Introduction

Resveratrol (3,5,4′-trihydroxystilbene) is a widely recognized compound that is known for preventing and slowing the occurrence of some diseases, including cancer (Jang et al. 1997), cardiovascular disease (Bradamante et al. 2004), and ischemic injuries (Wang et al. 2002; Sinha et al. 2002). It has also been shown that resveratrol can enhance stress resistance (Howitz et al. 2003) and extend the lifespan of various organisms ranging from yeasts to vertebrates (Valenzano et al. 2006). Resveratrol has been found in grapes (Vitis vinifera), a variety of other berries, peanuts, and medicinal plants such as Japanese knotweed (Polygonum cuspidatum Siebold & Zucc.) (Baur and Sinclair 2006). The resveratrol found in grapes can be transferred to wine during winemaking and is directly related to the resveratrol content in wine (Casas et al. 2009). The pretreatment (crushing and pressing) and vinification process in winemaking could also affect the resveratrol content found in wine. Wine has attracted much attention because it is one of the primary sources of resveratrol for humans due to its high resveratrol content. Therefore, many efforts have been made to increase the resveratrol content in grape fruits by using short anoxic treatments or enriched ozone atmosphere in grape storage (Jiménez et al. 2007; Artés-Hernández et al. 2007), and in wine by using transgenic yeast for fermentation (Wang et al. 2011), respectively. Currently, resveratrol extracts are mainly produced in China using an extraction process from Japanese knotweed (Changsha Nutramax Inc. 2009).

Microbial fermentation has been used to produce various valuable compounds on a large scale due to its high efficiency, ease of operation, and low cost (Chemler and Koffas 2008). Therefore, researchers have explored ways to produce resveratrol using recombinant microorganisms (Beekwilder et al. 2006; Shin et al. 2011), chemical synthesis, and plant cell cultures (Fan et al. 2010). Identifying microorganisms able to produce resveratrol should provide new resources for genes, or new pathways for producing resveratrol.

At present, most genes involved in the trans-resveratrol biosynthesis pathway have not been described in microorganisms (Kiselev 2011). Currently, the genes used for producing trans-resveratrol in vitro are from plant origin, including grape, tobacco, and peanut (Beekwilder et al. 2006; Shin et al. 2011). These genes encode 4-coumarate CoA-ligase (4CL) and stilbene synthase (STS), which are required for resveratrol biosynthesis when 4-coumaric acid is added to the culture medium as a precursor. So far, only one microbial gene for a putative 4CL has been identified in Aspergillus flavus, and it is available in the National Center for Biotechnology Information database (NCBI) (Genbank accessing number: XM002380137). Homology searches using the encoded protein sequences indicate that 4CL genes might be abundant in fungi.

Many endophytes exhibit the capability to produce the same functional compounds as their hosts (Kumaran et al 2010; Suryanarayanan et al. 2009; Strobel and Daisy 2003) and live asymptomatically within plant tissues (Schulz et al. 1993). Therefore, endophytes have been studied for new medicine development and plant disease management (Arnold et al. 2003; Azevedo et al. 2000; Berg et al. 2004; Murray et al. 1992). However, endophytes able to produce resveratrol have not been reported.

The aim of this study was to identify microorganisms that can produce resveratrol and optimize the conditions for producing resveratrol through microbial fermentation. Plant Merlot wine grape (V. vinifera L. cv. Merlot), wild Vitis (Vitis quinquangularis Rehd), and Japanese knotweed (P. cuspidatum Siebold & Zucc) were used as the sources for isolation of these microorganisms.

Materials and methods

Materials

Three different plants were used to isolate resveratrol-producing endophytes: grape cultivar Merlot (V. vinifera L. cv. Merlot), wild Vitis (V. quinquangularis Rehd), and Japanese knotweed (P. cuspidatum Siebold & Zucc) grown in Yangling Town, Xianyang City of Shaanxi Province; Luocheng Town, Hechi City of Guangxi Province; and Taochuan Town, Taibai County, Baoji City of Shaanxi Province, China, respectively. The fruits of the plants were picked during harvest. The stems were also collected together with the fruits for both Merlot and wild Vitis, and the whole stem tuber was collected for Japanese knotweed. The fruits were placed in sterilized bags and kept on ice during transportation. Samples were analyzed within 24 h of acquisition.

Media

Three media were used for the isolation of endophytes according to that of Liu et al. (2010): Gause medium G-1 (GA1: soluble starch 20 g, KNO3 1 g, K2HPO4 0.5 g, MgSO4·7H2O 0.5 g, NaCl 0.5 g, FeSO4·7H2O 0.01 g, agar 20 g, and distilled water 1 L; pH 7.4–7.6), beef-protein medium (BP: beef extract 3 g, peptone 10 g, NaCl 5 g, agar 20 g, and distilled water 1 L; pH 7.4–7.6), and potato dextrose agar (PDA: potato (peeled and diced) 200 g, dextrose 20 g, agar 20 g, and water 1 L). The liquid phases of these media (without agar) were used to determine if the obtained isolates could produce resveratrol. The broth of potato dextrose medium (PDB) was used to study the optimum conditions for Alternaria sp. MG1 to produce resveratrol.

Czapek yeast extract agar medium (sucrose 30 g, yeast extract 5 g, NaNO3 3 g, KCl 0.5 g, MgSO4·7H2O 0.5 g, FeSO4·7H2O 0.01 g, K2HPO4 1 g, agar 20 g, and water 1 L) was used to identify strains of the genera Penicillium and Aspergillus according to the method reported by Pereyra et al. (2011) and Pitt and Hocking (1997). Potato carrot agar medium ((PCA: potato (peeled and diced) 20 g, carrot (peeled and diced) 20 g, agar 20 g, and water 1 L) was used to identify Alternaria sp. (Sørensen et al. 2009). Other isolates were observed on PDA plates.

Isolation of endophytes

Endophytes were isolated by cultivating the tissues of berries, cobs, and stems of Merlot and wild Vitis, and the root and stem tissues of Japanese knotweed. All samples were cultivated on GA1, BP, and PDA after the sample surfaces were sterilized and rinsed with sterile water (Larran et al. 2002). The rinse water was collected after the last rinsing and also cultivated to ascertain a complete surface sterilization (Naik et al 2009). All inoculated plates were cultivated in darkness at 28 ± 1 °C. During cultivation, colonies were transferred to a fresh medium from which the isolates were obtained for purification. All experiments for each test were conducted in triplicate.

Screening resveratrol-producing endophytes

In order to produce comparable results, spore suspensions of 104/mL (measured using a hemacytometer) from all of the endophytic fungi were prepared. Samples of 1-mL suspension were used to inoculate three parallel cultures of 100 mL of a liquid medium that corresponded to the solid medium used for isolation. After incubation in darkness at 28 ± 1 °C with a rotation speed of 100 rpm for 7 days, the liquid broth and cells were separated by centrifugation at 3500 × g for 15 min. The collected cells were crushed into powder after freezing in liquid nitrogen and then extracted with 80 % ethanol (15 mL ethanol per 1 g cell weight) three times (10 h for each extraction) to obtain the resveratrol inside cells. The liquid broth and ethanol extracts of the cells were combined and concentrated to 100 mL at 45 °C using a vacuum evaporator (R-200, BUCHI, Flawil, Switzerland). The 100-mL sample was partitioned three times with 50 mL ethyl acetate each time. The three samples of the ethyl acetate phase were combined and then partitioned with three volumes 20 mL 3 % NaHCO3 to remove water. All the obtained ethyl acetate phase was evaporated to dryness at 40 °C and then dissolved in 2-mL methanol. Finally, to determine resveratrol content, the methanol solution was analyzed by high performance liquid chromatography (HPLC) after being filtered through a Millex-HV filter membrane (0.45-μM, 13-mm diameter; Millipore, Billerica, MA, USA). The resulting resveratrol content in the liquid culture (in micrograms per liter) was determined and reported as the average of the three parallel cultures.

Conditions for HPLC measurement of resveratrol concentration

Resveratrol concentration in the methanol extracts obtained above was measured using a Shimadzu Essentia LC-15 C analytical HPLC system (Shimadzu, Kyoto, Japan) equipped with LC-15 C pumps, a SIL-10AF automated sample injector, a SPD-15 C dual-wavelength detector, a Shimadu Wondasil C18-column (250 × 4.6 mm), and LC solution software. The column was operated at a temperature of 35 °C. The mobile phase was acetonitrile (Sigma, St. Louis, MO, USA) (eluant A) and double-distilled water (eluant B). A multistep gradient was used for all tests at a flow rate of 1 mL/min according to the following steps: 0–28 min: 95 % (v/v) B; 28–33 min: 40 % B; 33–40 min: 15 % B; and 40–50 min: 95 % B. The sample injection volume was 20 μL. Trans-resveratrol (≥99 % ([GC], Sigma) was used as the standard for measurements.

Identification of resveratrol-producing endophytes

Colony appearance and sporulation of the isolates were observed on the corresponding identification media for morphological identification (Xing and Guo 2011). Specimens for light microscopy (BA 400; Motic, Richmond, Canada) were mounted in 3 % KOH or sterile distilled water and observed (Yuan et al. 2011). The sequence of the internal transcribed spacer (ITS) regions of ribosomal DNA (rDNA), including ITS1, 5.8S rDNA, and ITS2 (GenBank accession number: JN102357), was determined for molecular identification of strain Alternaria sp. MG1. For sequencing, cultivated mycelium was freeze-dried and ground with liquid nitrogen (Dey et al. 2011). Genomic DNA was extracted from the obtained mycelia using a Fungus Genomic DNA extraction kit (BioFlux, Kyoto, Japan), and the amplification of the whole ITS region of rDNA was performed by using a Fungi Identification PCR Kit (TaKaRa, Kyoto, Japan) and primers, 5′-GAGCGGATAACAATTTCACACAGG-3′ and 5′-CGCCAGGGTTTTCCCAGTCACGAC-3′, according to the manufacturer’s instructions.

Profiles of cell growth and resveratrol production of Alternaria sp. MG1

The strain Alternaria sp. MG1 was selected to obtain profiles of cell growth and resveratrol production in vitro in 100-mL liquid PDB in a 250-mL flask in an incubator at 28 °C with a rotation speed of 100 rpm. Dry cell weight was measured every day after drying the mycelia at 60 °C for 48 h, and the resveratrol produced was measured using the above described methods from days 2 to 11 of cultivation. The resveratrol contents in the cell-free broth and inside the cells were also measured separately at days 5 and 7 to determine the distribution of resveratrol within the cultures. For separate analysis of the resveratrol accumulation in liquid broth and inside cells, the liquid broth and cells were treated separately after centrifugation. Similar to the treatment described above, cells were extracted with ethanol and ethyl acetate, while the liquid broth was extracted with ethyl acetate directly, dried with 3 % NaHCO3 and finally dissolved in 2-mL methanol for HPLC measurement. The resulting resveratrol content in the liquid culture (in micrograms per liter) and within the cells (in micrograms per gram) was determined and reported as the average of three parallel samples.

Optimum production conditions of resveratrol by Alternaria sp. MG1

The single-factor design and the Box–Behnken design were used to optimize the production conditions of resveratrol by Alternaria sp. MG1. The four parameters of temperature, rotation speed, inoculum size, and medium volume of flask were investigated. In the single-factor design, unless specified, four parameters were set as follows: temperature = 28 °C, rotation speed = 100 rpm, inoculum size = 5 %, and medium volume of flask = 100 mL in a 250 mL flask. The inoculum size was measured as the volume ratio of spore suspension (1 × 104 spore/mL) to the liquid medium. The levels of each parameter in the single-factor design were set at 14, 21, 28, 35, and 42 °C for temperature test; 60, 90, 100, 120, 140, 160, and 180 rpm for rotation speed test; 1, 5, 10, and 15 % for inoculum size test; and 100, 125, 150, 175, and 200 mL for medium volume of flask test.

Based on the results obtained from the single-factor experiments, a set of 29 experiments were designed and carried out according to the Box–Behnken design with four factors (see Table S1 in the supplementary materials) to determine the optimum levels for each parameter under interaction. The results of the Box–Behnken design are explained by Eq. (1) (Khajeh and Ghanbari 2011):

$$ Y = \beta 0 + \sum\limits_{{i = 1}}^k {{\beta_i}{\chi_i}} + \sum\limits_{{i = 1}}^k {\beta ii\chi_i^2} + \sum\limits_{{i = 1}}^k {} \sum\limits_{{j = 1}}^k {{\beta_{{ij}}}{\chi_i}{\chi_j}} + \varepsilon $$
(1)

The statistical software package Design-Expert (Version 7.0.0; Stat-ease Inc., Minneapolis, MN, USA) was used to make the response surface methodology analysis and analyze the experimental results.

Results

Isolated endophytes and resveratrol-producing endophytes

A total of 65 isolates were obtained from the three tested plants: 30 from Merlot, 15 from wild Vitis, and 20 from Japanese knotweed (Table 1). Only five isolates were identified as bacteria, and the remaining 60 were fungi. In total, 21 strains, including seven from Merlot, six from wild Vitis, and six from Japanese knotweed, produced resveratrol in the range of 6–123 μg/L (Table 2). Figure S1 (see supplementary materials) shows the HPLC chromatogram of standard trans-resveratrol and the culture of Alternaria sp. MG1. The HG6, MP1, MP15, YG3, and MG1 strains produced >90 μg/L resveratrol (Table 2). The resveratrol production capability of most isolates, however, declined and even disappeared for some isolates after being subcultured three times. However, only the MG1 strain (isolated from Merlot using GA1 medium) remained the most stable and retained the highest production capability (Table 2).

Table 1 Endophytes isolated from Merlot, wild Vitis, and Japanese knotweed
Full size table
Table 2 Reveratrol production of the obtained resveratrol-producing fungi
Full size table

Identification of resveratrol-producing isolates

All resveratrol-producing isolates were identified and confirmed based on their specific morphological and sporulation characteristics. As shown in Table 2, the resveratrol-producing fungi belonged to seven genera including Alternaria (nine strains), Botryosphaeria (one strain), Penicillium (two strains), Cephalosporium (five strains), Aspergillus (one strain), Geotrichum (one strain), and Mucor (one strain). All the resveratrol-producing strains isolated from Merlot belonged to genus Alternaria. The strain MG1 was studied further to provide information for potential future commercial production.

The strain MG1 showed felty and brown colony morphology with a dark color in reverse side of the medium after cultivation on PCA for 5 days (Fig. 1). The colony growth showed a concentric appearance (Fig. 1a) with the mycelium displaying smooth subhyaline branches consisting of septate hyphae and 3.5–4 μm in width. Conidiophores were abundant, subhyaline, erect or ascending from both submerged and aerial hyphae, simple or branched, septate, 50–70 μm × 3–3.5 μm in length and width, with three to six uniperforate geniculations (Fig. 1b). The conidia were obclavate, ovoid, obpyriform, with three to seven transverse septa (commonly four), zero to three longitudinal septa, smooth, brown to dark brown, and 4.3–12.5 × 18.2–40 μm in length and width. Conidial chains appeared after 7 days of cultivation and were multi-branched with sizes of 1–15 conidia (see Fig. 1c). These morphological characteristics were consistent with that of Alternaria sp. (von Arx 1981).

Fig. 1
figure 1

Colony (a), conidiophore (b), and conidial chain (c) of Alternaria sp. MG1. Pictures were taken after 5 days (a), after 3 days (b), and after 7 days of incubation (c). The cultivation was performed on potato carrot agar medium (PCA)

Full size image

The DNA sequence of the ITS regions of strain MG1, together with the 5.8 S rDNA, was 622 bp long and archived in the GenBank database under the accession number JN102357. The sequences were subjected to a sequence similarity search performed through the NCBI database using the Basic Local Alignment Search Tool (BLAST) (http://www.ncbi.nih.gov/BLAST/). BLAST results showed a highest identity value of 91 % between the ITS sequence data of MG1 and the best hit to a sequence available at GenBank. The ITS sequence data for strain MG1 and closest related species available at GenBank were selected for the construction of a phylogenetic tree using the neighbor-joining method in the MEGA (version 4.0) program after sequence alignment with ClustalX (version 1.8) (Thompson et al. 1997; Saitou and Nei 1987; Tamura et al. 2007). The confidence values for individual branches were determined by bootstrap analyses (1,000 replications) and maximum parsimony. As shown in Fig. S2 (supplementary materials), the sequences with the highest identity (91 %) to MG1 were Alternaria sp. Hf3 (GenBank accession number: GU183167.1) with a bootstrap value of 76. Thus, isolate MG1 was identified as Alternaria sp. consistent with the morphological identification. Isolate MG1 was deposited in the China Center for Type Culture Collection (Wuhan, China) and coded Alternaria sp. CCTCC M 2011348.

Profiles of cell growth and resveratrol production by Alternaria sp. MG1 in liquid culture

The peak cell density of strain MG1 in potato dextrose medium was reached after continuously growing for 5 days (Fig. 2). A decrease in dry cell weight was observed after 8 days of cultivation. The resveratrol accumulation, both within cells and in the medium, started from the first day of cultivation and approached its highest value of 376 ± 13 μg/L after continuously increasing for 7 days, and resveratrol amounts sharply decreased. In the early stage of cultivation, the cell growth and total amounts of resveratrol within cells and in the medium increased almost simultaneously (Fig. 2). However, in a parallel analysis of resveratrol amounts within cells and in the medium, no resveratrol was detected in the medium. These results revealed that the intracellular amounts of resveratrol on days 5 and 7 of cultivation were 224 ± 25 μg/L medium and 353 ± 24 μg/L medium, respectively, corresponding to 69 ± 8 μg/g dry cell weight and 113 ± 6 μg/g dry cell weight, respectively. Therefore, resveratrol may be a constitutive product that accumulates within cells of Alternaria sp. MG1.

Fig. 2
figure 2

Profiles of resveratrol production (closed square) and cell growth (open square) in liquid fermentation of Alternaria sp. MG1. The cultivation was carried out in broth of potato dextrose medium (PDB) at 28 °C with a rotation speed of 100 rpm. The profile of cell growth was drawn according to the dry cell weight. The bars show the standard deviation of the mean of three replicates

Full size image

Optimum conditions for resveratrol production by Alternaria sp. MG1

The effect of each factor (inoculum size, medium volume of flask, rotation speed, and temperature) on resveratrol production from the single-factor tests is shown in Fig. 3. All factors had a similar effect on the resveratrol production showing an initial increase followed by a decrease. According to the highest values of resveratrol production obtained from the single-factor tests, the middle values for the Box–Behnken design were set as 5 % for inoculum size, 125 mL for medium volume of flask, 100 rpm for rotation speed, and 28 °C for temperature. The test results are listed in the supplementary materials (Table S2).

Fig. 3
figure 3

Effect of inoculum size (a), medium volume of flask (b), rotation speed (c), and temperature (d) on the resveratrol production of Alternaria sp. MG1 in liquid cultivation. The experiment was carried out at conditions of (a) 28 °C, rotation speed of 100 rpm, and medium volume of flask of 100 mL/250 mL, (b) 28 °C, rotation speed of 100 rpm, and inoculum size of 5 %, (c) 28 °C, inoculum size of 5 %, and medium volume of flask of 100 mL/250 mL, and (d) inoculum size of 5 %, medium volume of flask of 100 mL/250 mL, with a rotation speed of 100 rpm. All the cultivations were carried out in liquid potato dextrose medium (PDB) and in dark environment by using a sheet of black paper to cover the incubator. The bars show the standard deviation of the mean of three replicates

Full size image

Based on the obtained data, a statistical model (Eq. 2) was obtained to determine the yield trends of resveratrol production.

$$ {R_{{1}}} = {413}.{88}{X_{{1}}} - {1}.{8}0{X_{{2}}} - {3}.{62}{X_{{3}}} - {85}.{36}{X_{{4}}} - {4}0.{84}{X_{{1}}}{X_{{2}}} + {93}.0{9}{X_{{2}}}{X_{{3}}} + {44}.{55}{X_{{2}}}{X_{{4}}} - {4}0.{78}X_1^2 - {6}0.{1}X_2^2 - {79}.{61}X_3^2 - {1}0{5}.{4}0X_4^2 $$
(2)

where R 1 is the amount of resveratrol (in micrograms per liter), X 1 is the inoculum size (percentage), X 2 is the medium volume of flask (in milliliters per 250 mL), X 3 is the rotation speed (in revolutions per minute), and X 4 is temperature (in degrees Celsius).

Analysis of variance for the model demonstrated that the model is highly significant based on Fisher F test (F = 36.2789) and a low probability value (P = 0.0001). The high coefficient of determination (R 2 = 0.9591) and adjusted coefficient of determination (adjusted R 2 = 0.9327) indicated the high significance of the model. The linear coefficients (X 1, X 4), four quadratic term coefficient (X 21 , X 22 , X 23 , and X 24 ), and cross product coefficients (X 1 X 2, X 2 X 3, and X 2 X 4) were statistically significant (P < 0.05). The coefficients of other parameters were not significant (P > 0.05). Temperature (X 4) showed the most significant effect on resveratrol production, followed by inoculum size (X 1), rotation speed (X 3), and medium volume of flask (X 2). By solving the model, the optimum conditions for the highest resveratrol production were obtained as 6 % (v/v) inoculum size, 125 mL/250 mL medium volume of flask, 101 rpm rotation speed, and 27 °C temperature. Under these conditions resveratrol production was 451 μg/L by predicted value and 422 ± 18 μg/L experimentally.

Response surface methodology is usually used to study the effects of several factors at different levels and their interactions (Priya and Kanmani 2011). The two-dimensional contour plots of the parameters show the effect of two factors on resveratrol production at a time (Fig. 4). The resveratrol production values in all figures were obtained along with two variables, while the other two variables were kept at level zero (middle value of the testing ranges). Elliptical contours mean that there is a significant interaction between the independent variables. It can be seen that the interaction between the medium volume of flask and rotation speed was the most significant (Fig. 4b), followed by medium volume of flask and temperature (Fig. 4c), and then inoculum size and medium volume of flask (Fig. 4a).

Fig. 4
figure 4

Contour plots of the effects of interactions between medium volume of flask and inoculum size (a), between medium volume of flask and rotation speed (b), and between medium volume of flask and temperature (c) on resveratrol production of Alternaria sp. MG1 in liquid cultivation. The cultivation was carried out in liquid potato dextrose medium (PDB)

Full size image

Discussion

Fungal endophytes are gaining increased importance because of their enormous potentials for the production of novel bioactive compounds for medicine and agriculture (Aly et al. 2011). Among the various secondary metabolite-producing endophytes, resveratrol-producing fungi were discovered for the first time from different plant tissues and fruits. Resveratrol-producing endophytes were isolated from plants with a high content of resveratrol, which is consistent with the coevolution theory reported by other research on endophytes (Amna et al. 2006; Eyberger et al. 2006; Kusari et al. 2009a, 2009b; Puri et al. 2005, 2006; Stierle et al. 1993). In addition, isolation of resveratrol-producing isolates was influenced when different plant tissues and media were used (as shown in Tables 1 and 2). Arnold et al. (2003) and Sun et al. (2011) have also found that there was a wide diversity of endophytic fungi in different plants and in different parts of a same plant.

The strain MG1 became the principal focus of this study since it exhibited a high and stable resveratrol-producing capability during subculture. For several other isolates, the resveratrol-producing capability decreased and was even abolished after subculture. Similar decrease and loss of the capability for production of other compounds has also been reported in other metabolite-producing endophytes isolated from plants (Wang et al. 2010). The instability of resveratrol production could possibly be caused by the absence of the host plant in vitro culture of the endophytes, which would cause unstable expression of the related genes and the transformation of metabolite pathway within cells. Studies on the biosynthesis pathways and discovery of the key genes encoding these pathways would provide essential information for the mechanism of loss of production.

During the liquid cultivation of the strain MG1, resveratrol accumulation was mainly found inside the cells and was shown to increase with increased biomass. This was different from the biosynthesis of secondary products by most microorganisms, which began at the late stage of exponential phase of cell growth (Chomcheon et al. 2009; Kornsakulkarn et al. 2011; Papagianni 2003). The profiles of cell growth and resveratrol accumulation showed that resveratrol should be a constitutive product in Alternaria sp. MG1.

The discovery of resveratrol-producing fungi in this study revealed the existence of resveratrol-producing microorganisms in nature. The genus Alternaria is normally reported as a pathogen for many plants (Akamatsu et al. 1999; Akamatsu 2004). Alternaria spp. have also been widely found in grapes as an endophyte or a fungus causing rotting or toxin production (Gonzalez and Tello 2011; Tournas and Stack 2001; Ostry 2008). Discovery of the resveratrol-producing capability of Alternaria sp. in this study indicates the possibility of a new pathway or new genes for the biosynthesis of resveratrol in microorganisms.

The effect of inoculum size, medium volume of flask, and temperature on resveratrol production in liquid fermentation apparently relates to the cell growth of Alternaria sp. MG1. In the batch cultivation method used in this study, low inoculum sizes resulted in a slow increase of cell growth, while inoculum sizes that were too large possibly exhibited cell degradation and thus resulted in low cell weight within a given period. A low medium volume of flask leads to a high relative inoculum size, while a high medium volume of flask corresponds to a low relative inoculum size. Therefore, similar cell growth data were obtained due to the influence of both inoculum size and medium volume of flask. The effect of temperature on resveratrol production is expected to be related to the temperature adaptability of the strain. The optimum temperature for resveratrol production was the same as that for the cell growth of Alternaria sp. MG1 (28 °C). The decrease in resveratrol production at a high rotation speed could be due to resveratrol being unstable or produced in lower amounts in the presence of oxygen, since high rotation speeds resulted in high oxygen content in the culture (King et al. 2006; Santos et al. 2011).

Recombinant microorganisms have been created for resveratrol production. For example, Escherichia coli strain JM109 has been transformed with the 4CL gene from Arabidopsis thaliana and the STS gene from Arachis hypogea, upon which it was demonstrated to be able to convert 4-coumaric acid into resveratrol with a yield of 100 mg/L (Watts et al. 2006). However, when E. coli received the 4CL gene from Lithospermum erythrorhizon and the STS gene from A. hypogea, and coumaroyl-CoA was used as a substrate, resveratrol production reached 171 mg/L (Katsuyama et al. 2007). Significant progress in increasing the resveratrol content in plant cells has been achieved by transforming grape cells with the rolC gene of Agrobacterium rhizogenes (Dubrovina et al. 2010) and treating the cell cultures with methyljasmonate and cyclodextrins that resulted in a resveratrol content of 1,600 μmol/g (Lijavetzky et al. 2008). In the current study, we found that the resveratrol production by Alternaria sp. MG1 was 353 μg/L in total culture and 113 μg/g within cells after 7 days of cultivation in PDB. The yield by fermentation of Alternaria sp. MG1 was much lower than that of plant cells and genetically modified bacteria in total culture, but comparable when the cells were collected for extraction. Collectively, these data show potential for improving the resveratrol production capability may be achievable given appropriate metabolic controls and the development of overproducing mutant strains.