Polygalacturonase gene pgxB in Aspergillus niger is a virulence factor in apple fruit

Aspergillus niger, a saprophytic fungus, is widely distributed in soil, air and cereals, and can cause postharvest diseases in fruit. Polygalacturonase (PG) is one of the main enzymes in fungal pathogens to degrade plant cell wall. To evaluate whether the deletion of an exo-polygalacturonase gene pgxB would influence fungal pathogenicity to fruit, pgxB gene was deleted in Aspergillus niger MA 70.15 (wild type) via homologous recombination. The ΔpgxB mutant showed similar growth behavior compared with the wild type. Pectin medium induced significant higher expression of all pectinase genes in both wild type and ΔpgxB in comparison to potato dextrose agar medium. However, the ΔpgxB mutant was less virulent on apple fruits as the necrosis diameter caused by ΔpgxB mutant was significantly smaller than that of wild type. Results of quantitive-PCR showed that, in the process of infection in apple fruit, gene expressions of polygalacturonase genes pgaI, pgaII, pgaA, pgaC, pgaD and pgaE were enhanced in ΔpgxB mutant in comparison to wild type. These results prove that, despite the increased gene expression of other polygalacturonase genes in ΔpgxB mutant, the lack of pgxB gene significantly reduced the virulence of A. niger on apple fruit, suggesting that pgxB plays an important role in the infection process on the apple fruit.


Introduction
Pectinases are the most important pathogenic factor in plant pathogenic bacteria and fungi [1][2][3][4]. They are responsible for pathogens to decompose pectin in plant cell wall. Pectin hydrolysis not only weakens the cell wall to facilitate penetration and colonization of the host, it also provides the fungus carbon sources for its growth [5]. Pectinases are consisted of polygalacturonase (PG), pectin lyase (PNL), pectate lyase (PL), pectinesterase (PE), pectin methyl esterase (PME). Polygalacturonase is one of the major members of pectinases which cleaves α-1,4-glycosidic of D-galacturonic acid in pectin and it is classified into endo-and exopolygalacturonase on the basis of the way of eliminating galacturonic acid [6,7]. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 The production of PG by pathogenic fungi is critical for their success and survival during host infection [8]. It has been confirmed that the loss of a polygalacturonase gene in some fungi would result in decreased pathogenicity. Shieh et al [9] showed that a polygalacturonase gene is related to the infection of Aspergillus flavus in cotton bolls. The disruption of endopolygalacturonase gene Bcpg1 or Bcpg2 in Botrytis cinerea reduced its virulence on different hosts [10]. It is also reported that the loss of pectin methyl esterase gene Bcpme1 reduced virulence on apples, and pectin lyase pelB was an important virulence factor in Colletotrichum gloeosporioides when attacking avocado [11,12]. PG is also required for infection by Phytophthora capsici and Alternaria citri [13,14]. However there are also studies demonstrated that disruption of some polygalacturonase genes in fungi did not directly affect virulence, for example, deletion of PG1, PG5, PGX4 in Fusarium oxysporum led to no virulence difference in tomato [15,16] and endopolygalacturonase PGN1 is not required for pathogenicity of Cochliobolus carbonum on maize [17]. Mutants lacking both polygalacturonase genes cppg1/cppg2 in Claviceps purpurea did not affected vegetative properties, but they are nearly nonpathogenic on rye [18].
Aspergillus niger is a saprophytic fungus which degrades plant cell wall polysaccharides and leads to the decay of fruits and vegetables [19]. Since the 1990s, technical advances in molecular biology speed up the operation mechanism of A. niger and A. niger gene sequencing has been completed [20,21]. Deletion of kusA gene in A. niger dramatically improved homologous integration efficiency and facilitated gene knockout in A. niger [22]. However, whether polygalacturonase contributes to the pathogenicity of A. niger on fruit is still unclear. Here we constructed pgxB deletion mutant in A. niger MA 70.15 via homologous recombination and the pathogenicity were evaluated in ΔpgxB strain.

Fungal strains and growth conditions
A. niger MA 70.15 (ΔpyrG, pyrG encodes orotidine-5-phosphate decarboxylase, cell lacking this enzyme cannot grow without exogenous uridine, but can resist toxicity of 5-Fluoroorotic acid) was used as wild type strain in this work. A. niger was grown on potato dextrose agar medium (PDA) (per liter: 200 g of potato; 20 g of agar; 20 g of glucose and 10 mM uridine) at 28˚C. Therefore all medium used in this study were supplemented with 10 mM uridine except where stated. For in vitro growth evaluation, A. niger spores were suspended in sterile water and adjusted to 10 6 spores per mL. To test whether the absence of pgxB affected mycelial growth, PDA and pectolytic enzyme-inducing medium (PEIM) (per liter: 20 g of agar; 20 g of pectin; MM medium (5 g of KNO 3 [23] were spot-inoculated with 5 μL spore suspension of wild type and ΔpgxB mutant. All strains were inoculated onto three plates of each medium and colony diameter was measured daily.
For the determination of growth curve, 1 mL spore suspension of wild type and ΔpgxB mutant were inoculated in 100 mL Erlenmeyer flasks containing 30 mL potato dextrose medium or PEIM at 30˚C, 150 rpm. Mycelium was harvested on a quantitative filter paper at the time of 12, 24, 36, 48, 60, 72, 84, 96 h and weighed.

Vector construction and transformation
The exo-polygalacturonase gene pgxB in A. niger was deleted following gene knockout method of Delmas et al [24]. Upstream and downstream DNA fragments AB (548 bp) and CD (564 bp) flanking pgxB gene were amplified by polymerase chain reaction (PCR) from A. niger MA 70.15 genomic DNA. Primers were designed using the genome database A. niger CBS 513.88 and the upstream and downstream fragments contained a common HindIII restriction site to ligate them together and EcoRI and XhoI restriction sites were used for cloning the joined fragment into the plasmid pC3 [24] to create pC3-An_ΔpgxB integrative plasmid. Primers are shown in Table 1.
A. niger protoplast preparation and transformation were carried out by the method of Baltz et al [23]. 4-day-old mycelia grown from PDA were harvested and were digested with 0.4 g Lyticase (Sigma) to obtain fungal protoplasts. Protoplasts were transformed with 10 μg pC3-An_ΔpgxB (un-linearized) in 50 μL polyethylene glycol 6000. Transformations were inoculated on upper layer of transformation medium without uridine (per liter: upper medium: MM; 6 g of agar; 0.95 M sucrose; lower medium: MM; 12 g of agar; 0.95 M sucrose) to select for the integration of the plasmid carrying pyrG on the chromosome. Transformants were purified by breeding them twice successively on the same transformation medium but lacking sorbitol (per liter: MM, 20 g of agar). Transformants were then propagated twice on PDA medium containing 10 mM uridine to release the selective pressure on the integrated plasmid. For selecting clones that had excised the plasmid (ΔpyrG), spores were then spread on MM-Uri-5-FOA medium (per liter: MM; 20 g of agar; 10 g of glucose; 1.6 mM uridine; 750 μg/mL 5-fluoro-orotic acid). Clones from last medium were cultured in MM-Uri-5-FOA liquid medium at 250 rpm at 30˚C for 3 d, and mycelia were harvested for genomic DNA extraction.

DNA extraction and PCR confirmation of ΔpgxB strain
Genomic DNA was extracted using Master Pure Yeast DNA Purification Kit (Epicentre). Primers pgxB-A, pgxB-D and another primers internal to the pgxB gene pgxB-E1, pgxB-E2 (PCR product size: 953 bp) were designed using Primer Premier 5.0 software. Primers are shown in Table 1.

Determination of PG activity
For determination of polygalacturonase activity, 100 μL of conidia (1×10 6 spores per mL) were inoculated into a 100 mL Erlenmeyer flask containing 30 mL of liquid PEIM and cultured at 30˚C for 5 days. Culture medium after suction filtration was used for PG activity assay following the method described by Miller [25]. Reaction mixture was consisted of 1 mL of 50 mM acetic acid-sodium acetate at pH5.5, 0.5 mL of 10 g/L pectin solution and 0.5 mL of crude enzyme or enzyme boiled for 5 min followed by incubation at 40˚C for 30 min. Reactions were terminated by adding 1.5 mL of DNS (3,5-dinitrosalicylic acid) followed by a 5 min incubation in a boiling water bath. The reaction mixture was cooled to room temperature, and distilled water was added to a final volume of 25 mL. Absorbance at 540 nm was measured. One unit of PG was defined as 1 μg of galacturonic acid produced by pectin per hour and expressed as U/mL enzyme extract.
PG activity was also determined by plate assay. The wild type and ΔpgxB mutant were inoculated on PEIM and cultured for 3 days at 30˚C. Thereafter the colonies were rinsed off the plates with distilled water before staining the plates with 0.05% ruthenium red. Pectinase production was evaluated by ratio of diameter of clear zone formed around colonies relative to diameter of mycelia [26,27].

Virulence assay of A. niger on apple fruit
Apple fruits were washed with tap water and then surface-sterilized with 75% ethanol. Five different sites on the surface of apples were wounded (2 mm diameter and 5 mm deep) and injected with 5 μL wild type or ΔpgxB mutant spore suspension (10 6 spores per mL) respectively. After air-drying, three replicates were put in sealed container with H 2 O at the bottom at 25˚C. The necrosis diameter were measured daily after inoculation. Spores grown from wounds were used for RNA extraction and quantitative PCR.

RNA extraction and quantitative PCR
RNA was extracted from frozen mycelium grown on PDA, PEIM or apples ground in liquid nitrogen with TRNzol RNA Reagent kit (Tiangen). After DNase treatment, cDNA were obtained according to the manufacturer's instruction of PrimeScript RT Master Mix (TaKaRa, Japan). Primers used for quantitative PCR are shown in Table 1. Relative quantification was processed using the method of Delta-Ct.

Statistical analysis
Statistical significance was tested by one-way analysis of variance (ANOVA), and the results are expressed as the mean values ± standard deviation (SD) of three independent experiments. Fisher's least significant differences (LSD) were calculated following a significant (P < 0.01 or P < 0.05) t test.

Results
Gene disruption of pgxB in A. niger 548-bp upstream and 564-bp downstream DNA fragments AB and CD were amplified by PCR from A. niger MA 70.15 genomic DNA. Fragment ABCD (1094 bp) by the ligation of AB and CD were cloned into plasmid pC3 to generate the recombinant plasmid pC3-An_ΔpgxB (6959 bp). Then the plasmid pC3-An_ΔpgxB was transformed and integrated into A. niger MA 70.15 protoplast and the ΔpgxB mutant was confirmed by genomic PCR. As shown in Fig 1, genomic DNA was used as template for PCR confirmation with primers pgxB-A and pgxB-D which flank pgxB gene and primers on the ORF of pgxB gene pgxB-E1 and pgxB-E2. 1100 bp band for ΔpgxB and 3500 bp for the wild type were obtained by the primers pgxB-A and pgxB-D, and a 953-bp band for the wild type and no amplification for ΔpgxB with the primers pgxB-E1 and pgxB-E2. All of these results confirmed that the pgxB gene was disrupted in ΔpgxB.

Growth analysis of ΔpgxB mutant strain
Pectinases in some fungi can be induced by pectin [28]. To evaluate whether the lack of pgxB would affect its growth on pectin medium, we compared growth of ΔpgxB and A. niger MA 70.15 on PDA (no pectin) and pectin medium (PEIM). The radial growth was measured on solid media. It was found that growth rate, estimated as colony diameter, showed no difference between the ΔpgxB mutant and wild type on PDA (Fig 2A and 2B)  and 2D). Besides, no significant difference in growth curve was found between ΔpgxB mutant and the wild type in liquid medium with or without pectin (Fig 2E and 2F).

Inducing effect of pectin on the expression of pectinase genes in A. niger
To study whether the expression of pectinase genes were induced by pectin, relative expression of different pectinase genes in ΔpgxB and A. niger MA 70.15 on PDA and pectin medium (PEIM) were determined by quantitative PCR. As shown in Fig 3A, expression of pgxB gene in wild type was significantly enhanced by pectin medium, while the expression was not detected in ΔpgxB, further confirming that pgxB gene was deleted in the mutant. For wild type, most of pectinase genes like PG genes pgaII, pgaB, pgaD, pgaE, pgaX, pgxA, pgxC, PL genes pelA, pelC, pelD, pelF, plyA, PME gene An04g09690, PE gene An02g12505 showed enhanced expression on PEIM than those on PDA, suggesting that the expression of pectinases genes were induced by pectin (Fig 3B and 3C). Similarly, enhanced gene expression of PG genes pgaII, pgaA, pgaB, pgaD, pgaE, pgaX, pgxA, pgxC, PL genes pelA, pelC, pelD, pelF, plyA, PME gene An04g09690, PE gene An02g12505 was also observed in ΔpgxB mutant (Fig 3D and 3E).

Polygalacturonase activity analysis
pgxB gene is a member of polygalacturonase family found in A. niger. To study whether pgxB gene actually contributes to the whole activity of polygalacturonase in A. niger, polygalacturonase activity was determined in ΔpgxB mutant. The ΔpgxB mutant and A. niger MA 70.15 were grown in liquid PEIM for 3 days and the activity of secreted polygalacturonase in the medium was analyzed. As shown in Fig 4A, ΔpgxB exhibited lower PG activity (by 5.8%) compared with the wild type. Secreted PG activity by A. niger was also assayed on plate. As shown in Fig 4B and  4C, the ratio of diameter of clear zone (Dc) which showing the degradation of pectin by secreted PG and mycelia (Dm) produced by ΔpgxB was smaller (by 6.5%) than that of wild type, indicating that ΔpgxB mutant produced less pectinase than the wild type A. niger MA 70.15.

Pathogenicity assay on apple fruit
Whether pectinase was a pathogenic factor in A. niger infection on fruit is still unknown, thus we studied the virulence of ΔpgxB mutant on fruit. Apple fruits were inoculated with conidial suspension of ΔpgxB and the wild type A. niger MA 70.15. Lesion development was monitored daily and the diameter was measured. As shown in Fig 5A and 5B, pgxB disruption resulted in a reduction of decay development as the lesion diameter caused by ΔpgxB was about 20% smaller than that of wild type on 4, 5, 6, 7 days and 15% smaller on 10, 11 days, suggesting that the virulence of ΔpgxB on apple fruit was significantly lower than that of wild type.
To study why the virulence was decreased in ΔpgxB, expression of various pectinase genes in ΔpgxB and wild type in the process of infecting apples was determined by quantitative PCR. Expression of PG genes pgaI, pgaII, pgaA, pgaC, pgaD, pgaE, pgaX in ΔpgxB were higher in ΔpgxB than those in wild type, while there was no significant difference in expression of PG genes pgaB, pgxA, pgxC, PL genes pelA, pelB, pelC, pelD, pelF, plyA and PME genes pmeA, An04g09690 between ΔpgxB mutant and wild type (Fig 5C and 5D). The increased expression of PG genes in ΔpgxB suggested a possible compensation effect in pgxB deletion mutant.
In this paper, we described the construction of a mutant disrupted in the pgxB gene in A. niger. ΔpgxB exhibited no growth rate reduction on PDA and pectin medium compared with wild type, which means the disruption of pgxB did not weaken its ability of decomposing pectin as carbon source (Fig 2). In contrast, Bcpme1 mutant in B. cinerea and pelB mutant in Colletotrichum gloeosporioides showed reduced growth on pectin medium [11,12]. Pectinase was induced by pectin, polygalacturonic acid or galacturonic acid and was repressed by glucose and polygalacturonase-inhibiting protein (PGIP) [37,38]. Quantitative PCR results showed that expression of most of pectinase genes such as PG genes pgaII, pgaA, pgaB, pgaD, pgaE, pgaX, pgxA, pgxC, PL genes pelA, pelC, pelD, pelF, plyA were up-regulated on pectin medium  15 and ΔpgxB mutant on PDA and PEIM; B and C, relative expression of pgaI, pgaII, pgaA, pgaB, pgaC, pgaD, pgaE, pgaX, pgxA, pgxC, pelA, pelB, pelC, pelD, pelF, pmeA, An04g09690, plyA and An02g12505 in A. niger MA 70.15 on PDA and PEIM medium respectively for 4 days; C and D, relative expression of pgaI, pgaII, pgaA, pgaB, pgaC, pgaD, pgaE, pgaX, pgxA, pgxC, pelA, pelB, pelC, pelD, pelF, pmeA, An04g09690, plyA and An02g12505 in ΔpgxB mutant on PDA and PEIM medium respectively for 4 days. * and ** in this figure and following ones stand for a significant difference between two data at P < 0.05 and P < 0.01, respectively. compared that on PDA (Fig 3), confirming that the expression of pectinase genes were induced by pectin. We also found that ΔpgxB mutant grown in liquid and solid PEIM showed significant lower PG activity than the wild-type strain, suggesting that the loss of pgxB reduced the production of PG (Fig 4).
Virulence assay on apple fruit showed that the deletion of pgxB gene has a profound effect on lesion development in the infection of apple as lesion diameter caused by ΔpgxB was smaller than that of wild type. A similar reduction in maceration ability has been observed with pectinase-deficient mutants of phytopathogenic bacteria such as Erwinia, Pseudomonas and Ralstonia species [39,40]. Besides, Oeser et al. [18] found that mutant in both cppg1 and cppg2 are nearly non-pathogenic on rye using a gene-replacement approach. However, pectinases are usually encoded by multiple genes, thus mutation in one pectinase gene might not affect the virulence on fruits [15][16][17]. The disruption of either the pelA or pelD gene in Nectria hematococca alone causes no detectable decrease in virulence, whereas disruption of both pelA and pelD drastically reduces virulence [41]. In order to understand the decreasing virulence of ΔpgxB mutant, expression of some pectinase genes were assayed, and we found that PG genes pgaI, pgaII, pgaA, pgaC, pgaD, pgaE and pgaX were expressed moore highly in ΔpgxB mutant than in the wild type. Deletion of one gene in a gene family might result in higher expression of other genes with the same function as polygalacturonases of A. niger are encoded by a family of diverged genes [42]. A similar phenomenon that disruption of serine proteinase caused an increase in metalloproteinase has also been found in Aspergillus flavus [43]. Even so, the lack of pgxB still dramatically reduced lesion diameter on apples. Our results demonstrate that pgxB is a virulence factor which partially contributes to the virulence of A. niger on apple fruit, thus highlighting the need for further research to elucidate the roles of other pectinase genes in A. niger.