EP1117775A2 - Tissue binding peptides, identification, production and utilization thereof - Google Patents

Abstract

The present invention relates to tissue binding peptides having amino acid sequence GEGRTVVLSF, AWCRGGILGDAM, GNLVDLVVGFDD, RVSPPKKSGGGV, GSSKWGLTXKCG, RGGVRQRSRGRR, GEGRTVVCRS or SQRWTALWQWIG and to variants of said peptides. The invention also relates to the identification of said peptides with the aid of tissues, to the production and utilization of said peptides as medicaments or diagnostic agents.

Description

 Tissue-binding peptides, their identification, production and use

The present invention relates to tissue-binding peptides, their identification with the aid of tissues, their production and use as medicaments or diagnostic agents.

The pathogenesis of arteriosclerosis is increasingly attributed to damage to the endothelium, which results in endothelial dysfunction. Endothelial dysfunction is mainly reflected in a changed expression pattern of surface proteins (selectins, adhesion molecules) but also in a change in the synthesis rates of growth hormones. So z. B. in the early stage of arteriosclerosis an increased expression of the E- and P-selectins and the cellular adhesion molecules VCAM and ICAM for the immigration of monocytes into the vessel wall. The conversion of the monocytes into macrophages leads to an altered synthesis of growth factors, which in particular stimulate smooth muscle cells to proliferate and to release mediators which have an inflammatory effect.

Drug therapy for arteriosclerosis tries to reduce the causes at an early stage, if possible before the manifestation. A main goal is to increase the plasma cholesterol level by increasing the excretion of cholesterol metabolites (bile acids) or by reducing cholesterol biosynthesis. Due to the counter-regulation initiated by the organism, drug therapy can only lower the total cholesterol level by approx. 30%. Since this method is also not suitable for inducing regression of an existing arteriosclerosis, percutaneous, transluminal angioplasty (PTA), ie widening of the stenosed section of the vessel using a balloon catheter, is the therapy today of choice. The disadvantage of this form of therapy, however, is that restenosis, ie reclosure, occurs in 30-50% of the patients treated. As an alternative, surgical treatment is therefore carried out by placing a bypass. However, this therapy also causes restenosis of the bypass vessel in about 50% of the patients within 3-6 months after the treatment.

A number of gene therapy methods have therefore been developed, in which a second catheter is used, via which a vector with a gene therapy effect is applied locally to the affected vessel section (see, for example, WO95 / 27070).

Various strategies based on modified balloon catheters have been developed for the local application of the therapeutic gene, which should allow direct application of a substance or the gene into the vessel wall. After local application with a double balloon catheter, e.g. B. Nabel, ER et al. (1990) Science 249, 1285 demonstrate transient expression of the β-galactosidase gene in transfected cells of the porcine femoral artery by liposomal and retroviral transfection. However, long-term expression (up to 5 months) could only be achieved after retroviral transfection, whereas expression in liposomally transfected cells was only detectable up to 4 weeks. In addition to the transfection time, the low transfection efficiency is a limiting factor of liposomal transfer. The transfection efficiency could be increased, for example, by using HVJ liposomes (Morishita, R. et al. (1994) Gene 149, 13). Inactivated Sendai viruses (hemagglutinating virus of Japan, HVJ) are fused with DNA-containing liposomes in this system. Due to the viral fusion and binding proteins (F and HN protein), the genetic material can be efficiently introduced directly into the cell's cytoplasm bypassing the lysosomes. Finally, there is the possibility of coupling cell-specific antibodies to liposomal (Vingerhoeds, H. et al. (1994) Immunometh. 4, 259) or viral (Wickha, TJ et al. (1996) J. Virol. 70, 6831) vectors, under the control of a strong viral promoter to be able to target the therapeutic gene only into the desired cell type.

However, the disadvantage of these methods is that, on the one hand, additional intervention is necessary and, on the other hand, the efficiency of local gene transfer in the organism is generally not very high. After all, not all vessels can be reached with a catheter, especially not the small-caliber arteries and capillaries. This is important because arteriosclerotic wall changes not only affect the large vessels, but also occur in smaller arteries.

The optimization of a tissue-specific expression is of great medical importance in particular for the therapy of vascular diseases. This not only opens up opportunities for direct vascular diseases, such as B. to treat arteriosclerosis and its complications (stenosis, restenosis, heart attack), but also all organs and tumors that can be reached via the bloodstream. Ultimately, tissue-oriented gene transfer into vascular cells can also be used to express therapeutic proteins which are not or are no longer present in sufficient amounts in the target organism due to pathological or genetic changes, e.g. B. Factor VIII deficiency, insulin deficiency etc.

An essential goal of somatic gene therapy is therefore to introduce a therapeutic gene into the target cells of the organism as specifically as possible after systemic or local administration and to express the therapeutic gene essentially only in these cells. For this purpose, the therapeutic gene should on the one hand be packaged in such a way that it is essentially not degraded by nucleases during transport through the blood vessel system, and on the other hand it should be in a form which enables cell-specific gene expression in the target cells. The specific recognition of cells is also referred to as "targeting". There are basically two possibilities for this targeting: on the one hand, antibodies against receptors on the Use cell surface, which are integrated into viral or liposomal vector systems (Vingerhoeds, H. et al (1994) supra; Wickham, TJ (1996) supra), and on the other hand peptides with high binding affinity for receptors on the cell surface can be used.

With certain antibodies, for example, certain tumor cells could be recognized and even detected in the intact tissue association using a suitable label. A major disadvantage of antibodies, however, is that for technical reasons they are usually produced in the mouse and often lead to serious side effects on patients due to an immune reaction. Even so-called humanized antibodies are not free from this problem. In individual cases, such pronounced immune complexes can occur, which drastically restrict the blood flow to the kidneys and thus lead to fatal kidney failure.

Interactions of proteins with other macromolecules such as sugars, antibodies, lipids or proteins could be studied in vitro with peptide-presenting banks. Peptide presenting libraries that are chemically produced have been developed e.g. B. von Geysen, H.M. et al. (1996) Mol. Immunol. 23, 709 or de Koster, H.S. (1995) J. Immunol. Methods 187, 179-188. In the peptide synthesis, amino acid mixtures were used on the solid phase, so that the amino acid sequence of the resulting peptides is statistical, i. H. was randomly distributed. The peptide banks could also be used in resin-bound form (de Koster, H.S. et al. (1995) supra).

With the use of peptide-presenting banks, which are based, for example, on the presentation of the peptides on the surface of phages, it has become possible to present a large number of different peptides (up to approximately 10 "peptides) at the same time. at the N-terminus of the pIII coat protein of filamentous phages (see, for example, Kay, BK et al. (1993) Gene 128, 59 or Jellis, CL et al. (1993) Gene 137, 63-68) presented peptides varied from 6 to 38 amino acids were used, among other things, for epitope screening of antibodies (Cortese, R. et al. (1994) Trends Biotechnol. 12, 262; Grihalde, ND et al. (1995) Gene 166, 187), for the characterization of protein- or peptide-binding domains ( Adey, NB & Kay, BK (1996) Gene 169, 133; Blond-Elguindi, S. et al. (1993) Cell 75, 717-728; Balass, M. et al. (1993) Proc. Natl. Acad. Sci. USA 90, 10638-10642), for the discovery of DNA-binding motifs (Wang, B. et al. (1995) J. Biol. Chem. 270, 23239), for the identification of unknown receptors which e.g. B. to viral binding proteins (Hong, SS & Boulanger, P. (1995) EMBO J. 14, 4714), for the identification of protein-binding peptides (Daniels, DA & Lane, DP (1994) J. Mol. Biol. 243, 639-652; Hong, SS & Boulanger, P. (1995) EMBO J. 14, 4714-4727, No. 19) or single-chain Fv (scFv) antigen binding fragments (Vaughan, TJ et al. (1996) Nature Biotechnology 14, 309-314) and proposed for targeting cells (Barry, MA et al. (1996) Nature Medicine 2, 299-305, No. 3).

In addition to peptide-presenting phage banks, peptide-presenting bacterial banks are known. An example of this are fusions of a combinatorial peptide library with thioredoxin, which led to the identification of peptide aptamers which recognize the cyclin-dependent kinase 2 (Colas, P. et al. (1996) Nature, 380, 548-550). Browsing a combinatorial peptide library was e.g. B. also with fusions to the C-terminus of the alpha subunit of G, (340-350) to identify rhodopsin-binding sequences (Martin, EL et al. (1996) J. Biol. Chem. 271, 361-366 , No. 1). Another example is peptide insertions in a thioredoxin-flagellin fusion protein, which is presented on the E. coli cell surface, with which protein-protein interactions can be investigated (Lu, Z. et al. (1995) Bio / Technology 13, 366- 372; U.S. Pat. No. 5,635,182).

The applications of the peptide-presenting banks described above, however, were limited to in vitro investigations with isolated and purified components (sugar, proteins, antibodies). The object of the present invention was therefore to find a means which enables targeted and as gentle as possible treatment and diagnosis of diseased organs.

One object of the present invention is therefore a tissue-binding peptide selected from a peptide with the amino acid sequence

GEGRTVVLSF, AWCRGGILGDAM, ^' GNLVDLVVGFDD, RVSPPKKSGGGV, GSSKWGLTXKCG, RGGVRQRSRGRR, GEGRTVVCRS or SQRWTALWQWIG and variants thereof.

According to the present invention, variants are understood to mean deletions, additions or substitutions of one or more amino acids, the tissue-specific binding of the peptide essentially not being changed, i. H. the peptide has retained its ability to bind better to one tissue than another.

In the case of a deletion, preferably no more than 4 amino acids, especially no more than 2 amino acids, are deleted. The deletion is preferably carried out at the N- or C-terminus of the peptide, in particular at the N-terminus.

According to the present invention, the term addition means not only the addition of one or more amino acids, for example 1-25 amino acids, but also fusion proteins of the peptide according to the invention with another C- and / or N-terminal peptide or protein. For example, a fusion of a tissue-specific peptide with a peptide that contains several histidines, for example the peptide GLFHAIAHFIHGGWHGLIHGWYG, leads not only to tissue-specific binding, but also to permeabilization of the cell membrane at a slightly acidic pH (see, e.g., Midoux, P. et al (1998) Bioconjugate Chem. 9, 260-267). A fusion protein consists, for example, of the peptide according to the invention and several lysines, for example 16 lysines in succession or the nucleocapsid protein (NCp) 7 derived from HIV-1 or NCp7 Peptides, a nucleic acid, for example a DNA containing a therapeutic gene, can be complexed to the peptide according to the invention in a particularly simple manner (see, for example, Harbottle, RP et al. (1998) Human Gene Therapy 9, 1037-1047 or Bachmann, AS et al. (1998) J. Mol. Med. 76, 126-132). According to the present invention, it is thus also possible, for example, to produce a fusion protein which contains a tissue-specific peptide, a peptide which brings about permeabilization of the cell membrane and a nucleic acid-binding peptide, which is particularly advantageous.

In the case of substitutions, conservative substitutions are preferably understood, i. H. Substitutions in which, for example, one or more amino acids with a hydrophobic side chain are replaced by one or more other amino acids with another hydrophobic side chain. Examples of amino acids with hydrophobic side chains are glycine, valine, leucine, isoleucine, methionine, phenylalanine, tryptophan or proline. Amino acids with a polar neutral residue can also be interchanged. Examples include serine, threonine, cysteine, tyrosine, asparagine or glutamine. Furthermore, amino acids with a polar acidic residue can be interchanged. For example, this includes aspartic acid or glutamic acid. Amino acids with a polar basic residue can also be interchanged, such as. B. lysine, arginine or histidine. Preferably no more than up to 4 amino acids, in particular no more than up to 2 amino acids, especially no more than one amino acid, are substituted.

According to the present invention, the term amino acid means not only the amino acids naturally occurring in proteins, but also other amino acids which can be incorporated into a peptide chain, such as, for. As hydroxylysine, hydroxyproline, ornithine, citrulline or homocysteine.

The individual amino acids of the peptide according to the invention can also be modified. For example, the peptide can be marked and / or modified to be protected. A suitable label is, for example, the exchange of the hydroxyl group of threonine or serine with the isotope ¹⁸ F, which has a half-life of 2 hours and can therefore be used well in clinical practice. Alternatively, the peptide can also be labeled with ¹⁸ F at the carboxy terminus via a fluoroethyl ^{amide bond} . The C-terminal fluoramidation takes place, for example, with fluoroethylamine and a urea derivative (e.g. TBU) for activation.

Another suitable label is the labeling of cysteine or methionine with the sulfur isotope ³⁵ S. The isotope ³⁵ S has the advantage over the isotope ¹⁸ F that it has a longer half-life.

The peptide according to the invention can also be labeled with other compounds, for example with β-galactosidase or biotin or avidin / streptavidin. However, radioactive labeling with in particular short-lived isotopes has the advantage that the structure of the peptide remains essentially unchanged and the distribution of the peptide in the organism can be analyzed in an advantageous manner, for example with the help of positron emission tomography (PET).

For labeling, it is generally advantageous if the functional side chains of the amino acids are protected beforehand. For example, the peptide on the solid phase is removed from the C- to the N-terminus using an acid-labile resin (e.g. SASRIN, a 2-methoxy-4-alkoxybenzyl alcohol resin, Frechet, JMJ et al. (1979) Polymer 20, 675) using the FastMocΘ method (Merrifield, RB (1963) J. Am. Chem. Soc. 95, 1328; Wang, SS (1973) J. Am. Chem. Soc. 95, 1328; Perkin-Elmer Applied Biosystems, Foster City , CA) synthesized. In this way, the peptide can be cleaved from the resin under conditions in which both the side chain protecting groups and the N-terminal protecting group (Fmoc, N-fluorenylmethoxycarbonyl) remain stable. The marking can then be done, for example, with ¹⁸ F. This is followed by the elimination of the protective groups and the purification via HPLC. Another object of the present invention is a nucleic acid coding for a tissue-binding peptide according to the invention. A nucleic acid is particularly preferably selected from a nucleic acid with the nucleotide sequence

GGCGAGGGGCGAACAGTCGTATTGTCGTTCG,

GCCTGGTGTCGGGGGGGTTCCCCGGGCGACGCTATG, GGAAACCTGGTGGATCTAGTTGTGGGTTTTGACGAC, CGGGTGAGTCCGCCAAAGAAGTCGGGGGGGGGGGGGGGGGGGGGGGAAGGGGGGGGAAAGGGGGTGGAAAA

GGCGAGGGGCGAACAGTCGTATGTCGTTCG or TCCCAGAGGTGGACTGCACTCTGGCAATGGATCGGG and variants thereof.

According to the present invention, the term variants is understood in particular to mean variants which code for the same peptide due to the degeneracy of the genetic code. However, the term variants also means the ribonucleic acids (RNA) corresponding to the deoxyribonucleic acids (DNA) listed. The DNA or RNA can also be modified to e.g. B. to be better protected against nuclease degradation. Suitable modifications can be found e.g. B. Uhlmann, E. & Peyman A. (1990) Chemical Reviews 90, 543-584 No. 4th

Another embodiment of the present invention relates to a vector containing one or more nucleic acids according to the invention. A vector can, for example, be an expression vector for the production of the peptides according to the invention in prokaryotic or eukaryotic cells. Examples of prokaryotic expression vectors are for expression in E. coli z. B. the vectors pGEM or pUC derivatives and for eukaryotic expression vectors for expression in Saccharomyces cerevisiae z. B. the vectors p426Met25 or p426GALl, for expression in insect cells z. B. Baculovirus vectors as disclosed in EP-B1-0 127 839 or EP-B1-0 549 721, and for expression in mammalian cells, for. B. the vectors Rc / CMV and Rc / RSV or SV40 vectors, which are all generally available.

In general, the expression vectors also contain suitable promoters for the respective host cell, e.g. B. the trp promoter for expression in E. coli (see, for example, EP-B1-0 154 133), the ADH2 promoter for expression in yeast (Rüssel et al. (1983), J. Biol. Chem. 258, 2674-2682), the baculovirus polyhedrin promoter for expression in insect cells (see e.g. EP-B1-0 127 839) or the early SV40 promoter or LTR promoters e.g. B. from MMTV (mouse mammary tumor virus; Lee et al. (1981) Nature 214, 228-232).

Examples of vectors which are active in gene therapy are virus vectors, preferably adenovirus vectors, in particular replication-deficient adenovirus vectors, or adeno-associated virus vectors, e.g. B. an adeno-associated virus vector, which consists exclusively of two inverted terminal repeat sequences (ITR).

For example, in adenoviral vectors, in particular of type 5 (sequence see Chroboczek, J. et al. (1992) Virol. 186, 280-285) and especially of subgroup C, the El gene region is generally replaced by a foreign gene with its own Promoter or replaced by the nucleic acid of the invention. By exchanging the El gene region, which is a prerequisite for the expression of the downstream adenoviral genes, a non-replication-capable adenovirus is created. These viruses can then only multiply in a cell line that replaces the missing El genes.

Replication-deficient adenoviruses are therefore generally formed by homologous recombination in the so-called 293 cell line (human embryonic kidney cell line), which has a copy of the El region stably integrated in the genome. For this purpose, the nucleic acid according to the invention is cloned into recombinant adenoviral plasmids under the control of a suitable promoter. This is followed by homologous recombination with an El-deficient adenoviral genome, such as, for example, dl327 or dell324 (adenovirus 5), in the Helper cell line 293. If recombination is successful, viral plaques are harvested. The replication-deficient viruses produced in this way are used in high titers (for example 10 ° to 10 "" plaque-forming units or plaque-forming units) for infection of the cell culture or for somatic gene therapy.

In general, the exact insertion site of the nucleic acid according to the invention into the adenoviral genome is not critical. It is e.g. It is also possible to clone the nucleic acid according to the invention in the place of the deleted E3 gene (Karlsson, S. et al. EMBO J. 5, 1986, 2377-2385). However, preferably the El region or parts thereof, e.g. the EIA or EIB region (see e.g. WO 95/00655) is replaced by the nucleic acid according to the invention, especially if the E3 region is also deleted. Also suitable are third generation adenoviral vectors, i. H. Vectors which, apart from the ITR and the packaging signal, contain no further sequences coding for viral proteins (Kochanek, S. (1996) Proc. Natl. Acad. Sci. USA 93, 5731). The recombination in the helper cell line takes place e.g. B. with the help of a helper plasmid.

However, the present invention is not limited to the adenoviral vector system, but adeno-associated virus vectors are also particularly suitable in combination with the nucleic acid according to the invention, since the transfer of the nucleic acid according to the invention into resting, differentiated cells can be carried out by AAV, which is for the treatment of vascular diseases is particularly advantageous. The two approximately 145 bp long inverted terminal repeat sequences (ITR: inverted terminal repeats; see, for example, WO 95/23867) are generally sufficient for the vector functions. They carry the signals necessary for replication, packaging and integration into the host cell genome. The ability to integrate ensures long-lasting gene expression in vivo, which in turn is particularly advantageous. Another advantage of AAV is that the virus is not pathogenic to humans and is relatively stable in vivo. The nucleic acid according to the invention is cloned into the AAV vector or parts thereof by methods known to the person skilled in the art, as described, for example, in WO 95/23867, by Chiorini, JA et al. (1995), Human Gene Therapy 6, 1531-1541 or Kotin, RM (1994), Human Gene Therapy 5, 793-801.

Vectors with gene therapy effects can also be obtained by complexing the nucleic acid according to the invention with liposomes (neutral or cationic), i.e. the nucleic acid is essentially included in the liposome, since it enables a very high transfection efficiency, in particular of cardiac muscle cells, to be achieved (see, for example, WO95 / 27070) and the nucleic acid is essentially protected against DNases. Transfection with nucleic acid-liposome complexes with the aid of Sendai viruses in the form of so-called HVJ liposomes (virosomes) is particularly advantageous, since this can further increase the transfection rate.

In lipofection, small unilamellar vesicles are made from cationic lipids by ultrasound treatment of the liposome suspension. The nucleic acid is bound ionically on the surface of the liposomes in such a ratio that a positive net charge remains and the nucleic acid is 100% complexed by the liposomes. In addition to those described by Feigner et al. (Feigner, PL et al. (1987) Proc. Natl. Acad. Sei USA 84, 7413-7414) used lipid mixtures DOTMA (1,2-dioleyloxypropyl-3-trimefhylammonium bromide) and DOPE (dioleoylphosphatidylethanolamine) meanwhile numerous new lipid formulations synthesized and tested for the efficiency of transfection of various cell lines (Behr, JP et al. (1989), Proc. Natl. Acad. Sci. USA 86, 6982-6986; Feigner, JH et al. (1994) J. Biol. Chem . 269, 2550-2561; Gao, X. & Huang, L. (1991), Biochim. Biophys. Acta 1189, 195-203). Examples of the new lipid formulations are DOTAP N- [l- (2,3-dioleoyloxy) propyl] -N, N, N-trimethylammonium ethyl sulfate or DOGS

(TRANSFECTAM; dioctadecylamidoglycyl spermine). An example of the production of DNA-liposome complexes from phosphatidylcholine, phosphatidylserine and cholesterol and their successful use in the transfection of vessel walls with the aid of Sendai viruses is described in WO95 / 27070. It is particularly advantageous if the nucleic acid-liposome complex contains nucleic acid binding proteins, for example chromosomal proteins, preferably HMG proteins (high mobility group proteins), in particular HMG-1 or HMG-2, or nucleosomal histones such as H2A, H2B, H3 or H4 , since this can increase the expression of the desired nucleic acid at least 3-10 times. The chromosomal proteins can be isolated, for example, from calf thymus or rat liver by generally known methods or can be produced by genetic engineering. Human HMG-1 can be genetically engineered using the human cDNA sequence from Wen, L. et al. (1989) Nucleic Acids Res., 17 (3), 1197-1214.

Another object of the present invention is therefore also a method for producing a peptide according to the invention, in which the peptide is either chemically synthesized or, as already explained above, produced by genetic engineering. The chemical synthesis takes place, for example, with the aid of the generally known Merrifield technique.

However, tissue-binding peptides can also be obtained according to the present invention by bringing one or more peptides into contact with a tissue and then isolating the bound peptide (s).

Another object of the present invention is therefore a method for finding a tissue-binding peptide comprising the following steps:

(a) contacting a tissue with one or more peptides, and

(b) isolating one or more tissue-binding peptides.

After contacting a tissue with one or more peptides according to step (a), non-bound peptides are preferably, for example, by Wash the tissue with a suitable buffer solution.

According to the present invention, peptide is preferably understood to mean peptides with a length of approximately 5 to approximately 40 amino acids, in particular with a length of approximately 5 to approximately 20 amino acids, especially with a length of approximately 10 to approximately 20 , particularly preferably with a length of about 10 to about 15 amino acids.

According to the present invention, tissue is understood to mean cell dressings, in particular cell dressings including their intercellular substance, for example epithelial tissue, connective tissue, supporting tissue, muscle tissue or nerve tissue. According to the present invention, the term tissue also includes whole organs or parts thereof and vessels. The organs or vessels are preferably still "living" organs or vessels.

The intercellular substance is a substance embedded in the intercellular spaces, which contains a structure substance that appears to be without structure and fibrous connective tissue fibers. It was therefore completely surprising that tissue-binding peptides were found by bringing tissue into contact with one or more peptides, which peptides were essentially specific for the tissue used and were not predominantly based on non-specific structures, such as, for example. B. the intercellular substance were bound.

The specificity of the tissue-binding peptides can preferably be increased by repeatedly bringing the peptides isolated according to step (b) into contact with the tissue one or more times. After their respective isolation, the peptides are preferably brought into contact with the corresponding tissue up to 10 times. In this way, cross-specific peptides, which are specific for several tissue types, can also be found if the peptides are brought into contact with different tissues in succession. If, for example, the peptides are brought into contact in portions with both healthy and diseased tissue, a subtractive comparison can be used to find the peptides required for healthy or diseased tissue are specific. Peptides that are essentially only specific for diseased tissue are particularly advantageous, for example, for diagnosis, for example in the context of a PET examination.

It is also known that by external circumstances such. B. stress or inflammation, tissue can be pathologically changed. So change e.g. smooth muscle cells in restenotic vascular areas their entire phenotype and change from the resting, secretory to the proliferative phenotype. With the help of the present invention, tissue-binding peptides can consequently also be found which can recognize tissue which is specifically pathologically altered under in vivo conditions. It is also advantageously possible to characterize pathological cell types in the tissue structure, which have not yet been adequately characterized, using the peptides selected according to the method of the invention, and to treat them, for example, with gene therapy.

To find tissue-binding peptides according to the method according to the invention, it is advantageous if a peptide bank, for example a combinatorial peptide bank (see eg Colas, P. et al. (1996) supra), is used. A peptide bank according to the present invention is understood to be a collection of several different peptides, which are present free or in bound form.

For example, a peptide bank of free peptides contains a collection of chemically synthesized peptides. (see e.g. Geysen, H.M. et al. (1996) supra). Amino acid mixtures were used on the solid phase in peptide synthesis so that the amino acid sequence of the resulting peptides was statistically distributed. The peptide banks could also be used in resin-bound form (de Koster, H.S. et al. (1995) supra).

Peptides in bound form are also so-called, for example peptide presenting banks, e.g. B. a peptide-presenting phage bank or preferably a peptide-presenting bacterial bank, as already described in more detail above. At the peptide-presenting banks, the peptides are presented in the form of a fusion protein on the surface of, for example, phages or bacteria. Preferred fusions are insertions, preferably fusions with thioredoxin, with thioredoxin (TrxA) -flagellin (FliC), with the alpha subunit of G _t , as already described in more detail above, or z. B. with LamB (Charbit, A. et al. (1988) Gene 70, 181-189) or Lpp-OmpA (Francisco, JA et al. (1993) Proc. Natl. Acad. Sci. USA 90, 10444-10448 ).

For the present invention, peptide-presenting bacterial banks and in particular those with a thioredoxin fusion protein are particularly preferred especially for selection in an in vivo approximate perfusion model of isolated vessels and organs for the following reasons:

1. The peptides are presented as fusion proteins together with a protein, for example thioredoxin, the three-dimensional structure of which is generally known. The randomized peptides are generally presented in a very accessible domain, which is on the surface of the total protein, such as the so-called "Cys-Cys active site loop" of thioredoxin. Because the peptide is generally sufficiently far from the bacterium and thus is presented in an easily accessible manner, a good interaction with the desired binding partner is generally made possible.

Consequently, it can generally be ensured that, for example, in the case of viral gene transfer into vascular cells, the peptides can also develop their effect in the form of fusion proteins with a protein having a therapeutic effect without essentially losing their binding capacity. Peptides fused to the terminus of a protein and presented in this way in phage libraries generally lose their binding capacity as soon as they are expressed as fusion proteins together with a therapeutically active protein, which is for use in gene therapy is disadvantageous. An important reason for this is that the peptides themselves have a free N or C terminus during the presentation which is available for binding. For this reason, it is particularly advantageous if the peptides are presented in the form of inserts into selected surface proteins by a peptide-presenting bank.

2. For example, the E.coli thioredoxin is stable against proteases. Consequently, a protease-resistant protein in the form of a fusion protein with the peptide is particularly preferred, for example, for selection in the organ model perfused ex vivo, in particular since proteases can be located on the cells in the vessels. Especially organs, e.g. isolated perfused liver, but also tumors, release a high level of proteases that could cleave the fusion protein in the event of protease sensitivity.

3. Due to their size, bacteria cannot get into the extravascular space as quickly as phages and thus reach cells in the perfused organ, for example, which are not accessible for subsequent therapy. In contrast to phages, corresponding bacterial banks can also be incubated at body temperature without the results being essentially falsified by endocytosis. As a result, the in vivo conditions can also essentially be maintained ex vivo, which is particularly advantageous, for example, in the case of “living” organs or vessels.

According to the present invention, the term “living” means organs or vessels that are still functional. To maintain functionality over a longer period of time, it is therefore advantageous if the organ or vessel is adequately supplied with oxygen and nutrient substrates A nutrient solution which contains Ca ²⁺ , Mg ²⁺ and a protein, for example an albumin such as BSA (bovine serum albumin) is preferably suitable for this purpose, preferably the concentration of calcium ions 2.5 mM and magnesium ions 1.1 raM is preferably present in the nutrient solution in a concentration of 1% for the perfusion of isolated The so-called Krebs-Henseleit solution, which is composed as follows (data in raM), is particularly suitable for vessels and isolated organs: NaCl 116; KC1 4.6; MgSO ₄ 1.1; NaHCO ₃ 24.9; CaCl ₂ 2.5; KH ₂ PO ₄ 1.2; Glucose 8.3; Pyruvate 2.0 and 1% BSA equilibrated with 95% O ₂ and 6% CO ₂ (pH 7.4; 37 °) (see e.g. Decking, UKM et al. (1997) Circulation Research 81, 154-164 , No. 2)

In a preferred embodiment, for example, a peptide-presenting bacterial bank was used in an in-vitro vascular perfusion model after a restenosis had previously been induced in a vessel on the whole animal (rat) by denuding the endothelium by means of a balloon catheter. This method has the decisive advantages that the changes in the expression patterns of membrane-associated or membrane-bound proteins which occur as a result of pathological vascular processes are retained, the cell polarity does not change and all cells remain in their natural tissue structure.

The peptides found with the aid of the method according to the invention can, as already described in more detail above, be modified. A suitable modification is, for example, a radioactive or non-radioactive label for use in a test system or diagnostic.

There are numerous applications for the tissue-specific peptides that can be found using the method according to the invention, for example the tissue-specific transfer of substances, in particular pharmaceutically active compounds, especially for tissue-specific gene transfer. Tissue-specific gene transfer is used, for example, for diagnosis and / or therapy of vascular and / or organ diseases with the aid of, for example, viral and / or non-viral vectors, preferably with the aid of liposomes, as already described in more detail above.

For tissue-specific gene transfer, the peptide can be physically, chemically or be genetically linked to the desired nucleic acid (see above). For example, the peptide can either be combined with a therapeutic protein, e.g. B. nitric oxide synthase (see e.g. WO95 / 27020), insulin (see e.g. EP-B1-0 001 929), erythropoietin (see e.g. EP-B1-0 148 605), or blood coagulation factors , such as B. factor VIII, interferons, cytokines, hormones, growth factors, etc. are expressed, or coupled to a nucleic acid, which encodes a therapeutic gene, for example a gene coding for the exemplified therapeutic proteins or for an antisense nucleic acid for the targeted switching off of expression products become. Suitable coupling methods are e.g. B. about lysine residues or nucleocapsid proteins, as already explained in more detail above (see, for example, Harbottle, RP et al. (1998) supra or Bachmann, AS et a 1. (1998) supra). It when one or more tissue-binding peptides via a positively charged domain, for. B. polylysine, are bound to one or more nucleic acids.

A peptide which increases membrane permeation is also suitable for increasing membrane permeation during gene transfer, e.g. B. a hisitidine-containing polypeptide (see above). Preferably, a so-called polyfection solution containing a nucleic acid with the desired therapeutic gene, a fusion protein of tissue-specific peptide and a DNA-binding part, e.g. B. a positively charged domain, and a peptide that increases membrane permeation. A suitable polyfection solution is described, for example, by Midoux, P. (1998) supra. Furthermore, a coupling of the peptide to the liposomes via, for example, a C-terminal cysteine introduced to an activated lipid component, e.g. B. via MPB-PE,

Maleinimidophenylbutyroylphosphatidylethanolamine, (Zuidam, N.J. et al. (1995) Biochim. Biophys. Acta 1240, 101).

The peptides according to the invention can also be used to change the tropism of viruses, preferably adenoviruses, for example for tissue-specific gene transfer. This is done, for example, by attaching the peptide via a suitable linker to the C-terminus of the Knob domain (see, e.g., Douglas, JT & Curiel, DT (1997) Neuromuscular Disorders 7, 284-298) or by exchanging the region responsible for binding in the knob domain of the fiber protein. Alternatively, the RGD motif in the hypervariable region of the penton base (eg Adl2 19AS, Ad2.5 82 AS) can be replaced by a cell-specific peptide analogous to the experiments carried out by Wickham with the FLAG epitope, preferably also the natural one Tropism can be switched off (see Wickham et al. (1996) J. Virol. 70, 6831). The tropism can be changed, for example, via a shuttle plasmid which codes for the corresponding construct.

Another possible use of the tissue-binding peptides is the use for the diagnosis of pathologically altered tissue, especially in a pathologically altered vessel, for inflammation, arteriosclerosis, vessels that supply tumor tissue, and tissue with proliferating smooth muscle cells of the vascular system, as well as the representation of different sections of the Vascular system, such as. B. small / large vessels, veins, organ-specific endothelium.

The diagnosis can either indirectly via z. B. Peptide-recognizing antibodies, for example, in a generally known ELISA test (see, for example, Voller, A. et al. (1976) Bull. World Health Organ., 53, 55-63) or directly by the tissue-binding Peptide is additionally marked. The label can be a non-radioactive label, e.g. B. about biotin or avidin / streptavidin, or a radioactive label (see above).

A radioactive label is particularly useful for a whole-body examination of the patient using imaging methods such as e.g. B. PET (positron emission tomography) or SPECT (single photon emission computer tomography) for the diagnosis of, for example, inflammation, vascular changes or tumors, advantageous. For this purpose, the labeled peptide is administered intravenously to the patient and the distribution of the peptide in the organism is analyzed, for example, using “whole-body PET”. The present invention therefore furthermore relates to a medicament and / or test system, for example in the form of a diagnostic agent, comprising one or more peptides according to the invention or one or more peptides found by the method according to the invention, or one or more nucleic acids according to the invention, and, if appropriate, suitable ones Auxiliaries and / or additives.

Another object of the present invention also relates to a composition, for example in the form of a transfection or polyfection system, containing one or more peptides according to the invention or one or more peptides found by the method according to the invention, or one or more nucleic acids according to the invention and one further substance, preferably a pharmaceutically active compound, as already described in more detail above, and, if appropriate, suitable auxiliaries and / or additives.

Suitable auxiliaries and / or additives are, for example, generally known protease or nuclease inhibitors.

The following figures, tables and examples are intended to describe the invention in more detail without restricting it:

FIGURES AND TABLES

Fig. 1 shows the carotid model from Clowes. After ligating the internal and external carotid arteries, a balloon catheter is inserted through the external carotid artery into the common carotid artery, dilated and the vessel denuded over a distance of 1 - 1.5 cm. The catheter is then removed again, the external carotid artery ligated and the internal carotid artery reopened.

Fig. 2 shows the selection of the peptides in the ex vivo perfused rat carotid. 3 shows the relative binding strength of the peptide P36 to a dilated rat carotid, which was removed on the 8th postoperative day, in comparison to the untreated control carotid.

Figure 4 shows the binding constants of peptide P36 to primary aortic porcine endothelial cells (PAEC), primary porcine smooth muscle cells (VSMC) and monkey kidney cells (COS). The endothelial cells were stimulated by LPS and TNFα (PAEC stimulated).

Tab. 1 shows the DNA sequences and peptide sequences derived therefrom of the isolated clones.

ABBREVIATIONS

BSA bovine serum albumin

DCM dichloromethane, synonymous: methylene chloride

DIEA diisoethylamine EDT ethanedithiol

HBTU / HOBt hydroxybenzotriazole

IMC medium, contains (lxM9 salts, 0.2% (casamino acids, 0.5%

Glucose, 1 mM MgCl ₂ ) 10xM9 salts: 0.42 M Na ₂ HPO ₄ , 0.22 M KH ₂ PO ₄ , 85 mM NaCl, 187 mM NH ₄ C1, pH 7.4) NMP N-methyl-2-pyrrolidone

PBS phosphate buffer

SASRIN "super acid sensitive resin"

TFA trifluoroacetic acid

EXAMPLES

1. Prepare a peptide-presenting bacterial bank

A peptide-presenting bacterial bank was used, which is commercially available from Invitrogen BV, the Netherlands (FliTrx ™ Random Peptide

Display Library, Catalog No. Kl 125-01; Z. Lu, et al. (1995) Biotech. 13, 366). In this bacterial bank, the bacteria contain an expression cassette which codes for a fusion protein of flagellin and thioredoxin. The flagellin serves to anchor the fusion protein in the bacterial membrane, while the thioredoxin presents a foreign 12 amino acid long peptide in an exposed domain, the coding 36 base pair long oligonucleotide of which was cloned at the appropriate position in the coding sequence of the thioredoxin. The oligonucleotide used for cloning was synthesized at random, so that the bacterial bank represents a total of 10 ⁹ different peptides. The expression cassette is under the control of the tryptophan promoter, so that the expression of the fusion protein can be switched on by the administration of tryptophan at the time of an exponentially growing bacterial culture. Since each bacterium only takes up one plasmid, only one type of peptide with a defined amino acid sequence is expressed from each bacterium and presented on the surface.

2. Growing the bacterial culture

50 ml of IMC medium were mixed with 10 ⁹ bacteria and incubated overnight at +25 ° C for 15 hours. The cell number was then determined by measuring the OD ^, 10 ° cells were incubated in 50 ml IMC medium with 100 μg / ml ampicillin and the expression of the bacterial bank was induced by adding 100 μg / μl tryptophan. The cells were incubated for another 6 hours at 25 ° C before being used for selection.

3. In vitro vascular perfusion model

Restenosis was induced in an established restenosis model in the rat (see, for example, Clowes, AW et al. (1983) Lab. Invest. 49, 327). As shown in FIG. 1, the common carotid artery in an anesthetized rat was freed from the endothelial cells by introducing a balloon catheter (denudation). This procedure stimulates the smooth vascular muscle cells to proliferate and changes their phenotype from the resting, secretory smooth muscle cell to the proliferating smooth muscle cell.

On the day of the strongest proliferation of smooth muscle cells (10th postoperative day), the vascular segment marked in FIG. 1 was dissected out and perfused in a special ex vivo perfusion system with a saline medium with the addition of glucose and gassing with oxygen. After an equilibration time of 10 min, which was used to wash out any blood cells and blood components, this vascular segment with a total of 10 ¹⁰ peptide-presenting bacteria in oxygen-gassed saline medium with a flow of 0.2 ml / min. Recirculated perfused for 60 min. The non-binding bacteria were detached by treatment with saline medium and the specifically binding bacteria were eluted by perfusion for ten minutes under stringent conditions with 0.1 M glycine pH 2.0. The eluted bacteria were cultivated again and subjected to a further five selection cycles (see FIG. 2) in order to increase the specificity.

4. Test protocol for the perfused rat carotid

Rats weighing approximately 500 g were anesthetized by administering 100 μl / 100 g body weight Dormicum and 100 μl / 100 g body weight Hypnorm. The carotid artery was then dissected cranially to the bifurcation, the internal carotid artery and all branches branching from it were ligated and the carotid communis was denuded starting from the external carotid about 2 cm along the common carotid with a 2F balloon catheter at 1 bar. The internal carotid artery ligature was reopened postoperatively. On the 10th postoperative day, the denuded common carotid artery was dissected and ex vivo with 0.1 ml / min PBS perfused with 0.1 M glucose. Meanwhile, 10 ml of the tryptophan-induced bacterial suspension was centrifuged at 1000 rpm for 3 min, washed in PBS, centrifuged again and the sediment in 6 ml total volume of PBS, 0.1 M glucose, 0.1% BSA, 1 mM CaCl ₂ , 10 mM MgCl ₂ suspended. The carotid was then perfused recirculating with this suspension with gassing of oxygen at 0.2 ml / min for 1 h. The suspension was then washed out by perfusion with 2-3 ml of PBS at 0.2 ml / min before the bound bacteria were eluted by elution with 3 ml of 0.1 M HCl (pH 2.2 adjusted with glycine) and in 50 ml of IMC Medium, 100 μg / ml ampicillin were grown overnight at +25 ° C.

5. Result of the selection attempts

According to the experiments described above, bacteria were obtained which express peptides on their surface which bind specifically to proliferating smooth muscle cells. The bacteria obtained after six selections were then plated on nutrient medium and cultures were grown from 60 individual colonies. From these cultured bacteria, the plasmid DNA was isolated and sequenced. With this model, surprisingly, specifically binding peptides could be detected using a method closest to the in vivo situation (organ culture model).

After sequencing the DNA of the isolated bacteria, the DNA sequences and peptide sequences derived therefrom were obtained (see Table 1). The sequence of peptide 36 shows homologies to the α ₄ integrin, which is expressed on leukocytes and is responsible for the attachment of the leukocytes to activated vascular cells. The recognition is mediated by the vascular cellular adhesion molecule VCAM-1. The homology is not in the area that is characterized as the binding area of the integrin (RGD-Motif). 6. Attachment studies

The peptide from clone 36 (peptide 36) used for the binding studies was produced using the Fast-Moc solid-phase synthesis on a fully automatic peptide synthesizer from Applied Biosystems on a 0.1 mmol scale. The peptide was radioactively labeled to determine the binding affinities to the desired cell type.

7. Radioactive labeling of a tissue-specific peptide

Since the bacteria used for the selection presented the peptides on their surface in the context of a fusion protein (thioredoxin), the N-terminal amino acid (glycine) was used for the radioactive labeling. For this purpose, ¹⁴ C-gycin was protected with activated N-fluorenyl-methoxycarbonyl-succinimide, purified and coupled covalently N-terminally to the peptide located on the resin and protected in the side chains by means of dicyclohexyl-carbodiimide. The peptide was then cleaved from the resin, purified by HPLC and examined for binding to the target cells in the perfused vessel model and in the cell culture.

8. Experimental protocol for the synthesis of the radioactively labeled C-Fmoc glycine

20 μmol of the radioactively labeled ¹⁴ C-glycine (corresponding to 1 mCi; DuPont) in 5 ml of water were combined in one with 38 mg of glycine (0.5 mmol) and 53 mg of Na ₂ CO ₃ (0.5 mmol) 25 ml of the round-bottom flask were added and 165 mg (0.49 mmol) of N-fluorenylmethoxycarbonyl carbonate (Fmoc carbonate) dissolved in 5 ml of acetone were added dropwise with stirring. During the dropping (60 min), the pH was kept constant by adding 1 M Na ₂ CO ₃ . The mixture was then left to stand at room temperature overnight, the aqueous phase was acidified by adding 2 M HCl and extracted twice with 10 ml of ethyl acetate each time. The combined ethyl acetate phases were washed neutral with water several times, dried with MgSO ₄ and the product was precipitated by adding petroleum ether and filtered off. For purification, the product was dissolved in ethyl acetate and slowly precipitated by adding small amounts of petroleum ether.

Yield: 95 mg (corresponding to 68% of theory)

0.32 mmol specific activity: 0.0064 MBq / μmol

9. Test protocol for the synthesis of radioactively labeled peptides

The labeled Fmoc-Glycine was used for the synthesis of a total of 0.4 mmol peptide, so that each peptide was labeled 75%.

The appropriate amount of the resin / peptide mixture (0.1 mmol) was weighed out and placed in a glass frit with a stopper and a tap. 25 mg of the ^I4 C-Fmoc glycine (0.075 mmol) were dissolved in 1 ml of NMP, 155 μl of 0.45M HBTU / HOBt and 750 μl of 2 M DIEA in NMP were added to the resin / peptide mixture and added for one hour incubated with gentle shaking at room temperature. To complete the coupling, 150 mg of Fmoc-glycine (0.5 mmol) were then dissolved in 2 ml of NMP, 2 ml of 0.45M HBTU / HOBt and 1 ml of 2M DIEA were added and the mixture was added to the reaction mixture. The mixture was 30 min. incubated at room temperature, the unreacted Fmoc-glycine filtered off and the mixture washed four times with 6-10 ml of NMP each. The protective group was split off by incubation for 30 minutes with 22% piperidine in NMP with stirring. The piperidine was filtered off, and the resin / peptide mixture was then washed four times with NMP and DCM.

10. Test protocol for splitting off the protective groups

To remove the protective groups, the mixture was ice-cooled and treated with a solution of 0.75 g phenol, 0.25 ml EDT, 0.5 ml thioanisole, 0.5 ml water and 10 ml of TFA are added with stirring. The mixture was then incubated for two hours at room temperature with stirring, the peptide released was filtered off and precipitated by adding five times the volume of ice-cooled diethyl ether. After centrifugation, the peptide in the sediment was washed several times with dieethyl ether and centrifuged again.

The peptides were chromatographed using an Äkta Purifier 100 over an RPC resource column (3 ml, Pharmacia) with a 20 minute gradient of water, 0.1% TFA to 95% acetonitrile, 0.07% TFA at a flow rate of 2 ml / min cleaned and then lyophilized.

Yields: 65.6 and 70.8% (depending on the peptide) specific activity: 0.0165 to 0.0182 MBq / μmol

11. Attachment studies

For the binding studies in the perfused vessel model and in the cell culture for binding to the target cells, the vessels or the cells were incubated for different time intervals with different concentrations of the peptides to be examined and the amount of bound peptide per cm vessel or per cell number was determined.

11.1 Binding to the perfused carotid

It can be seen from FIG. 3 that the peptide 36 binds approximately four times more strongly to the dilated carotid than to the untreated control in the ex vivo perfusion.

11.2 Binding to pig endothelial cells in cell culture

Since peptide 36 also has homologies to the porcine α ₄ integrin sequence, cell culture studies were also carried out on porcine endothelial cells (PAEC), which were stimulated by treatment with lipopolysaccharides (LPS) and the tumor necrosis factor (TNFα). This stimulation corresponds to an in vivo activation of cells which are located in vascular areas exposed to stress or inflammatory stimuli and thus also in restenotic vascular areas.

The peptide was able to bind specifically to smooth muscle cells (VSMCs) as well as to activated PAECs. Surprisingly, the peptide binds even more strongly to activated PAECs than to VSMCs. In addition, the binding constants of the peptide differ from activated and non-activated PAECs by almost 3 orders of magnitude (FIG. 4). The peptide practically does not bind to control cells (COS).

Claims

claims

1. tissue-binding peptide selected from a peptide having the amino acid sequence GEGRTVVLSF, AWCRGGILGDAM, GNLVDLVVGFDD,

RVS PPKKSGGGV, GSSKWGLTXKCG, RGGVRQRSRGRR,

GEGRTVVCRS or SQRWTALWQWIG and variants thereof.

2. Nucleic acid coding for a tissue-binding peptide according to claim 1.

3. Nucleic acid according to claim 2 selected from a nucleic acid with the nucleotide sequence

GGCGAGGGGCGAACAGTCGTATTGTCGTTCG, GCCTGGTGTCGGGGGGGTATCCTGGGCGACGCTATG, GGAAACCTGGTGGATCTAGTTGTGGGTTTTGACGAC, CGGGTGAGTCCGCCAAAGAAGTCGGGGGGCGGCGTG, GGGAGTAGCAAGTGGGGATTGACTTAAAAATGTGGG, CGCGGGGGAGTCCGCCAAAGAAGTCGGGGGCGGCGT, or GGCGAGGGGCGAACAGTCGTATGTCGTTCG

TCCCAGAGGTGGACTGCACTCTGGCAATGGATCGGG and variants thereof.

4. Vector containing a nucleic acid according to claim 2 or 3.

5. A method for producing a tissue-binding peptide according to claim 1, characterized in that the peptide is either chemically synthesized or produced by genetic engineering.

6. A method for finding a tissue-binding peptide comprising the following steps:

(a) contacting a tissue with one or more peptides, and (b) isolating one or more tissue-binding peptides.

7. The method according to claim 6, characterized in that the peptides isolated according to step (b) are repeatedly brought into contact one or more times with the same or with another tissue.

8. The method according to claim 6 or 7, characterized in that the tissue is a pathological tissue.

9. The method according to any one of claims 6-8, characterized in that the peptide or peptides are contained in a peptide bank.

10. The method according to claim 9, characterized in that the peptide bank is a peptide-presenting bank, preferably a peptide-presenting bacterial bank.

11. Use of a tissue-binding peptide according to claim 1 or obtained by the method according to any one of claims 6-10 for the tissue-specific transfer of substances, in particular of pharmaceutically active compounds, especially for tissue-specific gene transfer.

12. Use according to claim 11, characterized in that the gene transfer is viral and / or non-viral, preferably with the aid of liposomes.

13. Use according to claim 11 or 12, characterized in that one or more tissue-binding peptides are bound to one or more nucleic acids via a positively charged domain.

14. Use of a tissue-binding peptide according to claim 1 or obtained by the method according to any one of claims 6-10 for Changes in the tropism of viruses, especially adenoviruses.

15. Use of a tissue-binding peptide according to claim 1 or obtained by the method according to one of claims 6-10 for the diagnosis of pathologically altered tissue and / or for the representation of different sections of the vascular system, preferably small / large vessels, veins or organ-specific endothelium.

16. Use according to claim 15, characterized in that the pathologically altered tissue is selected from a pathologically altered vessel; with inflammation; Arteriosclerosis; Vessels that supply tumor tissue; Tissue with proliferating smooth muscle cells of the vascular system.

17. Use according to claim 15 or 16, characterized in that the tissue-binding peptide is labeled, preferably radioactively labeled.

18. Medicament and / or diagnostic agent containing one or more tissue-binding peptides according to claim 1, one or more nucleic acids according to one of claims 2-4 or one or more tissue-binding peptides obtained by the method according to one of claims 6-10, and, if appropriate, suitable auxiliaries - and / or additives.

19. Composition containing one or more tissue-binding peptides according to claim 1, one or more nucleic acids according to one of claims 2-4 or one or more tissue-binding peptides obtained by the method according to one of claims 6-10, and a further substance, preferably a pharmaceutical active compound, and, if appropriate, suitable auxiliaries and / or additives.