Whole-Cell Display of Phosphotransferase in Escherichia coli for High-Efficiency Extracellular ATP Production

Adenosine triphosphate (ATP), as a universal energy currency, takes a central role in many biochemical reactions with potential for the synthesis of numerous high-value products. However, the high cost of ATP limits industrial ATP-dependent enzyme-catalyzed reactions. Here, we investigated the effect of cell-surface display of phosphotransferase on ATP regeneration in recombinant Escherichia coli. By N-terminal fusion of the super-folder green fluorescent protein (sfGFP), we successfully displayed the phosphotransferase of Pseudomonas brassicacearum (PAP-Pb) on the surface of E. coli cells. The catalytic activity of sfGFP-PAP-Pb intact cells was 2.12 and 1.47 times higher than that of PAP-Pb intact cells, when the substrate was AMP and ADP, respectively. The conversion of ATP from AMP or ADP were up to 97.5% and 80.1% respectively when catalyzed by the surface-displayed enzyme at 37 °C for only 20 min. The whole-cell catalyst was very stable, and the enzyme activity of the whole cell was maintained above 40% after 40 rounds of recovery. Under this condition, 49.01 mg/mL (96.66 mM) ATP was accumulated for multi-rounds reaction. This ATP regeneration system has the characteristics of low cost, long lifetime, flexible compatibility, and great robustness.


Introduction
Adenosine-5'-triphosphate (ATP), is one of the essential molecules in living systems [1], plays a central role in many biochemical reactions, and has the potential to synthesize many high-value products [2,3]. The synthesis and consumption of ATP play an important role in many aspects of cell metabolism, such as active transport mechanisms [1,2], ATP-binding cassette (ABC) transporters [4], and ATP as the precursor to synthesize DNA, RNA, and NAD(P) [5]. ATP is necessary for the biosynthesis of cyclic adenosine monophosphate (cAMP), that is a significant second messenger in signal transduction; ATP also can serve as a signal ligand for ATP-sensitive or purinergic ionotropic and G-protein coupled receptors [6]. ATP provides energy for the biosynthesis of a large number of biological compounds, such as amino acids, proteins, and lipids [6]. Large amounts of ATP are also needed in another important field that is cell-free protein expression, as they are involved in complex reaction cascades [1,7,8], such as the production of S-adenosyl-homocysteine (SAH), glutathione (GSH), and S-adenosyl-methionine (SAM) [9].
Traditional ATP synthesis is chemically synthesized from the substrate AMP or ADP with compounds with phosphoric acid groups as the phosphorylation reagents [10]. In some cases, ATP was regenerated by the direct transfer of a phosphoryl group of another phosphorylated compound to ADP and AMP for energy metabolite [2]; this is the most significant way of ATP regeneration for anaerobic microorganisms and cells during anoxia [2]. It provides a quicker, less energy-efficient source of ATP, and it does not require the FoF1 ATP synthase driven by proton motive forces (PMF) across the membranes compared with oxidative phosphorylation and photophosphorylation to regenerate ATP [5,11]. At the Figure 1. Schematic presentation of polyphosphate-based ATP regeneration from AMP or ADP catalyzed by phosphotransferase.

Construction of Recombinant Plasmids for Cell Surface Display
The PAP-Pb (WP_013692547.1.) was synthesized and cloned into the vector pET-23a(+) and pET-23a(+)-sfGFP as shown in Figure 2. The recombinants were identified by sequencing.

Construction of Recombinant Plasmids for Cell Surface Display
The PAP-Pb (WP_013692547.1.) was synthesized and cloned into the vector pET-23a(+) and pET-23a(+)-sfGFP as shown in Figure 2. The recombinants were identified by sequencing.

Secretory Expression of the Target Protein
The plasmids and strains used in this research are shown in Table 2. The recombinants pET23a-PAP-Pb and pET23a-sfGFP-PAP-Pb were transformed into E. coli BL21(DE3) competent cells. One of the positive clones was inoculated into 5 mL LB medium with 50 µg/mL Ampicillin and cultured at 220 rpm and 37 • C. Then the overnight cultures were transferred to 100 mL LB medium with 50 µg/mL ampicillin and cultured at 220 rpm and 37 • C. When the optical density at 600 nm (OD 600 ) reached to 0.7-1.6, cells were harvested by centrifugation (15,000× g, 5 min, 4 • C ), the pellet was completely resuspended in 10 mL of phosphate-buffered saline (PBS, pH 7.4) and then treated with ultrasonication. After centrifugation at 15,000× g for 10 min, the supernatant and pellet were analyzed with SDS-PAGE.

Secretory Expression of the Target Protein
The plasmids and strains used in this research are shown in Table 2. The recombinants pET23a-PAP-Pb and pET23a-sfGFP-PAP-Pb were transformed into E. coli BL21(DE3) competent cells. One of the positive clones was inoculated into 5 mL LB medium with 50 μg/mL Ampicillin and cultured at 220 rpm and 37 ℃. Then the overnight cultures were transferred to 100 mL LB medium with 50 μg/mL ampicillin and cultured at 220 rpm and 37 ℃. When the optical density at 600 nm (OD600) reached to 0.7-1.6, cells were harvested by centrifugation (15,000× g, 5 min, 4 ℃ ), the pellet was completely resuspended in 10 mL of phosphate-buffered saline (PBS, pH 7.4) and then treated with ultrasonication. After centrifugation at 15,000× g for 10 min, the supernatant and pellet were analyzed with SDS-PAGE.

Cell Fractionation
Cell fractionation was performed according to the method described by Quan [32,33]. Cells were harvested from 10 mL culture broth (OD600 ≈ 1.6) by centrifugation (5000 g, 10 min, 4 ℃) and treated with one-third volume of Tris-EDTA-NaCl solution (TEN) containing 50 mM Tris-HCl, 5 mM EDTA, and 50 mM NaCl (pH 8.0). The mixture was incubated at 4 ℃ overnight processing [11]Error! Reference source not found. , then the solution was centrifuged at 5000 g at 4 ℃ for 20 min. The supernatant was regarded as outer membrane fraction, and then supernatant was collected, and protein samples were detected by 10% SDS-PAGE.

Cell Fractionation
Cell fractionation was performed according to the method described by Quan [32,33]. Cells were harvested from 10 mL culture broth (OD600 ≈ 1.6) by centrifugation (5000 g, 10 min, 4 • C) and treated with one-third volume of Tris-EDTA-NaCl solution (TEN) containing 50 mM Tris-HCl, 5 mM EDTA, and 50 mM NaCl (pH 8.0). The mixture was incubated at 4 • C overnight processing [11], then the solution was centrifuged at 5000 g at 4 • C for 20 min. The supernatant was regarded as outer membrane fraction, and then supernatant was collected, and protein samples were detected by 10% SDS-PAGE.

Biosynthesis of ATP
The 1 mL reaction mixture used for the synthesis of ATP containing 20 mM poly(P), 4 mM ADP or AMP, 30 mM Mg 2+ , 50 mM Tris-HCl (pH 8.5), and appropriate amounts of PAP-Pb or sfGFP-PAP-Pb (supernatant or intact cells). The reaction was performed in a shaker at 37 • C and terminated by heating the mixture at 80 • C for 10 min, followed by detecting the products with high-performance liquid chromatography (HPLC, Welch) after diluting five-fold with sterile dH 2 O.

Detection and Analysis by High-Performance Liquid Chromatography (HPLC)
ATP, ADP, and AMP were monitored with HPLC (Welch), equipped with a Welch Ultimate LP-C18 column (4.6 × 250 mm, 5 µm, Welch Materials, Inc, Shanghai, China) after 0.22 µm filtration. The mobile phase was methanol with 50 mM ammonium formate buffer (pH 4.5) (95% v/v) at a flow rate of 0.5 mL/min [34], and the effluent was detected at a wavelength of 260 nm. The detection times were as follows: ATP (7.453 s), ADP (8.489 s), and AMP (14.009 s).

Effect of pH and Mg 2+ Concentration ATP Synthesis
A single variable was maintained to determine the effect of the optimum pH and Mg 2+ concentration. The reaction mixture was prepared as previously described. The optimal pH for the reaction was determined by experiments using a pH range of 7.0-9.0 at 37 • C with the buffer of 20 mM poly(P), 4 mM ADP or AMP, 30 mM Mg 2+ , 50 mM Tris-HCl; the optimal Mg 2+ concentration was determined using the concentration from 0-100 mM at 37 • C with the buffer of 20 mM poly(P), 4 mM ADP or AMP, 50 mM Tris-HCl (pH 8.5). All experiments were performed for at least three biological replicates.

Effect of Polyphosphate Concentration on ATP Synthesis
A single variable was maintained to determine the effect of the optimum polyphosphate concentration. The reaction mixture was prepared as previously described. The optimal polyphosphate concentration was determined using the concentration from 0-150 mM at 37 • C with the buffer of 4 mM ADP or AMP, 30 mM Mg 2+ , 50 mM Tris-HCl (pH 8.5). The experiment was performed for at least three biological replicates.

Effect of the Cell State on the Production of ATP
The recombinant pET23a-sfGFP-PAP-Pb was transformed into E. coli BL21(DE3) competent cells. One of the positive clones was inoculated into 5 mL LB medium with 50 µg/mL ampicillin and cultured at 220 rpm and 37 • C. Then the overnight cultures were transferred to 100 mL LB medium with 50 µg/mL ampicillin and cultured at 220 rpm and 37 • C until the OD 600 reached to 0.7. The OD 600 was measured every hour until OD 600 reached to 1.7, then the OD 600 was measured every 5 h. The same amount of cells with different OD 600 were used in the reaction system using AMP or ADP as substrates to detect the enzyme activity, respectively.

Effects of Temperature and Stability of Whole-Cell Catalysts
The optimal temperature was determined using the temperature range of 18-50 • C at optimal pH and Mg 2+ concentration. The reaction mixture was prepared as previously described, containing 20 mM poly(P), 4 mM ADP or AMP, 30 mM Mg 2+ , 50 mM Tris-HCl (pH 8.5) in 1 mL reaction mixture.
For detecting the stability of whole-cell catalysts, the same amount of cells were incubated at 37 • C, 40 • C, and 45 • C for different lengths of time. Then they were used in the reaction system using AMP or ADP as substrates to detect the residual activity of the enzyme.

Reuse of Whole-Cell Catalyst
The whole-cell catalyst was tested at optimal pH, temperature, Mg 2+ , and poly(P) concentration. The 1 mL reaction mixture contains 20 mM poly(P), 4 mM ADP or AMP, 30 mM Mg 2+ , 50 mM Tris-HCl (pH 8.5), and appropriate amounts of sfGFP-PAP-Pb intact cells. After reaction at 37 • C for 20 min, the reaction mixture was centrifuged at 15,000 g for 5 min, and the pellet was used for the next reaction.

Expression of the Recombinant Proteins
Recombinant strains E-PAP-Pb and E-sfGFP-PAP-Pb were cultured as described in the methods section. Proteins were extracted and analyzed by SDS-PAGE. As shown in Figure 3a, the PAP-Pb (61.0 kDa)was mainly detected in the precipitate of the cell lysate whether it was induced, while the fusion protein sfGFP-PAP (87.0 kDa)was mainly observed in the supernatant of the cell lysate. As indicated, the sfGFP-PAP-Pb was more soluble compared to PAP-Pb. The results indicate sfGFP enables proper folding of PAP-Pb proteins in E. coli.
membrane. This is consistent with previous reports that sfGFP can promote protein tion outside the cell membrane [27,28].
Laser scanning confocal microscopy (LSCM) was used to detect the distribut sfGFP-PAP-Pb in E. coli BL21(DE3) without any dyeing process. Under the eyepiec objective lens magnification was adjusted to find the cells that needed to be observed LSCM was switched to scan mode, and the laser intensity parameters were adjus obtain a clear confocal image. The images obtained by scanning in the bright-fiel dark-field fields of view are shown in Figure 3c. It was observed that sfGFP-PAP-P terial cells had bright autofluorescence, and sfGFP could also be found in the backgr  To further verify whether PAP-Pb and sfGFP-PAP-Pb were successfully displayed on the cell surface, the cell fraction of outer membrane was separated and analyzed via SDS-PAGE. As shown in Figure 3b, sfGFP-PAP-Pb mainly appeared in the fraction of the outer cell membrane. While few of PAP-Pb was detected in the fraction of the outer cell membrane. This is consistent with previous reports that sfGFP can promote protein secretion outside the cell membrane [27,28].
Laser scanning confocal microscopy (LSCM) was used to detect the distribution of sfGFP-PAP-Pb in E. coli BL21(DE3) without any dyeing process. Under the eyepiece, the objective lens magnification was adjusted to find the cells that needed to be observed. The LSCM was switched to scan mode, and the laser intensity parameters were adjusted to obtain a clear confocal image. The images obtained by scanning in the bright-field and darkfield fields of view are shown in Figure 3c. It was observed that sfGFP-PAP-Pb bacterial cells had bright autofluorescence, and sfGFP could also be found in the background.

Biosynthesis of ATP with Fusion Enzyme Displayed on the Surface of E. coli
We cultured E. coli BL21(DE3) strains harboring psfGFP-PAP-Pb and pPAP-Pb plasmid to generate ATP, respectively. The strain containing psfGFP was used as a control. To ensure the same number of cells were applied in the reaction, the cells expressed with sfGFP-PAP-Pb, PAP-Pb or sfGFP were collected and diluted to the same OD 600 . To evaluate the effect of the cell-surface display of PAP-Pb, we compared the conversion rate of ATP with displayed sfGFP-PAP-Pb and cytosol PAP-Pb. As shown, the fusion sfGFP-PAP-Pb possessed better catalytic activity than the PAP-Pb (Figure 4). For the supernatant of sfGFP-PAP-Pb and PAP-Pb, the conversion rate of ATP catalyzed by sfGFP-PAP-Pb was higher than that by PAP-Pb whether the substrate was AMP (p < 0.05) or ADP (p < 0.01), that is consistent with the protein expression level shown in Figure 3a. However, for the intact cells, the conversion rate of ATP catalyzed by sfGFP-PAP-Pb was significantly higher than that of PAP-Pb whether the substrate was AMP or ADP (p < 0.01). The results demonstrate the advantage of the cell-surface display to improve productivity [35]. The increased productivity may be due to a shortened ATP reaction step. In the cytosol-expressed PAP-Pb strain, AMP or ATP needs to be transported into the cell to react with PAP-Pb and be converted to ATP. Then, ATP is secreted back to the extracellular medium. Whereas with surface-displayed PAP-Pb strain, AMP or ADP encounters with sfGFP-PAP-Pb directly on the cell surface and can be converted to ATP in the extracellular medium. Therefore, the conversion of AMP or ADP into ATP by displayed sfGFP-PAP-Pb is more efficient, resulting in a higher ATP conversion rate.

Effects of pH and the Concentration of Mg 2+
A single variable was used to determine the effects of pH and the concentration of Mg 2+ for ATP production. As shown in Figure 5, the optimal pH was 8.5 for the synthesis reaction of ATP respectively whether using ADP or AMP as the substrate (Figure 5a). The optimal concentration of Mg 2+ was 30 mM (Figure 5b) using ADP or AMP as the substrate, that was different to the data reported [1,12,13,17,34]. The results showed a relatively broad pH and concentration ranges of Mg 2+ for the reaction. The conversion rate of ATP seems to be comparable when the pH ranged from 7.0 to 9.0 and the Mg 2+ concentration ranged from 30 mM to 100 mM.
Because only one enzyme is associated with the whole reaction, the temperature, pH, and the Mg 2+ concentration of the reaction are well controlled; moreover, a relatively broad temperature, pH, and the Mg 2+ concentration ranges of the reaction. This ATP regeneration system is robust and convenient, so it is suitable for industrial large-scale application.

Effects of the Concentration of Polyphosphate
In this ATP regeneration system, the adenosine scaffold could be recycled in the reaction, without a large amount of AMP or ADP [36]. While, polyphosphate as a phosphate donor, is a consumable for ATP regeneration [36]. Therefore, a large amount of polyphosphate (poly(P)) is required to improve the conversion rate of ATP regeneration. We carried out the reaction with various concentrations of poly(P) at optimal temperature, pH, and concentration of Mg 2+ . As shown in Figure 6, the optimal concentration of poly(P) was 10-20 mM for ATP regeneration. However, when the concentration of poly(P) reached to 40 mM or even more, the conversion rate of ATP regeneration dropped rapidly. The main reason may be that the high concentration of poly(P) was the inhibitor of the ATP regeneration reaction [36].  The total volume was brought up to 1000 μL with double-distilled water. Different enzymes or cells were added as shown, sfGFP and BL21(DE3) were as negative controls. The reaction solution was incubated at 37 ℃ for 20 min. A small amount of reaction solution was terminated in 80 °C water bath for 10 min and diluted five-fold for HPLC analysis. Results are means ± SD of three parallel replicates.

Effects of pH and the Concentration of Mg 2+
A single variable was used to determine the effects of pH and the concentration of Mg 2+ for ATP production. As shown in Figure 5, the optimal pH was 8.5 for the synthesis reaction of ATP respectively whether using ADP or AMP as the substrate (Figure 5 a). The optimal concentration of Mg 2+ was 30 mM (Figure 5 b) using ADP or AMP as the substrate, that was different to the data reported [1,12,13,17,34]. The results showed a relatively broad pH and concentration ranges of Mg 2+ for the reaction. The conversion rate of ATP seems to be comparable when the pH ranged from 7.0 to 9.0 and the Mg 2+ concentration ranged from 30 mM to 100 mM. Because only one enzyme is associated with the whole reaction, the temperature, pH, and the Mg 2+ concentration of the reaction are well controlled; moreover, a relatively broad temperature, pH, and the Mg 2+ concentration ranges of the reaction. This ATP regeneration system is robust and convenient, so it is suitable for industrial large-scale application.

Effects of the Concentration of Polyphosphate
In this ATP regeneration system, the adenosine scaffold could be recycled in the reaction, without a large amount of AMP or ADP [36]. While, polyphosphate as a phosphate donor, is a consumable for ATP regeneration [36]. Therefore, a large amount of polyphosphate (poly(P)) is required to improve the conversion rate of ATP regeneration. We carried out the reaction with various concentrations of poly(P) at optimal temperature, pH, and concentration of Mg 2+ . As shown in Figure 6, the optimal concentration of poly(P) was 10-20 mM for ATP regeneration. However, when the concentration of poly(P) reached to 40 mM or even more, the conversion rate of ATP regeneration dropped rapidly. The main reason may be that the high concentration of poly(P) was the inhibitor of the ATP regeneration reaction [36].

Effects of the Cell State on the Production of ATP
In this reaction system, ATP was catalyzed by the intact cells of sfGFP-PAP-Pb, as described above, so the cell state should affect the synthesis reaction. In order to improve the cell performance, we performed the growth curve of cells after OD600 to 0.6 ( Figure 7a). The same amount of cells with different OD600 were used in the reaction system using AMP or ADP as substrates, respectively. As shown in Figure 7b, the conversion rate of ATP maintained a higher level when the OD600 reached to 0.7-1.6, and the conversion rate of ATP reached to 97.5% and 80.1% respectively when AMP or ADP was used as sub-

Effects of the Cell State on the Production of ATP
In this reaction system, ATP was catalyzed by the intact cells of sfGFP-PAP-Pb, as described above, so the cell state should affect the synthesis reaction. In order to improve the cell performance, we performed the growth curve of cells after OD 600 to 0.6 ( Figure 7a). The same amount of cells with different OD 600 were used in the reaction system using AMP or ADP as substrates, respectively. As shown in Figure 7b, the conversion rate of ATP maintained a higher level when the OD 600 reached to 0.7-1.6, and the conversion rate of ATP reached to 97.5% and 80.1% respectively when AMP or ADP was used as substrates. For OD 600 over 1.6, the conversion rate decreased rapidly. Figure 6. Effect of polyphosphate concentration on ATP synthesis. Productivity was assayed in 4 mM ADP or AMP, 30 mM Mg 2+ , and 50 mM Tris−HCl buffer (pH 8.5) at 37 °C with the concentration of polyphosphate from 0-150 mM. Results from three independent experiments were quantified. Error bars represent SD of three independent experiments.

Effects of the Cell State on the Production of ATP
In this reaction system, ATP was catalyzed by the intact cells of sfGFP-PAP-Pb, as described above, so the cell state should affect the synthesis reaction. In order to improve the cell performance, we performed the growth curve of cells after OD600 to 0.6 ( Figure 7a). The same amount of cells with different OD600 were used in the reaction system using AMP or ADP as substrates, respectively. As shown in Figure 7b, the conversion rate of ATP maintained a higher level when the OD600 reached to 0.7-1.6, and the conversion rate of ATP reached to 97.5% and 80.1% respectively when AMP or ADP was used as substrates. For OD600 over 1.6, the conversion rate decreased rapidly.  The reaction solution was incubated at 37 • C for 20 min. A small amount of reaction solution was terminated in 80 • C water bath for 10 min and diluted five-fold for HPLC analysis. Results are means ± SD of three parallel replicates.

Effects of Temperature and Stability of Whole-Cell Catalysts
The temperature result of the sfGFP-PAP-Pb for ATP production is shown in Figure 8a. Though the optimal temperature was 42 • C for the synthesis reaction of ATP when using ADP or AMP as substrates. The results showed a relatively broad temperature range for the reaction. The difference of the conversion rate of ATP is comparable when the temperature ranged from 37 • C to 50 • C.
The instability of anchoring motif on the cell surface is a critical problem of surface display in E. coli. Thus, we tested the stability of cell-surface-displayed sfGFP-PAP-Pb. Cells were incubated at 37 • C, 42 • C, and 50 • C for different times using ADP or AMP as the substrates respectively, and the residual activity of sfGFP-PAP-Pb was detected. As shown in Figure 8b,c, the sfGFP-PAP-Pb remained significantly stable at 37 • C. The residual activity of 78% and 60% was detected after incubation for 12 h at 37 • C when using AMP and ADP respectively. However, when at 42 • C and 50 • C, the enzyme quickly lost most of their activity. Although a little higher conversion rate of ATP at 42 • C was observed than that at 37 • C, the stability of sfGFP-PAP-Pb at 37 • C was much higher than that of 42 • C. Considering the reuse of intact cells, we chose the reaction condition at 37 • C in subsequent experiments.

Effects of Temperature and Stability of Whole-Cell Catalysts
The temperature result of the sfGFP-PAP-Pb for ATP production is shown in Figure  8a. Though the optimal temperature was 42 °C for the synthesis reaction of ATP when using ADP or AMP as substrates. The results showed a relatively broad temperature range for the reaction. The difference of the conversion rate of ATP is comparable when the temperature ranged from 37 °C to 50 °C. The instability of anchoring motif on the cell surface is a critical problem of surface display in E. coli. Thus, we tested the stability of cell-surface-displayed sfGFP-PAP-Pb. Cells were incubated at 37 °C, 42 °C, and 50 °C for different times using ADP or AMP as the substrates respectively, and the residual activity of sfGFP-PAP-Pb was detected. As shown in Figure 8 (b, c), the sfGFP-PAP-Pb remained significantly stable at 37 °C. The residual activity of 78% and 60% was detected after incubation for 12 h at 37 °C when using AMP and ADP respectively. However, when at 42 °C and 50 °C, the enzyme quickly lost most of their activity. Although a little higher conversion rate of ATP at 42 °C was observed than that at 37 °C, the stability of sfGFP-PAP-Pb at 37 °C was much higher than Next, we examined the reaction time for the generation of ATP. The result is shown in Figure 8d, the conversion rate of ATP rose rapidly to 60% after reaction for 5 min, and reached to the plateau period when the reaction time reached 20 min.

Reuse of Whole-Cell Catalyst
Unlike free enzyme, intact cells as biocatalysts in the reaction mixture can be recycled as they are easily separated from the solution [17]. The fusion enzyme was displayed on the cell surface and contacted the substrate directly, and the cells could release the fresh enzyme continually. Using whole cells of E. coli that displayed sfGFP-PAP-Pb on the surface, we tested the possibility of reuse, as shown in Figure 9a. After a 20-min reaction, the intact cells were recovered and used as biocatalysts for the next synthesis of ATP. The conversion rate of ATP was 79.42% in the first reaction when using AMP as substrate, and after the cells were reused up to 40 times, the yield of ATP remained above 43%. When using ADP as substrate, the conversion rate of ATP was 80.51% in the first reaction, and after the cells were reused for 40 times, the residual activity was 46% of the first time.
face, we tested the possibility of reuse, as shown in Figure 9a. After a 20-min reaction, the intact cells were recovered and used as biocatalysts for the next synthesis of ATP. The conversion rate of ATP was 79.42% in the first reaction when using AMP as substrate, and after the cells were reused up to 40 times, the yield of ATP remained above 43%. When using ADP as substrate, the conversion rate of ATP was 80.51% in the first reaction, and after the cells were reused for 40 times, the residual activity was 46% of the first time. The fusion protein with sfGFP tag has its own green fluorescence [27], which can be directly visible without any equipment, that is a very important advantage compared to other tags. Moreover, the fusion enzyme sfGFP-PAP-Pb can be displayed on the surface of the E. coli cells, so the intact cells of sfGFP-PAP-Pb are also visible with green fluorescence. As shown in Figure 9b, the green fluorescence of the intact cells became weaker after every reaction. When the fluorescence was very weak or invisible, it indicated that the intact cells were no longer available. So, we can judge from the intensity of fluorescence whether the recycle reaction should be ended without the detection of enzyme activity. That is a very convenient and economical strategy, especially for large-scale industrial applications.

Conclusions
An efficient, visible, simple, and economical method to prepare ATP from AMP or ADP with whole-cell PAP-Pb catalysts was developed by using the sfGFP to display the enzyme to the cell surface. This method is attractive because the process can be carried The fusion protein with sfGFP tag has its own green fluorescence [27], which can be directly visible without any equipment, that is a very important advantage compared to other tags. Moreover, the fusion enzyme sfGFP-PAP-Pb can be displayed on the surface of the E. coli cells, so the intact cells of sfGFP-PAP-Pb are also visible with green fluorescence. As shown in Figure 9b, the green fluorescence of the intact cells became weaker after every reaction. When the fluorescence was very weak or invisible, it indicated that the intact cells were no longer available. So, we can judge from the intensity of fluorescence whether the recycle reaction should be ended without the detection of enzyme activity. That is a very convenient and economical strategy, especially for large-scale industrial applications.

Conclusions
An efficient, visible, simple, and economical method to prepare ATP from AMP or ADP with whole-cell PAP-Pb catalysts was developed by using the sfGFP to display the enzyme to the cell surface. This method is attractive because the process can be carried out without the subsequent enzyme extraction and purification steps. Moreover, only one enzyme in the reaction system and the sfGFP-PAP-Pb is visible, the visible fusion sfGFP-PAP-Pb fixed on the surface of the bacteria is very stable under the reaction condition. So, the reaction condition is easy to control, especially for large-scale industrial applications. The catalytic activity of sfGFP-PAP-Pb intact cells was 2.12 and 1.47 times higher than that of PAP-Pb intact cells, when the substrate was AMP and ADP, respectively and the conversion rate of ATP from AMP and ADP reached to 97.5% and 80.1% respectively when catalyzed by the surface-displayed enzyme at 37 • C only for 20 min. Further, the whole cell could be reused for more than 40 rounds keeping the conversion rate of ATP above 40%. Under this condition, 49.01 mg/mL (96.66 mM) ATP was accumulated for multi-rounds reaction. This resting cell reaction system could be used for ATP regeneration or coupled with other ATP-dependent enzymes for the synthesis of high-value products. Data Availability Statement: All relevant data of this study are presented. Additional data will be provided upon request.