Determination of fusion protein location
To quickly and conveniently detect the expression and observe the self-assembly form of proteins LdcI, AsnC, IadA, FucU PanB and AdiA in the CFPS system, YFP was first fused to them. However, the influence of the fusion protein YFP on protein self-assembly was unknown. Therefore, YFP was linked to the N-terminus or C-terminus of the proteins, respectively. It could be seen in Fig. 2 that the position of the fusion protein YFP had an effect on the protein expression, and the expression of proteins with YFP at the N-terminus was higher than that at C-terminus. As shown in Fig. 2B and Additional file 1: Table S2, the fluorescence value of proteins with YFP at the N-terminus was 0.5–5 times higher than that at C-terminus. The solubility rate of proteins with YFP at the N-terminus was 1–18 times higher than that at C-terminus. Among different constructs, the protein FucU showed the best expression yield and solubility. Similarly, western blot analysis in Fig. 2C further confirmed that the expression and solubility of proteins with YFP at N-terminus were much better than that at C-terminus, and the protein FucU was great.
Subsequently, the total proteins of these 12 protein constructs were observed through a confocal microscope for analyzing the self-assembly (Fig. 2D, Additional file 1: Fig. S11). The proteins with YFP at N-terminus could self-assemble into 1–10 µm polymers, which were 1–10 times larger than that with YFP at C-terminus. It could also be seen that YFP-FucU could self-assemble into a 10 µm polymer. The other five proteins with different symmetry also showed different assemblies. Overall, by comparing the construct expression, western blot analysis, and confocal microscope observation, the fusion proteins were more suitable for connecting to the N-terminus of scaffold proteins. In addition, the expression and self-assembly of these six proteins fused with N-terminal YFP in the CFPS system were also demonstrated by size-exclusion chromatography. Compared with the YFP protein that did not form a polymer, the other six proteins showed a shorter retention time or a smaller elution volume, which meant that all of them formed high-molecular-weight polymers. (Additional file 1: Fig. S10).
Analysis of self-assembly protein solubility
It has been proved that the solubility of these proteins could reach 60–90%, and the proteins could form large polymers of 10 µm. However, it was not confirmed whether the insoluble precipitate contained large polymers. To understand the self-assembly status of these six proteins more precisely, the protein constructs in the CFPS products were subsequently purified for further analysis. Observed by the confocal microscope, fluorescent macromolecules were seen in both the purified soluble protein and the precipitate after centrifugation (Fig. 3A, B). The biomacromolecules larger than 10 μm were observed in the precipitates of YFP-LdcI and YFP-AsnC. To gain structural insights, these purified proteins were observed under the transmission electron microscope (TEM) (Fig. 3C). However, the polymers were only shown in the purified total proteins and were not observed in the purified precipitates. As shown in Fig. 3C, YFP-AsnC, YFP-FucU, and YFP-PanB could form 100–500 nm protein polymers.
Effects of CFPS conditions
After determining the expression of proteins in the CFPS system, the reaction environment was further explored to improve the expression and assembly. Redox environment affects protein folding by affecting disulfide bond formation (Yin and Swartz 2004; Chakraborty et al. 2021), and therefore a total of 8 different molar ratios of GSSG to GSH (1: 2, 1: 4, 1: 9, 2: 1, 4: 1, 9: 1, 1: 1, and 0: 0) were added into the CFPS system to obtain different redox potentials, and their effects on protein expression and self-assembly were explored. It could be seen that the presence of glutathione increased the expression of YFP-AsnC, YFP-IadA and YFP-FucU (Fig. 4A, B), but had few effects on the expression of other constructs. The presence of glutathione had few effects on the solubility of six proteins. Then YFP-FucU was selected to observe the assembly morphology under the confocal microscope (Fig. 4C). The folded state of YFP-FucU did not change much. It could be seen from the results that the presence of glutathione may affect some elements of the CFPS system, thereby affecting the synthesis and expression of the three proteins YFP-AsnC, YFP-IadA and YFP-FucU. However, changes in these elements had little effect on the expression of the other three proteins and the self-assembly of all proteins. This result also reflected that these six proteins could form more stable states in cell-free systems, thus showing greater potential for forming stable biomaterials.
It was possible to weaken or enhance the intermolecular electrostatic interactions and hydrophobic interactions by altering the salt concentration of the solution, thereby affecting the protein assembly (Donnarumma et al. 2020; Lefevre et al. 2020; Zhang et al. 2021). So the effect of different concentrations of NaCl (0 mM, 100 mM, 250 mM, and 500 mM) on cell-free protein self-assembly was explored in this article. It could be seen from Fig. 5A that higher salt concentrations could decrease the expression of protein constructs and had few effects on their solubilities. As shown in Fig. 5B, the high concentration of NaCl (500 mM) might disrupt the assembly of proteins. This result showed that these proteins could still form polymers at lower salt concentration, showing a relatively stable state, but the self-assemble effect was weakened at 500 mM salt concentration, which may be due to the fact that the higher salt concentration disrupts the intramolecular interactions of the protein.
Universal applicability of self-assembly proteins
To explore the effects of replacing the fusion proteins on the self-assembly of proteins, YFP was replaced by sfGFP or mCherry. By measuring the expression fluorescence value (Fig. 6A, B) and observing the self-assembly morphology under the confocal microscope (Fig. 6C, D, Additional file 1: Fig. S11), the universal applicability of these proteins was verified. With the confocal images of proteins with YFP (Additional file 1: Figs. S12–S17), AsnC, FucU, LdcI, and AdiA self-assembled into large polymers of about 10 µm, and IadA and PanB formed small polymers of less than 5 µm. Some proteins were slightly different under confocal microscopy imaging. The reason for this difference is related to the noise generated by the cell-free system and the instrument. From the whole field of view, the fusion of sfGFP and mCherry had no effect on the assembly of the six proteins. All of these results proved that the scaffold proteins in this study could be fused with different functional molecules to form polymers in the CFPS systems for various applications.
Functional verification of self-assembly proteins
Although the self-assembly properties of these proteins have been characterized, the functionality of the protein polymer needs to be confirmed when the enzymes were fused to the polymer scaffold. To demonstrate their performance, xylanase (TFX) was fused to the N-terminus of the proteins to test its catalytic activity. The TFX is an important enzyme that converts lignocellulose (a high-potential renewable resource) into xylose, and xylose has received widespread attention as an alternative carbon source to replace glucose or starch for different applications, such as biofuels (Sun et al. 2019).
The proteins carrying TFX were expressed and self-assembled in the CFPS system, and the protein expression analysis (Fig. 7A, Additional file 1: Fig. S18) and catalytic activity analysis (Fig. 7B) were performed. The specific enzyme activities of 6 constructs were analyzed, as shown in Fig. 7C. The specific activities of TFX enzymes fused with different proteins were different, which might be attributed to different structural effects, but they were all higher than that of TFX alone. This may be due to the steric proximity effect caused by the ordered linkage with symmetric protein scaffold, the specific activity of the enzyme was related to its spatial structure, such that TFX fused to the six proteins more active than TFX alone (Huang et al. 2012). These results validated that the supramolecular assembly of proteins could be used as polymer scaffolds for biocatalysis applications.
Comparison of six proteins expressing and self-assembling in cell-free or cell systems
All six proteins selected in this work were also expressed and self-assembled in E. coli cells. However, these cells grew slowly, which demonstrated that these proteins might have toxic or inhibitory effects on the cell growth. According to the fluorescence level, it could be seen that the expression of these proteins in cells were low, as the highest fluorescence in cell culture was only 26,500 (YFP-AdiA), which was much lower than the fluorescence level of the proteins expressed in the CFPS system (Fig. 8A). Combined with the results of the western blot analysis, it was found that YFP-IadA and YFP-PanB were not expressed well (Fig. 8B). The intracellular self-assembly of these proteins was also observed under confocal fluorescence microscopy. Due to the tethering of the cell membrane, these proteins were only able to form aggregates no larger than 5 µm (Fig. 8C).Therefore, it was obvious that the CFPS system was more suitable for the controllable expression and forming larger polymers.