2.3 Regulation of metabolic reactions via mass action
Besides the enzyme concentration, another critical component influencing a biochemical reaction rate is the substrate concentration. If both enzymes and substrates are enriched in the same compartment, the reactions can be accelerated. If they are separated from each other, the enzymatic reactions can be shut down (Figure 1B). Considering that the main driving force of FUSLCD LLPS is pi-pi interaction, compounds enriched with pi-electrons could be included in the compartment as clients theoretically[13, 28]. To explore the regulatory effect of PhASE#1 on enzymatic reactions, the enzyme of interest (EOI) was tethered to the light-responsive module as the experimental group (EXP), so that it could be recruited to the membraneless compartments according to light-induced CRY2-CIB1 interaction. In the control groups (CTRL), EOIs were fused with mCherry to perform catalysis in a fully dispersed state, excluding the influence of protein fusion on enzymatic activity (Figure 4A). Then, both groups were treated with light and the reaction rate difference between them was derived.
As a proof of concept, the coelenterazine (CLZ) oxidation reaction catalyzed by Renilla luciferase (Rluc) was tested, which was a reporter system frequently harnessed for in vivo studies[43] (Figure 4B), and the luminescence generated during the process was used as a direct measurement of reaction rate[44] (Figure 4C). After the light induction, EOIs were successfully recruited into the membraneless compartments (Supplementary Figure 4A). In the first 20 minutes after CLZ addition, the EXP group displayed a significantly higher oxidation rate compared to that of the CTRL group, with 2.6 times of coelenteramide (CLM, the oxidative product of CLZ) accumulation difference. Since the overall substrate and enzyme concentration remained the same between these two groups (Figure 4B), we deduced that CLZ should have been enriched by the compartment from the cytosol, otherwise the limitation of enzyme movement in cells would not benefit the reaction efficiency. Remarkably, the concentration of bacteria did not have a major impact on this reaction at a proper range as long as the overall enzyme concentration remained at the same level (Supplementary Figure 5). With four biological replicates, it was concluded that PhASE#1 could indeed regulate this enzymatic reaction.
Subsequently, more biochemical reactions were conducted for possible practical applications. Catechol 2,3-dioxygenase (XylE) catalyzes the oxidation of catechol, which is a key metabolite in many organic degradation pathways[45, 46] (Figure 4D). Like the design above, a modified light-responsive module tethered with XylE was constructed and was demonstrated to be successfully enriched into the LLPS-based compartments (Supplementary Figure 4B). Both incubated under daylight, the EXP group showed a higher reaction rate compared to that of the CTRL group at the same enzyme and substrate levels, with 1.6 times 2-hydroxymuconate semialdehyde (2-HMS, the oxidative product of catechol) accumulation difference 4.5 minutes after the substrate addition (Figure 4E). From these results, the regulatory ability of PhASE on enzymatic reactions was demonstrated. Additionally, they also provided evidence for the principle of regulating metabolic reactions via changing the distribution correlation of enzymes and substrates.
Furthermore, we have built a quantitative model based on mass action and Michaelis-Menten equation to explore the parameters affecting the regulatory capacity of a de-mixed system on biochemical reactions, in which the acceleration of the reaction via the colocalization of the substrate and enzyme has been verified. Interestingly, we also found that the size of the compartments had significant effect on the regulatory capacity.
Discussion
Tuning metabolic reactions via regulations on the enzyme expression level has been studied comprehensively[1, 47]. Yet in many cases, changing the concentration of protein can cause additional metabolic burden and a waste of time[2, 40]. Here, we engineered protein-based condensates into a reaction regulator named PhASE, changing reaction rate via the mechanism of mass action. Without the need for protein synthesis during regulation, our tool enabled the control of enzymatic reactions at a higher temporal resolution, completing in a few seconds. Also, the reversibility of PhASE allows the dynamic regulation of reaction velocity, while the traditional regulations via chemical inducers and non-inducible scaffolds can hardly be reversed[3, 6]. Additionally, our exploration of chemical interactions between small molecular client and protein scaffold provided the potential of regulating vast metabolic reactions via condensate-based approach[24] (Figures 1D, 4B and 4E).
However, there are still some considerations that should be taken into using PhASE. It can be difficult to find such protein scaffolds for every specific chemical to enrich them at arbitrary folds. Remarkably, too strong interactions between chemical client and protein scaffold may increase reaction efficiency either since the scaffold will compete substrate binding with enzymes. Besides, to design those compartments in specific sizes can be difficult, except for tuning the expression level of scaffold proteins (Figure 2). Inappropriate fractional volume of the condensates can lead to limited regulatory capacity (Figure S7).
Nevertheless, our system has many potential applications. With a very high regulatory temporal resolution and reversibility, the PhASE system allows the switch of enzymatic reactions in a higher frequency, while the pulsatile activation of different pathways has been proved to enhance biosynthetic production[12]. Besides metabolic regulations, PhASE system also has the potential to regulate and study various biological activities. By quickly changing the spatial correlation of substrates and enzymes, it is suitable for studying transient biochemical reactions and the switch of signaling pathways (Figure 2). Some of these processes last only for a few seconds, which can hardly be rebuilt and studied via transcriptional and translational control, which achieves the maximum regulatory effect in minutes or even hours[42, 48]. Additionally, the fold change of reaction rate between two regulatory states may be increased even more if our approach can be integrated with the proximity-based regulation of metabolic pathways[16].
Material and Methods