Culture of East Indian sandalwood tree somatic embryos in air-lift bioreactors for production of santalols, phenolics and arabinogalactan proteins

The Indian Sandalwood tree is globally acclaimed for the precious essential oil and heartwood. Over-exploitation, diseases, and habitat loss have posed significant challenges to find an alternative bioresource for biomass production. Here, we report the successful growth of in vitro grown somatic embryos in 10 L air-lift bioreactors. Additionally, we characterized arabinogalactan proteins and small molecule constituents such as phenolics and terpenoids that are secreted by the suspended somatic embryos into the culture media. In parallel to the biochemical characterisation, we followed the entire developmental progression of proembryogenic masses into matured cotyledonary embryos during a single run of the bioreactor.


Introduction
The East Indian sandalwood tree, Santalum album, is a woody and tropical forest tree that belongs to the family Santalaceae. The species is globally acclaimed for its very costly heartwood and essential oil obtained from matured individuals. The steam-distilled commercially available essential oil is rich in sesquiterpenoid constituents known as 'santalols', i.e. a-santalol and b-santalol together contributing to .90 % (Howes et al. 2004). Unmet anthropogenic needs for use in perfumery (fragrance) and as food additives (flavour) have led to the decline of natural populations due to illegal trade and harvesting, and overexploitation. The estimated global annual requirement is 10 000 tons of wood (equivalent to 200 tons of oil), involving a trade of approximately $125 million, of which only 10 % is met from natural resources. Its natural enemy in the form of the mycoplasmal 'spike disease' has led the species into the IUCN (International Union for Conservation of Nature and Natural Resources) Red List of Threatened Species (IUCN 2012). Sandalwood finds extensive applications in traditional medicinal systems such as Ayurveda, and is gaining increasing importance in modern pharmacological investigations as a source of anticancer (Bommareddy et al. 2012), anti-Helicobacter pylori (Ochi et al. 2005) and antiviral (Benencia and Courreges 1999) biomolecules. However, the havoc caused by the epidemic spike disease and the hemiparasitic and slow-growing nature of the tree necessitated research towards the development of biotechnological means of in vitro production as early as 1963, i.e. callusing (Rangaswamy and Rao 1963). Further efforts on somatic embryo (SE) production (Bapat et al. 1990) and maturation in air-lift bioreactors (Das et al. 1998(Das et al. , 1999 have also been successful. Plant cell cultures have yielded valuable natural products in the form of pharmaceuticals, flavours and fragrances, and agricultural, cosmetic, bioherbicidal and fine chemicals, with 2000 new plant chemicals added annually. The global market for plant-derived drugs was worth an estimated US$18 billion in 2005, with an expected annual growth rate of 6.6 % to US$26 billion by 2011, with the USA accounting for 50 % of the global plant-derived drug market (Saklani and Kutty 2008). These phytomolecules are proven immunomodulating, antiviral, antimicrobial, antiparasite, antitumour, anti-inflammatory, hypoglycaemic, tranquillizer and antifeedant agents. Furthermore, the homogeneous and synchronous nature of cell suspension cultures makes them amenable to large-scale production of phytochemicals. Successful reports include those from Japan on the production of shikonin, ginseng and berberine at commercial scales using bioreactors (Misawa 1991). In 1984, shikonin was the first product that was produced from cell cultures of Lithospermum erythrorhizon in 750-L bioreactors by Mitsui Petrochemical Co., Japan (Tabata and Fujita 1985). Advances made towards commercialization of plant cell culture processes for the synthesis of biomolecules have been reviewed recently (Wilson and Roberts 2012). To this end, excellent reviews have been published underscoring the efficient production of secondary metabolites (Kolewe et al. 2008;Weathers et al. 2010) and recombinant proteins (Huang and McDonald 2009), with an emphasis on the development of suitable bioreactor configurations for plant cell culture-based processes (Georgiev et al. 2009). Additionally, somatic embryogenesis offers a potential system for large-scale plant propagation in automated bioreactors (Paek et al. 2005). Furthermore, disposable bioreactors have been used for micropropagation of plant tissues, undifferentiated bioactive cells and for expression of secondary metabolites and glycoproteins (Eibl et al. 2009). Similarly, successful product/platform pairs leading to synergies during production and in clinical trials have also been reviewed (Paul and Ma 2011). Additionally, polyphenols and isoprenoids have been identified as major classes of antioxidants produced by the plant cells in vitro and were reviewed (Matkowski 2008). Similarly, Xu et al. (2011) have identified the use of arabinogalactan proteins (AGPs) as 'tags' for facilitation of production of recombinant pharmaceutical proteins in plant cell suspension cultures on a large scale. Additionally, somatic embryogenesis offers a potential system for large-scale plant propagation in automated bioreactors (Paek et al. 2005).
Thus, to meet the challenges of the commercial viability of S. album, a protocol with high and consistent production of SEs is necessary. To this end, a bioreactor-based production system using a liquid medium was developed (Das et al. 1998). However, the scale-up study reported earlier (Das et al. 1999) did not address the detailed investigation of biochemicals. In this study, we compare the culture of SEs in 2-and 10-L air-lift bioreactors as a biotechnological means of propagation for this endangered species in terms of the growth and biochemical characteristics.

Bioreactors
A 2-L-volume air-lift bioreactor (Bio Stat, B. Braun Biotech. International, Lancaster, PA, USA) and a 10-L culture vessel (Cat. No.: 2227-0020, Nalgene Nunc TM International, Rochester, NY, USA) were used in this investigation. The 10-L culture vessel (Nalgene) was modified to perform as an air-lift-type bioreactor. For comparison, the earlier reported 2-L (Bio Stat, B. Braun) air-lift bioreactor (Pal et al. 2003) was studied for growth of sandalwood SEs. For the 10-L vessel we used a modified internal draft tube of 20 mm diameter attached to silicone tubing for air passage through 0.21-mm filters as inlet and outlet. The constant aeration sources were two oil-free air compressors. Agitation was maintained at a constant rate, i.e. 4 L min 21 of air flow. Air-lift through an internal draft tube allowed the cultures to agitate at 60-80 rpm. The ambient temperature of the laboratory was maintained at 25 + 2 8C. The light conditions were set at 3000 lux (16-h light/8-h dark cycles). The data presented in this study are the outcome of four independent runs conducted over several weeks.

Plant materials
Callus developed from a highly proliferating cell line (IITKGP/91) was maintained in the laboratory as suspension cultures in liquid media, i.e. WPM (Woody Plant Medium) media (Lloyd and McCown 1981) supplemented with 100 mg L 21 myo-inositol, 1 mg L 21 (4.52 mM) 2,4-dichlorophenoxyacetic acid and 3 % sucrose. Cultures were maintained by subculturing 2-week-old suspensions by transferring each culture into five 250-mL-volume Erlenmeyer flasks containing 50 mL of fresh media.

Preparation of inoculum
The bioreactors carrying 1.5 and 7.5 L of WPM media (as described in the 'Plant materials' subsection) were sterilized in vertical autoclaves for 35 and 45 min, respectively, at 121 8C and 15-psi pressures. The rest of the media were sterilized separately and allowed to cool down, and fresh inoculum was mixed with sterile media in a laminar air flow cabinet. Two-week-old suspension cultures yielding the proembryogenic mass (PEM) of cells were used as the inoculum for both the air-lift-type bioreactors at a dose of 2 % (v/v) fresh weight (FW).

Growth rate and biomass estimation
The specific growth rate (m) and doubling time (t d ) of SE suspension culture were calculated with the following formulae: m ¼ (ln DW 2 2 ln DW 1 )/(T 2 2 T 1 ) and t d ¼ ln 2m 21 , where DW 1 and DW 2 are corresponding dry weight masses in days and during exponential growth, and T1 and T2 are the culture initiation and harvest days in numbers.
For FW determination, the suspensions were filtered with a Whatman No. 1 filter paper, gently pressed on filter papers to remove excess water and weighed. Cells were dried in an oven at 60 8C to a constant weight for the determination of cell DW. The initial culture volumes were 2 and 10 L, respectively, for the B. Braun and the Nalgene air-lift reactors ( Fig. 1A and B) [see Supporting Information -Videos]. Growth in terms of biomass was expressed as FW and DW upon harvesting at the end of 28 days, for the 2-and 10-L bioreactors, respectively.

Determination of SE counts
Although the embryogenic suspension also contained embryos of other stages, the majority were globular stage embryos. Data were recorded on the basis of number of globular embryos per 50 mL of medium produced at the end of 14 days of culture. The number of globular embryos was counted manually for representative smaller aliquots under a Leica Wild (M3Z) (Leica, Wetzlar, Germany) stereomicroscope.

Extraction of santalols from extracellular medium
The sesquiterpenoids were extracted following protocols described elsewhere (Harborne 1998) and were further analysed by high-performance thin-layer chromatography (HPTLC).

HPTLC analysis for santalols
Sesquiterpenoids from the suspension culture extracts were analysed by HPTLC as described previously (Misra and Dey 2012). Sesquiterpenoids were identified in comparison with constituents obtained from authentic steam-distilled sandalwood oil samples from Cauvery TM (Government of Karnataka, India).

Extraction of phenolics from extracellular medium
The extracellular medium (ECM) was collected, filtered with four layers of miracloth, then with laboratory-grade filter paper and finally with Whatman No. 1 filter paper, and then centrifuged at 5000 g for 20 min to remove the particulate matter. This was repeatedly extracted three times with diethyl ether using a separatory funnel in a AoB PLANTS www.aobplants.oxfordjournals.org ratio of 1 : 1 of solvent and ECM. The diethyl ether fraction was evaporated to dryness under vacuum at 37 8C and redissolved in 50 % aqueous methanol to yield the phenolics-rich fraction. The ECM thus extracted was saponified with 2 N NaOH at pH 2 for 24 h to release the ester-and ether-linked phenolics, neutralized with 2 N HCl, and then re-extracted as above and pooled together for quantification as well as reverse-phase high-pressure liquid chromatography (RP-HPLC) analysis.

Folin -Ciocalteau assay for quantification of phenolics
The total phenolic content of the ECM was quantified as described previously (Julkunen-Tiitto et al. 2005). Phenolic content was determined using the Folin-Ciocalteau assay and a standard curve prepared with gallic acid.

RP-HPLC of phenolics
Reverse-phase high-pressure liquid chromatography analyses were performed on a Phenomenex TM (Torrance, CA, USA) C18 column (RP-Hydro, 4 mm, 250 × 4.6 mm) using a Waters HPLC system (Milford, PA, USA) equipped with a dual-absorbance UVdetector. Chromatograms were monitored simultaneously at 254 and 310 nm and analysed on a Windows XP TM platform with BREEZE TM software version 3.2 (Waters), and separations were achieved following the method of Sachan et al. (2004). The phenolic acids were identified by comparing their retention times with those from authentic standards and co-elution experiments.

Extraction and quantification of AGPs in ECM
After clarification of the ECM obtained from suspension cultures by high-speed centrifugation, AGPs were precipitated by excess Yariv reagent in 2 % v/v NaCl. The Yariv-AGP complex was then dissociated via reductive cleavage of the diazo linkage with sodium dithionite, followed by dialysis and freeze-drying as described previously (Pal et al. 2003). Arabinogalactan protein was determined using b-glucosyl Yariv reagent binding in a single radial gel diffusion assay (Kitazawa et al. 2013). Yariv reagent was synthesized in the authors' laboratory as described previously (Pal et al. 2003).

Histology and microscopy of SEs
The SEs collected after final harvest were observed under a Wild (M3Z) (Leica, Wetzlar, Germany) stereomicroscope at ×20 magnification and compared. Wiesner staining of whole SEs for the detection of lignin was performed in 1 % phloroglucinol (prepared in 92 % ethanol) for 2 min (Pomar et al. 2002).

Statistical analyses
Statistical analysis was performed with the SPSS software package (version 17) (SPSS Inc., Chicago, IL, USA). Statistical analyses were performed by Student's t-test.  P values less than 0.05 (*) and 0.01 (**) were considered to be statistically significant. Results were expressed as mean + SD (error bars) for data obtained from four independent batches of bioreactor run.

Yield of biomass from the bioreactors
There were some heart and torpedo stage embryos in most of the batches (,1 % of the number of globular embryos) and these were neglected in the counts for normal and abnormal embryos. The total SE yield in terms of count was 3200 and 5500 L 21 of medium for the 2-and 10-L reactors, respectively. Post-cultivation, the media volumes of the 2-and 10-L reactors were reduced to 65 and 75 % of starting culture volumes, respectively. In addition, the pH had decreased to 4.2 -4.6 at the time of final harvest. The biomass yields (both in FW and DW) were 2.9-to 3-fold higher for the 10-L bioreactor compared with the 2-L bioreactor. The specific growth rate (m) was 0.0675 and 0.0549 day 21 , and the doubling time (t d ) was 10.26 and 12.62 days, for the 2-and 10-L air-lift reactors, respectively (Table 1).

RP-HPLC-based identification of ECM phenolics
Various types of phenolics obtained from the ECM of SEs were identified by RP-HPLC (Table 2). Air-lift bioreactor-grown SE culture ECM showed the presence of several phenolics, i.e. 4-hydroxybenzyl alcohol, 4-hydroxybenzaldehyde, and protocatechuic, p-coumaric and ferulic acids. Only vanillin was detected exclusively in the ECM of the 10-L bioreactor. Moreover, the RP-HPLC chromatograms indicated that 15 -19 phenolic constituents remained unidentified in both the ECM samples ( Fig. 2A and B).

Total yield of biomolecules (santalols, phenolics and AGP) from the ECM
High-performance thin-layerchromatography-based quantification indicated that the santalol yield from the ECM was comparable for both the air-lift reactors. In addition, results obtained from single radial agar diffusion assays ( Fig. 4A and B) of the ECM demonstrated AGP   yields of 33 -39 mg L 21 for both the air-lift bioreactors. Furthermore, total yields of phenolics from ECM were quantified at 29 -31 mg gallic acid equivalence L 21 .

Morphological evaluation of SEs
Single-cell suspension cultures (Fig. 5A) from shake-flasks were the starting material for induction of embryogenesis to obtain the embryogenic inoculum. Matured SEs in various stages of development were obtained after 28 days culture of suspended PEMs in cultivation medium (Fig. 5B), although the majority were in globular stages. Different stages of SE development were observed in the final harvest, i.e. globular (Fig. 5C), heart-shaped (Fig. 5D) and torpedo stages (Fig. 5E), while the heart stage embryos were less frequent than others. The mature dicotyledonary embryos were 2 -5 mm long (Fig. 5F). Thus normal SEs representative of a progressive somatic embryogenesis process were obtained at a high frequency. It was observed that the Wiesner (phloroglucinol-HCl) staining was located towards the root poles of the torpedostage matured SEs (Fig. 5G).

Discussion
Air-lift bioreactors have wide applications due to their unique hydrodynamic characteristics, better oxygen transfer, good growth, reasonable mixing, less contamination due to the absence of moving parts and low operating cost due to simple design. Moreover, large-scale SE culture is an attractive alternative to the conventional method of plantation or cell culture. The use of large-scale liquid cultures and automation offers the potential to resolve manual handling of the various stages of micropropagation, thereby decreasing the production cost significantly. In addition, in vitro-grown plants are free from the effects of seasonal variations, microbial infestations and soil-borne contaminants that can affect the medicinal value of the harvested tissues. The successful production of SEs in bioreactors has been reported in a number of species (Paek et al. 2005). The growth and accumulation of secondary metabolites in cell cultures in bioreactors are influenced by various physical and chemical factors, like oxygen supply. Given that a metabolite is growth associated, a single-step reactor is sufficient to grow the cells and recover the molecules at the same time.
This investigation focused on the large-scale biomass production of sandalwood SEs with concomitant production of biomolecules of industrial importance. Therefore, we used a 10-L culture vessel as an air-lift bioreactor and compared the biomass and metabolite yield with that of a 2-L bioreactor. The simplicity of the 10-L air-lift bioreactor used in this study results in the low cost and efficient production of viable and matured SEs ranging from heart to torpedo stages. Furthermore, efficient top-to-bottom mixing makes this 10-L culture vessel and mixer system ideal for suspension cell culture applications. In fact, the agitation had a less pronounced effect on the 10-L vessel than the 2-L vessel, attributed to a larger culture volume, and hence the larger vessel was less prone to shearing stress (data not shown). The lowered pH of the ECM at harvest suggested the possible release of phenolics and sugar acids into the ECM during the growth phase, thereby influencing the pH. This investigation is a supplemental study on the production of peroxidase and AGP as by-products from SE cultivation in a 2-L air-lift bioreactor (Pal et al. 2003). Hence, we have not only obtained similar results to those reported before, but have also shown for the first time the production of santalol in vitro by means of HPTLC analysis. Furthermore, we demonstrated the production of phenolics in the ECM and confirmed their identities by RP-HPLC analysis.
In fact, air-lifts are generally considered unsuitable for high-density (.20 g L 21 DW) plant cell cultures due to the serious problems encountered during mixing and  AoB PLANTS www.aobplants.oxfordjournals.org oxygen transfer (Dornenburg and Knorr 1995). However, we report SE biomass densities of 8.2-14 g L 21 by DW in this investigation, for both the bioreactors. Additionally, we show the accumulation of sesquiterpenoid alcohols in the ECM of SEcultures for the first time. In fact, the scale-up of plant cell suspensions to large culture volumes while maintaining their biosynthetic potential has been recognized as critical (Godoy-Hernandez et al. 2000).
In this investigation, the cell growth and biomass accumulation increased gradually with time and maximum biomass levels were reached after 28 days. Apart from the biomass, we screened the ECM for several metabolites such as phenolics, santalols and AGPs. In fact, phenolic acids are implicated in embryo development and are involved in alterations of the cell wall composition during differentiation and morphogenesis (Cvikrova et al. 1998). Additionally, hydroxycinnamates are involved in alterations of the cell wall composition (Lozovaya et al. 1996). Interestingly, we report the presence of hydroxycinnamates such as ferulic, protocatechuic and p-coumaric acids in the ECM of both the bioreactors, thus implying their role in SE development and maturation. Moreover, the presence of vanillin in the ECM of a 10-L-capacity bioreactor is notable. The Wiesner stain is known to react with cinnamaldehyde residues in lignin and the colour intensity roughly reflected the total lignin content. The observations in this study indicated lignin development towards the roots of SEs, thereby suggesting normal development of SEs. The staining reaction also indicated the biosynthesis of cinnamaldehydes in the SEs.
We further demonstrated the accumulation of AGPs in the ECM. Arabinogalactan proteins are cell wall proteoglycans that bind specifically to a synthetic probe, Yariv phenylglycoside, and are located at the plasma membrane, cell wall and in the media of cell cultures. Arabinogalactan proteins are present in both the periplasm and growth medium. Moreover, AGPs are known to promote somatic embryogenesis in cotton cell cultures (Poon et al. 2012). In fact, conditioned ECM is known to stimulate cell growth and accumulation of metabolites, such as anthocyanin production by strawberry suspension cells (Sakurai and Mori 1996). Probably, in the case of sandalwood SE suspension cultures in air-lift bioreactors, the inoculum carrying some amount of conditioned medium may be responsible for inducing the secretion of AGPs, santalols and phenolics.

Conclusions
The most relevant and critical finding of this study is the development of a large-scale cultivation protocol optimized for a 10-L-volume bioreactor system. Secondary metabolites are currently being obtained commercially by extraction from whole plants or tissues. Moreover, cultured multipotent cambial meristematic cells have been identified as potential resources for the largescale production of certain natural products (Roberts and Kolewe 2010). Large-scale SE culture in a bioreactor may provide an attractive alternative to the traditional method of sandalwood plantation. Although our results do not show significant increases in terms of metabolite production, the greater yield of biomass in 10-L bioreactors does indicate the commercial feasibility of the process for production of sandalwood biomolecules. Moreover, as a general rule, and due to their genetic stability, SEs are less prone to erratic metabolite production and display a lower sensitivity to shear stress than undifferentiated cells. Our results indicate that an efficient protocol for the mass production of sandalwood biomass can be achieved by a bioreactor-based cultivation of SEs and can simultaneously be used as a source of raw medicinal by-products of industrial importance.

Sources of Funding
Funding for the presented investigation was obtained from the Ministry of Human Resources Development (MHRD), Government of India, India.

Conflicts of Interest Statement
None declared.

Supporting Information
The following Supporting Information is available in the online version of this article -File 1. Movie. Demonstrates the 2-L air-lift bioreactor (B. Braun) under running conditions supporting the growth of sandalwood somatic embryos.
File 2. Movie. Demonstrates the 10-L air-lift bioreactor (Nalgene) under running conditions supporting the growth of sandalwood somatic embryos.