Recombinant Protein Expression for Structural Biology in HEK 293F Suspension Cells: A Novel and Accessible Approach

The expression and purification of large amounts of recombinant protein complexes is an essential requirement for structural biology studies. For over two decades, prokaryotic expression systems such as E. coli have dominated the scientific literature over costly and less efficient eukaryotic cell lines. Despite the clear advantage in terms of yields and costs of expressing recombinant proteins in bacteria, the absence of specific co-factors, chaperones and post-translational modifications may cause loss of function, mis-folding and can disrupt protein-protein interactions of certain eukaryotic multi-subunit complexes, surface receptors and secreted proteins. The use of mammalian cell expression systems can address these drawbacks since they provide a eukaryotic expression environment. However, low protein yields and high costs of such methods have until recently limited their use for structural biology. Here we describe a simple and accessible method for expressing and purifying milligram quantities of protein by performing transient transfections of suspension grown HEK (Human Embryonic Kidney) 293F cells.


Introduction
The rapid progress of molecular cell biology and the constant need for improved drugs in medicine has created a need for structural biologists to look at increasingly more complex protein structures. These can often require particular post-translational modifications, molecular chaperones and co-factors to support their elaborate folding and enzymatic activity. Whilst the vast majority of structures present in the Protein Data Bank have been obtained using bacterial expression systems, prokaryotes are unable to perform a large number of these modifications and lack many essential eukaryotic co-factors. This can be an issue for the study of large multi-subunit complexes that are activated by small signaling molecules as well as for nuclear, cell surface and secreted proteins that require elaborate folding machineries. A number of E. coli strains have been engineered to overcome some of these limitations 1 . In recent years, however, the use of mammalian expression systems has been increasing because they reliably produce eukaryotic proteins that are otherwise problematic to express in other systems 2 . Today it is possible to obtain stable and transient mammalian expression cell lines through a variety of techniques that range from viral transduction to chemicalmediated transfection, physical gene-transfer such as electroporation and direct injection 3 . While all of these methods have their own set of advantages and disadvantages, only a few of them are suitable for structural studies being either too expensive and/or time consuming.
Here we describe a very simple, fast, inexpensive yet highly efficient method for expressing protein complexes for structural biology in suspension grown mammalian cells. The approach uses transient co-transfection of a Human Embryonic Kidney (HEK) cell line (e.g., Freestyle HEK 293F cells). These cells have been derived from the HEK 293 cell line, are adapted to grow in suspension cultures, reaching high densities using serum-free media (such as FreeStyle 293 Expression Medium). Cells are then transiently transfected using a branched version of polyethylenimine (PEI), an inexpensive polymeric reagent that has been reported to function for a large range of mammalian cells 4 by forming DNA/PEI complexes that enter the host cell by endocytosis 5 . This method is suitable for both small scale (30 ml) and large-scale (up to 300 ml) experiments and can produce high levels of purified protein complexes. It is particularly useful for studying proteins that require complex folding machineries, co-factors or particular post-translational modifications that cannot be performed by bacteria, yeast and insect cells.

Low number of cells transfected.
Make sure cell density is approximately 1.0 x 10 6 before adding the transfection reaction mixture to the culture.
Cells might have been subcultured too many times.
Use fresh stock cells after approximately 90 passages.
DNA may be degraded or have a high amount of impurities.
Make sure the plasmid DNA being used has a 260/280 ratio between 1.8 and 2.0. Running the DNA on an agarose gel is advisable to assess its quality.

Discussion
We have developed a straightforward and cost effective method (PEI costs much less than commercially available lipophilic transfection reagents) for expressing and purifying large amounts of recombinant proteins and multi-subunit complexes from mammalian cells. Optimal transfection and expression efficiency can be reached if highly pure plasmid DNA (260/280 between 1.8 and 2.0) is used in combination with PEI as described in the protocol section. Cells must be cultured in serum-and antibiotic-free media, therefore, sterile technique is strictly required for passaging and transfecting cells in order to avoid costly and time-consuming infections. Cell viability should be 90% or higher and cultures should not be grown to densities greater than 2.5 x 10 6 cells/ml immediately before a transfection, as doing so will reduce protein yield. For a successful transfection, cultures should only contain single or dividing cells. Clusters may be broken up by vigorous vortexing 20 to 30 sec ( Figure 5). The ratio of each plasmid used in the co-transfection can be varied by the user according to the stoichiometry of the complex being studied. Expression efficiency can be optimized for the protein of interest, so that a suitable expression time can be established. Purification should be performed keeping the protein sample cold at all times to reduce the risk of unwanted proteolytic degradation. The choice of tags and purification buffers is critical, since they may interfere with structural elements or active sites of certain proteins, often resulting in reduced solubility and/or loss of enzymatic activity. Small scale transfections are particularly useful for testing different constructs for the purification of large protein complexes.
Today, a large variety of alternative methodologies are available for the expression of recombinant eukaryotic proteins ( Table 2). For example, Baculovirus expression in insect cells is widely used due to its high transduction efficiency and its lack of cytotoxicity in comparison to other viral species 6 . However, the process of making the virus is time-consuming and its instability does not allow the virus to be stored for long periods. On the other hand, yeast expression systems offer the possibility of growing cells to very high densities in fermenter cultures, resulting in high protein yields , which is not found in prokaryotic cells. This method has also allowed us to purify and solve the crystal structure of HDAC1 interacting with MTA1 in the NuRD complex, which is similarly activated by IP4 9 . Ideally, we would always like to express eukaryotic proteins in their natural, physiological environment. Mammalian expression systems employing HEK 293-EBNA1 cells in bioreactors [10][11][12] have been described and yield very high levels of protein, but these can be complex to use.
Our method is a simple and accessible alternative to expression in bacteria, yeast and bioreactor systems. The expression method is fast and does not require the use of expensive equipment, particularly if the roller bottle or flask atmosphere is replaced with 5% CO 2 and standard shaking incubators are used. We have used these protocols to co-transfect multiple plasmids and purify complexes with up to five proteins with yields of >1 mg/L of culture. Interestingly, the system helps the identification of stable well-behaved complexes. For instance, expression of the binary complex of HDAC1 and Sin3A gave limited yields of protein, yet addition of SDS3 resulted in 5-fold higher amounts and hence guides the structural biologist in choice of suitable constructs and stable complexes for crystallization.

Disclosures
The authors declare that they have no competing financial interests.