Top tips for Challenging Proteins

Lei Zheng, Paul Scherrer Institute, Switzerland

Tips to crystallisation of membrane protein:

• Try to use our detergent fingerprint approach to examine your targets and decide which family you would like screen at first priority. This approach can save you lots of time and works.
• Try to prolong your linker at terminus of your protein if your protein could not give nice crystal at all after all detergent screen trial. We have found that the resolution of crystal was improved from 20 angstrom to 2.6 angstrom just with a 20 residue flexible linker even we can not see it after solved the structure. It might stabilize the detergent micelle in next layer of crystal contact with unknown dynamic reason. So it might be worth to give a try with uncleavaged tag and even prolonged tag as well.
• To play the lipid as the additive during crystallization and further improvement.

Rita Casadonte, University of Magna Graecia, Italy

• Great reference for comparing different approaches in highly represented proteins from serum: K. Bjorhall, T. Miliotis, P. Davidsson, Proteomics. 5 (2005) 307.
• During the cICAT labelling procedure, we suggest to use 100 mM formic acid to lower pH of the sample prior to injecting it into the cartridge instead of the cation exchange buffer-load provided by the company.
• We strongly suggest to wash extensively the C18 trapping column. Shorter trapping column washing times result, in fact, in visible salt deposits at the MS orifice and subsequent decreased performance of the on-line system after running just a few 2D LC-MS/MS cycles.

Irina Cirpus, University of Technology, The Netherlands

Here my golden rules for lab work. Very simple, but not self-evident for everyone:

• Get motivated, use your imagination, don't give up: What motivates you determines your consistency and desired results.
• Read other literature reviews in your discipline, talk to colleagues about your scientific problem, consult internet fora (But also share your knowledge!!!): Combined knowledge will inspire you to tackle the challenge.
• Use fresh biomass for protein purification, do not freeze protein if somehow avoidable, freeze cells and protein in a cryoprotectant: The quality of your starting material determines your results.
• Know the composition of your solutions and buffers and the function for the ingredients: Basic knowledge about the chemistry enables you to troubleshoot.
• Prepare your solutions freshly: Degraded or ineffective chemicals will sabotage your experiment.

Annamaria de Laurentiis, University "Magna Graecia" of Catanzaro, Italy

• The glycoprotein that we are studying is detectable by Western blotting but not by coomassie blue staining; thus, the protein immunoprecipitated was separated on SDS-PAGE and a small fraction of the sample was separated on the same gel and detected by immunoblotting analysis. A band corresponding to the size of the protein as revealed by Western blotting was in situ digested and eluted from gel for mass spectrometry analysis.
• To quantify the protein in the different purification steps, we used Western blotting followed by densitometry analysis of the specific signal; the amount of protein was calculated as pixels/µg.
• In order to increase the purity of the sample for mass spectrometry, the protein was eluted from the immunocomplex using a peptide mimicking the epitope recognized by the specific mAb. This peptide was previously selected by screening a random peptide library exposed on phages using the same mAb.

Said Eshaghi, Karlinska Institutet, Sweden

For working with membrane proteins:

• Fos-Choline 12 (Anatrace) is a very good choice of detergent for high-yield extraction
• Keep the ionic strength low to avoid protein aggregations
• Always screen a number of detergents when working with a new protein

Titus M. Franzmann, Technische Universitat Munchen, Germany

• If refolding in a single dilution step fails, sequential refolding varying the conditions more and more to native conditions might be the method of choice. It is always good to know the thermodynamic behaviour of the protein. So carrying out analytical equilibrium refolding transitions will give insight into what happens at the corresponding denaturant strength.
• On column refolding might be applicable for several proteins. Its advantage is the quick and convenient buffer exchange for scouting.
• Oligomeric proteins should be partially refolded. First the smallest building block should be refolded. Than association can be carried out.

Melanie M. Hoefer, Biotechnology Institute Thurgau, Taegerwilen, Switzerland

Avoidance of inclusion bodies
Lowering the temperature (30 °C to 20 °C) and increasing the aeration during expression helps to avoid inclusion body formation. Or: Addition of 2% ethanol to the culture when the O/N culture is inoculated into fresh medium.

Purification of insoluble GST-tagged proteins
Insoluble proteins may be GST-purified by dissolving the pellet in a very small volume urea or SDS buffer, and very quickly injecting it into a large volume of PBS (+ protease inhibitors), spinning down and clearing through a filter before loading onto the GST-column. Loss is inevitable, but a decent amount of protein will be soluble.

"Invisible" proteins
Proteins that are not immediately detectable may have diffused from the gel (especially if <10 kDa) . glutaraldehyde instead of 50% methanol fixation may be necessary. Further, alternative stainings may be considered, e.g. reverse zinc and copper stainings, "stains all"-solution or combinations of Coomassie blue with zinc or silver staining.

Martin Kollmar, Max-Planck-Institut for Biophysical Chemistry, Germany

Use of VariPerm series dialysis columns.
These columns can be used to directly dialyse the eluate of a preceding column (affinity column, ion exchanger, etc.) thus saving the time of a normal dialysis, that would last some hours or go over night. The columns consist of several (~ 70) capillaries and work as counter flow dialysis. It is also possible to concentrate the sample during dialysis adding e.g. PEG 20,000 to the dialysis buffer that is not able to diffuse through the dialysis membrane. The method has originally been described in: Schwarz, T., Sindern, S., Bartholmes, P., Kaufmann, M. (1994) The use of a hollow fiber membrane module in sample conditioning prior to electrophoresis. Electrophoresis 15, 1118-9.

Use of homologous genes from different species.
If it is intended to solve the structure or to perform biochemical investigations of an unknown full-length protein, often the N-terminal and/or C-terminal ends of the genes are cut to stabilize the product or enhance the expression level. This most often ends in small domains, and long proteins are separated in many parts. Thus the interaction between the domains, or even between the short 20 aa N- or C-terminus and the core-domain is lost. Instead of cutting the gene I would try to get homologous genes from different species. Over 100 genomes have been sequenced so far and it is easy to get clones from most of them.

Create your own protease inhibitor mix.
Instead of buying the expensive protease inhibitor mixes from companies try to figure out which protease inhibitors are really necessary for the purification of your protein. In most cases only two or three inhibitors are needed, and in the cases, in which strong proteolysis is observed, the mixes from companies often also do not help.

Viktoria Kukhtina, Free University Berlin, Germany

Protein Prediction Programs
Before starting to work with a new protein of the unknown structure try to use the bioinformatics prediction servers to get to know your protein better. Some of the useful servers are the following: DisEMBL and PONDR (prediction of unstructured regions); JPred, PSIPRED and PROF (secondary structure and domain prediction); Eukaryotic Linear Motif Search (prediction of short binding motifs), Robetta (modeling of a tertiary structure ab initio).

Protein Refolding
Finding the right conditions for the protein refolding from inclusion bodies can be done fast with the help of Hampton Research FoldIt Kit. "FoldIt is a fractional factorial protein folding screen which evaluates 12 factors in 16 unique solutions". One can buy a kit from the company, but can also prepare all the solutions by him/herself using the User Guide from Hampton Research.

Soluble Expression
In our hands pET SUMO Expression System from Invitrogen was the best one for production of difficult proteins in a soluble form. "The System utilizes a small ubiquitin-related modifier (SUMO) fusion to enhance the solubility of expressed fusion proteins. After expression, the 13-kd SUMO moiety can be cleaved by the highly specific and active SUMO (ULP-1) protease at the carboxyl terminal, producing a native protein".

Gabriela Lack, Universite de Geneve, Switzerland

The review of S.C.Makrides [1] summarizes the strategies for achieving high level expression in E.coli. In our case, expression of the protein with a highly soluble fusion partner (e.g. NusA-Tag) and the addition of sorbitol and betaine to the growth medium to increase the osmotic pressure [2] made it possible to obtain soluble and active protein.
[1] S.C. Makrides, Microbiological Reviews 60 (1996) 512.
[2] J.R. Blackwell, R. Horgan, FEBS Letters 295 (1991) 10.

Thorsten Lamla, Boehringer Ingelheim Pharma, Germany

• The NanoDrop (ND-1000 Spectrophotometer) from NanoDrop Technologies Inc. The NanoDrop is a photometer which needs only 1 µl of a DNA sample or 2 µl of a protein sample to determine the concentration and to take a spectrum. The machine works very quick and therefore helps to save time.
• Keep the conditions during the purification process. Don't change the required temperature, pH value and salt concentrations of the particular protein if not necessary. Don■ft purify your protein more than necessary. The purest protein is often not the most active one and also no guarantee for crystals.
• Jpred is a tool for protein secondary structure prediction from the university of Dundee. http://www.compbio.dundee.ac.uk/~www-jpred/

Dirk Linke, Max Planck Institute for Developmental Biology, Germany

Protein FTIR
We use a Bruker Tensor 27 FTIR system with AquaSpec cuvette assembly. To my knowledge, this system (including specialized protein software) is the only FTIR system up to date where you can easily mesure protein samples in aqueous environments (without D₂O). It needs protein concentrations of >1mg/ml and a perfect buffer for background measurements, ideally the reservoir buffer from a dialysis experiment. We found it very useful for secondary structure estimation of membrane proteins, as detergents and other UV-absorbing substances do not matter as long as you have the background buffer available (in contrast to CD measurements, where UV absorption of detergents, buffer substances and even of Cl^- Ions is crucial).

Membrane resuspension
Depending on the type of membrane preparation needed, it is sometimes advisable to resuspend membranes to a very disperse suspension. I ususally use a luer lock syringe with a thin needle and suck the membrane pellet in question through that needle. Some
pellets are so tough and sticky though, that we asked our workshop to produce a lever apparatus for us. The apparatus is very simple in design. It holds a 50 ml plastic syringe that can be exchanged. A long lever pulls the stamp of the syringe up, and the operator needs only a fraction of the force compared to manual pulling. A strong temporary vacuum can be created in the syringe that resuspends even the toghest pellets in no time at all.

Phase separation using different detergents
Phase separation is a very simple, cheap and efficient way of separating hydrophobic from hydrophilic proteins. Classical phase separation for proteins usually involves Triton X-114 ("Phase separation of integral membrane proteins in Triton X-114 solution", C.Bordier, JBC 265, 1981, pp 1604-1607) which separates from the water phase at room temperature but is homogeneous at 4■ °C. Similar effects can be induced with almost any detergent by adding salts to the buffer system. C8POE gives nice phase separation with 20% Ammonium sulphate. C12E8 has been shown to "phase separate" in the prescence of sodium citrate. Membrane proteins are usually found in the hydrophobic upper phase, and are not only purified but also concentrated in the process.

Sinead T. Loughran, Dublin City University, Ireland

• The purification efficiency of the affinity tags currently available for purification of recombinant proteins varies, selection of the most suitable affinity tag for protein purification can impact greatly on the cost of the project, the efficiency of the tag should be considered at the project outset using comparison studies, published comparison studies such as those by Lichty et al. 2005 and Fang and Ewald, 2004 can be useful.
• Screening of the open reading frame of the gene to be expressed and purified for rare codons at the initial stages of the project will make selection of the correct expression host and strain straightforward and may save time overcoming expression difficulties by host swapping
• Insertion of a spacer or linker peptide between the two moieties of a Histidine-tagged fusion protein may enhance the flexibility of the histidine-tag to adopt a range of conformations and maximise the opportunity for interaction with resin binding sites during IMAC.

Dunja Lukovic, University of Valencia, Spain

Experimental tips: Recombinant expression and purification of Surfactant Protein C (SP-C)

Concentration of organically extracted SP-C in the presence of lipids
The novelty of our approach lies in the use of organic extraction as a strategy to separate the SP-C moiety from the more hydrophilic one, Staphylococcal nuclease A (SN). The following purification step involves gel filtration chromatography in order to eliminate the components that could have co-purified with SP-C. This requires a previous concentration step in order to minimise the sample volumes generated during organic extraction that are to be loaded onto the lipophilic gel filtration chromatography column. Large scale extractions generate up to 200 mL volumes that have to be concentrated to approximately 2 mL. Concentrating rSP-C in organic solution would change the proportion of organic solvents in favour of methanol, since chloroform evaporates faster. This, together with the accumulation of traces of water, which are difficult to remove, can induce protein precipitation. In order to prevent protein self-aggregation and precipitation, we concentrated the organically extracted rSP-C in the presence of egg yolk phosphatidylcholine added at 1:5 protein: lipid ratio (w:w). This proteo-lipidic mixture is loaded onto the lipophilic Sephadex LH-20 resin and eluted with chloroform: methanol (2:1) solution. In parallel with the elimination of co-extracted peptides, egg yolk lipids used for the concentration of rSP-C-enriched organic solution are removed (Fig 3, peak II).

N-Lauroylsarcosine does not interfere with thrombin activity
Purification of recombinant SP-C has to be performed in the presence of detergent in order to maintain it in the soluble form. The presence of detergent is also required when the peptide is cleaved from its chimera to prevent protein aggregation due to its high hydrophobicity. Therefore, it is necessary to use a detergent which will not interfere with protease activity. We have tested several detergents (CHAPS, SDS, TritonX-100) at different concentrations for their permissiveness for thrombin activity. Samples with N-lauroylsarcosine demonstrated the highest degree of digestion at all the times and conditions tested.

Overexpression of rSP-C as a fusion to Staphylococcal nuclease
Overexpression of membrane proteins is largely known to be difficult. Due to their high hydrophobicity membrane proteins can either integrate into the host cell membranes or hide form the aqueous environment by inclusion into inclusion bodies. In the first case cell membranes collapse and the cultures stop growing while in the later complicated procedures are required for protein refolding.
To avoid any of these possibilities our strategy is based on the fusion of the SP-C protein to the nuclease A (SN) from Staphylococcus aureus, an 18 kDa monomeric soluble protein that has been successfully used previously for membrane protein overexpression. SN is a highly basic protein with numerous lysine residues exposed to the solvent precluding protein aggregation of fused proteins. In addition, in contrast to other frequently employed fusion partners, SN is devoid of cysteine residues, which allows disulfide-mapping of protein-protein domains, or single cysteine spin-labelling for the design of site-specific spectroscopic experiments of the fused protein in membranes. Due to its globular structure, it migrates according to its predicted molecular mass in electrophoresis and in size exclusion chromatography, thus allowing analysis of multimeric structures. All in all, besides designing a successful procedure for obtaining rSP-C, we expect that our strategy will become useful for overexpression and purification of other highly hydrophobic proteins.

Per Nilsson, Uppsala University, Sweden

• In the first step of the purification of a recombinant protein, make sure that the expression is maximised and you will decrease drastically unspecific binding to the matrix you are using.
• Investigate different expression vectors; using the same expression system but with different vectors will give different results.
• Do you have problems with your protein precipitating? If it is a DNA/RNA binding protein it is always a good idea to try a phosphate buffer since the protein needs a lot of charged phosphate groups for its natural folding.

Desiderio Ordoño, Centro Nacional de Biotecnologia, Spain

Deglycosylation of Glycoproteins for Good.
Glycoproteins exposed in membranes are commonly highly glycosilated. This glycosylation is thought to be essential for a correct folding in most of the cases. In X ray crystallization, a highly glycosilated protein is a problem in two ways: on one side, difficults the crystal formation and on the other side glycan chains can reduce the resolution of a diffraction pattern once a crystal has been obtained. These problems can be avoided using mutant celular lines producing homogeneous glycosylation pattern, easily removable from the protein using a specific enzyme. In CHO-Lec 3.2.8.1 production system, enzymes that glycosilate glycoproteins have four mutations, so that it only makes N-glycosylation and excrets glycoproteins with a unique high mannose glycosylation pattern. This pattern is eliminated by digestion with endoglycosidase H, leaving only one N-acetylglucosamine residue per glycosylation site and removing the rest.

Detection of Low Concentration of Glycoproteins in Celular Supernatant.
In a protein production process is very important to check the level of expression obtained, in order to assure an acceptable amount of initial protein for the purification event. An ELISA (enzyme-linked immunoadsorbent assay) is an essential tool to check and quantify the protein of interest prior to its purification. Antibodies both again the protein and engineered tags can be used for protein detection.

Solubilization of Protein Aggregates.
In some cases a pure protein can aggregate and precipitate, for instance, upon removal of glycans, during protein concentration, etc. The problem can be circumvent by the addition of growing concentrations of NaCl, which very often results in the solubilization of protein precipitates.

Phyllis Quinn, University of Nottingham, United-Kingdom

• The biology of IgE and the basis of allergic disease¹ is an excellent review on IgE. It will provide a thorough account on the biology of IgE and the interaction with IgE Fc receptors.
• For those individuals affinity purifying protein from bulk tissue culture supernatant, always do a trial run with a small volume of supernatant to ensure the system is fully functional. The trial run will save you time and effort in the event of a problem with the FPLC machine.
• Avoid purifying IgE using thiophilic chromatography (T-gel). There is no evidence to suggest that IgE has been successfully purified using this method and I had no success when purifying recombinant IgE on T gel

Theres Redeby and Herman Carr, Royal Institute of Technology and Stockholm University, Sweden

• From chloroplast membranes, dissolve proteins directly (without any intermediate steps like acetone precipitation) in the chosen solvent(s) to extract high protein amounts.
• For protein extraction and solubilization, use several different solvents and detergents in parallel, to cover a broader range of proteins.
• Use the fluorinated solvent hexafluoroisopropanol (HFIP) to solubilize highly hydrophobic proteins and/or peptides.

Noelia Sainz Pastor, University College London, United Kingdom

• Staining of proteins transferred onto a PVDF membrane with Coomassie Blue. It detects total protein with a higher sensitivity than the usual for Coomassie Blue staining on polyacrilamide gel. It is very useful when the amount of protein is low.
• Diaconcentration.
  The diaconcentrator dialyses a given molecule by passing the solution through a filter. It retains the molecule and discards a certain amount of solvent, which is replaced by the same volume of the new solution being dialysed against. It is faster than dialysis and also allows concentration.
• Affinity chromatography as a concentration step.
  High capacity matrices can be loaded with big volumes of diluted proteins and can elute them in much lower volumes (usually 1 or 2 column volumes). It avoids loss of protein during an additional concentration step.

Elisa Sala, National Cancer Institute, Italy

• The first tip is the following paper: Gunby, R., C.J.Tartari et al. (2005). "An enzyme-linked immunosorbent assay to screen for inhibitors of the oncogenic anaplastic lymphoma kinase." Haematologica.90(7):988-90 This paper describes an useful test for the screening of small molecule inhibitor of kinases. The ELISA assay presented is easy to optimize, in terms of protein concentration, buffers, salt concentration. It needs the availability of a specific substrate- peptide, containing some residues phosphorylable (serine, threonine, tyrosine). The results usually are consistent with the values of IC₅₀ reported in literature or obtained from other tests.
• The second tip is the protocol of Calcium phosphate- mediated transfection of eukaryotic cells
  This protocol has been recently published on Nature Methods (April 2005, vol. 2, number 4). It's a classical protocol of cellular biology, but is a good and cheap alternative to other common techniques. Normally, we use this procedure to transfect human adherent cell lines, i.e. renal carcinomas, melanomas, papillary thyroid carcinomas with good results. We follow the protocol proposed in Nature Met. and we treat the cells, as final step, with glycerol 15% in HEPES 1M for one minutes and thirty seconds, to increase the percentage of transfected cells.
• The third tip is the Vivaspin Column
  We use Vivaspin column to concentrate our samples containing recombinant purified proteins. The membrane of polyethersulfone avoids the loss of proteins, despite other materials that interact with recombinant proteins. This tool is useful to concentrate large volume in few hours, maintaining the stability of the proteins.

Susana Selles and Juan Casado, University of Allicante, Spain

• To prevent fast oxidation of endogenous phenolics and undesired non-specific interactions with the enzymes of interest, loquat flesh was homogenized in cold buffer containing 50 mM ascorbic acid.
• A cocktail of proteinase inhibitors was added in the homogenate to avoid the proteolysis and activation of the enzyme during the purification process.
• The DEAE anionic exchange cromatography of the detergent-depleted upper phase (particulate PPO) was carried out using Triton X-100 instead of Triton X-114 to avoid the block of the columns.

Neelam Shahani, University of Heidelberg, Germany

Tip: Its important to remember that while purification of small polypeptides, sometimes the conventional traditional methods are very helpful in improving the yield. Its always advantageous to use as many steps as possible.
As an example from our work on purification of MINUS, we introduced the conventional Ammonium sulfate precipitation step to greatly reduce the contaminating proteins, and MINUS polypeptide was present in the enriched soulble fraction at highly saturated (80%) ammonium sulfate solution. We did not use the tubulin affinity chromatography step as was used in the original procedure.

Maarit Takatalo, University of Helsinki, Finland

• French press in cell lysis
  French press is a mechanical cell lysis system that disrupts cells with high pressure. Benefits are: the sample does not warm up helping to keep the protein solution (especially helpful if the protein has tendency to aggregate); there is little variation between different cell lysis times unlike e.g. with sonication; it lysis bacterial and other cells very efficiently and no pretreatments are needed.
• Glycerol in the purification steps
  Glycerol can be used in protein isolation to prevent hydrofobic interactions and to improve folding and solubility as well as stability. Relatively easy modification for buffers that can help in isolation as well as stabilize in storage.
• Rosetta-gamiTM and other modified bacterial expression strains
  Rosetta-gami bacterial strain (Novagen) is developed to help disulphide bond formation in bacterial cytoplasm and to express proteins with different codons from E. Coli. Produced in it, many proteins have higher yields and are more stable. Besides Rosetta-gami, several new expression strains have been introduces that help with difficult proteins.

Miro Venturi, Pharmacia Italia S.p.A., Italy

Concerning top 3 tips for working with difficult proteins, let mine be summarized in one single referral which is the following book:
Membrane Protein Purification and Crystallization, a practical guide, ed. Hunte, von Jagow, Schaegger (2nd edition) Academic Press, 2003

Sebastian Wandinger, Technische Universitat Munchen, Germany

• It might be helpful to add possible co factors during renaturation experiments.
• It might he helpful to try homologous proteins from different organisms, in case that renaturation of a particular protein is not possible.
• It is always helpful to scout a matrix of various conditions to reach effective refolding.