Anne S. Robinson (Professor, Chemical Engineering)
Membrane proteins represent over one-third of all proteins in the human genome, yet are less than one percent of all structures found in the protein data bank. Since membrane proteins span the lipid bilayer in their native environment, the need to be produced and stabilized in a native-like membrane environment presents a major challenge in their heterologous expression and biophysical characterization for scientists and engineers. In fact, all but three structures obtained for membrane proteins (128 unique coordinates as of August 31, 2007, http://blanco.biomol.uci.edu/Membrane_Proteins_xtal.html are for naturally abundant proteins. Since most membrane proteins are not naturally abundant, solving the structures for many pharmacologically interesting membrane proteins, such as the superfamily of G-protein coupled receptors, will require developing improved methods for heterologous expression. Grisshammer and Tate (2003) underscored the magnitude of this daunting problem: “Membrane protein structure determination remains one of the great challenges in structural biology … Only once the overproduction of functional membrane proteins becomes easier, will we see exponential growth in the number of atomic resolution structures of integral membrane proteins.”
We have made considerable progress on these goals to date. Publications related to this work are listed below.
Niebauer, R. T., and A.S. Robinson (2006) “Exceptional total and functional yields of the human adenosine (A2a) receptor expressed in the yeast Saccharomyces cerevisiae”, Prot. Exp. Purif., 46, p. 204-211.
Wedekind, A.L., O’Malley, M., Niebauer, R.T., and Robinson, A.S. (2006) Optimization of the Human Adenosine A2a Receptor Yields in Saccharomyces cerevisiae, Biotechnology Progress, 22(5):1249-55.
Bane, S.E., Velasquez, J.E., and A.S. Robinson (2007) “Expression and purification of milligram levels of inactive G-protein coupled receptors in E. coli”, Protein Expression and Purification, 52(2):348:355.
McCusker, E., O’Malley, M., Bane, S.E., and A.S. Robinson (2007), “Heterologous GPCR expression: A bottleneck to obtaining crystal structures”, Biotech Progress, May-Jun;23(3):540-7.
O’Malley, M., Lazarova, T., Britton, Z.T., and Robinson, A.S. (2007) “High-level expression in Saccharomyces cerevisiae enables isolation and spectroscopic characterization of functional human adenosine A2a receptor”, J. Struct Biol., 159(2): 166-178.
Almost every cell in the human body contains membrane proteins at its surface. These proteins sample their cellular environment by recognizing specific extracellular molecules. Upon binding the molecule, the membrane protein, or receptor, undergoes a conformational change, enabling transmission of the signal to the interior of the cell. G-protein coupled receptors, or GPCRs, represent a diverse family of receptors whose ligand molecules include biogenic amines, amino acids, odorants, lipids, ions, proteases, nucleotides, peptides and even photons (Bockaert and Pin 1999). Rhodopsin, the only crystal structure obtained to date from this protein superfamily (~1000 exist in the human genome) was purified from bovine retina, where it is naturally abundant. This will not be possible for the numerous physiologically relevant GPCRs in the human genome as they are typically expressed at low levels in native tissue.
The major advantages of using E. coli as a heterologous expression host are its low cost, ease of use, the variety of strains with different genetic backgrounds, and the ability to generate proteins with isotopic labeling to facilitate NMR studies. However, to date, soluble expression of GPCRs has been quite low, with the best yields obtained from maltose-binding protein fusions to the GPCR of interest. In this case, fusions to the human adenosine A2A receptor yielded ~ 150 µg/L of culture (Weiss and Grisshammer 2002) and the rat neurotensin receptor, ~150 – 200 µg/L of culture (Grisshammer et al. 1999; White et al. 2004).
In contrast to previous attempts at soluble expression, our goal was to utilize the high-level expression power of E. coli by producing GPCRs as protected aggregates, called inclusion bodies. These inactive, insoluble aggregates must then be solubilized and refolded to obtain active receptor protein. We first subcloned nine different GPCRs, and expressed them under a number of different conditions, varying temperature, induction conditions, and strains. Surprisingly, although proteins in general readily express to inclusion bodies, most of the receptors in this study were difficult to express under all conditions tested. However, the tachykinin NK1 receptor was expressed in high yields to inclusion bodies. Following high-level expression, several approaches were investigated in order to identify a method to solubilize the inclusion bodies, and then, using immobilized metal affinity chromatography (IMAC), the NK1 receptor was isolated from inclusion body material and other cellular debris. Using this approach, ~5 mg/L of inactive NK1 receptor can be obtained for use in characterization and refolding studies (Bane et al. 2007). The SDS-PAGE gel below shows the analysis of human NK1R purification by IMAC. Soluble protein refolded from these inclusion bodies shows native-like alpha helicity, as determined by circular dichroism spectroscopy. Current efforts are focused on understanding why many GPCRs are not well expressed as inclusion bodies, as well as determining detergent systems that may promote formation of active NK1 receptor.
Yeast systems offer some of the advantages of E. coli microbial culture yet also have some mammalian-like properties such as a eukaryotic secretory pathway, and much of the canonical G-protein signaling pathway. GPCRs have been expressed in baker’s yeast, Saccharomyces cerevisiae, and the methylotrophic yeast Pichia pastoris, which is capable of growing to higher cell densities. We chose S. cerevisiae for our studies due to its fully sequenced genome, the existence of its three native GPCRs, and the potential to re-engineer the downstream GPCR signaling pathway (Pausch 1997).
By utilizing an integrating vector system, and screening for high expressing cells using flow cytometry, we were able to obtain yeast cells expressing ~4 mg/L of functional human adenosine A2A receptor (A2AR) (Niebauer and Robinson 2006). By addition of a decahistadine tag to the expression cassette, this GPCR was readily purified as described above for the NK1 receptor. Cholesterol addition to detergent micelles was found to be essential for ligand-binding activity in purified fractions, and this activity was also correlated with increased alpha helicity in the presence of cholesterol (O'Malley et al. 2007). Current studies are focused on further biophysical characterization of the A2AR as well as extending this expression system to other GPCRs of interest.
 |
 |
 |
 |
|---|
| ©2008 University of Delaware Chemical Engineering, All Rights Reserved | Last revised: August 15, 2008 |