Enhancement of Monascus Azaphilone Pigments Production by Activating the cAMP Signalling Pathway in Monascus Purpureus HJ11

Background: Monascus azaphilone pigments (MonAzPs), which were produced by Monascus species, have been used as important food colorant and food supplements for more than one billion people during their daily life. Moreover, MonAzPs recently have received more attention because of their diverse physiological activities. However, the high microbial production of MonAzPs is still not always guaranteed. Herein, the aim of this study was to develop an ecient biotechnological process for MonAzPs production. Results: In this study, exogenous cyclic adenosine monophosphate (cAMP) treatment not only induced MonAzPs production, but also stimulated the expression of a cAMP phosphodiesterase gene, named as mrPDE, in M. purpureus HJ11. Subsequently, MrPDE was identied as a cAMP phosphodiesterase by in vitro enzymatic reaction with puried enzyme. Further, a gene knockout mutant of mrPDE was constructed to verify the activation of cAMP signalling pathway. Deletion of mrPDE in M. purpureus HJ11 improved cAMP concentration by 378 % and enhanced PKA kinase activity 1.5-fold, indicating that activation of cAMP signalling pathway was achieved. The ΔmrPDE strain produced MonAzPs at 8563 U/g, with a 2.3-fold increase compared with the WT strain. Moreover, the NAPDH/NADP + ratio of the ΔmrPDE strain was obviously higher than that of the wild type strain, which led to a higher proportion of yellow MonAzPs. With fed-batch fermentation of the ΔmrPDE strain, the production and yield of MonAzPs achieved 332.1 U/mL and 8739 U/g. Conclusions: A engineered M. purpureus strain for high MonAzPs production was successfully developed by activating the cAMP signalling pathway. This study not only describes a novel strategy for development of MonAzPs-producing strain, but also provides a roadmap for engineering efforts towards the production of secondary metabolism in other lamentous fungi. U/g, which was 2.3-folds higher than that of WT strain. In order to improve the MonAzPs production, the fed-batch fermentation was performed in a 10-L stirred fermenter. The MonAzPs production was signicantly enhanced to 332.1 U/mL, with the MonAzPs yield of 8739 U/g DCW. This study describes a promising method for high production of MonAzPs and provides a strategy for metabolic engineering of secondary metabolism in other lamentous fungi.


Background
Monascus azaphilone pigments (MonAzPs) are large group of secondary metabolites produced via polyketide biosynthesis mainly by Monascus spp. strains [1,2]. MonAzPs have been generally classi ed into yellow, orange, and red pigment subclasses on the basis of color [3,4]. MonAzPs have been extensively used as natural food coloring agents and food supplements for more than one billion people [5,6]. These pigments have also been found to exhibit a wide range of biological activities, such as anticancer, anti-in ammation, and anti-obesity [7][8][9].
The annual MonAzPs production has been estimated to exceed 20,000 metric tons in China alone [6]. Many approaches have been applied for improvement of MonAzPs production, such as mutation breeding, process control, genetic engineering and extraction fermentation [10][11][12][13][14].Among them, the highest MonAzPs production was achieved with 211 U/mL using a mutant M. purpureus strain M183 in liquid-state (submerged) fermentation [15]. However, these microbial productions of MonAzPs still do not fully meet the demand for cost-effective production.
The cyclic adenosine monophosphate (cAMP) is a second messenger in eukaryotic and prokaryotic cells [16]. cAMP was produced by adenylyl cyclase (AC) with ATP as substrate [17]. The concentrations of cAMP in microbial cells directly governed the activity of protein kinase A (PKA) and indirectly modulated metabolic and transcriptional processes ( Fig. 1) [18][19][20]. The cAMP-PKA pathway is important for various growth and developmental processes in different fungi [21]. Furthermore, cAMP signalling has been implicated in regulating secondary metabolisms in several fungi, including Fusarium graminearum and Aspergillus species [22,23]. It has been reported that MonAzPs production in M. ruber M7 was improved by addition of exogenous cAMP to the culture medium [24].
In cAMP metabolism, it is rapidly and continuously converted into adenosine 5′ monophosphate (5′-AMP) by cAMP phosphodiesterases (PDEs) [25]. Reduction or elimination of PDE activity could improve cAMP concentration, which led to regulation of secondary metabolites (SMs) biosynthesis [26,27]. PDEknockout mutant of Aspergillus avus CA14PTs exhibited 2-fold higher cAMP level and SM a atoxin production [28]. In F. graminearum, Pde2 was found to be the major PDE responsible for negative regulation of deoxynivalenol production [23]. On the basis of these literatures, it was speculated that knockout of PDE gene in Monascus spp. might improve MonAzPs production by increasing cAMP concentration (Fig. 1). However, the PDE gene and its regulation on cAMP in Monascus spp. is still unknown.
In this study, a novel PDE gene, mrPDE, from M. purpureus strain HJ11 was identi ed by in vitro reaction. Then, the mrPDE gene knockout strain was constructed by genetic engineering. The cAMP concentration of ΔmrPDE strain signi cantly increased, which resulted in a signi cant improvement of MonAzPs production. Further, a high MonAzPs production was achieved by fed-batch fermentation. This study could provide a new insight for development of e cient MonAzPs-producing strain.

Strains and plasmids
Monascus purpureus strain HJ11 was isolated from rice wine starter and preserved in our laboratory. Escherichia coli BL21 (DE3) was used as gene expression host. M. purpureus HJ11 strain was coincubated with Agrobacterium tumefaciens EHA105, containing recombinant pCAMBIA3300, to obtain deletion mutants. pCAMBIA3300 vector was used as deletion vector and pET30a(+) was used as heterologous gene expression in E. coli BL21 (DE3).

Flask cultivation
M. purpureus strains were pre-cultured in 10 mL of glucose mineral salt (GM) medium at 30 o C, 120 rpm for 5 days to obtain the seed culture (GM medium in g/L: glucose 30, (NH 4 ) 2 SO 4  Heterologous expression and puri cation of MrPDE RNA extraction and cDNA synthesis was performed according to our previous study [29]. The mrPDE gene fragment, Monpu1|448456|e_gw1.142.20.1 (ID number of the JGI database), was cloned from cDNA with speci cdesigned primers using PCR reaction. The mrPDE DNA fragment and vector pET30a(+) were digested by endonucleases EcoRI and HindIII, respectively. The recombinant vector pET30a(+)-mrPDE was constructed with T4 DNA ligase and used for expression of mrPDE. All sequences were con rmed by DNA sequencing. The recombinant vector pET30a(+)-mrPDE was transformed into E. coli BL21(DE3) for expression. The resultant E. coli strain was inoculated in 100 mL LB medium, containing 50 μg/mL kanamycin, in 500-mL ask and incubated at 37 o C and 200 rpm. Isopropyl-β-D-thiogalactopyranoside (IPTG) was added to the medium at a nal concentration of 0.5 mM, when the optical density of culture reached 0.6-1.0 at 600 nm. Then, the culture was incubated at 18 o C and 200 rpm for 16 h. The Histagged mrPDE protein was puri ed using Nickel-NTA Agarose (Qiagen, Valencia, CA, USA), as described in our previous study and was examined on 12% (w/v) SDS-PAGE gel [30]. Protein concentration was estimated by the Bradford method using bovine serum albumin as standard.

PDE activity detection
Phosphodiesterase activity was determined using a PDE activity assay kit (Colorimetric) following the manufacturer's instructions (Abcam, Cambridge, United Kingdom) with slight modi cation [31]. The PDE activities were determined in puri ed mrPDE and recombinant E. coli lysates. The lysates were obtained by disrupting the cells using sonication (W140D, Heat System-Ultrasonics, Inc., NY), followed by centrifugation at 12,000 × g for 10 min to remove cell debris. One unit of PDE activity for 3'5'-cAMP was de ned as the amount of enzyme required to release 1.0 μmol 5'-AMP from 3'5'-cAMP per minute at 30 o C.

PKA activity determination
Suspended cells were lysed by a high-pressure homogenizer. Cell lysates were incubated at 30 °C as speci ed by the protocol of PKA (Protein Kinase A) Colorimetric Activity Kit (Thermo Fisher Scienti c). The PKA kinase activity were measured by the colorimetric method at 450 nm on a spectrophotometer.
The PKA kinase activity was calculated using a calibration curve, which was constructed using ve PKA standard solutions.
qRT-PCR RNA extraction, cDNA synthesis, and qRT-PCR were performed as previous work described [29]. Brie y, total RNA was extracted from mycelial samples, and then checked for purity and integrity. cDNA synthesis was performed using the PrimeScript RT master mix (TaKaRa) according to the manufacturer instruction. qRT-PCR was conducted with a two-step thermal producure (step 1, 95°C for 10 s, and step 2, 40 cycles of 95°C for 3 s and 60°C for 25 s) on a 7500 Fast real-time PCR system (Applied Biosystems). The cycle number was used for the quantitation of the expression level. Relative expression level of target cDNA was obtained by the 2 -ΔΔCT method via normalization to β-actin (GenBank accession no. AJ417880).

cAMP concentration measurement
The cAMP concentration was determined according to the previous work with modi cation [32]. M. purpureus HJ11 strains were cultivated in GM medium at 30 o C. Fresh mycelia were harvested, frozen and lyophilized and ground into powder using liquid nitrogen. The powder sample was kept in chilled 6% (w/v) trichloro acetic acid (TCA) solution and incubated for for 20 min. After centrifugation of the mixture at 3500 × g at 4 o C for 20 min, the supernatant was extracted 3 times with 10 volumes of diethyl ether to remove TCA residues. The resulting extract was dried before further analysis. The cAMP levels were quanti ed using cAMP Enzyme Immunoassay Kit, Direct (Sigma-Aldrich, St. Louis, MO) following the manufacturer's instructions. In total, assay was repeated three times independently with three biological replicates for each strain. After 4 days of fermentation, a certain amount of feed medium was added to the 10-L stirred tank bioreactor to maintain glucose concentration at 30 g/L in the fermentation medium, and same amount of culture was withdrawn to analyze residual glucose concentration, biomass and MonAzPs concentration.
The fermentation was performed similar to the batch fermentation.

MonAzPs analysis
Intercellular MonAzPs concentration was estimated following our previous method [29]. In this study, the concentration of MonAzPs was measured using UV-1700 spectrophotometer (Shimadzu, Tokyo, Japan) at speci c wavelength of 410, 470, and 510 nm that corresponded to the characteristic absorbance of yellow, orange and red pigments, respectively. Total MonAzPs was calculated as the sum of yellow, orange, and red pigments.
Gene knockout in M. purpureus HJ11 Targeted gene knockout and complementation of mrPDE in M. purpureus HJ11 was performed as described in our previous study [29].

Data availability
Sequences of the genes mentioned in this article are available in GenBank [29].

Improvement in MonAzPs production and mrPDE expression by exogenous cAMP addition
The effect of exogenous cAMP on M. purpureus HJ11 was evaluated. The MonAzPs production was determined under different concentrations (0, 1.0, 2.0, 3.0 and 4.0 mM) of cAMP in GM medium. In the presence of exogenous cAMP, the strain showed higher MonAzPs yield (Fig. 2). At 2.0 mM of cAMP concentration, maximum MonAzPs yield of 6065 U/g DCW was achieved, compared to the yield of 2606 U/g DCW without cAMP ( Fig. 2A). This was consistent with a previous study, in which addition of 1.0 mM exogenous cAMP promoted MonAzPs yield in M. ruber strain M7 [24]. In Fusarium graminearum, mycotoxin deoxynivalenol production was approximately increased by 40-folds in cultures treated with 4 mM cAMP [23]. The yields of red, orange, and yellow MonAzPs were 1812 U/g DCW, 1787 U/g DCW and 2466 U/g DCW, respectively (Fig. 2B, 2C and 2D). Among them, maximum rate of increase (179%) was observed in the yield of yellow MonAzPs (Fig. 2D). These results indicated that the cAMP concentration played an important role in the growth and development of fungi and biosynthesis of secondary metabolites. Moreover, it was ascertained that the MonAzPs yield could be improved by increasing the concentration of cAMP.
When the concentration of cAMP was 2.0 mM, maximum dry cell weight (DCW) of 4.6 g/L was achieved, while maximum DCW without cAMP was only 4.2 g/L (Fig. 2E). Similar results were found in M. ruber strain M7 [24]. We evaluated the effects of exogenous cAMP on the transcription of genes relative in the cAMP-PKA pathway by quantitative reverse transcription-PCR (qRT-PCR) analysis. There were no marked changes in the expression levels of mrpigA and AC genes. The expression of global regulator gene laeA, which can regulate the MonAzPs biosynthetic gene cluster [14], was a little increased. The MonAzPs polyketide synthase gene mrpigA, which was responsible for MonAzPs precursor synthesis [29], showed an obvious increase in expression, explaining the improved production of MonAzPs. Unexpectedly, incubation with cAMP led to a strong induction of expression of a hypothetical PDE gene, named as mrPDE ( Fig. 2F and S1). So, we speculated that this gene was responsible for negative regulation of cAMP in M. purpureus HJ11.

Identi cation Of Mrpde As A Camp Phosphodiesterase
To the best of our knowledge, there was no report about PDE gene in Monascus spp. It was imperative to identify the PDE gene in M. purpureus HJ11. Comparison of the amino acid sequence of MrPDE with known sequences in the NCBI protein databases showed that a hypothetical PDE gene, XP_001264269.1 from A. scheri, displayed the highest similarity of 64% with MrPDE. To further analyze MrPDE, several known PDEs (4OJV_A from S. cerevisiae, Q5AGE4 from Candida albicans, P12019 from Dictyostelium discoideum, and P36599 from Schizosaccharomyces pombe) were selected. A multiple alignment of 6 sequences mentioned above was also performed. The alignment analysis revealed two highly conserved amino acid motifs, which contained catalytic residues (Fig. 3A). This result indicated that MrPDE may be a fungal PDE.
To verify the function of MrPDE, the enzymatic activity was evaluated through in vitro reaction. MrPDE was heterologously expressed in E. coli BL21(DE3) strain. After induction with 0.5 mM IPTG at 18 o C for 12 h, SDS-PAGE analysis showed that MrPDE was successfully expressed in E. coli under the control of the T7 promoter (Fig. 3B, S lane). The enzyme was puri ed using Ni a nity chromatography column ( Fig. 3B, P lane). The catalytic activity of MrPDE was determined with 3',5'-cAMP as substrate. MrPDE was found to e ciently catalyze the hydrolysis of 3',5'-cAMP (Fig. 3C). The speci c activity of puri ed MrPDE was 15.4 U/mg, which was close to that of PDE1 (20.5 U/mg) from S. cerevisiae S288C [33]. These results indicated that MrPDE was indeed a fungal PDE.

Activation of cAMP signalling pathway by mrPDE knockout
To construct a MonAzPs high-producing strain, a mrPDE gene knockout strain ΔmrPDE and a mrPDE complemented knockout strain CΔmrPDE were successfully engineered through homologous recombination technology. The colony diameter of ΔmrPDE was a little bigger than those of WT and CΔmrPDE, which indicated that the growth of ΔmrPDE strain was not inhibited by mrPDE gene knockout (Fig. 4A). Meanwhile, the morphology of ΔmrPDE colonies grown on GM plate showed more intense color than those of WT and CΔmrPDE, implying that the ΔmrPDE strain might have a high production on MonAzPs (Fig. 4A).
The cAMP concentration was determined in each strain at different times. Notably, the cAMP concentration in ΔmrPDE strain was signi cantly higher than those of WT and CΔmrPDE strains, attaining a maximum of 11145 pmol/g on sixth day (Fig. 4B). The cAMP concentration in WT strain increased to 3219 pmol/g at the fourth day before gradually decreasing (Fig. 4B). Similar cAMP trend was observed in CΔmrPDE strain. This result suggested that knockout of mrPDE could inhibit the cAMP degradation, which led to a higher intracellular cAMP concentration. It has been reported that deletion of PDE gene in Ustilaginoidea virens and Magnaporthe oryzae both resulted in double-fold increase in cAMP concentration [21,34]. Thus, it was ascertained that deletion of PDE gene could e ciently increase cAMP concentration in M. purpureus HJ11.
In cAMP signalling pathway, PKA activity is essential for regulation of primary and secondary metabolism [35,36]. Our data showed that knockout of mrPDE could induce PKA kinase activity (Fig. 4C), suggesting the increase of cAMP concentration triggered PKA activation [37]. The DCW of ΔmrPDE strain reached 5.1 g/L, which was higher than those of WT (4.2 g/L) and CΔmrPDE strains (4.2 g/L) (Fig. 4D). These results indicated that the cAMP signalling pathway was activated by knockout of mrPDE.

Improvement of MonAzPs yield in Δ mrPDE strain
For further con rmation of the MonAzPs production in ΔmrPDE strain, shake-ask fermentation was performed. After fermentation for 10 days, the MonAzPs yield of ΔmrPDE strain reached 8563 U/g DCW, which was 2.3-fold higher than that of WT strain (Fig. 5). The yields of red, orange, and yellow pigments of ΔmrPDE strain were 2377 U/g DCW, 2245 U/g DCW, and 3941 U/g DCW, respectively (Fig. 5B, 5C and 5D). These yields were 1.58-times, 1.80-times, and 3.46-times higher than those of WT strain, respectively. Similar study has been reported in F. graminearum, the knockout of PDE gene pde1 resulted in increased production of secondary metabolite deoxynivalenol [23]. The deletion of PDE gene pdeH from A. avus led to an increased production of a atoxin to 110 mg/mL from 48 mg/mL [38].
To explain the reason of increasing MonAzPs yield, the expression of each genes from Monascus azaphilone pigments biosynthetic gene cluster (MPBGC) was determined by quantitative reverse transcription-PCR (qRT-PCR) analysis. All MPBGC genes were expressed in the WT and ΔmrPDE strains. Unexpectedly, as shown by the ΔmrPDE strain, transcriptional activation of large parts of the genes in the MABGC was achieved, apart from mrpigH, mrpigI, mrpigL and mrpigP, which didn't participate in MonAzP biosynthesis (Fig. S2) [29,39]. The downstream targets of PKA include transcriptional regulators and other effectors to control gene transcriptional expression [40]. These suggested that PKA, activated by cAMP, improved the transcriptional expression of MPBGC genes through indirect activation [41].
It was worth noting that the increased proportion of yellow MonAzPs yield was signi cantly higher than that of total MonAzPs yield in ΔmrPDE strain (Figs. 2 and 5). In our previous study, orange MonAzPs were found to be converted into yellow MonAzPs in the presence of adequate NADPH [42]. So, it was speculated that the increased cAMP concentration led to higher rate of NADPH/NADP + . There is little knowledge about the in uence of cAMP on NADPH/NADP + rate in microorganism. Herein, it was found that the ratio in ΔmrPDE strain was 0.91, which are much higher than those of WT and CΔmrPDE strains (0.55 and 0.54, respectively) (Fig. S3). This might be the main reason for the signi cantly higher yield of yellow MonAzPs than those of red and orange MonAzPs.

High-density Fermentation For High Monazps Production
In order to enhance the production of MonAzPs, batch fermentation of ΔmrPDE strain was performed in a medium, containing initial glucose concentration of 100 g/L. After 10 days of cultivation, total MonAzPs production reached a maximum of 158.9 U/mL (Fig. 6A), and corresponding total MonAzPs yield was 6782 U/g (Fig. 6B). However, this yield was lower than that in shake ask (8563 U/g). This might be attributed to the insu cient glucose supply (< 5 g/L) in the middle and later stages of fermentation, where mycelia accumulated the main part of MonAzPs (Fig. 6A).
For improvement of glucose supply, fed-batch fermentation was also performed in a 10-L stirred fermenter [43]. The ΔmrPDE strain displayed rapid growth, and DCW reached 38.1 g/L, which was signi cantly higher than that of batch fermentation (Fig. 6C). In another study, fed-batch fermentation of Monascus anka strain GIM 3.592 achieved a DCW of 39.77 g/L [44]. Subsequently, a signi cant increase in MonAzPs production was observed, reaching a maximum of 332.1 U/mL. The total MonAzPs yield reached 8739 U/g DCW, slightly higher than that in shake ask fermentation (Fig. 6D). It has been reported that a mutant M. purpureus strain M183 produced MonAzPs at 211.6 U/mL [15]. To the best of our knowledge, this study achieved highest MonAzPs yield and production.

Conclusion
In summary, mrPDE was identi ed as a PDE from M. purpureus HJ11. Subsequently, knockout of the mrPDE gene was performed to enhance cAMP concentration. The MonAzPs yield in ΔmrPDE strain was achieved at 8563 U/g, which was 2.3-folds higher than that of WT strain. In order to improve the MonAzPs production, the fed-batch fermentation was performed in a 10-L stirred fermenter. The MonAzPs production was signi cantly enhanced to 332.1 U/mL, with the MonAzPs yield of 8739 U/g DCW. This study describes a promising method for high production of MonAzPs and provides a strategy for metabolic engineering of secondary metabolism in other lamentous fungi.

Declarations
Ethics approval and consent to participate Not applicable Consent for publication All authors have approved publication of the manuscript.

Availability of data and material
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.     Comparison of MonAzPs production between the WT, ΔmrPDE, and CΔmrPDE strains. Time courses of total MonAzPs (A), red MonAzPs (B), orange MonAzPs (C), yellow MonAzPs (D) production of WT, ΔmrPDE and CΔmrPDE strains, respectively. The strains were cultivated in GM medium at 30 oC for 10 days. The means and standard deviations calculated from measurements from three biological replicates for each strain were shown.