Colour bio-factories: Towards scale-up production of anthocyanins in plant cell cultures

Anthocyanins are widely distributed, glycosylated, water-soluble plant pigments, which give many fruits and flowers their red, purple or blue colouration. Their beneficial effects in a dietary context have encouraged increasing use of anthocyanins as natural colourants in the food and cosmetic industries. However, the limited availability and diversity of anthocyanins commercially have initiated searches for alternative sources of these natural colourants. In plants, high-level production of secondary metabolites, such as anthocyanins, can be achieved by engineering of regulatory genes as well as genes encoding biosynthetic enzymes. We have used tobacco lines which constitutively produce high levels of cyanidin 3-O-rutinoside, delphinidin 3-O-rutinoside or a novel anthocyanin, acylated cyanidin 3-O-(coumaroyl) rutinoside to generate cell suspension cultures. The cell lines are stable in their production rates and superior to conventional plant cell cultures. Scale-up of anthocyanin production in small scale fermenters has been demonstrated. The cell cultures have also proven to be a suitable system for production of 13C-labelled anthocyanins. Our method for anthocyanin production is transferable to other plant species, such as Arabidopsis thaliana, demonstrating the potential of this approach for making a wide range of highly-decorated anthocyanins. The tobacco cell cultures represent a customisable and sustainable alternative to conventional anthocyanin production platforms and have considerable potential for use in industrial and medical applications of anthocyanins.


Generation of suspension cultures
Cell cultures were generated as described in (Mustafa et al., 2011) with the following modifications. Callus formation was induced from darkly pigmented young leaves for approximately 21-28 days on MS agar plates with high auxin to low cytokinin ratio (induction medium, the composition of all media is given in Supplementary Table 1).
Explants with developing callus were sub-cultured 2-3 times for 14 days on the same plates.
Friable callus was removed from the explants and subsequently transferred to MS plates with reduced phytohormone concentrations (transition medium) and sub-cultured 4-6 times for 14 days. Thereafter, soft callus was kept and continuously sub-cultured for 14 days on MS agar without cytokinins and elevated myo-Inositol concentration (tobacco growth medium). In some cases, callus was transferred directly from induction plates to the final medium, dependent on the formation of soft callus. Cell suspensions were initiated from friable callus and grown as batch cultures at 23°C in the dark at 95 rpm in a rotary shaker (New BrunswickTM Innova® 2000), and were sub-cultured every 7 days.

LC-UV/MS and LC-MS/MS analysis of anthocyanins
For LC-UV/MS analysis, 100 µL of the extract were mixed with 25 µL of solvent A (water + 0.5% formic acid) and incubated for one hour at 4°C prior to injection to allow for formation of any potential precipitates. After centrifugation, 5µL of the supernatant were injected per sample. Anthocyanins were analysed by LC-UV/MS as described in Oertel et al. (2017), using an ultra-performance liquid chromatography (UPLC) instrument (H-CLASS, Waters, Milford, MA, USA) with photo diode array (PDA) detection coupled to an ultra-high resolution time of flight mass spectrometer (UHR-TOF-MS, maXis Impact, Bruker Daltonics, Bremen, Germany) for MS detection. Separation was performed on a CSH Phenyl-Hexyl column (2.1x100mm, 1.7 µm, Waters) equipped with an CSH Phenyl-Hexyl VanGuard pre-column (130 Å, 2.1 x 5 mm, 1.7 µm, Waters) using a gradient from 2% solvent B (acetonitrile + 0.5% formic acid) to 20% solvent B over 4 minutes, followed by a cleaning step from 20% to 98% solvent B over 3 minutes. Column temperature was 35°C and solvent flow rate was 500 μL min -1 . UV spectra were acquired from 210 nm to 800 nm. MS were detected after electrospray ionization (ESI) in positive ion mode at 220°C dry temperature, four bar nebulizer gas pressure, 4000 V capillary voltage and a dry gas flow of 11 L min -1 . The instrument was operated in MS full scan mode at 3 Hz acquisition speed using the following settings: mass range m/z 50-1500, hexapole radio frequency (RF) voltage 100 V peak-topeak (Vpp), collision energy 8 V, funnel 1 RF 300 Vpp, funnel 2 RF 600 Vpp, prepulse storage time 15 μs, transfer time 50 μs and collision cell RF 500 Vpp. Instrument calibration was performed with 10 mM sodium formate solution (12.5 mL H2O, 12.5 mL isopropanol, 50 µL concentrated formic acid, and 250 µL 1M NaOH).
For MS/MS fragmentation analyses the instrument was operated in MS full scan mode and detected by collision-induced dissociation (CID) with argon at 1.5 mTorr in the quadrupole collision cell of the instrument using the following settings: molecular mass 300 Da (width 5 Da) with 27, 22, and 17 eV for charge state 1, 2, and 3 respectively; molecular mass 500 Da (width 6 Da) with 32, 27, and 22 eV for charge state 1, 2, and 3 respectively; molecular mass 1000 Da (width 8 Da) with 45, 40, and 35 eV for charge state 1, 2, and 3 respectively; molecular mass 1500 Da (width 10 Da) with 45, 40, and 35 eV for charge state 1, 2, and 3 respectively.

Preparative isolation of anthocyanins from tobacco suspension cultures
Freeze dried cell suspension cultures were ground finely and sequentially extracted with 80% methanol, 2% formic acid (FA) to a final ratio of 100 mL solvent g -1 cell dry weight (DW).
Insoluble material was removed by centrifugation for 30 min at 13,000 x g. The resulting supernatant was sequentially filtered through 100 µm and 20 µm nylon net filters. The volume of the filtered extract was reduced under vacuum using a rotary evaporator.
Degreasing was undertaken by phase separation with two volumes (v/v) of n-heptane. Any remaining organic solvent was removed under vacuum using a rotary evaporator. Prior to chromatography, the degreased extract was sequentially filtered through nylon net filters down to 0.2 µm. Filtered extract was loaded on a solid phase extraction column (RediSep RF Supplementary Table 2 Isotope pattern. Comparison of isotope pattern of C3R from control and 13 C-sucrose-labelled cell cultures. For evaluation of the carbon status we extracted the sum of mass spectra across the chromatographic peak for C3R at a retention time of 3.8 minutes for both samples (control and 13 C-sucrose). The resulting mass spectra were then compared regarding the isotopic pattern of C3R (chemical formula of the cation C27H31O15 + ). The specified carbon state neglects the incorporation of naturally low abundance 2 H (0.0115% relative abundance) and 3 H (<0% relative abundance), as well as 17 O (0.038% relative abundance) and 18 O (0.205% relative abundance). The relative natural abundancies for carbon are 98.93% ( 12 C), 1.07% ( 13 C) and <0% ( 14 C).