Performance Characterization of HisPur Cobalt Superflow Agarose

Purity, dynamic binding capacity, reusability and chemical compatibility properties of a cobalt-charged IMAC resin

by Scott Meier, M.S.; Ramesh Ganapathy, Ph.D.; Aaron McBride, M.S.; Barb Kaboord, Ph.D.; Paul Haney, Ph.D.; Chris Etienne, Ph.D. - 05/02/13

The study of protein function and structure has been greatly enhanced by recombinant protein expression strategies that incorporate fusions of affinity tags to aid in purification. The poly-histidine fusion tag (His tag) is especially popular because it is small (6-12 amino acids) and provides for simple and gentle purification by immobilized metal affinity chromatography (IMAC), yielding large quantities of >90% pure protein in one step.

A variety of IMAC supports are commercially available, differing in base support resin, the specific chelator molecule, the method of chelator immobilization, and the metal ion that is chelated (i.e., charged to the chelator). These differences affect the binding capacity, reusability, flow characteristics and chemical compatibility of each IMAC resin, ultimately determining the quality of results (yield, purity, speed) in particular applications. For instance, IMAC resins charged with cobalt typically result in cleaner purifications compared to resins charged with nickel, although nickel usually provides higher binding capacity than cobalt. Also, resins with tetradentate chelators such as nitrilotriacetic acid (NTA) result in less metal ion leaching during purifications compared to tridentate chelators such as iminodiacetic acid (IDA). Careful resin selection and method optimization are required for specific applications and scales.

Thermo Scientific HisPur products are IMAC-based supports for his-tagged protein purification. These include magnetic beads and two kinds of agarose resin, cobalt and nickel varieties, and various sizes and formats of bottles and columns. HisPur Cobalt Superflow Agarose uses a proprietary cobalt-charged tetra-dentate chelator immobilized onto highly crosslinked (Superflow) 6% beaded agarose. In this article, we detail the performance characteristics of HisPur Cobalt Superflow Agarose, which is especially suited for small- and large-scale purifications of recombinant poly-histidine tagged proteins by FPLC.

Purity

To demonstrate purification efficiency of HisPur Cobalt Superflow Agarose, we overexpressed and then used the resin to purify several recombinant His-tagged fusion proteins by FPLC: green fluorescent protein (GFP), lipoic acid ligase A (LplA), and beta-galactosidase (β-Gal). Purifications resulted in elution fractions with 85 to 98% purity levels, as assessed by SDS-PAGE and densitometry analysis (Figure 1). We also tested and compared purification performance between HisPur Cobalt Superflow Agarose and alternative sources of cobalt- and nickel-based IMAC purification resins (Figure 2). Similar levels of purity were achieved with both cobalt IMAC resins; however, HisPur Cobalt Superflow resin yielded significantly greater purity than similar 6x-His tagged protein purifications on a nickel-NTA IMAC resin from Qiagen.

Figure 1. Purification of His-tagged fusion proteins using Thermo Scientific HisPur Cobalt Superflow Agarose. Clarified over-expression E. coli lysates (40mL) were loaded onto columns packed with 10mL of HisPur Cobalt Superflow Agarose. Unbound protein was removed with wash buffer. Bound protein was then recovered using elution buffer (see Methods). Pooled fractions of load lysate (L), flow-through (FT), wash (W) and eluate (E) samples were separated by SDS-PAGE and stained with Thermo Scientific GelCode Blue Stain Reagent (Part No. 24590). Yield was determined using the Thermo Scientific Pierce 660nm Protein Assay (Part No.22662). Purity was determined by densitometry analysis of the stained eluate lanes with Thermo Scientific myImage Analysis Software (Part No. 62237). A total of 10µg protein was loaded from each elution fraction; for other fractions, 1µL of sample was loaded.

Figure 2. Purity of 6xHis-GFP on different IMAC resins. Aliquots (8mL) of an E. coli over-expression lysate corresponding to approx. 16mg 6xHis-GFP were loaded onto 1mL columns (diameter = 0.7cm) of three different IMAC resins: Thermo Scientific HisPur Cobalt Superflow Agarose, Clontech™ TALON™ Superflow Metal Affinity Resin (Cat. 635507), Qiagen™ Ni-NTA Superflow (Cat. 30410). Columns were washed with 10mL (10 column-volumes) of wash buffer containing 15mM or 30mM imidazole for cobalt- and nickel-charged resins, respectively. Bound protein was eluted with 10mL of elution buffer containing 150mM or 300mM imidazole for cobalt and nickel charged resins, respectively. Fractions were collected at 1mL intervals. Aliquots of load lysate (L), flow-through (FT), wash (W) and eluate (1 to 10) fractions were separated by SDS-PAGE, stained with Thermo Scientific Imperial Protein Stain (Part No. 24615). Purity of 6xHis-GFP across all elution fractions in each gel was determined by densitometry with Thermo Scientific myImage Analysis Software (Part No. 62237).

Dynamic binding capacity

Maximizing the performance of protein purification in FPLC protocols is critical and requires a good understanding of the binding characteristic of a resin under a variety of flow rates. The longer the protein sample is in contact with the resin, the higher the binding capacity will be until it reaches the theoretical maximum capacity for the support (Figure 3). Greater binding capacities are possible with slower flow rates, but these conditions necessarily require increased processing time. Therefore, the profile of a resin’s dynamic binding capacity is an important consideration when optimizing a purification process.

Dynamic binding capacity of a column run at a given specific flow rate (a value that is inversely proportional to residence time) is typically reported as the amount of protein that will bind to the resin before 10% of the target protein accumulates in the flow-through fraction (called the “breakthrough”). Maximizing dynamic binding capacity enables the use of higher flow rates and less resin, while still minimizing target protein loss and controlling process time.

We determined a profile of dynamic binding capacities for HisPur Cobalt Superflow Agarose by applying highly purified N-terminal 6xHis-tagged green fluorescent protein (GFP) to a 1mL resin bed at several flow rates (Figure 3). At a rapid flow rate of 1mL/min (equivalent to a short, 1-minute residence time), the resin exhibited a binding capacity of approximately 20mg 6xHis-GFP per mL of packed resin. At a slow flow-rate of 0.1mL/min (equivalent to a long, 10-minute residence time), the resin had a capacity of 65mg of 6xHis-GFP per mL of packed resin.

Variations in tag accessibility between recombinant proteins, as well as the presence of other proteins and biological molecules in a complex lysate may affect the resin’s overall binding capacity. Therefore, it is important to determine the appropriate balance between flow rate (production run speed) and the capacity (production yield) for each process being developed.

Reusability and compatibility

Depending upon the laboratory setting, affinity purification may be performed at scales ranging from sub-milliliter volumes to many liters. Although it is often easier and cleaner to pack and use a new column each time for small-scale purification, this is not practical or necessary for large-scale processes involving repeated cycles of purification from the same bulk sample. HisPur Cobalt Superflow Agarose is a robust highly crosslinked resin designed for use at a wide range of scales. The resin can withstand linear flow rates as high as 1260cm/hr without compressing. This is almost double the linear flow rate at which Sepharose™ 6B Agarose (GE Healthcare) begins to compress (data not shown).

To demonstrate reusability of HisPur Cobalt Superflow Agarose, we graphed the profiles of repeated cycles of chromatography on a single column (Figure 4). No significant decline in binding capacity or elution efficiency occurred after 5 cycles of purification from a lysate sample and 25 cycles of washing and regeneration. HisPur Cobalt Superflow Agarose can be reused multiple times without stripping the metal or recharging the support between purifications because practically none of the cobalt leaches from the immobilized tetra-dentate chelate column during normal purification and regeneration protocols.

Figure 4. Reuse of HisPur Cobalt Superflow Agarose with a standard MES buffered saline regeneration procedure. A single, 1mL column (column diameter = 0.7cm) of HisPur Cobalt Superflow Agarose was processed through 26 cycles of chromatography at a flow rate of 1mL/min. Graph displays the results of every fifth cycle (including the first), which included purification of over-expressed 6xHis-GFP from an E. coli lysate: 5mL lysate, then 5mL wash buffer, then 10mL elution buffer. All 25 intervening cycles included regeneration to remove imidazole and then wash and equilibrate the column: 10mL MES-buffered saline (20mM MES pH 5.0, 300mM NaCl), 10mL ultrapure water, 10mL binding buffer. Yields were identical for all six purification cycles.

HisPur Cobalt Superflow Agarose has low non-specific binding, and very little residual protein remains on the column after purification. In most cases, a simple regeneration protocol is sufficient: wash with MES-buffered saline, pH 5, to remove bound imidazole, then equilibrate with binding buffer. However, depending upon the clarity, solubility and quality of the sample or cell lysate, precipitated proteins and other hydrophobic substances can accumulate on the resin over time. This leads to an increase in column back-pressure, and decreases protein purity and yield during purification. If this situation does occur, resin performance can be restored with a clean-in-place protocol: wash with 6M guanidine-HCl and 1% nonionic detergent solution. A slight decrease in protein yield occurs after multiple runs when a guanidine/detergent clean-in-place protocol is inserted (data not shown). EDTA stripping and regeneration with cobalt is not necessary for cleaning the resin.

Finally, we evaluated the compatibility of HisPur Cobalt Superflow Agarose with various chemicals commonly used during purification, cleaning, or storage of IMAC supports (Figure 5). We stored the resin as a 50% slurry in the indicated solutions for an extended period of time (2 hours or 1 week). Following incubation, we equilibrated each resin sample in binding buffer and compared its static batch-binding capacity to the untreated resin sample. None of these chemicals significantly affected binding capacity. More stringent clean-in-place protocols using HCl (pH < 2) or NaOH (pH > 12) did dramatically decrease the binding capacity of the resin (data not shown), and should be avoided. Users should also avoid strong chelators such as EDTA and EGTA in lysis, binding, washing and elution buffers because these will strip metal off the resin.

Figure 5. Chemical compatibility of HisPur Cobalt Superflow Agarose. 1mL aliquots of HisPur Cobalt Superflow Agarose were stored as 50% slurries in various solutions and then used for purification. Following incubation in each solution, triplicate 0.1mL samples of each resin-aliquot were added to 2mL spin columns (for subsequent processing by centrifugation at 700xg), washed 3 times with 1mL of binding buffer and then incubated for 30 minutes with 5mg of 6xHis-GFP by end-over-end mixing at 22°C. After centrifugation to remove the protein solution, columns were washed 3 times with 1mL wash buffer; finally, bound protein was eluted with 3 x 1mL aliquots of elution buffer. Yield in the combined elution fractions of each sample were determined in triplicate using the Pierce 660nm Protein Assay (Part No. 22662). Error bars are standard deviations of the triplicate samples of each condition.

All purification and binding experiments were performed using the following buffer conditions, unless noted otherwise:

• Binding buffer (25mM sodium phosphate pH 7.4, 0.3M sodium chloride, 5mM imidazole)
• Wash buffer (25mM sodium phosphate pH 7.4, 0.3M sodium chloride, 15mM imidazole)
• Elution buffer (25mM sodium phosphate pH 7.4, 0.3M sodium chloride, 0.15M imidazole)

All wash and elution steps were done with 5 to 10 column volumes (i.e., resin-bed volumes). Buffers can also be supplemented with 8M urea or 6M guanidine-HCl for purification under denaturing conditions. The results outlined in this article are considered typical; however, protein yield and purity are influenced strongly by recombinant fusion protein expression level, conformation, solubility and fusion tag accessibility. In some cases, variation of the buffer conditions and flow rates may be required. All recombinant proteins described herein contain a single 6x-His fusion tag and were expressed using E. coli BL21(DE)3 cells grown in LB media and induced for less than 16 hours at 30°C with IPTG. Cells were harvested and lysed with binding buffer containing 0.005% NP-40 and 1x Thermo Scientific Halt Protease Inhibitor Cocktail (Part No. 78438) using a microfluidizer.

1. Jansen, J-C. (editor). (2011). Protein Purification: Principles, High Resolution Methods, and Applications. 3rd edition. Volume 151 of Methods of Biochemical Analysis. John Wiley & Sons, Hoboken, NJ
2. Bornhorst, J.A. and Falke, J.J. (2000). Purification of Proteins Using Polyhistidine Affinity Tags. Methods Enzymol. 326: 245-254.