Part:BBa_K3730035
NOTOC__ Anti-hGPC3-scfv,a single chain variable antibody of human GPC3 protein. In our program, it is fused
Anti-hGPC3-scfv,a single chain variable fragment of human GPC3 protein. In our program, it is fused with fiber protein. This design allows combination to GPC3 receptor which specifcally exists on the surface of the hepatic cancer cell.
Part 2.1 Our fusion strategy
(1) Design of fusion protein
Our aim is to fuse the single-chain variable antibody(scFv) of GPC3 to the C-terminal of virus coat protein, fiber. In order to achieve this, we used (G)4S flexible joint to connect the heavy chain variable region and light chain variable region of anti-GPC3 antibody, resulting in scFv of anti-GPC3. Then scFv is fused to the C-terminal of fiber protein through (G)4S flexible joint so that scFv will be expressed together with the fiber protein of the virus and present on the surface of the virus (as is shown in Figure 2-1).
Figure. 2-1 Strategy of fiber and scFv fusion.
(2) Protein and DNA sequence of scFv
The following figure shows the amino acid sequence of the fusion protein of fiber knob region (C-terminal) and scFv. The blue part shows the fiber protein knob region, and the red part shows VH and VL regions of anti-GPC3 antibody.
Figure. 2-2 The amino acid sequence of fiber knob-scfv protein.
The following figure shows the DNA sequence of scFv. The red part is the VH and VL regions of anti-GPC3 antibody.
Figure. 2-3 The DNA sequence of scFv.
(3) 3D structural prediction of fusion protein
We know that the strategy of modifying the virus fiber protein may greatly affect the assembly of fiber trimer, further affect the efficiency of virus intracellular assembly and amplification, and affect the virus titer[2]. Therefore, we made a de novo prediction on the structure of the fusion protein by an online website[3]. Fig. 2-4 shows a three-dimensional structure of the knob region trimer of adenovirus type 5 fiber protein downloaded from the PDB structure database[4]. The results (Fig. 2-5) show that the fusion protein did not affect the structure of the fiber protein knob region.
Figure. 2-4 3D structure (trimer) of knob region of adenovirus type 5 fiber protein.
(a) Bottom view, (b) side view. The spherical model shows the C terminal.
Figure. 2-5 Protein structure ab initio prediction (I-TASSER).
(a) The prediction of 3D structure of fiber knob. Red shows the N terminal, green shows
the main body of the fiber knob, and yellow shows the C terminal. (b) The prediction of
3D structure of fusion protein. Red shows the N terminal of the fiber knob, green shows
the main body of the fiber knob, and yellow shows the C terminal of the fiber knob. Grey
shows the flexible joint, purple shows the VH of scFv, and orange shows the VL of scFv.
(4) Construction of prokaryotic expression plasmid of anti-GPC3 scFv protein
Since the virus packaging takes a long time and the virus experiment has higher requirements for the laboratory, we decided to express scFv protein in prokaryotic first, and then verify its activity. The prokaryotic expression vector we selected is pET28a (+), which has a T7 promoter (controlled by lac operon) and kanamycin resistance. Then we constructed scFv into the target vector by enzyme digestion (EcoR I and Hind III restriction endonuclease) and T4 ligation. As is shown in Fig. 2-6, we successfully inserted the DNA sequence (801bp) of scFv into the target vector.
Figure. 2-6 Construction of scFv prokaryotic expression plasmid.
Lane1: DNA ladder, lane2: pET28a(+) empty plasmid,lane3: pET28a(+) vector was digested
by enzyme EcoR I and Hind III, lane4: scfv-pet28a(+) expression plasmid, lane5:
scfv-pet28a(+) vector was digested by enzyme EcoR I and Hind III.
(5) Solution to rare codon problem
After the construction of scFv expression plasmid, we transformed it into shuffle T7 competent cells for protein expression, but it was found that scFv protein was not successfully expressed by SDS-PAGE. After checking our sequence, we found what the problem was. It turned out that although we optimized the codons of VH and VL regions of scFv, we ignored whether the flexible joints connecting them had rare codons. Unfortunately, it did. Therefore, we know that Rosetta competent cell contains a plasmid that can express rare codons, pRARE [5](as shown in Fig. 2-7). So this may help us express scFv successfully.
Figure. 2-7 The map of pRARE helper plasmid.
(6) Dual plasmid co-transformation strategy
Since shuffle T7 strain constitutively expresses disulfide isomerase DsbC compared with Rosetta strain, this may help more soluble expression of foreign proteins. Therefore, we co-transformed scFv expression plasmid and pRARE helper plasmid into shuffle T7 strain. After dual antibiotics screening(Kan and Cm), we successfully obtained the strain carrying double plasmid (as shown in Fig. 2-8).
Figure. 2-8 ScFv-pET28a(+) expression plasmid and helper plasmid pRARE were co-transformed
into shuffle T7 strain. Lane1: DNA ladder, lane2: pRARE helper plasmid, lane3: scfv-pET28a(+)
expression plasmid, lane4: pRARE helper plasmid and scfv-pET28a(+) expression plasmid
were co transformed into shuffle T7 strain, and then the plasmid was extracted.
(7) Detection of Protein expression——selection of expression strain
We compared the expression of scFv protein in several different hosts, and the results are shown in Fig. 2-9. Before IPTG induction, scFv protein was almost not expressed, indicating that our expression plasmid has strong robustness. After induction, scFv protein was successfully expressed. Compared with shuffle T7 strain carrying only expression plasmid, Rosetta strain carrying both pRARE plasmid and expression plasmid and shuffle T7 strain carrying both expression plasmid and pRARE plasmid had significantly enhanced protein expression. In shuffle T7 strain carrying only expression plasmid, the protein mainly existed in soluble form, but the amount was a little small. When two plasmids exist in the same host, the expression of total protein and soluble protein increased significantly, but many proteins also appeared in the inclusion body. However, only in Rosetta strain, the expression of soluble protein was very small. Our results also confirmed the rationality of the dual plasmid transformation strategy [6].
Figure. 2-9 Expression of scFv in different hosts.
S: Shuffle T7 strain carrying expression plasmid, R: Rosetta strain carrying expression
plasmid, RS: Shuffle T7 strain carrying both pRARE plasmid and expression plasmid.
Bef.ind: before induction, Af.ind: after induction, Af.ind.sou: souble protein after induction,
Af.ind.inc: inclusion protein after induction.
(8) Exploration of optimal expression conditions
Although we successfully expressed scFv protein, most of the protein still existed in the form of an inclusion body. We extracted and purified the inclusion body protein and verified that it had no activity. In addition, the refolding steps of inclusion body protein are very complex and can not guarantee 100% recovery of activity, so we chose soluble protein. Therefore, we explored the best conditions, including temperature, inducer concentration, and induction time, in order to obtain the most soluble expression.
Firstly, the protein can have time enough to be folded by reducing the speed of protein synthesis. Therefore, it can be achieved by reducing the temperature and the induced IPTG concentration. We selected four temperature gradients of 10 ℃, 20 ℃, 30 ℃ and 37 ℃. The results showed that with the decrease of temperature, the expression of total scFv protein decreased, but the expression of soluble protein first increased and then decreased (as shown in Fig10 a). We obtained that the optimum expression temperature for soluble protein was 20 ℃. That may be because there is higher dissolved oxygen at low temperature to prevent the anaerobic growth of bacteria from producing a large amount of acetic acid; Secondly, the half-life of antibiotics was longer and the plasmid stability was higher when cultured at low temperature. Moreover, protein has a lower degradation rate and soluble protein has higher stability at low temperatures. Finally, some researchers also believe that low temperature can change the dynamics of peptide folding and increase the amount of correctly folded proteins.
Then, we investigated the effects of different IPTG concentrations on soluble protein expression at 20 ℃. Results, as shown in Fig10 b, the expression of total protein, increased with the increase of IPTG concentration within the given four gradient ranges. However, the detection of soluble protein showed that there was no significant relationship between soluble protein expression and IPTG concentration. This shows that even at low IPTG concentration, the expression of total protein is still excessive, so there are still enough proteins to enter the correct folding process.
Finally, we explored whether the induction time would significantly affect the expression of soluble protein. Results, as shown in Fig10 c, the expression of total protein and soluble protein, increased with the increase of induction time, but the increase of protein expression was not obvious after 18h.
In addition, more strategies to improve the soluble expression of foreign proteins are mentioned in the literature [7]. Including adding special factors to the culture medium to help protein folding; Select appropriate vectors, such as vector pet32b (+) with soluble fusion label or vector with a secretory label; Some metal ions can also improve the soluble expression of protein; Change the composition of the medium or add protease inhibitors. However, due to lack of time, we did not make further exploration.
Figure. 2-10 Optimization of scFv expression conditions.
a. The expression of total scFv protein and soluble scFv protein at different temperatures.
b. The expression of total scFv protein and soluble scFv protein at different IPTG concentrations.
c. The expression of total scFv protein and soluble scFv protein at different induction times.
The histogram shows the relative value obtained by gray value analysis of the strip with Image J software.
(9) Purification of soluble scFv protein from Escherichia coli
After optimizing the expression conditions of soluble protein, we need to purify the protein. pET28a (+) vector has double His-tags, which means that we can purify it by Ni column affinity chromatography. The purification effect is shown in Fig. 2-11. Most of the miscellaneous proteins are washed off by washing a solution of low imidazole concentration, and then the adsorbed scFv protein is eluted off by eluent of high imidazole concentration. By collecting the eluent tube by tube, we get a relatively pure scFv protein. In the future, we can explore more suitable purification conditions, such as increasing the amount of washing solution or using gradient elution. The purified protein sample was added with equal volume glycerol and stored at - 20 ℃.
Figure. 2-11 Purification of soluble scFv protein.
Lane1: DNA ladder, lane2: bacterial lysate before column passing, lane3: bacterial
lysate after column passing, lane4-5: washing liquid, lane6-11: elution liquid.
(10) Determination of concentration of scFv protein after purification
Because the purified scFv protein is not very pure, the actual scFv protein concentration obtained by BCA protein concentration quantitative method will be very inaccurate. Therefore, we used Coomassie Blue Staining to determine the protein concentration. Use a series of standard protein samples with known concentrations and purified protein samples to run SDS-PAGE gel and stain with Coomassie brilliant blue (as shown in Fig 2-12 (a)), use Image J for gray analysis, make a standard curve (as shown in Fig. 2-12 (b)), and calculate the protein concentration represented by the target band in the sample through the standard curve. Finally, the concentration of scFv in the purified sample was 26.74ug/ml(as shown in Fig. 2-12 (c)).
Figure. 2-12 Determination of concentration of scFv protein after purification.
(a) SDS-PAGE results of standard samples and purified samples.
(b) Standard curve of corresponding relationship between protein concentration and gray density.
(c) Concentration calculation of scFv in purified samples.
（11）The activity of scFv protein was detected by Western blot
After obtaining the purified soluble scFv protein, we need to verify its activity to ensure that it can recognize GPC3 antigen. Firstly, after running SDS-PAGE gel with the recombinant human GPC3 sample, we detected the recombinant GPC3 antigen with the commercially available polyclonal antibody against GPC3 and anti-GPC3 scFv respectively (as shown in Fig. 2-13 (a)). Next, we prepared HepG2 cell lysate and detected its GPC3 antigen by Western blot. The commercial polyclonal antibody against GPC3 successfully detected GPC3 core protein(full length, 66kD) and the C-terminal of the cleaved GPC3 protein(30kd) (as shown in Fig. 2-13 (b)), and scFv detected GPC3 core protein(full length, 66kD), the N-terminal of the cleaved GPC3 protein(40KD), and some glycated GPC3 proteins (as shown in Fig13 (c)). These results are consistent with those in the literature[8-10]. The full-length GPC3 can be cut into small subunits of 40KD and 30kd and can be glycated. Meanwhile, It shows that anti-GPC3 scFv has high affinity and specificity.
Figure. 2-13 The activity of scFv protein was detected by Western blot.
(a)The detection of the recombinant GPC3 antigen with the commercially available polyclonal
antibody against GPC3 and anti-GPC3 scFv respectively.(b)The detection of the endogenous
GPC3 antigen of hepG2 cell with the commercially available polyclonal antibody against
GPC3. (c)The detection of the endogenous GPC3 antigen of hepG2 cell with the anti-GPC3 scFv.
（12）The detection of Kd value by ELISA
Next, we want to make some quantitative characterization of scFv to prove that scFv has high affinity. One of the most powerful representatives is determine the Kd (ligand concentration that binds to half the receptor sites at equilibrium) and Bmax (maximum number of binding sites). Therefore, we can use hill model (Y= Bmax*X^h / (Kd^h+X^h)) to equate the kinetics of specific binding[11].
We selected the polyclonal antibody of GPC3 and the nucleic acid aptamer specific for liver cancer cells to compare with scFv. The concentration dependence of their specific binding was characterized by ELISA. The determination results of polyclonal antibody, scFv and aptamer are shown in Fig. 2-14, Fig. 2-15 and Fig. 2-16, respectively. The Kd value of the polyclonal antibody of GPC3 is several hundreds ng/ml (as shown in Fig. 2-14), while the Kd value of scFv reaches several tens ng/ml (as shown in Fig. 2-15), which shows that the affinity of scFv for GPC3 is dozens of times that of the polyclonal antibody. Although the reported Kd values of TLS11a aptamer for hepatoma cells reached 4.51nm/ml and 7.16nm/ml for the mouse hepatoma cell line (MEAR cell line) and human hepatoma cell line LH86 [12], our results showed that TLS11a did not specifically bind to HepG hepatoma cell line (as shown in Fig. 2-16). In conclusion, the results show the high affinity of scFv.
Figure. 2-14 Determination of Kd value of anti-GPC3 polyclonal antibody.
(a) (b) (c) Three ELISA results. (d) Parameters in hill model.
Figure. 2-15 Determination of Kd value of anti-GPC3 scFv.
(a) (b) (c) (d) Four ELISA results. (e) Parameters in hill model.
Figure. 2-16 Determination of Kd value of TLS11a. (a) (b) (c) (d) Four ELISA results.
（13）Cellular immunofluorescence
In order to prove that anti-GPC3 scFv can bind to the membrane of HepG2 cells, we did immunofluorescence experiment. The results show that scFv has good membrane binding characteristics, which can help our recombinant adenovirus bind to the cell surface by the interaction between scFv and GPC3. In addition to the fluorescence on the membrane, the group of polyclonal antibodies also has strong fluorescence in the cytoplasm, especially in Golgi, which may be related to the nonspecific binding of polyclonal antibodies and the cleavage of GPC3 by Furin protease in Golgi [13]. Although TLS11a is considered to bind to the membrane protein of hepatoma cells, the results of cellular immunofluorescence do not show good membrane binding characteristics, which is consistent with the reported literature [14].
Figure 2-17 Immunofluorescence of polyclonal antibody, scFv and TLS11a.
（14） Flow cytometry
In addition to the above experiments, we also analyzed the binding of scFv to HepG2 cells by flow cytometry. As shown in Fig. 2-18, compared with the control group, the probability of FITC positive cells in His-scFv group increased significantly and increased in a concentration dependent manner.
Figure 2-18 The binding of scFv to HepG2 cells was analyzed by flow cytometry.
</p> </html>
Sequence and Features
- 10COMPATIBLE WITH RFC[10]
- 12COMPATIBLE WITH RFC[12]
- 21COMPATIBLE WITH RFC[21]
- 23COMPATIBLE WITH RFC[23]
- 25INCOMPATIBLE WITH RFC[25]Illegal AgeI site found at 223
- 1000INCOMPATIBLE WITH RFC[1000]Illegal BsaI site found at 56
| None |