Differential effects of environment on potato phenylpropanoid and carotenoid expression

BACKGROUND: Plant secondary metabolites, including phenylpropanoids and carotenoids, are stress inducible, have important roles in potato physiology and influence the nutritional value of potatoes. The type and magnitude of environmental effects on tuber phytonutrients is unclear, especially under modern agricultural management that minimizes stress. Understanding factors that influence tuber secondary metabolism could facilitate production of more nutritious crops. Metabolite pools of over forty tuber phenylpropanoids and carotenoids, along with the expression of twenty structural genes, were measured in high-phenylpropanoid purple potatoes grown in environmentally diverse locations in North America (Alaska, Texas and Florida). RESULTS: Phenylpropanoids, including chlorogenic acid (CGA), were higher in samples from the northern latitudes, as was the expression of phenylpropanoid genes including phenylalanine ammonia lyase (PAL), which had over a ten-fold difference in relative abundance. Phenylpropanoid gene expression appeared coordinately regulated and was well correlated with metabolite pools, except for hydroxycinnamoyl-CoA:quinatehydroxcinnamoyl transferase (HQT; r = -0.24). In silico promoter analysis identified two cis-acting elements in the HQT promoter not found in the other phenylpropanoid genes. Anthocyanins were more abundant in Alaskan samples and correlated with flavonoid genes including DFR (r = 0.91), UFGT (r = 0.94) and F3H (r = 0.77). The most abundant anthocyanin was petunidin-3-coum-rutinoside-5-glu, which ranged from 4.7 mg g(-1 )in Alaska to 2.3 mg g(-1 )in Texas. Positive correlations between tuber sucrose and anthocyanins (r = 0.85), suggested a stimulatory effect of sucrose. Smaller variation was observed in total carotenoids, but marked differences occurred in individual carotenoids, which had over a ten-fold range. Violaxanthin, lutein or zeaxanthin were the predominant carotenoids in tubers from Alaska, Texas and Florida respectively. Unlike in the phenylpropanoid pathway, poor correlations occurred between carotenoid transcripts and metabolites. CONCLUSION: Analysis of tuber secondary metabolism showed interesting relationships among different metabolites in response to collective environmental influences, even under conditions that minimize stress. The variation in metabolites shows the considerable phenotypical plasticity possible with tuber secondary metabolism and raises questions about to what extent these pathways can be stimulated by environmental cues in a manner that optimizes tuber phytonutrient content while protecting yields. The differences in secondary metabolites may be sufficient to affect nutritional quality.