Anaplerotic flux into the Calvin–Benson cycle: hydrogen isotope evidence for in vivo occurrence in C3 metabolism

Summary As the central carbon uptake pathway in photosynthetic cells, the Calvin–Benson cycle is among the most important biochemical cycles for life on Earth. A carbon flux of anaplerotic origin (i.e. through the chloroplast‐localized oxidative branch of the pentose phosphate pathway) into the Calvin–Benson cycle was proposed recently. Here, we measured intramolecular deuterium abundances in leaf starch of Helianthus annuus grown at varying ambient CO2 concentrations, C a. Additionally, we modelled deuterium fractionations expected for the anaplerotic pathway and compared modelled with measured fractionations. We report deuterium fractionation signals at H1 and H2 of starch glucose. Below a C a change point, these signals increase with decreasing C a consistent with modelled fractionations by anaplerotic flux. Under standard conditions (C a = 450 ppm corresponding to intercellular CO2 concentrations, C i, of 328 ppm), we estimate negligible anaplerotic flux. At C a = 180 ppm (C i = 140 ppm), more than 10% of the glucose‐6‐phosphate entering the starch biosynthesis pathway is diverted into the anaplerotic pathway. In conclusion, we report evidence consistent with anaplerotic carbon flux into the Calvin–Benson cycle in vivo. We propose the flux may help to: maintain high levels of ribulose 1,5‐bisphosphate under source‐limited growth conditions to facilitate photorespiratory nitrogen assimilation required to build‐up source strength; and counteract oxidative stress.


Notes S2. Estimation of intercellular CO2 concentrations during growth chamber experiments
shows Ca and Ci data obtained by gas exchange analysis (black circles). Their relationship is well described by a quadratic equation (dotted line, R 2 =0.999, p<0.001, n=5).

Figure S1 Ambient and intercellular CO2
concentrations in Helianthus annuus leaves (Ca, and Ci, respectively

Notes S3. Dilution of isotope signals by remnant starch
The plants studied here were grown at a CO2 concentration of 450 ppm over 7 to 8 weeks. H 1 and H 2 of starch glucose synthesised under these conditions have δD values of -247‰ and -557‰, respectively (Fig. 3). To drain the starch reserves and avoid dilution by 450 ppm isotope signals, all plants were kept in darkness for 24 hours after transfer to growth chambers.
Reportedly, this treatment led to a starch reduction from about 8 to almost 0 μmol glucose g -1 FW in Arabidopsis thaliana (Smith & Stitt, 2007). Similarly, we found a starch remnant of <0.6 μmol glucose g -1 FW in our samples. After two days of acclimation to 1500 ppm, we found a starch content of >171 μmol glucose g -1 FW. Thus, at 1500 ppm CO2, dilution of isotope signals due to remnant (450 ppm) starch is negligible (0.6/171=4‰).

Notes S4. Deuterium fractionation by glucose-6-phosphate dehydrogenase
Modelling of fractionations by glucose-6-phosphate dehydrogenase, G6PD, assumed an open system at steady state and followed published basic procedures (Hayes, 2002). Incoming glucose 6-phosphate, G6P, has two fates, starch biosynthesis or anaplerotic reinjection into the Calvin-Benson cycle via G6PD. D fractionation between the reaction product, 6phosphogluconolactone (6PGL), and remaining educt, G6P', is given as where R denotes D/H ratios, and α denotes D isotope effects of G6PD (αD=2.97) (Hermes et al., 1982). Isotope mass balance of the system is given as where F denotes fractional abundances (D/(H+D)), and f denotes the 6PGL commitment (f was varied between 0 and 1). F and R relate to each other as Here, we substitute the unknown D/H ratio of incoming G6P, RG6P, by the known ratio of Vienna Standard Mean Ocean Water, RVSMOW (155.76*10 -6 ) (Hagemann et al., 1970).

Notes S5. Deuterium fractionation by phosphoglucose isomerase
In the F6P to G6P direction, spinach PGI has a tritium isotope effects, αT, of 3 (Noltmann, 1972). The corresponding deuterium isotope effect, αD, can be estimated from αT as where kH and kD denote the reaction rates of the protium and deuterium isotopologues of F6P, respectively (Melander & Saunders, 1980). Based on this relationship, we estimate an αD of 2.14 which may cause D depletions at G6P H 2 of about 534‰ (1/αD-1).
For rabbit muscle PGI, Rose and O'Connell (1961) (Noltmann, 1972). To our knowledge, isotope effects of hydrogen loss to the medium and hydrogen uptake from the medium are unknow as is the hydrogen isotope composition of the medium. Furthermore, hydrogen exchange rates are temperature dependent and differ between the forward and backward reaction of PGI (Rose & O'Connell, 1961;Noltmann, 1972). Thus, isotope fractionation by hydrogen exchange requires further attention.