose was used as the carbon source, TNΔXR produced 34.19 ± 1.57 g/L pullulan, which was 28.39% lower than the wild-type (Fig. 4b). Similarly, the pullulan titer on l-arabinose was 30.20 ± 1.98 g/L for TNΔXR, representing a 26.04% reduction compared to the wild-type strain (Fig. 4c). Consistent with the reduced product synthesis, TNΔXR also exhibited lower biomass accumulation on both d-xylose (16.46 ± 0.56 g/L vs. 20.43 ± 1.28 g/L for wild-type) and l-arabinose (16.98 ± 1.06 g/L vs. 19.15 ± 0.64 g/L for wild-type). Genetic complementing AmXLNR in AX8 restored pullulan production and biomass accumulation to similar levels of the wild-type strain. Furthermore, overexpression of AmXLNR in EX5 led to a significant enhancement in pullulan production from pentose, achieving titers of 56.75 ± 1.51 g/L from d-xylose and 47.81 ± 0.83 g/L from l-arabinose, with yields of 0.52 g/g and 0.43 g/g, respectively. Notably, the pullulan titer of EX5 achieved from d-xylose was comparable to that with glucose as the carbon source. These results suggest that AmXlnR plays an important role in pullulan production from xylose and overexpressing AmXLNR alone is sufficient to effectively boost pullulan biosynthesis from xylose in A. melanogenum.

Fig. 4.

Impact of AmXLNR deletion and overexpression on pullulan production and pentose catabolism. (a–c) Pullulan production and biomass of the wild-type strain TN2-1-2, AmXLNR mutant TNΔXR, and overexpression strain EX5 on d-glucose, d-xylose, and l-arabinose. (d) Heatmap showing the relative expression (log₂ fold-change) of key pentose catabolic pathway genes in TNΔXR and EX5 compared to those of TN2-1-2 when cultured on d-xylose and l-arabinose. Three independent replicates were performed for the statistical analysis. ∗P < 0.05,∗∗P < 0.01.

To elucidate the mechanism underlying the enhanced pullulan production, we examined the expression of key genes involved in the pentose catabolic pathway (PCP). When grown on either d-xylose or l-arabinose, deletion of AmXLNR significantly downregulated the transcription of genes encoding l-arabinose reductase (LAR), d-xylose reductase (XYR), xylitol dehydrogenase (XDH), d-xylulokinase (XKI), and d-xylose isomerase (XI) (Fig. 4d, Tables S6 and S7). Conversely, overexpression of AmXLNR in EX5 led to a substantial upregulation of these core PCP genes compared to the wild-type strain, showing induction levels ranging from 3.5-fold to 13.8-fold on d-xylose and 4.8-fold to 12.6-fold on l-arabinose (Fig. 4d, Tables S6 and S7). This activation is consistent with the presence of our previously identified activating AmXlnR-binding motif 5′-GGNT(A/t/g)A-3′ in the promoters of these genes (Fig. S5a). However, it is significant to note that, under pentose growth conditions, the transcriptional levels of genes encoding l-arabinitol 4-dehydrogenase (LAD) and l-xylulose reductase (LXR) remained unaffected by either the deletion or overexpression of AmXLNR under growth conditions with either pentose. This lack of regulation could potentially be attributed to the absence of AmXlnR-binding motif in the promoters of the LAD and LXR genes. These results confirmed that AmXlnR plays a key role in the activation of core PCP genes in A. melanogenum, therefore its overexpression contribution to the enhanced pullulan production from pentose.

The widespread presence of xylan in low-cost lignocellulosic biomasses, such as corn cobs, straw, and hardwoods, makes its monomeric sugar, xylose, a crucial target for bioconversion to support sustainable development and improve process economics [43]. However, efficient xylose utilization in fungi is hindered by several critical bottlenecks, including slow xylose uptake, significant carbon catabolite repression by glucose, cofactor imbalances in xylose reductase (XYR)/xylitol dehydrogenase (XDH) pathways [6,44]. Current strategies for enhancing xylose utilization efficiency primarily rely on labor-intensive metabolic engineering, such as sequential overexpression of the xylose isomerase (XI), xylulokinase (XKI) and transketolase (TKL) genes in A. pullulans TKL-4 [45]; the combined overexpression of the XI, xylulose kinase gene (XKS1), and xylose transporter genes (HXT), along with the knockout of competing pathway genes (GRE3, YPR1, ADH6, GPD1) in Saccharomyces cerevisiae FJ13 [7]; and co-expression of XYR/XDH and XK in Y. lipolytica XYL+ [46]. In this study, the sole overexpression of the transcription factor gene AmXLNR could globally upregulate the expression of genes involved in the pentose catabolic pathway, including XYR, XDH, XKI, LAR, and XI (Fig. 4d, Tables S6 and S7), resulting in a pullulan titer of 56.75 g/L from xylose, which was comparable to that obtained from glucose (Fig. 4b). In contrast, Guo et al. reported a pullulan titer of 13.72 g/L from xylose in Aureobasidium pullulans P4 by expression the XYR and XDH genes [47]. In another study, the engineered strain A. melanogenum XI-15, with a deleted XYR gene and overexpressed xylose isomerase gene (XI), accumulated 53.74% (w/w) lipids from xylose, which was still lower than the 59.4% lipid content observed in the wild-type strain grown on glucose [8]. Therefore, the overexpression of AmXLNR in this study presents a simple yet highly effective strategy for boosting bioconversion from xylose.

3.5. Universality of XlnR as a xylanolytic transcription factor

A previous study has indicated that XlnR orthologs are present in almost all filamentous Ascomycota [48]. In this study, we characterized the regulatory function of XlnR in Aureobasidium, a yeast-forming fungi group previously unexplored in this context. Thus, a comprehensive phylogenetic analysis of XlnR orthologs associated with conservative domain analysis was performed based on BLAST search using the NCBI database. As indicated in Fig. 5 and Fig. S6, XlnR homologs, harbor a zinc cluster Zn(II)₂Cys₆ DNA-binding domain (DB) and a fungal-specific TF domain (TF-MHR), and are widely distributed within Ascomycota, particularly in the Pezizomycotina (across seven classes) and Taphrinomycotina subphyla, but absent in Saccharomycotina subphylum. Moreover, no XlnR homologs were found in Basidiomyceta. This absence may be attributed to genome reduction and niche specialization in Saccharomycotina yeasts during evolution [49]. A similar genome simplification is also observed in Basidiomyceta, such as Ustilaginomycotina subphylum [49]. In contrast, Pezizomycotina fungi, which have not undergone significant genome reduction, have developed complex and efficient machinery for plant biomass degradation [48,49]. These findings highlight that Aureobasidium a model yeast-forming fungi for investigating XlnR-mediated regulation.

Fig. 5.

Phylogenetic tree of XlnR homologs from Ascomycota associated with conservative domain analysis. DB: Zinc cluster Zn(II)₂Cys₆ DNA-binding domain. TF-MHR: fungal-specific TF domain.

In addition, based on the findings of this study, overexpression of AmXLNR represents an effective strategy for enhancing product synthesis from xylan-derived monosaccharides. This approach also holds broader applicability for other industrial strains possessing XlnR homologs from Pezizomycotina and Taphrinomycotina (Fig. 5), such as T. reesei and A. niger. Our work thus offers a valuable reference for optimizing the performance of xylose-utilizing industrial microorganisms, facilitating their application in lignocellulosic biomass conversion and related biotechnological processes.

4. Conclusion

Aureobasidium spp. are promising microbial cell factories, characterized by their broad xylanolytic activities, efficient xylose utilization, and capacity for producing valuable metabolites. In this study, AmXlnR was identified as the key transcriptional regulator controlling both xylan degradation and pentose catabolism in A. melanogenum. Notably, AmXlnR exhibits a dual regulatory functionality, capable of simultaneously activating and repressing different xylanolytic genes by recognizing distinct promoter motifs. Furthermore, overexpression of AmXLNR alone was sufficient to significantly enhance the pullulan titer from xylose to 56.75 ± 1.51 g/L, a titer comparable to that from glucose, primarily by upregulating the pentose catabolic pathway. These findings provide valuable insights into the function of XlnR in yeasts and offer a new and feasible regulatory strategy to construct engineered strains for the efficient conversion of lignocellulosic biomass-derived pentose into valuable biochemicals.

CRediT authorship contribution statement

Wen-Hui Zhao: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation. Ya-Jun Liu: Writing – original draft, Investigation, Funding acquisition, Data curation. Lu-Lu Zhang: Validation, Data curation, Conceptualization. Ayaka Uke: Writing – review & editing. Akihiko Kosugi: Writing – review & editing, Supervision. Zhe Chi: Writing – review & editing, Funding acquisition. Zhen-Ming Chi: Writing – review & editing, Supervision. Guang-Lei Liu: Writing – review & editing, Visualization, Supervision, Resources, Funding acquisition, Conceptualization.

Declaration of competing interest

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

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Data availability

Data will be made available on request.