Site News  |   Monitor Keywords  |   Monitor Archive  |   Organizer  |   Account Info  |

Protein crystallization method

This application is a Division of application Ser. No. 10/520,690, filed Aug. 9, 2005, incorporated herein by reference, which is a 371 to PCT Application No. PCT/EP2003/008369 filed Jul. 29, 2003, which claims the benefit of European Application No. 02017135.1 filed Jul. 30, 20002, Belgium Application No. 2002/0509, filed Sep. 1, 2002 and German Application No. 10301342.3, filed Jan. 16, 2003.

The invention relates to methods for the crystallization of macromolecules in a three-phase system, appropriate devices, a three-phase system, and the use thereof in automated crystallization methods.

Knowledge about the three-dimensional structure of biological macromolecules has a dramatic impact both on the development of modern biotechnology and in the therapeutic treatment of diseases. Thus, for example, the determination of the three-dimensional structure of a protein target can provide critical information for the identification and development of novel active substances which interact with the respective protein target.

The elucidation of three-dimensional structures of biological macromolecules is effected most precisely by X-ray structural analysis. A precondition for structural elucidation is that the macromolecules, such as proteins or their complexes, are available as single crystals.

Further, crystallization methods can contribute to the derivation and generation of crystalline formulations of recombinant active substances, and crystal morphology can have a significant influence on the pharmacokinetic properties of a formulation. The examination of polymorphisms of low-molecular weight compounds of synthetic or natural origin may also be realized through crystallization experiments. Another field of application of crystallization is its property as a purification method for biological macromolecules as well as of low-molecular weight compounds. Crystallization methods may be performed on both a small and a large scale.

Usually, it is tried to obtain crystals from saturated solutions of highly purified proteins in the presence of a concentrated solution of corresponding salts or polymers (polyethylene glycols etc.). The water necessary for solubility is more or less withdrawn from the protein solutions during a particular time course, so that associations of different sizes may form from fluctuations of the protein molecules. The formation of nuclei is a precondition of the formation of crystals. These processes are normally examined empirically by selecting different protein concentrations, different salts under varying ionic strength conditions, pH values, buffers, polymers, organic solvents, temperatures etc., which is a multi-parameter problem (McPherson, A., Crystallization of Biological Macromolecules, Cold Spring Harbor Press, New York, 1999).

Different crystallization methods are available. In so-called batch methods, the protein to be crystallized is added as a concentrated solution to a previously stored aqueous solution, or vice versa, in a suitable sample support (Chayen, N. E.; Stewart, P. D. S.; Maeder, D. L.; Blow, D. M.; J. Appl. Cryst., 1990, 23, 297-302; Chayen, N. E.; 3. Crystal Growth, 1999, 196, 434-41). Since in this system there is no enrichment of the precipitation reagents over an extended period of time, but the final concentration is set directly when the protein solution is applied, this is called a “batch method”. This method does not allow for a continuous entry into the metastable region of a phase diagram. Thus, the greatest advantage of classical crystallization methods is lost (kinetic properties and dynamic end-point determination, see FIG. 1).

A modified method of the batch method is to use a paraffin oil to which poly(dimethylsiloxane) (DMS), for example, has been added in stead of the pure paraffin oil, so that a continuous diffusion of water from the solution takes place (D'Arcy A.; Elmore, C.; Stihle, M.; Johnston, J. E.; J. Crystal. Growth. 1996, 168, 175-80). However, this method does not allow for setting an end point because the system is not a closed one. A great drawback is the fact that the protein solution will dry completely over a particular period of time. Thus, it can be used only conditionally for crystallogenesis (see FIG. 1).

In contrast, in classical methods (hanging/sitting drop), one drop of a protein solution (with added reservoir solution) is usually incubated in a closed vessel with a reservoir of an aqueous solution of a predetermined higher concentration with exclusion of air. Over time, water evaporates from the drop, so that the protein concentration increases continuously. An end point is reached when the drop is in equilibrium with the reservoir. Thus, the end point can be set through the concentration of the excess reservoir.

However, the provision of crystals suitable for structural elucidation is often associated with great difficulty or not possible at all by the known methods. It frequently happens that ordered crystallization does not occur for unknown reasons. The interplay of physico-chemical processes occurring during crystallogenesis has not been described clearly to date because crystallization has escaped a consequent analysis. When crystals are obtained, they often fail to be of sufficient size or quality.

In view of the great progress achieved in recent years in the elucidation/identification of gene and protein sequences by novel and automated methods, it is desirable to subject the numerous novel proteins which can now be expressed to systematic examination using automated methods. Also, subsections of X-ray structural analysis have experienced huge increases of efficiency in recent years. Contributions were made, inter alia, by the provision of X-ray sources of high brilliance (synchrotrons) and by the provision of efficient hardware and software for the interpretation of data. However, to date, the known methods for the crystallization of biological macromolecules could not be performed satisfactorily as automated high-throughput methods (Chayen, N. E., and Saridakis, E.; Acta Cryst. 2002, D58, 921-27). Therefore, there is a great need for methods for the crystallization of proteins which are readily automated and thus allow for the screening of the proteins having become available in the meantime and the systematic production of crystals.

In automated diffusion experiments according to the hanging drop or sitting drop methods, closed systems are sealed, for example, by siliconized glass covers or by self-adhesive transparent sheets. In the hanging drop method, a drop must be pipetted respectively to the bottom side of a cover, the cover inverted and placed onto a liquid reservoir without adversely affecting the drop, and an air-tight seal must be achieved. For a robot, this is connected with a relatively high number of tedious operations. Such automated methods are not attractive in terms of cost, speed and reproducibility and are not readily realized with ultrasmall sample volumes (see also Chayen and Saridakis, 2002).

Especially when ultrasmall sample volumes are employed, it is to be taken care that the solutions are protected from exhaustive evaporation as long as the systems are not sealed, for example, by the application of self-adhesive sheets or by covers, i.e., mechanically.

A large number of different devices for the crystallization of biological macromolecules exist. Thus, for example, the U.S. Pat. No. 5,096,676 describes a device for crystallization in a sitting drop which can be sealed, for example, by means of a self-adhesive sheet. Further devices are described in U.S. Pat. Nos. 5,130,105; 5,221,410; 5,400,741; 5,419,278; 6,039,804, and in the applications US-2002/0141905; US-2003/0010278, WO-00/60345, WO-01/88231 and WO-02/102503.

All these devices allow for classical vapor-diffusion crystallization experiments and must be sealed mechanically, for example, by means of a self-adhesive sheet.

In automated batch methods, robots place the solutions used in the corresponding screening setups under paraffin oil, for example (Chayen, N. E., and Blow, D. M.; J. Crystal Growth, 1992, 122, 176-80). Thus, the aqueous solution is directly protected from exhaustive evaporation. This is important, in particular, when ultrasmall sample volumes are employed, because the slightest losses of water can cause the set concentrations to vary significantly and thus affect reproducibility (interference sources). Such automated methods are connected with the usual disadvantages of batch methods, because no change of concentration and, in case of a change in concentration, no end point can be set. However, the application of adaptive crystallization methods based on statistical analysis requires an exact knowledge of the final concentration.

The present invention is based on the problem of providing a crystallization method which overcomes the known drawbacks of the prior art methods. In particular, it should be made possible to obtain crystals of macromolecules as homogenous and large as possible at a relatively small expenditure and with a relatively high probability of success.

The method according to the invention should also be suitable as an automated method for the screening of crystallization setups. In particular, the automated method according to the invention should be readily performed by robots with relatively few operations and overcome the described drawbacks of automated crystallization methods.

The object of the invention is surprisingly achieved by a method for the crystallization of macromolecules in a three-phase system using a vessel containing a lower aqueous phase, a middle phase and an upper hydrophobic phase having a lower density than that of the lower aqueous phase, wherein an aqueous solution of the macromolecules is added to the middle phase to form a fourth phase, followed by incubation.

The invention also relates to methods according to claims 1 to 16, a device according to any of claims 18 to 23, a three-phase system according to claims 24 and 25, crystals according to claim 17, the use according to claim 26, and the structures according to claim 27.

The invention is based on the provision and use of a three-phase system. The aqueous solution of the macromolecules forms a fourth phase which does not immediately mix with the lower phase. Generally, the fourth phase is added in such an amount as to form a drop. Preferably, the fourth phase will migrate to the phase boundary between the lower and middle phases after having been introduced into the vessel. Mixing with the lower phase, which is also an aqueous phase, does not occur. Rather, the fourth phase remains in the system at least until the crystallization process begins. In alternative embodiments, the fourth phase becomes positioned at the phase boundary between the middle and upper phases or within the middle phase. What is important is that no immediate mixing with the lower aqueous phase takes place. The term “fourth phase” generally designates the aqueous solution of the macromolecule, even if the three-phase system consisting of the lower, upper and middle phases has not yet formed completely.