Dynamic Screening Experiments to Maximize Hits for Crystallization† Sahir Khurshid, Lata Govada, and Naomi E. Chayen* Department of Bio Molecular Medicine, DiVision of Surgery, Oncology, ReproductiVe Biology and Anaesthetics, Faculty of Medicine, Sir Alexander Fleming Building, Imperial College London, London SW7 2AZ, United Kingdom.
CRYSTAL GROWTH & DESIGN 2007 VOL. 7, NO. 11 2171–2175
ReceiVed July 24, 2007; ReVised Manuscript ReceiVed August 16, 2007
ABSTRACT: In the first step of crystallization screening, the protein is exposed to a wide variety of reagents at different concentrations. Once a “hit” deemed to be conducive to crystallization is identified, parameters such as precipitant concentration, pH, and temperature are used to produce crystals suitable for analysis by X-ray diffraction. Crystals, crystalline precipitate, and phase separation are usually considered leads that are worth pursuing. Clear drops are mostly disregarded. This paper presents a screening technique that makes use of clear drops. Clear drops are subjected to evaporation with the aim of driving them to supersaturation. The findings reported bring a new dimension to screening and open up the scope for utilizing a potential wealth of crystallization conditions that are currently being ignored. Furthermore, this technique enables the utilization of far less protein sample and allows us to obtain the hits in shorter times. Introduction Searching for crystallization conditions of a new protein requires numerous experiments. There is generally no indication that one is close to crystallization conditions until a crystalline precipitate or the first crystals are obtained. The first step is to screen the protein solution with numerous different crystallization agents in order to find a lead of conditions that may be suitable for crystal growth.1–4 The common way of screening is to set up the trials, incubate them, and then observe them at regular intervals. A lead that is considered worth pursuing exhibits crystals, crystalline precipitate, or phase separation. Clear drops are mostly disregarded. When drops remain clear after a two week incubation, it implies that the protein solution is either undersaturated or in a metastable condition. If instead of letting the experiments just sit and incubate, one can adopt a dynamic approach by initiating evaporation of the drops, which is thereafter arrested, this may drive the trial to supersaturation, resulting in nucleation and subsequent crystal growth. The aim of this work was to devise a means of utilizing such clear drops, which are usually considered a “dead-end” in screening experiments. In addition, this technique was also tested on existing hits, using protein samples at much lower concentrations than those required to produce hits by conventional experiments. Experimental Section Materials. EasyXtal Tools were purchased from Qiagen (cat. no. 132023). Alpha crystacyanin from lobster carapace was supplied by Dr. Zagalsky (Royal Holloway College, London) and concentrated to 10 and 20 mg/mL in 0.1 M TRIS-HCl and 1 mM EDTA, pH 6.1. The cardiac muscle protein at 1.0, 2.0, 3.0, 4.0, 6.0, and 10 mg/mL in TRISHCl, pH 7.0, was provided by Dr. C. Redwood (University of Oxford). Protein PXMn, from Dr. R. Leone (University of Sienna, Italy), was concentrated to 0.5, 1.0, 2.0, 3.0, and 4.0 mg/mL in 10 mM HEPES, pH 7.0, and 50 mM zinc sulphate. The obesity-related peptide was supplied by E. Thompson (Imperial College London) and prepared at † Part of the special issue (vol 7, issue 11) on the 11th International Conference on the Crystallization of Biological Macromolecules, Quebec, Canada, August 16-21, 2006 (preconference August 13-16, 2006). * Corresponding author. E-mail:
[email protected]. Telephone: 44 20 7594 3240. Fax: 44 20 7594 3169.
Figure 1. EasyXtal Tool with its screw caps (courtesy of Qiagen). 1.0, 2.0, and 4.0 mg/mL in deionized water. The membrane protein was concentrated to 0.5 and 1.0 mg/mL in imidazole–TRIS-HCl, pH 7.3, and 1% β-octylglucoside and was provided by Dr. C. Martin and Dr. K. Linton (Imperial College London). Hen egg white lysozyme was purchased from Sigma L-6876 and prepared at 10 and 20 mg/mL in sodium acetate buffer pH 4.7. All chemical reagents were from BDH Chemical Ltd. Poole, Dorset, U.K. Five commercial crystallization screens were used: Crystal Screen (Hampton Research, cat. no. HR2–110) PEG-Ion (Hampton Research, cat. no. HR2–126), Index Screen (Hampton Research, cat no. HR2–144), PACT Suite (Qiagen cat. no. 130718), and Classic suite (Qiagen cat. no. 130701). Design of the Crystallization Plate and Experimental Procedure. The EasyXtal Tool (Qiagen) that was used in this work is a crystallization plate for setting up vapor-diffusion hanging-drop experiments. It consists of wells that are sealed by screw caps and O-rings that replace the traditional coverslips and grease used in vapordiffusion hanging-drop set-ups (Figure 1). The caps have two ridges parallel to each other at regular intervals. When placed over the reservoirs, the screw caps can be aligned at different angles to the plate (Figure 2a–c). A turn of 180° fully seals the cap over the well. Any angle less than 180° loosens the cap and will allow some evaporation 5,6
10.1021/cg700686k CCC: $37.00 2007 American Chemical Society Published on Web 10/18/2007
2172 Crystal Growth & Design, Vol. 7, No. 11, 2007
Khurshid et al.
Figure 2. Schematic diagram of the screw cap showing the ridges. (a) Screw cap at sealed position. x marks the starting point of the cap when sealed. (b) Screw cap when turned by 90°. (c) Screw cap when turned by 180°. which can then be arrested by resealing the cap back to 180°. The position of the ridges on the cap is noted at the start of the experiment and the extent of the seal can be set for each individual reservoir. The screw cap design thus enables controlled evaporation by allowing variable amounts of evaporation of the crystallization drop without exposing it to the air and contamination. This is an advantage over previous evaporation experiments, which had exposed the drops in a vapor diffusion set up, 7 and in microbatch where the drops evaporate until they dry out, as there is no way to arrest the evaporation process.1 All experiments were set up as hanging-drops in the EasyXtal Tool. Two microliter drops were set up over 400 µL of reservoir solution by mixing 1 µL of the protein solution plus 1 µL of reservoir solution. The screw caps were then inverted over the well and tightened. Trials were set up at 20 °C, either in an incubator or in a temperaturecontrolled room. However, the humidity was not controlled. Each set of experiments was performed using the same batch of protein on the same day in the same trays. The first set of experiments involved screening of lysozyme and R-crustacyanin using Crystal Screen, Index Screen, and PEG-Ion Screen (Hampton Research). All observations were recorded for 2 weeks by initially observing them every day in the first week and then every other day the following week. All drops that remained clear after two weeks were then subjected to evaporation by loosening the screw caps by a 90° turn for 18 h and then resealing. The obesity related peptide was set up with the Classic suite and PACT suite screens from Qiagen. Observations were recorded for a week. Evaporation of the clear drops was initiated by loosening the screw caps by a 180° turn for 2 h, and then resealing.
The drop and the reservoir are exposed to the ambient laboratory or incubator conditions by restricted evaporation through the loosened cap. The extent to which the trial was allowed to evaporate was determined by (i) the length of time that the cap was left loosened and (ii) the amount that the cap was loosened (i.e., its angle in relation to the plate). After loosening the caps they were subsequently retightened by turning them to their original sealing position. This study investigated a variation of evaporation times, but for practical purposes, we advise retightening the caps when the drop visually shrinks. In the second set of experiments, previously established hits were explored utilizing between 25 and 50% of the protein concentration used to obtain the original hit. Evaporation times ranged between 2 and 72 h at different positioning of the caps. A duplicate set of experiments was set up for each protein as controls where the drops were not subjected to the evaporation procedure. The trials were monitored initially every hour for 6 h and then daily for at least 1 month. The crystals of four out of the six test cases (five proteins and one peptide) tested using this technique were subjected to X-ray diffraction in order to confirm that they were protein. Two of the cases were not subjected to diffraction, the R-crustacyanin and the peptide. In the case of alpha crustacyanin, the blue coloration of the crystals confirmed their proteinaceous nature whilst for the peptide, optimization to improve crystal sizes is underway. X-ray Diffraction Measurements. The crystals were mounted on cryo-loops (Hampton Research) and tested on:
Maximizing Hits for Protein Crystallization
Crystal Growth & Design, Vol. 7, No. 11, 2007 2173
Table 1. Novel Hits, Not Previously Reported, Obtained from Clear Drops Using the Evaporation Method; Lysozyme and Alpha-Crustacyanin were Evaporated for 18 h and the Peptide was Evaporated for 2 h
protein
protein concentration (mg/mL)
lysozyme
alpha-crustacyanin peptide
appearance of hit post evaporation (h)
crystallization reagents
screen
20
PEG ION
20
PEG ION
10
CRYSTAL SCREEN
20 20
CRYSTAL SCREEN CRYSTAL SCREEN
10
INDEX
20 10
INDEX INDEX
20 20
INDEX INDEX
10
INDEX
20
INDEX
10 4
INDEX PACT SUITE
0.2 M potassium sodium tartrate, 20% PEG 3350 0.2 M diammonium tartrate, 20% PEG 3350 0.8 M KNa tartrate, 0.1 M Na HEPES pH 7.5 as above 1.6 M NaK phosphate, 0.1 M HEPES pH 7.5 0.1 M bisTris pH 5.5, 0.5 M magnesium formate as above 0.1 M bisTris pH 6.5, 0.3 M magnesium formate as above 0.1 M bisTris pH 8.5, 0.3 M magnesium formate 1.0 M succinic acid pH 7.0, 0.1 M HEPES pH 7.0, 1% w/v PEG MME 2000 0.1 M HEPES pH 7.0, 30 & w/v Jeffamine M-600 pH 7.0 1.0 M Na-K phosphate pH 8.2 0.2 M NaI, 20% PEG3350
24 24 48 24 72 48 24 24 24 24 24 48 24 72
Table 2. Comparison of Conventional Screening versus Screening Using the Evaporation Procedure; Conventional Implies Setting up the Trials with Previously Established Conditions and Leaving Them Undisturbed. Evaporation Implies Loosening of the Screw Caps and Then Resealing Them conventional method
evaporation method
protein
conc (mg/mL)
time to obtain hit (h)
conc (mg/mL)
cap opened (deg of turn)
evaporation time
time to obtain hit (h)a
R-crustacyanin cardiac muscle protein lysozyme membrane Protein PXMn obesity peptide
20.0 10.0 20.0 1.0 4.0 4.0
72 168 72 72 48 16
10.0 3.0 10.0 0.5 1.0 1.0
90 90 90 180 90 180
18 48 18 4 72 2
42 96 42 48 96 19
a
Total time including evaporation time for crystals to appear.
The home X-ray source using a Rigaku-RU-H3RHB X-ray generator equipped with Osmic mirrors, a Rigaku X-stream cryogenic system and a Mar 345 detector. Beam line 14.1 at the SRS-Daresbury using a focused, collimated, monochromatic, X-ray beam from a multipole wiggler. The station was equipped with a Quantum 4 ADSC detector, a single axis rotation camera, and a cryojet at 100K.
Results Five proteins and one peptide listed in Tables 1 and 2 were used in this study to test two hypotheses: (a) if controlled evaporation of clear drops could lead to new hits and (b) if hits could be obtained using lower concentrations of samples than those required to obtain hits by standard vapor diffusion methods. Utilizing Clear Drops To Produce New Hits. In the first part of the work, experiments using commercial screens were set up. Two proteins and a peptide were screened using the EasyXtal Tools in the usual hanging drop method.8 Lysozyme was used first, in order to test the method and benchmark the technique. Before the screening trials commenced, an initial test was performed to elucidate the optimal evaporation time that was required to drive the trials into the nucleation zone. Three different conditions giving clear drops were set up in triplicate and subjected to 12, 18, and 24 h evaporation periods,
respectively. In each instance, 18 h proved optimal. Evaporation of 12 h did not have any effect; all the drops remained clear. Twenty-four hour evaporation caused all the drops to precipitate. This is not to say that 18 h is an optimal evaporation time for all conditions; however, it serves as a reference point from which the evaporation period can be optimized further. Following that, 242 conditions each were set up with 10 and 20 mg/mL of lysozyme using three Hampton screens (listed in the Materials and Methods section). The crystallization drops were incubated and monitored for up to 6 weeks. Ninety-seven drops that remained clear after 2 weeks were subjected to evaporation by loosening the caps to a 90° turn for 18 h. Thereafter, they were resealed. The trials then continued to incubate for another 7 days and were monitored daily. Of the 97 clear drops of lysozyme, 12 drops yielded crystals following the 18 h evaporation period (Table 1). In all cases except one, the crystals appeared within 48 h after arresting the evaporation (one drop took 72 h). The crystals obtained were X-rayed and gave diffraction patterns indicating that they were protein. Of these 12 drops that yielded crystals, 8 drops were new conditions, not previously reported in the PDB for hen egg white lysozyme. On the basis of these promising results with the lysozyme experiments, the screening experiments were repeated with R-crustacyanin at 10 mg/mL using the three Hampton Research screens. Fifty drops that remained clear after 2 weeks were
2174 Crystal Growth & Design, Vol. 7, No. 11, 2007
Figure 3. New condition, not previously reported, for R-crustacyanin at 10 mg/mL obtained using the evaporation method. The condition is listed in Table 1.
evaporated by loosening the caps to a 90° turn for 18 h, after which they were resealed. Out of these 50 clear drops, one new hit was obtained (Table 1) that is totally different from the previously reported condition.6 Figure 3 shows a photograph of the new hit for R-crustacyanin. The natural blue color of the crystals confirms that they are protein. On the basis of the results of the two proteins above, an obesity related peptide of 4 kDa with 37 amino acid residues was screened at 4 mg/mL using 192 conditions from two different Qiagen screens (specified in Materials). Ten drops at different conditions remained clear but they remained unchanged when the caps were loosened at 90° for 18 h. Our aim was to expedite the experiment. To this end, evaporation was initiated 4 days after set up and the screw caps were loosened by 180° in order to allow a higher rate of evaporation. The 10 drops that remained clear were set up in triplicate (i.e., 30 drops were set up). Each set of 10 drops was evaporated for 1, 2, and 4 h, respectively and then resealed. Evaporation of 1 h did not have any effect; all the drops remained clear, whereas 4 h of evaporation caused all the drops to precipitate. One condition produced a new hit of tiny crystals 72 h after resealing the drop following 2 h of evaporation (Table 1). The crystals obtained are too small to be X-rayed; optimization trials are currently underway. When the caps were loosened, no apparent gap between cap and reservoir could be seen. The amount of dehydration from the drops and the reservoirs was not measured; however, the volumes of the drops and of the reservoir solutions were measured in the controls (which remained sealed throughout the duration of experiments) and in the loosened ones. The volumes of the control drops remained unchanged throughout the duration of the experiments. The drops over reservoir solutions of both salts and PEGs showed a decrease in volume when the caps were loosened. In the case of caps loosened by 90° for 2 h, the drops were 80% of the volume of the control drops, whereas the volumes of the drops loosened by 180° for the same period of time was reduced to 50%. Twelve hours after resealing the drops following the evaporation period, it was observed that the drops regained almost their original size. The volumes of the reservoir solutions, the controls, and the evaporation trials remained unchanged throughout the duration of the experiments. Control experiments were also performed whereby the screw caps were disturbed by loosening them for a few seconds and then resealing. All such drops remained clear throughout the duration of the trials.
Khurshid et al.
Figure 4. Hit obtained for cardiac muscle protein using the evaporation method at lower protein concentration than that required by standard procedure.
Obtaining Crystals at Lower Protein Concentrations. In the second part of this work, five proteins and the peptide were screened at several concentrations as described in Materials and Methods. The objective here was to test if the evaporation technique would enable the experimenter to utilize lower protein concentrations than were originally required to obtain the same hits by conventional screening experiments. Trials were set up using these previously known hits at lower concentrations of the proteins and peptide. Table 2 demonstrates that the evaporation procedure gave rise to hits at much lower initial concentrations of the samples than that required by the standard procedures and often within a shorter period of time. Previous screening of the cardiac muscle protein produced a lead at the concentration of 10 mg/mL, below which the trials remained indefinitely clear. However, following a 48 h evaporation period, the same lead (which would have stayed clear otherwise) produced that hit at 3 mg/mL (Figure 4). The peptide and the PXMn protein yielded even more noteworthy results where leads were obtained at a quarter of the initial concentration. The initial conventional screening of the peptide, which gave a hit at 4 mg/mL, produced the same hit at 1 mg/mL as a result of loosening the cap for 2 h. For PXMn, the hit was obtained at 1 mg/mL (as opposed to 4 mg/mL) after loosening the caps for 72 h. R-Crustacyanin and lysozyme produced hits at half the concentrations required to produce hits by conventional screening. The most striking result was with the membrane protein, which could not be concentrated beyond 1 mg/mL. The evaporation method resulted in a hit consisting of tiny crystals at 0.5 mg/mL after an evaporation period of 4 h. For all the proteins, duplicate drops were set up at the same low concentrations and left undisturbed (i.e., they did not undergo the evaporation procedure). All control drops remained clear for 4 weeks after setting up trials. Further optimization of the cardiac muscle protein led to its successful structure determination. Discussion The results reported in this paper describe a dynamic means of screening by actively influencing the crystallization trials as they take place. The purpose of loosening the screw caps is to promote nucleation in the clear drops with the intention of driving the system into the nucleation zone and then arresting the evaporation by tightening the cap seal before excessive
Maximizing Hits for Protein Crystallization
nucleation occurs. The initiation of nucleation is not influenced by the nature of the reservoir solution. Because of the large difference in volume between the drop (2 µL) and the well (400 µL), the loss of liquid takes place from the drop. In our test panel of 5 proteins and 1 peptide, the evaporation procedure gave rise to 10 new hits, not previously reported nor detected by traditional screening methods, out of a total of 157 clear drops. This technique opens up a new scope for screening that goes far beyond that of finding new hits from the clear drops. The other major feature of this technique is its ability to employ a fraction of the protein concentration required for standard methods, and for the high throughput experiments that utilize nanolitre volumes. High throughput experiments are very successful in the miniaturization of samples leading to the use of less material; however, setting up nanoliter volumes does not necessarily mean using low concentrations of the protein. Moreover, after hits are attained, problems are frequently encountered (e.g., see ref 9, in scaling up the volumes for crystal optimization). It is often difficult to produce protein samples at sufficient concentrations that are required for crystallization trials. Using the evaporation method, experiments were set up at lower protein concentrations than those applied to produce hits in conventional hanging-drops. The evaporation procedure has enabled us to obtain hits at much lower concentrations of all six test cases investigated in this study. This includes a membrane protein that could not be concentrated beyond 1 mg/ mL and has now produced a hit at 0.5 mg/mL after evaporation. The results demonstrate that such experiments can be performed with standard volume set-ups without the need to scale up. Setting up screens with dilute protein and altering the ratio of the protein solution to the reservoir solution away from the 1:1 can also produce hits that might be otherwise missed.2 However, this requires setting up multiple screening experiments, unlike the evaporation method, which manipulates previously set trials. Conclusions In summary, the evaporation procedure has many advantages: (a) It facilitates the detection of leads that had not been found by standard screening procedures that ignore clear drops. (b) It is a valuable method when a sample is difficult to overexpress or purify or when it is not possible to reach a sufficient
Crystal Growth & Design, Vol. 7, No. 11, 2007 2175
concentration of the sample for crystallization trials. (c) The hits are obtained in a shorter time than under standard conditions. (d) It provides a means to salvage screening experiments in vapor-diffusion, which to date is still the most common way of screening, especially when experiments are performed manually. Acknowledgment. We thank Dr. P. Zagalsky, Dr. C. Redwood, Dr. R. Leone, Dr. C. Martin, Dr. K. Linton, and Dr. E. Thomson for providing us with purified proteins, and Dr. G. Nneji and A. Backshall for a photograph of alpha crustacyanin. The UK Engineering and Physical Sciences Research Council (EPSRC) and the European Commission EMeP Project LSHGCT-2004-504601 and OptiCryst Project LSHG-CT-2006-037793 are acknowledged for financial support.
References (1) D’Arcy, A.; Mac Sweeney, A.; Stihle, M.; Haber, A. The advantages of using a modified microbatch method for rapid screening of protein crystallization conditions. Acta Crystallogr., Sect. D 2003, 59, 396– 399. (2) Dunlop, K. V.; Hazes, B. When less is more: a more efficient vapourdiffusion protocol. Acta Crystallogr., Sect. D 2003, 59, 1797–800. (3) Luft, J. R.; Collins, R. J.; Fehrman, N. A.; Lauricella, A. M.; Veatch, C. K.; DeTitta, G. T. A deliberate approach to screening for initial crystallization conditions of biological macromolecules. J. Struct. Biol. 2003, 142 (1), 170–179. (4) Wooh, J. W.; Kidd, R. D.; Martin, J. L.; Kobe, B. Comparison of three commercial sparse-matrix crystallization screens. Acta Crystallogr., Sect. D 2003, 59, 769–72. (5) Chayen, N. E. Methods for separating nucleation and growth in protein crystallization. Prog. Biophys. Mol. Biol. 2005, 88 (3), 329–37. (6) Nneji, G. A.; Chayen, N. E. A crystallization plate for controlling evaporation in hanging drops. J. Appl. Crystallogr. 2004, 37, 502–503. (7) Jovine, L. A simple technique to control macromolecular crystal nucleation efficiently using a standard vapour-diffusion setup. J. Appl. Crystallogr. 2000, 33 (2), 988–989. (8) Bergfors, T., Protein Crystallization:Techniques, Strategies and Tips, A Laboratory Manual; International University Line: LaJolla, CA, 1999. (9) Brown, J.; Walter, T. S.; Carter, L.; Abrescia, N. G. A.; Aricescu, A. R.; Batuwangala, T. D.; Bird, L. E.; Brown, N.; Chamberlain, P. P.; Davis, S. J.; Dubinina, E.; Endicott, J.; Fennelly, J. A.; Gilbert, R. J. C.; Harkiolaki, M.; Hon, W. C.; Kimberley, F.; Love, C. A.; Mancini, E. J.; Manso-Sancho, R.; Nichols, C. E.; Robinson, R. A.; Sutton, G. C.; Schueller, N.; Sleeman, M. C.; Stewart-Jones, G. B.; Vuong, M.; Welburn, J.; Zhang, Z.; Stammers, D. K.; Owens, R. J.; Jones, E. Y.; Harlos, K.; Stuart, D. I. A procedure for setting up high-throughput nanolitre crystallization experiments. II. Crystallization results. J. Appl. Crystallogr. 2003, 36, 315–318.
CG700686K
