The labeling of lung cancer circulating tumor cells (CTCs) is an important part of specific detection. To this end, we decided to fuse mGreenLantern fluorescent protein (mGL) with a dodecapeptide (namely C7) specifically recognizing CTCs through membranous folate receptors α (FRα) (Xing et al. 2018).
To ensure that C7 recognizes CTCs, we must keep the flexibility of C7 in mGL-C7 fusion protein. A previous report indicated that a linker between the host protein and peptide could effectively enhance the flexibility of the peptide. To determine the proper length of the linker, we decided to model the protein structure of mGL-C7 fusion protein with different linker lengths and simulate the docking of these fusion proteins on FRα.
Find the proper linker that maintains the flexibility of C7 in the mGL-C7 fusion protein.
From the docking result of C7 on the FRα receptor (Xing et al. 2018), we learned that the C-terminal of the C7 peptide is important for recognizing the FRα receptor. Therefore we decided to fuse mGL protein to the N-terminal end of C7 .
▲ Figure 1. Binding mode of C7 peptide to FRα by molecular docking (from Xing et al. 2018)
(A) Overlay of the crystal structures of C7-FRα complexes
(B) 2D interaction of C7 and FRα
In general, Alanine can be used as a linker between the host protein and peptide to enhance the flexibility of the peptide. To determine the apporpriate length of linker that maintains the flexibility of C7, We separately fused 0 to 4 Alanines at the N-terminal of the C7 dodecapeptide as linker to mGL. Subsequently, we used I-TASSER (Zhou et al. 2022) to perform structural modeling of fusion proteins with different linker lengths for further docking procedures.
To simulate the docking of the fusion protein with the FRα (4KM6), we employed the Frodock method (Ramírez-Aportela et. al, 2016 ). This method combines 3D grid-based potentials with the efficiency of spherical harmonics (SH) approximations for docking model simulation. We utilize this method to generate probable docking models of different linker length (0A-4A) fusion protein with FRα (4KM6) for us to observe and further discuss.
Top view | side view | |
---|---|---|
mGL-C7 |
|
|
mGL-1A-C7 |
|
|
mGL-2A-C7 |
|
|
mGL-3A-C7 |
|
|
mGL-4A-C7 |
|
|
▲ Table 1. The docking result of mGL-C7 fusion protein with FRα (4KM6)
with different length of linker (Alanine)
The cylindrical shape is mGL, the red strip on the side connected to mGL is C7 dodecapeptide, and the irregular shape on the right is FRα.
The modeling result showed that the mGL protein and C7 dodecapeptide tend to closely connect without a linker. This would cause a lack of flexibility and hinder the binding of C7 to FRα. However, by introducing 1~4 Alanines as a linker between the mGL molecule and the C7 dodecapeptide (mGL-1A/2A/3A/4A-C7), we observed an expansion of the space between mGL and C7, which keeps the flexibility of the C7 dodecapeptide in the fusion protein.
The docking result of mGL-1A-C7 and mGL-3A-C7 shows that the C7 (red helix) dock on the wrong position of FRα (green helix). On the other hand, the docking of mGL-2A-C7 and mGL-4A-C7 is on the correct blue region, which is near the open pocket of FRα. The docking of C7 dodecapeptide in mGL-4A-C7 is closest to that of original C7. Therefore, we selected the mGL-4A-C7 for subsequent expression and labeling.
We conducted experiments to label SKOV3 cells (FRα positive, a mimic of CTCs) with mGL-4A-C7. As compared to the eGFP control protein, the result shows that the mGL-4A-C7 fusion protein successfully labels SKOV3 cells.
control | 50 μg/mL | 100 μg/mL | 200 μg/mL | |
---|---|---|---|---|
nucleus |
|
|
|
|
mGL-4A-C7 |
|
|
|
|
Merge |
|
|
|
|
Ramírez-Aportela E., J.R. López-Blanco, and P. Chacón (2016). FRODOCK 2.0: Fast Protein-Protein docking server. Bioinformatics, 32(15), 2386-2388
Xing, L., Xu, Y., Sun, K., Wang, H., Zhang, F., Zhou, Z., Zhang, J., Zhang, F., Caliskan, B., Qiu, Z., & Wang, M. (2018). Identification of a peptide for folate receptor alpha by phage display and its tumor targeting activity in ovary cancer xenograft. Scientific reports, 8(1), 8426.
Zhou, X, W Zheng, Y Li, R Pearce, C Zhang, EW Bell, G Zhang, Y Zhang. I-TASSER-MTD: a deep-learning-based platform for multi-domain protein structure and function prediction. Nature Protocols, 17: 2326-2353 (2022)