Patent Grant
7091028
This application is the national phase under 35 U.S.C. §371 of PCT/EP00/10878, filed Nov. 3, 2000, which claims priority to German patent application No. 19952982.5, filed Nov. 3, 1999.
The present invention relates to a procedure for packaging of molecular substances in protein shells.
The in vitro packaging of nucleic acids or other drugs in protein shells has a high impact for biotechnological usage, especially for the insertion of heterologous DNA in eukaryotic cells for studies of the cell biology, or for medical usage for therapeutic purposes within a gene therapy. Thereby often simple, recombinantly produced viral envelopes are used for in vitro-packaging, since they represent natural nucleic acid packaging units; corresponding systems are termed as virus-like gene transfer systems. They are counted to the physical gene transfer systems, in contrast to the liposomal systems (based on different formulations of mostly cationic lipids and detergents which are assembled to liposomes and related structures) and the viral vector systems. The viral vector systems for gene transfer and gene therapy are today mostly derived from retroviruses, but increasingly other types of viruses (herpes virus, adenovirus, adeno-associated virus) are used, too. A summary of the different gene transfer systemes can be found in the first TAB report “Stand und Perspektiven naturwissenschaftlicher und medizinischer Problemlösungen bei der Entwicklung gentherapeutischer Heilmethoden” of the “Büro für Technikfolgenabschätzung” at the German Parliament (TAB report No. 25, 1994).
The natural packaging of nucleic acids in viral envelopes is a complex process, in which many factors, among others also some from the host cell, interact with each other. The process is in most of the cases still poorly understood. That is why most of the in vitro-packaging systems based on it work only with comparatively low efficiency and yield. Apart from this, a size limit concerning the DNA which has to be packed is usually observed for the in vitro procedure. The low efficiency of the in vitro method in comparison to the natural packaging process at virus formation is often caused by the absence of specific host factors in the artificial in vitro system.
A special obstruction is also the length of the uncondensed DNA of usual size in aqueous solution. Under these conditions, plasmids reach (due to the repulsion of negative charge of the comparatively stiff polyphosphate backbone), at a coding length of 5 kbp, a linear (hydrodynamic) length in the range of micrometers. With typical icosaedric viral envelope diameters of 30 to 100 nm it is clear that an efficient packaging is hardly possible without DNA condensation. On the other hand, the simple addition of condensation agents to the capsid or capsomer solution does not cogently lead to improved packaging efficiency. DNA condensed in such a way tends in typical concentration ranges to the formation of so called toroids, which are ring-shaped, high-molecular aggregates, that have a diameter of several micrometers and therefore, due to their size, cannot be packed in common virus-like protein shells.
Finally, there are only few described procedures for packaging of nucleic acid in virus-like particles in vitro, that means without the usage of special packaging cell lines. A well-characterized system with respect to this is the murine Polyomavirus (cf. S. N. Slilaty & H. V. Aposhian, “Gene transfer by Polyoma-like particles assembled in a cell-free system”, Science 220, S. 725–727, 1983; J. Forstova, N. Krauzewicz, V. Sandig, J. Elliott, Z. Palkova, M. Strauss & B. E. Griffin, “Polyoma virus pseudocapsids as efficient carriers of heterologous DNA into mammalian cells”, Hum. Gene Ther. 6, S. 297–306, 1995). The pentameric coat protein VP1 of the Polyomavirus (PyVP1) can form under suitable solvent conditions (increase of the ionic strength, addition of calcium or ammonium sulphate, choice of a suitable pH value) in vitro virus-like structures, which are in structure mostly identical to the virus capsids originating under biological conditions. A loading of a capsid produced in this way can be initiated by a procedure known for a long time as the osmotic shock method (S. M. Barr, K. Keck & H. V. Aposhian, “Cell-free assembly of a Polyoma-like particle from empty capsid and DNA”, Virology 96, S. 656–659, 1979; S. N. Slilaty, K. I. Berns & H. V. Aposhian, “Polyoma-like particle: characterization of the DNA encapsidated in vitro by Polyoma empty capsids”, J. Biol. Chem. 257, S. 6571–6575, 1982). Here, nucleic acid molecules incubated with capsids are protected against nuclease digestion. Probably this happens less for a packaging into the interior of the capsid, but rather as an apposition of the nucleic acid onto the surface of the capsid. Moreover, this method of the osmotic shock is very inefficient. Recent analyses show that in contrast to former assumptions with respect to the method of the osmotic shock, indeed single nucleic acids in solution can be encapsidated statistically into the capsid (H. Braun, K. Boller, J. Lower, W. M. Bertling & A. Zimmer, “Oligonucleotide and plasmid DNA packaging into Polyoma VP1 virus-like particles expressed in Escherichia coli”, Biotechnol. Appl. Biochem. 29, S. 31–43, 1999); the efficiency and specificity of the encapsidation is low here, too, due to the lack of a directed but instead a rather statistic nature of the method in solution.
Therefore, the task of the present invention is to remove the named disadvantages of the present state of the technology.
This is solved according to the invention by a procedure according to claim 1 for inclusion of molecular substances in protein shells with the following steps:
Advantageous configurations of the method result from the subclaims and from the description.
The present invention relates to the production of transport systems, among other also suitable for gene transfer, transfer of drugs, and gene therapy, which are derived from in vitro produced protein shells. In methods according to the current state of the technology a type of coat protein or several coat protein species of—mostly icosaedric—viruses or phages are forced by suitable choice of the solvent conditions to assemble to regular structures which enclose in their interior a protected cavity. The cavity formed by the protein shell can be used for the transport of molecular substances. In the case of a conventional gene transfer applications a suitable DNA plasmid which codes for one or more genes is used as the substance which has to be transported in cells. The in vitro packaging of nucleic acids in virus-like particles, for instance in Polyomavirus particles in solution, has been analyzed for a long time (cf. S. N. Slilaty, K. I. Berns, & H. V. Aposhian, “Polyoma-like particle: characterization of the DNA encapsidated in vitro by Polyoma empty capsids”, J. Biol. Chem. 257, S. 6571–6575, 1982). Apart from viruses, phages can also be used in principle for this artificial packaging.
The directed insertion of drugs, especially of DNA-based expression plasmids, in virus-like particles is actually ineffective according to the current state of the technology. In case of the Polyomavirus coat protein VP1, for example, the current procedure for the loading of virus-like particles with DNA is the usage of the osmotic shock procedure which is inefficient and by which possibly only an external loading of the DNA onto the artificial virus-like capsids occurs. The present invention relates to a new procedure for directed packaging of molecular substances, for instance of nucleic acid, peptides, proteins or other active substances, into the protected interior of protein shells.
The method which is described in the present invention, can in principle be used for directed and highly efficient packaging of molecular substances of many different kinds into shells, for example derived from viruses or phages. The encapsidation of such molecular substances can be used for many technical or medical and therapeutical processes.
In the method according to the invention, a region of the protein shell (in this invention termed as protein shell fragment) is bound, respectively immobilized, by a first part (region) to a matrix. A second region of the protein shell fragment, to which the molecular substance can bind, is at the same time unfixed and is freely accessible for the molecular substance which has to be packaged. The protein shell fragment bound, respectively immobilized, to the matrix is now brought into contact with the molecular substance which has to be packaged. Then the separation of the loaded protein shell fragment from the matrix is induced. Before this, simultaneously, or afterwards, the assembly or complexation of the loaded protein shell fragment with other protein shell fragments occurs. The molecular substance can be packaged with this procedure into the coat in a directed way. Also, with the method according to the invention molecular substances can be packaged which tend to aggregation in solution or show other unfavourable features.
For molecular substances which have to be packed according to the invention single-stranded or double-stranded DNA, single-stranded or double-stranded RNA, peptides, peptide hormones, proteins, protein domains, glycoproteins, ribozymes, PNA (Peptide Nucleic Acid), pharmaceutical active compounds, for example with hydrophilic or hydrophobic or amphiphilic character, nucleotides, hormones, lipids, or carbohydrates can be used. Of advantage is DNA in the form of linear or circular plasmids, single-stranded or double-stranded oligonucleotides, chromosomes or chromosome fragments, or proteins in the form of antibodies, single-chain antibodies, enzymes or marker proteins, or RNA in form of antisense-RNA, ribozymes, catalytic RNA, or coding mRNA.
If needed, further additions, as for example free subunits of the coat protein, variants of it, or nucleic acid-condensation agents, can be used in the progress of the method.
In case of the packaging of DNA as a molecular substance, which is in uncondensed form too large for a packaging in protein shells, can, according to the invention, one or more condensation substances be added after the contact of the the matrix-bound protein shell fragment with the molecular substance, in order to get a more compact structure of the molecular substance.
The procedure described in the present invention for packaging of molecular substances is not limited to icosahedral viruses and phages as protein shells, but can also be applied on morphologically different shaped viruses and phages as well as on macromolecular associations with an inner cavity like proteasomes or chaperones. The following table 1 shows the viruses and phages with their respective morphology summed up which can be used according to the invention. Especially well analysed is in this case the group of icosahedral viruses.
One example for such protein shells is shown hereafter on the basis of the Polyomavirus pseudocapsid (protein shell consisting of VP1 subunits of the Polyomavirus). Among the documented examples in previous table 1 for usable viruses and phages, the SSV1 particle (Fuseolloviridae), which infects the Archaebacterium Sulfolobus shibatae, has to be stressed. This representative of the phages is hyperthermophilic due of its host specificity, consequently also resistant at high temperatures and due of that it can be also used advantageously for a large number of application in the sectors of the biotechnology and medicine. It can form a very stable protein shell, whereby the building blocks are easy to produce recombinantly. Similar representatives of thermophilic respectively hyperthermophilic phages are found for instance in the Lipothrixviridae, too. Not further classified are the thermophilic and hyperthermophilic representatives of the Bacilloviridae and the Guttaviridae, which can also be used in such processes where the stability of a protein shell (made out of phage proteins) is relevant.
According to the invention monomeric subunits or dimers or oligomers of subunits of the above-mentioned protein shells can be used as protein shell fragments.
Protein shell fragments can be modified in different ways according to the invention, as long as they are still able to assemble with other protein shell fragments to a protein shell.
The first region, with which the protein shell fragments bind to the matrix, can be for example a C-terminal, a N-terminal, or another region of the protein shell fragment, for instance a loop region. Advantageously, the first region is situated in such a way that it lies on the exterior side of the protein shell after the assembly.
The first region can be modified according to the invention in such a way that this first region shows an improved binding affinity to the matrix. These modifications can be, for example, the fusion of a peptide or a peptide domain, which has the known binding properties, to the protein shell fragment with the help of genetic methods. Such a binding-mediating molecule can also be covalently bound chemically, for example, by specific biotinylation after the production of the protein shell fragment. Examples for such modifications in the first region are the fixation of GST (Glutathione-S-Transferase) to glutathion-containing matrices, the use of a His-tag for coupling to nickel-chelate matrices, chitin binding domains in combination with chitin matrices, cellulose binding domains in combination with cellulose matrices, polyionic peptide sequences in combination with oppositely charged matrix surfaces (peptide loaded matrices or typical matrices from the area of ion-exchange chromatography), WW domains respectively SH3 domains in connection with proline-containing matrices, polyproline peptides in combination with matrix-immobilized WW domains or SH3 domains, antigenes in combination with antibody-loaded affinity matrices (for instance from immuno-affinity chromatography), lectins in combination with carbohydrate-loaded matrices, or protein A, protein M, protein G, or protein Z in combination with immobilized antibodies. The binding-mediating molecules can also be connected via linker segments to the protein shell fragments.
The second region of the protein shell fragments, to which molecular substances can bind, can be for example a C-terminal, a N-terminal, or another non-terminally situated region of the protein shell fragment, preferred a loop-region. Advantageously, the second part is located in such a way that after the assembly it is located on the inside of the protein shell.
The second region can, according to the invention, be modified in such a way that this second region shows an improved binding affinity to the molecular substance. These modifications can be, for example, a site-specific mutagenesis of a segment of the protein shell fragment, a fusion with a protein fragment, or the insertion of a protein fragment (protein domain) or a peptide. In the same way, the modification of the protein shell fragment can take place by chemical modification with molecular substances which show a suitable binding affinity as desired. Examples for the modification of the second region are the binding or insertion of substrate analogs for the binding of enzymes, the use of lecitins or lecitin-like domains for the binding of carbohydrates, the use of the bacterial protein A, protein G, protein M, or protein Z for unspecific binding of antibodies, the use of C1 complexes of the complement system for binding of antibodies of specific classes, the binding of antigens for specific binding of antibodies, the use of proline-containing peptides for the binding of WW domains or SH3 domains, the use of SH3 domains or WW domains for the binding of peptides rich in proline, the use of polyionic peptides for binding of peptides or chemical associates (polymers) with the respective opposite charge, the use of oligonucleotides for binding of complementary nucleotide sequences, as for example the use of Poly-T-Tags for binding of the Poly-A-tails of eukaryotic mRNA.
Further, functional modifications in further parts of the protein shell fragment are also possible according to the invention.
The binding of the first region of the protein shell fragment to the matrix can be reversible or irreversible. According to the invention, binding is to be understood in this context as covalent bindings, ionic interactions, van der Waals interactions or other interactions between protein shell fragment and molecular substance. In case of an irreversible binding of the protein shell fragment with the first region to the matrix, after the binding of the molecular substance to the second region of the protein shell fragment, only one region of the protein shell fragment will be separated from the matrix which contains the second region of the protein shell fragment and which can be assembled with other protein shell fragments to a protein shell.
As matrix, according to the invention, every matrix can be used to which protein shell fragments can be bound, for example a chitin matrix, sepharose matrix, dextran matrix or diethylaminoethyle matrix can be used. The matrix can be for instance a solid matrix or it can be present in the form of a gel.
The separation of the protein shell fragments or of the assembled protein shells from the matrix occurs according to known procedures. For example, a linker segment can be separated from the protein shell fragment by a specific, for instance enzymatic cleavage by proteinases, or by activation of an intein. Further possibilities, for example a change of the salt concentration (ionic strength), the pH value, the temperature, or an addition of solvent additives are at one's disposal.
According to one model of the invention one protein shell fragment, for example a capsomer subunit of the polyomavirus VP1 capsid, is reversibly immobilized via the (in this case C-terminal) region of the protein shell fragment to a solid matrix. A specially constituted—in a later example, for instance, a nucleic-acid-binding—segment at the aminoterminal end of the capsomer is not fixed hereby and is freely accessible for the molecular substance which has to be packaged. The capsomer immobilized in this way is now incubated with the nucleic acid which has to be packaged and this binds to the capsomer. Then, the separation of the nucleic acid-loaded capsomer from the matrix follows. Simultaneously or afterwards, the assembly of the nucleic acid-loaded protein component with other capsomers to a virus-like protein shell (capsid) is induced. With this, under suitable, defined molar proportions of loaded and unloaded capsomers, a maximum yield of nucleic acid-containing capsids will be reached. These proportions, for example, can be defined by the expert as follows: There is in each case one experiment carried out in which the concentration of the loaded capsomer is constant, while the concentration of the unloaded capsomers varies. As a measure for the range of variation, the stoichiometry of the capsid can be used, in case of the polyomavirus-hell which consists of 72 capsomers, the variation of the concentration can be for example in the range of 1:71 to 71:1. The evaluation of this experiment should show an optimum curve for the relative quantities of loaded and unloaded capsomeres. If it is required, free subunits of the protein or functional varieties of it can for example be inserted during the assembly. In case of DNA packaging, additionally nucleic acid condensation agents can be used.
In one model of the procedure according to the invention, prior to the packaging of the appropriate molecular substances, for example nucleic acid, into the protein shell of the virus-like capsid, one region of the not yet assembled shell (a protein shell fragment) is immobilized reversibly to a matrix: A second region of the protein shell fragment which has a sufficiently high nucleic acid binding affinity (respectively an adequate affinity to the in the respective case chosen molecular substance) is not fixed and freely accessible for the substance in solution. Then, the molecular substance (for instance a nucleic acid) can be added to the in this way immobilized protein shell fragment (capsomer), for example in the form of a usual circular plasmid which codes for a gene. This nucleic acid binds to the immobilized capsomer via the unfixed nucleic acid binding segment, for example in a stoichiometry of 1:1 (one nucleic acid binding segment per immobilized capsomer). According to the invention, binding is to be understood in this context as covalent binding, ionic interaction, van der Waals bond or other interactions between protein shell fragment and molecular substance. The nucleic acid bound in this way, respectively active agents, are now spatially separated. As a special advantage of the invention the molecular substances cannot interact respectively aggregate with each other due to this spatial separation.
In case of the usage of usual active substances like proteins, peptides, single-stranded and sufficiently short oligonucleotides, etc., the elution step can be immediately performed afterwards. In case of complex and large agents as for example large DNA plasmids, one further step can be necessary. In case of DNA, this is usually the usage of a condensation agent. For example, using histones, histon-like proteins, or adequate polycationic molecules, the condensation of the fixed DNA is induced. The disturbing formation of toroids or aggregates is avoided in this case by spatial separation and indirect binding of the nucleic acid molecules to the matrix.
For condensation of DNA, using it as a molecular substance which has to be packaged using the procedure according to the invention, any condensation agent can be used in general, so for example histones, histone-like proteins, polycations, polyarginine, polylysin, spermidine, methylized spermidine, CTAB (cetyl trimethyl ammonium bromide), cationic lipids, lipospermine, polyethylene glycol, polyethylene imin, cobalt-amine-compounds or manganese-compounds.
From simple experiments the expert can find out if the condensation agent competes with the binding of the nucleic acid to the capsomer and due of this the nucleic acid is prematurely separated from the capsomer. Analogously, this applies to all active agents that can be used.
The complete protein shell subsequently can be completely assembled around the active agent respectively the condensed nucleic acid, whereby the bound capsomer is the initiation-building-block and defined amounts of free capsomers (or variations of it) are added. The assembly optionally can already be done during the elution, consequently still in the direct environment of the matrix, or after the release of the complex consisting of capsomer and active agent and separation from the matrix. For some processes, an assembly at the matrix, still prior to the release of the capsomers, can be advantageous; this is also possible by usage according to the invention.
The capsomer can be arranged in such a way that at an interior part of the capsomer (for example the N-terminal region when using the protein PyVP1 as capsomer) a high affinity nucleic acid-binding segment is located. This can be the natural N-terminal end of the wild type protein in case of PyVP1 which has known nucleic acid-binding properties. In other viral envelopes or phage coats this can be the C-terminal end or a loop structure on the inside of a capsid, too, which can be recognized on the basis of the tertiary structure of the coat proteins. Of advantage is the use of modified segments which bind with unusual high affinity to DNA or specific other nucleic acid types. One example for this are the well analysed protamine-related sequences from the capsid protein of the hepatitis B virus p21.5 which show DNA and RNA binding properties (cf. T. Hatton, S. Zhou & D. N. Standring, “RNA- and DNA-binding activities in Hepatitis virus capsid protein: a model for their roles in viral replication”, J. Virology 66, S. 5232–5241, 1992). In the following, examples according to this are a series of N-terminally modified capsomeres are described which show such high affinity binding segments. However, the usage according to the invention should not be restricted to these examples. Rather, many other high affinity binding segments can be inserted within the usage according to the invention. For the binding of other active agents, the respective method can be used; proteins and peptides can be bound by using affinity binding partners to the capsomere structure, as it is demonstrated in the following examples. Examples for this are the avidin-biotin interaction, the interaction of WW domains respectively of SH3 domains with peptide sequences rich in proline, the interaction of oppositely charged polyionic sequences, etc.
For the invention, the use of an intein-based fusion construct in the described procedure can be used with special advantage. Here, the separation of the immobilized complexes from the matrix by change of the redox conditions in the solution takes place, for example by simple addition of Dithiothreitol (DTT) or hydroxylamine, consequently by cleavage of a covalent binding. These redox conditions can be changed easily without the presence of loaded capsomers or active substances like DNA or other nucleic acids having a negative effect on it.
Especially advantageous according to the invention is the use of an oligomer capsomer, since the binding of an oligomer to the matrix is much stronger (due to cooperative effects) than the one of single monomers.
Furthermore, the procedure offers the advantage that at the immobilized part (preferably a capsomer from a protein shell) allows a screening for optimization of the packaging properties of the system. Such a screening can lead among others to improvements of the physicochemical or solvent conditions, the DNA sequence, or the molecular binding module, which are used in the packaging procedure. Furthermore, the properties of the molecular substance as for instance its maximum size permitted for the packaging can be analysed. Such a limit in size is especially of importance for the enclosure of DNA in virus-like vector systems. The elucidation of the limit can simply occur in such a way that a heterogeneous DNA population with respect to the size of the fragments is added to the immobilized capsomer. These heterogeneous fragments can be produced enzymatically, by use of restriction enzymes or by fractions of large DNA molecules. After the usage of the described procedure according to the invention and after benzonase digestion of the DNA not packaged or peripherally attached, the packed DNA protected in the virus-like particle is analyzed on an agarose gel. The largest DNA still packed into particle marks the limit in size which can be packaged with this procedure using the respective protein shell.
An essential advantage of the procedure is that the protein shell fragment bound to the matrix can be of different chemical or physical nature than the other building blocks used for protein shell formation. This has the special advantage that a mixed assembly of the protein shell is possible. The immobilized part has the specific function of binding and internalization of the molecular substance to be packed. Further functions for the production of a vector system with nearly any properties can be anchored in the other regions of the protein shell. The mosaic-like mixed assembly thus permits the generation of heterogeneous particles, whereby the directed insertion of a substance in the inner cavity can be guaranteed by the immobilized part.
Finally, the immobilized protein region or the molecular substance to be packed can be labelled with, for instance, a fluorescent or radioactive molecule, a specific binding peptide segment (tagging), a biotinylation or other chemical labelling, or even on another way and with this—after the assembly to virus analogous protein shells—it can be identified afterwards. This has the special advantage that with this a determination of the specific packaging efficiency by analysis of the assembled protein shell is possible, the produced protein shells which have a label contain the capsomer as a part of the shell which is responsible for the packaging of the molecular substance and consequently, with the choice of suitable conditions, with high probability also the molecular substance itself. In contrast, empty protein shells, that means shells which have formed spontaneously in the solution from the remaining parts of the protein shell subunits added and which exclude the molecular substance which has to be packed, and loaded shells can be distinguished. With this special labelling (if necessary also in combination with a special binding segment or a biotinylation), a selection of protein shells loaded with the molecular substance from empty protein shells is possible, and with that the production of homogeneous populations of virus-like protein shells with packaged molecular substance in each case.
The following examples show models of the present invention without limiting the protection of it.
The examples and the description refer to the following figures:
Production of PyPV1-Capsomers with Intein/CBD-Fusion Moiety for Immobilization to a Chitin-Matrix
The expression and purification of PyVP1 occurs as fusion protein with a C-terminally fused intein domain and a chitin binding domain (CBD) added to it. For that purpose, a plasmid is generated which is based on the commercially available vector pCYB2 of the IMPACT-system (company: New England Biolabs). Via the multiple cloning site of the pCYB2 by the restriction sites NdeI—XmaI a DNA-fragment produced by PCR amplification and in regard to this digested with the same restriction enzymes is cloned which codes for the VP1 gene of the mouse polyomavirus. As basis for this, the polyomavirus variant is used which does not show any cysteine in the sequence; the six cysteine of the wildtype protein have been replaced before by serine. This variation of PyVP1 has the advantage that the redox conditions of the solution have no influence on the condition of the protein, and by that it is easier to handle in many applications. For the purpose of labelling, a new cysteine instead of a threonine is introduced into the protein at the position 249 which makes, for example, a specific labelling possible with fluorescent dyes; the variant used is thus called PyVP1-CallS-T249C or short PyVP1 in the following. For the PCR, the following oligonucleotides are used as primers: 5′-TAT ACA TAT GGC CCC CAA AAG AAA AAG C-3′ (SEQ ID NO:11), and 5′-ATA TCC CGG GAG GAA ATA CAG TCT TTG TTT TTC c-3′ (SEQ ID NO:12). In this PCR simultaneously the C-terminal amino acids of the wildtype-VP1-protein are transformed from Gly383-Asn384 into Pro383-Gly384, since a C-terminal asparagine is very disadvantageous for the intein cleavage system. The named point mutation does not influence the essential properties of the PyVP1 protein in the following. The tac-promoter of the pCYB2-vectors yields only a small amount of expression of the fusion protein, therefore the fusion construct PyVP1-intein-CBD is isolated by a further PCR from the pCYB2-vector and cloned into a highly expressing pET-vector with a T7lac-promoter (pET21a, Company: Novagen), via NdeI—EcoRI—restriction sites. Oligonucleotides: 5′-TAT ACA TAT GGC CCC CAA AAG AAA AAG C-3′ (SEQ ID NO:13), and 5′-ATA TGA ATT CCA GTC ATT GAA GCT GCC ACA AGG-3′ (SEQ ID NO:14).
The created vector permits the expression of the fusion protein PyVP1-intein-CBD with the help of the high-expressing T7lac-promoter in E. coli BL21(DE3)-cells (company: Novagen). For this, the transformed cells are cultivated in 5 1 Erlenmeyer flasks, each containing 2 1 LB-medium, at 37° C., until OD600 of the culture is between 2.0 to 2.5. For induction of the protein expression, IPTG is added up to a final concentration of 1 mM. The cultures are incubated for protein expression for further 20 hours at 15° C.; the low temperature prevents the intein part of the fusion protein from cleavage under in vivo-conditions. Then, the cells are harvested by centrifugation, solved in 70 ml resuspension buffer (20 mM HEPES, 1 mM EDTA, 100 mM NaCl, 5% (w/v) glycerol, pH 8.0) and lysed by high pressure homogenization. After centrifugation of the crude extract for 60 min at 48 000 g, a clear cell extract is obtained. It is immediately loaded on a 10 ml chitin-affinity matrix (New England Biolabs) with a flow rate of 0.5 ml/min at 10° C. Then, the column is washed with 3 column volumes of the resuspension buffer, 15 column volumes of a washing buffer with high ionic strength (20 mM HEPES, 1 mM EDTA, 2 M NaCl, 5% (w/v) glycerol, pH 8.0) and again with 3 column volumes of the resuspension buffer; by which all undesired E. coli-host proteins are removed from the chitin matrix.
At this stage of the experiment the PyVP1-protein is immobilized by the intein fusion part and the chitin-binding-domain (CBD) to the chitin matrix with high affinity and cannot be eluted from the matrix even by buffers of high ionic strength, for instance the previously mentioned washing buffer with 2 M NaCl. It turned out to be advantageous that the PyVP1 protein is a pentamer in solution which is consequently fixed as fusion protein PyVP1-intein-CBD with 5 binding moieties with high affinity to the chitin matrix. This part of the fusion does not disturb the pentamerization of the protein, but it prevents the premature assembly of the pentamers to capsids, for which a free C-terminal end of the protein is essential. Under the previously described fixed conditions, the loading of the PyVP1-pentamer via the free N-terminal end can take place (cf. the following examples 2 and 3).
Using a pulse (3 column volumes) with 50 mM dithiothreitol (DTT), 50 mM hydroxylamine, or 30 mM DTT with 30 mM hydroxylamine, respectively, in the resuspension buffer, the cleavage of the PyVP1 capsomer from the fusion protein by the self-splicing intein activity can be induced. For that purpose a loaded chitin matrix with one of the solutions mentioned is incubated at 10° C. for 14 hours. The PyVP1 is completely set free and can be eluted and separated from the chitin matrix and the other matrix-bound components of the fusion protein by column-chromatographic procedures. For this, a linear salt gradient with a concentration between 0.1 and 2.0 M NaCl is used appropriately. The regeneration of the chitin matrix takes place according to the recommendation of the manufacturer by washing of the chitin material with 3 column volumes of a SDS containing buffer (1% SDS (w/v) in resuspension buffer) at room temperature.
Binding of Plasmid DNA to a Matrix-Immobilized PyVP1 Coat Protein
It is known that PyVP1 after recombined expression, in which it is present as pentamer protein, can be assembled in vitro to a virus-like protein shell.
The association of heterologous DNA to the protein shell has been also shown by measuring the protection of the DNA against nuclease digestion. This happened of course with only little yield. Above this, a size limit of approximately 1 to 2 kbp length of protected DNA is observed.
It is supposed that the N-terminus of the PyVP1 protein shows nucleic acid binding properties. These natural binding properties are possibly not sufficient for some applications. For that reason, alternatively to the application of the wildtype-PyVP1, a strong DNA-binding peptide sequence which is derived from the DNA binding region of the capsid protein p21.5 of the hepatitis B virus (so called repeats II, III and IV; in the following termed as 234-sequence) is fused by gene technological methods to a PyVP1 shortened by 18 amino acids at the N-teminus. Therefore, a synthetic double-stranded 78-bp-oligonucleotide of the sequence 5′-ATG GCC AGC CCG CGT CGT CGT ACC CCA AGC CCA CGT CGT CGT CGT AGC CAG AGC CCG CGG TCG TCC GGT CGT AGC CAG-3′ (SEQ ID NO:15) is synthesized and inserted into the expression vector from example 1 using standard cloning procedures; with that, the first amino acids of the PyVP1 protein (sequence: MAPKRKSGVS KSETKSTK; SEQ ID NO:16) are replaced by the 234 sequence (coded sequence of the 78-bp-oligonucleotides; sequence: MASPRRRTPS PRRRRSQSPR RRRSQ; SEQ ID NO:17). The production of the 234-PyVP1 protein is analogous to the conditions described in example 1. Surprisingly, the stable and defined binding of a plasmid-DNA (pEGFP-N1, company: Clontech) to a 234-PyVP1 immobilized in this manner succeeded.
In the standard buffer with low ionic strength (20 mM HEPES, pH 8.0, 100 mM NaCl, 1 mM EDTA) the DNA does not bind to the chitin matrix and occurs in the flow-through after it is loaded on a chromographic column containing the chitin material. The loading of the chitin matrix with 234-PyVP1 is carried out according to the conditions from example 1. After washing of the column with a buffer of high ionic strength (standard buffer, see above, with a total of 2.0 M NaCl), the column is equilibrated again with the standard buffer. Then, 30 μg of the plasmid are injected. This corresponds to 10 pmol of the plasmid used; at a protein amount of 2–4 mg (10–20 nmol 234-PyVP1-pentamers) on the matrix, the plasmid is in an approximately thousand times molar shortage. The DNA completely binds to the immobilized 234-PyVP1-capsomeres under these conditions. An identical experiment can be carried out using the wildtype-PyVP1 and results in analogous yields.
For analysis the DNA can afterwards be eluted again with a linear NaCl gradient between 0.1 and 2.0 NaCl (cf.
Analogous to the described example, a RNA-binding sequence can be used. For that purpose the first repeat of this binding region (repeat 1) is used instead of the termed 234-sequence of the DNA-binding region of the capsid protein p21.5 of the hepatitis B virus, which is repetitively used one after the other in three copies (111-sequence; sequence: MARRRDRGRS RRRDRGRSRR RDRGRS; SEQ ID NO:18) and again replaces the first N-terminal amino acids of the wildtype-PyVP1-protein. For this, the synthetic, double-stranded oligonucleotide of the sequence 5′-ATG GCG CGT CGT CGT GAT CGT GGC CGT AGC CGT CGT CGT GAT CGT GGT CGT AGC CGT CGT CGT GAT CGT GGT CGT AGC-3′ (SEQ ID NO:19) is cloned into the corresponding expression vector by standard methods and the resulting 111-PyVP1-protein is produced in analogy to example 1. The repetitive sequence of the N-terminus binds RNA with high affinity. The 111-PyVP1-protein produced in the same manner as the 234-PyVP1 shows analogous properties to the 234-PyVP1 with the exception that specifically RNA instead of DNA can be packed as nucleic acid.
Co-Elution of DNA and VP1 after Cleavage from the Intein Fusion Moiety
After the nucleic acid is loaded on and it is completely bound to the immobilized capsomers, the cleavage buffer is added to the mixture described in example 2 (using 234-PyVP1) which induces the removal of the capsomer from the fusion moiety. The cleavage buffer contains the components of the standard buffer, and additionally 50 mM DTT or 50 mM hydroxylamine (cf. example 1) for separation of the intein fusion construct according to the recommendations of the manufacturer (company: New England Biolabs), the condensation agent cetyl trimethyl ammonium bromide (CTAB) in a concentration of 20 μM, and 0.5 mM CaCl2 for the induction of the assembly of the protein shell of the virus-like capsid from the capsomeres. After incubation of the mixture on the column at 15° C. overnight it is eluted by a linear NaCl gradient and the separated fractions are analyzed in three ways:
The evaluation of the experiment as well as
Transfection of Matrix-Packaged DNA in Eukaryotic Cells
The system described in example 3 is used for gene transfer of plasmid pEGFP-N1 into NIH3T3 cells. The capsids fractionated after elution from the column are added for that purpose in the presence of benzonase (5 units) for digestion of free plasmid DNA still present to the NIH3T3 cells which are seeded the day before (20 000 cells and 40 μg complex per experiment). After 48 h incubation of the cells at 37° C. and 5% CO2 the cells are washed three times with PBS, incubated with trypsin/EDTA-solution for 5 min, and then loosened from the bottom. Upon the expression of the eGFP gene the green fluorescent protein GFP is produced whose fluorescence can be measured in the fluorescence activated cell sorter (FACS). An analysis of the transfected cells by FACS shows the total amount of GFP fluorescence which is 170% of the basic fluorescence of cells not transfected. With this it is shown that the virus-like capsids produced according to example 1 and 3 at the matrix are able to efficiently transport molecular substances, as for instance nucleic acid, into cells. The components used are not toxic; in the cells the nucleic acid can be brought to expression. Therefore, a productive gene transfer takes place by the system used.
Packaging of Peptides, Proteins or Pharmaceutical Agents
Apart from the use of nucleic acids described in examples 2 and 4, also a directed packaging of other active agents, for example proteins or peptides, can take place. For that purpose, a binding site at the inside of the capsomer is modified in such a way that a fixation of the protein or peptide agent by an affinity binding to this binding site takes place.
This fixation can take place for example by ionic interactions, that means by two oppositely charged peptide parts, or by a binding with hydrogen bonds between the capsomer on its inside and the peptide or protein. Finally, even natural, known interactions like the one between SH3 domain and sequences rich in proline or the one between the protein avidin and the peptide sequence known as strep-tag can be used. For that purpose, one binding partner is provided during the production with an adequate binding segment, the active substance to be packed analogous to tis with a complementary binding segment. The fixation of the binding segments with recombinant production of the capsomer with the help of gene technological standard methods can occur especially simple when peptides, proteins or protein domains are used as binding segments. These binding segments can be produced according to standard procedures. In the same manner this is valid for the active substance to be packed with the complementary binding segment located on it. After immobilization of the capsomer to the matrix and the fixation of the peptide or protein under arrangement of the two complementary binding segments, again the steps of separation from the matrix and assembly of the capsomers to the protein shell take place, whereby the order of the two steps can vary according to the process. In this manner in analogy to the previously described nucleic acid packaging the directed packaging of other active substances, as for instance proteins or peptides, can be carried out, too.
Packaging of GFP (Green Fluorescent Protein)
In this example it is shown that by advantageous placing of adaptor segments a directing of the molecular substances, in this case proteins, into the inside of viral envelopes or virus-like coats (capsids) can take place with the help of the matrix-supported procedure. In consideration of the three-dimensional structure of polyomavirus VP1 it is known according to the state of the technology that the N-terminus of the protein is located at the inside of the shell after the assembly to a capsid. Consequently, a variant of the PyVP1 protein is produced which contains so called WW-domains at the amino-terminus of the native wildtype protein (variation PyVP1-WW1). The basis for this variation is as in example 1 a VP1 protein of the polyomavirus which has replaced all natural cysteines by serines and in which a new cysteine is additionally inserted (PyVP1-CallS-T249C, shortened PyVP1). WW domains are small protein domains which have a high affinity to proline-rich sequences; the WW-domain of the 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:20) (from the FBP11-gene product of the mouse) shows especially high affinity to the peptide sequence Pro-Pro-Leu-Pro (SEQ ID NO:21).
In first instance an amplification of a WW-domain is carried out by PCR; the FBP11-gene of the mouse serves as template. As oligonucleotides for the PCR 5′-AAT ATA TCA TAT GTC CAT CAT CCG GCT TTT CCC AGG TAG ACT-3′ (SEQ ID NO:22) (with NdeI-restriction site), and 5′-TAT TAA TCA TAT GAG CGG CTG GAC AGA ACA TAA ATC ACC TGA TGG-3′ (SEQ ID NO:23) are used. The PCR-product obtained with that is afterwards cloned by the restriction sites Nde I—Nde I inserted by the oligonucleotides into the expression vector pET21 a from example 1 which contains the gene for a fusion protein PyVP1-intein-CBD; at the 5′-end of the gene a singular Nde I restriction site is located. The expressed gene product of this vector is the desired protein PyVP1-WW1. The expression and purification of the protein takes place in accordance with example 1. Here, as in example 1 the protein is immobilized to the chitin matrix and the components not bound are removed by the washing solution.
In analogous manner, the production and purification of a GFP variant is performed. GFP is a protein which shows under native condition a green fluorescence (absorption maximum at 490 nm). Consequently, it is excellently suitable for labelling of complexes and associations.
For production of a GFP variant with a terminal sequence rich in proline first a PCR-based amplification of the GFP gene takes place, whereby the template of the plasmid pEGFP-N1 (company: Clontech) is used. Contemporary suitable restriction sites are inserted into the PCR product. The PCR takes place with the oligonucleotides 5′-TTA TTT ACA TAT GGT GAG CAA GGG CGA GGA G-3′ (SEQ ID NO:24) (with NdeI-restriction site), and 5′-ATA TCT TAA GTA CAG CTC GTC CAT GCC G-3′ (SEQ ID NO:25) (with AfIII-restriction site). The PCR-product obtained in this way is cloned by the restriction sites into the vector pTIP and is expressed there; this vector is a derivative of the documented intein-cleaning vector from example 1 basing on pET21 a with additionally inserted sequences rich in proline. The vector is constructed in such a way that optionally at the 5′- or 3′-end of a gene inserted by a multiple cloning site, a sequence rich in proline is fused. The sequence rich in proline contains mainly Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro (SEQ ID NO:26), and with that the strongly binding sequence rich in proline which is necessary for the fixing to the WW domain. The production and purification of the GFP-PPLP protein takes place by chitin affinity chromatography in agreement with the example 1. The GFP-PPLP-protein, which has on the C-terminus the sequence rich in proline Pro-Pro-Pro-Pro-Pro-Pro-Pro-Pro-Leu-Pro (SEQ ID NO:26), can be solubly produced in large amounts. The green colour of the protein solution shows at the same time that the protein can fold to its native structure.
The PyVP1-WW1-protein which is immobilized on the chitin column is at first incubated with the GFP-PPLP (in the molar relation 1:6) on the matrix for 20 min (solvent: 10 mM HEPES, 1 mM EDTA, 150 mM NaCl, 5% glycerol, pH 7.2). Then, the cleavage of the PyVP1-WW1-fusion protein from the intein domain is induced by addition of reducing agents according to example 1. The PyVP1-WW1-protein is eluted together with the GFP-protein bound via WW-polyproline-affinity from the column and immediately afterwards a solution is added with unmodified PyVP1-proteins in the stoichiometric relation of 1:1 to 1:10. The capsid formation of the PYVP1 variants is induced by dialysis against a buffer which contains 0.5 mM CaCl2. On the basis of gel filtration examinations (column TSKGel G6000PWXL, company: TosoHaas) it can be shown that a part of the native GFP-PPLP-protein (which can be identified on the basis of the specific absorption at 490 nm) is included in the capsid fraction (at elution volumes between 9 and 10 ml). This means that during the incubation of the GFP-PPLP-protein with the matrix-associated variant PyVP-WW1 on the column a binding of both proteins to each other took place, by which the GFP has been directed into the inside of the virus-like particle at the following capsid assembly.
In summary, the experiment in this example shows that variants of PyVP1 with the WW-domain fusioned N-terminally to it are able to bind sequences rich in proline under matrix-associated conditions and to direct these as well as the molecular substances connected to it during assembly into capsids under suitable conditions into the inside of virus-like shells. With that, the procedure described is suitable to cause a directed packaging of molecular substances into viruses respectively virus-like capsids. An essential advantage which has to be mentioned over the packaging in solution is the avoidance of aggregation of the substance which has to be packed, which can be used in sufficiently little concentration. By immobilization after binding to the matrix-bound PyVP1-WW1 protein the aggregation can be completely avoided and the molecular substance is concentrated on the column without aggregation coming into action.
Loading over Biotin/Avidin (Streptavidin)—Interaction
A further case of application is the use of the widespread system for production of interactions also on basis of avidin (or streptavidin respectively other derivatives) with biotin. This interaction is extraordinarily strong; molecules that interact with each other via this interaction are almost inseparably bound to each other. This property is often used for coupling of biomolecules. A disadvantage of the system is that avidin has a total of four binding sites for biotin. The simple incubation of biotinylated substances with avidin often leads to the undesired formation of (homo-)tetramers in solution or to formation of aggregates. Within the present invention this disadvantage according to the state of the available technology can be removed by matrix immobilization. For that—as in the previous examples shown—for example a modified viral coat protein is used which has one singular cysteine on one exposed site (in example 1 PyVP1-CallS-T249C, compare sequence protocol). This coat protein, as described in the previous examples, can be immobilized to a matrix. Afterwards, a reagent is added which has on one side a reactive maleimide group and on the other site a functional group like for example biotin; both these groups are separated by a suitably long spacer. Such reagents are commercially available, for example from the company Pierce. If this reagent is incubated with the immobilized PyVP1-CallS-T249C the maleimid group forms a stable and specific covalent bond with the singular cysteine of the PyVP1-CallS-T249C. In the following, avidin respectively streptavidin can be added which binds with high affinity to the biotin group. Afterwards, the capsomers can be separated according to the procedure described in example 1, and one obtains capsomers which carry avidin respectively strepavidin. These can be incubated before or after separation from the matrix with a further biotinylated substance by which one receives capsomers which are bound stably by the biotin-avidin-biotin-interaction to the molecular substance, which for example can be packaged in the inside of the capsids. By the described matrix-protected procedure it is avoided that aggregate formation or premature formation of homotetramers takes place and an uniform loading of the capsomers is induced.
| Number | Date | Country | Kind |
|---|---|---|---|
| 199 52 982 | Nov 1999 | DE | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/EP00/10878 | 11/3/2000 | WO | 00 | 10/21/2002 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO01/32852 | 5/10/2001 | WO | A |
| Number | Date | Country |
|---|---|---|
| 36 36 540 | Apr 1988 | DE |
| WO 9925324 | May 1999 | WO |
