A Novel Endo-Polygalacturonase from Penicillium oxalicum: Gene Cloning, Heterologous Expression and Its Use in Acidic Fruit Juice Extraction

An endo-polygalacturonase (endo-PGase) exhibiting excellent performance during acidic fruit juice production would be highly attractive to the fruit juice industry. However, candidate endo-PGases for this purpose have rarely been reported. In this study, we expressed a gene from Penicillium oxalicum in Pichia pastoris. The recombinant enzyme PoxaEnPG28C had an optimal enzyme activity at pH 4.5 and 45°C and was stable at pH 3.0–6.5 and < 45°C. The enzyme had a specific activity of 4,377.65 ± 55.37 U/mg towards polygalacturonic acid, and the Km and Vmax values of PoxaEnPG28C were calculated as 1.64 g/l and 6127.45 U/mg, respectively. PoxaEnPG28C increased the light transmittance of orange, lemon, strawberry and hawthorn juice by 13.9 ± 0.3%, 29.4 ± 3.8%, 95.7 ± 10.2% and 79.8 ± 1.7%, respectively; it reduced the viscosity of the same juices by 25.7 ± 1.6%, 52.0 ± 4.5%, 48.2 ± 0.7% and 80.5 ± 2.3%, respectively, and it increased the yield of the juices by 24.5 ± 0.7%, 12.7 ± 2.2%, 48.5 ± 4.2% and 104.5 ± 6.4%, respectively. Thus, PoxaEnPG28C could be considered an excellent candidate enzyme for acidic fruit juice production. Remarkably, fruit juice production using hawthorn as an material was reported for the first time.


Protein Purification and Determination
The recombinant protein PoxaEnPG28C in the culture broth was concentrated and applied to size exclusion chromatography using a HiLoad 16/600 Superdex 75 column (GE Co., Ltd., Sweden). Sodium dodecyl sulphate polyacrylamide gel electrophoresis [SDS-PAGE, 5% (w/v) stacking gel and 10% (w/v) separating gel] was used to determine the purity and molecular weight of the enzyme [13], and the Bradford method with bovine serum albumin protein as the standard was used to determine the protein concentration [14]. The purified enzyme was desalted by dialyzing against ultrapure water, then biochemically characterized and used for fruit juice extraction.

Enzyme Assay
The enzyme activity of PoxaEnPG28C was determined by the same method as used for the previously reported enzymes PoxaEnPG28A [12] and PoxaEnPG28B [12], and the concentration of the released reducing sugar was calculated using a standard curve constructed with D-galacturonic acid as the standard. One unit of enzyme activity was defined as the liberation of 1 μmol of reducing sugar from the enzyme-catalyzed reaction system in 1 min.

Mode of Action of PoxaEnPG28C
The purified and desalted PoxaEnPG28C was incubated with 0.5% PGA at 45°C for 0.25, 0.5, 1, 2, 8, 12, 18, and 24 h, and then the liberated products were analyzed by thin-layer chromatography according to the previously reported method [11].

Effects of pH and Temperature on the Enzyme Activity and Stability
To determine the optimal pH, the enzyme activity of PoxaEnPG28C was determined at different pH values and at 50°C; for the optimal temperature, the enzyme activity of PoxaEnPG28C was determined at different temperatures (at the optimal pH). The highest enzyme activity was defined as 100%, and other enzyme activities were calculated as relative values.
For pH stability, 0.1 mL of PoxaEnPG28C was mixed with 0.9 mL of buffers with different pH values (0.1 M citric acid-Na 2 HPO 4 for pH 3.0-7.0 and 0.1 M Tris-HCl for pH 7.0-9.0) at 4°C for 24 h, and then the residual activities were determined. For the thermal stability, PoxaEnPG28C was incubated at 45, 50, 55, and 60°C for 15, 30, 45, and 60 min, and the residual enzyme activities were measured. The enzyme activity of un-incubated enzyme was defined as 100%, and the residual enzyme activities were calculated as relative values.

Substrate Specificity and Michaelis-Menten Constant of the Enzyme
For substrate specificity, the enzyme activities of PoxaEnPG28C against 0.5% (w/v) PGA, citrus pectin and apple pectin were determined under optimal reaction conditions (pH 4.5 and 45°C). The enzyme activity against PGA was defined as 100%, and other enzyme activities against other substrates were calculated as relative values. For the Michaelis-Menten constant, enzyme activities of PoxaEnPG28C against 0.2-2.0 g/l PGA were determined under the optimal reaction conditions (pH 4.5 and 45°C). The K m and V max values were then obtained by calculation using the Lineweaver-Burk plotting method.

Effects of Metal Ions on Enzyme Activity
The enzyme activities of PoxaEnPG28C were determined in the presence of each of the following metal ions: 2 mM Na + , K + , Mg 2+ , Ca 2+ , Mn 2+ , Ba 2+ , Co 2+ , Zn 2+ , Fe 2+ , or Cu 2+ . The enzyme activity without extra additive was defined as 100%, and other enzyme activities were calculated as relative values.

Use of PoxaEnPG28C for Acidic Fruit Juice Extraction
Fresh oranges, lemons, strawberries and hawthorn were purchased from the local market. Oranges and lemons were peeled, and the hawthorns were cored. Fresh fruits were then mixed with ultrapure water and pulped. Twenty-five grams of pulp were treated with 0 or 100 U of PoxaEnPG28C per gram of pulp (at pH 4.5 and 45°C) for 2.5 h, and then centrifuged in a tubular filter (400 mesh) at 500 rpm for 2 min. The filtered juice was centrifuged at 5,000 ×g for 10 min, and the parameters of the supernatant were determined according to previously reported methods [12]. The parameters of the juices without addition of PoxaEnPG28C were set as the control values (100%).

Sequence Analysis and Nucleotide Sequence Accession Number
The SignalP 4.1 online server (http://www.cbs.dtu.dk/services/SignalP/) and SMART online server (http:// smart.embl-heidelberg.de/) were used to predict the signal peptide and catalytic domain, respectively. ExPASy online server (http://web.expasy.org/protparam/) was used to calculate the hypothetical molecular weight of the protein. DNAMAN 6.0 software was used to carry out the multiple sequence alignment. The nucleotide sequence of the gene encoding PoxaEnPG28C has been deposited in the GenBank database under the Accession No. MZ614864.

Gene Cloning, Heterologous Expression and Identification of the Recombinant Protein
After prediction using the SMART online tool, a putative signal peptide including the initial 20 amino acids and a GH 28 catalytic domain were found in the putative protein EPS32977 of P. oxalicum 114-2 [10]. The amino acid sequence encoded by the amplified cDNA from P. oxalicum CZ1028 showed 100% identity with the putative protein EPS32977 of P. oxalicum 114-2, and showed 78.35% identity with a functional identified endo-PGase from Penicillium janthinellum IF0 7719 [15]. In comparison with a functionally mutated endo-PGase [16], Asp 169 , Asp 190 , Asp 191 , and His 212 were expected to be the catalytic residues of the amplified putative endo-PGase, and Arg 245 and Lys 247 of the putative endo-PGase were expected to be involved in the binding of substrate. Four putative disulfide bonds were formed by Cys 18 and Cys 33 , Cys 192 and Cys 208 , Cys 318 and Cys 323 , and Cys 342 and Cys 351 of the amplified putative endo-PGase (Fig. 1). Compared with two previously reported endo-PGases from P. oxalicum CZ1028 (PoxaEnPG28A [12] and PoxaEnPG28B-Pp [9]), PoxaEnPG28C had a closer phylogenetic relationship with endo-PGases from Aspergillus niger (RePgaA from A. niger JL-15 [17] and endo-PgaA from A. niger SC323 [18]), emphasizing the diversity of the genome of P. oxalicum (Fig. 2).
After 4 days of cultivation, the enzymatic activity of the pectinase in the culture broth of the recombinant P. pastoris reached 1057.98 ± 28.31 U/ml, which was higher than the activities of endo-PGI produced by P. pastoris (6.2 U/ml) [19], endo-PGA1 produced by P. pastoris (50 U/ml) [20], PG1 produced by Saccharomyces cerevisiae (50 U/ml) [21], and PG7fn produced by P. pastoris (678.1 U/ml) [22]. After purification, a protein showing a sharp band at about 36 kDa on the SDS-PAGE gel was obtained (Fig. 3), and the apparent molecular weight matched the theoretical molecular weight of 35.87 kDa.
At the early stage of hydrolysis, galacturonic acid, digalacturonic acid and trigalacturonic acid appeared among the products (Fig. 4). Tetramers and pentamers of galacturonic acid appeared later, but as the hydrolysis time became prolonged, the amounts of tetramers and pentamers of galacturonic acid decreased. Meanwhile, the amount of galacturonic acid did not increase during the hydrolysis process, indicating the bond connecting the terminal galacturonic acid was not the prime target of the recombinant enzyme. The variations in the amounts of mono-and oligo-galacturonic acid in the hydrolysis product confirmed an endo-acting pattern of the recombinant enzyme [3]. Because the enzyme was the third reported endo-PGase from P. oxalicum [12], it was named as PoxaEnPG28C.

Substrate Specificity and Michaelis-Menten Constant of the Enzyme
PoxaEnPG28C showed higher enzyme activity towards PGA (4,377.65 ± 55.37 U/mg, 100.00 ± 1.26%) than towards citrus pectin (1,321.32 ± 43.13 U/mg, 30.18 ± 0.99%) and apple pectin (704.39 ± 51.61 U/mg, 16.09 ± 1.18%), indicating that its hydrolysis activity was hindered by the methylation of the backbone [12]. As the concentration of PGA increased, the specific enzyme activity of PoxaEnPG28C increased and then remained constant (Fig. S1A). Moreover, the K m and V max values were calculated to be 1.64 g/l and 6127.45 U/mg, respectively (Fig. S1B). The K m of PoxaEnPG28C was in the range of previously reported K m values of endo-PGases (from 0.32 g/l to 19.5 g/l) [12]. The V max value of PoxaEnPG28C was higher than those of some reported   endo-PGases, 55.55 U/mg [33], 103.58 U/mg [25] and 240 U/mg [34], but lower than that of PoxaEnPG28B-Pp (77,882.2 U/mg) [12].

Use of PoxaEnPG28C for Acidic Fruit Juice Extraction
The pH of oranges (Citrus reticulata Blanco) used in this study (pH 4.44) was different from that of a previous report of pH 3.5 [23]. PoxaEnPG28C reduced the viscosity by 25.7 ± 1.6%, increased the light transmittance by 13.9 ± 0.3% and increased the yield by 24.5 ± 0.7% (Table 1). In a previous report, a purified endo-PGase PG1 also increased the yield of citrus juice from 2.5 ml to 3.75 ml [35]. Lemon [Citrus limon (L.) Burm. f.] belongs to the same genus as orange, but it is more acidic than orange. The pH of lemons used in this study (pH 2.58) was lower than that of a previous report of 3.50 [23]. PoxaEnPG28C reduced the viscosity by 52.0 ± 4.5%, increased the light transmittance by 29.4 ± 3.8% and increased the yield by 12.7 ± 2.2% (Table 1). In previous reports, endo-PGases PG2 and PgaB increased the light transmittance of lemon juice by 43% [23] and about by 40% [25], respectively. However, some other important improvements in the juice such as a reduction in viscosity and an incremental yield of the juice were not mentioned in those reports [9].  The pH of strawberries (Fragaria vesca) used in this study (pH 3.41) was close to that of a report of 3.9 [36]. PoxaEnPG28C reduced the viscosity by 48.2 ± 0.7%, increased the light transmittance by 95.7 ± 10.2% and increased the yield by 48.5 ± 4.2% (Table 1). The performance of PoxaEnPG28C was better than those of other endo-PGases including NfPGI (reduction of viscosity: 32.4%; increment of light transmittance: 28.72%; increment of yield: 6%) and NfPGII (reduction of viscosity: 20.6%; increment of light transmittance: 46.41%; increment of yield: 9%) [36]. Moreover, the enzyme dosage quantity of PoxaEnPG28C was 100 U/g juice pulp, which was lower than the dosages of 7,500 U/ml (for NfPG I) and 3,000 U/ml (for NfPGII) /g juice pulp (the density of strawberry pulp was 1.02 g/ml) [36].
Hawthorn fruit (Crataegus pinnatifida Bunge) is high in nutritional value; it contains (100 g fresh weight basis) 23.688 g citric acid, 18.378 g fructose, 13.893 g glucose and 9.418 mg vitamin C, and has a unique taste [37]. However, it is seldom used to produce juice. PoxaEnPG28C reduced the viscosity of hawthorn juice by 80.5 ± 2.3%, increased the light transmittance by 79.8 ± 1.7% and increased the yield by 104.5 ± 6.4% (Table 1). The excellent performance of PoxaEnPG28C might aid the development of hawthorn juice as a commercial product.
Endo-Pgase should function well in several types of fruit juice since the varied hydrolysis environment of pectin is a challenge for the enzyme [7]. The pH values of fruit pulp used in this study were different from previous reports, which might be due to the differences in the cultivars and the maturity of the fruits [7]. However, the differences in catalytic environment in the different fruit juices go well beyond pH and include different cell-wall compositions and polymer structures, which can have a major influence on the catalytic efficiency of endo-Pgase. The degradation of natural pectin in fruits is much more difficult than the degradation of purified pectin because natural pectin takes part in and is embedded in the network of the plant cell wall [2]. The structure of plant cell wall and catalytic environments of juices differ from those of fruits [38] and are dependent on the ripening periods of the fruit [39]. Thus, compared with other endo-Pgases which only function well in one kind of fruit juice, the performance of PoxaEnPG28C in four acidic fruit juice preparations emphasizes the considerable application potential of this enzyme ( Table 2).
In this study, a gene encoding a putative endo-polygalacturonase was cloned from P. oxalicum and expressed in P. pastoris, and the recombinant protein PoxaEnPG28C was purified and identified as an endo-polygalacturonase. The enzymatic characteristics of PoxaEnPG28C were determined, and it was found to be an acidic enzyme. Remarkably, PoxaEnPG28C functioned well in orange, lemon, strawberry and hawthorn juice production. Thus, PoxaEnPG28C could be considered an excellent candidate enzyme for acidic fruit juice production.

Acknowledgments
This work was financially supported by the Middle-aged and Young Teachers' Basic Ability Promotion Project of Guangxi (2019KY0924), the Doctoral Scientific Research Fund of Nanning University (2021DSRF01), and the Science Foundation of Nanning University (2018XJ42).