Specific reduction of chloroplast glyceraldehyde-3-phosphate dehydrogenase activity by antisense RNA reduces CO₂ assimilation via a reduction in ribulose bisphosphate regeneration in transgenic tobacco plants

Abstract

The reduction of 3-phosphoglycerate (PGA) to triose phosphate is a key step in photosynthesis linking the photochemical events of the thylakoid membranes with the carbon metabolism of the photosynthetic carbon-reduction (PCR) cycle in the stroma. Glyceraldehyde-3-phosphate dehydrogenase: NADP oxidoreductase (GAPDH) is one of the two chloroplast enzymes which catalyse this reversible conversion. We report on the engineering of an antisense RNA construct directed against the tobacco (Nicotiana tabacum L.) chloroplastlocated GAPDH (A subunit). The construct was integrated into the tobacco genome by Agrobacterium-mediated transformation of leaf discs. Of the resulting transformants, five plants were recovered with reduced GAPDH activities ranging from 11 to 24% of wild-type (WT) activities. Segregation analysis of the kanamycin-resistance character in self-pollinated T₁ seed from each of the five transformants revealed that one plant (GAP-R) had two active DNA inserts and the others had one insert. T₁ progeny from GAP-R was used to generate plants with GAPDH activities ranging from WT levels to around 7% of WT levels. These were used to study the effect of variable GAPDH activities on metabolite pools for ribulose1,5-bisphosphate (RuBP) and PGA, and the accompanying effects on the rate of CO₂ assimilation and other gasexchange parameters. The RuBP pool size was linearly related to GAPDH activity once GAPDH activity dropped below the range for WT plants, but the rate of CO₂ assimilation was not affected until RuBP levels dropped to 30–40% of WT levels. That is, the CO₂ assimilation rate fell when RuBP per ribulose-1,5-biphosphate carboxylase-oxygenase (Rubisco) site fell below 2 mol·(mol site)⁻¹ while the ratio for WT plants was 4–5 mol·m(mol site)⁻¹. Leaf conductance was not reduced in leaves with reduced GAPDH activities, resulting in an increase in the ratio of intercellular to ambient CO₂ partial pressure. Conductance in plants with reduced GAPDH activities was still sensitive to CO₂ and showed a normal decline with increases in CO₂ partial pressure. Although PGA levels did not fluctuate greatly, the effect of reduced GAPDH activity on RuBP-pool size and assimilation rate can be interpreted as being due to a blockage in the regeneration of RuBP. Concomitant gas-ex change and chlorophyll a fluorescence measurements indicated that photosynthesis changed from being Rubisco-limited to being RuBP-regeneration-limited at a lower CO₂ partial pressure in the antisense plants than in WT plants. Photosynthetic electron transport was down-regulated by the build-up of a large proton gradient and the electron-transport chain did not become over-reduced due to a shortage of NADP. Plants with severely reduced GAPDH activity were not photoinhibited despite the continuous presence of a large thylakoid proton gradient in the light. Along with plant size, Rubisco activity, leaf soluble protein and chlorophyll content were reduced in plants with the lowest GAPDH activities. We conclude that chloroplastic GAPDH activity does not appear to limit steady-state photosynthetic CO₂ assimilation at ambient CO₂. This is because WT leaves maintain the ratio of RuBP per Rubisco site about twofold higher than the level required to achieve a maximal rate of CO₂ assimilation.