Genetic modifications of critical regulators provide new insights into regulation modes of raw-starch-digesting enzyme expression in Penicillium

Background Starch is a very abundant and renewable carbohydrate and an important feedstock for industrial applications. However, most starch-based products are not cost-efficient due to the high energy input needed in traditional enzymatic starch conversion processes. Raw-starch-digesting enzymes (RSDEs) from filamentous fungi have great commercial value in starch processing. However, the regulatory mechanisms associated with their production in filamentous fungi remain unknown. Results In this study, we reported the novel finding that cellulolytic fungus Penicillium oxalicum 114-2 has broad RSDE activity. Four regulators, including the amylase transcription activator AmyR, the catabolite repression repressor CreA, the group III G protein α subunit PGA3, and the nonhistone chromosomal protein HepA, have been found to play a crucial regulatory role in RSDE expression. Enzymatic assays revealed that RSDE production significantly increased after the overexpression of AmyR and HepA, the deletion of CreA and the dominant activation of PGA3. RT-qPCR analysis demonstrated that there is a mutual regulation mode between the four regulators, and then formed a cascade regulation mechanism that is involved in RSDE expression. Comparative transcriptomic analysis between the wild-type strain and genetically engineered strains revealed differentially expressed genes that may mediate the RSDE expression. Conclusions The four different types of regulators were systematically investigated and found to form a regulatory network controlling RSDE gene expression. Our results provide a new insight into the regulatory mechanism of fungal amylolytic enzyme expression and offer a theoretical basis to rationally improve the RSDE yield in the future. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-022-02162-6.

starch structure in starch-based industrial processes. Briefly, raw starch was sequentially gelatinized at approximately 100 °C, liquefied with thermophilic α-amylases at approximately 95 °C, and treated with glucoamylases at 50-60 °C [5,6]. Therefore, the conventional conversion of starch to glucose is energy intensive and time-consuming, thus increasing the production cost of starch-based products. In particular, the large amount of energy consumed in starch processing is also inconsistent with the current global strategy of "carbon peaking" and "carbon neutrality".
To reduce the energy costs associated with traditional starch biorefineries, raw-starch-digesting enzymes (RSDEs) could directly degrade raw starch granules into oligosaccharides or glucose below the gelatinization temperature of starch [7]. It has potential applications in food processing and biofuel preparation [8]. The RSDEs derived from Aspergillus sp. and Rhizopus sp. and Bacillus sp. are the most widely used and possess the following advantages: low production cost, good consistency, small time and space required for production, and easy process modification and optimization [9]. Although RSDEs are mainly secreted by filamentous fungi in the natural environment, they have low yields, which have not yet met the quantitative and cost requirements for large-scale industrialization of raw starch biorefining. The optimization strategy regarding the modification of RSDEs is mainly focused on two aspects. One approach to reduce the cost of raw starch biorefining is mining and identification of novel RSDEs that hydrolyse high concentration raw starches with high efficiency at a lower temperature. Recently, different α-amylases with high specific raw starch hydrolysing activity have been identified and characterized from bacteria and fungi [5,[10][11][12][13]. Wei Fang et al. identified a novel raw starch-digesting α-amylase AmyZ1 derived from marine bacterium Pontibacillus sp. ZY, and the AmyZ1 could efficiently hydrolyse raw starches at a relatively low temperature [14]. Additionally, a novel raw starch-digesting glucoamylase, PoGA15A, from P. oxalicum, was heterologously expressed by P. pastoris, displaying a high capacity for raw starch degradation and remarkable stability [6]. On the other hand, the improvement of RSDE production by using biotechnology approaches is a feasible strategy. Compared to those traditional strategies, such as random mutagenesis and fermentation optimization in the pregenomic era [15], rational genetic engineering transformation was predominantly employed to improve the yield of RSDEs. Generally, the expression of fungal RSDE genes is strictly controlled by upstream regulators, particularly transcription factors (TFs) at the transcriptional level. Some TFs have been proven to be significantly related to amylolytic enzyme production, such as Zn(II)2Cys6 zinc finger protein AmyR in Aspergillus species [16,17]. Recently, a comparative transcriptomics study on P. oxalicum HP7-1 in the presence of glucose and starch revealed five transcription factors exhibiting significant regulation in the production of RSDEs. Among them, a transcription factor containing a pair of SANT/Myb domains controls the expression of an important RSDE gene, POX01356/ PoxGA15A, in P. oxalicum [18]. In addition to these regulators, the HMG-box protein PoxHmbB and the extracellular protease activator PrtT, were identified as RSDE repressors in P. oxalicum [19,20]. However, their research is insufficient to elucidate the regulatory mechanism behind the RSDE gene expression. Therefore, to further reduce the cost of enzyme production, it is rational to breed fungal strains based on an understanding of the molecular mechanisms controlling RSDE gene expression.
Apparently, P. oxalicum exhibits good prospects for raw amylase production and application in recent years [6,15,18]. The filamentous fungus P. oxalicum 114-2 has received substantial attention due to its highly efficient apparatus for cellulolytic enzyme synthesis. It has been under investigation for more than 30 years in China [21]. According to the genome annotation and analysis, five amylolytic enzyme genes were found to encode Amy15A (PDE_09417), Amy15B (PDE_05527), Amy13A (PDE_01201), Amy13B (PDE_01021), and Agl31A (PDE_03966) in P. oxalicum. However, secretome analysis performed through LC-MS/MS showed that Amy15A and Amy13A, which occupied 28.85% and 10.95% of the total protein, respectively, were the two major amylolytic enzymes in the wild-type strain [21,22]. Numerous reported studies have focused on the regulator network of cellulase expression in P. oxalicum114-2 over the years [23]. Regulators, such as the classic TFs AmyR and CreA, nonclassical TFs, including a casein kinase CK2, proteins heterotrimeric G protein, and nonhistone chromosomal protein, have been identified to be involved in cellulase expression pathways [24][25][26]. Although there is indirect evidence that the cellulase expression regulators may mediate amylolytic enzyme expression, their function in RSDE expression has not been systematically elaborated. In this study, we found that amylases secreted by the WT 114-2 strain on starch medium display efficient enzymatic activity against different raw starches. Moreover, four engineered strains were constructed by genetic modification of the amylase transcription activator AmyR, the catabolite repression repressor CreA, the group III G protein α subunit PGA3 and the nonhistone chromosomal protein HepA in P. oxalicum. By employing RT-qPCR and RNA-seq technologies, the regulatory network was revealed and provided deep insights into the RSDE expression mode of P. oxalicum. Additionally, the suggested mechanisms behind regulators of RSDE expression might be general properties in amylolytic fungi and broadly enable engineering strategies for the protein hyperproducers.

Raw starch-digesting enzyme activity and substrate specificity
To analyse the raw starch-digesting abilities of extracellular protein from WT 114-2, the strain was cultivated in Vogel's minimal medium with starch as a sole carbon source and amylolytic enzyme expression inducer. Soluble starch and different raw starches derived from rice, corn, potato, buckwheat, and cassava were used as reaction substrates. The extracellular amylases showed different raw starch-digesting enzyme activities. Compared with soluble starch as a substrate, the extracellular enzymes exhibited higher RSDE activity towards raw rice starch, but lower RSDE activity towards the other four raw starches (Fig. 1). RSDEs secreted from the WT 114-2 strain preferentially hydrolysed raw rice starch, and the result is comparable to the single RSDE characteristic from the marine bacterium Pontibacillus sp. and P. oxalicum GXU20 [6,14]. Then, the substrate specificity of RSDE towards raw starches was investigated using the abovementioned substrates. The results are shown in Table 1. RSDEs from the WT 114-2 strain had a broad range of substrate specificities, in particular, the activity of RSDEs against raw rice starch was almost equal to the activity against soluble starch as the enzymatic reaction substrate. It exhibited the highest specific activity of 451.4 ± 12.6 U/mg towards soluble starch. Its enzymatic activity towards rice raw starch (94.1%), were much higher than those towards tested raw starches, including corn (60.3%), potato (26.9%), buckwheat (62.5%), and cassava (47.7%).

Scanning electron microscopy of raw rice starch hydrolysed by P. oxalicum RSDEs
Photographs of native and enzyme-hydrolysed starch were detected using scanning electron microscopy (SEM). Figure 2 shows that the surface of untreated raw rice starch granules was round and smooth ( Fig. 2A). This shape is similar to the shape of raw cassava starch observed by SEM [6]. In contrast, when RSDEs attacked the starch, the shape of starch collapsed rapidly in 1 min (Fig. 2B), which was different from the observations of a large hole extending into the granule interior. With the extension of reaction time, most of the residual starch was further degraded within 5 min (Fig. 2C). Surprisingly, the residual starch granule almost disappeared after incubation with RSDEs for 10 min (Fig. 2D). This observation indicates that degradation of starch with mixed RSDEs is faster than that of a single enzyme, which displayed a hole-digging degradation pattern [6].

Enhanced expression of RSDE production by genetic manipulation of four critic regulators
The reported studies and our primary research provided some clues that two TFs AmyR and CreA, a group III G protein α subunit PGA3 possibly mediate amylolytic enzyme expression [23,25]. By employing genetic engineering techniques, we constructed a series of engineered mutant transformants including OamyR (amyR overexpressed under the ubiD strong constitutive promoter), ΔCreA (deletion of the creA gene), and Mpga3 (expressing a dominantly activated PGA3 with substitution of Gln208 by leucine), and OhepA (hepA overexpressed under the ubiD strong constitutive  promoter). In particular, HepA was first identified for its role in regulating amylolytic enzyme expression in P. oxalicum. The four transformants were purified and identified by Southern blot and gene sequencing (Additional file 1: Fig. S1). The WT 114-2 and engineered strains were incubated in VMM plus soluble starch as a sole carbon source for 5 days. Expectedly, by using raw rice starch as a reaction substrate, markedly higher amylolytic enzyme activities were detected for both engineered strains compared with WT 114-2 during the fermentation period. The overexpression strains OamyR and OhepA showed 2.0-to 2.1-fold and 1.6to 2.5-fold increases, respectively, compared with WT 114-2. On the fourth day, the RSDE activity of the CreA deletion strain ΔCreA reached 80.4 U/mL, which was 3.2-fold higher than that of WT 114-2. In particular, after dominant activation of the G protein PGA3, the level of RSDE activity increased by 2.4-to 2.9-fold in Mpga3 compared with the WT strain (Fig. 3). Moreover, amylolytic enzyme activity was further measured by using soluble starch as a reaction substrate, and the results were quite consistent with the results obtained by using raw rice starch as a reaction substrate (Additional file 1: Fig. S2). In addition, the HepA deletion strain (ΔhepA) and PGA3-deletion strain (ΔPGA3) were constructed in a preliminary study. Neither ΔHepA nor ΔPGA3 was measured with increased yield of RSDEs (Additional file 1: Fig. S3); therefore, the two transformants have not been further investigated. To examine the transcriptional levels of RSDE genes in the four engineered strains, a RT-qPCR was carried out to compare the transcription patterns of the major RSDE gene amy15A [22]. As shown in Fig. 4, the transcript levels of amy15A in both engineered strains were obviously increased when compared with the WT strain. For OamyR, the transcript level of amy15A increased by 65.0-, 5.4-and 2.8-fold at 3, 21 and 48 h compared with the WT, respectively. The transcript levels of amy15A in OhepA were increased by 15.4-, 31.4-and 8.9-fold, respectively. The amy15A was upregulated remarkably in ΔCreA, and the expression of amy15A in ΔCreA was 5.2-, 2.8-, and 220.5-fold higher than that in WT at 3 h, 21 h and 48 h, respectively. The transcript abundance of amy15A was upregulated to 10.4-, 2.4-, and 3.5-fold in Mpga3 compared with that of the WT strain at the three detection points. The results suggested that overexpression of AmyR and HepA, deletion of CreA and dominant activation of PGA3 significantly activated RSDE gene expression at the transcript level.

Influence on spore germination and hyphal extension of the mutant strains
To determine the effect of the genetic modifications on P. oxalicum, phenotypic analyses were conducted. Equal amounts of fresh spores collected from P. oxalicum 114-2, OamyR, ΔCreA, OhepA and Mpga3 were inoculated on solid-medium plates in the presence of starch as the sole carbon source and PDA solid medium, and cultured at 30 °C for 6 days. The colonies of OamyR and Mpga3 showed no obvious change when compared with the WT 114-2 strain on PDA and starch plates. The colony of ΔCreA on PDA was smaller than that of the WT strain. Since spore germination of P. oxalicum comes from white hyphae, the colonies gradually turn pale green with the development and maturation of spores. Therefore, we can infer that the deficiency of CreA would affect spore germination even in conditions with abundant nutrition (PDA medium) while restricting mycelium extension in nutrient limiting conditions (starch medium). Spores of the OhepA strain germinated and grew slower on PDA, while a similar phenotype on starch was observed compared with the WT strain (Fig. 5).

Mutual regulation among the four regulators at the transcriptional level
Given the dose-controlled or additive regulation of glycoside hydrolases present in filamentous fungi [23,25,27], we assessed whether the four regulators, AmyR, CreA, HepA and PGA3 could affect each other's expression at the transcriptional level by RT-qPCR detection. RT-qPCR was performed using total RNA from all strains grown on raw rice starch medium for 3 h, 21 h and 48 h after a shift from glucose. Four genes, amyR, creA, hepA  and pga3 (or Mpga3), were selected for evaluation in WT 114-2 and four engineered strains. As shown in Fig. 6, the expression of amyR on rice starch was upregulated to different extents in the four engineered strains relative to the WT (Fig. 6A), implying that the four regulators may activate RSDE gene expression by promoting amyR expression. The transcription level of the creA gene in OamyR was slightly higher, while no obvious trend could be summarized for its expression in the other three engineered strains (Fig. 6B). The results demonstrate that the transcript pattern of the carbon catabolite repressor gene creA was constitutive expression. The transcription of hepA and pga3 was significantly upregulated in ΔCreA and exhibited no differential expression in the OamyR and Mpga3 strains ( Fig. 6C and D). The results indicated that pga3 and hepA gene expression was inhibited by carbon catabolite repressors in P. oxalicum.

Transcription profiling of P. oxalicum 114-2 and the mutant strains
To further understand the regulatory mechanism of RSDE expression on starch, global gene expression changes in the WT, OamyR, ΔCreA, OhepA and Mpga3 strains were evaluated by RNA-Seq with three biological replicates. Total RNA was extracted from the mycelia of P. oxalicum grown on rice starch for 21 h and then sequenced. Every sample produced more than 50 million clean reads, with a < 0.025% error rate and > 92.5% mapped into the genome of the WT strain (Additional file 1: Table S1). The high consistency of samples among the three biological replicates made the transcriptome data more reliable. Transcriptome data showed that the expression levels of 1781, 1659, 1684, and 1832 genes exhibited significant differences (twofold or greater, FDR < 0.001) in the OamyR, ΔCreA, OhepA and Mpga3 strains when compared with WT 114-2 (data not shown). The GO terms that included the most abundant genes in molecular function, cellular component, and biological process were metabolic process (GO:0008152), membrane part (GO:0044425) and catalytic activity (GO:0003824), respectively ( Fig. 7A-D). Furthermore, KEGG annotation indicated that upregulated and downregulated genes were most enriched in the process of amino acid metabolism and carbohydrate metabolism, demonstrating that synthesis of enzyme proteins and degradation of carbohydrates  responded simultaneously in four engineered strains ( Fig. 7E and F). Among the above DEGs, we were interested in the genes that were simultaneously upregulated and downregulated among the four engineered strains related to the genetic modification of RSDE regulators. Consequently, by using an FDR ≤ 0.01 and fold change ≥ 2 as the threshold, we determined that 702, 784, 793, and 996 genes were upregulated in response to the OamyR, ΔCreA, OhepA, and Mpga3 strain, respectively, compared with the WT 114-2 strain. Additionally, comparative transcriptome analyses identified 1079, 975, 891, and 836 downregulated genes in the four engineered strains compared with the WT strain, respectively. The Venn diagram shown indicates that 9 genes and 36 genes were found to be simultaneously upregulated and downregulated, respectively, in the four engineered strains ( Fig. 8A and B). In particular, three amylolytic enzyme genes Amy15A (PDE_09417), Amy13A (PDE_01201) and Agl31A (PDE_03966), were found to be involved in the 9 co-upregulated genes ( Table 2). Additionally, many genes annotated as uncharacterized and hypothetical proteins were included in those co-downregulated genes, demonstrating that those proteins may act as negative regulators of RSDE expression in P. oxalicum (Table 3). Generally, glycoside hydrolase expression is regulated by upstream transcription factor(s) in filamentous fungi [18,23,28]. TFs are considered essential players in signal transduction pathways, and are the last link between signal flow and target gene expression. Almost all putative transcription factor transcript levels in the four engineered strains were comparatively analysed with their transcript levels in WT 114-2 as a standard. We are interested in those transcription factors that are co-upregulated or co-downregulated in the four engineered strains. As a result, 25 transcription factors, most of which contained zinc-related structures (Zn_clus, C2H2, Zn2Cys6, etc.), were co-upregulated or co-downregulated in at least two or more mutants. As shown in Fig. 8C, Cluster 1 (C1) consisted of 7 TF genes displaying upregulation while C2 consisted of 14 TF genes displaying downregulation in three or four engineered strains. Among these 25 transcription factors, 5 TF genes (C3) were co-upregulated in the two mutants and co-downregulated in the other two strains. These data clearly indicated that the transcription of the putative TFs was regulated in different modes. In particular, of these genes, PDE_06057 (Zn_clus), PDE_06088 (Zn_clus) and PDE_09536 (d-galacturonic acid-responsive regulator) were found to be co-upregulated, while only PDE_02772 (Zn_clus) was found to be co-downregulated in the four engineered strains.

Discussion
In the present study, we systematically investigated the degradation ability of P. oxalicum 114-2 on different raw starches. The results showed that the amylolytic enzymes secreted by the 114-2 strain had comprehensive degradation activity towards raw starch from different sources. The RSDEs from P. oxalicum 114-2 showed better amylolytic activity against rice raw starch, while they exhibited a faster deconstruction rate when degrading starch granules ( Figs. 1 and 2). This may be due to the presence of other amylolytic enzymes, such as Amy13A, which may generate synergistic effects with Amy15A in the raw rice starch degradation process (Additional file 1: Fig. S4). The mixed RSDEs that mainly contained Amy15A and Amy13A showed higher activity against rice raw starch and lower activity against potato raw starch. The results are in agreement with those of PoGA15A from P. oxalicum GXU20 [6], demonstrating that the major RSDE Amy15A plays a dominant role in the degradation of raw starch. Unlike the heterologously expressed PoGA15A in Pichia pastoris, we obtained Amy15A by homologous expression in an engineered P. oxalicum host strain [29].
Although amino acid sequence alignment shows the high sequence identity between Amy15A and PoGA15A, some differences in enzymatic properties were found for the two amylolytic enzymes. Their pH stability was similar, but the temperature stability of the homologousexpressed Amy15A was not as high as that of the heterologous-expressed PoGA15A (Additional file 1: Figs. S5 and S6). This may be due to differences in posttranslational modifications in Pichia pastoris, which resulted in protein structure changes that were unexpectedly beneficial for enzyme temperature stability. The expression of glycoside hydrolases in filamentous fungi is generally regulated by upstream regulators, with transcription factors being the most studied. Overexpression and knockdown of a single transcription factor are usually used to positively and negatively reveal the regulatory mechanism involved in glycoside hydrolase expression in filamentous fungi [30][31][32]. However, an increasing number of reports demonstrate that extracellular enzyme expression is not determined by a single regulator, but may be synergistic and dose-controlled by an upstream regulatory network consisting of multiple regulators [23,27]. The four regulatory factors AmyR, CreA, HepA, and PGA3, which were previously confirmed as regulators involved in cellulase expression, were newly identified for their crucial regulatory function in RSDE expression in this study. Additionally, unlike those positive and negative studies of single transcription factors by application of gene overexpression/knockout technology, we tried to reveal the mode of RSDEs regulatory network from a positive insight by modifying the four regulators to make the downstream RSDEs production increased. We can speculate about the possible regulatory network model involved in RSDE expression. The plasma membrane-associated G protein binding of an extracellular ligand to transmembrane G protein-coupled receptor (GPCR) leads to GDP-GTP exchange on the Gα subunit and its activation [25,33]. The Gα subunit PGA3 was dominantly activated and enhanced the expression of AmyR, which in turn increased the expression of RSDEs (Figs. 4 and 6A). Therefore, we can speculate that the G protein signalling pathway, a central signalling cascade with crucial functions in all organisms, may mediate the expression of RSDEs by transmitting external signals. CreA, as a homologue of the major carbon catabolite repressors in phylogenetically diverse fungi, played a negative role in RSDE expression [34,35]. This regulator allows an organism to utilize a preferred carbon source but hinders it from metabolizing complex carbon sources, including raw starch. Deletion of CreA to relieve catabolite repression not only increased RSDE gene expression but also enhanced the expression of their upstream activators. It is suggested that CreA might have a cascade regulation because it repressed the activator genes of AmyR, PGA3 and HepA as well as the structural genes whose expression was upregulated by the three regulators (Fig. 6); therefore, the significant improvement of RSDE yield may be due to the "double-enhancement" effect after deleting CreA (Figs. 3  and 4). In addition, overexpression of HepA, a protein associated with a highly compact chromatin state [36,37], may alter the structure of chromosomes to make it accessible for transcription machinery to attach to target genes. In particular, the regulators CreA and HepA are conserved in filamentous fungi [34][35][36][37]; therefore, the possible existence of an intimate crosstalk among certain developmental processes, such as sporulation and RSDE production pathways, is mediated by those regulators in ascomycete fungi (Fig. 5). AmyR, a master and key transcription factor, was exceptionally identified and positively regulated amylolytic enzyme gene expression in P. oxalicum. We speculated that the activation of RSDE expression by other regulators (i.e., CreA and HepA) may be partially due to enhanced AmyR expression (Fig. 6A). Taken together, in contrast to investigation of the regulation of multiple transcription factors, the expression of RSDE genes was found to be regulated by various global regulators in this study (Fig. 9). We further expanded our understanding of the regulatory network of RSDE expression by analysing regulators in different regulatory manners. Although comprehensive and intensive analyses are required, this study may provide a "stereoscopic" or possible temporal regulation mode for RSDE expression in P. oxalicum.

Conclusions
While RSDE-assisted starch industrial processing has been a topic of study over the years, the gaps in our knowledge regarding the RSDE expression and the regulation of the pathways could be considered actual bottlenecks in the development of more effective genetic engineering strategies. In this study, we identified a number of common genes and metabolic pathways that mediate RSDE expression through global transcriptome analysis of four engineered strains. These common upregulated or downregulated genes are most likely to be involved in the regulation of RSDE expression, which will provide reduced but valuable targets for the next step of optimizing the regulatory network. These sophisticated regulatory mechanisms may also function  to optimize carbon and energy flux. We will further investigate these common genes to explore more signalling transmitters or regulatory factors to improve the regulatory network of RSDE expression, thus providing theories for the most fruitful means of genetic modification to obtain higher industrial RSDE yield in the future.

Strains and culture conditions
The wild-type strain 114-2 was deposited at the China General Microbiological Culture Collection Center (CGMCC) under the number CGMCC 5302 and used throughout this work as a control. Raw starches (rice, corn, potato, buckwheat and cassava) were purchased from a local market in Changsha, China. The raw starch substrates used for the enzymatic activity assay were washed at least three times with ultra-pure water. Soluble starch was purchased from Sigma-Aldrich (St. Louis, MO, USA). The strains listed in Table 4 were routinely grown in wheat bran extract slants at 30 °C to obtain fresh conidia. For phenotypic analyses of strains on agar plates, Vogel's minimal medium (VMM) supplemented with 1% (wt/vol) soluble starch and PDA was used [38], and the conidia of strains were spotted in 1-μL aliquots. All strains were grown on solid plates at 30 °C for 6 days. For microscopic observation and RSDE activity determination, conidia of strains at a final concentration of 10 7 per mL were inoculated into and grown in liquid glucose minimal medium for 24 h. Then, mycelia were collected by vacuum filtration, and 0.5 g wet mycelia were resuspended in 100 mL liquid VMM supplemented with 1% (wt/vol) soluble starch for further growth. The cultures were maintained at 30 °C on a rotary shaker at 200 rpm. The fermentation supernatant was used for microscopic observation of starch degradation and amylolytic enzyme activity determination.

Construction of targeting cassettes and transformation of P. oxalicum
The genomic DNA used as a PCR template was extracted from young hyphae grown on PDA liquid medium. The knockout or overexpression cassettes for each candidate gene were constructed by fusion PCR [39]. The cassettes contained a 1.9-kb hph gene and DNA fragments approximately 2.0 kb upstream and downstream of the target gene, and the DNA fragments were amplified by PCR using the corresponding primer pairs (Additional file 1: Table S2), For site-directed mutagenesis of pga3, the coding sequence of Mpga3 (expressing a dominantly activated PGA3 with substitution of Gln208 by leucine), which contains a mutant site, was synthesized. The template of the pga3 mutagenesis cassette was obtained by fusing the Mpga3 coding region, hph cassette, and 3' flanking region (Additional file 1: Fig. S1). For all transformations of P. oxalicum, protoplasts were prepared based on a modified method as described by Gao et al. [40].

Southern blot and sequencing
To determine the integration type of targeting cassettes in transformants, Southern blot analyses were performed (Additional file 1: Fig. S1). The mycelia were collected and genomic DNA was extracted as described previously [41]. The prepared genomic DNA (approximately 30 μg) of the OamyR, OhepA, ΔCreA, Mpga3 and 114-2 strains was digested by restriction enzymes, respectively, and then separated using agarose gel electrophoresis. Three probes were amplified for strain OamyR/OhepA, ΔCreA and Mpga3 analysis. A DIG-High Prime DNA Labelling and Detection Starter Kit I (Roche, Basel, Switzerland) was used for probe preparation, fragment hybridization, and immunological detection according to the manufacturer's instructions. The coding region sequence of Mpga3 was amplified from the Mpga3 transformants, and the mutations were confirmed by Sanger sequencing (BGI, Shenzhen, China).

Enzyme assays and SDS-PAGE analysis
The raw starch-digesting enzyme activity was measured according to the DNS method [42]. The absorbance of the reaction system was measured at 540 nm. Up to 1.5 mL of starch solution was added for RSDE activity assays, and the reaction was incubated at 40 °C for 10 min. One unit of enzymatic activity (U) was defined as the amount of enzyme required to produce 1 μmol of reducing sugars per min from the reaction substrate. SDS-PAGE was performed using 12% polyacrylamide to determine protein purity. The protein profile was analysed by staining gels

RNA isolation, cDNA synthesis and quantitative RT-PCR
We validated the expression pattern of related genes by RT-qPCR. Total RNA was extracted using TRIzol reagent (TaKaRa, Dalian, China), and complementary DNA was generated using a reverse transcription kit following the manufacturer's instructions. Quantitative real-time PCR was performed using SYBR Premix Ex Taq ™ (Perfect Real Time) (TaKaRa, Dalian, China) with the primers listed in Additional file 1: Table S2. Each amplification reaction was in a total reaction volume of 20 µL. The thermal cycling protocol was as follows: initial denaturation for 2 min at 95 °C followed by 40 cycles of 10 s at 95 °C and 30 s at 61 °C. The fluorescence signal was measured at the end of each extension step at 80 °C. The reactions were performed in triplicate, and average transcription levels were determined and normalized to corresponding actin gene expression levels as an internal control.

RNA-sequencing
Fresh spores of each strain were inoculated into 100 mL of 1% glucose medium and incubated at 200 rpm and 30 °C for 24 h. Equal amounts of mycelia were transferred to 1% raw rice starch medium and cultured for 21 h. Total RNA was extracted from the different samples by using a TRIzol total RNA extraction kit (TaKaRa, Dalian, China) and treated with DNase I (TaKara, Dalian, China) to remove the DNA. Transcriptome assays based on Illumina sequencing technology were performed at Shanghai Majorbio Biopharm Technology Co., Ltd (Shanghai, China). After quality control, the generated clean reads were mapped against predicted transcripts from the P. oxalicum 114-2 genome. Transcript abundance (fragments per kb per million reads, FPKM) genes with significantly different expression levels were identified through a significance test with combined thresholds (FDR ≤ 0.01 and fold change ≥ 2). Pearson's correlation coefficient was used to evaluate transcriptome reliability, and three biological replicates were used in each sample. Differential gene expression was analysed by using the Majorbio cloud computing platform, included a series of DESeq software packages.