Whitepaper 3:layout

From Screen to Structure using the
Crystal Former
Vivian Stojanoff1, Jean Jakoncic1, Morten O. A. Sommer2, .
PURPOSE: Innovative, simple and effective methods are needed for INTRODUCTION: Obtaining high quality crystals of biological
growing beam-ready protein crystals. The Crystal Former
macromolecules remains a bottleneck in structural biology. The fluid (Microlytic, Denmark) is a device designed to facilitate crystalliza- physics, at the microscale, allows gentle mixing of solutions by diffu- tion by combining unique surface chemistry with highly efficient sion, facilitating the crystallization process. While various microfluidic fluid mixing in a simple to use format. The Crystal Former can be devices have been demonstrated to improve protein crystallization out- used in initial screening efforts using conventional sparse matrix put (1-5), wide scale adoption of these devices has been slow due to screens leading to improved crystallization hit rates. In this study we issues related to reliable crystal harvesting and obtaining crystals that evaluate the feasibility of harvesting and cryo-protecting crystals are immediately suitable for X-ray diffraction studies. The Crystal grown in the Crystal Former channels using a sparse matrix screen Former is designed to enable gentle diffusive mixing of protein and for subsequent X-ray diffraction studies. precipitant inside microchannels resulting in significantly higher crys-tallization hit rates. In addition to the improved mixing kinetics, the METHODOLOGY: The protein Xylanase XYNII (Hampton Crystal Former allows harvesting of crystals from individual reaction
Research, USA) was diluted to a concentration of 18 mg/mL in
chambers without disturbing other reaction chambers. In this way deionized water. Using the JCSG plus screen (Molecular crystals can be harvested from individual experiments and be subject- Dimensions, England) crystallization experiments were set up using ed to X-ray diffraction analysis. In principle, these advantages enable the Crystal Former. Crystallization experiments were incubated at growth of crystals suitable for solving the structure of a biological room temperature for 7 days and crystallization hit rate was evaluat- macromolecule without need for using other crystal growth methods.
ed by manual inspection. Selected channels containing crystals were We wanted to test if we could obtain a high resolution structure for the harvested from the Crystal Former. Crystals were cryo-protected and protein Xylanase XYNII using the Crystal Former as the sole device for extracted using both cryo loops (Hampton Research, USA) and microtools (Mitegen, USA). Crystals were mounted at the X6Abeamline at NSLS at Brookhaven National Labs and diffraction data METHODS: Xylanase XYNII of Trichoderma sp. (Hampton
Research, USA) was diluted in deionized sterile water to a concentra-tion of 18 mg/mL (final protein buffer contained: 21.5 % glycerol, RESULTS: We obtained 11 crystal hits using the Crystal Former for 90mM Na/K phosphate pH 7). Using the sparse matrix screen JCSG
the protein Xylanase XYNII. Crystals grown under varying conditions
plus (6) (Molecular Dimensions, UK) crystallization trials were set up were harvested, cryo-protected and full datasets were obtained. The for Xylanase using 6 Crystal Former devices. Devices were incubated resolution of the refined structure was 1.45 Å. at room temperature for 1 week and inspected for crystal formation.
Crystals produced in specific channels of the Crystal Former were har- CONCLUSIONS: Using a set of 6 Crystal Former chips we were vested by simply peeling back the film covering the underside of the
able to go directly from initial screening using a standard sparse
channels. Scouring the film with a scalpel allows the film to be easily matrix screen to the structure of the protein Xylanase XYNII to 1.45 lifted from the channel area exposing the channel and crystal for Å resolution. The Crystal Former can be used as an efficient and self extraction. Once the channel is exposed, 5 µL of cryo-protectant is sufficient crystallization tool allowing users to more efficiently added to the open channel to prevent drying of the crystal during explore the protein crystallization phase space, while enabling users manipulations and protect the crystal during the subsequent flash to collect full diffraction datasets from crystals grown within the cooling. Cryo-protectant was made up by combining equal volumes of the precipitant condition yielding crystals and 50 % glycerol. Crystalswere harvested from microchannels using conventional nylon loops 1 NSLS, Brookhaven National Labs, Upton, NY 11973, USA 2 Microlytic North America Inc., 300 Trade Center, Suite 3650, Woburn, MA 01801, USA.
Correspondence should be directed to MOAS: [email protected] (Hampton Research, USA) as well as various microtools (Mitegen, Table 1 | X-ray crystallographic data collection statistics.
USA). In general it is best to use loops or microtools with diameters of100 µm or less to manipulate the crystals within the 150 µm channel.
Data reduction
Crystals were flash frozen and stored in liquid nitrogen for eventualdata collection. Diffraction properties were evaluated at the X6A beamline at Brookhaven National Laboratory and full data sets werecollected for suitable crystals.
All crystals were randomly mounted on the goniostat. A complete data set consisting on 300 frames, 1 degree oscillation each was recorded using the ADSC Q210 detector on a single crystal. The exposure time was 30 s and the distance was set at 200 mm. Data processing and scal- ing was performed with HKL3000 (7). The structure was solved from data from a single crystal by the molecular replacement method using the program MOLREP (8) and the 1.50 Å resolution Xylanase-IIstructure with PDB ID 2JIC as the search model (9). Refinement con- sisted of repeated cycles of model building in COOT (10) and refine-ment in REFMAC (11). On the Ramachandran plot, 91% of the residues were in the most favored regions and 9% in the additionallyallowed regions. Detailed data collection and refinement statistics are RESULTS: Xylanase crystals were grown in 11 out of the 96 condi-
tions tested (JSCG plus # 7, 15, 23, 52, 57, 58, 60, 70, 71, 74, 76).
All crystals were verified to be protein crystals by X-ray diffraction.
Resolution ranged from 1.4 to 4 Å. A crystal from JCSG plus condi- tion 76 (0.2 M trimethylamine N-oxide, 0.1 M TRIS pH 8.5, 20 %w/v PEG 2000mme) was used to collect a full data set (Figure 1 and Refinement
2). This crystal diffracted to 1.45 Å resolution and belongs to primi-tive monoclinic P21 space group. Final refined structure had a resolu-tion of 1.45 Å (Table 1 and Figure 2). There are 2 complete xylanase Figure 1 | Crystal of Xylanase XYNII of Trichoderma sp. grown in the
Crystal Former channel using 0.2 M trimethylamine N-oxide, 0.1 M TRIS
pH 8.5, 20 % w/v PEG 2000mme as the precipitant after 7 days of incuba-
tion at room temperature. The crystal was harvested from the Crystal Former
device and used for collection of full diffraction data set (Figure 2).
molecules per asymmetric unit, 3 glycerol molecules were identified inthe sugar binding domain of one of the 2 monomers and 2 glycerolmolecules in the second. The root mean square deviation on the C-αbetween the 2 monomers is 0.15 Å.
CONCLUSIONS: We used the Crystal Former to grow crystals of
Xylanase XYNII of Trichoderma sp. using a commercially available
sparse matrix screen. We have demonstrated that crystals grown in the
microchannels can be readily harvested and flash frozen using standard
cryo loops or microtools. Crystals grown in and harvested from the
Crystal Former showed excellent diffraction properties and allowed for
collection of full datasets from individual crystals and subsequent high
resolution structure determination. This work demonstrates the utility
of using the Crystal Former as a self contained crystallization screening
tool for growing high quality crystals immediately suitable for X-ray
diffraction studies.
Figure 2 | Structural model of Xylanase XYNII of Trichoderma sp. Full data
set collected on individual crystal grown in the initial crystallization screen
and harvested directly from Crystal Former channel. Final structure has a res-
olution of 1.45 Å.
References
1. C. L. Hansen, E. Skordalakes, J. M. Berger, S. R. Quake, Proc Natl Acad Sci U S A 99, 16531 (2002). 2. C. L. Hansen, M. O. Sommer, S. R. Quake, Proc Natl Acad Sci U S A 101, 14431 (2004).
3. B. Zheng, L. S. Roach, R. F. Ismagilov, J Am Chem Soc 125, 11170 (2003). 4. K. Dhouib et al., Lab Chip 9, 1412 (2009). 5. J. D. Ng, P. J. Clark, R. C. Stevens, P. Kuhn, Acta Crystallogr D Biol
Crystallogr 64, 189 (2008). 6. J. Newman et al., Acta Crystallogr D Biol Crystallogr 61, 1426 (2005). 7. W. Minor, M. Cymborowski, Z. Otwinowski, M. Chruszcz, Acta Crystallographica Section D-Biological
Crystallography 62, 859 (2006). 8. A. Vagin, A. Teplyakov, Journal of Applied Crystallography 30, 1022 (1997). 9. R. Moukhametzianov et al., Acta Crystallographica Section D-Biological Crystallography 64,
158 (2008). 10. P. Emsley, K. Cowtan, Acta Crystallographica Section D-Biological Crystallography 60, 2126 (2004). 11. G. N. Murshudov, A. A. Vagin, E. J. Dodson, Acta Crystallographica Section D-
Biological Crystallography 53, 240 (1997).

Source: http://www.microlytic.com/sites/default/files/CF_Whitepaper_FromScreenToStructure.pdf