Efficient biotransformation for preparation of pharmaceutically active ginsenoside Compound K by Penicillium oxalicum sp. 68

Abstract

Pathogenic fungus Penicillium oxalicum sp. 68 was screened from soil and identified by ITS sequencing. The strain was found to be able to transform protopanaxadiol-type ginsenosides to produce a series of bioactive metabolites. Glycosidase from the culture of P. oxalicum sp. 68 was partially purified with a simple two-step procedure consisting of DEAE-cellulose chromatography and ammonium sulfate precipitation. Bioactive ginsenoside Compound K was prepared selectively and efficiently by biotransformation of ginsenosides Rb1, Rb2, Rc and Rd using the partially purified glycosidase. The optimal conditions for transforming Rb1 into Compound K were pH 4.0, 55 °C and 0.5 mg mL⁻¹ Rb1. The sole product is Compound K and the maximum yield reached 87.7 % (molar ratio). The transformation pathways of Rb1, Rb2, Rc and Rd are Rb1→Rd→F2→Compound K, Rb2→CO→CY→Compound K, Rc→Mb→Mc→Compound K and Rd→F2→Compound K, respectively. This biotransformation method showed great potential for preparing minor bioactive ginsenosides, especially Compound K, in the pharmaceutical industry because of its high specificity and favorable environmental compatibility.

Introduction

Biotransformation is a powerful tool in the pharmaceutical industry. Many natural products have been transformed via biotransformation into pharmaceutically active compounds (Muffler et al. 2011). Ginsenosides are the major active components of ginseng. The major protopanaxadiol-type saponins, Rb1, Rb2, Rc and Rd, are abundant in ginseng but are absorbed poorly in the intestine after oral administration (Tawab et al. 2003; Han and Fang 2006). These compounds are metabolized initially to Compound K by human intestinal bacteria after oral administration. Compound K has been proven to be the major form of protopanaxadiol-type saponins absorbed in the intestines (Akao et al. 1998a; Hasegawa 2004; Lei et al. 2007; Ruan et al. 2009). Recent experimental results have demonstrated several remarkable biological properties of Compound K, such as anti-tumor, anti-allergic and anti-inflammatory activities (Lee et al. 2005; Zhou et al. 2006; Choi et al. 2007; Zhou et al. 2009). Therefore, Compound K has a potentially broad application in medicine.

However, Compound K is rarely naturally present in ginseng. Based on current technology, it is impossible to synthesize from simple starting materials. Some transformations have been carried out to produce Compound K from structurally similar compounds by chemical methods (Han et al. 1982; Chen et al. 1987). Application of acid or alkaline chemical hydrolysis of the high content ginsenoside Rb1 to produce Compound K is limited not only by undesirable side reactions, but also by the resulting considerable environmental pollution.

Biotransformation is thought to be more promising for the preparation of Compound K because of its high specificity, mild conditions and low pollution potential. In previous studies, some microorganisms, such as Paecilomyces Bainier sp. 229 (Zhou et al. 2008; Yan et al. 2010), Fusarium sacchari (Han et al. 2007), Bifidobacterium sp. Int57, Bifidobacterium sp. SJ32, Aspergillus niger (Chi et al. 2005), and Caulobacter leidyia (Cheng et al. 2006), have been used to transform major protopanaxadiol-type saponins into Compound K. However, the yields of these microbial conversions were relatively low, the transformation procedures were quite long, and the enzymes involved in the transformation remain unknown. Several enzymes, such as lactase from Penicillium sp., β-galactosidase from Aspergillus oryzae (Ko et al. 2007) and β-glycosidase from Sulfolobus solfataricus (Noh et al. 2009), have been used to produce Compound K from protopanaxadiol-type saponins. However, the purification process of these enzymes and their specificity for this transformation process still limit their large-scale industrial application. Therefore, compared with microbes or purified enzyme, transformation by partially purified enzyme would be potentially preferable due to easier preparation, higher transformation efficiency and simpler procedure for product separation. In this study, we report a simple procedure for Compound K preparation by biotransformation of Rb1 using a partially purified enzyme preparation from the phytopathogenic fungus Penicillium oxalicum sp. 68.

Materials and methods

Materials

Standard ginsenosides were purchased from Chengdu Mansite Biotechnology (Chengdu, China). The mixture of ginsenosides Rb1, Rb2 and Rc and their pure individual compounds were prepared from Chinese white ginseng roots (5 years old, cultivated in Fusong, Jilin Province of China) as previously described (Cheng et al. 2006) and identified by HPLC and ¹³C-NMR spectrometry. p-Nitrophenyl-β-d-glucopyranoside (pNPG) was purchased from Sigma (St. Louis, MO). DEAE-cellulose gel was provided by Shanghai Hengxin (Shanghai, China). All other reagents were of analytical or HPLC grade.

Analytical methods

TLC was carried out on a silica gel G60 plate (3.0 cm × 5.0 cm) using a mixture of chloroform: methanol: water (65: 35: 10, v/v/v, lower phase) as the developing solvent. After developing, the plates were stained with 5 % (v/v) sulfuric acid in ethanol and then heated at 110 °C for 5 min. Under the above conditions, the relative mobility rates (R _f) of standard Rb1, Rb2/Rc, Rd, F2 and Compound K were assigned at 0.2, 0.3, 0.46, 0.70 and 0.88, respectively. The transformed products were identified primarily by comparing the R _f values on the TLC plate to those of the standards.

HPLC was carried out using a Shimadzu HPLC system composed of two LC-10AT pumps, a UV-visible detector and a data-processing system (Zhejiang University, N2000). An analytic Shim-pack PREP-ODS (H) Kit column (4.6 mm × 250 mm, 5 μm) was used to analyze the hydrolyzed products and to calculate the yield of Compound K. The following gradient programs were used: 0–12 min, 37.5 % acetonitrile (in distilled water, v/v); 12–32 min, 37.5 %–70 % acetonitrile; and 32–40 min, 70 % acetonitrile. The flow rate was 1 mL min⁻¹, and the elution process was monitored at 203 nm.

Microorganism cultivation and identification

The filamentous fungus, sp. 68, was collected from the soil in the Changbai Mountain in China and maintained on V8-juice agar medium (per liter: 200 mL V8 juice, 2 g CaCO₃ and 15 g agar). The strain was identified by 18S rDNA and ITS sequencing using the PCR method (Lindsley et al. 2001). NS1 (5′-GTAGTCATATGCTTGTCTC-3′) and NS8 (5′-TCCGCAGGTTCACCTACGGA-3′) were used for 18S rDNA amplification, and the ITS sequence was amplified using the primers ITS1 (5′-GGAAGTAAAAGTCGTAACAAGG-3′) and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′). Sequencing of 18S rDNA and ITS were performed using an ABI PRISM^TM 3730XL DNA Analyzer and an ABI PRISM^TM 377XL DNA Sequencer from Takara Biotechnology (Dalian, China). The 18S rDNA and ITS rDNA sequences were compared to database sequences retrieved from GenBank.

Transformation of the ginsenoside mixture by Penicillium oxalicum

Strain P. oxalicum was first cultured on V8-juice agar medium for 8 days, after which the mature spores were transferred into V8-juice liquid medium (per liter: 200 mL V8 juice and 2 g CaCO₃). After being cultured at 28 °C and 130 rpm for 48 h, the spores were washed twice with 25 mM acetate buffer (pH 5.0, buffer A) to remove residual medium. The remaining cells were used for biotransformation.

To test the hydrolysis of major protopanaxadiol-type saponins by P. oxalicum, the spores were re-suspended in buffer A to a final concentration of 5 × 10⁶ spores mL⁻¹ and then shaken at 28 °C and 130 rpm for 24 h. Afterwards, the ginsenoside mixture of Rb1, Rb2 and Rc (filtered through a 0.22 μm filter) was added to the pre-incubated spores to reach a final concentration of 0.25 mg mL⁻¹, and the culture was incubated for 10 days. Aliquots were withdrawn every day, and the products were extracted by n-butanol. The n-butanol phase was concentrated to dryness under vacuum and analyzed by TLC.

Partial purification of the enzymes from P. oxalicum sp. 68

The strain P. oxalicum was cultured in V8 juice liquid medium at 28 °C and 130 rpm. During the fermentation process, the β-glucosidase activity of the fermentation liquor was measured every 12 h using pNPG as a substrate. The reaction mixture containing 2 mM pNPG, 25 mM acetate buffer (pH 6.0) and appropriate enzyme solution was incubated at 37 °C for 30 min. The reaction was then terminated by adding 0.25 M NaOH and the absorbance was read at 405 nm. One unit (U) β-glucosidase activity was defined as the amount of enzyme liberating 1 μmol p-nitrophenyl per minute under assay conditions. When β-glucosidase activity reached maximum, the fermentation liquor was filtered through absorbent gauze and then centrifuged at 10,000 g at 4 °C for 15 min. The supernatant was used to prepare the partially purified enzymes.

The supernatant was first purified by DEAE-cellulose chromatography. Briefly, 100 mL (bed volume) of DEAE-cellulose was added to approximately 4 L supernatant. The gel slurry was rotated overnight at 4 °C and subsequently applied onto an empty column (2.5 cm × 25 cm). The unbound materials flowed through the column during the column-packing process. The bound materials were eluted stepwise with 200 mL portions of 0, 0.25, 0.50, 0.75 and 1.0 M NaCl in buffer A at a flow rate of 4 mL min⁻¹. The active fractions eluted by 0.25 M NaCl were combined and precipitated with ammonium sulfate (30–90 %) at 0 °C. After being centrifuged at 10,000 g at 4 °C for 20 min, the precipitate was re-dissolved in buffer A and then used in the biotransformation.

Transformation of ginsenosides by partially purified enzymes

The mixture of ginsenoside Rb1, Rb2 and Rc was dissolved in buffer A, final concentration 0.25 mg mL⁻¹ and incubated with the partially purified enzymes (0.25 U mL⁻¹, 5 U) at 37 °C. The ginsenoside Rb1, Rb2, Rc or Rd were dissolved singly in buffer A, final concentration: 0.25 mg mL⁻¹ and incubated with the partially purified enzymes (0.25 U mL⁻¹, 5 U) at 37 °C. Aliquots were withdrawn at different times and the products were extracted with n-butanol. The n-butanol phase was concentrated to dryness under vacuum and analyzed by HPLC. The ginsenosides without enzyme were used as control.

Preparation and identification of transformed products

To prepare intermediate and final transformed products of Rb1, Rb2, Rc and Rd, each ginsenoside (100 mg, 0.5 mg mL⁻¹) was incubated with the partially purified enzymes (0.25 U mL⁻¹, 50 U) at pH 5.0 and 37 °C for 24 h. The incubation culture was extracted twice with n-butanol. The n-butanol phases were combined, concentrated to dryness under vacuum to obtain the mixture of transformed products. The products were separated by preparative HPLC same as in analytical method, with a preparative Shim-pack PREP-ODS (H) Kit column (20 mm × 250 mm, 5 μm), flow rate 5 mL min⁻¹. The purified products were identified by ¹³C NMR, carried out on a Bruker Av 600 NMR spectrometer at 150 MHz using MeOD as solvent.

Optimization of the biotransformation for Compound K production

The effects of pH, temperature and substrate concentration on preparation of Compound K by biotransformation of ginsenosides using partially purified enzymes were evaluated. The ginsenosides Rb1, Rb2, Rc and Rd (final concentration: 0.25 mg mL⁻¹) were as substrate individually and incubated with the partially purified enzymes (0.25 U mL⁻¹, 20 U) at 37 °C and pH 5.0 for 2 h. Ginsenoside Rb1 was chosen as the substrate in the subsequent optimization. The effect of pH on Compound K production was measured by incubating Rb1 (final concentration: 0.25 mg mL⁻¹) with the enzymes (0.25 U mL⁻¹, 20 U) at 37 °C for 2 h in Na₂HPO₄-citrate buffer at different pH values (25 mM, pH 2.0–8.0). To optimize the temperature, ginsenoside Rb1 (final concentration: 0.25 mg mL⁻¹) was incubated with the enzymes (0.25 U mL⁻¹, 20 U) in 25 mM Na₂HPO₄-citrate buffer (pH 4.0, buffer B) for 2 h at different temperatures ranging from 25 °C to 80 °C. The concentration of ginsenoside Rb1 for Compound K production was optimized by incubating different concentrations of ginsenoside Rb1 (0.1, 0.25, 0.5, 0.75 and 1 mg mL⁻¹) with the enzymes (0.25 U mL⁻¹, 20 U) in buffer B at 55 °C for 2 h. In these experiments, ginsenoside substrates without the enzyme were used as controls under the corresponding conditions. The reactions were stopped by the addition of n-butanol. The n-butanol phases were concentrated to dryness under vacuum and analyzed by analytical HPLC. Compound K was quantified according to a standard curve, which was linear throughout the integration area between 0.05 and 2.0 mg mL⁻¹ $\left( {Y = 3.01 \times {{10}^6}X + 239520} \right.$, R ² = 0.982; in which X = concentration of Compound K, mg mL⁻¹; Y = peak area). The highest yield of Compound K in each experiment was defined as 100 %.

Results and discussion

Identification of fungus sp. 68 and biotransformation of ppd-type saponins

More than 50 pathogenic fungi were isolated from the soil from Changbai Mountain where ginseng is cultivated, and screened for their transformation of ginsenosides. Of these, fungus sp. 68 was found to be able to transform the major protopanaxadiol-type saponins into Compound K selectively and efficiently. Strain sp. 68 was identified by comparing its 18S rDNA and ITS sequences with those published in the GenBank database. The 18S rDNA sequence of strain sp. 68 (accession number: GU078431) appeared homologous to Penicillium sp. and the ITS sequence (accession number: GU078430) appeared highly homologous to Penicillium oxalicum (accession number: GQ376104, FJ977097). Therefore, the strain sp. 68 was identified as P. oxalicum.

Penicillium oxalicum is a pathogenic fungus that can infect sorghum to cause sorghum glume blight and infect maize ears to cause rotting. Phytopathogenic fungi are very abundant in nature. They usually secrete extracellular glycosidases to digest the defensive glycosides produced by plant hosts. Therefore, phytopathogenic fungi are very promising sources of glycosidase for transforming the major ginsenosides into highly active metabolites. In our previous studies, we have tested more than 50 kinds of fungi for their ability to transform the major protopanaxadiol-type saponins (Zhao et al. 2009). The results show that more than 90 % of the tested strains could transform major protopanaxadiol-type saponins into different products. The fungus P. oxalicum was found to transform major protopanaxadiol-type saponins Rb1, Rb2 and Rc to Compound K, which has remarkable pharmacological activities.

As shown in Fig. 1a, P. oxalicum sp. 68 transformed the mixture of Rb1, Rb₂ and Rc to produce three primary metabolites: Rd, F2 and Compound K. The fungus transformed Rb1 more efficiently than Rb2 and Rc. Rb1 was transformed significantly after a 2-day incubation with the fungus, and was totally converted into the products after 8 days, whereas Rb2 and Rc still remained after 8 days incubation. After 2 days, the desired products Rd and F2 were observed, and, after 4 days the third product, Compound K, was observed. Rd and F2 are precursors of Compound K. Therefore, it was deduced that the final biotransformation product should be Compound K. Penicillium oxalicum sp. 68 could be a potential resource for preparing the enzymes that can transform the major protopanaxadiol-type saponins into bioactive Compound K.

Fig. 1

Biotransformation of protopanaxadiol-type saponins. a By fungus sp. 68. b By enzymes partially purified from strain sp. 68. Products were identified by a TLC or b HPLC by comparison to standard ginsenosides

Enzyme preparation

To improve transformation efficiency, the partially purified enzymes secreted by P. oxalicum sp. 68 were prepared as described in Materials and methods. Fermentation was stopped after 108 h, when glycosidase activity reached a maximum (Fig. 2a). The culture filtrates were collected and applied to DEAE-cellulose so that the enzymes were concentrated on the DEAE-cellulose matrix. A major active glycosidase peak, eluted with 0.25 M NaCl, was observed in the elution profile (Fig. 2b). The collected active fraction was precipitated with 30 % ammonium sulfate to remove impurities and subsequently with 90 % ammonium sulfate to obtain the glycosidases. The specific activity of the glycosidases was 64.5 U mg ⁻¹ using pNPG as substrate to measure the enzyme activity and the Bradford method to determine the protein concentration (Table 1). Based on the activity of the collected enzymes, the yield of glycosidases was 79.1 % with respect to the fermentation culture.

Fig. 2

Enzymatic activity in fermentation and partial purification of the enzymes. a Enzymatic activity curve in the fermentation of Penicillium oxalicum sp. 68; activity was measured using p-nitrophenyl-β-d-glucopyranoside (pNPG) as substrate. b DEAE-Cellulose chromatogram profile of the enzymes from P. oxalicum sp. 68; ■ enzymatic activity using pNPG as the substrate; □ protein concentration at 280 nm; −− NaCl concentration

Table 1 Partial purification of enzymes from Penicillium oxalicum sp. 68

Biotransformation of protopanaxadiol-type saponins by partially purified enzymes from P. oxalicum sp. 68

The partially purified enzymes from P. oxalicum sp. 68 transformed the mixture of Rb1, Rb2 and Rc to the final product Compound K more efficiently than the fungus (Fig. 1b). Rb1 was transformed significantly after a 1-h incubation and completely converted after a 2-h incubation. The intermediate product, Rd, appeared first, followed by F2. The final product, Compound K, appeared after a 4-h incubation. Besides the two main intermediate products, Rd and F2, four minor intermediate products, ginsenosides CO, CY, Mb and Mc, were observed in this enzyme-catalyzed biotransformation. Intermediates CO and CY were from Rb2 and intermediates Mb and Mc were from Rc. Rb1 decreased more quickly than Rb2 and Rc, which indicated that the enzymes transformed Rb1 more efficiently. As shown in Fig. 1b, after a 12-h transformation, there was still a small amount of Rb2 and Rc remaining. After a 24-h incubation, all of the substrates and intermediate product were transformed to the main product, Compound K.

The biotransformation by the fungus strain was slow, and some intermediate products still remained in the final solutions, which led to difficulties with product isolation and low yields. Using the enzymes partially purified from P. oxalicum sp. 68, the transformation progressed more efficiently than with the fungus strain. The transformation rate by the partially purified enzymes was about 92 % after 24 h, while it was only about 58 % by spores after 8 days. Therefore, the partially purified enzymes have greater potential in minor ginsenosides preparation compared with the parent fungal strain.

Pathways of individual ginsenosides transformed by the partially purified enzyme

To verify the biotransformation by the partially purified enzyme, the individual ginsenosides were used as substrates for the biotransformation. As shown in Fig. 3a, Rb1 was transformed successively by the enzymes into three products, identified as Rd, F2 and Compound K by HPLC. These three compounds are the main products produced by the transformation of the ginsenoside mixture (Fig. 1b), which is consistent with the content of Rb1 in the mixture. Rb1 was completely transformed into the final metabolite Compound K via Rd and F2 after 24 h. The transformation pathway of Rb1 was Rb1→Rd→F2→Compound K. When Rd was used as starting substrate, it was completely transformed into Compound K via F2 after a 12-h transformation. As expected, the transformation pathway of Rd was Rd→F2→Compound K (Fig. 3b). As shown in Fig. 3c, Rb2 was transformed into Compound K via the ginsenosides CO and CY, which were different than those of the Rb1 and Rd transformation. The corresponding transformation pathway was Rb2→CO→CY→Compound K. Rc was converted into Compound K via ginsenosides Mb and Mc. The transformation pathway of Rc was Rc→Mb→Mc→Compound K (Fig. 3d). The different pathways of ginsenosides Rb1, Rb2, Rc and Rd are outlined in Fig. 4.

Fig. 3

Time course of the transformation of each ginsenoside by the partially purified enzymes from P. oxalicum sp. 68; the transformation was monitored by HPLC, attached to a Shim-pack PREP-ODS (H) Kit column (4.6 mm × 250 mm, 5 μm). a, b, c and d show the time course of Rb1, Rd, Rb2 and Rc, respectively

Fig. 4

Transformation pathways of Rb1, Rb2 and Rc by the partially purified enzymes from P. oxalicum sp. 68

The hydrolysis order of sugar reflects the selectivity of the partially purified enzymes for different glycosidic bonds. The hydrolysis preference for glycosidic bonds in the major protopanaxadiol-type saponins was as follows: outer β-1, 6-glucosidic linkage at C-20 > outer β-1, 2-glucosidic linkage at C-3 > inner glucose at C-3 > outer α-1, 6-arabinosyl linkage at C-20 (Arap for Rb2 and Araf for Rc). The inner glucose at the C-20 site could not be hydrolyzed by the partially purified enzymes. This hydrolysis selectivity resulted in different transformation pathways of different ginsenosides to produce different intermediate products, but the same final product, Compound K, which has been demonstrated to be one of the main pharmacologically active metabolites found in blood after oral administration of ginsenosides Rb1, Rb2, or Rc (Akao et al. 1998b; Bae et al. 2002; Qian and Cai 2010). Transformations from protopanaxadiol-type ginsenosides to Compound K by bacteria or fungi have been reported previously. Some microorganisms, such as A. oryzae (Rb1 and Rb2), S. solfataricus (Rb1), C. leidyia (Rb1), A. niger and A. usamii (Rb2), use similar pathways to P. oxalicum. Several microorganisms use pathway Rb1, Rb2, Rc→Rd→F2→Compound K (Penicillium sp., Aspergillus sp. g48p). The different pathways are summarized and listed in Table 2. The partially purified enzymes from P. oxalicum, would in any case have greater potential for Compound K production because of easier enzyme preparation, efficient transformation and simple product separation.

Table 2 Production of Compound K by microbial transformation

Preparation and identification of the transformed products

The transformed products were identified initially by TLC and HPLC while monitoring transformation progress using standard samples as references. To confirm the products, each product was prepared by biotransformation, purified by preparative HPLC, and identified by ¹³C NMR. By controlling the time course of the biotransformation, all products, including intermediate and final products, were obtained. The yields of each product are listed in Table 3. The ¹³C NMR spectra of obtained compounds are shown in Fig. 5. The ¹³C NMR data of Rb1, Rb2, Rc, Rd, F2, Mc and Compound K have been reported in the literature (Han et al. 2007; Ko et al. 2007; Zhao et al. 2009). The ¹³C NMR spectra of ginsenosides CO, CY and Mb are reported for the first time in this paper. By comparing our data with the known signals of C-1, we could assign the C-1 signals of CO, CY and Mb. All assignments of the C-1 signals of the ten ginsenosides are shown in Fig. 5. The changes in the C-1 signals of sugar rings in the NMR spectra clearly confirmed the products and the pathways of the biotransformation of Rb1, Rb2, Rc and Rd; Rb1 was transformed to Rd, then to F2 and finally to Compound K. The C-1 signals of the intermediate products of Rb2 and Rc were different from the signals of Rd and F2, which indicated that the Ara residues in C-20 were hydrolyzed in the final step to give the same final product, Compound K, as Rb1. Thus, ginsenoside CO is the first intermediate product of the transformation of Rb2, followed by CY. For Rc, the first intermediate product is Mb, followed by Mc.

Table 3 Yields of biotransformation products after preparative HPLC

Fig. 5

¹³C NMR spectra of related ginsenosides. The ¹³C NMR spectra were collected on a Bruker Av 600 NMR spectrometer at 150 MHz using MeOD as solvent

Optimization of pH, temperature and substrate concentration for Compound K production by the enzyme from P. oxalicum sp. 68

To improve the yield of Compound K, the biotransformation conditions by the partially purified enzymes were optimized. The enzyme concentration was evaluated at 0.1, 0.25, 0.5 and 1 U mL⁻¹; 0.25 U mL⁻¹ was found to be optimal since lower enzyme concentration caused slow transformation and higher concentration would waste enzyme (data not shown). Furthermore, the effects of pH, temperature and substrate concentration for Compound K production were evaluated using Rb1 as the substrate because Rb1 is the most abundant among the protopanaxadiol-type saponins (Fig. 6a). The optimum pH was 4.0 (Fig. 6b), and the optimum temperature 55 °C (Fig. 6c). Compound K yield at the optimum pH and temperature reached its maximum at 0.5 mg mL⁻¹ of Rb1. Higher or lower concentrations both decreased the yield (Fig. 6d). Under optimal conditions, the biotransformation reached the highest Compound K yield of 87.7 % (molar ratio).

Fig. 6

Optimization of conditions for Compound K production by enzymatic transformation. a Substrate, b pH, c temperature, d substrate concentration. The highest yield of Compound K was defined as 100 % in each experiment. The results are presented as mean ± standard deviations (n = 3)

It has been reported that Rb1 can be transformed to Compound K by some enzymes from bacteria, yeasts or fungi (Park et al. 2010), such as β-glycosidase from S. solfataricus (Noh et al. 2009) and Paecilomyces Bainier sp. 229 (Yan et al. 2010), lactase from Penicillium sp. and β-galactosidase from A. oryzae (Ko et al. 2007). Partially purified enzymes from P. oxalicum might be more suitable for producing Compound K from the main protopanadaxiol-type saponins Rb1, Rb2 and Rc. One reason for this is that the enzymes from P. oxalicum can be prepared easily from the culture by DEAE-cellulose chromatography and ammonium sulfate precipitation in high yield (79.1 %). The partially purified enzymes are highly active in transforming major protopanadaxiol-type saponins, including Rb1, Rb2, Rc and Rd to Compound K in relatively mild conditions. Thus, the biotransformation for producing Compound K would be more efficient because the major protopanadaxiol-type saponins do not need to be separated in this transformation. As the result, the preparation procedure of Compound K would be shorter, the cost would decrease and the environmental pollution caused by producing Compound K would be less. Therefore, partially purified enzymes from P. oxalicum have potential for the large-scale production of Compound K. This biotransformation is very significant in the utilization of ginseng resources and might be useful for preparing Compound K and other pharmaceutically active compounds from naturally abundant glycosides.

Conclusions

The pathogenic fungus Penicillium oxalicum sp. 68 can transform the major protopanaxadiol-type saponins into bioactive Compound K. Partially purified enzymes from P. oxalicum sp. 68 can transform protopanaxadiol-type saponins into Compound K more efficiently than the parent fungus strain. Under optimum conditions (pH 4.0; temperature, 55 °C and substrate concentration, 0.5 mg mL⁻¹), the enzymes transform Rb1 to Compound K in 87.7 % yield (molar ratio). The transformation pathways for Rb1, Rb2, Rc and Rd are determined to be Rb1→Rd→F2→Compound K, Rb2→CO→CY→Compound K, Rc→Mb→Mc→Compound K and Rd→F2→Compound K. Because the enzymes can be prepared simply and can transform protopanaxadiol-type saponins efficiently into the pharmaceutically active Compound K, this process has potential for industrial-scale Compound K production and significance for protection of the environment.