Introduction of polyphosphate as a novel phosphate pool in the chloroplast of transgenic potato plants modifies carbohydrate partitioning

Potato plants (Solanum tuberosum L., cv. Désirée) were transformed with the polyphosphate kinase gene from Escherichia coli fused to the leader sequence of the ferredoxin oxidoreductase gene (FNR) from Spinacea oleracea under the control of the leaf specific St-LS1 promoter to introduce a novel phosphate pool in the chloroplasts of green tissues. Transgenic plants (cpPPK) in tissue culture developed necrotic lesions in older leaves and showed earlier leaf senescence while greenhouse plants showed no noticeable phenotype. Leaves of cpPPK plants contained less starch but higher concentrations of soluble sugars. The presence of polyphosphate in cpPPK leaves was demonstrated by toluidine blue staining and unambiguously verified and quantified by in vitro 31P-NMR of extracts. Polyphosphate accumulated during leaf development from 0.06 in juvenile leaves to 0.83 mg P g-1 DW in old leaves and had an average chain length of 18 residues in mature leaves. In situ 31P-NMR on small leaf pieces perfused with well-oxygenated medium showed only 0.036 mg P g-1 DW polyphosphate that was, however, greatly increased upon treatment with 50 mM ammonium sulfate at pH 7.3. This phenomenon along with a yield of 0.47 mg P g-1 DW polyphosphate from an extract of the same leaf material suggests that 93% of the polyphosphate pool is immobile. This conclusion is substantiated by the observation that no differences in polyphosphate pool sizes could be discerned between darkened and illuminated leaves, leaves treated with methylviologen or anaerobis and control leaves, treatments causing a change in the pool of ATP available for polyPi synthesis. Results are discussed in the context of the chelating properties of polyphosphates for cations and its consequences for the partitioning of photoassimilate between starch and soluble sugars.