Technical Highlight - October 2008
Short description: Proc. Natl Acad. Sci. USA 105, 14371-14376 (2008)
Production of membrane proteins for in vitro experiments can be very difficult. Escherichia coli is the most commonly used organism for overexpression of recombinant proteins, and although high concentrations of recombinant membrane proteins can be achieved, they often end up in inclusion bodies.
Wagner et al. have a solution. They have engineered a strain of E. coli, called Lemo21(DE3), in which overexpression can be precisely controlled. With it they achieved high yields, not through ramping up overexpression levels in individual bacteria, but by ensuring good overall growth and ultimately a higher biomass than achieved with other strains.
In the commonly used E. coli strain BL21(DE3), the bacteriophage T7 RNA polymerase (T7RNAP) usually drives recombinant protein production. T7RNAP is in turn controlled by the strong promoter lacUV5. But most overexpressed membrane proteins end up in the cytoplasm and in aggregates, which are often toxic to the cell. The accumulations are thought to be because the Sec protein, which mediates the translocation and integration of secreted proteins across or into the cytoplasmic membrane, becomes swamped and stops working.
So Wagner et al. turned to the Walker strains of E. coli. These are known to produce a good yield of membrane protein, and the overexpressed proteins do not seem to be particularly toxic to the cell. But the reason for this was not clear.
Analysis of Walker strains overexpressing the membrane protein YidC revealed the key mutations in the lacUV5 promoter that led to improved protein expression. Wagner et al. then used this information to engineer a BL21(DE3) derivative strain – Lemo21 (DE3) – in which activity of T7RNAP can be precisely controlled by its natural inhibitor T7 lysozyme (T7Lys). They put T7Lys under the control of an L-rhamnose-inducible promoter (rhaBAD), so that adding L-rhamnose in different amounts to the cultures resulted in different and scalable concentrations of YidC.
Use of Lemo21(DE3) should simplify membrane protein overexpression. It requires only a single strain and a simple titration of L-rhamnose and therefore could be well suited to high-throughput applications. Notably, ongoing studies indicate that Lemo21(DE3) also performs very well for the overexpression of 'difficult' soluble proteins.
Samuel Wagner, Mirjam M Klepsch, Susan Schlegel, Ansgar Appel, Roger Draheim et al. Tuning Escherichia coli for membrane protein overexpression.
Proc. Natl Acad. Sci. USA 105, 14371-14376 (2008). doi:10.1073/pnas.0804090105