Metabolomic and transcriptomic analyses identify external conditions and key genes underlying high levels of toxic glycoalkaloids in tubers of stress-sensitive potato cultivars

Introduction High levels of toxic steroidal glycoalkaloids (SGAs) in potato tubers constitute a recognized food quality problem. Tuber SGA levels vary between potato cultivars and can increase after post-harvest stresses such as wounding and light exposure. A few cultivars, e.g., ‘Magnum Bonum’ and ‘Lenape,’ have been withdrawn from commercial sales due to excessive SGA levels during some cultivation years. However, these sudden SGA increases are diffucult to predict, and their causes are not understood. To identify external and genetic factors that underlie sudden SGA increases in certain potato cultivars, we have here in a 2-year study investigated ‘Magnum Bonum’ and five additional table potato cultivars for their SGA levels after wounding and light exposure. Results and methods Results showed that ‘Magnum Bonum’ has an unusual strong SGA response to light exposure, but not to wounding, whereas ‘Bintje’ displayed an opposite regulation. Levels of calystegine alkaloids were not significantly altered by treatments, implicating independent metabolic regulation of SGA and calystegine levels also under conditions of high SGA accumulation. Metabolomic and transcriptomic analyses identified a small number of key genes whose expression correlated with SGA differences between cultivars. Overexpression of two key genes in transgenic low-SGA potato cultivars increased their leaf SGA levels significantly. Discussion The results show that a strong response to light can underlie the SGA peaks that occasionally occur in certain potato cultivars and indicate that a between-cultivar variation in the expression of single SGA key genes can account for cultivar SGA differerences. We propose that current attempts to mitigate the SGA hazard will benefit from an increased consideration of cultivar-dependent SGA responses to post-harvest conditions, particularly light exposure. The identified key SGA genes can now be used as a molecular tool in this work.


FIGURE S1 .
FIGURE S1.Outline of sterol and SGA biosynthesis in potato.

FIGURE S5 .
FIGURE S5.Calystegine levels in potato tubers after wounding or light exposure.

FIGURE S12 .
FIGURE S12.GAME9 cDNA sequence and amino acid composition in Bintje and Magnum Bonum.

FIGURE S14 .
FIGURE S14.Relative gene expression of StHMGR1 and StTAM2 in transgenic Bintje plants.

FIGURE S2 .
FIGURE S2.General metabolic responses in tubers from six potato cultivars subjected to wounding.Tubers were obtained from one or two years of cultivation, stored or not during 6 months, and treated or not by wounding.(A) Orthogonal partial least-square discriminant analysis (OPLS-DA) score scatter plot of the first two predictive components of metabolite analyses after wounding for 0 h (green), 12 h (yellow), 24 h (blue), or 48 h (red).(B) Corresponding loading scatter plot for peak areas in LC-MS (green circles).A set of quality control samples were run in parallel (blue circles).(C) Plot of features with a variable importance in projection (VIP) value > 1,5.Model statistics: R2X=0.64;R2Y=0.60;Q2=0.41; ellipse indicates Hotelling´s T2 (95 %).

Figure xx .
Figure xx.Calystegine levels in potato tubers subjected to wounding or a light-exposure.Total glycoalkaloid levels were in a two-year study analysed in tubers at different time points after wounding (upper) or white light exposure (lower), either at harvest (blue bars) or after a cold storage for 6 months in darkness (yellow bars).Mean level of total glycoalkaloids ± range for the two years, Desiree and Princess were only analysed year one.

FIGURE S7 .FIGURE S8 .FIGURE S9 .
FIGURE S7.Differential gene expression in tubers from the potato cultivars Bintje and Magnum Bonum after wounding or light exposure.(A) Numbers of differentially expressed genes as determined by RNA sequencing of tubers from the potato cultivars Bintje (Bj) and Magnum Bonum (MB), that had been wounded or exposed to white light for 24 h and 48 h, respectively.(B) Venn display of upregulated genes in the two cultivars.(C) Venn display of down-regulated genes in the two cultivars.

FIGURE S10 .
FIGURE S10.Clustered gene expression patterns of 58 sterol-and SGA-related genes in tubers from the potato cultivars Bintje and Magnum Bonum after wounding or light exposure.The same set of 58 genes as in Fig. 3 are displayed.Average RPKM fold-change values relative to untreated control samples were subjected to k-means clustering into 8 different clusters based on the relation to the endpoint SGA levels in the experiment.Gray lines represent individual induction values of each transcript in a cluster, black lines show the median induction pattern for each cluster.Ctrl, control; Bj, Bintje; MB, Magnum Bonum.SGA indicate the total SGA levels (mg kg -1 FW) measured in untreated control tubers (Bj/MB) or in tubers 48h after wounding (W) or 8 days after light exposure (L) as described in Fig. 2. The genes in each cluster are listed in TableS9.

FIGURE S11 .FIGURE S12 .
FIGURE S11.Relative gene expression levels of the transcription factor StGAME9 in tubers from the potato cultivars Bintje and Magnum Bonum after wounding or light exposure.Pools of four tubers from cultivars Bintje (Bj) and Magnum Bonum (MB), were at harvest during two cultivation years subjected to a wounding (24 h) or light exposure (48 h) treatment.Mean transcript levels ± range of the two years.Supplemental FigureS8.Relative transcript level of StGAME9 transcript factor in potato tubers from Bintje and Magnum Bonum cultivars subjecte wounding (W24) and light exposure (L48).The StGAME9 Ct values were normalized against two reference genes, TUBULIN (TUB) and 60s rRNA was estimated its relative expression levels by using the Livak method (2 -Δ is shown the mean transcript levels ± SD of a pool of slices from four inde tubers from two different years, with two technical replicates each.Asteris indicate significant differences in the expression level relative to its basal expression in control non-stress tubers for each cultivar (student's t-test, * *** P<0.0001)

FIGURE S14 .FIGURE S15 .FIGURE S16 .
FIGURE S14.Relative transgene expression levels in leaf samples from wild-type potato plants (cv Bintje) and derived 35S:StHMGR1MB and 35S:StTAM2MB transformants.(A) Transgene expression in leaves of 35S:StHMGR1MB transgenic lines H1 to H4. (B) Transgene expression in leaves of 35S:StTAM2MB transgenic lines T1 to T4. Gene expression was monitored by QPCR analysis, and Ct values of the target gene were normalized against the TUBULIN reference gene.The transcript expression level relative to its basal expression in the Bintje wild type was calculated according toLivak & Schmittgen (2001).Mean ± range or SD of triplicate QPCR analyses of the plants used for the corresponding SGA analyses in Figure6.
General appearance of tubers from six potato cultivars subjected to light exposure.Potato tubers were after harvest kept in darkness for two weeks (control), and then exposed to constant white fluorescent light (100 µmol m -2 s -1 ) for 8 days in a controlled light cabinet.The red colour at harvest of the Desiree skin, and the red zones of King Edward, are normal for the cultivars.Pictures do not reflect size differences between cultivars.
Orthogonal partial least-square discriminant analysis (OPLS-DA) score scatter plot of the first two predictive components of metabolite analyses after wounding for 0 h (green), 12 h (yellow), 24 h (blue), or 48 h (red).(B)Correspondingloading scatter plot for peak areas in LC-MS (green circles).A set of quality control samples were run in parallel (blue circles).(C)Plot of features with a variable importance in projection (VIP) value > 1,5.Model statistics: R2X=0.64;R2Y=0.60;Q2=0.41;ellipseindicates Hotelling´s T2 (95 %).FIGURE S5.Calystegine alkaloid levels in tubers from six potato cultivars subjected to wounding or light exposure.Total calystegine levels were in a two-year study analyzed by LC-MS in tubers at time zero, or at different time points after wounding (12 h, 24 h, and 48 h; upper row) or white light exposure (2 d, 4 d, and 8 d; lower row), either at harvest (blue bars) or after a cold storage for 6 months in darkness (yellow bars).Total levels were calculated from the sum of calystegines A3, B2, and B4.Mean level of total calystegine ± range of two years (MB, Bi, Ju, KE), or one year (De, Pr).