Prokaryotic soluble expression and purification of bioactive human fibroblast growth factor 21 using maltose-binding protein.

Human fibroblast growth factor 21 (hFGF21) has been characterized as an important regulator of glucose and lipid metabolism homeostasis. Here, to produce hFGF21 efficiently in Escherichia coli, the expression and solubility of hFGF21 were tested and optimised by fusing the protein with one of eight tags: hexahistidine (His6), thioredoxin (Trx), small ubiquitin-related modifier (Sumo), glutathione S-transferase (GST), maltose-binding protein (MBP), N-utilisation substance protein A (NusA), human protein disulphide isomerase (PDI), and the b'a' domain of PDI (PDIb'a'). Each tag increased solubility of the protein when the expression temperature was 18°C. Unlike many other tags that were tested, MBP significantly enhanced the solubility of the protein also in the culture condition at 37°C. Thus, the MBP-hFGF21 construct was further pursued for optimisation of affinity chromatography purification. After tag removal, 8.1 mg of pure hFGF21 was obtained as a final product from 500 mL of starting culture. The protein was then characterised by mass spectroscopy and an in vitro functional assay using NIH-3T3 cells transfected with a ß-klotho reporter gene. These characteristics are similar to those of commercial hFGF21. Thus, the MBP tag is useful for efficient prokaryotic production and purification of bioactive hFGF21.

Granulocyte colony-stimulating factor (GCSF) fused with Fc Domain produced from E. coli is less effective than Polyethylene Glycol-conjugated GCSF.

Human granulocyte colony-stimulating factor (GCSF) is a well-known cytokine for neutropenia treatment. However, daily injections are required due to the short circulating half-life of the protein. To overcome this bottleneck, we fused GCSF with the Fc domain of IgG1 at the C terminus (GCSF-Fc) and with the maltose binding protein (MBP) tag at the N-terminus and expressed it as a soluble protein in the cytoplasm of E. coli. We also conjugated PEG aldehyde to GCSF to make PEG-GCSF. The bioactivities of GCSF-Fc and PEG-GCSF were similar to native GCSF using the mouse M-NFS-60 myelogenous leukemia cell line. The EC50 dose-response curves for GCSF, GCSF-Fc and PEG-GCSF were 37 ± 12 pM, 75 ± 13.5 pM and 46 ± 5.5 pM, respectively. When the proteins were injected into neutropenic rats, the group injected with PEG-GCSF showed the highest and fastest recovery of neutrophils, followed by GCSF-Fc and GCSF. ELISA assay revealed the PEG-GCSF had the longest plasma circulation (>72 h), followed by GCSF-Fc (>48 h) and GCSF (~24 h), which is consistent with the in vivo activities of the proteins. In summary, the GCSF-Fc purified from E. coli was not as efficient as PEG-GCSF in treating neutropenic rats.