nature communications Article https://doi.org/10.1038/s41467-022-33033-1 Kinetic compartmentalization by unnatural reaction for itaconate production Received: 2 November 2021 Dae-yeol Ye1,4, Myung Hyun Noh1,4, Jo Hyun Moon 1,4 , Alfonsina Milito2, Minsun Kim3, Jeong Wook Lee 1,3, Jae-Seong Yang 2 & Accepted: 29 August 2022 Gyoo Yeol Jung 1,3 Check for updates Physical compartmentalization of metabolism using membranous organelles in eukaryotes is helpful for chemical biosynthesis to ensure the availability of 1234567890():,; substrates from competitive metabolic reactions. Bacterial hosts lack such a 1234567890():,; membranous system, which is one of the major limitations for efficient metabolic engineering. Here, we employ kinetic compartmentalization with the introduction of an unnatural enzymatic reaction by an engineered enzyme as an alternative strategy to enable substrate availability from competitive reactions through kinetic isolation of metabolic pathways. As a proof of con- cept, we kinetically isolate the itaconate synthetic pathway from the tri- carboxylic acid cycle in Escherichia coli, which is natively separated by mitochondrial membranes in Aspergillus terreus. Specifically, 2-methylcitrate dehydratase is engineered to alternatively catalyze citrate and kinetically secure cis-aconitate for efficient production using a high-throughput screen- ing system. Itaconate production can be significantly improved with kinetic compartmentalization and its strategy has the potential to be widely applicable. Diverse intracellular biochemical reactions have evolved to efficiently synthesized from a decarboxylation reaction of cis-aconitate by cis- supply energy and synthesize essential metabolites required for aconitate decarboxylase11,13. However, cis-aconitate is an intermediate organism survival1–4. For example, highly orchestrated enzymatic metabolite that is transiently generated in the tricarboxylic acid (TCA) reactions involving substrate channeling by multienzyme complex or cycle via the enzyme aconitase, which converts citrate into isocitrate8. consecutive enzyme reactions have evolved to rapidly consume The strong affinity of aconitase enzyme for cis-aconitate causes its intermediates, avoiding the formation of byproducts that undermine rapid conversion into isocitrate8, reducing the cis-aconitate availability the efficiency of the desired pathway5–9. However, such an efficient for itaconate biosynthesis11,13,14. metabolic reaction chain may limit metabolic engineering when the One solution used by nature to avoid competing reactions is spatial newly introduced metabolic pathway must use the metabolic inter- compartmentalization13,15,16, which utilizes the physical separation of an mediate as a substrate, as the accessibility of the intermediate is lim- intermediate to block kinetic competition among diverse chemical ited by kinetic competition with the native reaction chain10. For reactions13,16. For example, Aspergillus terreus, a native itaconate pro- example, in itaconate biosynthesis, substrate availability can limit the ducer, facilitates spatial compartmentalization of cis-aconitate by biotechnological production of these valuable compounds11. Itaconate pumping out mitochondrial cis-aconitate into the cytosol by exchan- is a dicarboxylic acid used in the resin and plastic industry12 and can be ging cytosolic oxaloacetate via an antiporter, MttA (Supplementary 1 Department of Chemical Engineering, Pohang University of Science and Technology, 77 Cheongam-Ro, Nam-Gu, Pohang, Gyeongbuk 37673, Republic of Korea. 2Centre for Research in Agricultural Genomics (CRAG), CSIC-IRTA-UAB-UB, Campus UAB, Bellaterra, Barcelona 08193, Spain. 3School of Inter- disciplinary Bioscience and Bioengineering, Pohang University of Science and Technology, 77 Cheongam-Ro, Nam-Gu, Pohang, Gyeongbuk 37673, Republic of Korea. 4These authors contributed equally: Dae-yeol Ye, Myung Hyun Noh, Jo Hyun Moon. e-mail: jaeseong.yang@cragenomica.es; gyjung@postech.ac.kr Nature Communications | (2022)13:5353 1 Article https://doi.org/10.1038/s41467-022-33033-1 Citrate Citrate PrpD* cis- Aconitate cis- Aconitate AcnB Cad AcnB* Cad Isocitrate Isocitrate TCA cycle Itaconate TCA cycle Itaconate Fig. 1 | Schematic diagram of overall strategies for kinetic compartmentaliza- to secure cis-aconitate, resulting in increased itaconate production. The non- tion using the non-natural enzyme reaction. 2-Methylcitrate dehydratase (PrpD) natural enzyme exhibits a kinetically compartmentalized effect without actual was evolutionarily engineered to alter the substrate preference and catalytic effi- spatial separation. Citrate (blue); isocitrate (orange); cis-aconitate (green); ciency towards citrate. The desired PrpD mutant has higher activity towards citrate itaconate (red). Fig. 1)17. This physical separation allows the fungus to accumulate Results cis-aconitate in the cytosol and efficiently produce itaconate by avoid- Selection of target enzyme for compartmentalization ing competition with the TCA cycle. Further, itaconate production can To separate cis-aconitate production from the TCA cycle, we first be efficiently regulated via MttA activity18. However, because of the reviewed the detailed reaction mechanisms of cis-aconitate synthesis insufficiently implemented genetic engineering tools and complicated in E. coli. The precursor of itaconate, cis-aconitate, can be synthesized culture conditions of A. terreus, research focused on itaconate pro- by aconitase encoded by acnA and acnB in E. coli. Dehydration of duction has been performed in well-known workhorse cells such as citrate is sequentially coupled to rehydration of cis-aconitate by the Escherichia coli11,14. Mimicking the spatial compartmentalization in E. coli same enzymes, which appear to catalyze the isomerization of citrate to would be an efficient approach for itaconate production; however, isocitrate. Thus, cis-aconitate is temporarily synthesized as a reaction prokaryotic hosts lack a membranous system, leading to challenges in intermediate (Supplementary Fig. 2a). Furthermore, the catalytic effi- building a compartmentalized system11,14,15. ciency (kcat/Km) of aconitase is much higher for cis-aconitate than for In this work, we adapt a kinetic compartmentalization approach citrate8. Consequently, cis-aconitate accumulates at very low levels in to overcome the inapplicability of spatial compartmentalization in native E. coli11,14, resulting in low itaconate production. Aconitase may bacteria. We introduce an unnatural enzymatic reaction to kinetically only conduct a dehydration reaction, however, the dehydration and separate competitive reactions by releasing the intermediate from the rehydration reactions are coupled with two main catalytic residues, native reaction chain without creating a physical barrier. Particularly, H444 and S24423. H444 functions as a proton donor, whereas S244 acts we kinetically compartmentalize the itaconate production reaction as a proton acceptor in the first dehydration reaction23. In the rehy- from the TCA cycle by introducing a non-natural enzyme that can dration reaction, H444 acts as a proton acceptor, and S244 functions separate consecutive biochemical reactions catalyzed by aconitase. as a proton donor, switching the roles of these residues23. After these We hypothesize that introducing a non-natural biochemical reaction two reactions, the catalytic residues return to the initial state to con- into E. coli cells, which can effectively synthesize cis-aconitate from duct another round of reactions. Thus, it is difficult to isolate the citrate, would increase the intracellular cis-aconitate level, thus trig- dehydration reaction. gering its conversion into itaconate through catalysis by cis-aconitate We screened enzymes with dehydration activity; their known decarboxylase. substrates are similar to those of citrate structures. Among all enzyme In particular, endogenous 2-methylcitrate dehydratase, PrpD19 is candidates, the endogenous enzyme of E. coli, PrpD, converts successfully engineered that originally converts 2-methylcitrate into 2- 2-methylcitrate into 2-methyl-cis-aconitate in a one-step dehydration methyl-cis-aconitate, to catalyze the conversion of citrate into cis- reaction without rehydration to 2-methylisocitrate (Supplementary aconitate. PrpD is evolutionarily engineered to switch its specificity Fig. 2b)19. This enzyme has a much lower affinity and catalytic efficiency from 2-methylcitrate to citrate by computational simulation design for citrate and cis-aconitate compared to aconitases19. Engineering and high-throughput screening20–22. Based on the non-natural enzyme, PrpD to alternatively catalyze citrate into cis-aconitate may be a useful an unnatural itaconate production pathway is constructed that is strategy for enhancing the pool of cis-aconitate for itaconate kinetically isolated from the TCA cycle (Fig. 1). Using our approach, we production. obtain a significantly increased itaconate yield of up to 10.6-fold compared to the parental strain at 46.0% of the theoretical maximum14. Development of itaconate-specific screening system This implementation of kinetic compartmentalization in bacteria Recently, an itaconate-responsive LysR-type transcription factor, ItcR, allows overcoming the limitation of being a spatially inseparable host from Yersinia pseudotuberculosis was shown to regulate the expression by introducing an unnatural biochemical reaction able to efficiently of the itaconate degradation pathway based on the presence and separate competitive reactions. This method shows potential for amount of itaconate24. We exploited ItcR and its cognate promoter, kinetic compartmentalization in prokaryotes, considering the sub- Pccl, to construct an itaconate-responsive screening system. The stantial number of promiscuous enzymes that can be engineered using screening system was designed to regulate the expression of the this approach. antibiotic resistance gene25,26 according to itaconate concentration. Nature Communications | (2022)13:5353 2 Article https://doi.org/10.1038/s41467-022-33033-1 a b Itaconate 0.25 0 mg/L 3 mg/L 7 mg/L 10 mg/L 0.20 15 mg/L Specific growth rate (h -1) 0.15 tetA itcR Pccl PJ23106 0.10 Itaconate responsive screening system 0.05 0.00 Tetracycline 0.01 0.1 1 Added itaconate (g/L) Fig. 2 | Design and validation of itaconate-responsive screening system. accordance with itaconate concentration under varied tetracycline concentrations. a Schematic diagram of itaconate-responsive screening system based on the tet- Data were presented as mean values and error bars indicate the standard deviations racycline resistance gene. b Specific growth rates of WS strains were plotted in from three biological replicates. Source data are provided as a Source Data file. Specifically, ItcR was constitutively expressed using a synthetic pro- confirmed that the methyl group of 2-methylcitrate toward W110 and moter (PBBa_J23106), and the tetracycline resistance gene (tetA) was G111 with our docking simulation (Supplementary Fig. 5). So we deci- controlled under Pccl. Collectively, the system was intended to provide ded to mutate W110, G111, and I331 residues to switch the catalytic a growth advantage under tetracycline pressure in environments with activity from 2-methylcitrate to citrate, which has no methyl group on high itaconate concentrations or in high-producing strains (Fig. 2a). the second carbon. A mutant library of PrpD (W110, G111, and I331 To validate the screening system, the itaconate-responsive residues) was constructed for expression in moderate strength screening system was transformed into an acid-tolerant E. coli W (BBa_PJ23108) based on structural analysis (see Methods) with theore- strain14 to produce the WS strain (Supplementary Data 1). We tical 203 variants numbers in size and transformed. demonstrated the effectiveness of the screening system and a further To increase the intracellular level of citrate and ensure the activity improvement in itaconate production using a previously developed of PrpD mutants, the catalytic efficiency of the competing enzyme, itaconate-producing E. coli strain with acetate as the sole carbon AcnB, was decreased through site-directed mutagenesis (AcnBW482R, source14. The tetracycline concentration as the selection pressure was Table 1)28. Mutagenesis of AcnB efficiently lowered the catalytic effi- varied up to 15 mg/L and growth retardation were confirmed in ciency to citrate by 3.75-fold (Table 1). Notably, by knocking out the accordance with the increased tetracycline concentration, indicating iclR-encoding transcriptional repressor for the glyoxylate shunt path- tight regulation of ItcR/Pccl (Fig. 2b). In addition, the gradual increase in way, the pathway was activated to increase anaplerosis, ensuring the the specific growth rate was validated by extracellular addition of ita- level of oxaloacetate, the different substrate for citrate, which is able to conate under selection pressure. Specifically, the specific growth rate result in enhancing itaconate production14,29. The resulting WAIC strain increased as the concentration of itaconate increased to 2 g/L. Addi- (E. coli W (acnB_W482R) ΔiclR harboring cad expressing-plasmid tionally, the growth rate according to itaconate showed different (pCAD)) showed a lower cell biomass (2.16 g DCW/L) and itaconate tendencies depending on the tetracycline selection pressure (Fig. 2b). production (0.26 g/L) by 1.17- and 1.22-fold, respectively, compared to Collectively, these results indicate that the itaconate-responsive the WCI strain (with wild-type acnB) (Supplementary Fig. 6)14. Never- screening system was successfully constructed and is widely applic- theless, the accumulation of citrate was significantly enhanced by up to able for itaconate production. 0.51 g/L, as expected. Overall, an environment was created in which the PrpD mutant with altered substrate specificity produced increased Screening PrpD mutant with altered substrate specificity itaconate and showed a sufficient growth advantage in the selection To select enzyme variants that can produce cis-aconitate, PrpD protein condition. The screening was conducted by increasing the selection engineering was conducted using the itaconate-responsive screening pressure from 7–15 mg/L of tetracycline over four rounds. system. Target residues for mutagenesis were selected based on structural analysis of tartrate-bound MmgE (PDB code: 5MUX [https:// Characterization of enriched PrpD mutants www.rcsb.org/structure/5MUX]), the homolog of PrpD derived from After enrichment, ten isolated mutants were analyzed, and five types of Bacillus subtilis27. As the catalytic residues are all conserved in PrpD mutants were characterized (Supplementary Table 1). We initially (Supplementary Fig. 3) and the tartrate has a similar structure with 2- validated itaconate production for each mutant (WAICPNNN strains, methylcitrate, we hypothesized that the tartrate-bound structure Supplementary Data 1) at the test tube scale (Supplementary Fig. 7). could give useful information about substrate-residue interactions. Then we superimposed 2-methylcitrate on the L-tartrate and selected residues near the methyl group of 2-methylcitrate. As our objective is Table 1 | Kinetic parameters of AcnB and AcnBW482R to find mutants to recognize citrate instead of 2-methylcitrate, we were Enzyme Substrate kcat (s−1) Km (mM) kcat/Km (mM−1 s−1) looking for more hydrophobic and small residues. The methyl group of AcnB Citrate 4,488.62 ± 696.56 3.32 ± 1.03 1,402.61 ± 249.94 2-methylcitrate in our superimposed model heading toward W110 and AcnBW482R Citrate 2,232.31 ± 758.05 6.34 ± 3.12 374.31 ± 67.74 G111. These residues make a hydrophobic pocket with I331 that seems Source data are provided as a Source Data file. to allow to bind methyl group (Supplementary Fig. 4). We also Nature Communications | (2022)13:5353 3 Article https://doi.org/10.1038/s41467-022-33033-1 a b c 12 12 12 2.0 2.0 2.0 Cell biomass (g DCW/L) Cell biomass (g DCW/L) Cell biomass (g DCW/L) 10 10 Itaconate & Citrate (g/L) 10 Itaconate & Citrate (g/L) Itaconate & Citrate (g/L) 1 1 1 Acetate (g/L) Acetate (g/L) Acetate (g/L) 8 1.5 8 1.5 8 1.5 6 6 6 1.0 1.0 1.0 0.1 0.1 0.1 4 4 4 0.5 0.5 0.5 2 2 2 0 0.01 0.0 0 0.01 0.0 0 0.01 0.0 0 12 24 36 48 0 12 24 36 48 0 12 24 36 48 Time (h) Time (h) Time (h) Fig. 3 | Fermentation profiles of itaconate production. a WAICP, b WAICPVTL, and up-triangles, acetate; squares, itaconate; down-triangles, citrate. Data were pre- c WAICPVL strains. The left y-axis and y-offset represent the cell biomass (g DCW/L) sented as mean values and error bars indicate the standard deviations from three and acetate (g/L), respectively. The right y-axis indicates the production of itaco- biological replicates. Source data are provided as a Source Data file. nate and citrate (g/L). The x-axis denotes time (h). Symbols: circles, cell biomass; The WAICPVTL strain (PrpDVTL with W110V, G111T, and I331L) showed a (15.81 mM of Km) was confirmed in PrpDVL compared to wild-type PrpD 1.50-fold increase in itaconate production, whereas most mutants (66.39 mM), as expected. However, the mutant displayed a 1.35-fold showed a decreased level of itaconate production compared to the lower turnover rate (kcat) for citrate (7,804.72 s−1), improving the cata- WAICP strain with wild-type PrpD. lytic efficiency by 3.11-fold (Table 2). An additional culture was conducted at the flask-scale to validate A noticeable decrease in the affinity for cis-aconitate was also the WAICPVTL strain (Fig. 3a, b). Interestingly, the cell biomass was observed in PrpDVL (1.52 mM of Km) compared to wild-type PrpD remarkably reduced in the WAICPVTL strain (Fig. 3b) compared to that (0.70 mM, Table 2). The change in the binding pocket following of the WAICP strain (Fig. 3a) by 1.80-fold (1.01 g DCW/L), and citrate mutagenesis may have altered the affinity for cis-aconitate and accumulation was reduced by 2.00-fold (0.15 g/L), similar to that of the citrate26,30. In addition, the turnover rate of PrpDVL was reduced WCI strain (Fig. 3b and Supplementary Fig. 6a). These results indicate (96.34 s−1), leading to a 2.68-fold decrease in catalytic efficiency. Col- that PrpDVTL exhibits increased catalytic efficiency toward the citrate to lectively, these results indicate that PrpDVL efficiently extracted citrate produce cis-aconitate, which induced kinetic separation and redir- from the TCA flux and induced kinetic compartmentalization by con- ected more citrate for use in itaconate production. Indeed, the verting it into cis-aconitate to facilitate itaconate production (Fig. 1). WAICPVTL strain showed a 2.56-fold increase in itaconate production to Computational simulations of 2-methylcitrate into PrpD enzyme 0.86 g/L compared to the WAICP strain. Moreover, the itaconate yield showed that in the wild-type, single mutated (PrpDV, PrpDT, and PrpDL), was significantly enhanced in the WAICPVTL strain by up to 4.76-fold and double mutant (PrpDTL) enzymes, the methyl group faced the (0.27 g/g), indicating that a kinetic compartmentalization strategy is a residues W110 and G111 (wildtype or mutated), while in the double powerful approach for enhancing the cis-aconitate availability and mutants PrpDVT and PrpDVL, and in the triple mutant PrpDVTL, the itaconate production. conformation shifted the methyl group in the opposite direction To validate the effect of each residue on PrpDVTL, mutants at single (Supplementary Fig. 5). Moreover, docking simulations revealed a residues and a combination of double mutants were generated as the slight decrease in hydrogen bonds between 2-methylcitrate and PrpD WAICPV, WAICPT, WAICPL, and WAICPTL strains (Supplementary Data 1); residues in the double mutants PrpDVT and PrpDVL and triple mutant these mutants showed no significant difference in itaconate produc- PrpDVTL compared to simulation into the active pocket of wild-type tion compared to the WAICP strain (Supplementary Fig. 8). In contrast, (Table 3). In contrast, docking predictions of citrate interactions the WAICPVT and WAICPVL strains showed 3.55- (1.19 g/L) and 3.84-fold showed an increased number of hydrogen bonds between this ligand (1.28 g/L) increases in the itaconate titer at 48 h (Supplementary Fig. 8 and PrpD residues in all mutants compared to the bonds in the wild- and Fig. 3c), respectively, which are even higher than that of WAICPVTL type enzyme (Supplementary Fig. 9 and Table 3). Interestingly, (Fig. 3b). These results agree with those of structural analysis, as W110 mutants showing a higher substrate specificity for citrate (PrpDVT, is the closest residue among the target residues to the methyl group of PrpDVL, and PrpDVTL) exhibited a greater increase in hydrogen bonds the second carbon on 2-methylcitrate (Supplementary Table 2), com- between citrate and PrpD residues, with W110V potentially acting as bined with the rationale for the initial multi-residue library design. the most important mutation determining the substrate shift, which is supported by the fact that mutating this single amino acid increased Kinetic and structural analysis of PrpD mutants the number of hydrogen bonds of citrate compared to 2-methylcitrate, The enzyme kinetics of wild-type PrpD, PrpDVTL, PrpDVT, and PrpDVL were in the previously mentioned double and triple mutants (Table 3). characterized to evaluate the enhancement of itaconate production Finally, docking on the double mutant PrpDTL gave the same amount of (Table 2). First, significant decreases of affinity to the natural substrate hydrogen bonds with 2-methylcitrate and citrate, in accordance with of PrpD, 2-methylcitrate, were observed in mutants as we desired. For the data showing the lowest enzymatic activity of this mutant towards example, PrpDVTL showed the lowest affinity (1.96 mM of Km) toward 2- citrate. methylcitrate, which contributed to lowering the catalytic efficiency (0.013 mM−1 s −1) by 5.77 times compared to the wild-type PrpD Further flux optimization for increasing production (0.075 mM−1 s−1). On the other hand, wild-type PrpD shows promiscuity The WAICPVL strain was further optimized. First, PrpDVL expression was for catalyzing citrate19 as demonstrated previously. However, this optimized by employing constitutive promoters (Supplementary enzyme showed 12.21-fold lower catalytic efficiency compared to AcnB Fig. 10)31. Itaconate production was significantly affected by PrpDVL (Tables 1, 2), supporting the slight decrease in citrate and increase in expression, as expected, indicating that efficient kinetic compart- itaconate of the WAICP strain (Fig. 3a) compared to the WAIC strain mentalization can allow metabolic flux to be regulated by changing the (Supplementary Fig. 6b). In contrast, the higher affinity for citrate activity of PrpDVL like spatial compartmentalization in A. terreus, where Nature Communications | (2022)13:5353 4 Article https://doi.org/10.1038/s41467-022-33033-1 Table 2 | Kinetic parameters of PrpD, PrpDVTL, PrpDVT, and PrpDVL Enzyme Substrate kcat (s−1) Km (mM) kcat/Km (mM−1 s−1) PrpD 2-methylcitrate 0.045 ± 0.025 0.96 ± 0.85 0.075 ± 0.054 Citrate 10,561.95 ± 3,427.82 66.39 ± 22.03 159.37 ± 2.36 cis-Aconitate 119.03 ± 13.42 0.70 ± 0.10 170.61 ± 5.65 PrpDVTL 2-methylcitrate 0.024 ± 0.008 1.96 ± 0.21 0.013 ± 0.005 Citrate 9,586.41 ± 1,193.47 35.32 ± 5.17 272.17 ± 11.72 cis-Aconitate 81.83 ± 14.14 0.72 ± 0.10 113.60 ± 14.61 PrpDVT 2-methylcitrate 0.023 ± 0.008 1.93 ± 0.21 0.012 ± 0.006 Citrate 6,093.22 ± 1,122.92 14.82 ± 2.38 410.08 ± 22.85 cis-Aconitate 172.08 ± 19.98 1.72 ± 0.25 100.52 ± 3.18 PrpDVL 2-methylcitrate 0.035 ± 0.009 1.72 ± 0.54 0.021 ± 0.002 Citrate 7,804.72 ± 693.10 15.81 ± 1.89 494.97 ± 15.22 cis-Aconitate 96.34 ± 4.55 1.52 ± 0.08 63.63 ± 3.33 Source data are provided as a Source Data file. To further validate the potential of the WAICPG5 strain, 5 L-scale Table 3 | Docking simulations summary fed-batch fermentation was performed (Fig. 5). To obtain enough cell biomass at the initial stage (~8 h), the acetate concentration was kept Receptor Ligand Binding energy (kcal mol−1) H bonds low at 5 g/L and in the subsequent production stage, it was adjusted to PrpD 2-Methylcitrate −5.0 4 10 g/L. As a result, itaconate was rapidly produced until 80 h, up to Citrate −5.0 4 5.06 g/L. In addition, the yield could be maintained (0.33 g/g), similar PrpDV 2-Methylcitrate −4.9 4 to the flask-scale culture during the fermentation, which is equivalent Citrate −4.8 6 to 46% of the theoretical maximum yield from acetate. Considering the PrpDT 2-Methylcitrate −5.2 4 characteristics of acetate as a carbon source33,34, these results are Citrate −5.1 5 remarkably high, and it far exceeds the previous titer (3.57 g/L) and yield (0.09 g/g) of itaconate production from acetate14. We also noted PrpDL 2-Methylcitrate −5.1 4 that itaconate productivity decreased sharply after 80 h in our initial- Citrate −5.0 5 stage reactor study, which is considered to be a spontaneous genetic PrpDVT 2-Methylcitrate −4.9 3 escapee due to acetate toxicity34,35. This will be intensively addressed Citrate −4.8 6 through future reactor study with various approaches, including PrpDVL 2-Methylcitrate −4.9 3 preventing-escapee strategies31,36 and reactor-parameter study37. Citrate −4.8 6 PrpDTL 2-Methylcitrate −5.1 5 Applicability of kinetic compartmentalization to itaconate Citrate −5.0 5 production Although kinetic compartmentalization was successfully applied to the PrpDVTL 2-Methylcitrate −4.9 3 itaconate production from acetate, this strategy may not work for Citrate −4.8 6 other carbon sources such as glucose as the cellular metabolism sig- nificantly differs between carbon sources38,39. Therefore, we validated the potential of our strategy to be broadly applied to the utilization of the precise flux distribution is regulated by MttA activity. The other carbon sources, and it was re-built for the most widely studied WAICP100VL strain with the highest PrpDVL expression showed the glucose-based production (Supplementary Data 1)11,40. highest itaconate production (1.35 g/L) after 48 h of cultivation. Next, The WCAD strain was initially constructed, which harbors cad the TCA cycle and glyoxylate shunt were additionally activated by expressing-plasmid (pCAD) (Supplementary Data 1). Thereafter, the overexpression of citrate synthase and isocitrate lyase encoded by gltA isocitrate dehydrogenase (encoded by icd) knockout strategy, which and aceA for further flux amplification and to facilitate the anaplerotic has been most widely and effectively utilized11,40, was preferentially reaction to maximize itaconate production, respectively14,32 (Fig. 4a, b). applied. Then, phosphate acetyltransferase (pta), acetate kinase (ackA), The expression of aceA was varied to determine the optimized flux and glucose-specific PTS component (pts) were additionally inactivated. distribution between the glyoxylate shunt and TCA cycle14,32. The ita- Finally, phosphoenolpyruvate carboxylase (ppc) was overexpressed for conate titer was increased along with increased expression of aceA the efficient anaplerosis41, resulting in WIAPPC strain (Supplementary (Fig. 4a), and the WAICPG5 strain with the highest expression level of Data 1) as a control strain. The WCAD strain (Supplementary Data 1) aceA showed a 1.52-fold increase in itaconate production (2.03 g/L) consumed most of the glucose (10 g/L) for cell biomass and the for- compared to the WAICP100VL strain after 48 h of cultivation. Finally, mation of acetate and then utilized secondary metabolites to produce itaconate production reached up to 3.13 g/L at 96 h; the yield was itaconate (0.26 g/L, Fig. 6, Supplementary Fig. 11a). On the other hand, maintained at a significant level (0.31 g/g, 43.0% of theoretical max- WIAPPC strain could consistently produce itaconate from the beginning imum yield)14 throughout cultivation and was driven by the sub- and showed a 2.03-fold increase (0.53 g/L) compared to the WCAD stantially increased acetate uptake rate (Fig. 4b)14. Also, it should be strain, however, severe retardations in cell growth and glucose con- noted that other metabolites were not detected meaningfully by HPLC sumption were observed (Supplementary Fig. 11b) without additional in the medium except itaconate and citrate as products. These results TCA supplements (the culture medium already contains 2 g/L of yeast indicate that in addition to the strongly kinetically compartmentalized extract) due to TCA imbalance caused by icd knockout42. flux toward the itaconate, the flux distribution could be optimized for In order to apply the kinetic compartmentalization strategy con- itaconate production. sidering higher TCA flux exists from glucose compared to from Nature Communications | (2022)13:5353 5 Article https://doi.org/10.1038/s41467-022-33033-1 a b 12 10 3.0 Cell biomass (g DCW/L) 2.0 Citrate & Itaconate (g/L) 10 2.5 Itaconate (g/L) Acetate (g/L) 1.5 8 1 2.0 6 1.0 1.5 4 0.1 1.0 0.5 2 0.5 0.0 0 0.01 0.0 0 12 24 36 48 60 72 84 96 WA WA PG1 WA PG2 WA PG3 WA PG4 Time (h) IC IC IC IC ICP Fig. 4 | Further optimization of itaconate production. a Comparison of the G5 y-axis indicates the production of citrate and itaconate (g/L). The x-axis denotes itaconate titer of strains with varied aceA expression for 48 h. White dots indicate time (h). Data were presented as mean values and error bars indicate the standard actual data. b Fermentation profile of WAICPG5 strain. The left y-axis and y-offset deviations from three biological replicates. Source data are provided as a Source represent the cell biomass (g DCW/L) and acetate (g/L), respectively. The right Data file. 20 10 6 have been detected even in some prokaryotes. However, in most pro- karyotes, simultaneous reactions involving numerous substrates within Cell biomass (g DCW/L) Consumed acetate (g/L) 5 a single space are determined only by the kinetic properties of the 15 enzymes, which have evolved to be well-coordinated to maximize cell Itaconate (g/L) 4 growth with high precision. Inspired by these characteristics, we 10 3 developed a non-natural enzymatic reaction, resulting in successful kinetic compartmentalization in prokaryotes for itaconate production. 1 2 Given that around 40–50% of enzymes with known functions have 5 multiple substrates and 10–20% of these multi-substrate-specific 1 enzymes can mediate consecutive reactions (Supplementary Fig. 12 0 0 and see Supplementary Data 2), our approach can be applied to other 0 12 24 36 48 60 72 consecutive reactions to enable metabolic engineering of pathways for Time (h) hardly obtainable intermediates. Numerous studies of the heterologous production of itaconate Fig. 5 | Fed-batch fermentation profile of WAICPG5. The left y-axis and y-offset have consistently focused on substrate availability11,14,49 but could not represent the cell biomass (g DCW/L) and consumed acetate (g/L), respectively. The right y-axis indicates the production of itaconate (g/L). The x-axis denotes time imitate or recapitulate the spatial compartmentalization strategy of its (h). The representative fermentation profiles was plotted. Source data are provided native producer, A. terreus, pumping out cis-aconitate into an inde- as a Source Data file. pendent space13. Therefore, we introduced kinetic compartmentali- zation to provide an efficient supply of cis-aconitate. Rather than using existing aconitase, which is highly reactive to cis-aconitate, the pro- acetate, the activities of competing enzymes of PrpDVL were initially miscuous enzyme PrpD was semi-rationally engineered and used to adjusted43; aconitate hydratase 1 (acnA) was inactivated, and the improve the itaconate production. We carried out the itaconate pro- expression of aconitate hydratase 2 (acnB) was downregulated (Sup- duction in E. coli both using acetate, which is a non-preferred carbon plementary Data 1). As a result, the WBAPAPF strain in which PrpDVL source requiring intensive engineering for its efficient conversion14, was introduced could efficiently produce itaconate (1.68 g/L, Fig. 6 and and glucose, widely utilized in fermentation. Supplementary Fig. 11c). These results correspond to a 6.46-fold In addition to the modified catalytic characteristic of PrpD, the increase compared to the WCAD strain and a 3.16-fold increase com- non-natural cis-aconitate synthesis reaction can be regulated by the pared to the WIAPPC strain with a comparative icd knockout strategy. expression level of the PrpD mutant. Overall, the itaconate titer was Besides increasing itaconate production, the WBAPAPF strain showed increased when a stronger promoter was used to express PrpDVL robust cell growth compared to the WIAPPC strain, suggesting that (Supplementary Fig. 10), indicating that more carbon flux was kineti- additional supplements such as yeast extract or glutamate are not cally compartmentalized according to PrpDVL expression. Surprisingly, required, unlike the previous strategy. Overall, our early-stage valida- the resulting strain WAICPG5 produced up to 5.06 g/L itaconate at tion for glucose-based production further demonstrates the potential 46.0% of the theoretical maximum yield from acetate (Fig. 5), which of our strategy. That is expected to be widely applicable to itaconate showed a 2.86-fold increase in yield compared to the previous study14, production systems in the future44,45. indicating the efficient kinetic compartmentalization. An itaconate-responsive screening system was successfully con- Discussion structed in this study. The initial library was constructed to sufficiently In nature, efficiently designed spatial compartmentalization systems cover all combinations; however, while determining the effect of each have been introduced to preserve substrate availability13 or prevent residue, double mutants with more desired characteristics were iden- damage caused by toxic intermediates15,46. Naturally assembled protei- tified. In addition, the enriched population clearly contained mutants naceous organelles, including carboxysomes47 and metabolosomes48 that improved itaconate production; however, not all mutants showed Nature Communications | (2022)13:5353 6 Article https://doi.org/10.1038/s41467-022-33033-1 a b of PrpD were obtained by site-directed mutagenesis and the Gibson 2.5 2.0 assembly method54. The terminators and promoters for all vectors were obtained from the Registry of Standard Biological Parts (http:// Cell biomass (g DCW/L) 2.0 parts.igem.org). Synthetic 5′ untranslated regions were computation- 1.5 ally designed using UTR Designer (http://sbi.postech.ac.kr/utr_ Itaconate (g/L) 1.5 designer)55. 1.0 1.0 Cell cultivation and enrichment for screening Cells were cultivated in modified minimal acetate medium containing 0.5 0.5 g/L MgSO4·7H2O (M2773), 2.0 g/L NH4Cl (A9434), 1.0 g/L NaCl 0.5 (S5886), 2.0 g/L yeast extract (70161), and 100 mM potassium phos- 0.0 0.0 phate buffer (P8584 and P8709, pH 7.0)14. As a carbon source, 10 g/L neutralized acetate (A6283, pH 7.0) was added to the medium. To WC WI D WB PC WC WI D WB PC maintain the plasmids, antibiotics were added to the medium (50 µg/ AP AP AP AP A A mL streptomycin, 50 µg/mL kanamycin, 34 µg/mL chloramphenicol, AP AP and 100 µg/mL ampicillin). F F For itaconate production, single colonies of each strain were Fig. 6 | Application of kinetic compartmentalization approach for itaconate inoculated into 3 mL of medium in a 15 mL test tube. After 12 h, the cell production from glucose. Comparison of the a cell biomass and b itaconate titer with glucose-utilizing strains after 60 h. The right y-axis indicates the cell biomass cultures were inoculated into 3 or 20 mL of fresh medium in the test and production of itaconate (g/L), respectively. The x-axis denotes strains. Data tube or 300-mL Erlenmeyer flasks, respectively, to an optical density at were presented as mean values and error bars indicate the standard deviations 600 nm (OD600) of 0.05. Isopropyl-β-D-thiogalactopyranoside (I5502) from three biological replicates. White dots indicate actual data. Source data are was initially added to a final concentration of 0.1 mM for induction. provided as a Source Data file. Cultures were performed in biological triplicate with continuous shaking (200 rpm) at 30 °C14. The pH was adjusted to 7.0 by adding an appropriate amount of 5 M HCl solution (H1758). Culture samples were the desired phenotype. After the PrpDVTL mutant dominates during the periodically collected and stored at −80 °C for analysis. The theoretical initial enrichment, hitchhikers may have been enriched together maximum yield from acetate to itaconate was calculated to be 0.72 g/g through crosstalk effects50. Nonetheless, we screened for effective referring to the stoichiometric balance of the previous study14. A 5 mutants in four rounds of enrichment and observed an unintended L-scale reactor study (Marado-PDA, BIOCNS, Daejeon, South Korea) decrease in catalytic efficiency toward cis-aconitate, supporting the was utilized. The recombinant strain was inoculated to 1.2 L of mod- effectiveness of the high-throughput screening system22,25,26,51,52. This ified minimal medium containing 5 g/L of acetate. Sterile air was method can be further applied in diverse strategies for itaconate pumped at a flow rate of 5vvm and 0.1% (v/v) of antifoam 204 (A8311) production, including adaptive laboratory evolution, the evolution of was treated to prevent the foaming. Acetate concentration was inter- cis-aconitate decarboxylase, and even the de novo itaconate produc- mittently added to be maintained above 10 g/L. tion pathway. To validate the screening system and enrichment for screening PrpD mutants, we exerted selection pressure using tetracycline Methods (T7660). The concentration of tetracycline varied from 0–50 mg/L to Oligonucleotides and reagents determine the initial selection pressure. The toxicity of itaconate was Oligonucleotides were obtained from Cosmo Genetech (Seoul, South determined using various concentrations of itaconate (I29204) from Korea). For routine DNA manipulation, Takara PrimeSTAR™ HS DNA 0–2 g/L. The mutant library was initially enriched with 7 mg/L of tet- Polymerase (R040A, Shiga, Japan) was used for DNA amplification. racycline and increased to 15 mg/L over four rounds of enrichment. All Plasmid DNA and genomic DNA were extracted using the AccuPrepR cultures except for the enrichment culture were conducted in Nano-Plus Plasmid Mini Extraction Kit (K-3111G, Bioneer, Daejeon, triplicate. South Korea) and GeneAllR ExgeneTM Cell SV Kit (106-101, GeneAll, Seoul, South Korea), respectively. Enzymes, including ligase and Analytical methods to detect cellular metabolites restriction enzymes, and Gibson assembly were purchased from New Cell biomass (OD600) was measured using a UV-1700 spectro- England Biolabs (Ipswich, MA, USA). Other chemical reagents were photometer (Shimadzu, Kyoto, Japan), and the dry cell weight (DCW) obtained from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise was calculated by converting 1 unit of OD600 to 0.31 g/L56. Metabolites stated. were measured using an Ultimate 3000 high-performance liquid chromatography system (Dionex, Sunnyvale, CA, USA). Filtered sam- Bacterial strains and plasmids ples were analyzed using an Aminex HPX-87H column (Bio-Rad All bacterial strains and plasmids used in this study are listed in Sup- Laboratories, Hercules, CA, USA). In the mobile phase, 5 mM H2SO4 plementary Data 1. Escherichia coli Mach1-T1R (Thermo Fisher Scien- (SX1248) was used at a flow rate of 0.6 mL/min; the temperature of the tific, Waltham, MA, USA) was utilized as a cloning host, and the acid- column oven was maintained at 14 °C29. The refractive index and tolerant E. coli W strain was utilized as an expression host14 and source absorbance at a UV wavelength of 210 nm were monitored using a of wild-type prpD. The WA, WAI, WAIPPC, and WBAPAPF strains were Shodex RI-101 detector (Shodex, Klokkerfaldet, Denmark) and a vari- constructed using the Lambda-Red recombination system with plas- able wavelength detector (Dionex). mids pM_FKF, pKD46, and pCP2053. The pCDF_CAD and pET-P7 plas- mid were used to construct pCAD and pCPHLF, respectively. And Characterization of enzyme kinetics variants of the pCOGA plasmid (pCOGA, pCOGA1,3,4,5) were used to To validate the enzyme kinetics, cells were cultivated for 9 h after amplify the glyoxylate shunt from our previous study14. The itaconate- induction, and cell pellets were resuspended in 40 mM Tris-HCl buffer responsive screening system, a codon-optimized itcR fragment from Y. (Bioneer, C-9006, pH 8.0). The cells were lysed using a Qsonica soni- pseudotuberculosis, was synthesized and assembled with amplified cator (Sonics & Materials, Newtown, CT, USA) for 3 min. To avoid pETduet-1 and tetA fragments24,25. pPRPD was constructed by assem- oxidation of an iron-sulfur cluster of enzymes, dithiothreitol (43816) bling amplified pACYCduet-1 and prpD from E. coli W. Genetic variants and (NH4)2Fe(SO4)2 (09719) were added to final concentrations of 2.5 Nature Communications | (2022)13:5353 7 Article https://doi.org/10.1038/s41467-022-33033-1 and 0.25 mM, respectively8. The cell lysates were centrifuged for reaction, we categorized them as multiple substrate-specific enzymes. 10 min at 13,000×g at 4 °C. The supernatants were utilized to purify the Then, we further characterize enzymes that can conduct consecutive 6X His-tagged enzymes with a MagListoTM His-tagged protein pur- reactions by themselves by checking if the same enzyme can take their ification kit (Bioneer, K-7200) under anaerobic conditions to prevent product in one reaction as substrate in the other reactions. During the the inactivation of aconitase activity8. The elutes were treated with analysis we excluded the following molecules because they can work as 1 mM dithiothreitol, 0.14 mM (NH4)2Fe(SO4)2, and 0.12 mM Na2S cofactors or common substrates: “H2O”, “H+”, “ATP”, “GTP”, “CTP”, (407410) to prevent the oxidization of iron-sulfur clusters. The amount “TTP”, “ADP”, “GDP”, “CDP”, “TDP”, “NADH”, “NADH+”, “NADPH”, “CO2”, of purified enzyme was quantified and adjusted to the same amount “NAD”, “NADP+”, “AMP”, “GMP”, “CMP”, “TMP”, “Fe2+”, “Fe3+”, and using the Bradford assay (B6916). “phosphate”. The enzyme assay was conducted in 2 mM Tris-HCl buffer with In order to find an enzyme that can be used to make kinetic varying amounts of substrate: 1, 2, 5, 10, and 20 mM for citrate (251275); compartmentalization, same as our strategy, we first choose con- 0.05, 0.1, 0.2, 0.5, and 1.0 mM for cis-aconitate (A3412); 0.1, 0.2, 0.4, secutive enzyme reactions of E. coli from KEGG. Then we searched the 0.6, and 1 mM for 2-methylcitrate (59464). This reaction was con- assigned EC (Enzyme Commission) number of enzymes that facilitate ducted at 37 °C for 10 min followed by the inactivation of the enzyme such consecutive reactions. If the enzymes have more than two EC at 90 °C for 5 min. The samples were analyzed using a high- numbers, we took the first EC number and manually find the other performance liquid chromatography system with an XTerraR RP18 enzyme from the super-class of the selected EC number with a similar column and the Aminex HPX-87H column (Bio-Rad Laboratories, substrate structure. As EC number is defined as enzyme classes, sub- Hercules, CA, USA). H2SO4 (5 mM) was utilized as the mobile phase at a classes have similar functions. For example, AcnB is assigned both flow rate of 0.6 mL/min. Absorbance was monitored at a UV wave- EC:4.2.1.3 and EC:4.2.1.99, so we searched the enzymes that have length of 215 nm. All assays were conducted in triplicate. similar substrate structures to AcnB. PrpD (EC: 4.2.1.79) was selected as it converts 2-methylcitrate that has a similar structure to citrate, a Structural analysis for PrpD mutant library design substrate of AcnB. The suggested enzyme list is provided as Supple- To identify the catalytic sites of PrpD, a structural model of PrpD was mentary Data 2. first generated using the structure of the homologous protein MmgE (PDB code: 5MUX)27 from B. subtilis. From the ligand of 5MUX, the Reporting summary catalytic site of MmgE was identified, and well-conserved residues Further information on research design is available in the Nature were found in both PrpD and MmgE (Supplementary Fig. 3)57. A 3D Research Reporting Summary linked to this article. model of 2-methylcitrate was generated from the downloaded SDF format of (2 S,3 S)−2-methylcitrate from PubChem. The SDF file was Data availability converted to PDB format using the online SMILES translator and Data supporting the findings of this work are available within the paper structure file generator (https://cactus.nci.nih.gov/translate/). Finally, and its Supplementary Information files. A reporting summary for this the 3D model of 2-methylcitrate was aligned with the ligand of 5MUX Article is available as a Supplementary Information file. Source data are using LS-Align. From this final model structure, the distance between provided with this paper. the catalytic sites of PrpD and nearby residue from the methyl group of 2-methylcitrate was calculated (Supplementary Table 1). References Among the residues in the catalytic site, those key for catalytic 1. Zhu, J. & Thompson, C. B. Metabolic regulation of cell growth and reactions and important for the interaction with the carboxyl group of proliferation. Nat. Rev. Mol. Cell Biol. 20, 436–450 (2019). citrates were preserved to maintain catalytic activity. For example, 2. Cooper G. M. The Cell: A Molecular Approach (Sinauer Associ- histidine often acts as a proton donor and acceptor, which is critical in ates, 2000). the catalytic reaction, and arginine is an open-form salt bridge with the 3. Chubukov, V., Gerosa, L., Kochanowski, K. & Sauer, U. Coordination anion form of a carboxyl group. Hydrophobic and small residues near of microbial metabolism. Nat. Rev. Microbiol. 12, 327–340 the methyl group of 2-methylcitrate were selected as target residues (2014). for the mutant library. 4. Fani, R. The origin and evolution of metabolic pathways: why and how did primordial cells construct metabolic routes? Evol. Educ. Docking simulations of the mutants Outreach 5, 367–381 (2012). Both the PrpD natural substrate 2-methylcitrate (PubChem ID: 5. De la Fuente, I. M. Elements of the cellular metabolic structure. 5460420) and citrate (PubChem ID: 31348) were docked into the active Front. Mol. Biosci. 2, 16 (2015). pocket of the PrpD enzyme using the crystallographic structure of apo- 6. Zhang, Y. H. P. Substrate channeling and enzyme complexes for protein 2-methylcitrate dehydratase from E. coli (PDB: 1SZQ [https:// biotechnological applications. Biotechnol. Adv. 29, 715–725 (2011). www.rcsb.org/structure/1SZQ])27 using Chimera software58. This pro- 7. Alber, B. E. & Fuchs, G. Propionyl-coenzyme a synthase from tein structure included two chains, one of which was deleted before Chloroflexus aurantiacus, a key enzyme of the 3-hydroxypropionate the docking simulation. The protein was prepared by eliminating water cycle for autotrophic CO2 fixation. J. Biol. Chem. 277, molecules and adding hydrogen and charge. The ligands were also 12137–12143 (2002). charged before analysis. As the presence of the methyl group of 8. Jordan, P. A., Tang, Y., Bradbury, A. J., Thomson, A. J. & Guest, J. R. 2-methylcitrate is an important factor in the substrate specificity of Biochemical and spectroscopic characterization of Escherichia coli PrpD, the orientation of this group was evaluated in all docking aconitases (AcnA and AcnB). Biochem. J. 344, 739–746 (1999). simulations along with the number of hydrogen bonds between the 9. Yang, H. L. & Kessler, D. P. Genetic analysis of the leucine region in ligand and protein residues resulting from each analysis. Escherichia coli B/r: gene enzyme assignments. J. Bacteriol. 117, 63–72 (1974). Characterization of multi-substrate and consecutive enzymes 10. Wang, J. et al. Microbial production of branched-chain dicarbox- We collected the enzyme reactions of E. coli from KEGG (https:// ylate 2-methylsuccinic acid via enoate reductase-mediated bior- pubmed.ncbi.nlm.nih.gov/10592173/) and BioCyc (https://pubmed. eduction. Metab. Eng. 45, 1–10 (2018). ncbi.nlm.nih.gov/29447345/). To characterize enzymes that have mul- 11. Okamoto, S. et al. Production of itaconic acid using metabolically tiple substrate specificities, we counted the number of known reactions engineered escherichia coli. J. Gen. Appl. Microbiol. 60, for each enzyme. If the enzyme is involved in more than one enzymatic 191–197 (2014). Nature Communications | (2022)13:5353 8 Article https://doi.org/10.1038/s41467-022-33033-1 12. Werpy, T. & Petersen, G. Top Value Added Chemicals from Biomass: 34. Seong, W. et al. Adaptive laboratory evolution of Escherichia coli Volume I - Results of Screening for Potential Candidates from Sugars lacking cellular byproduct formation for enhanced acetate utiliza- and Synthesis Gas. https://doi.org/10.2172/15008859 (2004). tion through compensatory ATP consumption. Metab. Eng. 62, 13. Tevž, G., Benčina, M. & Legiša, M. Enhancing itaconic acid pro- 249–259 (2020). duction by Aspergillus terreus. Appl. Microbiol. Biotechnol. 87, 35. Pinhal, S., Ropers, D., Geiselmann, J. & de Jong, H. Acetate meta- 1657–1664 (2010). bolism and the inhibition of bacterial growth by acetate. J. Bio- 14. Noh, M. H., Lim, H. G., Woo, S. H., Song, J. & Jung, G. Y. Production technol. 201, 1–19 (2019). of itaconic acid from acetate by engineering acid-tolerant Escher- 36. Rottinghaus, A. G., Ferreiro, A., Fishbein, S. R. S., Dantas, G. & Moon, ichia coli W. Biotechnol. Bioeng. 115, 729–738 (2018). T. S. Genetically stable CRISPR-based kill switches for engineered 15. Chen, A. H. & Silver, P. A. Designing biological compartmentaliza- microbes. Nat. Commun. 13, 1–17 (2022). tion. Trends Cell Biol. 22, 662–670 (2012). 37. Soini, J., Ukkonen, K. & Neubauer, P. High cell density media for 16. Hammer, S. K. & Avalos, J. L. Harnessing yeast organelles for Escherichia coli are generally designed for aerobic cultivations - metabolic engineering. Nat. Chem. Biol. 13, 823–832 (2017). Consequences for large-scale bioprocesses and shake flask cul- 17. Scarcia, P. et al. Mitochondrial carriers of Ustilago maydis and tures. Microb. Cell Fact. 7, 1–11 (2008). Aspergillus terreus involved in itaconate production: same phy- 38. Oh, M. K., Rohlin, L., Kao, K. C. & Liao, J. C. Global expression pro- siological role but different biochemical features. FEBS Lett. 594, filing of acetate-grown Escherichia coli. J. Biol. Chem. 277, 728–739 (2020). 13175–13183 (2002). 18. van der Straat, L. et al. Expression of the Aspergillus terreus itaconic 39. Lim, H. G., Lee, J. H., Noh, M. H. & Jung, G. Y. Rediscovering acetate acid biosynthesis cluster in Aspergillus niger. Microb. Cell Fact. 13, metabolism: its potential sources and utilization for biobased 1–9 (2014). transformation into value-added chemicals. J. Agric. Food Chem. 19. Brock, M., Maerker, C., Schütz, A., Völker, U. & Buckel, W. Oxidation 66, 3998–4006 (2018). of propionate to pyruvate in Escherichia coli: Involvement of 40. Harder, B.-J., Bettenbrock, K. & Klamt, S. Model-based metabolic methylcitrate dehydratase and aconitase. Eur. J. Biochem. 269, engineering enables high yield itaconic acid production by 6184–6194 (2002). Escherichia coli. Metab. Eng. 38, 29–37 (2016). 20. Rix, G. et al. Scalable continuous evolution for the generation of 41. Song, C. W., Kim, D. I., Choi, S., Jang, J. W. & Lee, S. Y. Metabolic diverse enzyme variants encompassing promiscuous activities. Nat. engineering of Escherichia coli for the production of fumaric acid. Commun. 11, 1–11 (2020). Biotechnol. Bioeng. 110, 2025–2034 (2013). 21. Leveson-Gower, R. B., Mayer, C. & Roelfes, G. The importance of 42. Joyce, A. R. et al. Experimental and computational assessment of catalytic promiscuity for enzyme design and evolution. Nat. Rev. conditionally essential genes in Escherichia coli. J. Bacteriol. 188, Chem. 3, 687–705 (2019). 8259–8271 (2006). 22. Yeom, S. J. et al. A synthetic microbial biosensor for high- 43. Enjalbert, B., Letisse, F. & Portais, J.-C. Physiological and molecular throughput screening of lactam biocatalysts. Nat. Commun. 9, timing of the glucose to acetate transition in Escherichia coli. 1–12 (2018). Metabolites 3, 820–837 (2013). 23. Williams, C. H. et al. E. coli aconitase B structure reveals a HEAT-like 44. Yang, Z., Wang, H., Wang, Y., Ren, Y. & Wei, D. Manufacturing domain with implications for protein – protein recognition. 4, multienzymatic complex reactors in vivo by self-assembly to 447–452 (2002). improve the biosynthesis of itaconic acid in Escherichia coli. ACS 24. Hanko, E. K. R., Minton, N. P. & Malys, N. A transcription factor-based Synth. Biol. 7, 1244–1250 (2018). biosensor for detection of itaconic acid. ACS Synth. Biol. 7, 45. Jeon, H. G., Cheong, D. E., Han, Y., Song, J. J. & Choi, J. H. Itaconic 1436–1446 (2018). acid production from glycerol using Escherichia coli harboring a 25. Yang, J. et al. Synthetic RNA devices to expedite the evolution of random synonymous codon-substituted 5′-coding region variant of metabolite-producing microbes. Nat. Commun. 4, 1413 (2013). the cadA gene. Biotechnol. Bioeng. 113, 1504–1510 (2016). 26. Seok, J. Y. et al. Directed evolution of the 3-hydroxypropionic acid 46. Kulagina, N., Besseau, S., Papon, N. & Courdavault, V. Peroxisomes: production pathway by engineering aldehyde dehydrogenase a new hub for metabolic engineering in yeast. Front. Bioeng. Bio- using a synthetic selection device. Metab. Eng. 47, 113–120 (2018). technol. 9, 7–12 (2021). 27. Berman, H. M. et al. The protein data bank. Nucleic Acids Res. 28, 47. Cameron, J. C., Wilson, S. C., Bernstein, S. L. & Kerfeld, C. A. Bio- 235–242 (2000). genesis of a bacterial organelle: the carboxysome assembly path- 28. Seok, J. Y. et al. Synthetic biosensor accelerates evolution by way. Cell 155, 1131 (2013). rewiring carbon metabolism toward specific metabolite. Cell Rep. 48. Liu, L. N. Bacterial metabolosomes: new insights into their structure 36, 109589 (2021). and bioengineering. Microb. Biotechnol. 14, 88–93 (2021). 29. Jo, M. et al. Precise tuning of the glyoxylate cycle in Escherichia coli 49. Otten, A., Brocker, M. & Bott, M. Metabolic engineering of Cor- for efficient tyrosine production from acetate. Microb. Cell Fact. 18, ynebacterium glutamicum for the production of itaconate. Metab. 1–9 (2019). Eng. 30, 156–165 (2015). 30. Yao, J. et al. Developing a highly efficient hydroxytyrosol whole-cell 50. Flachbart, L. K., Sokolowsky, S. & Marienhagen, J. Displaced by catalyst by de-bottlenecking rate-limiting steps. Nat. Commun. 11, deceivers: prevention of biosensor cross-talk is pivotal for suc- 1515 (2020). cessful biosensor-based high-throughput screening campaigns. 31. Kang, C. W. et al. Synthetic auxotrophs for stable and tunable ACS Synth. Biol. 8, 1847–1857 (2019). maintenance of plasmid copy number. Metab. Eng. 48, 51. Noh, M. H., Cha, S., Kim, M. & Jung, G. Y. Recent advances in 121–128 (2018). microbial cell growth regulation strategies for metabolic engi- 32. Noh, M. H., Lim, H. G., Park, S., Seo, S. W. & Jung, G. Y. Precise flux neering. Biotechnol. Bioprocess Eng. 25, 810–828 (2020). redistribution to glyoxylate cycle for 5-aminolevulinic acid pro- 52. Nguyen, N. H., Kim, J. R. & Park, S. Application of transcription duction in Escherichia coli. Metab. Eng. 43, 1–8 (2017). factor-based 3-hydroxypropionic acid biosensor. Biotechnol. Bio- 33. Song, H. S. et al. Enhanced isobutanol production from acetate by process Eng. 23, 564–572 (2018). combinatorial overexpression of acetyl-CoA synthetase and ana- 53. Datsenko, K. A. & Wanner, B. L. One-step inactivation of chromo- plerotic enzymes in engineered Escherichia coli. Biotechnol. somal genes in Escherichia coli K-12 using PCR products. Proc. Natl Bioeng. 115, 1971–1978 (2018). Acad. Sci. USA 97, 6640–6645 (2000). Nature Communications | (2022)13:5353 9 Article https://doi.org/10.1038/s41467-022-33033-1 54. Gibson, D. G. et al. Enzymatic assembly of DNA molecules up to Additional information several hundred kilobases. Nat. Methods 6, 343–345 (2009). Supplementary information The online version contains 55. Seo, S. W. et al. Predictive design of mRNA translation initiation supplementary material available at region to control prokaryotic translation efficiency. Metab. Eng. 15, https://doi.org/10.1038/s41467-022-33033-1. 67–74 (2013). 56. Ko, Y. et al. Coenzyme B12 can be produced by engineered Correspondence and requests for materials should be addressed to Escherichia coli under both anaerobic and aerobic conditions. Jae-Seong Yang or Gyoo Yeol Jung. Biotechnol. J. 9, 1526–1535 (2014). 57. Sievers, F. et al. Fast, scalable generation of high-quality protein Peer review information Nature Communications thanks Matthias Stei- multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. ger, Xiao-Xia Xia and the other, anonymous, reviewer(s) for their 7, 539 (2011). contribution to the peer review of this work. 58. Pettersen, E. F. et al. UCSF chimera - a visualization system for exploratory research and analysis. J. Comput. Chem. 25, Reprints and permission information is available at 1605–1612 (2004). http://www.nature.com/reprints Acknowledgements Publisher’s note Springer Nature remains neutral with regard to This research was supported by the C1 Gas Refinery Program jurisdictional claims in published maps and institutional affiliations. (NRF-2018M3D3A1A01055754, G.Y.J.) and the grants (NRF- 2019R1A2C2084631, G.Y.J. and NRF-2021R1A6A3A03043982, M.H.N.) Open Access This article is licensed under a Creative Commons from the National Research Foundation (NRF) of Korea. We also Attribution 4.0 International License, which permits use, sharing, acknowledge that this work was supported by grants SEV‐2015‐0533 adaptation, distribution and reproduction in any medium or format, as and CEX2019-000902-S funded by MCIN/AEI/10.13039/501100011033, long as you give appropriate credit to the original author(s) and the and by the CERCA Programme / Generalitat de Catalunya (J.-S.Y.). Also, source, provide a link to the Creative Commons license, and indicate if this project has received funding from the European Union’s Horizon changes were made. The images or other third party material in this 2020 research and innovation programme under the Marie Skłodowska- article are included in the article’s Creative Commons license, unless Curie grant agreement No 945043 (A.M.). indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended Author contributions use is not permitted by statutory regulation or exceeds the permitted D.Y., M.H.N., J.H.M., M.K., A.M., J.W.L., J.-S.Y., and G.Y.J. conceived the use, you will need to obtain permission directly from the copyright project, conducted data analysis and interpretation, and wrote the holder. To view a copy of this license, visit http://creativecommons.org/ manuscript. D.Y., M.H.N., and J.H.M. designed and conducted the licenses/by/4.0/. experiments. J.-S.Y. and G.Y.J. supervised the project. All the authors read and approved the final manuscript. © The Author(s) 2022 Competing interests The authors declare no competing interests. Nature Communications | (2022)13:5353 10