Introduction

With recent advances in crystallography, the bottleneck for X-ray structure solution projects lies with obtaining suitable samples for crystallization. Technologies developed by structural genomics (SG) initiatives for parallel processing of thousands of targets has profoundly changed the way structural biology projects are managed. Streamlined, miniaturized, automated high throughput (HTP) protocols have become the SG standard. However, coverage takes precedence over success on individual targets. Notoriously intractable targets (membrane proteins/complexes) have only recently been added to the SG portfolio. In non-SG groups with a hypothesis-driven, small portfolio of difficult targets, the majority of project time is spent optimizing cloning, expression and purification protocols. Employing technologies and protocols adopted from SG can considerably reduce the time frame required.

The HTP platform

By screening a wide parameter space of target sequences, plasmids and conditions, one can determine high yield expression conditions for most targets much faster than traditional approaches. The idiosyncratic biochemistry of such targets often demands a higher work input during the expression, purification, and stabilization stages than acceptable for an SG-type project. To combine HTP screening methodology with dedicated, in-depth characterization, we have built up a HTP technology platform encompassing a liquid handling robot (Tecan Freedom Evo II) for cloning and expression/solubility screening, an automated FPLC (Äkta quadExpress), analytical devices, a crystallization robot (Phoenix) and a robotic crystal imager (Rock Maker 182).

Target sequence, expression vector, host, and external expression condition are all screenable variables. Starting from a given cDNA, the HTP platform’s task is to optimize their values and to find expression conditions suitable for milligram-level production within about a month. Downstream HTP-format purification/protein characterization is used to obtain pure and stable protein for functional studies and structure solution. We subject the resulting protein samples to standard crystallization screens on a nanoliter-drop robot. An automated imaging system records crystal growth. Hits are refined through grid screens rationally designed with dedicated software.

Tecan Freedom Evo II: The state of the art liquid handling workstation includes 2 8-tip liquid handling arms, a plate moving arm, a PCR cycler, a UV/VIS/fluorescence/luminescence microplate reader, an automatic plate washer, a temperature controlled incubator with automatic doors, 3 shakers (heatable/1 coolable), an E-gel base and a temperature controlled pipetting rack.

The SLS PX (Protein Crystallography) team builds up an automated crystallization/imaging/diffraction quality screening platform at the third PX beamline.

Our primary host is E. coli. Since a number of our targets require eukaryotic expression, we adapt our protocols for expression in Pichia pastoris, insect cells, and mammalian tissue culture cells to an HTP format. We start with the construction of a set of expression vectors for the chosen host. Only recombination based cloning systems present the target sequence independent versatility to suit an HTP scheme. An additional requirement is the easy switch between different hosts,. A comprehensive set of Gateway™destination vectors has been constructed encoding a wide set of fusions. A generic, cleavable His 6-tag allows affinity purification and immunodetection. A major disadvantage of Gateway are the att recombination sites, which lead to incorporation of artificial amino acids. In order to create expression vectors encoding only authentic target sequence and to minimize costs, we have ported the mating assisted genetically integrated cloning (MAGIC) system by Li & Elledge [1] to our robotics platform.

We screen for soluble, high yield expression combining autoinduction media [2] with a dot blot protocol [3] in a microscale format.

Dot blot showing total (T) and soluble (S) expression levels of different target proteins (left column) in microscale E. coli cultures grown at 30 °C (left panel) and 37 °C (right panel). Different strains are marked as coloured boxes.

Preparative scale cell lysates are purified on an ÄKTAxpress™ automated FPLC. We carry out buffer optimization screens to optimize stability against aggregation using either a Thermal Shift Assay [4] , or an Optimum Solubility Screen [5] .

Äkta Express: 4 automated chromatography machines allows multidimensional chromatographic purification of up to 16 different targets in parallel.

Initial crystallization screens are performed on a Phoenix nanoliter-drop dispensing robot. It sets up a 96-well plate in <3 minutes. Crystal growth is monitored on a RockImager™ imaging robot. Refinement screens designed with RockMaker software are run on the Phoenix.

Phoenix: Nanoliter-drop liquid handling robot for crystallisation.

The fully automated robotics crystallization, imaging, and plate mounting facility setup under construction at the SLS third protein crystallography beamline will allow the screening for diffraction quality directly in the drops.

Formulatrix Rock Maker 182: Automated Plate Imager

Several projects have already been carried out or initiated on our HTP platform. The first centers on enzymes processing micro-RNAs [6]. ~80 target variants (domains/domain combinations) have led to the construction of ~300 expression clones, ~50 of which led to soluble expression. About 20 of these could be produced at large scale, a figure that is about to rise with protocol optimization. Several targets have proceeded to the crystallization stage.

Crystals of a target from the miRNA processing project. They diffract to 1.7 Å resolution.

References

[1] Li, M.Z. and Elledge, S.J. (2005) Nat Genet,37, 311-319.

[2] Studier, F.W. (2005) Protein Expression and Purification,41, 207-234.

[3] Knaust, R.K. and Nordlund, P. (2001) Anal.Biochem.,297, 79-85.

[4] Vedadi, M., Niesen, F.H., lali-Hassani, A., Fedorov, O.Y., Finerty, P.J., Jr., Wasney, G.A., Yeung, R., Arrowsmith, C., Ball, L.J., Berglund, H., Hui, R., Marsden, B.D., Nordlund, P., Sundstrom, M., Weigelt, J. and Edwards, A.M. (2006) Proc.Natl.Acad.Sci.U.S.A,103, 15835-15840.

[5] Jancarik, J., Pufan, R., Hong, C., Kim, S.H. and Kim, R. (2004) Acta Crystallogr.D.Biol.Crystallogr.,60, 1670-1673.

[6] He, L. and Hannon, G.J. (2004) Nat.Rev.Genet.,5, 522-531.

[7] Kambach, C. (2007) Curr. Protein Pept. Sci. 205 - 217