Provitamin A biofortification of cassava enhances shelf life but reduces dry matter content of storage roots due to altered carbon partitioning into starch

Summary Storage roots of cassava (Manihot esculenta Crantz), a major subsistence crop of sub‐Saharan Africa, are calorie rich but deficient in essential micronutrients, including provitamin A β‐carotene. In this study, β‐carotene concentrations in cassava storage roots were enhanced by co‐expression of transgenes for deoxy‐d‐xylulose‐5‐phosphate synthase (DXS) and bacterial phytoene synthase (crtB), mediated by the patatin‐type 1 promoter. Storage roots harvested from field‐grown plants accumulated carotenoids to ≤50 μg/g DW, 15‐ to 20‐fold increases relative to roots from nontransgenic plants. Approximately 85%–90% of these carotenoids accumulated as all‐trans‐β‐carotene, the most nutritionally efficacious carotenoid. β‐Carotene‐accumulating storage roots displayed delayed onset of postharvest physiological deterioration, a major constraint limiting utilization of cassava products. Large metabolite changes were detected in β‐carotene‐enhanced storage roots. Most significantly, an inverse correlation was observed between β‐carotene and dry matter content, with reductions of 50%–60% of dry matter content in the highest carotenoid‐accumulating storage roots of different cultivars. Further analysis confirmed a concomitant reduction in starch content and increased levels of total fatty acids, triacylglycerols, soluble sugars and abscisic acid. Potato engineered to co‐express DXS and crtB displayed a similar correlation between β‐carotene accumulation, reduced dry matter and starch content and elevated oil and soluble sugars in tubers. Transcriptome analyses revealed a reduced expression of genes involved in starch biosynthesis including ADP‐glucose pyrophosphorylase genes in transgenic, carotene‐accumulating cassava roots relative to nontransgenic roots. These findings highlight unintended metabolic consequences of provitamin A biofortification of starch‐rich organs and point to strategies for redirecting metabolic flux to restore starch production.

binary vector was prepared by cloning the patatin promoter-crtB-3' nos UTR cassette into the Mlu I site and the patatin promoter-DXS-3' nos UTR cassette into the Asc I site of pKAN2 to generate DXS//PS-pKAN2 (hereafter named DXS//PS). A second plant expression vector was prepared that contained only the crtB cassette cloned into the MluI site of pKAN2 to generate PS-pKAN2 (hereafter named as PS). The DXS//PS and the PS constructs were used to generate transgenic lines in cultivar 60444 background. For recovery of transgenic lines in TME 7 (Okoiyawo), TME 7S, and TME 204 the same DXS and crtB expression cassettes described above were moved to a binary vector p5000 (Beyene et al., 2016a;Beyene et al., 2016b) to generate a construct named pEC20. The p5000 is a modified pCAMBIA2300. A large number of provitamin A accumulating lines in TME 204 cultivar were also generated using a construct designated p8001 (Chauhan et al., 2015) that combined expression cassettes of DXS and crtB for provitamin A accumulation and an inverted repeats of coat protein genes of Ugandan cassava brown streak virus (UCBSV) and Cassava brown streak virus (CBSV) fused in tandem. The p8108 was a variation of the pEC20 described above except that the nos terminator used in both crtB and DXS cassettes was replaced with a 425 bp homologues UTR obtained from the patatin type-I gene 3' UTR. The p8108 was used generate transgenic potato lines.
The plant expression vector containing the mutant cauliflower Orange (Or) genes was the same as previously described for potato carotene biofortification (Lu et al., 2006). The mutant Or gene in this vector is under control of the potato granule-bound starch synthase promoter. All binary vectors were electroporated into Agrobacterium strain LBA4404 and used for transformation of cassava cultivars.

Production of transgenic cassava plants and establishment in greenhouse and field
Transgenic lines were produced by Agrobacterium-mediated transformation of friable embryogenic callus produced from cassava cultivars 60444, TME 7S, TME 204 and Oko-iyawo (TME 7) (Chauhan et al., 2015;Taylor et al., 2012). Regenerated plants were tested for presence and expression of the target transgenes by PCR and RT-PCR respectively using primer listed in Table S1. Lines which tested positive for presence and expression of the transgenes were micropropagated established and grown under greenhouse as previously described (Beyene et al., 2016a;Taylor et al., 2012). Plants were harvested and evaluated after 12-16 weeks. To establish plants in the field in vitro-derived plants were established in 50 mL Falcon tubes (Ogwok et al. 2012) and shipped to the University of Puerto Rico, Mayaguez, USA. Upon arrival plants were transferred to soil and acclimated for 4 weeks under greenhouse and another 4 weeks in screenhouse before establishment in the field.

Production of transgenic potato lines
In vitro stock plants were used as a source of material for transformation of Solanum tuberosum Desiree with Agrobacterium tumefaciens LBA4404 containing the p8108 construct according to methods described by Van Eck et al. (2007). The reported method has since been modified with the substitution of 300 mg/l timentin (Gold Biotechnology, Inc., St. Louis, MO) for carbenicillin.
Transgenic lines were verified by PCR analysis for the presence of nptII. For PCR analysis DNA was isolated from leaf material by homogenization in a buffer (0.2M Tris, 0.25M NaCl, 25 mM EDTA, 5 mg/ml SDS) and precipitation in isopropanol. The resultant pellet was washed in 70% ethanol and air dried. Primers used for PCR detection of nptII are listed in Table S1. Ten nptIIpositive lines were transferred to the greenhouse. For acclimation to greenhouse conditions, 6week-old in vitro plants were removed from the selective rooting medium, the medium was washed from the roots, and the plants were placed in 4-inch plastic pots containing Metro-Mix 360 (Griffin Greenhouse Supply, Auburn, NY). The plants were immediately covered with small, transparent plastic containers as they were transferred to soil and kept in a growth chamber for 1 week. The containers were removed and the plants were kept in the chamber for one additional week before being transferred to a greenhouse where they were maintained at 20 to 22 o C on a 16-hr photoperiod. Approximately 1 month after the acclimation period, plants were transferred to 3-gallon pots containing Metro-Mix 360. Tubers were harvested 14 weeks after plants were transferred to the greenhouse. Upon harvest tubers were separated into size categories (small, medium and large) and representative samples of tubers from medium and large size categorizes were peeled, chopped and immediately frozen for further analysis. Dry matter content, total carotenoid, starch, glucose, sucrose and fatty acids were determined following the same procedure described for cassava storage roots.

Field evaluation of transgenic cassava lines
Transgenic 60444 lines and TME 204 lines co-expressing crtB and DXS genes, were field tested at

Fatty acid and triacylglycerol analyses
Approximately 100 mg of lyophilized cassava storage root flour was extracted in 3 mL of chloroform:methanol (1:2 v/v) supplemented with 500 µg of triheptadecanoin (17:0triacylglycerol, Nu Chek Prep) as an internal standard in a 13 X 100 mm glass screw cap tube.
Following 30 min of incubation with agitation on a nutating mixer, 1 mL of chloroform and 1.8 mL of water was added to each sample and mixed thoroughly. Following centrifugation (1,000 g for 5 min), the lower lipid phase was transferred to another glass tube and dried under nitrogen.
The lipid extract was resuspended in 100 µL of chloroform:methanol (6:1 v/v), and 25 µL of the extract dried under nitrogen and transesterified in 1 ml of 2.5% sulfuric acid/methanol (v/v) and 250 µL of toluene (Msanne et al., 2012) for measurement of total fatty acids. The remainder of the extract (75 µL) was dried under nitrogen and redissolved in 1 mL of heptane for purification of triacylglycerols using a procedure similar to that previously described (Zhu et al., 2016). The extract was applied to a 3 ml Supelco LC-Si solid phase extraction column equilibrated in heptane.
The column was eluted with 500 µl of heptane, followed by 1 mL of heptane:diethyl ether (95:5 v/v). The column was then eluted with 2.5 mL of heptane:diethyl ether (80:20 v/v) and collected in a glass screw cap test tube. This fraction, containing eluted triacylglycerols, was dried under nitrogen and transesterified in 1 mL of 2.5% sulfuric acid/methanol (v/v) and 250 µl of toluene.
Samples for transesterification were heated at 95 °C for 30 min in glass test tubes capped under nitrogen. The resulting fatty acid methyl esters were recovered and analyzed by gas chromatography with flame ionization detection as described (Msanne et al., 2012). Fatty acid methyl esters were quantified in the total lipid and triacylglycerol fractions by comparison of sample peak areas relative to that of methyl heptadecanoic acid from the internal standard.

HPLC conditions for carotenoid analyses
Analyses were conducted using an Agilent 1200 HPLC with a binary pump and detection by absorbance at 455 nm using a diode array detector. Separation of carotenoid species was achieved using a C30 ProntoSIL column (250 mm length, 4.6 mm inner diameter, 5 µm particle size) or a C30 Dionex Acclaim column (150 mm length, 4.6 mm inner diameter, 3 µm particle size) with an isocratic solvent system consisting of 80% methanol/20% methyl-tert-butyl ether at a flow rate of 1.4 ml/min, essentially as described (Rodriguez-Amaya and Kimura, 2004).  Table S2. Selected differentially expressed genes associated with generation of carbon precursors for fatty acid biosynthesis (pyruvate dehydrogenase, acetyl-CoA carboxylase), de novo fatty acid biosynthesis (β-ketoacyl acyl carrier protein synthase I, III), or fatty acid storage as triacylglycerols (diacylglycerol acyltransferases 1, 2, 3) between provitamin A accumulating lines (DXS//PS-20 and DXS//PS-37) and wild-type controls of cassava storage roots at 12 months after planting. Fold-changes are presented relative to the wild-type control. Cassava orthologs of Arabidopsis genes are presented with TAIR IDs.      . Total carotenoid and dry matter content in transgenic p8108 potato lines co-expressing crtB and DXS. (a) total carotenoids and, (b) correlation between total carotenoids and dry matter content. Transgenic potato lines were generated using p8108 that harbors the crtB and DXS transgenes each driven by patatin type I promoter and the homologues untranslated region for transcript termination. Data were generated from tubers harvested at 14 weeks after planting in the greenhouse. Bars are SE of 4-6 biological replicates per line.