Efficient Expression of Soluble Recombinant Protein Fused with Core-Streptavidin in Bacterial Strain with T7 Expression System

The limited amount of fusion protein transported into cytosol milieu has made it challenging to obtain a sufficient amount for further applications. To avoid the laborious and expensive task, T7 promoter-driving pET-30a(+) coding for chimeric gene of thymidine phosphorylase and core streptavidin as a model system was constructed and transformed into a variety of E. coli strains with T7 expression system. Our results demonstrated that the pET-30a(+)-TP-coreSA/Lemo21(DE3) system is able to provide efficient expression of soluble TP-coreSA fusion protein for purification. Moreover, the eluted TP-coreSA fusion protein tethered on biotinylated A549 carcinoma cells could effectively eliminate these malignant cells after administrating prodrug 5′-DFUR.


Introduction
The bacterium Escherichia coli is the most convenient and widely used system for expression of recombinant proteins [1]. The E. coli strains with the presence of T7 promoter-based plasmids have been extensively employed to express proteins. In this system, the T7 RNA polymerase gene is located in λ DE3 prophage of the chromosome under the control of the IPTG-inducible LacUV5 promoter [2,3]. To facilitate the immobilization of protein on biotinylated materials, expression of chimeric proteins fused with the streptavidin affinity domain has been reported using bacteria [4][5][6]. However, full length streptavidin (SA) tends to form higher order aggregates and thus has poor solubility. In contrast, core streptavidin (coreSA) is more soluble than the full-length SA. Moreover, coreSA has enhanced biotin-binding features compared to full-length SA because the removal of non-functional terminal regions promotes the binding with less steric hindrance [7]. Some chimeric proteins such as scFv and CD47 fused with coreSA affinity tag have been reported to be produced in E. coli [8][9][10][11]. However, a large amount of coreSA-tagged fusion proteins after expression in E. coli was present in an insoluble form found in inclusion bodies. These insoluble protein aggregates in E. coli cells could only be dissolved under strong denaturing conditions that are time-consuming and costly [12,13]. Moreover, the limited amount of fusion proteins translated into cytosol made it difficult to purify sufficient quantities necessary for further applications. Attempts have been made to express and purify coreSA-associated fusion proteins from bacterial lysate [10,11]; however, the yields were low.
In this study, we constructed a T7 promoter-based pET-30a(+) vector encoding with thymidine phosphorylase (TP)-coreSA chimeric gene as a model system for the expression of TP-coreSA fusion 1.

Equipment
Purify the cloned PCR product of TP and coreSA with Monarch PCR & DNA cleanup kit and analyze with 1% agarose gel electrophoresis.

Expression of TP-coreSA Encoding Plasmid
1. Transform the constructed pET-30a(+)-TP-coreSA vector into different strains of E. coli competent cells (i.e., BL21(DE3), NiCo21(DE3), Lemo21(DE3), and SHuffle ® ) according to manufacturer standard protocols. 2. Spread on separate agar plates supplemented with 50 µg/mL kanamycin. 3. Incubate overnight at 37 °C to screen the positive clones. 4. Pick a single colony from each agar plate and resuspend it into 5 mL LB media supplemented with 50 µg/mL kanamycin. 5. Grow overnight with shaking at 220 rpm and 37 °C to produce starter culture. 6. On the next day, dilute the starter cultures to 100 mL for SHuffle ® , BL21(DE3), and NiCo21(DE3) separately with fresh media and keep shaking at 37 °C. 7. Induce 400 µM IPTG when optical density at 600 nm (OD600) of the cultures reaches 0.5, and shake for 5 h at 30 °C to induce expression. 8. Harvest the cells by centrifugation at 4500× g for 15 min at 4 °C. 9. Re-suspend each cell pellets in 2 mL of lysis buffer that was made by mixing B-PER with 50 mM Tris-HCl (pH 7.5) and EDTA-free protease inhibitor (1 tablet for 50 mL). 10. Wait for 10 min at room temperature and sonicate 10 times (pulse of 10 s with 10 s rest each time) with output at seven to maximize the protein extraction efficiency. 11. Finally, centrifuge lysates at 17,000× g for 20 min and collect the supernatant named as a soluble crude fraction and the re-suspend the pellets in 1x PBS named as an insoluble fraction. 12. To determine the optimal expression level of recombinant TP-coreSA protein with Lemo21(DE3), use 25 mL of culture media with different concentrations of L-rhamnose (0, 100, 250, 500, 750,

3.
Incubate overnight at 37 • C to screen the positive clones.

4.
Pick a single colony from each agar plate and resuspend it into 5 mL LB media supplemented with 50 µg/mL kanamycin.

5.
Grow overnight with shaking at 220 rpm and 37 • C to produce starter culture. 6.
On the next day, dilute the starter cultures to 100 mL for SHuffle ® , BL21(DE3), and NiCo21(DE3) separately with fresh media and keep shaking at 37 • C. 7.
Induce 400 µM IPTG when optical density at 600 nm (OD 600 ) of the cultures reaches 0.5, and shake for 5 h at 30 • C to induce expression. 8.
Harvest the cells by centrifugation at 4500× g for 15 min at 4 • C. 9.
Re-suspend each cell pellets in 2 mL of lysis buffer that was made by mixing B-PER with 50 mM Tris-HCl (pH 7.5) and EDTA-free protease inhibitor (1 tablet for 50 mL). 10. Wait for 10 min at room temperature and sonicate 10 times (pulse of 10 s with 10 s rest each time) with output at seven to maximize the protein extraction efficiency.  (For instance, 2 mL of resin was used in our experiment, it was washed with 10 mL with wash buffer and eluted with 6 mL of elution buffer). 5. Measure OD280 of each collected fraction with SpectraMax M2e microplate reader to plot the elution profile. CRITICAL STEP: L-rhamnose concentration is very important and it can differ for every fusion protein of interest. So, researchers need to identify the best concentration through titer usually in the range of 0 to 2000 µM. Once the optimal concentration of L-rhamnose is identified, 100 mL of culture medium can be set up for the protein expression. (For instance, 2 mL of resin was used in our experiment, it was washed with 10 mL with wash buffer and eluted with 6 mL of elution buffer). 5. Measure OD280 of each collected fraction with SpectraMax M2e microplate reader to plot the elution profile.
CRITICAL STEP: Always place lysates, crude proteins, insoluble fraction, and purified protein on ice during the experiments.

OPTIONAL STEP: Concentrate the eluted protein with Protein Concentrators PES (MWCO = 50 kD) if needed. Based on the size of purified protein of interest, different MWCO might be used.
6. For the control group study using coreSA protein only, Lemo21(DE3) is transformed with the constructed pET-30a(+)-coreSA plasmid, expressed, and purified by the aforementioned methods. elution protein (obtained from purification).

CRITICAL STEP:
The IPTG concentration should remain the same, only change L-rhamnose concentration for optimal expression of fusion protein of interest.

2.
Mix the soluble crude fraction of Lemo21(DE3) in 1:1 ratio with binding buffer and let it bind to HisPur™ cobalt resin for 1 h with gentle shaking at 4 • C.

3.
Load into the column and collect the flow-through.

4.
Wash with 5 resin bed volume of wash buffer and elute with 3 resin bed volume of elution buffer.
(For instance, 2 mL of resin was used in our experiment, it was washed with 10 mL with wash buffer and eluted with 6 mL of elution buffer).

5.
Measure OD 280 of each collected fraction with SpectraMax M2e microplate reader to plot the elution profile. (For instance, 2 mL of resin was used in our experiment, it was washed with 10 mL with wash buffer and eluted with 6 mL of elution buffer). 5. Measure OD280 of each collected fraction with SpectraMax M2e microplate reader to plot the elution profile.
CRITICAL STEP: Always place lysates, crude proteins, insoluble fraction, and purified protein on ice during the experiments. CRITICAL STEP: Always place lysates, crude proteins, insoluble fraction, and purified protein on ice during the experiments.
OPTIONAL STEP: Concentrate the eluted protein with Protein Concentrators PES (MWCO = 50 kD) if needed. Based on the size of purified protein of interest, different MWCO might be used.

6.
For the control group study using coreSA protein only, Lemo21(DE3) is transformed with the constructed pET-30a(+)-coreSA plasmid, expressed, and purified by the aforementioned methods.  (For instance, 2 mL of resin was used in our experiment, it was washed with 10 mL with wash buffer and eluted with 6 mL of elution buffer). 5. Measure OD280 of each collected fraction with SpectraMax M2e microplate reader to plot the elution profile.

4.
For the second Western blot, redo the 1st and 2nd step and use anti-TP (PD-ECGF) monoclonal antibody for the 3rd step.

5.
Wash the membranes again 3 times to remove any excess of primary antibodies and incubate for 1 h at room temperature in secondary antibody solution (1:100 dilution of mouse IgG HRP-conjugated antibody in blocking buffer).

6.
Finally, wash the membranes again and detect the fusion protein with horseradish peroxidase activity using ECL substrate by taking imaging through chemiluminescence imager.

2.
Measure the initial absorbance as a baseline at 305 nm with SpectraMax M2e microplate reader.

3.
Incubate the reaction mixture at 37 • C for 2 h and terminate the reaction by the addition of 500 mM sodium hydroxide.

5.
For enzyme kinetics, repeat all steps (from 1 to 4) with various concentrations of 5 -DFUR to react with TP-coreSA fusion protein for 2 h and obtain enzyme activities. 6.
Use the Michaelis-Menten equation to do a curve fitting of the data points. (For instance, 2 mL of resin was used in our experiment, it was washed with 10 mL with wash buffer and eluted with 6 mL of elution buffer). 5. Measure OD280 of each collected fraction with SpectraMax M2e microplate reader to plot the elution profile.

Expression of pET-30a(+)-TP-coreSA
The pET-30a(+)-TP-coreSA vector was constructed by inserting gene sequences of core streptavidin (coreSA) and thymidine phosphorylase (TP) in between EcoRI/XhoI and EcoRV/EcoRI respectively, and verified using agarose gel as shown in Figure 1. Furthermore, the cloned sequence of TP and coreSA in pET-30a(+)-TP-coreSA was confirmed by the Sanger sequencing performed by GenScript. The molecular weight of fusion protein is estimated to be 75 kDa as the TP being 1458 bp (~55 kDa), coreSA being 387 bp (~15 kDa), and others (S tag, thrombin site, 6xHis, enterokinase site) being 135 bp (~5 kDa). The pET-30a(+)-TP-coreSA vector was then transformed into different strains of competent cells-Lemo21(DE3), Shuffle ® , BL21(DE3), and NiCo21(DE3)-to check the expression of the fusion protein in soluble and insoluble fraction. As shown in Figure 2 of TP-coreSA fusion protein in an insoluble form (lane 6) was lower than the one from BL21(DE3) (see lane 6). Additionally, there was nominal no detection of fusion protein in soluble fraction (lane 5). Taken altogether, large amount of TP-coreSA fusion protein was presented in inclusion bodies (i.e., cell pellets) when using NiCo21(DE3), BL21(DE3), and SHuffle ® bacterial strains. However, the expression with Lemo21(DE3) led to more soluble fraction (lane 7) as compared to its insoluble counterpart (lane 8).  Generally, T7 expression of proteins targeted to the Sec translocase often leads to accumulation of inclusion bodies, hence making the protein of interest in an insoluble form [12]. According to the results shown in Figure 2, the issue of TP-coreSA fusion protein prone to insoluble expression can be alleviated by using Lemo21(DE3), which is a variant strain of BL21(DE3) that offers tunable T7 expression by controlling the level of T7 lysozyme which is a natural inhibitor of T7 RNA polymerase [15]. The level of T7 lysozyme is modulated by varying L-rhamnose concentration in expression culture. A variety of concentrations (0, 100, 250, 500, 750, 1000, 2000 µM) of L-rhamnose were titrated to optimize the expression level of soluble fractions. As illustrated in Figure 3, L-rhamnose concentration ≥ 500 µM was able to finely tuned high level of TP-coreSA fusion protein expressed in soluble fractions. For each lane in Figure 3, the amounts of soluble crude protein and its corresponding eluted TP-coreSA fusion protein were determined by BCA assay. The percentages of TP-coreSA fusion protein in soluble crude protein were calculated to be 10%, 10.9%, 10.7%, 16.8%, 15.3%, 15.9%, and 15.3% for 0, 100, 250, 500, 750, 1000, and 2000 µM of L-rhamnose treatment, respectively. It is noteworthy that the eluted fusion protein from 500 µM or higher of L-rhamnose soluble fraction was about 3-fold higher than the one without L-rhamnose, although the background was also expressed more for other proteins (Figure 3 lane 4-7).

Purification of TP-coreSA Fusion Protein
The crude protein was extracted using Lemo21(DE3) competent cells with 500 µM of L-rhamnose as explained in the previous section. The TP-coreSA fusion protein was purified from crude fraction by affinity chromatography using a disposable column packed with cobalt resin. Although the yield of the fusion protein is lower using cobalt resin instead of nickel resin, it will help to reduce the nonspecific binding and results in more purified protein. Moreover, it also eases the purification process as it does not need to optimize the imidazole concentration in binding and washing buffer, which is usually needed while using nickel resin. The elution profile of the purification system shows the peak OD 280 at elution 2 as given in Figure 4A.

Purification of TP-coreSA Fusion Protein
The crude protein was extracted using Lemo21(DE3) competent cells with 500 µM of L-rhamnose as explained in the previous section. The TP-coreSA fusion protein was purified from crude fraction by affinity chromatography using a disposable column packed with cobalt resin. Although the yield of the fusion protein is lower using cobalt resin instead of nickel resin, it will help to reduce the nonspecific binding and results in more purified protein. Moreover, it also eases the purification process as it does not need to optimize the imidazole concentration in binding and

SDS-PAGE Analysis, Western Blot, TP-coreSA Enzyme Activity and Kinetics
The purified TP-coreSA fusion protein was concentrated five to ten folds by a protein concentrator PES (MWCO = 50 kD). As shown in Figure 4B, SDS-PAGE displayed the purified single band of TP-coreSA fusion protein (75 kDa) along with soluble crude fraction. It was further characterized through Western blot; the purified fusion protein was detected against anti-TP and anti-SA monoclonal antibodies confirming the fusion protein at 75 kDa location ( Figure 4C). The concentration of fusion protein was calculated as 217 ± 6.5 µg/mL using the BCA assay. SpectraMax M2e microplate reader was used to calculate the enzyme activity of TP-coreSA fusion protein by incubating with 5 -DFUR as described in the Section 3.4.4. As shown in Figure 4D, the change in absorbance at 305 nm is used to calculate the 5-FU produced from the conversion of 5 -DFUR by TP-coreSA enzyme using 5-FU calibration curve at 305 nm. The obtained absorbance changes after 30, 60, and 120 min were used to calculate 5-FU produced as 10.1 ± 0.33, 21.6 ± 0.52, and 37 ± 0.98 µmole, respectively. Since 20 µL of purified fusion protein was used, the TP-coreSA enzyme activity was calculated to be 4.98 ± 0.13 µmole/µg/h. It should be noted that the enzyme activity obtained here is much higher than the reported values ranged from 0.269 to 3.383 µmole/mg/h [14,[16][17][18]. This is because the literature-reported values of TP enzyme activity were obtained by dividing total amount of crude protein; however, in this study, the enzyme activity of TP-coreSA was calculated by dividing the purified protein, hence, it is much higher than the values reported. For the study of enzyme kinetics, enzyme activities (µmole/µg/h) plotted with their corresponding concentrations of substrate (i.e., 5 -DFUR) were curve fitted well with the Michaelis-Menten equation (data not shown). washing buffer, which is usually needed while using nickel resin. The elution profile of the purification system shows the peak OD280 at elution 2 as given in Figure 4A.

SDS-PAGE Analysis, Western Blot, TP-coreSA Enzyme Activity and Kinetics
The purified TP-coreSA fusion protein was concentrated five to ten folds by a protein concentrator PES (MWCO = 50 kD). As shown in Figure 4B, SDS-PAGE displayed the purified single band of TP-coreSA fusion protein (75 kDa) along with soluble crude fraction. It was further characterized through Western blot; the purified fusion protein was detected against anti-TP and anti-SA monoclonal antibodies confirming the fusion protein at 75 kDa location ( Figure 4C). The concentration of fusion protein was calculated as 217 ± 6.5 µg/mL using the BCA assay. SpectraMax M2e microplate reader was used to calculate the enzyme activity of TP-coreSA fusion protein by incubating with 5′-DFUR as described in the Section 3.4.4. As shown in Figure 4D, the change in absorbance at 305 nm is used to calculate the 5-FU produced from the conversion of 5′-DFUR by TP-coreSA enzyme using 5-FU calibration curve at 305 nm. The obtained absorbance changes after 30, 60, and 120 min were used to calculate 5-FU produced as 10.1 ± 0.33, 21.6 ± 0.52, and 37 ± 0.98 µmole, respectively. Since 20 µl of purified fusion protein was used, the TP-coreSA enzyme activity was calculated to be 4.98 ± 0.13 µmole/µg/h. It should be noted that the enzyme activity obtained here is much higher than the reported values ranged from 0.269 to 3.383 µmole/mg/h [14,[16][17][18]. This is because the literature-reported values of TP enzyme activity were obtained by dividing total amount of crude protein; however, in this study, the enzyme activity of TP-coreSA was calculated by dividing

Surface Biotinylation of A549 Cells
The biotinylated A549 cells were incubated with SA-FITC for 1 h to allow streptavidin-biotin binding. After that, images were taken under a fluorescent microscope. Figure 5A shows phase contrast grayscale image of biotinylated A549 cells, and Figure 5B is the fluorescent image of the same location showing green illuminance of bound SA-FITC. The overlay image (normal grayscale plus fluorescent) ( Figure 5C) demonstrates that the approach used for cell biotinylation did not affect cell viability and its morphology.

Surface Biotinylation of A549 Cells
The biotinylated A549 cells were incubated with SA-FITC for 1 h to allow streptavidin-biotin binding. After that, images were taken under a fluorescent microscope. Figure 5A shows phase contrast grayscale image of biotinylated A549 cells, and Figure 5B is the fluorescent image of the same location showing green illuminance of bound SA-FITC. The overlay image (normal grayscale plus fluorescent) ( Figure 5C) demonstrates that the approach used for cell biotinylation did not affect cell viability and its morphology.

Cytotoxicity Studies
Several studies have shown the anti-cancer effectiveness of TP as it converts prodrug 5′-DFUR to cytotoxic chemotherapy drug 5-FU [19][20][21]. The functionality of purified TP-coreSA fusion protein was examined via its anti-cancer activity after treating malignant cells in the presence of prodrug 5′-DFUR. The A549 cancer cells were cultured up to 40% confluence, biotinylated, and tethered with

Cytotoxicity Studies
Several studies have shown the anti-cancer effectiveness of TP as it converts prodrug 5 -DFUR to cytotoxic chemotherapy drug 5-FU [19][20][21]. The functionality of purified TP-coreSA fusion protein was examined via its anti-cancer activity after treating malignant cells in the presence of prodrug 5 -DFUR. The A549 cancer cells were cultured up to 40% confluence, biotinylated, and tethered with TP-coreSA fusion protein. After administrated with 100 µM of 5 -DFUR, A549 cells were eradicated in four days (see Figure 6D). This indicated that 5 -DFUR (prodrug) was enzymatically catalyzed into 5-FU with the aid of TP-coreSA fusion protein anchored on the surface of A549 cells via biotin-SA binding. To confirm the anticancer effectiveness is attributed to the formation of toxic 5-FU from prodrug 5 -DFUR, three control experiments were conducted as following: (i) biotinylated A549 cells without any treatment, (ii) biotinylated A549 cells treated with 5 -DFUR, and (iii) biotinylated A549 cells tethered with coreSA protein and treated with 5 -DFUR. As shown in Figure 6A-C, cells in these three control groups grew continuously and reached 100% confluency in 4 days. This means the endogenous expression of TP (if there is any) in A549 cells was not enough to convert the 5 -DFUR to 5-FU to kill cells. Moreover, the positive control group ( Figure 6E) was separately carried out by direct administration of 100 µM of 5-FU and revealed similar cell morphological change and cell death pattern illustrated in Figure 6D. However, the elimination of cells in the positive control was slightly quick with most of the cells killed at day 2 and 3 ( Figure 6, E3-E4) as compared to 5 -DFUR that showed the eradication occurred at day 3 and 4 ( Figure 6, D3-D4). It is speculated that the conversion yield of 5 -DFUR into 5-FU was not 100%; hence, this inefficient conversion led to a delay in killing cancer cells [22].    Figure 7 showed the cell confluency on a daily basis. For all negative control groups, A549 cells with 38% confluency at the inoculation continued to grow and reached 100% confluency in four days. Conversely, the prodrug treated with TP-decorated cells showed little increase in confluency on day 1 and then started dying and ended up with < 6% of viable cells on day 4, while cells treated directly with 5-FU showed a decrease in cell viability on the very next day after cytotoxic drug administration and had about 3% cell confluency on day 4. Methods Protoc. 2020, 3, x FOR PEER REVIEW 13 of 15 Figure 7 showed the cell confluency on a daily basis. For all negative control groups, A549 cells with 38% confluency at the inoculation continued to grow and reached 100% confluency in four days. Conversely, the prodrug treated with TP-decorated cells showed little increase in confluency on day 1 and then started dying and ended up with < 6% of viable cells on day 4, while cells treated directly with 5-FU showed a decrease in cell viability on the very next day after cytotoxic drug administration and had about 3% cell confluency on day 4. Figure 7. Viable cell confluency of A549 cells after treatment with nothing, 5′-DFUR, 5′-DFUR + coreSA, 5′-DFUR + TP-coreSA, and 5-FU for 4 days of culture. Cell inoculation for all groups was about 38% confluency.

Conclusions
In this study, pET-30a(+) plasmid encoding with TP-coreSA fusion gene was constructed. The pET-30a(+)-TP-coreSA plasmid was transformed and expressed in a variety of E. coli strains with T7 expression system. Our results demonstrated that Lemo21(DE3) produced substantial amount of TP-coreSA fusion protein in soluble fraction which was then purified by affinity chromatography. The optimized pET-30a(+)-TP-coreSA/Lemo21(DE3) system is time-saving and cost-effective because it avoids denaturing and refolding processes involved with protein expressed in inclusion bodies. The purified TP-coreSA fusion protein was characterized by enzyme activity assay. The anticancer effectiveness of TP-coreSA fusion protein was revealed by tethering it on biotinylated A549 cancer cells and treated with prodrug 5′-DFUR.

Conclusions
In this study, pET-30a(+) plasmid encoding with TP-coreSA fusion gene was constructed. The pET-30a(+)-TP-coreSA plasmid was transformed and expressed in a variety of E. coli strains with T7 expression system. Our results demonstrated that Lemo21(DE3) produced substantial amount of TP-coreSA fusion protein in soluble fraction which was then purified by affinity chromatography. The optimized pET-30a(+)-TP-coreSA/Lemo21(DE3) system is time-saving and cost-effective because it avoids denaturing and refolding processes involved with protein expressed in inclusion bodies. The purified TP-coreSA fusion protein was characterized by enzyme activity assay. The anticancer effectiveness of TP-coreSA fusion protein was revealed by tethering it on biotinylated A549 cancer cells and treated with prodrug 5 -DFUR.