Callus Induction and Cellular Suspensions from Murtilla (Ugni molinae Turcz.) for trans-resveratrol Production

1Department of Plant Production, Faculty of Agriculture, University of Concepción, Chillán, Chile 2Laboratory of Natural Products Chemistry, Department of Botany, Faculty of Natural Sciences and Oceanography, University of Concepción, Concepción, Chile 3Department of Agrochemistry and Biochemistry, Faculty of Science, University of Alicante, Alicante, Spain 4Laboratory of Tissue Culture, Biotechnology Center, Faculty of Forestry Sciences, University of Concepción, Concepción, Chile


Introduction
Chile is a country with high biological diversity. Here high concentrations of antioxidants provided by natural compounds such as anthocyanins, flavonoids and phenolic acids, have been identified in native and/or endemic plant species (Ruiz et al., 2010;Rubilar et al., 2011). However, resveratrol, an important antioxidant compound, mainly foundin berries, has not been detected in native Chilean plants yet. One such species, Ugni molinae Turcz. (murta or murtilla in Chile), a perennial plant of the Myrtaceae family is known for its aromatic and sweet berries fruits. The, contents of flavonoids and other polyphenols has been found in these fruits with properties that indicate presence of compounds such as phytoalexins (Aguirre et al., 2006;Rubilar et al., 2006;Avello et al., 2009). Resveratrol (3,5,4'trihydroxystilbene; t-R) a phytoalexin that belongs to the stilbene family, is naturally occurring in its isomeric forms cis and trans, or bound to a glycoside (piceid). The trans isomer is a bioactive compound with several health benefits (antioxidant, anti-inflammatory, anticancer and anti-diabetic) prolonging cells longevity (Baur and Sinclair, 2006). These compounds have been identified in fruits of commercial interest, such as peanuts, pistachios, grapes and some berries (e.g., blueberries) (Rocha-González et al., 2008). t-R is typically obtained from these sources by conventional extraction and separation techniques that involve costly chemical methods and large volumes of plant tissue. This has a number of disadvantages, such as its content in raw material is low and accumulates slowly, there is a wide variability among individuals and abusive use may lead to eradication of plant species (Trejo-Tapia and Rodríguez, 2007;Rocha-González et al., 2008). Given these factors, callus culture and the selection of highly productive lines have attracted much interest in recent years to produce high contents of alkaloids, saponins, polyphenols and terpenes like shikonin, taxol and berberin, which are widely used in the pharmaceutical, agrochemical and food industries (Karuppusamy, 2009). Cellular suspensions for resveratrol production have been established from species such as Vitis spp. (Bru et al., 2006;Santamaria et al., 2011a;2011b;Belchi-Narravo et al., 2012), Arachis hypogaea L. (Ku et al., 2005) and Gossypium hirsutum L. (Kouakou et al., 2006). The level of resveratrol is shown to depend not only on the speciesor genotype of donor plant, but also on culture conditions and type of elicitor used (Morales et al., 1998;Donnez et al., 2009;Halder and Jha, 2015;Xu et al., 2015). Of these, the highest content has been found in cellular suspensions of Vitis vinifera with values up to 1,257 mg L −1 per day (Bru et al., 2006). Nevertheless, stilbenes have not yet been quantified and/or determined in a variety of plants and fruits, especially wild ones. This plant has never been studied in the biotechnology processes or for resveratrol quantification to select the most productive stilbenes lines. Given this situation, this study aimed to assess for the first time the t-R content in three ecotypes of Ugni molinae.

Plant Material
Three ecotypes aged 1 and 1.5 years, grown under nursery conditions by the Chilean National Institute of Agricultural Research (INIA, Carillanca), the Araucanía Region, were used for this study. These ecotype were obtained from three different sites; Porma (38º08, 73º16, 70 masl), the Araucanía Region (E1); Aucar-Quemchi (42º09, 73º29, 20 masl) the Los Lagos Region (E2) and Mehuin (39º26, 73º12, 10 masl) the Los Ríos Region (E3). Leaves, immature and mature fruits, were randomly collected by hand from each mother plant. Samples were taken to the laboratory under cold conditions and washed with tap water.

In vitro Establishment of Leaves and Fruits for Theinduction of Callus
Leaves were washed with sterile distilled water for 5 min, followed by a superficial aseptic in a laminar flow chamber with 70% ethanol (v/v) for 5 min and three rinses with sterile distilled water (3 min each). 0.1% mercury (II) chloride HgCl 2 solution (v/v) was applied as a disinfectant for 1 min, followed by five washes with sterile distilled water (3 min each), following the method by Liu et al. (2010). Explants segmented in sizes of approximately 50 mm 2 , keeping a part of the petiole, were arranged in an abaxial position on the culture medium. Both mature and immature fruits were washed with tap water and 2% (v/v) commercial dish soap (Quix™) for 5-10 min and continuously immersed under running water for 6 h. Explants were sterilized with 1.25% sodium hypochlorite NaClO for 30 s and rinsed five times with sterile distilled water in a laminar flow chamber (Liu et al., 2010). Then the exocarp was extracted and segmented into 50 mm 2 pieces and established with the outer exocarp surface on the culture medium.

Media and Culture Conditions
The growth medium used was BTM (Chalupa, 1983), supplemented with 30 g L −1 sucrose, 0.5 g L −1 Polyvinylpyrrolidone PVP (Calbiochem™), 2 ml L −1 Plant Preservative Mixture PPM (Nalgene™) and different concentrations and combinations of 2,4-D, KIN and NAA as growth regulators ( Table 1). The pH of the media was adjusted to 5.8 with HCl and NaOH (1M) prior to solidification with 7 g L −1 bacteriological agar (Merck ™ ). Media were autoclaved for 20 min at 1 atm and 121°C and then arranged in Petri dishes. Six explants were established, with five replicates per treatment for leaves and six replicates for fruits. Trials were incubated in a growth chamber at 25±1°C and 55% relative humidity in constant darkness. During callogenesis, six subcultures were performed to increase biomass. In the last subculture, cell growth by means of Fresh Weight (FW) and the contents of tresveratrol and t-piceid in the formed calli (µg g −1 FW) were assessed.

Cellular Suspensions
Friable calli (2 g FW) of selected lines were used to initiate cellular suspensions in 250 mL flasks, with 100 mL of the respective culture medium without agar and supplemented with 20 g L −1 sucrose, 0.5 g L −1 PVP, 2 mL L −1 PPM, 2 g L −1 potassium nitrate KNO 3 and 0.250 g L −1 hydrolyzed casein. Suspensions were maintained in an orbital shaker at 110 rpm in a growth chamber at 25±1°C and in constant darkness for 21 days. Three replicates were performed per each cell line. Cell growth was quantified every 2 days and expressed as packed cell volume (%PCV) after centrifugation at 4,000 rpm for 10 min in a swing bucket rotor. The viability of these aggregates was assessed every 7 days with 100 µL Trypan Blue (0.1%) under an Olympus CX31 optic microscope.

Extraction of Stilbenes from Calli and Cell Aggregates
One g of calli was maintained in the darkness with 4 mL of 80% ethanol for 12 h at 4°C with continuous stirring. Samples were centrifuged at 3,000 rpm for 10 min. Finally, the supernatant was collected and filtered through a 0.45 µm Millipore ™ membrane and stored at -20°C until analyzed in HPLC (Martínez-Esteso et al., 2009).

Extraction of T-Resveratrol from Plant Material
Fresh leaves (10-15 g) were extracted over 5 d in 95% ethanol at 4°C in the dark following the methodology of Rubilar et al. (2006) with some modifications. For the exocarps of the mature and immature fruits, 20-50 g of fruits were extracted for 7 d in methanol:acetone:water:formic acid (40:40:20:0.1 v/v) at 4°C in the dark, according to the methodology of Rimando et al. (2004) with some modifications. The obtained filtrates were homogenized in ultrasound equipment (Branson 1210 ™ ) and the organic solvent was removed by a rotary evaporator (Heidolph OB2000 ™ ) at 30°C. This method was performed in triplicate and samples were stored at -20°C until analyzed.

HPLC Analysis of T-Resveratrol and T-Piceid
Stilbenes were analyzed in a SPD-M10 Avp model chromatographer (Shimadzu™) equipped with a C18 reverse phase column (Kromasil, 250×4.6 mm, 5 µm 100 A C18, Phenomenex), a multifocal diode arrangement detector (190-800 nm) and an LC-10ATvp Shimadzu™ pump. The Class-vp software was run for data acquisition and instrument control. Methanol solutions of 1 mg mL −1 t-R (Calbiochem™) and 1 mg mL t-P (Sigma-Aldrich™) were used as standards which were detected at a fixed wavelength of 306 nm. Separation of previously filtered in vitro material (Millipore 0.22 µm), was prepared at 40°C using a mixture of 0.1% acetic acid in water (solvent A) and acetonitrile (solvent B) at a flow rate of 1.0 mL min −1 , in a 30 min segmented gradient consisting of 5-70% B (25 min), 95% B (0.1 min), 95% B (2 min) and 5% B (0.1 min). Under these conditions retention times were 17.4 min for t-R and 14.1 min for t-P. The plant material extracts were analyzed by HPLC according to the methodology by Vitrac et al. (2002), using 0.1% TFA in water as solvent A and of 0.1% TFA in acetonitrile:water (80:20 v/v) as solvent B at a flow rate of 0.7 mL min −1 and an injection volume of 20 µL.

Experimental Design and Statistical Analyses
An analysis of variance (ANOVA) with factorial arrangement using Tukey's multiple comparison test, was used to analyze the statistical differences for the concentrations of growth regulators. The results were statistically significant at P<0.05, The SAS System 9.2 software for Windows ™ was used.

Callus Induction and Growth
Callus formation from fruit exocarp occurred 60 days after explants were established on induction media. Significant differences were observed among the callus induction treatments in mature exocarps, where all the explants of ecotypes 2 and 3 formed calli with T1. In the immature fruits, the best response in ecotype 2 was obtained in T4 and in ecotype 3 both in T3 and T4 (Table  2). Induction was uneven among ecotypes because not all explants were able to de-differentiate and produce a response. The morphological appearance of the formed calli was strikingly similar between explants; a friable cell mass with whitish coloration and brown tones (Fig. 1).

Selection of Highly Productive Lines from the Established Calli
T-Rproduction in the callus lines from the immature fruits was higher than those from mature fruits. The highest t-R concentrations in the immature fruits were found in cellular line T3EI3 with 553.4 µg t-R. g −1 FW, as opposed to the most productive line of the mature exocarp line T1EM3, with 76.19 µg t-R g -1 FW (Fig. 2). The highest t-P values were found in the cell line T3EI2, with 52.6 µg t-P g −1 FW; followed by line T1EI3 with 47.8 µg of t-P g −1 FW.

Cell Aggregates from the Suspensions Culture
The t-R concentration at the beginning of suspensions (day 0) was 1.89 µg t-R g −1 FW for line T4EI3 and 2.70 µg t-R g −1 FW for line T3EI3, which increased with a maximum of 54.26 µg t-R g −1 FW on day 14 in line T4EI3 and of 30.56 µg of t-R g −1 FW on day 21 in line T3EI3. The t-P concentration was higher than t-R in both lines, particularly on day 14, when it reached 54.03 µg of t-P g −1 FW for line T3EI3 and up to 95.47 µg of t-P g −1 FW in line T4EI3. The results agree with the growth profile of the cell aggregates, where an exponential phase was present between days 4 and 14 for line T4EI3 and between 6 and 12 days for line T3EI3. These remained steady between days 14 and 16, with 82% maximum cell viability for T3El3 and 70% for T4El3.

Quantification of t-R and t-P in the Plant Material
The t-R and t-P concentrations were also assessed in plant material, particularly in leaf and exocarp of the mature and immature fruits analyzed by HPLC, thus being the first publication to report resveratrol levels in different Ugni molinae ecotypes. The highest t-R concentrations were detected in fruits, with biggest differences found between ecotypes. The same occurred with calli. The best results were obtained from the exocarp of the immature fruits from ecotype 3, with a total of 5,100 µg t-R g −1 , followed by the exocarp of the mature fruits of ecotype 2 and 3 (2,750, 2,600 and 1,400 µg t-R g −1 ).  These values were higher than the in vitro material. Nevertheless, the values obtained from the mature and immature calli were higher than those found in leaves (60 µg t-R g −1 ), especially in the calli formed from the immature fruit (245.9 t-R µg g −1 callus).

Callus Induction and Growth
The combination of 1 mg L −1 of ANA + 1 mg L −1 of 2,4-D and 0.5 mg L −1 of KIN as a growth regulator for callus induction, has been used in other berry cultures; e.g., Vaccinium macrocarpon Ait. and Vaccinium phalae (ohelo) (Madhavi et al., 1995;Fang et al., 1999). This allowed the subsequent development of suspension culture to produce anthocianins. However, this has been the only species reported for callus induction from berry plants, thus, our study becomes the third report of callus culture from berries. Furthermore, is one of a few that developed callus using a fruit explant from immature and mature exocarps (adult material). As shown here, the synthesis and location of secondary metabolites may vary among plant tissues by using different plant parts. In the study by Liu et al. (2010), three plant tissues in four genotypes of Vitis vinifera were used to successfully form calli. However, the best results occurred in young leaves and seeds rather than in the exocarps of fruits, where the response was slower and even showed necrosis in some genotypes. This response is expected in almost all occasions, because calli (undifferentiated) is best formed from young material such as leaves and nodal segments. Instead our research form friable calli from adult material (fruits explants and leaves), noting that the use of the hormonal combination was adequate.

Selection of Highly Productive Lines from the Established Calli
Vitis vinifera has been the most widely studied plant species for tissue culture-mediated production of stilbenes and phenolic compounds (Cai et al., 2011), being trans-piceidthe stilbene with the highest concentration found, from the Red Globe cultivar the most productive line (69.9 µg t-P g −1 FW) (Santamaria et al., 2011b). However, very few studies have reported stilbene content in callus. In the Arachis hypogaea 'Tainan' cultivar, callus elicited with ultraviolet light and microorganisms showed t-R contents between 1.03 and 7.08 µg g −1 FW and t-P contents from 0.71 to 9.61 µg g −1 FW (Ku et al., 2005;Yang et al., 2010). When comparing these results with those obtained here (cell line T3El2, with 52.6 µg t-P g −1 FW), Ugni molinae gave very similar t-P values compared the most productive line from the Red Globe cultivar and in addition, much higher t-R content values (line T3El3 with 553.4 µg t-R g −1 FW) compared to other plant species. These results are very promising and revealed that highly productive lines can be obtained from calli.

Cell Aggregates from the Suspensions Culture
The biomass growth of U. molinae cell aggregates in liquid culture followed a behavior similar to that reported for cellular suspensions of Vitis spp., starting with an initial growth phase between days 2-4 of culture until day 12-14, when plant cells finally entered a steady state that ends day 16-18 (Martínez-Esteso et al., 2009;Santamaria et al., 2011a;2011b;Belchi-Narravo et al., 2012). The end of the cell division phase and the start of cell expansion characteristic of the steady phase, are associated with the synthesis and accumulation of specific secondary metabolites, including some alkaloids, anthocyanins and other phenolic derivatives (Lindsey and Jones, 1989). This may explain the increase in the piceid and t-Rcontents in aggregates as they primarily accumulate in cellular compartments as a storage form. Piceid has also been assessed in 'Gamay' Vitis viniferacell suspensions with the highest values found of 280 mg L −1 (Aumont et al., 2004). It is the most abundant stilbenoid and accumulates with the biomass in normal growth medium (Martínez-Esteso et al., 2011). When elicitors are added, resveratrol becomes highly abundant and is released into the culture medium (Morales et al., 1997;Bru et al., 2006;Donnez et al., 2009;Martínez-Esteso et al., 2009;. Cellular suspensions of Gossypium hirsutum have been developed as well for the in vitro t-R production, with values of 7.2 µg g −1 FW and also in roots of Arachis hipogea, with values of 2 µg g −1 and between 0.81 and 1.5 mg g −1 FW (Medina-Bolivar et al., 2007;Kouakou et al., 2006;Kim et al., 2008). Cellular suspensions in Ugni molinae or Myrtaceas species have never been reported, thus, this quantification is the first to be published. The results obtained are comparable with those reported for other species and could be further improved if elicitation strategies are implemented in the future.

Quantification of t-R and t-P in Plant Material
In Chile, resveratrol from in vivo material only has been quantified in Vitis vinifera (0.007 and 0.26 µmol g −1 DW) and in Berberis buxifolia (0.98 µmol g −1 and 3.87 µmol g −1 FW) (Ruiz et al., 2010). The results obtained here from murtilla fruits, exceeded the values of the genotypes formerly studied in grapes, that ranged from 39.71 to 300 µg g −1 (Tobar-Reyes et al., 2009;Liu et al., 2013), in blueberries (32 ng g −1 DW), peanut (1.92 µg g −1 DW) and pistachios (1.67 µg g −1 DW) (Lyons et al., 2003;Tokusoglu et al., 2005), being these values interesting for further research. In all these studies, concentrations varied depending on genotype and sample collection area. Some authors have explained that this phenomenon may be due to the attribution of abiotic or biotic stress factors from the surrounding environment, time of harvesting, climatic conditions, geographic origin, plant development and even crop type (organic, wild or agricultural) (Vitrac et al., 2002;Li et al., 2006). The comparative studies between in vivo and in vitro material (calli) to quantify secondary metabolites have been conducted for species like Vaccinium macrosporum and Buddleja cordata (Madhavi et al., 1995;Estrada-Zuniga et al., 2009). Presently, however, very few successful examples of the commercial application of cellular suspensions exist due to the low biosynthesis of the compound, differentiation and compartmentalization of cells, lack of developing organelles, unstable cellular lines and the difficulty of scaling production (Kolewe et al., 2008;Estrada-Zuniga et al., 2009).
Even when t-resveratrol production in U. molinae did not increase when cell suspensions were used, our results demonstrated that native species from Chile could be a biological reserve of resveratrol, which remains unknown and could be highly competitive with other currently used extraction sources. Further research is required to identify the variables that affect final tresveratrol and t-piceid production.

Conclusion
This study is the first to report presence of resveratrol from in vitro cultures of Ugni molinae from different explants. The callus formed from the fruit exocarp was used to establish cell suspensions, where amounts of tresveratrol and t-piceid were identified in cell aggregates. The most productive line was ecotype 3 when 5 mg L −1 of 2,4-D + 0.5 mg L −1 KIN and 1 mg L −1 of NAA were used, obtaining 553.4 µg t-R g −1 . The t-R production in immature fruits was higher than in mature callus. The highest t-P values were found in cell line T3EI2, with 52.6 µg t-P g −1 FW, followed by line T1EI3 with 47.8 µg of t-P g −1 FW. Differences among ecotypes were observed in all stages. The amounts of resveratrol found in the in vitro callus and field material from the fruit exocarp exceeded those reported for other plant species. This indicates that Ugni molinae is a productive source of resveratrol that was not known until now.