Enhanced production of terrein in marine-derived Aspergillus terreus by refactoring both global and pathway-specific transcription factors.

BACKGROUND: Terrein, a major secondary metabolite from Aspergillus terreus, shows great potentials in biomedical and agricultural applications. However, the low fermentation yield of terrein in wild A. terreus strains limits its industrial applications.
RESULTS: Here, we constructed a cell factory based on the marine-derived A. terreus RA2905, allowing for overproducing terrein by using starch as the sole carbon source. Firstly, the pathway-specific transcription factor TerR was over-expressed under the control of a constitutive gpdA promoter of A. nidulans, resulting in 5 to 16 folds up-regulation in terR transcripts compared to WT. As expected, the titer of terrein was improved in the two tested terR OE mutants when compared to WT. Secondly, the global regulator gene stuA, which was demonstrated to suppress the terrein synthesis in our analysis, was deleted, leading to greatly enhanced production of terrein. In addition, LS-MS/MS analysis showed that deletion of StuA cause decreased synthesis of the major byproduct butyrolactones. To achieve an optimal strain, we further refactored the genetic circuit by combining deletion of stuA and overexpression of terR, a higher terrein yield was achieved with a lower background of byproducts in double mutants. In addition, it was also found that loss of StuA (both ΔstuA and ΔstuA::OEterR) resulted in aconidial morphologies, but a slightly faster growth rate than that of WT.
CONCLUSION: Our results demonstrated that refactoring both global and pathway-specific transcription factors (StuA and TerR) provides a high-efficient strategy to enhance terrein production, which could be adopted for large-scale production of terrein or other secondary metabolites in marine-derived filamentous fungi.

Introduction

( +)-Terrein, the major secondary metabolite produced by Aspergillus terreus, was firstly discovered in 1935 and its chemical structure was elucidated in the 1950s. Until recently, terrein has been reported to possess strong anti-tumor activities by targeting multiple pathways, including inhibition of melanin biosynthesis [1-3], anti-proliferation [4, 5], suppressing angiogenin production and inducing apoptosis of carcinoma cells [6]. And, terrein displays strong cytotoxicity against ABCG2-expressing breast cancer cells [7], further expanding its potentials in the development of effective therapeutic drugs of cancer. Also, terrein has anti-inflammatory properties in multiple cell lines [8-10], phytotoxicity [11], and antimicrobial activities [12]. Interestingly, terrein is the first reported dual inhibitor of QS and c-di-GMP signaling in the pathogenic bacteria Pseudomonas aeruginosa [13], showing potential in the development of novel antibiotics against multidrug-resistant (MDR) gram-negative bacteria. Therefore, terrein has important potential applications in medicine and agriculture.

The biosynthetic gene cluster for terrein has been identified, and its biosynthetic pathway has partially elucidated by genetic and biochemical analysis [11]. The cluster, which responsible for terrein synthesis, consists of 11 functional genes. Initially, two acetyl-CoA and four malonyl-CoA units are condensed and catalyzed by TerA, the non-reducing polyketide synthase, to generated 6-hydroxy-2,3-dehydromellein (2,3-dehydro-6-HM). Then, 2,3-dehydro-6-HM is reduced to 6-HM by the multi-domain protein TerB (Fig. 1A). Until now, the biosynthesis pathway from 6-HM to the end product of terrein remains unresolved. Markus Gressler et al. reported that nitrogen or iron starvation acts as the signal to induce terrein production, which depend on function of the global transcription factors AreA, AtfA and HapX [14]. Notably, elevated methionine level promotes terrein production even under non-inducing condition [14].

Fig. 1

Graphical abstract and summary (A) biosynthesis of terrein (B) development of a cell factory for terrein production

Like other fungal secondary metabolites, the production level is the bottleneck of further development and application of terrein. Screening of strains and media, and process optimization had been applied to improve terrein titers [15-18]. However, rational strain engineering has yet been used to improve terrein production until now. In this study, we achieved the high-producing and robust cell factory for terrein synthesis by modifying both the pathway-specific regulator TerR and the global regulator StuA (Fig. 1B). The modified mutants grow faster and produce terrein at gram level by using low-cost substance. We also found that the mutant strain can further reduce the costs by inhibition of biosynthesis of byproducts.

Results

Identification the biosynthetic gene cluster for terrein in marine-derived fungus A. terreus RA2905

( +)-Terrein is the major secondary metabolite produced by A. terreus, highlighted by its broad and excellent biological activities, including antimicrobial, anti-inflammatory, antitumor activities. Thus, its biosynthesis and regulatory mechanisms has always been a focus of attention in fungal secondary metabolism. A gene cluster of approximately 33 Kb, consisting of 11 functional genes, had been previously identified for terrein biosynthesis in A. terreus SBUG84411. By sequence alignment, we identified the terrein cluster in A. terreus RA2905 (Table 1). Overall, the terrein gene cluster was highly conserved between A. terreus RA2905 and A. terreus SBUG844. The exception is that a protein of 118 amino acids with unknown function specially exists in the terrein cluster of A. terreus RA2905, but not in A. terreus SBUG844. According to previous findings, the protein is not likely involved in biosynthesis of terrein. In the biosynthesis of terrein (Fig. 1A), TerA firstly assembles two acetyl-CoA and four malonyl-CoA to generate 6-hydroxy-2,3-dehydromellein (2). Then, the compound 2 is reduced to 6-hydroxymellein (3) by TerB. Other four unidentified enzymes (TerC, TerD, TerE and TerF) involve in the later biosynthesis pathway from 3 to the end product terrein (1). A Zn2Cys6-class transcription factor, named TerR, was presented in RA2905. It had been previously reported that TerR played a positive role in the regulation of expression of other terrein biosynthesis genes in SBUG844 [14].

Table 1

Terrein biosynthetic gene cluster

RA2905	NIH 2624	Identity (%)	Functional description
gene2371	TerA	96.52	Non-reducing polyketide synthase
gene2372	TerB	96.81	6-Hydroxymellein synthase
gene2373	TerC	99.26	FAD-dependent monooxygenase
gene2374	TerD	99.51	FAD-dependent monooxygenase
gene2375	TerE	98.58	Multicopper oxidase
gene2376	TerF	99.08	Kelch-like protein
gene2377	TerG	100.00	Efflux pump
gene2378			Unknown
gene2379	TerH	99.41	NAD-dependent epimerase/dehydratase
gene2380	TerI	100.00	Lactoylglutathione lyase-like protein
gene2381	TerJ	90.30	Efflux pump
gene2382	TerR	98.07	Transcription factor

To determine whether A. terreus RA2905 has the ability to synthesize terrein, compounds in the crude extracts from its rice cultures were chemically purified and structurally characterized. One of these compounds (30 mg) with an appearance of white power was obtained, and its molecular formula was determined as C8H10O3 by the HR–ESI–MS data at 309.2 [2 M + H]+, which was consistent with the formula of terrein reported previously. Moreover, our 1H and 13C NMR data further confirmed the obtained compound was terrein (Additional file 1: Figure S1 and Figure S2). Together, these results suggested that A. terreus RA2905 could synthesize terrein, but the production titer was low.

Overexpression of pathway-specific transcription factor TerR to improve terrein yield

To activate the terrein cluster expression, firstly, we over-expressed the terR gene in RA2905. The constitutive and strong promoter from glyceraldehyde-3-phosphate dehydrogenase of A. nidulans was PCR-amplified and used to drive terR overexpression. Then, the terR expression cassette was fused with the selection marker gene pyrg, and then transformed into uracil autography mutant ΔpyrG established previously [19]. A total of 11 transformants with uracil prototroph were obtained, and five of them have the whole expression cassette by diagnostic PCR analysis. After further verification by using quantative PCR, OEterR-1 and OEterR-6 showing increased expression level in comparison to that of WT, were used for further analysis. The fungal phenotype and conidial development of two terR overexpression mutants and WT were analyzed (Fig. 2). The results showed that overexpression of terR has a faint effect on fungal growth and asexual development in A. terreus RA2905 (Fig. 2). The OEterR-6 mutant showed slightly faster growth rate than that of WT on solid plates (Fig. 2). Both terR overexpression mutants produced largely equal amount of conidia on all plates tested (Fig. 2).

Fig. 2

Phenotype and conidial development of terrein-producing mutant. All colonies were cultured on various media at 28 ℃ for 7 d

By expected, the production of terrein was significantly improved in terR overexpression mutants compared to WT (Fig. 3A). Based on our preliminary data, a low-cost medium TFM, consisting of starch and peptone as carbon and nitrogen resource, was used for fermentation production of terrein. The 7-day fermentation, terrein were organically extracted and the titer was determined by HPLC. Our result showed that the titer of terrein was 0.5- and twofold improved in OEterR-1 and OEterR-6, respectively, as compared with that in WT strain (Fig. 3A). These results further demonstrated that TerR was a specific regulator of terrein, and its overexpression represented an efficient strategy to enhance terrein biosynthesis.

Fig. 3

Yields of terrein and biomass in all mutant and WT strains. Fungal mycelia (10 g/L) were cultured in TFM media at 28 °C, 150 rpm for 7d, and yields of terrein and biomass for individual strain were determined by HPLC and dry weight method. Titer of terrein were the mean of three biological replicates. ns. means no significance, *p ≤ 0.05 and **p ≤ 0.01

Substantially enhancement of terrein yield by deleting stuA

In addition to pathway-specific regulator, transcription of most of secondary metabolite gene clusters are also controlled by global regulators. These regulatory proteins not only involve in secondary metabolism, but also have roles in fungal growth, development, as well as primary metabolic process [20]. The APSES transcription factor StuA was previously known as the conserved morphological modifier, playing crucial roles in cellular differentiation, mycelial growth, as well as asexual and sexual development in filamentous fungi [21, 22]. Recent studies has suggested that StuA regulates biosynthesis of multiple SM gene clusters in different species [21, 23], implying that it might function as a global regulator of SM in fungi. We presumed that StuA might contribute to biosynthesis of terrein in A. terreus for its high transcription abundance during terrein inducing media. The stuA mutant has been established in our previous investigation [19], but its biological function remain unknown in A. terreus.

As in other filamentous fungi, there was little to no conidia could be observed in the ΔstuA mutant regardless of media used (Fig. 2), while WT produced abundant amounts of conidia on any medium. Unlike in other fungi, our results suggested that radial growth of the ΔstuA mutant was approximately 10% increased as compared with that of WT on PDA, GMM, and TFM. Interestingly, a pale yellow pigment was clearly observed in the ΔstuA mutant, especially on the TFM plate, but not in WT. Considering that multiple precursors in the terrein biosynthesis are yellow, it can be speculated that the terrein gene cluster was activated in the mutant. To investigate the effect of StuA on terrein production, both the ΔstuA mutant and WT strains were cultured in TPM media to induce terrein biosynthesis. The yield of terrein was detected at 5 day by HPLC. The results showed that the ΔstuA mutant produced 10.8 mg/ml terrein, which was 21 folds more than that of WT. These results showed that deletion of stuA greatly enhanced terrein production.

Construct a robust and high-producing cell factory for terrein synthesis by combing terR overexpression and stuA deletion

To further improve the terrein production, the pathway-specific transcription factor TerR and the global regulator StuA were simultaneously genetically modified in the marine-derived fungus A. terreus RA2905. The double gene mutant ΔstuA::OEterR were obtained by deletion of stuA in the OEterR-6 mutant with hph was used as the selectable marker gene described previously [19]. The resultant mutant ΔstuA::OEterR also lost the ability to develop conidia, which is the same as ΔstuA (Fig. 2), its terrein yield was further enhanced than that of ΔstuA or OEterR, reach the level of 12.6 g/L. More intriguingly, approximately 80% of fungal extract for ΔstuA::OEterR cultured in TFM media was terrein, which represent the highest yield to date.

Loss of StuA improve the biomass of A. terreus

To investigate the TerR and StuA on fungal growth under fermentation process, the biomass of the ΔstuA and OEterR mutants, as well as WT strains were analyzed by dry cell weight assay (DCW) (Fig. 3B). Considering that the ΔstuA mutant did not produce conidia, a shift experiment was designed to achieve the same initial inoculum dose. The results showed that two terR overexpression mutants develop the same biomass as that of WT, suggesting that TerR not involved in fungal growth. In contrast, the biomass of ΔstuA mutant (16 g/L) was significantly increased compared to WT (12 g/L), indicating that StuA play a negative regulatory role in fungal growth.

Expression analysis of terrein biosynthesis genes

Compared with the WT strain, a significant improvement in terrein yields was observed in OEterR, ΔstuA, ΔstuA::terR mutant mutants. We wondered whether this improvement was the result from up-regulation in gene transcription level. Therefore, the transcripts of the major terrein biosynthetic genes, including terA ~ terR, are analyzed by semi-quantitative PCR. Considering that no conidia were produced in ΔstuA or ΔstuA::terR, a shift experiment of mycelia was designed to ensure the same inoculation size and time among stuA mutants and WT. As expected, the results indicated that the transcript abundance of most of terrein biosynthetic genes were up-regulated significantly in both OEterR and ΔstuA mutant than that of WT under terrein induction media (Fig. 4), and the double mutant showed the highest expression level, which was consistent with terrein titer analysis above. These results demonstrated that all of these mutants improve the terrein production by upregulating the expression level of terrein biosynthetic genes.

Fig. 4

Expression level analysis of biosynthetic genes for terrein. Fungal strains were cultured in TFM media at 28 °C, 150 rpm for 48 h, and the gene expression of terA ~ terR were analyzed by semi-quantative PCR, and the expression level of actin gene was used as the control

Block the biosynthesis of other secondary metabolites in terrein-producing mutant

As a prolific producer in secondary metabolism, A. terreus wild type strain was reported to produce diverse secondary metabolites simultaneously. And, a variety of secondary metabolites, including butyrolactones, thiodiketopiperazine, and benzyl furanones and pyrones were isolated from the marine-derived A. terreus RA2905 [24, 25]. To determine the effect of terR overexpression, stuA deletion and combined mutation on the biosynthesis of other secondary metabolite, the extracts of OEterR, ΔstuA or ΔstuA::OEterR and WT were analyzed by LC–MS/MS (Fig. 5). The results further supported that the production of terrein in all mutants were enhanced, and ΔstuA::OEterR showed the highest yield, in turn by ΔstuA, and OEterR mutants. As expected, OEterR mutant produced comparable butyrolactones and other secondary metabolites as WT strain. Significantly, other secondary metabolites, specially butyrolactone, which was the major metabolite in WT, were completely abolished in ΔstuA or ΔstuA::OEterR. Therefore, ΔstuA::OEterR established here was a cell factory for robust and high-level producing terrein, which greatly reduced the following purification process.

Fig. 5

LC–MS/MS analysis of the extracts of both mutants and WT strains. Peak between two lines was confirmed to be the terrein based its molecular ion peak in MS

Discussion

The A. terreus secondary metabolite terrein had been demonstrated to possess extensive biological and pharmacological benefits, including anti-cancer, anti-inflammatory and antimicrobial activities. Furthermore, Matthias Brock et al. has reported that the terrein gene cluster was a high-performance expression system for biosynthesizing other fungal secondary metabolites [26]. However, the terrein yield and expression level of terrein genes were relatively low in the wild type strain of A. terreus. Therefore, identification and engineering of critical factor for terrein production is essential for providing resource of pharmacological research, as well as developing the more efficient and robust expression platform of fungal secondary metabolite gene cluster.

The APSES transcription factor StuA was initially found to be a regulator of fungal asexual and sexual development [22]. Recently, its regulatory role in secondary metabolism has been revealed in several fungi [21, 23]. In general, StuA positively regulates the SM gene expression, and its deletion resulted in reduction or even absence of secondary metabolite biosynthesis, such as aflatoxin [21]. However, it was interestingly found that deletion of stuA in A. terreus lead to reduced biosynthesis of other secondary metabolism, such as butyrolactones, but increase terrein production. However, the underlying mechanism remain further investigation.

Biosynthesis of secondary metabolites are controlled by both global and pathway-specific regulators in fungi. Global regulators, such as methyltransferase LaeA, were mainly more conserved than that of these pathway-specific regulators in the regulation of secondary metabolism [27, 28]. However, it was found that these global regulators also interfere with primary metabolic process in fungi, such as growth and conidial development, which caused unexpected defects [29]. While pathway-specific regulators are confined to the given BGC in a species, its function was specialized, could not deter the primary metabolism, as well as other secondary metabolite biosynthesis. Furthermore, these global regulators have regulatory roles in the expression of pathway-specific transcription factors, which establish a regulatory cascade for accurately and sharply control the secondary metabolite production. Thus, interpretation and engineering of the pathway-specific and global regulators, as well as its relationship, were essential for developing the efficient cell factory for a given and valued secondary metabolite production. Only alteration of either was limited for SM production engineering. Here, the regulatory program was redesigned by combing the pathway-specific regulator terR overexpression and the global regulator stuA deletion, the double mutant produced about 30 folds more terrein than that of WT, achieving ten-gram-level fermentation production of the anti-cancer compound terrein.

In conclusion, a genetic roadmap for engineering terrein production was established by combing redesign of both global and pathway-specific regulators, greatly improve the terrein production, not only help deepen the understand of regulatory mechanism of secondary metabolism, but provide a paradigm for yield improvement of other valued secondary metabolites in A. terreus and other filamentous fungi.

Materials and methods

Strains, primers and media

Strains of marine-derived A. terreus used in this study were listed in Table 2. All A. terreus strains were cultured on PDA (potato broth 20% and glucose 2% in sea water) media for propagation. GMM (Czapek–Dox salt 2% and glucose 2% in sea water) supplemented with 1 M sorbitol were used to for protoplast regeneration, or supplemented o.5 mM uridine and 1 mM uracil if necessary. TFM (starch 3%, peptone 1.5% and aginomoto 0.5%) media was used to ferment production of terrein. A. terreus mutants and WT strains were grown on PDA, GMM and TFM media supplemented with 1.5% agar for growth and conidial development analysis.

Table 2

Strains used in this study

Strain	Genotype	Reference
RA2905	WT	CCTCC AF 2,021,052
ΔpyrG	ΔpyrG::ptrA	[19]
OEterR-1	ΔpyrG::ptrA;PgpdA::terR::AfpyrG	This study
OEterR-6	ΔpyrG::ptrA;PgpdA::terR::AfpyrG	This study
ΔstuA	ΔpyrG::ptrA;ΔstuA::AfpyrG	[19]
ΔstuA::OEterR	ΔpyrG::ptrA;PgpdA::terR::AfpyrG	This study

Construction of mutants

To construct terR overexpression strains, open reading frame (ORF) and terminator of terR was amplified from A. terreus genomic DNA with primers GPDterRF and terR1. The constitutive promoter of gene gpdA from A. nidulans was amplified from plasmid pIG1783 with primer pair GPDF/ GPDR, and the selectable maker gene pyrG as amplified from A. fumigatus genomic DNA with primers PyrGF/PyrGR. The three fragments obtained above were fused by using Double-Joint PCR based on the homologous sequence in primers. Preparation of protoplast and transformation of A. terreus RA2905 were performed as established previously [19]. All primers used were listed in Table 3.

Table 3

Primers used in this study

Name	Sequence	Use
PyrGF	GCCTCAAACAATGCTCTTCACCC	Amplification of pyrG
PyrGR	CAGAAAGAGTCACCGGTCACTGT ACGTCTGAGAGGAGGCACTGATGC	Amplification of pyrG
GPDF	GTACAGTGACCGGTGACTCTTTCTG	Amplification of promoter
GPDR	GGTGATGTCTGCTCAAGCGGGGTAG	Amplification of promoter
terR2	TCCCTTCATACCTCGCTTTA	Amplification of terR
GPDterRF	CCGCTTGAGCAGACATCACCATGTTCGCCGAACTTAACGC	Amplification of terR
terR1	TTTGAGGCTTATACAACGAG	Amplification of terR
terR2	AGACGCTTGTAGCCCGTTAT	Amplification of terR
TerARTF	AGCAGCAAGTCATACAGCAA	Expression analysis
TerARTR	AACTCCTCGCATAACAAAGC	Expression analysis
TerBRTF	CAAGGATTGAAGGCGGGATC	Expression analysis
TerBRTR	AGAAGTTTCGGGCACAGGTT	Expression analysis
TerCRTF	GCATCTGCAAAGCCTCCTGT	Expression analysis
TerCRTR	CTCTGGCGTGTCTCATACCG	Expression analysis
TerDRTF	TCAGGCTGGAGACAGTAATC	Expression analysis
TerDRTR	AACGCATCTGGTATCTGGTC	Expression analysis
TerERTF	TGGTGCCTCCCGAGAAACAG	Expression analysis
TerERTR	CTCGCCGACCTTGGCTACAT	Expression analysis
TerFRTF	TACGACCAGGAAGGAACAGT	Expression analysis
TerFRTR	GAGCCGATAGTAGCAGAAGC	Expression analysis
TerRRTF	CCCGACTCAGCAAGACGAAG	Expression analysis
TerRRTR	TGAAAGCCGAAAGCCATCAC	Expression analysis

HPLC analysis of terrein

All strains were cultivated in TFM at 30 °C, 150 rpm for 7 days. The fungal cultures were extracted with ethyl acetate at least three times. HPLC analysis was performed on an Agilent InfinityLab LC Series 1260 system coupled with a Diode Array Detector WR detector and using an analytical Kromasil column (250 mm × 4.6 mm, 5 mm). The mobile phase was consisted of 0.1% formic acid in acetonitrile (A) and 0.1% formic acid in water (B), and the flow rate was 1 mL/min with 10 μL injection volume. Terrein was recorded at 281 nm.

LC–MS analysis of secondary metabolites

LC–MS/MS were analyzed on a UHPLC system (1290, Agilent Technologies) with a UPLC HSS T3 column (1.8 μm 2.1 × 100 mm, Waters) coupled to a quadrupole time-of-flight mass spectrometer 6545 (Q-TOF, Agilent Technologies) equipped with an ESI dual source in positive-ion mode. The ESI conditions were as follows: the capillary temperature at 350 °C, source voltage at 4 kV, and a sheath gas flow rate of 0.5 mL/min. The mass spectrometer was operated with an auto MS/MS mode. Mass spectra were recorded from m/z 200 to m/z 1000 in a speed of 6 spectra/sec, followed by MS/MS spectra of the twelve most intense ions from m/z 200 to m/z 1000 in a speed of 12 spectra/sec.

Gene expression analysis

All utilized A. terreus strains were inoculated into GMM liquid media and pre-cultured for 24 h, and then transferred into fresh LFM media at 28 °C for 48 h. The mycelia were collected via filtering, were then frozen in liquid nitrogen, and stored under − 80 °C conditions. RNA extraction and cDNA synthesis were performed using a RNA reagent (TRIzol reagent, Biomarker Technologies, Beijing, China) and a cDNA synthesis Kit (TransGen, Beijing, China) following the protocols of the manufactures, respectively. The quality and integrity of RNA samples was determined using Nanodrop 2000 (Thermo Fisher Scientific, USA) and Agilent 2100 bioanalyzer (Agilent Technologies, Palo Alto, CA), respectively, while the quantity was determined with a Qubit RNA assay kit (Life Invitrogen, USA). All utilized primers were listed in Table 3. The expression level of actin gene was used as the internal control.

Additional file 1: Determine the structure of terrain. Figure S1. 1H NMR of terrain. Figure S2. 13C NMR of terrain.