WO2002008965A2 - Method for determining spatial distances in polymers - Google Patents

Abstract

The invention relates to a method for determining spatial distances in polymer molecules of a known monomer sequence or complexes of such polymer molecules, whereby a) the polymer molecules or complexes are transformed by means of mixtures of homo or hetero-bifunctional cross-linker molecules of varying length and mass and possibly comprising varying chromophores, but having the same binding properties as the reactive terminal groups, b) after reacting with the mixture of cross-links, the polymer molecules or complexes are specifically split at known splitting positions of the polymer sequences, c) the resulting fragments are measured, the mass thereof being preferably determined, d) the resulting fragments are identified especially the fragments in which the cross-linkers connect two polymer or complex residues by comparing the measurement results from step c) with the expected values of theoretical fragments, said values being determined on the basis of known data on monomer sequences, splitting positions, cross-linkers etc., e) naturally available linkages (for example disulfide bridges in proteins) are taken into account, f) the cross-linker having the shortest length of all the cross-linkers of the mixture links the two residues, said length being a measurement for the distance between the two residues.

Description

Method for determining spatial distances in polymers or complexes of polymers with the aid of mixtures of cross-linker molecules

The invention relates to a method for determining spatial distances in polymers or complexes of polymers, in particular biopolymers of known monomer sequences, using informatic and physicochemical methods, in particular using mixtures of cross-linker molecules.

The spatial structure of polymers, i.e. the relative spatial arrangement of the residues, is of great importance in a large number of applications. A well-known example is the rational design of new active pharmaceutical ingredients. You are looking for a small molecule that fits in the binding pocket of the target protein (such as a virus protein) like a key in a lock. The binding of such an active substance molecule to the target protein then ideally leads to the medically desired blocking of the latter. Apparently, a requirement of the rational

Drug designs the knowledge of the 3D structure of the protein, ie the "lock".

The conventional methods (X-ray diffraction on single crystals or nuclear magnetic resonance) for clarifying the spatial structure of polymers may be able to provide high-resolution structural models, but on the other hand have serious disadvantages:

1. A relatively large amount of the respective polymer in high-purity form is required. This is precisely the case with proteins The production and cleaning of such quantities is often a major hurdle.

2. The polymers have to be brought into a certain form, for example in the form of single crystals. This step is also difficult in many cases and requires lengthy series of tests.

On the one hand, there is great interest in 3D structures and their economic use, as described above using the example of drug design. On the other hand, conventional methods for obtaining 3D structures are comparatively slow. In the case of proteins, this is clearly reflected in the growing discrepancy between explosively growing gene databases and the slowly increasing number of dissolved protein structures. A faster method would obviously bring significant economic benefits to the user.

In fact, there are already fast methods that primarily provide distances between pairs of functional groups. These include fluorescence transfer methods and chemical cross-linking methods. Fluorescence-based methods require the presence of chromophores. In contrast, cross-linking methods are more general, since a variety of different chemical techniques are available to link different functional groups. When the groups linked by the cross linker are identified, the cross linker serves as a molecular yardstick, the length of which provides the distance between the linked groups. For identification, the fragments are (sequentially) sequenced or determined via other characteristics, such as their mass. The main limitation of cross-linking methods is that the linked Groups for which cross-linking reagents must be accessible. Despite this limitation, cross-linking methods are becoming increasingly widespread. A known application is in particular the determination of spatial neighborhoods in complexes of several macromolecules such as the ribosome (Baranov et al (1999) Nucleic Acid Res

27: 184-5). Cross-linkers with different lengths and end groups were used, always one cross-linker per experiment. A major advantage of cross-linking methods is that comparatively small amounts of polymer are required and that no complex preparation is necessary.

Cross-linkers are usually somewhat flexible, so they can take an interval of lengths. The lengths determined by cross-linking are therefore uncertain. To reduce this uncertainty, several experiments can be done with overlapping

Uncertainty intervals are carried out (Nitao and Reisler (1998) Biochemistry 37: 16704-10). Due to the high complexity of the experimental results, only experiments in which one cross-linker was used per assay have so far been carried out.

The object of the present invention is to provide a method for determining spatial distances in polymers which no longer has the disadvantages of the prior art mentioned. In particular, the experimental effort of the method should be less than with the previously known methods, better resolution should be achieved and the method should be faster than the previous methods.

This task is accomplished by a spatial determination method Distances in polymers or polymer complexes of known monomer sequence solved, in which

a) the polymer molecules or complexes are reacted with mixtures of homo- or hetero-bifunctional cross-linker molecules of different lengths, masses and optionally with different chromophores with the same binding specificities of the reactive end groups,

b) the polymer molecules or complexes after reaction with the

Mixture of the cross-linkers are specifically cleaved at known cleavage sites of the polymer sequences,

c) the resulting fragments are measured, preferably their mass is determined,

d) the resulting fragments are identified - in particular those fragments in which cross-linkers link two residues of the polymer or complex - with the aid of a comparison of the measurement results from step c) with expected values of theoretically possible fragments based on the known ones Data on monomer sequence, cleavage sites, cross-linker, etc. are determined,

e) also naturally existing links (in

Proteins, for example disulfide bridges) are taken into account,

f) the length of the cross-linker from the mixture is a measure of the distance between two residues that has the shortest length among those cross-linkers of the mixture that link both residuals.

The progress of this method compared to the prior art is mainly due to two new parts of the method: firstly in the use of a mixture of cross-linkers (step a)) and secondly in the computational evaluation (step d)) of the experimental results ,

The method according to the invention is carried out in the following steps:

1. First, a solution of the polymer is reacted with a mixture of cross-linker molecules of different lengths. The cross-linkers are chosen so that they all react approximately equally strongly with the polymer under the given conditions (pH, temperature, etc.). In addition to the cross-linkers added, other linkages can be used, for example naturally occurring cystines in proteins. To increase the information content, both a deuterated and a protonated version can be used in a mixture for each type of cross-linker; instead of deuteration, other isotope labels are also possible. This marking can be used to split up isobaric signals in the mass spectrum measured later, and thus facilitate the assignment of masses to molecular fragments. In addition, each of the cross-linkers can be mixed with a different chromophore, so that the cross-linkers can be distinguished spectroscopically.

2. The polymer is split up specifically, that is, as precisely as possible positions given by the monomer sequence. A typical example is the tryptic cleavage of polypeptides behind arginines and lysines.

3. The fragments from step 2. are separated. Various methods such as high-performance liquid chromatography, liquid chromatography coupled to mass spectrometry, affinity chromatography or mass spectrometry can be used.

4. The fragments separated in step 3 are identified with parts of the polymer. A new computational procedure is used for this. Serve as input

a) the monomer sequence of the polymer,

b) the potential cleavage points in the chain,

c) the potential add-on parts for the cross-linker,

d) the experimentally separated spectrum of the fragments,

e) potential chemical modifications of residues or cross-linkers,

f) known linkages (e.g. disulfide bridges in proteins),

g) other parameters such as the maximum Number of unrecognized cleavage points or the maximum unsharpness that is accepted when comparing calculated and measured peak positions.

The calculation algorithm is based on the following algorithm ("SearchLinks"):

1 A <- 0

2 B <- set of base fragments

3 foreach non-empty subset b of B do

4 L <- set of all potential placements of an individual

Crosslinkers for b

5 U <- set of all potential unrecognized cleavage sites for b

6 foreach consistent subset I of L do

7 foreach allowed subset u of U do

8 c <- combine b, I, u into a combination fragment

9 if c is connected then

10 P <- Amount of all peaks matching c in the measured spectrum

11 foreach peak p from P do

12 a <- combine p, c into an assignment

13 A <- A combined with a

14 enddo

15 endif

16 enddo

17 enddo

18 enddo

Notes on selected steps of the algorithm:

1) At the end, A contains all the assignments found, including all combination fragments that are compatible with the measured spectrum under the given boundary conditions. 2) For the calculation of the base fragments, all conceivable polymer pieces between the potential cleavage sites are first generated. Parts of these polymer pieces may still be connected to one another by known connections. Each connected component from the number of conceivable polymer pieces forms its own basic fragment.

3) A stack is used to generate the non-empty subsets b of B. In practice, not all possible non-empty subsets are created. For example, subsets whose mass is sufficiently far above the peak with the largest mass found in the experiment are not generated. Furthermore, base fragments that cannot bind a cross-linker are only used for expanding the current subset if they are adjacent to at least one of the base fragments in b (and can therefore be "linked" via an unrecognized cleavage site).

4) A subset I of is "consistent" if and only if each residue from b is bound to at most one cross-linker. Exceptions are the terminal residues of the polymer, which may have more than one binding site. 5) A subset u of U is "allowed" if and only if its thickness does not exceed the predetermined maximum number of unrecognized cleavage points. 3-9) If there is only one bifunctional cross-linker type, the effort associated with steps (4), (6) and (7) can be significantly reduced. In this case it is possible to draw conclusions about the possible connection of the associated combination fragment from the thickness of the quantities and the total number of those residues in b that can bind a cross-linker. Accordingly, the loops in steps (6) and (7) then only run over those combinations of cross-linker or cleavage numbers for which a coherent combination fragment c can be formed from b. The query in step (9) is then omitted. Because the position of the cross-linker and unrecognized cleavage points in this special case at the moment of As a rule, assignment is not clearly defined, all possible related configurations are subsequently enumerated for the combination fragments from the assignments found (and only for these) at the end of the algorithm.

10) In this step all combinations of possible modifications of c are considered. In particular, a mixture of cross-linkers, all of which have the same binding functionality and can e.g. differ only in their length, map to the case of a single cross linker type with a corresponding number of modifications. Cross-linkers that only reacted once can also be treated as modifications of the residuals in question.

5. From the set of combination fragments found, there is a linkage pattern which shows which residues are connected by which cross-linker or other linkages. The minimum length of a cross link between two residuals is an accurate measure of the distance between the residuals.

6. If necessary, new information can be obtained by repeating the above steps, varying the experimental conditions. For example, the mixtures of the cross-linker (other lengths and functions), cleavage reagents, etc. can be varied further using matrix-assisted laser desorption / ionization

Mass spectrometry to change the crystalline matrix, the intensity of the laser, etc.

If there are sufficient spatial distances, a spatial model of the polymer can be calculated. The intrinsic parallelism of the described method through the use of cross-linker mixtures leads to a significantly higher efficiency compared to the prior art. The use of this parallelism is made possible by the combination of high-resolution experimental techniques such as mass spectrometry with new algorithms, which are able to extract the essential information from complex measurement results.

The advantages of the method according to the invention lie in the efficient use of experimental resources: the experimental effort (time, amount of polymer used) is approximately inversely proportional to the number of components of the cross-linker mixture; For example, the use of a mixture of 10 different cross-linkers reduces the experimental effort to 1/10 compared to the prior art. Effort tends to be shifted from the more expensive experimental part to the cheaper computational part.

example

For the protein ribonuclease A (RNase) from the bovine pancreas, it should be shown by way of example how the method can be used. The 3D structure of this protein is already known (Howlin et al. 1989 Acta Cryst A 45: 851), so that the results of the process can be validated using an independent method.

The following groups can be linked by bifunctional cross-linkers to obtain distance information: The amino groups on the

n 10 lysines and at the N-terminus, the carboxylates on glutamates, aspartates and the C-terminus. In addition, the information from the 4 disulfide bridges can be used and partially non-specifically binding photo-activatable cross-linkers can be used.

For example, one can use a mixture of cross-linkers with imido esters as reactive groups (reaction with amino groups on protein) and different lengths. DMA (linker length 0.86 nm), DMS (linker length 1.10 nm) (from Pierce, Rockford, U.S.A.) and other cross-linkers with different linker lengths and the same reactive groups can be in the mixture.

Using the example of the linkage of the two lysines at positions 31 and 91 in the amino acid sequence of RNase, it is shown how the method provides an exact distance between two residues. In parallel, distances between other groups can be determined with the method according to the invention without additional experimental effort.

The above-mentioned mixture of cross-linkers is reacted with a solution of RNase (room temperature or 4 ° C., pH 8-9 borate or bicarbonate buffer, various molar ratios cross-linker / protein 1-10 / 1, variation the protein concentration 0.1 mg / ml to 2mg / ml, reaction time <30 min).

The reaction product is digested with trypsin or chymotrypsin (4 hours, pH 8, 37 or 25 ° C, molar ratio enzyme / substrate 1/50).

The digestion is separated by reverse phase high-pressure liquid chromatography using a C18 column, 0.1% TFA in a water / acetonitrile gradient of 5-65% acetonitrile. The factions are collected, dissolved in 33% acetonitrile, mixed with matrix solution (for example 10mg / ml 2,5-dihydroxybenzoic acid in 33% acetonitrile) and the peptides analyzed with matrix-assisted laser desorption / ionization mass spectrometry.

The mass spectra are analyzed using the algorithm described above and the linked or unlinked peptides are thereby identified. Tryptic peptides 11-33 and 86-91 (or 86-98 if modifications prevent proteolytic cleavage) appear, as well as chymotryptic peptides 31-35 and 80-97. This results in two peptides, each containing only one lysine, namely lysine 31 and 91, respectively. Via the mass, and if necessary using a post source decay mass spectrum, the two peptides or the molecule can be composed of both peptides and the linking cross Linkers can be identified without a doubt.

Lysines 31 and 91 are not linked by DMA or shorter cross-linkers, but they are linked by DMS and longer cross-linkers. This results in a distance between the amino groups of the two lysines of more than 1.16 nm and less than 1.40 nm (each linker length plus two bond lengths of 0.15 nm between amino-N and linker-C). So the distance is about 1.28 nm +/- 0.12 nm. In fact, there is a distance of 1.26 nm in the structure determined by X-ray crystallography, that is to say an excellent match.

Claims

claims 1. Method for determining spatial distances on polymer molecules of known monomer sequence or complexes of such polymer molecules a) the polymer molecules or complexes are reacted with mixtures of homo- or hetero-bifunctional cross-linker molecules of different length, mass and, if necessary, with different chromophores, but with the same Binding specificities of the reactive end groups, b) the polymer molecules or complexes are specifically cleaved after reaction with the mixture of the cross-linkers at known cleavage sites of the polymer sequences, c) the resulting fragments are measured, preferably their mass is determined, d) the resulting fragments are identified - in particular those fragments in which cross-linkers link two residues of the polymer or complex - with the aid of a comparison of the measurement results from step c) with expected values of theoretically possible fragments based on the known ones Data on monomer sequence, cleavage sites, cross-linker, etc. are determined, e) links which are already naturally present (in proteins, for example disulfide bridges) are also taken into account, f) the length of that cross-linker from the mixture is a measure of the distance between two residuals, which has the shortest length among the cross-linkers of the mixture that link the two residuals. 2. The method according to claim 1, characterized in that both protonated and deuterated cross-linker molecules, or cross-linker molecules differentiated by other isotope labeling, are used. 3. The method according to claim 1 or 2, characterized in that experimental methods, in particular chromatographic methods, absorption and fluorescence spectroscopic methods (ultraviolet and visible), and mass spectrometric Methods are used to separate the mixtures of the polymer fragments and to determine the input data for the computerized identification of the fragments. 4. The method according to any one of claims 1 to 3, characterized in that the differences in the measured values with cross-linkers to the measured values for the polymer or the complex without cross-linkers or other links are used in the mathematical identification of the fragments. 5. The method according to any one of claims 1 to 4, characterized in that a set of "base fragments" is constructed, consisting of the conceivable polymer sections between the potential cleavage points and possibly already known polymer linkages. 6. The method according to claim 5, characterized in that the base fragments are systematically combined to form a tree structure of composed fragments (combination fragments). 7. The method according to claim 6, characterized in that the tree structure is trimmed during the generation with the aid of experimental measured values. 8. The method according to claim 6 or 7, characterized in that a list of all possible quantities of ^' combination fragments is produced which are compatible with the measured values, and that the comparison between the measured masses and the calculated masses of the constructed combination fragments is taken into account, that a) potential cleavage sites may not be recognized by the cleavage reagent, b) cross-links occur within a combination fragment, c) the incorporation of a cross linker can prevent the cleavage before or after the affected residue, d) the polymer or the complex can be chemically modified. 9. The method according to claim 8, characterized in that from the list of combination fragments corresponding linkage patterns are derived and thus distances between residues of the polymer or complex are obtained, and that the corresponding measurement values for the polymer or the complex are taken into account without cross-linkers or other modifications. 10. The method according to any one of claims 1 to 9, characterized in that the comparison of the measurement results with the expected values of theoretically possible fragments contains the following steps: a) determination of an amount of polymer sections located between two adjacent cleavage sites, b) Determination of an amount of linked polymer sections (Base fragments), connected by links that may be present between the polymer sections from a), c) determining a set of potential links between base fragments from b), d) determination of an amount of cleavage sites that could not have been recognized by the cleavage reagent, e) if this results from basic fragments of b) when the Links of c) and links at the cleavage sites of d) are connected: assignment of this molecule (combination fragment) to the corresponding measured values, f) repetition of steps b) to e) until all possible Combo fragments have gone through. 11. The method according to any one of claims 1 to 10, characterized in that several experiments and calculations are carried out at different times with different Mixtures of cross-linkers or different cleavage reagents. 12. The method according to any one of claims 1 to 11, characterized in that with the help of the distances thus determined and independent information spatial models of the Polymers or complex can be calculated.