Troubleshooting Purification Methods

The troubleshooting guide below addresses problems common to the majority of purification methods as well as problems specific to a particular method. In the latter case, the relevant method is indicated.

Problem	Possible Cause	Solution
GST-tagged protein does not bind or binds very poorly.	The flow rate used during sample loading is too high.	Decrease the flow rate during sample loading. One of the most important parameters affecting the binding of GST-tagged proteins to Glutathione Sepharose is the flow rate. Due to the relatively slow binding kinetics between GST and glutathione, it is important to keep the flow rate low during sample loading for maximum binding capacity.
	GST-tagged protein denatured by mechanical lysis (e.g., sonication). Too extensive lysis can denature the tagged protein and prevent it from binding.	Use mild mechanical/chemical lysis conditions during cell lysis. Conditions for lysis must be empirically determined.
	GST-tagged proteins have aggregated in the sample, causing precipitation.	Add DTT to the sample prior to cell lysis and also add DTT to the buffers. Adding DTT to a final concentration of 1 to 20 mM may significantly increase the yield of some GST-tagged proteins.
	Concentration of tagged protein is too low.	Concentrate the sample. The binding kinetics are concentration dependent. Proteins with low expression may not bind as efficiently as highly expressed proteins; therefore, concentrating the sample may improve binding.
	The tagged protein may have altered the conformation of GST, thereby reducing the affinity for the GST-tagged protein.	Test the binding of GST from parental pGEX: Prepare a sonicate of cells harboring the parental pGEX plasmid and check binding to the medium. If GST produced from the parental plasmid binds with high affinity, the tagged protein may have altered the conformation of GST, thereby reducing the affinity for the GST-tagged protein. Adequate results may be obtained by reducing the temperature used for binding to 4°C, and by limiting washing.
	Equilibration time is too short.	Ensure that the column has been equilibrated with at least 5 column volumes of a buffer pH 6.5 to 8.0 (e.g., PBS).
	Binding of GST-tagged proteins is not efficient at pH less than 6.5 or greater than 8.	Equilibrate with a buffer pH 6.5 to 8.0 (e.g., PBS) before the clarified cell lysate is applied. Make sure that lysis is done at pH 6.5 to 8.0. Check that the sample has been adjusted to binding buffer conditions.
	GSTrap column: Column needs cleaning.	Clean the column according to the standard cleaning protocol (see Appendix 2, Characteristics of Glutathione Sepharose products). If the GSTrap column has already been used several times, it may be necessary to use a new one.
	Glutathione Sepharose medium has been used too many times.	Use fresh Glutathione Sepharose medium. Also see cleaning procedures in Appendix 2 (Characteristics of Glutathione Sepharose products).
	GSTrap columns. The column or system is clogged, leading to high back pressure and no binding.	Clogged column: Clean the column according to instructions. Make sure the sample has been centrifuged and/or filtered through a 0.45 µm filter. Clogged system: clean system according to manual.
GST-tagged protein is not eluted efficiently.	The volume of elution buffer is insufficient.	Increase the volume of elution buffer used. In some cases, especially after on-column cleavage of a tagged protein, a larger volume of buffer may be necessary to elute the tagged protein.
	The time allowed for elution is insufficient.	Increase the time used for elution by decreasing the flow rate during elution. With GSTrap columns, for best results use a flow rate of 0.2 to 1 ml/min (1 ml HiTrap column) and 0.5 to 5 ml/min (5 ml HiTrap column) during sample application. For centrifugation methods, decrease the centrifugation speed during elution.
	The concentration of glutathione is insufficient.	Increase the concentration of glutathione in the elution buffer: The 10 mM recommended in this protocol should be sufficient for most applications, but exceptions exist. Try 50 mM Tris-HCl, 20 to 40 mM reduced glutathione, pH 8.0 as elution buffer.
	The pH of the elution buffer is too low.	Increase the pH of the elution buffer: Increasing the pH of the elution buffer to pH 8 to 9 may improve elution without requiring an increase in the concentration of glutathione used for elution.
	The ionic strength of the elution buffer is too low.	Increase the ionic strength of the elution buffer: Adding 0.1 to 0.2 M NaCl to the elution buffer may also improve results.
	The glutathione in the elution buffer is oxidized.	Use fresh elution buffer. Add DTT.
	Nonspecific hydrophobic interactions cause nonspecific interaction with the medium or aggregation of proteins, preventing solubilization and elution of tagged proteins.	Add a nonionic detergent to the elution buffer: Adding 0.1% Tween-20 or 2% n-octylglucoside can significantly improve elution of some GST-tagged proteins.
Multiple bands are observed after electrophoresis/ Western blotting analysis of eluted target protein.	Mr 70 000 protein copurifies with the GST-tagged protein.	The Mr 70 000 protein is probably a protein product of the E. coli gene dnaK. This protein is involved in protein folding in E. coli. It has been reported that this association can be disrupted by incubating the tagged protein in 50 mM Tris-HCl, 2 mM ATP, 10 mM MgSO4, pH 7.4 for 10 min at 37°C prior to loading. Alternatively, remove the DnaK protein by passing the tagged protein solution through ATP-agarose or a similar purification medium, or perform ion exchange.
	Partial degradation of tagged proteins by proteases.	Add a protease inhibitor: Multiple bands may be a result of partial degradation of tagged proteins by proteases. Adding 1 mM PMSF to the lysis solution may improve results. A nontoxic, water-soluble alternative to PMSF is 4-(2-aminoethyl)-benzenesulfonyl fluoride hydrochloride (AEBSF), commercially available as Pefabloc SC from Roche Biochemicals. Note: Serine protease inhibitors must be removed prior to cleavage by thrombin or Factor Xa. PreScission Protease is not a consensus serine protease and is insensitive to many of the protease inhibitors tested at Cytiva. PMSF is toxic, with acute effects. Use Pefabloc whenever possible.
	Proteolysis in the host bacteria.	Use a protease-deficient host: Multiple bands may be the result of proteolysis in the host bacteria. If this is the case, the use of a host-deficient strain may be required (e.g., lon- or ompT). E. coli BL21 is defective in OmpT and Lon protease production.
	Cell disruption during mechanical lysis.	Decrease lysis time: Cell disruption is apparent by partial clearing of the suspension and can be checked by microscopic examination. Adding lysozyme (0.1 volume of a 10 mg/ml lysozyme solution in 25 mM Tris-HCl, pH 8.0) prior to mechanical lysis may improve results. Avoid frothing as this may denature the tagged protein. Over-lysis can also lead to the copurification of host proteins with the GST-tagged protein.
	Chaperones may have copurified.	Include an additional purification step: Additional bands may be caused by the copurification of a variety of proteins known as chaperones, which are involved in the correct folding of nascent proteins in E. coli. These include, but may not be limited to: DnaK (Mr ~ 70 000), DnaJ (Mr ~ 37 000), GrpE (Mr ~ 40 000), GroEL (Mr ~ 57 000), and GroES (Mr ~10 000). Several methods for purifying GST-tagged proteins from these copurifying proteins have been described
	Antibodies that react with various E. coli proteins may be present in the tagged protein sample.	Cross-adsorb antibody with E. coli proteins: Depending on the source of the anti-GST antibody, it may contain antibodies that react with various E. coli proteins that may be present in the tagged protein sample. Cross-adsorb the antibody with an E. coli sonicate to remove anti-E. coli antibodies from the preparation. Anti-GST antibody from Cytiva has been cross-adsorbed against E. coli proteins and tested for its lack of nonspecific background binding in Western blots.
Multiple bands are observed after electrophoretic analysis of cleaved target protein.	Proteolysis has occurred in the host bacteria.	Determine when the bands appear: Test to be certain that additional bands are not present prior to PreScission Protease, thrombin, or Factor Xa cleavage. Such bands may be the result of proteolysis in the host. Tagged partner may contain recognition sequences for PreScission Protease, thrombin, or Factor Xa: Check the sequences. See the GST Gene Fusion System Handbook, 18115758 for details