MODEL

We modeling team employed metabolic flux analysis (MFA) and MATLAB-based dynamic simulations to investigate the role of 2-oxoglutarate (2-OG) in flavonoid biosynthesis of modified engineering strain of Bacillus subtilis. The simulation results demonstrated that 2-OG can theoretically enhance the production of luteolin, a key flavonoid secreted by plant roots to attract rhizobia. As the core effector of our overall goal, increased luteolin levels are expected to strengthen rhizobial chemotaxis, thereby improving nodulation efficiency and nitrogen fixation in plants. Beyond confirming the stimulatory role of 2-OG, our modeling efforts also provided valuable insights into potential genetic engineering targets and metabolic regulation strategies for future experimental work.

The synthesis pathway of Luteolin is closely associated with 2-oxoglutarates. In this pathway, the Flavone synthase I (FNS I) acts as the executive enzyme for the conversion of naringenin to apigenin and eriodictyol to luteolin—both reactions are key steps in luteolin synthesis. FNS I is a typical 2-oxoglutarate-dependent dioxygenase (2-ODDs), which requires 2-oxoglutarate (2-OG) as a cofactor to exert its enzymatic activity[1].

Based on the flow diagram of the 2-oxoglutarate-related pathway as shown in Figure 1, the ordinary differential equations for the amount of substance and flux equations of each relevant metabolite can be obtained as follows.

(Notes: N: Naringenin; A: Apigenin; L: Luteolin; E: Dihydrokaempferol; O: 2-oxoglutarate)

According to an article published by Hausinger, R. P., 2004, the steady-state effective concentration of 2-OG is typically in the order of 0.1~1μM, which is much lower than that of the Michaelis constant(10~50μM) for 2-OG-dependent enzymatic reactions[3]. Therefore, this paper considers the concentration of 2-OG to be low.

Regard the denominator of in equation (6):

If the [O] satisfy the assumption

Then,

Similarly, If the [O] also satisfy the assumption

Then,

Assuming that [A] reaches a steady state quickly,

Then according to the equation (2),

Thus, under the hypothetical conditions of

The equation (5) is equivalent to

Solve this O.D.E, we can get

Substitude the values of β and γ, we can eventually get

Under the steady state,

The simulation process for the steady-state concentration of each substances is as follows.

(Notes: The values of each maximum reaction rate and the loss coefficient of Luteolin and Dihydrokaempferol were not found. For the convenience of simulation, this paper makes reasonable assumptions based on the common order of magnitude in MFA.)

Based on the parameters (Table 1) and equations (O.D.E and Flux Equations), we used the ode45 function in MATLAB to obtain the relationship graph (Figure 3) between the steady-state concentration of each substance and the amount of 2-oxoglutarate, where 2-oxoglutarate was set from 0.5 to 2 μM.

As shown in Figure 2, under steady state with existence of 0.001mM 2-OG in the system, the concentrations of each substance are as follows:

Additionally, as shown in Figure 2, when the concentration of 2-OG is 0.5, 1.0, 1.5 and 2.0 μM respectively, the concentration of Luteolin is correspondingly 0.000088, 0.000158, 0.000218, and 0.000270 mM. In other words, when the concentration of 2-OG is at a relatively low level, the concentration of Luteolin increases as the concentration of 2-OG rises.

Based on the Parameters shown in Table 1 and 2, the process of the verification of hypothetical conditions (Δ) is as follows:

Under relative steady state, d[A]/dt ≈ 0 holds true naturally, and thus no proof is provided here. Based on the above process, we successfully verified the validity of the hypothetical condition (Δ) .

When the concentration of 2-oxoglutarate is relatively low, the concentration of Luteolin [L] is proportional to the concentration of 2-oxoglutarate [O] under steady state.

In order to identify genes in B. subtilis whose overexpression can effectively promote 2-oxoglutarate (2-OG) synthesis, we utilized the iYO844 model corresponding to B. subtilis str. 168 used in the experimental group. First, we performed individual gene knockout simulations on the model bacterium in Matlab, and screened out 17 genes whose knockout affected 2-OG synthesis flux but did not substantially impact the biomass reaction (linked to biological growth rate). These genes were further screened according to specific criteria. Subsequently, we conducted overexpression of the candidate genes. The results showed that simultaneous overexpression of citB and icd could increase the 2-OG synthesis flux of wild-type B. subtilis by 19.62% while maintaining normal biomass reaction rate.

We performed individual knockout simulations for each gene in B. subtilis str. 168(844 genes totally), and screened out the following 17 genes, whose knockout reduces 2-oxoglutarate (2-OG) synthesis without significantly affecting the growth of the bacterial strain.

(Note: Wild-type 2-OG flux: 10.0827, biomass reaction flux: 0.6242)

Among them, the regulatory genes corresponding to essential central energy metabolism reactions and reactions that do not directly control 2-OG production were excluded to prevent unknown impacts on other normal metabolic processes of the engineered strain. The final candidate genes are odhB, odhA, pdhD, citB and icd.

The basic information table of the candidate genes is as follows:

We conducted overexpression of the 5 candidate genes via individual traversal and combinatorial traversal methods. The result is as follows:

As shown in Table 5 and Figures 4, overexpression of each group had no effect on the biomass reaction. Individual overexpression of each candidate gene did not affect 2-oxoglutarate (2-OG) synthesis, while simultaneous overexpression of citB and icd increased the flux of the 2-OG demand reaction (corresponding to the intracellular accumulation rate) by 19.62%. Additionally, further overexpression of the other three genes on this basis did not change the increase in demand reaction flux.

In conclusion, simultaneous overexpression of the citB and icd genes in Bacillus subtilis str. 168 can significantly enhance the intracellular 2-OG synthesis.

As shown in Figure 5, The citB gene regulates the ACONT enzyme (Aconitate Hydratase) , which catalyzes the conversion of citrate to isocitrate; the icd gene regulates the ICDH enzyme (Isocitrate Dehydrogenase) , which catalyzes the conversion of isocitrate to 2-OG. Simultaneously increasing the expression of these two genes will raise the biosynthesis of 2-OG.

In this study, we demonstrated through metabolic flux analysis and MATLAB simulations that 2-oxoglutarate (2-OG) can act as a potential booster of luteolin biosynthesis. Furthermore, gene-level modeling highlighted citB and icd as promising candidates for enhancing 2-OG production. Looking ahead, subsequent researchers can validate these findings experimentally and explore metabolic engineering strategies to sustainably elevate intracellular 2-OG levels, thereby reinforcing luteolin synthesis and secretion. This approach will not only strengthen plant-rhizobium interactions but also open up new possibilities for engineering root-associated signaling networks to support sustainable agriculture.

1. Dan-Dan Li, Rong Ni, Ping-Ping Wang, Xiao-Shuang Zhang, Piao-Yi Wang, Ting-Ting Zhu, Chun-Jing Sun, Chang-Jun Liu, Hong-Xiang Lou, Ai-Xia Cheng. (2020) Molecular Basis for Chemical Evolution of Flavones to Flavonols and Anthocyanins in Land Plants. Plant Physiol. 184(4):1731-1743.
2. Kanehisa, L. (2019, June 21). Flavonoid biosynthesis - Reference pathway https://www.kegg.jp/pathway/map=map00941&keyword=Nari
3. Robert, P. FeII/alpha-ketoglutarate-dependent hydroxylases and related enzymes (2004), Crit Rev Biochem Mol Biol., 39(1):21-68.
4. Britsch, L. Purification and characterization of flavone synthase I, a 2-oxoglutarate-dependent desaturase (1990), Arch. Biochem. Biophys., 282, 152-160.
5. Han, X.J.; Wu, Y.F.; Gao, S.; Yu, H.N.; Xu, R.X.; Lou, H.X.; Cheng, A.X. Functional characterization of a Plagiochasma appendiculatum flavone synthase I showing flavanone 2-hydroxylase activity (2014), FEBS Lett., 588, 2307-2314.
6. Thill, J.; Miosic, S.; Gotame, T.P.; Mikulic-Petkovsek, M.; Gosch, C.; Veberic, R.; Preuss, A.; Schwab, W.; Stampar, F.; Stich, K.; Halbwirth, H. Differential expression of flavonoid 3-hydroxylase during fruit development establishes the different B-ring hydroxylation patterns of flavonoids in Fragaria x ananassa and Fragaria vesca (2013), Plant Physiol. Biochem., 72, 72-78.
7. Khumkarjorn, N.; Thanonkeo, S.; Yamada, M.; Thanonkeo, P. Cloning and expression analysis of a flavanone 3-hydroxylase gene in Ascocenda orchid (2017), J. Plant Biochem. Biotechnol., 26, 179-190.
8. Gómez et al., 2002, Phytochemistry, 59: 187–193.
9. Kanehisa, L. (2024, October 12). HIF-1 signaling pathway - Reference pathway. KEGG Pathway Database. https://www.kegg.jp/pathway/map=map04066&keyword=2-OG