BG108222A - Method for the preparation of macromolecular crystals - Google Patents

Abstract

The present invention relates to a method of producing macromolecular crystals comprising the steps (i) dispensing a solution of macromolecule and crystallizing agent so that the solution is, or becomes under a layer of oil, wherein the oil layer is such that evaporation from the solution is permitted; (ii) incubating the solution for at least 1 minute; and (iii) subsequently administering additional oil to the oil layer such that evaporation from the solution is decreased, and materials useful in said method.

Description

The present invention relates to a technique for optimizing the crystallization of macromolecules and its application to automation and high throughput systems.

Automation is crucial for highly efficient crystallization as well as for other stages of structural genomics, since the need for good crystals requires many different crystallization conditions to be tested. The main efforts and resources in the field of crystallization are directed towards the automation of the screening procedure applied to find the conditions for crystallization. However, despite the ability to carry out numerous samples so far, only a small percentage of the proteins obtained have led to definitions of structures. This is because, usually, screening alone is not enough - it must be supplemented by an equally essential procedure for obtaining crystals, ie. of (crystal quality) optimization.

The topic of protein crystallization has recently gained new strategic importance in the next phase of the genomic project, in which X-ray crystallography will play a major role. There has been a great deal of progress in the stages of protein automation, as well as in X-ray analysis and bio-informatics, provided that crystals that are able to diffract are present. However, good methods for the crystallization process itself are not consistent with this progress.

Optimization techniques were a little underrated, mainly because it was wrong to expect that large-scale screening alone would produce the desired result. In addition, optimization was based on specific, individually applicable methods that are often difficult to automate and adapt to high efficiency. The real stumbling block in structural genomics is manifested in the various pilot projects currently underway. They show that the success rate in the transition from cloned proteins to definitions of structures is *

only about 5 to 10%. For example, the numbers taken from the Human Proteomic Structural Genomics pilot project (Brookhaven National Laboratory, Rockefeller University and Albert Einstein College of Medicine: http://proteome.bnl.gov/progress.html) indicate that 123 cloned after 59 were read routine screening 33 of these 59 give some crystals. However, only 15 of these crystals are of good enough quality to be used for defining structures, and to date only 10 structures have been found (Fig. 1). During the Cambridge Conference, September 2000, other structural genomics projects reported similar successes at: http://www.mgms.org/cambs2000. This obviously sheds light on the general problem that even when proteins can be cloned, expressed (i.e. obtained by DNA), solubilized and purified, and even when crystallization attempts lead to some crystals, this is not usually a guarantee that these crystals are good enough to allow the structure (of the protein molecule) to be determined. These problems need to be addressed in order for Structural Genomics to succeed.

There has been a lot of progress lately in both automation of production (via DNA) and protein purification methods [1] and in X-ray analysis [2] and modeling [3]. Nowadays, when commercially available kits of crystallizing agents and computer algorithms are available to outline potentially suitable crystallization conditions, it is no longer a big problem to automate crystallization attempts [4, 5]. The automatic screening of high-performance crystallization experiments is also approaching [2, 6]. However, there are some issues that still deserve attention. These include: (i) the need for much manual preparatory work before automation, (j) the need to clean hundreds of syringes, and (iii) the monitoring, monitoring and analysis of results. Some proteins will probably crystallize during this initial screening, however, most attempts are likely to yield microcrystals or low-order crystals. Transforming such crystals into useful samples requires intelligence and individual optimization techniques. Such techniques do not easily lend themselves to automation and must be adapted to handle the huge amount of experiments required by Genomic Projects. Therefore, the problem of optimization is too underestimated if we abstract from the obvious first steps - only to change the concentration or pH (of the solution) around the conditions of interest. The need for high efficiency at each stage of structural genomics has led us to begin designing automated optimization methods that go beyond the usual fine-tuning of conditions.

Manual optimization methods are known, some of which have led to the successful crystallization of proteins that could not be crystallized otherwise [e.g. 7, 8]. We have now developed an automated optimization technique.

The first aspect of the invention provides a method for growing macromolecular crystals, comprising the steps of:

(i) placing a solution of macromolecules and a crystallizing agent in such a way that the solution is located or subsequently falls under a layer of oil, the layer of oil allowing evaporation of the solution through it.

(n) allow the solution to age for at least 1 minute; and then (iii) an additional amount of oil is poured onto the oil layer to make the solution evaporation negligible.

The macromolecule can be any macromolecule, but preferably it is a biological macromolecule. Furthermore, it is even more preferred that the biological macromolecule is a polypeptide.

The crystallization agent and the macromolecule may be placed under the oil at the same time, or may be arranged individually or sequentially in any order.

The crystallization agents are known to the person skilled in the art and their composition can be optimized according to the nature of the macromolecule to be crystallized. Typical crystallization agents are salt and buffer.

The rate of evaporation of the solvent from the drop of sample or solution under the oil may be reduced in such a way that it is less than before the addition of oil in step (iii), or it may be reduced to a negligible degree.

By negligible, we mean evaporation from the sample droplet under the oil, which is undetectable for a period of at least one day, preferably for a period of at least two days, or 5 days or a week. Preferably, · negligible evaporation means loss of water from the solution, which is small enough to not be detected for a period of at least two weeks or 1 month or 2 months or 3 months. Evaporation from the sample droplet can be detected by any suitable means, including measuring its size or drying it.

The oil can be any suitable liquid oil. It is preferable for the oil to be less dense than that of the liquid which can be aspirated through the tip of the pipette,

capable of dosing between 0.1 μ 0.1 and 1 μΐ. This is so due to the fact that it is preferable oil to cover the crystallization drop. If the oil is denser than the liquid in the crystallization drop, it would not be able to cover the drop. Preferably the oil has a density of about 0.84 g cm ³ . It is also preferred that the oil used in the system has the properties of an inert coating and does not interact with the crystallization samples, for example one that does not cause crystallization. Therefore, it is preferable that the oil consists of / or contains paraffins. Even more preferably, the oil consists of a paraffin for glowing, a purified mixture of liquid saturated hydrocarbons obtained from oil.

According to one embodiment of this aspect of the invention, it is sufficient for the oil added in step (iii) to retard the evaporation from the solution without negligibly making it, and then the method is supplemented by additional step (iv) in which more oil is added in the oil layer, either in quantity or type, to make the evaporation of the solution negligible.

By first adding oil in an amount that slows the evaporation but does not stop it, the evaporation can be slowed down gradually - before finally stopping. This provides additional freedom to control the speed or mode of evaporation of the solution.

The biological macromolecule can be any biological macromolecule, including amino acid, complex polysaccharide and virus. Preferably, the biological macromolecules are polypeptides. The polypeptides contain at least one chain of amino acid residues that are covalently linked via a peptide bond. The polypeptide chain may have any number of amino acid residues, preferably at least two, more preferably at least 100, 500, 1000 or 2000. The polypeptide chain may have more than 2000 residues. The polypeptide may have particular or artificial residues, it may include non-peptide bonds, such as disulfide bonds. The residues may also be modified, for example, by including phosphate groups or saccharide chains (eg ologosaccharides) or oily units. The polypeptide may have more than one chain (for example, two chains linked by a sulfur disulfide bond in the cysteine side chain), and also include inorganic or organic co-factors or groups. The term polypeptides includes such modifications and additives.

Preferably, the concentration of the solution of biological macromolecules under oil in step (i) corresponds to a saturation or metastable state. In other words, the concentration of the solution is outside the area of germ formation according to the phase diagram of this solution.

Furthermore, it is preferable if the oil in step (iii) is added only when the concentration of the solution of the biological macromolecules reaches the nucleation zone according to the phase diagram of the solution. An example of a crystallization phase diagram is shown in the paper: Chayen et al (1996) Reviews of Biophysica 29: 227-278.

It is preferable if the oil layer in step (i) permits evaporation from the solution of biological macromolecules due to insufficient thickness of the oil layer, which cannot make the evaporation negligible. Typically, the thickness of the oily layer at which evaporation ceases to be negligible is below

3.5 mm. The thickness of the oily layer is preferably between 0.7 and 1.2 mm.

It should be noted that the oil layer in step (i) may be of only one type of oil or be a mixture of at least two oils. It is preferable if the oily layer is of only one type of oil and not a mixture, since this gives the advantage of easy estimation of the evaporation kinetics and allows complete evaporation of the sample (ie when the evaporation of the solvent is negligible). Suitable layer oil in step (i) includes paraffin. Preferably, the oily layer contains paraffin and preferably paraffin is supplied by Hampton Research, CA92677-3913 USA, Catalog Number HR3-411. When using a mixture of oils, it is preferable that the mixture is of paraffin and silicone. A suitable mixture of paraffin and silicone is '65% paraffin and 35% silicone. The use of oil and a mixture of oils to allow diffusion through the layer is described by D'Arcy et al. (1996) J. Crystal Growth 168, 175-180 and Chayen (1997) J. Appl. Cryst. 30: 198202.

The thickness of the * oil layer that allows evaporation of the biological macromolecule solution lying below it may vary according to the composition of the oil layer. Therefore, the type of oil and the thickness of the oil layer will determine the rate of evaporation and therefore the speed at which the germ formation zone approaches. One skilled in the art can readily determine the thickness of the oil and the composition of the oil layer in step (i) that are most suitable for the desired rate of evaporation of the solvent from the sample or the solution of biological macromolecules.

In step (ii), the oil may be mapped using any suitable means, however, it is preferable to use an automated system for dispensing liquids. Automated fluid dispensing systems are a well-known and suitable example are the IMP AX and Oryx 6 systems manufactured by Douglas Instruments Ltd (Douglas House, Hungerford, Berks UK). It is preferable for the IMP AX system to be modified in such a way as to include at least one additional hand with a nozzle capable of pumping volumes of oil between 10 μΐ and 1 ml. Even more preferred is the Oryx 6 system.

According to preferred embodiments of this aspect of the invention, the moment at which oil is added in step (iii), i. the incubation time, in step (j), is determined by pre-screening. Briefly, suitable pre-screening would include a set of crystallization tests (i.e., a series of crystallization samples) where the oil layer is such that they allow evaporation from the test sample. The appearance of crystals will be monitored at intervals after betting on the experiment, for example, after 24 hours, 2, 4, 6 and 12 hours. If no crystals are formed in such pre-screening, or if crystals are not expected to form within 24 hours of sample placement, crystal formation can be checked every 24 hours after samples have been placed. It is known to those in the know that the stage at which crystals appear is related to both the nucleation kinetics and the osmotic parameters. A person who is experienced in crystallography can evaluate these parameters when placing a screening or optimization attempt. The appearance of crystals can be recorded by any suitable method known in the art of crystallography. For example, the crystals can be observed by eye, or using light scattering described by Rosenberger et al. (1993) J. Crystal Growth 129: 1-12. If light scattering is used, scintillation of micron-sized crystals causes a significant increase in the signal supplied by the photodiode of the light scattering apparatus; This increase indicates the formation of crystals.

The time required for the macromolecule crystals to become distinguishable by pre-screening is the incubation time of step (ii). Therefore, this is the time when oil can be added in step (iii) to reduce evaporation from the crystallization sample, preferably

to a negligible degree and thus favor growth as opposed to emergence.

A second aspect of the invention is the use of an automated system for dispensing liquids in the method of the first aspect of the invention. An automated fluid dispensing system can be any suitable fluid dispensing system, including IMP AX and Oryx 6 systems manufactured by Douglas Instruments, Hungerford, Berks UK, provided that it can dispense oil. Preferably, the dosing system can be programmed to dispense additional oil to the crystallization samples at a particular time or series of moments.

A third aspect of the invention is the use of an oil to reduce or stop the evaporation of the macromolecule solution, which solution is below a layer of oil, which oil layer is such that it allows evaporation from the solution lying below.

Thus, according to this aspect, the oil is added to the oil already present above the solution. Preferably, the solution is one from which there is already some evaporation of the solvent and the addition of oil according to this aspect of the invention aims to reduce the evaporation or to retain it so that it becomes negligible. In other words, there is a delay of at least 1 minute, preferably 10 minutes, even more preferably at least 30 minutes or 1 hour between the first addition of oil (which is such that it allows evaporation from the solution below) and the subsequent addition of oil according to this aspect of the invention.

To avoid misunderstanding: The use of oil in this aspect is not intended to prevent evaporation of the sample on which oil has not previously been applied or oil has not come into contact with the sample or has been used / previously contacted with the sample for less of 1 min, or less than 5 min, or 10, or 30 min.

Any oil capable of stopping evaporation from the macromolecule solution may be used. Preferably, this oil is paraffin. It is preferable for paraffin to stop evaporation from the macromolecule solution rather than to reduce it. By stoppage we mean that the evaporation from the solution is negligible, as defined above.

Preferably, the macromolecule solution is a solution of bio-macromolecules. More preferably, the solution is a protein solution. Furthermore, the solution preferably contains crystallization agents.

In addition, it is preferable to use the paraffin oil at a stage when the macromolecules are generated and / or crystallized in the solution or sample.

A fourth aspect of the invention provides that the paraffin oil used allows the evaporation of the macromolecule solution, which solution is below the paraffin layer, with the paraffin layer thin enough to allow evaporation from the underlying solution.

Paraffin oil may be one of the components in a mixture of oils. A suitable second component in such a mixture is silicone oil. Preferably, the paraffin is added to the macromolecule solution as a stand-alone oil rather than used as a mixture with another oil.

The thickness of the paraffin to allow the evaporation of the solution or sample lying below it may be determined by a person skilled in the art using only routine operations and may vary according to the desired rate of evaporation, according to the sample or solution. For a crystallization sample, a thickness of 0.7 -1.2 mm is typical, which allows a usable rate of evaporation from the sample. This corresponds to 0.25 ml to 0.5ml of oil on a sample of 0.7 μΐ to 2μ1 in standard crystallization aids of 5.5x8x0.9 cm. As is known to those in the know, a thickness of 4 mm or more makes the evaporation of the sample or solution below that negligible. .

The invention will now be described in more detail with the help of the following figures and examples:

Description of the figures

FIG. 1 From cloning to structures - the success of structural genomics. The biggest problem is getting good protein crystals.

The order was plotted: the number of successful attempts, respectively, for the preparatory stages indicated on the abscissa: 1 cloning of proteins, 2 - expressed proteins (protein production by DNA), 3 - solubilized, 4 - purified, 5 crystals, 6 - structures. These are the stages leading to X-ray structures (structures) and their success. The histogram shows the success of each stage towards the end goal from cloning to structure. The data are taken from the report of a structural genomic project published on the Internet (see text).

Obtaining good crystals of purified proteins impedes rapid success.

FIG. 2 Controlled Evaporation (a) Microbatch droplets under a thin layer of oil, allowing the solution to be concentrated there (symbolized by arrows).

(b) Stop evaporation / concentration by adding more oil on top to form a thicker layer above the droplets. From here on, the experiment follows the conventional path.

Example 1.

Encourage embryo formation and retardation using microbatch oil.

Initial screening was performed with a set of drops containing lysozyme listed in standard crystallization aids, for example drops containing final concentrations of 20 mg / ml lysozyme in 50 millimoles of sodium acetate buffer, pH 4.7 and 6% NaCl, dosed by IMP AX. The drops were placed under a thin (0.7 mm) layer of oil. After waiting, it was determined when crystals appeared. If the first crystals appear after, for example, 8 hours, several other experiments have been set up under the same conditions, with the addition of oil (up to a layer thickness of more than

3.5 mm) and the oil was added at different intervals. The intervals were chosen so that in one experiment the oil was added to reduce evaporation and from there stop the nucleation after half an hour, the other after 1 hour, in the third - after 2 hours, 4 hours, 6 hours. etc. It is assumed that one of these intervals will give the best result in crystal formation. In the case of lysozyme, this time was 3 hours. In other words, the crystallization experiment was conducted with drops (1-2 μ () containing the macromolecule for crystallization, the droplets were coated with a thin layer of oil and then isolated - by adding more oil after 3 hours.

In cases where crystallization required one week, the addition of the oil to stop germ formation by reducing evaporation to a negligible level occurred at intervals of 12-24 hours.

In order to avoid the trial and error approach, the time at which germ formation occurs can be determined by light scattering, as in Rosenberger et al. (1993) J. Crystal Growth

129.1-12.

The first almost effective screening and optimization experiments were conducted in 1990 as an oil-based microbatch [4]. IMP AX automated system created droplets of 0.7 - 2 μΐ from a mixture of protein and crystallizing agent. They were distributed and incubated under oil to prevent evaporation. The IMP AH system has two modes of action: one for the automated screening of numerous potential crystallization conditions and the other for the optimization of the most promising conditions (resulting from screening) using a special systematic step [9, 10]. The microbatch method paved the way for the crystallization of macromolecules and many important proteins were successfully crystallized with its aid [eg 11, 12, 13]. In its current modification, an IMP AH machine can produce about 2,000 attempts a day. Since the batch method is the simplest crystallization method from a mechanical point of view, this procedure is suitable for highly efficient crystallization. In the USA, the microbatch method is already being applied to highly efficient crystallization experiments, using a large set of syringes that dose 1536 drops with a volume of 0.4-0.5 μΐ, located in standard micro-droplet guides [6].

Evaporation kinetics control

It is well known that embryo formation is a precondition for / and a first step in the growth of crystals. Excess nucleation produces a large number of small crystals instead of a small number of usable crystals. The microbatch method provides a way to control embryo formation, initially slowing it down and then stopping it - before it becomes redundant. This is done by controlling the evaporation (and therefore concentration) of the droplets by a thin layer of oil. By increasing the thickness of the oil layer, the evaporation is subsequently stopped.

The paraffin oil most commonly used in the standard microbatch method is not completely impermeable to the aqueous solution that forms the crystallization droplets. The standard microbatch method involves the use of a layer of oil that is thick enough (4 mm, corresponding to about 8 ml in standard 5.5x8x0.9 cm microbatch crystallization aids) to make evaporation through it negligible for the duration of the experiment. However, if controlled evaporation is required, the thickness of this layer may become an active process parameter. Instead of microbatch conditions being selected according to the phase diagram immediately in the nucleation region, the conditions of insufficiency or metastability were selected and the water was allowed to evaporate slowly through a thin layer of oil (0.7-1.2 mm corresponding to 0.25-0.5 ml; FIG. 2a). Therefore, the solution reaches the embryonic region in a controlled manner. Then the thickness of the oil layer increases, so that evaporation through it becomes negligible and the experiment proceeds in the conventional butch method (Fig. 26). The critical incubation time at which oil is added to create optimal conditions for crystal growth depends on the (specificity of) the system and is obviously related to both the nucleation kinetics and the osmotic parameters. This procedure can be performed automatically by the IMP AH system with the extra hand with a nozzle to dispense oil at set intervals.

So far, this procedure has been tried with three model proteins: trypsin, lysozyme and cyanine protein. In all cases, an increase in the average size and yield of diffraction-friendly crystals was obtained, compared to the results of the standard microbatch method. The standard method yielded many microcrystals in many of the droplets, something that was almost completely overcome by the application of the controlled evaporation technique.

Delaying the diffusion of money by means of an oil barrier

A way to slow down the rate of equalization of the vapor pressure above the hanging or sitting droplets in experiments carried out by these methods [23] - and thus to create a saturation slower to avoid the above-mentioned abundance of microcrystals, is by placing a barrier of mixed paraffin oil and silicone on the tank (by the hanging or sitting drop method). Paraffin and silicone can be varied according to the need for the oil mixture. The most effective volumes were found to be 250-500 μΐ, placed on a 1 ml tank in standard Limbro (micro) containers (corresponding to a layer thickness of 1.25-2.5 mm). This method has been shown to work well with several proteins [e.g. 7, 19]. The advantage of this technique is that no change in the crystallization conditions or the method used is required. It can be used in a variety of (micro) containers Limbro, VDX, Cryschem or any other, and can be easily automated by adding an additional stage to a robot program, such as Cyberlab.

Claims (17)

Patent Claims A method for the preparation of macromolecular crystals, comprising the steps of: (i) placing a solution of macromolecules and a crystallizing agent in such a way that the solution is located or subsequently falls under a layer of oil, the layer of oil allowing evaporation of the solution through it. (ii) allow the solution to age for at least 1 minute; and then (iii) an additional oil is poured onto the oil layer, which reduces the evaporation of the solution. The method of claim 1, wherein the evaporation from the solution becomes negligible. The method according to claims 1 and 2, wherein the concentration of the macromolecular solution under oil in step (i) is saturated or metastable. A method according to any one of claims 1 to 3, wherein the oil in step (ii) is added only when the concentration of the macromolecular solution reaches the nucleation zone according to the phase diagram of the solution. A method according to any one of claims 1 to 4, wherein the thickness of the oil layer is less than 3.5 mm. The method of claim 5, wherein the thickness of the oil layer is between 0.7 and 1.2 mm. A method according to any one of claims 1 to 6, wherein the oil is paraffin oil. A method according to any one of claims 1 to 7, wherein the oil in step (ii) is dosed by an automated dosing system. Use of an automated dosing system in the method of any one of claims 1 to 8. Use according to claim 9, wherein the automated dosing system is an Oryx 6 or IMP AX system. 11. Use of oil to reduce or stop the evaporation of a solution of macromolecules, which solution is located under a layer of oil, which layer of oil is such that it allows evaporation from the solution lying below. Use according to claim 11, wherein the oil stops evaporation from the solution. 13. Use of an oil which permits evaporation from a solution of macromolecules, which solution is located under a layer of oil, which layer of oil is sufficiently thin to allow evaporation from the solution lying below. Use according to any one of claims 10 to 13, wherein the oil is paraffin oil. The use according to claim 14, wherein the paraffin is a component of a mixture of oils. A method according to any one of claims 1 to 8 or use according to any one of claims 10 to 15, wherein the macromolecule is a biological macromolecule. The method or use of claim 16, wherein the biological macromolecule is a polypeptide.