A novel approach for the purification of aggregation prone proteins

Abstract

The protein aggregation is one of the major challenges of the biotechnological industry, especially in the areas of development and commercialization of successful protein-based drug products. The inherent high aggregation tendency of proteins during various manufacturing processes, storage, and administration has significant impact upon the product quality, safety and efficacy. We have developed an interesting protein purification approach that separates the functionally active protein from inactive aggregates using a detergent concentration gradient. The C-terminally His tagged nucleocapsid protein of Crimean Congo Hemorrhagic fever virus (CCHFV) has high aggregation tendency and rapidly precipitates upon purification by NiNTA chromatography. Using the new purification approach reported here, the freshly purified protein by NiNTA chromatography was further processed using a detergent gradient. In this new purification approach the active protein is retained in the low detergent concentration zone while the inactive aggregates are promptly removed by their rapid migration to the high detergent concentration zone. The method prevented further aggregation and retained the RNA binding activity in the native protein despite numerous freeze thaw cycles. This simple approach prevents protein aggregation by rapidly separating the preformed early aggregates and creating the appropriate microenvironment for correctly folded proteins to retain their biological activity. It will be of potential importance to the biotechnological industry and other fields of protein biochemistry that routinely face the challenges of protein aggregation.

Fig 1

(A) A modeled 3D structure of CCHFV N protein showing head and stalk domain. The positive and negative charged surfaces are colored in blue and red, respectively. This model was previously reported in [20] where the amino acid composition of RNA binding sites in the stalk and head domains is discussed. (B) A panhandle like secondary structure formed by the partially complementary nucleotides of the 5’ UTR of the CCHFV S-segment derived mRNA. (C) Elution profile of wild type N protein (red) and stalk domain (black) from the HisTrap NiNTA column using the AKTA pure protein purification system (GE Healthcare). The clear and turbid fractions are shown by arrows. (D) The clear fractions 2 and 9 from panel C were combined and tested for RNA binding activity. Shown are the Biolayer interferometry (BLI) sensograms of wild type N protein (black) and stalk domain (red), demonstrating the association and dissociation kinetics for protein-RNA interaction. (F) The precipitated wild type N protein and stalk domain were briefly centrifuged at room temperature. The protein in the precipitated pellet and the supernatant (soluble) were examined by SDS-PAGE.

Fig 2

(A) Elution profile of wild type N protein (red) and stalk domain (black) using the new purification method on AKTA pure protein purification system. (B) The fractions from panel A were 50% diluted with dilution buffer and assayed for RNA binding activity. Shown are the BLI sensograms of wild type N protein (red) and stalk domain (black) for binding to the RNA. (C) The diluted fraction of 10 ml volume is thawed at room temperature after overnight freezing at– 80°C. The number line shows the 1 ml markings on the tube. The 1 ml fractions from 10 ml tube are poured into 1 ml Eppendorf tubes without disturbing the concentration gradient, as shown. SDS-PAGE analysis of the 1 ml fractions is shown. (D & E) The representative BLI sensograms for RNA-protein interaction of some of the fractions in panel C are shown. Wild type N protein (D) and stalk domain (E) from respective fractions 1,4,10 and mixture 1–4 at a concentration of ~ 250 nM and ~ 150 nM were used to generate the shown BLI sensograms, demonstrating the association and dissociation kinetics for N protein-RNA interaction. The fraction 1 in panel D is also shown in Fig 4C as P1 for comparison.

Fig 3. The absorbance at 280 nm of the fractions 1–10 from buffer alone (red), wild type N protein (black) and stalk domain (green) were recorded and plotted verses fraction number.

The data points (fractions 1–7) and (fractions 8–10) were separately fit to straight lines, showing two different slops.

Fig 4

(A) Wild type N protein eluted from the column was diluted 50% by the dilution buffer. A diluted fraction (10 ml) was frozen overnight at -80°C and thawed next morning. The lysis buffer devoid of N protein was subjected to similar purification steps and a diluted fraction (10 ml) was similarly frozen overnight and thawed next morning, shown as buffer alone. (B) The defrosted N protein solution and buffer alone from panel A were poured into Eppendorf tubes in 1 ml fractions (P1-P10) and (B1-B10), respectively, from top to bottom of the concentration gradient. (C) BLI sonograms for the binding of N protein from fractions P1 (250 nM), P10 (250 nM), Mix 1 (125 nM) and Mix 2 (125 nM) with RNA are shown.

Fig 5. A mixture of wild type N protein generated by mixing equal volumes of fraction 1–4 (Fig 2) at a final concentration of 125 nM was tested for RNA binding after different freeze thaw cycles.

The BLI sensograms in different colors demonstrating association and dissociation kinetics of N protein with the RNA after different freeze thaw cycles are shown.

Fig 6

(A) The eluted fractions of the wild type N protein were 50% diluted and either frozen over night at– 80°C or left at room temperature or at 4 oC. Next morning that thawed solution were poured into 1 ml fractions from top of the gradient. Shown are the BLI sensograms of N protein from the fraction 1, demonstrating the association and dissociation kinetics with the RNA. (B). The OD measurements of each fraction at 280 nm was carried out as mentioned in Fig 3. The OD values corresponding to the fraction number were plotted and the data points were fit to a straight line.