Synthetic biology strives to create non-natural biological parts and systems, including new enzymes and metabolic networks. The CETCH cycle, a synthetic CO₂-fixation pathway, offers increased efficiency compared to natural photosynthesis. However, its biosynthetic capabilities are limited by a single output molecule, glyoxylate. Natural metabolic pathways exhibit dynamic flexibility through anaplerosis – the replenishment of central metabolic intermediates diverted to various biosynthetic routes. This flexibility is lacking in most synthetic *in vitro* networks. To enhance the CETCH cycle's potential, this study aims to develop anaplerotic modules that directly access and replenish core intermediates (propionyl-CoA and methylmalonyl-CoA) of the cycle, thereby enabling the synthesis of a broader range of complex molecules. The chosen benchmark molecule for evaluating the anaplerotic modules is 6-deoxyerythronolide B (6-DEB), the macrolide backbone of erythromycin, a polyketide requiring propionyl-CoA and methylmalonyl-CoA as building blocks. Successful implementation would establish a complex *in vitro* metabolic network capable of continuously producing complex molecules directly from CO₂.

The literature extensively describes anaplerosis in natural metabolic networks, particularly in the context of the citric acid cycle, where various pathways replenish intermediates like oxaloacetate. Studies have demonstrated the design and implementation of synthetic CO₂ fixation pathways, such as the CETCH cycle, which shows improved efficiency over the Calvin cycle. Prior research has explored using glyoxylate (the primary output of the CETCH cycle) as a precursor for various high-value products, including terpenes. However, directly accessing other CETCH cycle intermediates for diverse biosynthesis has been a challenge. This work draws inspiration from natural anaplerotic pathways, specifically the β-hydroxyaspartate cycle (BHAC), the serine cycle, and the reductive TCA cycle, among others, to design efficient anaplerotic modules.

The study involved designing and reconstituting four anaplerotic modules to replenish CETCH cycle intermediates. Module 1 represents the optimized CETCH cycle itself, modified to enhance CO₂ fixation efficiency by using a creatine phosphokinase-based ATP regeneration system and glycolate-based readout instead of a malate readout. Modules 2 and 3 convert glyoxylate (the output of Module 1) into oxaloacetate, malate, and acetyl-CoA. Modules 4a-4d represent four different anaplerotic pathways feeding back into Module 1: Module 4a utilizes the reductive TCA cycle for succinyl-CoA production; Module 4b uses the glyoxylate cycle for succinyl-CoA; Module 4c uses the 3-hydroxypropionate (3-OHP) cycle for propionyl-CoA regeneration; and Module 4d uses a portion of the EMC pathway for crotonyl-CoA production. Each module was reconstituted *in vitro* using various enzymes, and their performance was optimized by adjusting enzyme concentrations and cofactor levels. The effectiveness of each module was evaluated by measuring methylmalonyl-CoA accumulation, which serves as a proxy for 6-DEB production. Finally, the integrated system (Modules 1-3 and one of 4a-4d) was coupled with the 6-DEB polyketide synthase (DEBS, Module 5) to assess 6-DEB synthesis from CO₂. Enzyme activities were measured using various spectrophotometric and LC-MS methods. 6-DEB was quantified using HPLC-MS, and a standard curve was created using spectrophotometric measurements of NADPH oxidation correlated to HPLC-MS measurements of 6-DEB.

Optimization of Module 1 (CETCH cycle) increased CO₂ fixation efficiency by 1.4-fold compared to the original design. Modules 2 and 3 successfully converted glyoxylate into acetyl-CoA, oxaloacetate, and malate. Among the anaplerotic modules (4a-4d), Module 4c, based on the 3-OHP cycle, demonstrated the highest methylmalonyl-CoA production due to the higher thermodynamic driving force. The addition of adenylate kinase to Module 4c further increased the yield. Module 4d's performance was optimized by adjusting CoA concentration. When integrated with DEBS (Module 5), three out of four anaplerotic modules (4a, 4b, and 4c) yielded detectable 6-DEB, with Module 4c producing the highest yield (31.9 ± 1.6 µM). This yield was comparable to that obtained with isolated DEBS alone (39.1 ± 0.3 µM), demonstrating the successful carbon-positive synthesis of 6-DEB directly from CO₂ via more than 50 enzymatic reactions. The carbon conversion efficiency in the best scenario (Module 4c) reached 172%. The absence of anaplerotic feedback resulted in significantly reduced 6-DEB synthesis, indicating the importance of anaplerotic modules for the continuous operation of this complex *in vitro* network. Analysis of CoA esters revealed significant differences in the metabolic profiles of the systems with and without anaplerotic feedback loops, especially when DEBS was included. Enzyme kinetic studies were also conducted for several key enzymes (Bbd and Mtk).

This study demonstrates the successful construction of a complex *in vitro* metabolic network capable of producing a complex molecule (6-DEB) directly from CO₂. The key innovation is the incorporation of anaplerotic modules, which mimic the dynamic flexibility of natural metabolic pathways. The high yield of 6-DEB achieved with Module 4c highlights the significance of thermodynamic driving force and the strategic choice of re-entry points into the CETCH cycle. The observed differences in efficiency among the anaplerotic modules highlight the importance of careful pathway selection and optimization. The negative influence of DEBS on the CETCH cycle in the absence of anaplerotic feedback underlines the necessity of these modules for sustaining continuous operation. Further optimization of certain enzymes, particularly Frd in Module 4a, might further enhance the overall efficiency. Dynamic regulation mechanisms, like cofactor purge valves, could further improve the network's performance. This work represents a significant step towards creating more sophisticated *in vitro* systems that mimic the complexity and robustness of natural metabolic networks.

This research successfully demonstrates the feasibility of synthesizing complex molecules directly from CO₂ using a synthetic metabolic network augmented with anaplerotic modules. The 172% carbon conversion achieved for 6-DEB synthesis using the 3-OHP cycle-based anaplerotic module is a noteworthy achievement. Future work could focus on optimizing existing modules, exploring additional anaplerotic pathways, and incorporating dynamic regulatory elements to enhance the efficiency and robustness of the system. This approach holds promise for producing a variety of other valuable molecules, including polyketides and polyhydroxyalkanoates.

The study focused on a single model product, 6-DEB. While the results are promising, it remains to be seen how well this approach generalizes to other complex molecules. The *in vitro* system requires careful optimization of enzyme concentrations and cofactor levels, which might limit its scalability. The current system does not incorporate sophisticated regulatory mechanisms found in natural metabolic pathways, which may be necessary for further enhancements in efficiency and robustness. The relatively high error associated with the use of the fumarate reductase from *Trypanosoma brucei* indicates further enzyme engineering is required for improved precision and reliability of the 4a anaplerotic module.