Cross-crystallization as a new optimization tool of crystallization procedures


Ivana Tomčová1, 2 and  Ivana Kutá Smatanová1, 2


1Institute of Physical Biology, University of South Bohemia in České Budějovice,

Zámek 136, CZ-373 33 Nové Hrady, Czech Republic

2Institute of Systems Biology and Ecology, Academy of Sciences of the Czech Republic,

Zámek 136, CZ-373 33 Nové Hrady, Czech Republic


Keywords: Crystal morphology; Single crystal growth; X-ray diffraction; Biological macromolecules; Additives; Cupric compounds; Cross-crystallization


The effect of several metal cations (Cu2+, Cd2+, Co2+, Ba2+) was tested in attempts to improve crystallization procedure and verify a newly discovered cross-crystallization method with two selected proteins; di-heme cytochrome c4 from anaerobic purple sulphur bacterium Thiocapsa roseopersicina and sweet-tasting protein thaumatin from the African berry Thaumatococcus daniellii. Presence of Cu2+ ions promoted dramatic improvement in crystal morphology, internal packing and diffraction quality. This investigation qualitatively established the influence of cupric cations on the crystal growth by using the cross-crystallization procedure. It was found that influence of Cu2+ ions produced evidently different outer morphology and internal packing of thaumatin crystals (hexagonal prism). Usually their shape is presented as a tetragonal bipyramids. In the case of cytochrome, the good diffractable crystals were obtained only by using cross-crystallization method with metal-ion salts. Newly grown crystals (hexagonal prisms) of thaumatin and cytochrome displayed as the same primitive tetragonal system and diffracted up to 1.7 Å. Crystals were suitable for high-resolution structure analysis.


The determination of successful crystallization conditions for a particular protein remains a highly empirical process. Screening procedures are rapid and economical means to determine preliminary crystallization conditions. During optimization a variable set of parameters (i.e. pH, precipitant type, precipitant/protein concentration, etc.) is screened to determine appropriate conditions for the nucleation and growth of single crystals suitable for X-ray diffraction analysis. Unfortunately, in many cases this strategy will not produce suitable single crystals. Empirically we have explored another tool used in optimization strategy described by Tomčová [1]. We developed and tested a crystallization procedure to modify crystal morphology, internal packing and also to influence crystal growth. For the first time the metal ion salts were added simultaneously to the protein drop and even to neighbouring drops to allow cross-influence during crystallization experiment. Here we report the effects of selected additives on crystallization of two different proteins; one well-known “model” protein thaumatin [2] and one crystallographicaly unexplored di-heme cytochrome c4.


Description of cross-crystallization method.

Cross-crystallization is a procedure applied to standard vapor diffusion sitting and/or hanging drop method. This procedure is based on using a set of additives that influence the quality of crystal growth. In principle, the inclusion of other droplets (containing chemical substances) against the same reservoir slightly changes the vapor pressure of water over the neighboring drop including protein. As described previously [1], the Emerald BioStructures CombiClover Crystallization Plate (EBS plate, Emerald BioStructures, Bainbridge Island WA, USA) with one central reservoir connected to four satellite drop-chambers (a, b, c, d) via dedicated vapor diffusion channels, was used in this procedure (Fig. 1a, 1b). Each of drop-chambers a, b, c, d was filled with different additive (in this case; chloride salts of copper, cadmium, cobalt and barium) and equal volume of the precipitating agent. The protein was only added into the one drop-chamber containing cupric chloride. Additives and reservoir solution, and not protein, were placed to the three remaining drop-chambers to promote crystallization in the fourth drop-chamber. 

Fig. 1a, 1b: Schematic side and top view of Emerald BioStructures CombiClover Crystallization Plate (EBS plate) for sitting drop experiments. Grey color presents reservoir solution, strap areas indicate each additives and grid represents protein-containing solution.


Cross-crystallization experiments. Cytochrome crystallization.

The cross-crystallization method was used to further improve quality of crystals by addition of additives. Deep red well-shaped cytochrome crystals grew within 3–4 days at 20 °C in the presence of 5 mM cupric chloride and ammonium sulfate in citric acid buffer at pH 5. Those crystals were not reproducible unless the other metal salts (CdCl2, BaCl2, CoCl2) were present in the remaining drop chambers as was described above. These cytochrome cross-crystallization experiments have been tested several times, and in all cases the cytochrome crystals grew only in hexagonal prism form. The same outer shape of crystals was observed when a cytochrome was cross-crystallized by hanging drop (Fig. 2).

Thaumatin crystallization.

Thaumatin was crystallized using the standard sitting drop method [2, 3] with the polyethylene glycol (PEG) as a precipitating agent. Well-constructed tetragonal bipyramids were obtained from these crystallization conditions. The effect of metal salt ions on cross-crystallization was tested. Dramatic change in thaumatin crystal morphology and internal packing was observed when thaumatin was crystallized as hexagonal prisms (Fig. 2). In this case, cupric chloride caused the greatest change in crystal outer shape while the other additives showed no significant effect on crystal growth.

Fig. 2: Overview of thaumatin and cytochrome crystallization experiments show   crystal morphology and internal packing influenced by metal-ion salts.


X-ray diffraction experiments.

Both, cytochrome and thaumatin hexagonal prism crystals with dimensions of approximately 1.00 x 0.05 x 0.02 [mm] (Fig. 2) were tested at the synchrotron DESY/ EMBL. Complete data sets collections were executed at beamline X13 with tunable wavelength using Oxford cryo-system type magnets for crystal mounting. Crystals were removed from the drop with a loop and flash-cooled in a nitrogen stream (Oxford cryo-system) at 100 K at the goniometer part of beamline. A crystal to detector distance of 120 mm was used to collect at least 200 frames of each. The exposure time for each image was 30 sec and the oscillation angle was 1°. Diffraction data were collected to 1.72 Å resolution for cytochrome, to 1.70 Å for thaumatin (tetragonal bipyramid) and to 1.50 Å for thaumatin (hexagonal prism), using MAR CCD 165 mm detector at DORIS storage ring with triangular monochromator and bent mirror beam.


The cross-crystallization method includes several factors that can facilitate protein crystallization, from the promotion of intermolecular contacts by divalent metal cations, stabilization of the protein with salts, to changing the aggregation state with precipitating agents. In fact, any addition of a new substance into a crystallizing mix resulting in crystallization is usually classified as a new crystallization technique and handled as a hot tip. However, the effectiveness of any newly discovered method could not be statistically determined. For example, from previous studies it was found that cupric ions in phosphate buffers have a tendency to produce heavy precipitate and even salt crystals [4, 5]. Another example of an additive effect, which can be explained on a molecular basis, is a formation of intermolecular contacts by intercalated divalent transition metal cations [6]. Cadmium (in sulfate solutions) was long known as a crystallization inducing agent of horse spleen ferritin and has been re-discovered as a useful agent to promote crystallization or to increase diffraction quality in a number of cases [Trakhanov 1998]. However, even with a mechanistic explanation of this effect, no rational prediction regarding the probability of success – except statistical evidence – is available!

The specific morphology of thaumatin and cytochrome crystals may depend on factors such as the source of material used during crystal growth and chemicals in the crystallizing buffer in the mother liquor, or on the mother liquor itself. For a single crystal form the angles between the faces are constant, but this is not true if the crystals belong to the different crystal forms such as tetragonal bipyramids and hexagonal prisms as in thaumatin. Their appearance depends on the use of metal salt cations, such as cupric chloride, and partially on the buffer and the precipitating agent used. We assume these metal ions influence evaporation in the protein drop even if they are absent from that drop. As this effect was tested on two different proteins only, we cannot speculate about how universally applicable this will be. However, the influence of Cu2+ ions on cytochrome crystal growth appears to be specific, because no other successful combination of ion salts with cytochrome was found among these four salts singly or in pairs. A similar effect was observed even in thaumatin crystallization when conditions with cupric chloride produced thaumatin crystals with a different morphology. The combination of four particular salts that promote crystallization can be quite reproducible also with other chemicals or even other volumes of the same drop in the remaining drop chambers.


1.       I. Tomčová, R.M.M. Branca, G. Bodó, Cs. Bagyinka, I. Kutá Smatanová, Acta Cryst. F62 (2006) 820-824.

2.       A. McPherson, Crystallization of biological macromolecules, Cold Spring Harbor laboratory press, New York, 1999.

3.       T.M. Bergfors, Protein crystallization: Techniques, strategies and tips, International University Line, La Jolla, USA, 1999.

4.       A. McPherson, J. Weickmann, J. Biomolecular Structure & Dynamics 7 (1990) 1053-1060.

5.       J. Jancarik, R. Pufan, C. Hong, S.H. Kim, R. Kim, Acta. Crystal. Sec. D. D60 (2004) 1670-1673.

6.       H. Sigel, A. Sigel, Metal ions in biological systems, Marcel Dekker - Taylor & Francis - CRC, California, USA, 1990.



This work is supported by grants MSM6007665808 and LC06010 of the Ministry of Education of the Czech Republic and Institutional research concept AVOZ60870520 of Academy of Sciences of the Czech Republic to I.K.S.