14-PSIL MAGIC: a multi-tiered approach to gaining increased carbon

In the Calvin-Benson cycle of plants, the enzyme RuBisCO fixes CO2 to produce two molecules of 3-phosphoglycerate. RuBisCO evolved ~3.6bn years ago in an atmosphere of high CO2 and low O2, with little need to discriminate between the two gases. In today's atmosphere RuBisCO fixes both CO2 and O2. The latter generates phosphoglycolate, which is retrieved by photorespiration but at an energy cost that represents a significant loss in photosynthetic efficiency. One method to reduce O2 fixation by RuBisCO is to raise the partial pressure of CO2. Carbon concentrating mechanisms (CCMs) have evolved multiple times to this end. For example, C4 photosynthesis uses phosphoenol-pyruvate carboxylase (PEPC), an enzyme that does not possess oxygenase activity, to fix HCO3- temporarily in C4 acids; cellular specialization allows release and concentration of CO2 for refixing by RuBisCO. As much as a 50% increase in yield might be realized in crops were O2 fixation by RuBisCO to be bypassed in a similar manner. Significant resources have already gone into engineering RuBisCO for increased CO2 selectivity and into introducing a single-celled version of C4 photosynthesis in rice, but a step change in photosynthetic efficiency has not yet been achieved.

Investigators from Universities in the US (John Golbeck (JG), Penn State; and Cheryl Kerfeld (CK), Michigan State) and the UK (Mike Blatt (MB), Glasgow; Nigel Burroughs (NB), Warwick; and Julian Hibberd (JH), Cambridge) participated in an NSF/BBSRC Ideas Laboratory in 2010, at which they proposed a novel strategy to address this problem, a proposal that has since matured to the level of technological implementation. They are now joined by Nick Smirnoff (NS, Exeter) and Manish Kumar (MK, Penn State), who bring additional and key expertise to the project. The research has two themes: a light driven ion pump, composed of halorhodopsin and an anion/HCO3- exchanger, AE1; and the use of artificial scaffolds for channelling CO2 to RuBisCO. A parallel goal is to re-engineer the light-driven ion pump to transport HCO3- directly and to absorb light energy not used by photosynthesis. These efforts are underpinned with mathematical modelling of CO2 delivery and assimilation to direct experimentation based around the following components.

Light-Driven Pump. Halorhodopsin (HR) is an integral membrane protein and consists of 7 transmembrane alpha-helices and a bound retinal. The retinal undergoes light-driven bond rotation between 13-cis and all-trans conformations to drive ion transport. HR transports other halides as well, and ion selectivity appears to be a localized feature of the pHR transport site. pHR is sufficiently promiscuous to make engineering a light-driven HCO3- pump a possibility.

Anion/Bicarbonate Exchanger: The erythrocyte Band3 protein (AE1) facilitates Cl-/HCO3- exchange across the membrane. It generates a high flux close to equilibrium and is largely insensitive to pH, making it well suited to engineering a HCO3- accumulating mechanism. Most promising for synthetic engineering, the AE1 transporter is functional in mammalian cell cultures, Xenopus oocytes, and yeast without adverse effects on homeostasis or growth. The modular structure of AE1, offers a realistic strategy for coupling HCO3- pumping coupled to pHR-driven Cl- transport.

Artificial Scaffolds: CO2 diffusion needs to be constrained locally for sufficient time to allow it to be fixed by RuBisCO. Substrate channelling is found in several natural systems, including plants. Efficiency gains arise from physical proximity and 'sponge'-like buffering that enables transfer of intermediates and minimizes runoff of substrates.

We are building on progress that includes expression of pHR in E. coli, cyanobacteria and plants. These advances have put us in a strong position to deliver within the next few years. Mathematical modeling has validated the idea of using a light-driven ion pump for concentrating CO2; it now remains to assemble and express these pumps and validate function in chloroplasts. The idea to use scaffolds to concentrate CO2 at RuBisCO remains a goal, but our strategies have changed in light of new understanding of the interplay between diffusion and kinetics. We have successfully expressed scaffold proteins in cyanobacteria and plants, demonstrating that they can be both targeted to specific sites and that they function to recruit their respective substrates. Our mathematical models predict that the original idea of utilizing these constructs to enhance channeling of CO2 to RuBisCO will have negligible impact on CO2 assimilation. We need now to confirm this prediction in our cyanobacterial systems. A rethinking of the problem of concentrating CO2 at RuBisCO in C3 plants leads to development of a new approach. Our mathematical models highlight the poor CO2 capture probability of RuBisCO as a major constraint. Here, we propose designs to slow the diffusion rate of CO2 in the stroma and increase assimilation by introducing transient (stationary) binding sites near RuBisCO (a CO2 'sponge'), effectively enhancing the native characteristics recently identified in photosynthetic systems. We will use the cyanobacterial system to screen and optimize this approach and will use the scaffolds now proven in our hands to translate these to chloroplasts. Finally, we previously lacked the ability to quantify performance, ie. HCO3- concentration gains. This capability is now available through a lipid vesicle technique.

This proposal is for fundamental research to develop new conceptual approaches relevant to ideas emerging within the international plant, systems and synthetic biology communities. The research will stimulate thinking around strategies for modelling and for applications of synthetic biology in plants, especially in relation to photosynthesis, and it should strengthen methodologies relevant at many levels from cell to crop engineering. Thus, the research is expected to benefit fundamental researchers and, in the longer-term agriculture and industry, through conceptual developments and approaches to improving carbon capture by plants. The research will feed into higher education training programmes through capacity building at the postgraduate and postdoctoral levels. Additional impact is proposed through public displays and the development of teaching resources building on the background work for this proposal. Finally the research will help guide future efforts in applications to agricultural/industrial systems. The applicants have established links with industrial/technology transfer partners and research institutes to take advantage of these developments. Further details of these, and additional impacts will be found in Part 1 of the Case for Support and in the attached Impact Pathways.