Patent Application
20060252130
The present invention concerns a process for the connection of two or more molecular substances by adapter segments which cause a directed interaction, based on the affinity of proline-rich amino acid sequences and protein domains of the type WW.
The interaction of two or more molecular substances is a frequent problem within the realm of biotechnological and pharmaceutical-medical research, development, and application. In particular, the interactions of two or more proteins or peptides as molecular substances are usually considered thereby. Such interactions are often explored as part of biochemical and cellular biological research, for instance, as with intra- and intercellular communication, signal transduction on a molecular level, or analyses of protein-protein interactions (amongst others, in the usage of two-hybrid systems and processes derived therefrom). Moreover, the association of biomolecules, particularly of two or more proteins, for in vitro synthesis of fusion proteins is of great importance for many biotechnological processes. Fusion proteins generated in such a way can be, for instance, heterobifunctional (bivalent) antibodies (so-called diabodies; see O. Perisic, P. A. Webb, P. Holliger, G. Winters & R. L. Williams, Crystal structure of a diabody, a bivalent antibody fragment, Structure 2, pp. 1217-1226, 1994), which comprise the binding domains (Fab/Fv/scFv fragments) of two distinct antibodies. If thereby both valences are, for instance, directed respectively at tumor cells or natural killer cells, then the bivalent, hybrid fusion protein can accordingly mediate an attachment of killer cells onto tumor cells. In the case of immunotoxins, antibodies are coupled with toxic substances and the cytotoxin is directed through specific antigen-antibody interaction into predefined cell types (see M. A. Ghetie & E. S. Vitetta, Recent developments in immunotoxin therapy, Curr. Opin. Immunol. 6, pp. 707-714, 1994).
With the help of fusion constructs and assemblates of various proteins, fundamentally any effectors can be combined with each other, and through appropriate interactions with antigens or other biological effects, two functions or characteristics can be achieved in a hybrid molecule. A series of examples hereof are published (J. P. McGrath, X. Cao, A. Schutz, P. Lynch, T. Ebendal, M. J. Coloma, S. L. Morrison & S. D. Putney, Bifunctional fusion between nerve growth factor and a transferrin receptor antibody, J. Neurosci. Res. 47, pp. 123-133, 1997; J. M. Betton, J. P. Jacob, M. Hofnung, J. K. Broome-Smith, Creating a bifunctional protein by insertion of beta-lactamase into the maltodextrin-binding protein, Nat. Biotechnol. 15, pp. 1276-1279, 1997; Y. Maeda, H. Ueda, T. Hara, J. Kazami, G. Kawano, E. Suzuki & T. Nagamune, expression of a bifunctional chimeric protein A-Vargula hilgendorfii luciferase in mammalian cells, Biotechniques 20, pp. 116-121, 1996; W. Wels, I. M. Harwerth, M. Zwickl, N. Hardman, B. Groner & N. E. Hynes, Construction, bacterial expression and characterization of a bifunctional single-chain antibody-phosphatase fusion protein targeted to the human erbB-2 receptor, Biotechnology (N.Y.) 10, pp. 1128-1132, 1992).
Heterobifunctional constructs are frequently produced through synthesis of fusion proteins at the gene level. This generally presupposes suitable connection elements (linkers) between both partners, as well as accessible termini of the polypeptide chains. In unfavorable cases, fusion of the partners can lead to the product of the fusion being inactive, for example because the fusion protein cannot develop a correct three-dimensional folding topology. Thus, it is often desirable that the connection of both fusion partners occurs in vitro, that is, after the separate synthesis and folding of both partners. Such a process would also allow, for instance, the fast production and analysis of various combinations of single components, without requiring new genetic constructions each time. For the fusion of these components, adapter segments are necessary, through which the process of fusion or directed association of the partners involved is isolated from their production. Furthermore, it is thereby necessary that the adapter segments (domains or peptide sequences) are locked onto the involved partners firmly without otherwise changing their specific characteristics.
In other applications, it is desired that a short-term, but strong, interaction results between two molecular species. Thereby, peptides and small protein domains play an especially important role, since in the process of recombinant production of proteins, they can be placed comparatively easily on the desired target proteins. Applications of this are, for instance, the purification of recombinantly produced proteins through specific binding segments. Often, a polyhistidine peptide segment is utilized for binding to nickel chelate columns (see P. Hengen, Purification of His-Tag fusion proteins from Escherichia coli, trends Biochem. Sci. 20, pp. 285-286, 1995), or the binding of a peptide segment known as Strep-Tag to Streptavidin (T. G. Schmidt, J. Koepke, R. Frank& A. Skerra, Molecular interaction between the Strep-tag affinity peptide and its cognate target, streptavidin, J. Mol. Biol. 255, pp. 753-766, 1996). The His-Tag process has the disadvantage, however, that the polyhistidine peptide segment can only bind to structures containing nickel ions; the connection, for instance, of two natural proteins or peptides is not possible in this way. For the connection of molecular substances, the process is hence not, or only in exceptions, suitable. In preparations that are purified in this manner, one also often finds nickel ions in the solution, which makes the system unattractive for medical-therapeutic applications. With the Strep-Tag process, the region of the binding partners that mediates binding is relatively large, so that for steric reasons it is not suitable for many connections. Additionally, avidin and streptavidin each possess four binding sites, so that a regulated formation of two different linked molecular substances in solution is very difficult.
Apart from use in the purification of proteins labeled in such a way, the immobilization of proteins on a solid, inert matrix is also of high biotechnological significance, for instance with the refolding of proteins on a matrix for the prevention of aggregation processes during folding (see G. Stempfer, B. Höll-Neugebauer & R. Rudolph, Improved refolding of an immobilized fusion protein, Nat. Biotechnol. 14, pp. 329-334, 1996), or the immobilization of an enzyme in a bioreactor. Polyionic sequences which have been used up to this time in the aforementioned process have the disadvantage, however, that their interaction is significantly disturbed by the presence of polyions, for instance DNA in the solution, or also through various solvent additives.
The purpose of the invention at hand is to provide a process for the connection of molecular substances that does not exhibit the mentioned disadvantages of the current technology.
This is accomplished in accordance with the invention through a process based upon claim 1 for the connection of two or more molecular substances with each other across adapter segments, distinguished in that
one of the molecular substances is modified in such a way that as an adapter segment it displays, in at least one area, a WW domain or a structure derived therefrom,
another molecular substance is modified in such a way that as an adapter segment it displays, in at least one area, a proline-rich sequence which binds at the WW domain or a structure derived therefrom,
and the molecular substances, through the association of WW domains or structures derived therefrom and of a proline-rich sequence, come into interaction with each other in order to achieve binding of one another.
Advantageous forms of execution emerge in the secondary claims and description.
The following figures are referred to in the description and the examples.
The linkage of two or more different molecular substances (molecular species) into one—usually heterobifunctional—fusion construction is a process of high biotechnological and pharmaceutical interest. Usually, as part of the invention-compliant application, proteins and/or peptides are used as the molecule species to be joined, since the adapter segments of the invention at hand originate from this chemical class. According to the invention, other molecular substances that possess one of the adapter segments of this invention are also usable. For instance, in compliance with the invention, a solid matrix can be loaded with a molecular substance over the specified adapter segments. Often, both substances to be joined must be stably and covalently linked with each other. Conversely, for some applications it can also be desired that the interaction between both molecular species exists only for a limited time and can be quickly dissolved again, for instance through extraneous additives. In yet other applications, a molecule species must be immobilized for a limited time, hence interact in a specific way with a matrix, for instance for the purification of a protein from a crude cellular extract in the recombinant production of a protein, or for a matrix-supported refolding of the protein. For such applications, the invention at hand is appropriate.
Compliant with the invention, a protein can, for instance, be directed into the interior of a virus-like coat for wrapping, or two or more different proteins can be joined into a chimeric protein with new characteristics, for instance as bivalent antibodies. Analogously, this interaction can be used for the immobilization of a molecular species, for instance for the separation of this substance from a mixture of substances.
In addition to connection through adapter segments, which is based on the interaction between WW domain and proline-rich sequence, covalent linkage of the molecular substances with each other can take place. This covalent linkage, through disulfide bridging via cysteines artificially introduced at a suitable spatial site in both molecular substances, can thereby lead to a durable connection between both molecular substances. Via disulfide bridging, bifunctional fusion molecules can be generated which exist stably under physiological and all ordinary solvent conditions, and are thus also useful for medical, therapeutic, diagnostic, and biotechnological processes.
Possible application forms of the invention as described above are also presented in exemplary manner in
For the connection of two or more molecular substances in compliance with the invention, the highly specific interaction of protein segments known under the term WW domain with a proline-rich peptide sequence (with a proline content of more than 50% within a short peptide succession of 2 to 6 amino acids) is exploited. These two molecular species show an unusually strong interaction with each other (KD 20 to 100 nM), when they are incubated together. The slow dissociation of the partners leads to the fact that the interaction is at first only temporarily effective. If this is unwanted, the dissociation can be prevented through the fixation of the binding partners by means of a disulfide bridge. Cysteines are artificially introduced in suitable spatial position into both adapter segments or within the region of the adapter segments. After association of the partners, the cysteine pairs can be oxidized through suitable choice of redox conditions and are, in this way, durably combined covalently with each other. The emerging hybrid fusion protein can display essential characteristics of the respectively underlying molecule species.
The WW domain is a small, globular protein domain, which usually consists of 30 to 40 amino acids (see M. Sudol, The WW Domain Binds Polyprolines and is Involved in Human Diseases, Exp. & Mol. Medicine 28, pp. 65-69, 1996), yet shorter variations are also known. WW domains display a high natural affinity to proline-rich ligands, which are bound with dissociation constants of 20 to 100 nM. The proline-rich ligands possess the minimum length necessary for binding of 5 to 15 amino acids with a proline content of more than 50% within this segment, whereby the direct interaction usually appears within a local segment of 2 to 6 amino acids (with more than 50% proline content). Natural ligands are thereby almost exclusively proteins that contain proline-rich segments in their amino acid sequence, however proline-rich peptides are also specific ligands of WW domains.
The designation WW domain derives from the observation that two conserved tryptophan residues (abbreviated WW) appear with a spacing of 20 to 22 amino acids; the second tryptophan, and a series of chiefly likewise-conserved hydrophobic amino acids, thereby form the binding pocket for the proline-rich ligands. A conserved proline is often located with a spacing of 2 amino acids after the second tryptophan. A series of different WW domain-types are known, which are presently arranged in 4 classes and which distinguish themselves from one another, particularly in reference to the preferentially-bound peptic ligands. WW domains can, in principle, compete with the (structurally unrelated) SH3-domains for the binding of proline-rich ligands, yet the ligands of the SH3-domains display deviant consensus sequences, so that proline-rich peptide ligands can be derived, which are specifically bound by WW domains. Furthermore, the binding of WW domains to proline-rich ligands is usually stronger than that of SH3 domains. The following table gives an overview of types and ligand-binding qualities of the of WW domain proteins.
*in direct proximity to a proline-rich sequence (>50% proline content)
—
The transmutation of a WW domain from type I into one of type II, along with the consequent change in the specificity with reference to the proline-rich peptides, can be achieved, for instance, through the amino acid exchanges L14W and H16G in the WW-type I-domain sequence. The structure of an agent from class I (Yes associated protein, YAP) shows that this WW domain consists of three β-strands which form a β-sheet (see M. Macias, M. Hyvonen, E. Baraldi, J. Schultz, M. Sudol, M. Saraste & Mr. Oschkinat, The Structure of the WW Domain in Complex with a Proline-Rich Peptide, Nature 382, pp. 646-649, 1996). The ligand binding pocket is formed from the second β-strand of the beta sheet with cooperation from the second conserved tryptophan.
The most important biological role of WW domains evidently exists in intracellular signal transduction. WW domains furthermore have been implicated directly or indirectly with a number of diseases, for instance inherited Liddle's syndrome, muscular dystrophy, and Alzheimers disease; thus, they are the target of a series of therapeutic strategies. Finally, WW domains play a biological role in the embryonic development of kidneys and in the intracellular life cycle of retroviruses.
As part of this invention it was found that, astonishingly, WW domains can form a stable structure (folding topology) under ordinary solvent conditions, even if they are isolated from their original molecular context and are genetically fused in or, dependent on the situation, to other proteins, for instance a viral coat protein. This applies, for instance, to a WW domain from the class of formin binding proteins with an unusually small size of only 31 amino acids, which forms a stable structure (folding topology) under these conditions. Remarkably, the introduction under favorable conditions of the WW domain, along with linker segments consisting of the amino acids serine and glycine, into external loops of proteins evidently neither disturbs their folding, nor are the binding qualities of the WW domain thereby negatively influenced. It could be shown that this also applies to variants of the WW domain, in which for instance amino acids were exchanged with cysteine at specific positions. It can also be valid for additional structures derived from WW domains, as for instance several consecutive, strung-together WW domains whose contributions to bonding either respectively add up, or in favorable cases are synergistic, shortened or extended WW domains, or even WW domains with site-directed exchanges of individual amino acids which, dependent upon the desired application, can for example, bind more strongly or weakly at proline-rich sequences than the natural protein domains. Such altered domains can be obtained, for instance, through interaction screening by means of current phage display technology.
Proteins which possess an inserted (that is to say, introduced in suitable loop regions within the polypeptide chain of the host proteins) or fused on (located respectively at the N-terminus and/or at the C-terminus of the host protein) WW domain demonstrate a high affinity to proline-rich sequences. These proline-rich sequences can thereby be fused with other proteins, peptides, or other molecular substances. Two arbitrary molecule species can therewith be brought into contact by means of the appended interaction partners (WW domain and proline-rich sequence). This association rests first of all on a hydrophobic interaction mediated by the WW domain and the proline-rich ligand. This interaction can be adjusted, however, with regard to higher specificity and more flexible use, through enlistment of ionic interactions or the introduction of covalent links between WW domain and proline-rich ligand, for example. Thus through placement of additional amino acids which are differently charged, or through point mutations in or near the adapter segments, the aggregation of the proline-rich ligands with the WW domain can be strengthened or more specifically configured. A covalent disulfide bridging of both components allows, in turn, a lasting and firm bond of the adapter segments and the molecular substances joined thereto.
A connection of more than two molecular substances with each other in compliance with the invention is also possible.
The kinetic parameters of interaction, such as the dissociation constant (kD) could be determined as part of the investigations by interaction measurements based on surface plasmon resonance measurements. With these, it can be shown that the interactions between WW domains and proline-rich peptide sequence are fundamentally suitable for the applications described in the invention at hand.
The nature of the interaction of adapter segments leads exclusively to the development of a heteromeric hybrid species, consisting of a part with WW domain and a part with a proline-rich sequence. The formation of homofunctional molecules (homodimers) can be excluded. Compared to other systems with comparable characteristics, the utilized WW domain has the advantage of being extraordinarily small and compact. Thereby it is clearly superior for many applications, for instance antigen antibody interactions, to other ligand binding domains (for example lipocalins and anticalins).
Furthermore, it could be shown that the introduction of cysteine residues at specific locations within the WW domain and in the proline-rich substrate can be used, beyond the interaction between WW domain and proline-rich sequence, to cause covalent coupling of the association partner and as a result to bring together the protein parts fused onto the adapter segments in a stable connection. Thus, a dissociation of the interaction partners can not result even under unfavorable conditions, for instance especially high or very low salt concentrations, or under physiologically extreme temperatures. For that, an exchange with cysteine is undertaken for example at position Asp8 (numbering follows the WW domain from the formin-binding protein FBP11) or alternatively at position Lys19. These positions are merely selected exemplary; the introduction of specific cysteines can also be useful and successful at other sites of the WW domain or the surroundings hereof, or in the proline-rich sequence or the surroundings hereof.
The particular advantage is that again only heterobifunctional species will be formed (heterodimers), since due to the strong interaction of proline-rich peptide and WW domain, only associates between both of these two adapter segments can at first be formed. The subsequent disulfide bridging under oxidizing conditions then leads to the directed formation of covalently bridged, heteromeric species. Due to the high local concentration (approximation) of cysteines in the associated form, the disulfide bridging can also be successful under slightly reducing conditions and can therewith take place with particular specificity. In contrast, in case of accidental disulfide bridging, that is to say without the necessary strong affinity of the adapter segments to one another (that is, in a non patent-compliant application), undesirable homodimers of both interaction partners would also be formed as byproducts under oxidizing conditions.
The procedure described in the invention at hand is suitable to attach arbitrary interaction partners together in solution (in vitro), whereby both a temporary as well as a lasting link of both partners is possible. Likewise, the procedure can be used to specifically separate proteins, peptides or other molecular substances, those of which are equipped with one of the two adapter types (WW domain or proline-rich sequence), from a mixture of substances. This takes place through reversible binding to a matrix that has bound the respective interaction partners covalently. The strong bond is effective insomuch that the molecules also adhere to the matrix under stringent solvent conditions. The process thereby allows, for instance, the fast and efficient purification of recombinant proteins from the crude cellular extract of bacteria or eukaryotic cells, on the condition that the recombinant (to be purified) molecule carries one of the two adapter segments (WW domain or proline-rich sequence) in fusion or as an insertion, while the corresponding counterpart to the adapter segment is immobilized at the fixed phase.
Likewise, this immobilization procedure is suitable to implement specific modifications or a refolding of the immobilized protein on the matrix, avoiding aggregation processes. Finally, with the invention at hand, applications are also possible in which a simple and stable immobilization of a molecular substance plays a key role, for instance in biosensors or in bioreactors (see R. S. Phadke, Biosensors and enzyme immobilized electrodes, Biosystems 27, pp. 203-206, 1992; M. Abdul-Mazid, Biocatalysis and immobilized enzyme/cell bioreactors. Promising techniques in bioreactor technology, Biotechnology (N.Y.) 11, pp. 690-695, 1993).
Apart from proteins and peptides, other substances can be used for the procedure described in the invention at hand. Thus peptide derivatives, peptide antibiotics, proteins with modified side chains such as fluorescence labels, alkylation, acetylation, disulfide mixtures with thiol-containing substances, and analogous changes can be deployed in a similar manner. Peptide or protein conjugates with carbohydrate, nucleic acid or lipid content can also be utilized in the procedure. Nucleic acids such as DNA, RNA, ribozyme, synthetic nucleic acids such as, for example, peptide nucleic acids, or hybrids thereof can likewise be coupled with an adapter segment, for example with chemical means. They are then likewise suitable to partake in an interaction with an analogous interaction partner. The only requirement is the stable attachment of one of the utilized adapter segments.
Within the realm of invention-compliant application, antibodies, antibody-analogous substances, enzymes, structural proteins, and capsomers of viruses or phages come in particular into consideration as proteins.
The insertion or attachment of proline-rich sequence or WW domain, or a structure derived therefrom, into or in a molecular substance can, in principle, take place at every site of the molecular substance, if the structure of the WW domain is not substantially influenced thereby. If applicable, it can be advantageous to implement the attachment or insertion under utilization of suitable linker segments, as described in example 1 for the protein PyVP1-WW150. In the case of insertion in proteins, it is expedient to seek out such areas of the protein structure in which no periodic secondary structural elements like α-helix or β-sheet exist. The insertion of WW domains or proline-rich sequences in protein structures takes place most favorably where turn areas or random coil areas of conventional definition exist.
The binding of both adapter segments to each other can be considered under the aspect of different physical interactions. Thus, a hydrophobic effect can dominate the interaction during stabilization of the interaction, as will be demonstrated in the following example 7. Other forms of interaction can however also contribute to binding, such as ionic interactions, ion-dipole interactions, dipole-dipole interactions, hydrogen bridge bonds, van der Waals forces, or dispersion forces. Ultimately, besides the aforementioned examples for non-covalent connections, a covalent connection of both molecular substances can also be brought about. Thereby a chemically stable atomic bond between two atoms of the interaction partners is created, preferably in the form of disulfide bridging of two participating cysteine side chains.
For immobilization of one of the adapter segments (WW domain or proline-rich sequence) the matrix can be charged, for instance, through the N-terminus of the proline-rich sequence or the WW domain (coupling through N-hydroxysuccinimide ester of the matrix) or through a thiol group of one of the cysteines contained in the proline-rich sequence or the WW domain (coupling of the matrix through iodacetamide group). As matrices, for example, agarose and agarose derivatives, agarose beads, sepharose, dextrans, carbohydrates, or similar polymer material come into consideration based on current technology.
Applications of this invention are demonstrated in the following examples, through which the extent of protection of the invention should not, however, be limited.
In the first example, a WW domain of the amino acid sequence Gly-Ser-Gly-Trp-Thr-Glu-His-Lys-Ser-Pro-Asp-Gly-Arg-Thr-Tyr-Tyr-Tyr-Asn-Thr-Glu-Thr-Lys-Gln-Ser-Thr-Trp-Glu-Lys-Pro-Asp-Asp (SEQ ID NO:23) is inserted in a specific loop of a viral coat protein. At the same time, a linker is additionally inserted before and after the WW domain, consisting of alternating Gly-Ser amino acids. In the given example, the employed viral core protein is the pentameric polyoma virus VP1 core protein in solution, which based on current technology, is capable of assembly in vitro into a virus-like coat. Based on the crystal structure of the protein it can be recognized that a loop region in the structure near to amino acid position 150 is possibly suitable for the insertion of the WW domain, since this loop region is found on the outside of the coat when assembly of the pentamer protein to a virus-like coat takes place.
The expression and purification of PyVP1-WW150 takes place as fusion protein with a C-terminally fused intein domain and a chitin binding domain (CBD) connecting to that. For this, first of all, a plasmid is constructed based on the vector pCYB2 of the IMPACT-System (New England Biolabs). Through the multiple cloning site of pCYB2, with help of the restriction sites NdeI-XmaI (New England Biolabs), a DNA-fragment, coding for a variant of the VP1 gene of mouse polyoma virus, is amplified by PCR and cloned by standard methods.
As basis for this, a polyoma virus variant is used which displays no cysteines in the sequence whatsoever; the six cysteines of the wild type protein were previously replaced with serine with the help of conventional mutation techniques. This variant of PyVP1 has the advantage that the redox conditions of the solution have no influence on the condition of the protein; it is thereby easily manageable in many applications. Additionally, by the later introduction of a cysteine in the inserted WW domain, a specific disulfide bridge of WW domain and proline-rich sequence can be achieved. As a further variation, a modification to site 249 is used; the threonine found there in the wild-type protein is replaced with cysteine. At this site in the protein, a marking with help of fluorescence dyes is, based on current technology, advantageously possible. The protected localization in the pentamer allows the marking at this site without unwanted side-effects. The variant of polyoma virus VP1 used is correctly named PyVP1-CallS-T249C, subsequently abbreviated with the term PyVP1.
For the PCR, the following oligonucleotides are used as primer: vp1NImp (5′-TAT ACA TAT GGC CCC CAA AAG AAA AAG C-3′; SEQ ID NO:24), and vp1CImp (5′-ATA TCC CGG GAG GAA ATA CAG TCT TTG TTT TTC C-3′; SEQ ID NO:25). With this PCR, the C-terminal amino acids of the wild-type VP1 protein of Gly383-Asn384 are transformed simultaneously into Pro383-Gly384, since a C-terminally localized asparagine is very unfavorable regarding the cleavage characteristics for the intein cleavage system. The point mutations named do not further influence the essential characteristics of the PyVP1 protein. The tac promotor of the pCYB2 vector provides only slight expression quantities of the fusion protein, thus the fusion construction PyVP1-intein-CBD is isolated from the pCYB2 vector through an additional PCR, and cloned in the NdeI-EcoRI sites of a highly-expressing pET-vector with T7lac promotor (plasmid pET21a, Novagen. Oligonucleotide:
The cloning of the WW domain as insertion in the external Loop of PyVP1 between the amino acid positions 148 and 149 takes place in several steps. With the oligonucleotides FBP11-WWaN (5′-ATA CTC TTC AGG CAG CGG CTG GAC AGA ACA TAA ATC ACC TGA TGG-3′; SEQ ID NO:27) and FBP11-WWaC (5′-ATA CTC TTC TAC CAC TAC CAT CAT CCG GCT TTT CCC AGG TAG ACT G-3′; SEQ ID NO:28), a PCR is implemented on a DNA-fragment which contains the formin-binding protein 11 (FBP11) from the organism Mus musculus (mouse). Among other things, a WW domain is encoded in this gene sequence. The oligonucleotides simultaneously insert a short linker sequence of 5 amino acids apiece, consisting of alternating glycine-serine amino acids. A second PCR on the previously described vector amplifies the N-terminal fragment of PyVP1 between amino acids 1 and 148 with help of the oligonucleotide vp1NImp (see above) and vp1-150-WWaC (5′-ATA CTC TTC AGG TAG CGG CGT AAA CAC AAA AGG AAT TTC CAC TCC AG-3′; SEQ ID NO:29). Finally, a third PCR also amplifies the C-terminal fragment of PyVP1 between amino acids 149 and the C-terminal end of the protein, with help of the oligonucleotide vp1-150-WWaN (5′-ATA CTC TTC AGC CGC TGC CTG TAT CTG TCG GTT TGT TGA ACC CAT G-3′; SEQ ID NO:30) and vp1CImp (see above).
All three PCR products are subsequently digested with type IIS restriction enzyme EamI 104 I (Stratagene). The N-terminal and the C-terminal fragment of PyVP1 (PCR products 2 and 3, see above) are dephosphorylated with help of alkaline phosphatase (CIP, New England Biolabs) to create a gene sequence in the following ligation step from the three prepared PCR-fragments, with the following sequence: (PyVP1-N-terminus)-WW domain fragment-(PyVP1-C-terminus). A PCR afterwards with the oligonucleotides vp1NImp and vp1CImp (see above) amplifies the ligation product of the three fragments, subsequently abbreviated with the term PyVP1-WW150. The PCR product can then be cloned by means of the standard method in the vector pCR-blunt (Invitrogen). After cutting out the cloned fragment PyVP1-WW150 with help of the restriction enzyme Nde 1-Sma, the final cloning in the previously described plasmid pET21a can subsequently take place.
The lastly generated vector allows the expression of the fusion protein (PyVP1-WW150)-intein-CBD with help of highly expressive T7lac promotors in E. coli BL21 (DE3) cells (mfr.: Novagen). For this purpose, transformed cells in 5 l.-Erlenmeyer flasks that contain 2 l. LB medium each, are cultivated at 37° C., until the OD600 of the culture reaches 2.0 to 2.5. Protein expression is induced by 1 mM IPTG in the medium. The cultures are thereafter incubated at 15° C. for a further 20 hours; the low temperature minimizes the cleavage of the intein part in the fusion protein under in vivo conditions. The cells are harvested by centrifugation, dissolved in 70 ml resuspension buffer (20 mM HEPES, 1 mM EDTA, 100 mM NaCl, 5% (w/v) Glycerol, pH 8.0), and lysed through high pressure homogenization. After centrifugation of the crude extract for 60 min. at 48,000 G, a clear cellular extract is obtained. This extract is applied with a flow rate of 0.5 ml/min at a temperature of 10° C. on a 10 ml chitin affinity column (New England Biolabs). The column is subsequently washed with 3 column volumes of the resuspension buffer, 15 column volumes of a wash buffer with high ionic strength (20 mM HEPES, 1 mM EDTA, 2 M NaCl, 5% (w/v) glycerol, pH 8.0) and again 3 column volumes of the resuspension buffer; thereby all undesirable E. coli host proteins are removed from the chitin matrix.
The cleavage of the PyVP1-WW150 monomer from the fusion protein by means of self-cleaving intein activity is induced in resuspension buffer through a pulse (3 column volumes) with respectively 50 mM dithiothreitol (DTT), 50 mM hydroxylamine, or 30 mM DTT along with 30 mM hydroxylamine. For this, the loaded chitin matrix is incubated with one of the indicated solutions for 14 hours at 10° C. The PyVP1-WW150 protein is thereby completely released and can be separated, by means of column chromatographic standard methods, from the chitin matrix and the remaining elements of the fusion protein adhering to the matrix. For that purpose, a linear salt gradient is appropriately used with a concentration between 0.1 and 2.0 M NaCl. The regeneration of the chitin matrix takes place through washing of the chitin material with 3 columns volumes of an SDS-containing buffer (1% SDS (w/v) in resuspension buffer) according to manufacturer's instructions.
The PyVP1-WW150 protein is, in the described process, expressed as a soluble pentamer and is native.
The example shows that surprisingly, the WW domain can be inserted with correct folding under suitable conditions in loop regions of protein structures, without substantially disturbing their native structure. The PyVP1-WW150 protein, pentameric in solution, contains the native WW domain inserted in the polypeptide chain and presents these, after assembly, on the outside of the virus-like coat (see Example 2).
By means of example 1, a protein (PyVP1-WW150) can be produced which artificially has a WW domain inserted. The binding qualities of PyVP1-WW150 in relation to proline-rich ligands can be characterized with various processes. An advantageous method for this is provided by surface plasmon resonance; in the given example, the instrument Biacore X (Biacore AB) is used. A synthetic peptide of the sequence Cys-Ser-Gly-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro (SEQ ID NO:15) is coupled, according to manufacturer's instructions via a thiol or amino coupling, to a type CM5 sensor chip. In the process, initially a quantity of 80 resonance units of the indicated peptide (RU) is immobilized on the surface. The following measurements always take place at 25° C. and a flow rate of 20 μl/min.
Binding studies of PyVP1-WW150 at the sensor chip are carried out with immobilized proline-rich peptide under various solvent conditions. The first measurement takes place under standard solvent conditions, with 10 mM HEPES, 1 mM EDTA, 150 mM NaCl, pH 7.2. The protein concentration of PyVP1-WW150 is varied from 5 to 50 nM. In
To test the binding under conditions of physiological ionic strength, Dulbecco's PBS (Gibco) is used as solvent for the second measurement. The remaining experimental conditions are selected analogously to the first experiment previously described.
For determination of the specificity of the binding, Dulbecco's MEM medium with 10% FCS (fetal calf serum, Gibco) is used as running buffer in a third measurement. FCS here represents a model system for a mixture of different proteins and other substances that are relevant in biological systems.
In summary, these three analyses, with help of Biacore technology and a sensor surface with immobilized proline-rich peptide, show that binding between molecular substances that contain a WW domain and proline-rich ligands takes place with high affinity and specificity. The interaction is thereby reversible and not essentially dependent on the chosen solvent conditions. The dissociation takes place relatively slowly compared to the association; the dissociation constant lies at 20 nM.
A further method for the characterization of binding qualities is a reversible immobilization of the WW domain on an inert matrix. For this, a synthetic proline-rich peptide (sequence Cys-Ser-Gly-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro; SEQ ID NO:15) is connected through a thiol coupling to SulfoLink column material (Pierce) according to manufacturer's instructions. A chromatography column is packed with the matrix modified in this way. This permits a loading of the samples on the matrix and elution with bound proteins under different conditions. The PyVP1-WW150 protein purified as in example 1 is applied to the column (solvent 10 mM HEPES, 1 mM EDTA, 150 mM NaCl, 5% Glycerin, pH 7.2). As evident from
This experiment shows that the PyVP1-WW150 protein is able to reversibly bind to a matrix that carries a proline-rich peptide. An temporary immobilization can thereby take place. A detachment of the protein from the matrix is possible through use of additives in the running buffer.
In a further experiment, the binding of a fluorescence-marked peptide with a proline-rich sequence to the surface (exterior) of virus-like capsids is explored. The assembly of the protein takes place in analogy to conditions already described, based on current technology (see Salunke, Caspar & Garcea, Polymorphism in the assembly of polyoma virus capsid protein VP1, Biophys. J. 56, pp. 887-900, 1989). The virus-like capsids are obtained after dialysis of the protein against 10 mM HEPES, 50 mM NaCl, 0,5 mM CaCl2, 5% Glycerin, pH 7.2. The proline-rich peptide Cys-Ser-Gly-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro (SEQ ID NO:15) is labeled specifically at the N-terminal cysteine with a fluorescein-maleimide derivative (Molecular Probes) according to manufacturer's instructions. After the assembly into capsids of the virus protein variants, a tenfold molar excess of fluorescence-marked peptide is added. Through gel filtration (column TSKGel G5000PWXL, TosoHaas), virus-like capsid coats can be clearly detected and separated from free, non-assembled capsid elements as well as from surplus peptide and fluorescence dye. The peptide bound to the WW domain located on the surface of the capsid elutes in the capsid fractions and can be verified through the specific absorption of the fluorescence dye (
This example shows that the PyVP1-WW150 variant can form capsid structures (virus-like coats) under suitable conditions. These capsids are able to bind proline-rich peptide. Thus, molecular substances can be brought in a directed manner to the surface (exterior) of virus-like structures via the specific and strong interaction of WW domain and proline-rich sequence.
In this example, it is shown that through favorable positioning of adapter segments, a localization of molecular substances into the interior of viral coats, or of virus-like coats (capsids) can take place. Due to the three-dimensional structure of polyoma virus VP1 it is known, based on current technology, that the N-terminus of the protein is localized in the interior of the coat after assembly into the capsid. The first 14 amino acids of the protein are thereby possibly unnecessary, since they can not be detected in the x-ray structure of the capsids. Hence two different variants of the PyVP1 protein are produced, which contain a WW domain at the amino terminus of the native wild-type protein (variant PyVP1-3C-WW1) or carry the WW domain at an N-terminus shortened by 14 amino acids (variant PyVP1-3C-WW[N-14]), as well as a variant of the PyVP1 protein which carries a proline-rich sequence at the N-terminus (PyVP1-3C-[N-14]-PLP). The basis for these variants is a PyVP1 variant which contains the cysteines C19 and C114, and in which a specific new cysteine is additionally introduced (analogous to the variant PyVP1-CallS-T249C). This variant is hereafter abbreviated with PyVP1-3C.
First of all, an amplification of the WW domain is performed by means of PCR; the FBP11 gene of the mouse serves thereby as a template, analogous to example 1. As oligonucleotide for the PCR, 5′-AAT ATA TCA TAT GTC CAT CAT CCG GCT TTT CCC AGG TAG ACT-3′ (SEQ ID NO:31) (with NdeI interface), and 5′-TAT TAA TCA TAT GAG CGG CTG GAC AGA ACA TAA ATC ACC TGA TGG-3′ (SEQ ID NO:32) are thereby used. The PCR product obtained is subsequently cloned, through the cutting sites Nde I-Nde I introduced by means of the oligonucleotide, into the expression vector pET21a from example 1, which contains the gene for a fusion protein PyVP1-intein-CBD; at that 5′ end of the gene, a singular Nde I restriction site is found. The expressed gene product of this vector is the desired protein PyVP1-3C-WW1. Analogous thereto, cloning with a fragment of PyVP1-3C shortened by 14 amino acids (PyVP1-3C-WW[N-14]) is performed based on the standard method described in example 1. To this end, a PCR is performed on the PyVP1 genetic fragment, with 5′-GCG CGC GCA TAT GAG CAC CAA GGC TAG CCC AAG ACC CG-3′ (SEQ ID NO:33) and the oligonucleotide vp1CImp (see Example 1). The PCR product that results is digested with the restriction enzymes Nde I-Sma I, and the fragment cloned into the vector pET21a from example 1 using the standard method. Expression and purification of both proteins takes place in accordance with example 1. The purified proteins are compared to the variants PyVP1 and PyVP1-WW150 in
In an analogous manner, the manufacture and purification of a GFP variant is undertaken. GFP is a protein that displays a green fluorescence (absorption maximum at 490 nm) in the native condition. It is excellently suited for the marking of complexes and assemblies. For the manufacture of a GFP variant with a proline-rich terminal sequence, a PCR-based amplification of the GFP gene first of all is carried out, with the plasmid pEGFP-N1 (Clontech) serving as template. Simultaneously, suitable restriction sites are introduced into the PCR product. The PCR takes place by means of the oligonucleotides 5′-TTA TTT ACA TAT GGT GAG CAA GGG CGA GGA G-3′ (SEQ ID NO:34) (with Nde I-cutting site) and 5′-ATA TCT TAA GTA CAG CTC GTC CAT GCC G-3′ (SEQ ID NO:35) (with AflII cutting site). The PCR product so obtained is cloned, over the restriction site, into the vector pTIP and expressed there. This vector pTIP is a derivative of the intein purification vector documented in example 1, on basis of pET21a, with additionally introduced proline-rich sequences. The vector is constructed so that a proline-rich sequence can be selectively fused to the 5′ or 3′ ends of a gene integrated in a multiple cloning site. The proline-rich sequence thereby contains mainly Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro (SEQ ID NO:36). The manufacture and purification of the GFP-PLP protein takes place by means of chitin affinity chromatography, in accordance with the method in example 1. The successful manufacture and purification of GFP-PLP is documented in
The manufacture of the Py-VP1-3C-PLP variant takes place analogously. For manufacture of this PyVP1 variant, PyVP1-3C-[N-14] is cloned into the vector pTIP, so that the the proline-rich sequence contained in the vector is fused N-terminally to the Py-VP1-3C-[N-14].
For inspection of the functional characteristics of both PyVP1 variants with WW domain at the respective N-terminus of the proteins, both variants are incubated with proteins that contain proline-rich sequences. The PyVP1-3C-WW1 protein is incubated with the prior-produced protein GFP-PLP (molar relationship 1:6) for 10 min. (10 mM HEPES, 1 mM EDTA, 150 mM NaCl, 5% Glycerol, pH 7.2), and the capsid formation of the PyVP1 variants induced through dialysis against a buffer which contains 0.5 mM CaCl2 (see Example 4). The successful detection of capsids (
In a second experiment analogous to experiment 5, PyVP1-3C-WW[N-14] is incubated in a molar relationship of 1:10 with a proline-rich peptide that was previously fluorescence-labelled. The labeling of the peptide (Cys-Ser-Gly-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro; SEQ ID NO:15) here takes place by means of fluorescein maleimide (Molecular Probes) through a maleimide coupling of the dye to the N-terminal cysteine, according to manufacturer's instructions. Again, as in the example 5, it is shown that after assembly of the PyVP1-3C-WW[N-14] variant, PyVP1-3C-WW[N-14] is competent for assembly under suitable conditions. In addition, the protein is able to bind the proline-rich peptide and, during the assembly into a protein coat, to bring the proline-rich peptide into the interior of the capsid. This is shown with the gel filtration in
Moreover, using the variants of PyVP1-3C-[N-14]-PLP (proline-rich sequence at the N-terminus) and GFP-WW1 (WW domain at the N-terminus), an analogous assembly attempt can be carried out. This shows that a reciprocal disposition of WW domain and proline-rich ligands at the substances to be joined, that is, the placement of proline-rich sequence on the polyoma core protein (capsid) and the WW domain on the protein to be wrapped, also leads to a successful targeting of GFP into the interior of the virus-like coat.
Summing up, the experiments in examples 5 and 6 show that variants of PyVP1 with WW domain fused to the N-terminal are able to bind proline-rich sequences and to conduct these, as well as any molecular substances located on them, under suitable conditions into the interior of virus-like coats during assembly into capsids. The described process is thereby suitable to cause a directed wrapping of molecular substances in viruses or in virus-like capsids. It could likewise be shown that variants of PyVP1 with a proline-rich sequence fused to the N-terminal are able to bind WW domains and molecular substances located on them.
The investigations described in the previous examples 1 to 6 show that through the interaction of WW domain with a proline-rich peptide sequence, a temporally limited aggregation of both adapter segments can occur. A permanent bridging of interaction partners can be achieved by equipping both adapter segments with specifically-introduced cysteine amino acids, which allow, with suitable positioning, disulfide bridging after the association of the binding partners.
Through point mutations, performed in accordance with conventional processes based on current technology, individual amino acids that are not essential for the association of both adapter segments can be changed to cysteines, both within the WW domain as well as in the proline-rich sequences of the ligands. Under suitable redox (oxidizing) conditions, a specific disulfide bridging can be formed between the bound proline-rich ligands and the WW domain, each containing one or more cysteines. This bridging is thereby decisively favored through the interaction of WW domain and the ligands. The temporally limited interaction between unconnected WW domain and proline-rich ligands lasts long enough to form a covalent linkage through disulfide bridging. The interaction of both adapter segments becomes, in this way, temporally unlimited, since disulfide bridges under physiological conditions, as they exist for instance in extracellular space, are stable. If desired, the disulfide bridge between WW domain and proline-rich ligands can be eliminated again in vitro, under reducing conditions (for instance 50 mM DTT, DTE or -Mercaptoethanol); by removal of the reducing agent, a reconnection is also possible.
Based upon the variant PyVP1-WW150 of the core protein VP1 of murine polyoma virus, an aspartate amino acid (position 8 into the WW domain) is transformed through mutagenesis into a cysteine. The resulting cysteine-containing variant is subsequently named PyVP1-WW150-D8C. Through binding studies based on surface plasmon resonance, it can be shown that this variant of the WW domain binds the proline-rich ligands Cys-Ser-Gly-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro (SEQ ID NO:15), even without creation of disulfide bridging. The extent of the interaction is, however, somewhat slighter than with PyVP1-WW150. This can evidently be ascribed to the newly introduced cysteine. It can be shown that this accumulation can be strengthened through addition of 500 mM ammonium sulfate. Thereby, the hydrophobic interactions between proline-rich ligands and WW domain are presumably intensified. The intensity of the interaction can thus be modulated through the ammonium sulfate concentration in the solvent.
The formation of the disulfide bridge between proline-rich ligands and WW domain subsequently takes place under slightly oxidizing conditions. For this purpose a buffer is used which both contains ammonium sulfate and maintains defined redox conditions. The latter conditions are achieved through usage of 1 mM GSSG and 5 mM GSH in the redox buffer (50 mM Tris, pH 8.5, 1 mM EDTA, 500 mM ammonium sulfate); oxidized (GSSG) or reduced glutathione (GSH) functions thereby as a redox shuffling system for the formation of disulfide bridges (see R. Rudolph, In vitro folding of inclusion body proteins, FASEB J. 10, 49-56, 1996). The disulfide bridging is carried out at 15° C. for 24 h and completed through dialysis against 50 mM Tris, 1 mM EDTA, pH 7. Under the conditions mentioned last, no further disulfide exchange occurs; the formed disulfide bridges are stable.
Summing up, it can be said that the introduction of cysteine amino acid residues into the WW domain make possible the covalent bridging of polyproline-rich ligands which carry at least one cysteine, with the WW domain and thereby lead to a stable covalent linkage of WW domain and ligand (see
A further area of application of the invention at hand is the separation of molecular substances from mixed substances, as is typically done in the purification of proteins from crude extracts (cellular extracts). In the process, the affinity of the WW domain to proline-rich ligands is exploited to isolate proteins that contain a WW domain from a complex mixture (crude extract) of proteins (principle of affinity chromatography). For this purpose, a column is used as in example 3; the SulfoLink material (Pierce, the reactivity of the matrix with SH groups is based on the iodacetamide group at the end of a linker, consisting of 10 CH2-groups) is thereby loaded through a thiol coupling with the peptide Cys-Ser-Gly-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro (SEQ ID NO:15), according to manufacturer's instructions.
Analogously, the coupling of peptides to other matrices is possible, for instance AffiGel 10 (Biorad, the reactivity of the matrix with NH2 groups is based on the N-hydroxysuccinimide group at the end of a linker, consisting of 10 CH2-groups) through the N-terminus of the peptide. Likewise, the peptide coupling can take place at the N-terminus of the peptide to a matrix based on CH-sepharose 4B, (Sigma, the reactive group of the matrix is likewise an N-hydroxysuccinimide ester). A covalent binding of the proline-rich ligands to a carrier material also results here, which subsequently allows a purification of WW domain proteins.
The PyVP1 variant PyVP1-3C-WW1 from example 5 (WW domain at the N-terminus of the PyVP1 protein) is, analogous to the specifications from example 1, produced as a fusion protein with an intein and a chitin-binding domain ([PyVP1-3C-WW1]-intein-CBD). It is not, however, purified with the described standard method, by means of chitin-affinity chromatography. Instead, the cellular extract, after cellular breakdown and subsequent centrifugation, is applied to the previously-described column. 10 mM HEPES, 150 mM NaCl, 1 mM EDTA, 5% Glycerin, pH 8.0 serve thereby as running buffer. After application of the extract, it is washed with 10 column volumes of a buffer which contains 10 mM HEPES (pH 8.0), 1 mM EDTA, 5% Glycerin, and in addition 2 M NaCl. With this wash procedure, all non-specific adsorbed proteins and cellular components are removed from the SulfoLink matrix. Afterwards, elution of the bound fusion protein [PyVP1-3C-WW1]-intein-CBD takes place with a buffer which contains 2% SDS. As can be recognized in
In summary, this example shows that with the described system it is possible to separate and purify specific molecules from a mixture of substances (crude extract).
Using the interaction of adapter segments (WW domain and proline-rich peptide), the manufacture of bifunctional or bivalent hybrid molecules in vitro can also take place. To that end, two molecular substances are produced which, according to application, can have identical or different characteristics, and each of which always carries one of the adapter segments covalently linked. In the chosen example, the manufacture of easily-detectable dimers of the GFP protein is performed.
A variant of GFP is produced for this purpose, analogous to the production of PyVP1 and with help of the intein-based expression system from example 1, with a WW domain at the N-terminus of the GFP (GFP-WW1). First of all, a PCR on the vector pEGFP-N1 (Clontech, see Example 5) is implemented with the oligonucleotides 5′-TAT AGC TAG CGT GAG CAA GGG CGA GGA GCT GTT C-3′ (SEQ ID NO:37) and 5′-GGG AAT TAA GTA CAG CTC GTC CAT GCC G-3′ (SEQ ID NO:38). The PCR product is ligated through the cutting sites Nhe I-Sma I into the vector pET21a from example 5, which, at the 3′ end of the insertion site, contains the fusion protein of intein and chitin binding domain (CBD) described in example 1. On the 5′-side of the insertion site, the WW domain that is described in example 5 is found. Thus, a fusion product is made with the sequence WW domain-GFP-intein-CBD. The plasmid that encodes the fusion protein can be transformed into the E. coli strain BL21(DE3). Analogous to examples 1 and 3, the manufacture and purification of fusion proteins can then take place. With the described process, the protein GFP-WW1 can be produced in purified form.
A second variant of GFP is produced analogously to this, with a proline-rich segment (Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro; SEQ ID NO:36) at the C-terminus. The manufacture and purification of the GFP-PLP protein thereby takes place identically to the description for the protein in example 5.
Both GFP variants are subsequently incubated together. Both proteins are thereby brought into connection with each other through the adapter segments and as a result a GFP dimer is formed that can be distinguished from the GFP monomer through gel filtration in a TSK-PW2000XL gel filtration column (TosoHaas).
The example demonstrates that with help of the procedure described in the invention at hand, a connection of arbitrary molecular substances can take place, which carry appropriate adapter segments on the basis of WW domains or proline-rich peptide segments. Homofunctional or heterofunctional assemblies can thereby be formed.
In this example it is shown that a direct packaging of proteins (GFP-WW1) which contains a WW domain can be achieved by positioning a proline-rich sequence at a protein shell. This example therefore shows in a mirror fashion to the elaborately documented example 5 in this patent application that the WW domains rsp. proline-rich sequences used as anchoring molecules are interchangeable. The result of this experiment was already represented in a short form in the patent application (
For the production of the expression plasmid for the GFP-WW1 fusion protein the vector pyVP1-3C-WW1 described in example 5 in the patent application is used, which contains a NheI restriction site in the VP1 gene close to the 5′ end. The GFP gene was amplified by PCR using the oligonucleotides 5′-TAT AGC TAG CGT GAG CAA GGG CGA GGA GCT GTT C-3′(NheI) and 5′-GGG AAT TAA GTA CAG CTC GTC CAT GCC G-3′(SmaI), thereby introducing a NheI cutting site at the 5′ end and a SmaI cutting site at the 3′ end. The VP1 gene in the vector pyVP1-3C-WW1 is replaced by the GFP gene, thereby creating an open reading frame for the fusion protein GFP-WW1.
In an analogous way, the production of the VP1 coat protein with an N-terminally fused proline-rich sequence is performed. First, the VP1 gene from the vector pyVP1-3C-WW1 is amplified using PCR, whereby an AflIII restriction site is introduced by the oligonucleotide 5′-TAT ACT TAA GTA CAA AGG CTT GTC CAA GAC CCG C-3′ and an EcoRI restriction site is introduced using the oligonucleotide 5′-ATA TGA ATT CCA GTC ATT GAA GCT GCC ACA AGG-3′. Subsequently, the VP 1 gene is inserted into the vector pTIP described in example 5 of the patent application which results in an N-terminal fusion of the proline-rich sequence to VP1.
The production and purification of the fusion proteins described before occurs by chitin affinity chromatography in analogy to the procedure described in example 1 of the patent application.
In order to check the functional properties of the two proteins, they are incubated together and the assiciation of the proteins via their adapters is analyzed using gel filtration chromatography. For this, the GFP-WW protein is added in a fivefold molar excess to the VP1-3C-PLP protein and subsequently incubated for 10 min in a buffer in which the VP1 is contained in its pentameric form, and therefore the accessibility of the adapters is ensured (pentamer buffer 10 mM HEPES, 1 mM EDTA, 150 mM NaCl, 5% Glycerol). Subsequently, the capsid formation is induced by addition of 0.5 mM CaCl2. Gel filtration analysis (column TSK Gel G6000PWXL) demonstrate that the VP1-3C-PLP variant is assembly competent under suitable conditions. In addition, it is demonstrated that GFP-WW1 is contained in the capsid fraction (
The example shows therefore that WW domain and proline-rich sequence can be introduced in different molecular environments, respectively, and the molecular substances modified in this way can so be linked together. Additionally, in the present special case of this example one of the molecular substances is an assembly-competent capsid, so that moreover a direction of the molecular substance into a protein shell was possible.
| Number | Date | Country | Kind |
|---|---|---|---|
| 19952956.6 | Nov 1999 | DE | national |
This application is a continuation of U.S. application Ser. No. 10/129,315 filed Nov. 4, 2002, which is a US National Phase Application Under 35 U.S.C. 371 of International Application No. PCT/EP2000/010873 filed Nov. 3, 2000, which claims the priority of German Application No. 19952956.6 filed Nov. 3, 1999, and the contents of each of which are incorporated herein by reference in their entirety.
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10129315 | Nov 2002 | US |
| Child | 11446587 | Jun 2006 | US |
