β-Glucan Synthase Gene Overexpression and β-Glucans Overproduction in Pleurotus ostreatus Using Promoter Swapping

Mushroom β-glucans are potent immunological stimulators in medicine, but their productivities are very low. In this study, we successfully improved its production by promoter engineering in Pleurotus ostreatus. The promoter for β-1,3-glucan synthase gene (GLS) was replaced by the promoter of glyceraldehyde-3-phosphate dehydrogenase gene of Aspergillus nidulans. The homologous recombination fragment for swapping GLS promoter comprised five segments, which were fused by two rounds of combined touchdown PCR and overlap extension PCR (TD-OE PCR), and was introduced into P. ostreatus through PEG/CaCl2-mediated protoplast transformation. The transformants exhibited one to three fold higher transcription of GLS gene and produced 32% to 131% higher yield of β-glucans than the wild type. The polysaccharide yields had a significant positive correlation to the GLS gene expression. The infrared spectra of the polysaccharides all displayed the typical absorption peaks of β-glucans. This is the first report of successful swapping of promoters in filamentous fungi.


Introduction
Mushroom b-glucans are the major structural constituents of the mushroom cell wall. They also provide antioxidant activity [1,2] and can be used as immunological stimulators in medicine, such as antitumor, immunomodulating, antioxidant, radical scavenging, cardiovascular, antihypercholesterolemia, antiviral, antibacterial, antiparasitic, antifungal, detoxification, hepatoprotective, and antidiabetic effects [3]. Therefore, there has been growing popularity in developing mushroom b-glucans as drugs or dietary supplements and scientifically investigating their functions [4]. However, controlled by metabolic regulation, the production of b-glucans in mushroom is low. Furthermore, due to enzymatic degradation by glucanase activity during storage, the production from fruiting body or submerged fermentation liquid is only 20-50 mg/100 g dry matter of fruiting body [5] or 0.15-5.3 g/l ferment broth [6,7]. Accordingly, it is of great value to improve b-glucans productivity in mushrooms by genetic engineering.
Most mushroom b-glucans with immunological stimulation are b-(1R6)-branched b-(1R3)-linked [8], and are synthesized by b-1,3-glucan synthase (GLS) (UDP-glucose 1,3-b-D-glucan 3-b-Dglucosyl transferase, EC 2.4. 1.34). The GLS complex is composed of a catalytic subunit FKS and a regulatory subunit RHO. FKS is activated by RHO, a G protein, through GTP-dephosphorylation, and synthesizes the polymer of glucose monomers. RHO activity is regulated by a wall GDP-GTP exchange factor protein [9]. In fungi genomes, FKS and RHO genes are highly conserved and usually have only one or two copies [10,11]. The information about the transcriptional regulation of GLS has been scant [11].
In yeast, GLS activity, both at transcriptional and enzymatic level, is stimulated by stress-inducing compounds present in the media [12]. A similar mechanism is also present in Lentinula edodes [13][14][15]. Nevertheless, the expression of the GLS genes in Ustilago maydis, a fungus causing smut disease on maize, was constitutive during its infection in maize and in response to ionic and osmotic stress [11].
Oyster mushroom Pleurotus ostreatus is a widely cultivated edible and medicinal mushroom in China and East Asia due to its short growth time, high adaptability and productivity. Its b-glucans demonstrated efficacy in promoting the survival of mice susceptible to bacterial infections [16], high SOD-like activity and antitumor activity [17]. In this study, we obtained the sequences of its only GLS gene and the promoter from P. ostreatus PC15 v2.0 in JGI Genome Portal (http://genome.jgi-psf.org). Using homologous recombination, the GLS gene promoter of P. ostreatus was replaced by the promoter of glyceraldehyde-3-phosphate dehydrogenase (gpd) gene of Aspergillus nidulans; the transformants displayed high expression of GLS gene and high production of bglucans.

Strains and DNAs
Pleurotus ostreatus TD300, often used as a commercial cultivation strain in China, was obtained from Zhengzhou Composite Experiment station, China Edible Fungi Research System (Zhengzhou, China), and cultivated on PDA medium at 28uC for six days as described elsewhere [18].
The 1,015 bp UH was the upstream homologous sequence which matched to the 39 end partial sequence preceding the GLS promoter of P. ostreatus. It was cloned from the genomic DNA of P. ostreatus by PCR using primers UH-F and UF-R (Table 1), and its GeneBank accession number was JX889617. The primers were designed based on the genome sequence of P. ostreatus PC15 v2.0 in JGI Genome Portal (http://genome.jgi-psf.org). In the primer UF-R sequence, the last 15 nts (highlighted with underline) were complementary to the 59 end sequence of P gpd 1035 .
The 1,035 bp P gpd 1035 served as the upstream FLP recognition target (FRT) sequence; it was the partial 59 end sequence of P gpd , the promoter of gpd gene in Aspergillus nidulans. P gpd 1035 was generated from plasmid PAN7-1 by PCR using primers P gpd 1048 -F and P gpd 1048 -R ( Table 1). The first 15 nts of P gpd 1048 -F and last 16 nts (with underline) of P gpd 1048 -R were complementary to the 39 end sequence of UH and 59 end sequence of hph, respectively. The plasmid pAN7-1 (kindly provided by Prof. van den Hondel, Leiden University, Netherlands) contains the hygromycin B resistance gene of Escherichia coli controlled by the gpd promoter and the transcription termination signal of tryptophan synthetase gene (trpC) from A. nidulans [19].
The 2,822 bp hph was hygromycin B resistance gene of E. coli expression cassette, and was amplified from plasmid PAN7-1 by PCR using primers hph-F and hph-R ( Table 1). The first 16 nts of hph-F and last 14 nts (with underline) of hph-R were complementary to the 39 end sequence of P gpd 1035 and 59 end sequence of P gpd , respectively. The 2,206 bp P gpd was also amplified from plasmid PAN7-1 by PCR using primers P gpd -F and P gpd -R ( Table 1). The first 14 nts of P gpd -F and last 17 nts (with underline) of P gpd -R were complementary to the 39-end sequence of hph and 59-end sequence of GLS 1025 , respectively.
The 1,025 bp GLS 1025 was the downstream homologous sequence in the 59 end partial sequence of P. ostreatus GLS. It was cloned from the genomic DNA of P. ostreatus by PCR using primers GLS 1025 -F and GLS 1025 -R (Table 1), and its accession number in GeneBank was JX889617. The primers were designed   (Fig. 2).

Touchdown-overlap extension PCR
Two rounds of combined touchdown PCR and overlap extension PCR (TD-OE PCR) were performed to fuse the above five long DNA segments.
The first round of TD-OE PCR was carried out for the fusion of the three segments: hph, P gpd , and GLS 1025 ; it produced a long fusion segment hPG. This round includes two steps. In step I, 47 mL reaction solution contains 1 mL of each DNA segment at 0.5 mM, 5 mL of 106 PCR buffer, 8.0 mL of 2.5 mM of dNTP, and 0.5 mL of 5 U/mL LA Taq. Amplification started at 94uC for 40 sec; the annealing temperature of the reaction decreased from 61.5uC to a touchdown 57.5uC at the cooling rate of 0.5uC every cycle, followed by five cycles at 57.5uC, 4 min at 68uC, and 10 min at 72uC. In step II, the reaction solution from Step I was added to 1.0 mL of 0.1 mM hph-F and GLS 1025 -R separately, and 0.5 mL LATaq. PCR conditions are similar to that of step I: amplification: 94uC for 40 sec; annealation: 60uC to 56uC decreasing by 0.5uC per cycle, then 20 cycles at 56uC; 7 min at 68uC, and 10 min at 72uC. After completion, 5 mL of the PCR reaction aliquots were analyzed on 1% agarose gels stained with ethidium bromide.
The second round of TD-OE PCR was carried out to fuse the three segments, i.e. UH, P gpd 1035 , and hPG, to generate the homologous recombination fragment. The PCR procedure was similar to that in the first round except for the annealing temperature and primers. In step I, the annealing temperature decreased from 62uC to a touchdown 58uC; in step II, added primers were UH-F and GLS 1025 -R, the annealing temperature decreased from 58uC to 56uC. Protoplasts preparation and PEG/CaCl 2 -mediated transformation of P. ostreatus TD 300 were performed as described previously [20,21]. The introduced foreign DNA was the homologous recombination fragment for swapping GLS promoter. Protoplasts of P. ostreatus were suspended in MTC buffer at 10 6 protoplasts/mL, and 100 mL suspensions were mixed with 10 mg the DNA fragment. Transformants were subcultured on PDA with or without hygromycin B. A full description of this method is given in the Extended Methods S1. To verify the replacement of GLS promoter by the introduced fragment, two PCR reactions were performed using the genomic DNA of the transformants as template, primers hph-1 and hph-2 for the amplification of the hph gene, and primers P gpd -1 and GLS-1 ( Table 1) for amplifying the combination of P gpd and GLS. Both PCR conditions were 94uC for 1 min, 56uC for 1 min, and 72uC for 1 min in 30 cycles. 5 mL of the PCR reaction aliquots were analyzed on 1% agarose gels stained with ethidium bromide.

Semi-quantitative RT-PCR
To analyze the GLS expression in transformants, semi-quantitative RT-PCR was carried out as described elsewhere [21] with slight modifications. Total RNA was extracted from transformants. Primers were GLS-F and GLS-R for reverse transcription and amplification of GLS, and AC-1 and AC-2 for reverse transcription and amplification of housekeeping gene b-actin. Reverse transcription of the mRNA was carried out at 42uC for 60 mins by using MMLV Reverse Transcriptase 1st-Strand cDNA Synthesis Kit (Epicentre, Madison, WI, USA). The PCR conditions were 94uC for 40 sec, 49uC for 40 sec, and 72uC for 1 min in 30 cycles. 5 mL of the PCR reaction aliquots was analyzed on 1% agarose gels stained with ethidium bromide. The electrophoresis bands of RT-PCR reaction were photographed and the density of each band was quantified using image analysis software, UVI band V. 97 (UVI Tech, Cambridge, UK).

Analysis of b-glucan content and infrared spectrum
To measure the b-glucan yield of the transformants, the transformants were cultivated in PD broth (150 mL in a 500 mL flask) at 25uC for 12 d under 150 rpm shaking. The culture broth was extracted in 98uC water bath for 3 h and then filtered; the supernatant was concentrated 10-fold via evaporation. Four volume 95% ethanol was added and placed overnight, centrifuged at 5,000?g for 15 min. The precipitate was washed 2 times with 85% ethanol, then dissolved in hot water, and de-proteinized by Sevag method [22]. Equal volume of Sevag reagent (chloroform/butanol 4:1, v/v) was added, vigorously shaken for 30 min, the mixture was centrifuged at 5,000?g for 10 min. The upper layer was separated and Sevag reagent was again added. This process was repeated 3 times. The polysaccharide content was measured using phenol-sulfuric acid method [23]. The polysaccharide characteristic was determined by infrared spectroscopy (Tensor 37, Bruker, Ettlingen, Germany).

Statistical analysis
Statistical analysis was performed using standard ANOVA techniques. Comparisons among the means of three independent experiments were made using the least significant difference (LSD) test.

Construction of the homologous recombination fragment for swapping GLS promoter
The three segments of hph, P gpd , and GLS 1025 were fused successfully, and the expected fusion product hPG (5,873 bp) was produced by only one round TD-OE PCR (Fig. 3A); hPG was subsequently fused with the two upstream segments UH and P gpd 1035 in the second round of TD-OE PCR (Fig. 3B), and produced the homologous recombination fragment for swapping GLS promoter (8,103 bp). The sequencing result of the fusion product confirmed that all of the five segments were fused correctly in accordance with the design order.

Identification of homologous recombination transformant by PCR
The integration and the promoter swapping in the transformants were verified by PCR. Six transformants, A1, A4, A9, A15, A17, and A21, were randomly selected. The expected PCR amplifications were obtained from all the DNA samples of the second generation transformants using the primer pair hph-1 and hph-2, but not from the untransformed original strain (wild type) and the next generation transformants (Fig. 4A). The segment could be amplified from the DNA samples of the transformants from generation three to five by using the primer pair P gpd -1 and GLS-1, which was 1,320 bp in length and spanned the 39 end of P gpd and the downstream of the 59 end of GLS 1025 (Fig. 4B). The result revealed that hph gene in the homologous recombination fragment was deleted by the homologous recombination between P gpd 1035 and P gpd from the third generation of the transformants. In addition, the introduced P gpd replaced the GLS promoter and remained genetically stable.

GLS gene overexpression and b-glucan overproduction of the transformants
To determine whether GLS gene over-expressed after the swap of its promoter by P gpd , its mRNA expression level in the transformants was measured by semi-quantitative RT-PCR and was found to be two to four folds higher than that of wild type (Fig. 5).
The polysaccharide yields of the six transformants were determined and they were 32% to 131% higher than that of wild type (Fig. 6); including the wild type, the yield had positive correlation to the GLS gene expression (p,0.05). The result indicated that the GLS gene expression may correspond with its enzymatic activity and protein level. The infrared spectra (IR) of the polysaccharides produced by the six transformants and wild type had the same characteristic with close resemblance to each other as shown in Fig. 7. The absorption peaks at 3,457 cm 21 were ascribed to hydroxyl stretching vibration of the polysaccharide; peaks at 3,151 cm 21 were CH (CH3, CH2, or CH) stretching vibration; those at 1,626 cm 21 attributed to C = O asymmetric stretching vibration [24]; those at 1,398 cm 21 were due to C-H variable-angle vibration; peaks at 1,143 to 1,088 cm 21 were characteristics of C-O-C asymmetric vibration of pyranose ring [25]; finally the bands at 856 cm 21 mapped to b-glucan [26], suggesting that bglycosidic bond was present in all samples.

Discussion
Promoter engineering is the principal strategy for metabolic engineering. It employs mutagenic endogenous promoters or introduced heterologous promoters to increase the expression of key pathway genes and maximize target production, and has been successfully applied to bacteria and yeast [27][28][29][30]. Nevertheless, the application of promoter engineering in filamentous fungi has not been reported. In this study, we introduced a heterologous promoter to increase the GLS expression and significantly improved b-glucan production.
In previous studies, the promoters for replacing native promoters could be constitutive or inducible [28]. Compared with constitutive promoters, inducible promoters tightly control their downstream gene to achieve high level expression, maximize protein production and reduce toxicity during growth phase [31]; but it is limited in practice due to inducer cost and cell hypersensitivity to inducer concentration [32]. P gpd is a constitutive promoter used across fungal species and provides high levels of constitutive gene expression [33,34]; for example, the expression of genes under the control of P gpd was significantly higher than that by the commonly used alcohol oxidase 1 promoter (P AOX1 ) in methanol-grown cells of Pichia pastoris [35]. In this study, we employed P gpd from A. nidulans to swap the native promoter of GLS in P. ostreatus, and consequently GLS expression was improved by up to two folds and b-glucan production increased by up to 32% compared to the wild type strain. CaMV 35S is another constitutive promoter often used in filamentous fungi, which is considerably weaker than P gpd in P. ostreatus [21]. The other constitutive promoters used in filamentous fungi included adhA, gdhA, oliC, pgkA, etc [36], but the comparison among them has not been reported.
Construction of the homologous recombination fragment for swapping promoter requires five to six segments fusion [29,37]. Multiple segments fusion is usually performed by overlap extension PCR (OE-PCR) [38,39]. However, OE-PCR cannot fuse more than two DNA segments simultaneously [40]. Several modified OE-PCR procedures can fuse multiple segments at the same time, but require chimeric primers and high and close annealing temperatures in order to minimize mispriming [41,42]. Touchdown PCR (TD-PCR) is an efficacious solution to reduce mispriming and rapidly optimize PCR to increase specificity, sensitivity, and yield [43,44]. In this study, we combined TD-PCR and OE-PCR and used only two rounds to fuse the five segments. Our technique produced the homologous recombination fragment for swapping GLS promoter without sedulously adjusting the annealing temperature of primers. It showed that touchdown-overlap extension PCR (TD-OE PCR) was a fast and highly efficient method for promoter swapping and metabolic engineering.

Conclusion
By our knowledge, this is the first report of successful swapping of promoters using TD-OE PCR in filamentous fungi through Figure 4. PCR for the identification of the recombinant hph and P gpd sequences in the transformants of P. ostreatus. PCR amplification on total DNA from the second generation transformants using primers hph-1 and hph-2 which defined a 750 bp sequence across the hph gene (A). PCR amplification on total DNA from the fifth generation transformants using primers P gpd -1 and GLS-1 which defined a 1,320 bp sequence spanning the P gpd and GLS gene located far from GLS 1025 (B). Lane 1: WT; Lane 2-7: transformant A1, A4, A9, A15, A17, and A21. doi:10.1371/journal.pone.0061693.g004  rapid construction of homologous recombination fragment. The polysaccharide yields of the transformants were 32% to 131% higher than that of wild type, and had significantly positive correlation to the GLS gene expression levels. TD-OE PCR, a novel procedure combining touchdown and overlap extension PCR, was carried out for the fusion of five segments to construct the homologous recombination fragment for swapping GLS promoter. Our study supports that TD-OE PCR was a fast and highly efficient method for promoter swapping and metabolic engineering.

Supporting Information
Methods S1 Extended Methods. (DOC)