TECHNICAL FIELD The invention concerns peptides from vitronectin and specifically a nonapeptide 351-359 from human vitronectin, their therapeutic application for the promotion of osteoblast adhesion and peptide-containing therapeutic compositions.

BACKGROUND ART Tissue engineering, as multidisciplinary approach for the development of innovative therapeutic devices, is of growing importance in the odontoiatric and orthopaedic fields. In fact, many odontoiatric and orthopaedic pathologies require the necessity to operate by adequate means in order to promote bone tissue healing and, in the same time, to recover organ functionality. In this perspective, it is likely to appreciate that osteoblasts play a key rule: thus, the search for opportune therapeutic devices necessarily implies to develop compounds able to promote osteoblast adhesion without evoking adverse reactions such as fibrous tissue formation due to the presence of other kind of cells. Several drawbacks connected to the use of materials for implantology can be attributed to adverse host response, which also includes the formation of fibrotic capsule around implant surface. At first, the development of novel materials has been focused on searching for surfaces able to minimize undesirable effects, hi the past few years, the most appealing research seems to be aimed at developing bioactive surfaces suitable for the promotion of beneficial interactions between cells and tissues surrounding the implant; in other words, in order to improve tissue- implant integration many attempts are addressed to the exploitation of those biochemical processes which are known to mediate tissue healing and growth. From this point of view, several efforts have concerned the use of complex molecules, i.e. growth factors and adhesion factors, for the modification of implant surface to promote integration (Anselme, K. Osteoblast adhesion on biomaterials. Biomaterials 2000, 21, 667-681; Puleo, D.A. and Nanci, A. Understanding and controlling the bone-implant interface. Biomaterials 1999, 20, 2311-2321). The complex sequence of events taking place in the proximity of the anatomical site of implantation, that might lead to the healing of the surrounding tissues and the integration of the implant itself, depends on the biochemical phenomena mediating the adhesion of the cells to the implant surface. In fact, cell adhesion is mandatory for the functionality of cells such as bone cells. In the orthopaedic and odontoiatric fields, with specific regards to implantology, materials and compounds able to improve osteoblast adhesion are under investigation. Osteoblast adhesion takes place throughout different mechanisms: a first mechanism requires the interaction between the sequence of amino acids called RGD (arginine-glycine-aspartic acid) which is present in many fibrous proteins, and the integrins located on the cell membrane (Puleo, D.A. andNanci, cited ref.); a second mechanism takes place by the interaction between eparan sulphate proteoglycans of cell membrane and some proteins of the extra-cellular matrix (Dee, K.C., Andersen, T.T. and Bizios, R. Design and function of novel osteoblast-adhesive peptides for chemical modification of biomaterials. J. Biomed. Mater. Res 1998, 40, 371-377). The RGD sequence, that mediates the first mechanism, is well known and it originated a number of peptides already used in experimental applications. On the other hand, peptide sequences mediating the second mechanism have not been surely identified yet. Nowadays, it is likely to hypothesise that those sequences are characterized by a motif of basic amino acids such as B-B- X-B (B are basic amino acids, X are non-basic amino acids); recently, the KRSR peptide (Lysine- Arginine-Serine-Arginine) reproducing the B-B-X-B motif, has been proposed as a sequence able to promote osteoblast adhesion (Dee, K.C. et al, cited ref). It is worthy to mention that the first integrin-mediated adhesion mechanism is not osteoblast-specific, that is, RGD- containing sequences are able to promote the adhesion of other kinds of cells besides osteoblast, i.e. fibroblasts; on the other hand, the second mechanism via proteoglycans is osteoblast-specific. For applicative purposes, bioactive-materials research has recently proposed implant surfaces modified by the application of adhesion factors (RGD containing peptides and proteins) and growth factors (Anselme, K. cited ref.). These modified surfaces seem to be able to promote and improve the healing phase of the tissues around the implant, thus enhancing its integration with the biological environment. On the other hand, the clinical use of these macromolecules mediating the physiological processes of tissue healing and regeneration, can cause some drawbacks mainly due to their difficult supply, their cost, and their complex purification (Puleo, D.A. and Nanci, cited ref.) that can sometimes produce contamination with lethal pathogens, i.e. BSE. Furthermore, these macromolecules are often not stable, in particular when adsorbed and/or anchored to the implant surface, and the integrity of their tertiary structure is required to preserve their biological activity. These problems could be bypassed by utilising bioactive peptides that can be obtained by means of chemical synthesis and used under more drastic conditions, thus reducing costs and increasing applicability. In order to define which mechanism among those above described could be of higher interest and to identify compounds able to specifically promote osteoblast adhesion without incurring in the mentioned drawbacks, by analysing the sequences of many proteins presenting binding sites for heparin, the Inventors have identified few signal sequences of potential interest for a comparison with RGD containing peptides; these sequences have been recognized in the vitronectin that is a well known protein from the extra-cellular matrix, phylogenetically conserved in mammals, whose sequence is documented (Suzuki, S., Oldberg, A., Hayman, E.G., Pierschbachetr, M.D., Ruosklahti, E. Complete amino acid sequence of human vitronectin deduced from cDNA. Similarity of cell attachment sites in vitronectin and fibronectin. EMBO J. 1985, 4, 2519-2524). In vitro studies carried out by the Inventors allowed to determine that peptides from the human vitronectin, which contain one or more B-B-X-B motifs, are able to improve osteoblast adhesion but not the adhesion of other cells, i.e. endothelial cells, ha particular, the in vitro assays gave evidence of the ability of the following sequences: (339-364), (339-351) and (351-364) from human vitronectin (ITVP), to specifically promote the adhesion of osteoblast from newborn rats calvaria on polystyrene but with an adhesion efficacy that is lower than RGD containing peptides and than the entire fibronectin, used as positive control, but higher than the KRSR peptide (Dettin, M., Conconi, M.T., Gambaretto, R., Pasquato, A., Folin, M., Di Bello, C, Parnigotto, P.P. Novel osteoblast-adhesive peptides for dental/orthopaedic biomaterials. J. Biomed. Mater. Res 2002, 60, 466-471). Among the above mentioned peptides belonging to the human vitronectin, the (351-364)HVP sequence demonstrated the highest activity. It is worthy that one peptide, presenting both the B-B-X-B motif (due to the KRSR sequence) and the RGD motif, showed higher activity in promoting osteoblast adhesion than human vitronectin-derived peptides, but lower activity when compared with other RGD-containing peptides. This peptide exhibits analogous activity in promoting the adhesion of endothelial cells (Dettin, M. et al. cited ref). On the basis of these results, it has to be pinpointed that the experimental model used (osteoblasts from newborn rat calvaria) is not the ideal one in order to investigate the biological activity of adhesion peptides, hi fact, even if differentiated and with high proliferation capacity, these osteoblasts exhibit specific characteristics due to the kind of tissue (newborn) and the anatomical site (calvaria). From an experimental and applicative point of view, the main source of the osteoblasts is the bone marrow of adult mammals. At the moment, bone marrow represents the best approach for obtaining osteoblasts in vitro since it is currently used as cell source for: i) in vitro assays to evaluate osteoblast adhesion and growth on several substrates (van den Dolder, J. Spauwen, P.H., Jansen, J.A., Evaluation of various seeding techniques for culturing osteogenic cells on titanium fiber mesh. Tissue eng. 2003, 9(2), 315-325); ii) tissue engineering applications to produce bone substitutes in vitro to be grafted in vivo. It is necessary to remember that bone marrow represents an important source of autologous osteoblasts even for the availability of the sampling, the differentiation of the cells and the high yield in osteoblasts (Bruder, S. P., Fox, B. S. Tissue engineering of bone. Cells based strategies. Clin. Orthop. 1999, 367 suppl., S68-83; Cancedda, R., Mastrogiacomo, M., Bianchi, G., Derubeis, A., Muraglia, A., Quarto, R. Bone marrow stromal cells and their use in regenerating bone. Novartis Found Symp. 2003, 249, 133- 143; Ohgushi, H., Miyake, J., Tateishi, T. Mesenchymal stem cells and bioceramics: strategies to regenerate the skeleton. Novartis Found Symp. 2003, 249, 118-127).

DISCLOSURE OF INVENTION The object of the present invention are peptides of vitronectin containing the RHRNRK core sequence whose length does not exceed 12 amino acids and the (351-359)HVP peptide from human vitronectin called "sequence 2" (FRHRNRKGY). Further object of the invention is the exploitation ofthese peptides for the preparation of therapeutic devices able to treat those pathologies that require the promotion of osteoblast adhesion, such as in odontoiatric and orthopaedic applications, and therapeutic compositions containing the above mentioned therapeutic compounds. This detailed description explains the aims and the advantages of vitronectin-derived peptides of the invention and of the nonapeptide (351-359)HVP (FRHRNRKGY) (phenylalanine-arginine- hystidine-arginine-asparagine-arginine-lisine-glycme-tyrosin e) of sequence 2, and the aims and the advantages of their therapeutic applications in promoting osteoblast adhesion for treating odontoiatric and orthopaedic pathologies, and in particular in implantology. As in example but without any limitation, we will describe peptide synthesis, its characterization and its biological properties. The primary structure of the nonapeptide (351-359)HVP (FRHRNRKGY) of sequence 2 is characterized by the presence of a sequence of basic amino acids RHRNRK that is seemed to be relevant for the biological activity. Without exiting the limits of the invention and for its purposes, peptides from mammalian, and in particular human, vitronectin can be the object of the invention; these are the peptides containing the RHRNRK core sequence with a number of amino acids not exceeding 12, and the nonapeptide (351-359)HVP (FRHRNRKGY) of sequence 2. These peptides can include from 1 up to 3 amino acids close to the C-terminus and/or N-terminus, derived from the sequence of vitronectin itself, excluding peptides with the sequence of the (351-364) peptide of human vitronectin of sequence 1, or derived from sequences of other proteins or other polypeptides and representing, for example, target sequences or "tag" sequences usually added to the C- or N- terminus in order to improve, for example, purification (i.e., His-tag), immunological recognition (i.e., myc-peptide), or different specifically oriented targeting. Further, chimeric polypeptides where the above mentioned peptides with a number of amino acids not exceeding 12 comprising the RHRNRK core sequence and the nonapeptide (351- 359)HVP of sequence 2 (FRHRNRKGY), can be covalently bound to other proteins than vitronectin or to carrier molecules such as polysaccharides, polymers, DNA, composite materials and metals, can be also object of this invention. The above mentioned peptides and polypeptides can be formed by D-amino acids and can be mutant peptides and polypeptides bearing conservative substitutions with regard to the charge of the basic residues of the core sequence. In order to verify the experimental data obtained on osteoblasts from newborn rat calvaria, the Inventors tested the same peptides on osteoblasts obtained from the bone marrow of adults rats for the previously discussed reasons. Surprisingly, the Inventors found that peptides reproducing sequences from the human vitronectin and, in particular, the (351-364) sequence (sequence 1) have not been able to improve osteoblast adhesion in a way significantly different form the control (i.e., polystyrene). Moreover, the Inventors found that one 9-mer peptide belonging to the (351-364)HVP fragment, i.e. the (351-359)HVP peptide (sequence 2), is not only active, but more active than the RGD containing peptides and fibronectin. The Inventors also found that, surprisingly, the (351-359)HVP peptide, which presents RHRNRK core sequence, is not only more active than the (351-364)HVP peptide, but has a different secondary structure.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 : acellular bone matrix Fig. 2: acellular bone matrix, pre-treated with fibronectin and used for cell seeding Fig. 3: acellular bone matrix after seeding Fig. 4: acellular bone matrix, pre-treated with the (351-359)HVP peptide of sequence 2 (FRHRNRKGY) used for cell seeding. Fig. 5: secondary structure of the (351-359)HVP linear peptide. Figure represents the best overlapping of the 4-8 segment backbone among 10 selected structures after elaboration with DYANA/AMBER. Black ribbon shows the average structure obtained, that is characterized by an extended conformation close to the N-terminus region, and by a type I β-turn centered on residue 6 that is present on 4 of the 10 selected structures. Fig. 6: secondary structure of the (351-364)HVP peptide. Figure represents the best overlapping of the 4-8 segment backbone among 10 selected structures after elaboration with DYANA/ AMBER. Black ribbon shows the average structure obtained. Model peptide shows an extended conformation with an equatorial γ-turn involving Arginine 6 that is present on all the selected structures. Fig. 7: secondary structure of the (351-359)HVP cyclic peptide. Molecular model globally shows a rectangular structure with residues 1-2 and 6-7 on the vertexes. Two γ-turns can be identified: an equatorial one, involving residue 2, and an axial one, involving residue 6. 3-5 and 7-9 segments are in extended conformations. Fig. 8: CD spectra of the (351-359)HVP linear peptide (x trace), (351-359)HVP cyclic peptide (y trace) and (351-364)HVP peptide (z trace) into three different solvents: 10 mM phosphate buffer pH = 7 (A); 14 mM sodium dodecylsulphate in 10 mM phosphate buffer pH = 7 (B); 2,2,2- trifluroethanol (C). Note that CD curves, which depends on the secondary structure of the examined peptides, differ in all of the solvents used.

EXPERIMENTAL SECTION Preparation of the nonapeptide (351-359)HVP of sequence 2 (FREiKNRKGY) can be obtained by means of a number of methods known to expert in the field of peptide synthesis; in particular, the nonapeptide (351-359)HVP can be synthesized in solid-phase as illustrated in the following example.

Example 1: synthesis of the nonapeptide (351-359)HVP (phenylalanme-arginme-hystidine- arginine-asparagine-arginine-lisine-glycine-tyrosine) The nonapeptide (351-359)HVP (FRHRNRKGY) has been obtained by means of solid-phase peptide synthesis using an Applied Biosystems synthesiser (Model 431A), applying the Fmoc chemistry and using 0.25 mmoles scale. Peptide chain has grown on a SASRTN (Bachem) support. Each amino acid has been inserted by single coupling reaction using 4 equivalents of Fmoc-amino acid opportunely protected on its side chain (Pmc for Arg; Trt for Asn and His; tBu for Tyr; Boc for Lys). The 2-(lH-benzotriazol-l-yl)-l,l,3,3-tetramethyluronium hexafluorophospate (HBTU) has been used as condensation agent. At the end of the synthesis peptide has been cleaved from the solid support and deprotected by treating with trifluoroacetic acid:phenol:H20:etandiole tioanisole = 10 mL:0.75 g:0.5 mL:0.25 mL:0.5mL. Crude peptide has been purified by reverse phase chromatography. Product identity has be ascertained by mass spectrometry investigations. Homogeneity of the product (99.9%) has been determined by chromatographic analyses and capillary electrophoresis. The biological properties of the nonapeptide that is the object of the invention have been investigated in vitro on osteoblasts from adult rat bone marrow and, in order to verify cell specificity, on rat endothelial cells.

Example 2: covalent functionalization of titanium disks surfaces with the (351-359)HVP nonapeptide Three titanium disks (15 mm diameter, 2.5 mm height) have been sand-blasted and treated with chloridric and sulphuric acids (M. Genovese, Caratterizzazione di superfici in Ti e Ti6A14V trattate per promuovere l'adesione di osteoblast!, doctoral thesis in Chemical Engineering, University of Padova, AA 2001/2002). Disks have been washed with toluene, previously distilled and dehydrated on metallic sodium. Disk surfaces have been oxidized using 25 niL H2SO4 and 25 mL H2O2 30% solution, for 2 hours at room temperature. Disks silanization has been performed by treating with 10% solution of 3-aminopropyltriethoxysilane in toluene refluxing for 4 hours. Afterwards, disks have been placed into a solution of 1% glutaraldehyde in 0.1 M sodium phosphate buffer, pH = 7, for 4 hours at 25° C. Finally, disks have been immerged in 10 mL milliQ water containing 12 mg of the (351-359)HVP nonapeptide overnight.

Example 3: adult rat bone marrow osteoblast adhesion Osteoblasts have been sampled from bone marrow of femurs of adult Sprague-Dawley rats. Rats have been sacrificed and femurs have been taken under sterile conditions and put on Petri dishes with buffered physiological solution (PBS). After removal of adherent tissues, short washing with ethanol 70% has been performed and epiphyses have been cut with scalpel, hi order to collect the cells, medullar channel has been repeatedly washed with a syringe using DMEM (Sigma) supplemented with fetal bovine serum (Seromed) and antibiotics (Sigma). Cell suspension has been centrifuged at 1500 rpm for 5 minutes and the cell-containing pellet has been resuspended in α-MEM (Gibco) with a 1% antibiotic-antimycotic solution (Sigma) supplemented with the following factors: Fetal Bovine Serum (FBS) 10%; Glutamine (Sigma) 2 mM; Ascorbic acid (Sigma) 50 μM; Dexamethasone (Sigma) 10"8 M; Ultroser (Gibco) 2%. 5 days after seeding, the medium has been renewed and added with β-glycerophosphate (10 mM) (Sigma). The medium has been changed every other day after a short washing in PBS. Confluent cultures has been removed with 0.002% EDTA (Sigma) and 0.25% trypsin (Sigma) in PBS, 1 : 1 (v/v), and then centrifuged at 1500 rpm for 5 minutes. Osteoblast phenotype has been confirmed by alkaline phosphatase activity, by formation of mineral calcium salts, and by electron microscopy (SEM, scanning electron microscopy). Tertiary cultures have been utilized for in vitro adhesion assays. 96 wells cultures plates have been previously conditioned with the examined peptides (2 x 10"3 μmoles/cm2), using fibronectin as control, by 60 minutes incubation at 37° C. Cells have been seeded on the substrates with 3 x 104 cells/cm2 density (surface area of substrate) and allowed to adhere for 5 hours under standard conditions. At the end of the incubation time, cells have been fixed with 10% formaldehyde in phosphate buffer overnight. Fixed cells have been stained with 0.04% cresyl violet in methanol 20% for 30 minutes. Stain has been extracted with 10% citric acid and 50% ethanol solution in water. Absorbance has been measured at 600 nm under Microplate autoreader EL 13 BIO-TEK. Results obtained as means of four experiments have been converted into percentages: optical density values (OD) corresponding to the cultures seeded into non-conditioned wells have been taken as 100%. Results The following peptides have been tested: RGD containing peptides GRGDSPK, (GRGDSP)4K, MAP(RGDSP), and peptides that are fragments of the human vitronectin (339-364)HVP, (339- 35I)HVP, (351-364)HVP, (351-359)HVP, (351-359)HVP cyclic, and fibronectin as standard reference. The results obtained are illustrated in the following table (Table 1).

peptide % average SD 104 GRGDSPK 104 102 2.9 99 113 (GRGDSP)4K 122 113 9.5 103 112 MAP(RGDSP) 97 Ul 13.1 123 99 (339-364)HVP 98 0.4 111 (339-35I)HVP 97 102 7.8 98 107 (351-364)HVP 98 100 6.2 95 127 (351-359)HVP 129 121 12.2 107 114 (351-359)HVPc 118 114 4.0 110 116 Fibronectin 115 112 6.1 105

Example 4: adult rat bone marrow osteoblast adhesion on acellular matrix Femurs of adult Sprague-Dawley rats have been washed 4 times for 15 minutes in PBS with antibiotics and amphotericin B. Bones have been demineralised by treatment with 0.5 M sodium EDTA for 21 days at 4° C and then treated following Meezan et al. (Life Sciences 1975, 17, 1721) to obtain the acellular matrix. Samples have been treated with distilled water for 72 hours at 4° C, 4% sodium deoxycholate for 4 hours and 2000 Kunits deoxyribonuclease I (Dnase I) for 2 hours. Absence of cellular elements has been histologically confirmed (stains: hematoxylin and eosine) and the acellular matrices have been stored in PBS at 4° C until use. Acellular bone matrices have been conditioned with 2 nmol/cm2 peptides or fibronectin and then 8 x 104 osteoblasts/cm obtained as illustrated in the Example 3 were seeded. After 24 hour incubation, cultures have been fixed with 3% glutaraldehyde in cacodylate buffer, pH = 7.2. After drying at critical point and deposition by gold sputtering, cultures have been examined under SEM. The results corresponding to the acellular bone matrix, osteoblast cultures on acellular bone matrix not conditioned or pretreated with fibronectin or the (351-359)HVP FRHMSDRKGY, are reported in Figure 1. Example 5: adult rat bone marrow osteoblast adhesion on titanium disks covalently functionalized with the (351-359)HVP nonapeptide Osteoblasts of adult rat bone marrow have been seeded and allowed adhering on titanium disks of Example 2 and under the same conditions previously described without nonapeptide. The results obtained demonstrated that the covalently bound nonapeptide on the Ti disks is able to promote osteoblast adhesion in a very effective manner (Table 2). Table 2: adhesion of adult rat bone marrow osteoblasts on disks peptide OD average % control 0.579 100 ^351^359)HVP ~ 0.977 169 control 0.410 100 (351-359)HVP 0.585 143 control 0.336 100 (351-359)HVP 0.584 174

Example 6: rat endothelial cell adhesion Materials and methods : Endothelial cells have been obtained from adult Sprague-Dawley rat brain, that it was washed twice for 10 minutes in cold solution of phosphate buffer (PBS) and antibiotic-antimycotic (sodium penicillin G salt 100 WJmL, streptomycin sulphate 100 μg/mL, amphotericin B 0.25 μg/mL) (AF) (Sigma- Aldrich). After removing connective elements, sample has been opportunely crumbled and suspended in an enzymatic solution containing: 0.1% collagenase B/dispase (Boehringer), 10 mM HEPES (Sigma), 20 units/mL Dnase I (Sigma). Enzymatic digestion and micro-vessels separation from remaining cellular components have been carried out in the Dubnoff bath under stirring for 1 hour at 37° C. Supension has been centrifuges for 5 minutes at 2100 rpm and pellets have been resuspended in a gradient of 25% serum albumin (BSA) in PBS solution containing 10 nM HEPES and centrifuged for 5 minutes at 2740 rpm. Pellet has been resuspended in PBS and centrifuged for 5 minutes at 2100 rpm to eliminate the BSA gradient still present. Microvessel fragments have been resuspended in 5 mL enzymatic solution containing 0.1% collagenase B/dispase and 10 mM HEPES, then incubated under mild stirring, for 45 minutes at 37° C, and finally centrifuged for 5 minutes at 300 rpm. Pellet has been resuspended and seeded onto Petri dishes previously conditioned with collagen (1 μg/cm2) (Sigma) and human fibronectin (1 μg/cm2) (Boeringher). Cells have been cultured in a EBM medium (BioWhittaker) added with following factors: 1% antibiotic-antimycotic (Sigma); 10% bovine fetal serum (Seromed); glutamine (200 niM) (Sigma); Endothelial Cell Growth Factor (12.5 μg/mL) (Sigma); hydrocortisone (1 μg/mL) (Sigma); Endothelial growth Factor (0.01 μg/mL) (Sigma); heparin (75 U/mL) (Biologici Italia Laboratories). Medium has been changed every other day until confluence. The cultures contained not only endothelial cells but also fibroblasts that have been eliminated by immunomagnetic separation. Purification has been performed in two steps: a) conditioning with magnetic beads anchoring the CD31 antibody; b) immunoseparation. Polystyrene magnetic beads (dynabeads M-450 tosylactivated (Oxoid), have been used; they have been coated by a polyurethane layer activated with p-toluensulphonyl chloride able to proved the reactive groups for protein covalent binding (i.e., -Ac- antibodies) or other ligand with sulphidric groups. In this case, beads have been conditioned with a primary antibody against a surface antigen specific for the endothelial cells: CD31 antibody (S erotec). Treatment has been performed by washing the beads in a buffer (0.1 M PBS, pH = 7.4) and then treating them with the antibody (4-10 x 108 beads/mL final solution), addition of 0.1% BSA and incubation for 24 hours at 37° C under stirring. Conditioned beads have been stored at 4° C until use. Cells from sub-confluent cultures have been detached and suspended in an appropriate PBS volume (106 cells/mL), pH = 7.4, with 0.1% BSA (Buffer D). CD31 conditioned beads (5 beads/cell) have been added to the cell suspension, followed by incubation for 30 minutes at 4° C. After repeated washings, cells still anchored to the beads have been seeded onto Petri dishes previously conditioned with collagen (1 μg/cm2) and human fibronectin (1 μg/crn2), using the same medium described above and changed every other day. In order to obtain sub-cultures, cells have been detached using 1:1 (v/v) 0.05% trypsin and 0.02% ETDA solution. Cell phenotype has been confirmed by imrnunocytochemistry using an antiobody againt the von Willebrand factor. Adhesion assays have been performed with secondary cultures. 96 wells cell culture plates have been pre-conditioned with the peptides (2 x 10"3μmoli/cm2) under investigation, and 3 x 104 cells /cm were seeded and allowed to adhere for 5 hours under standard conditions. At the end of the incubation, cells have been fixed with 10% formaldehyde in phosphate buffer overnight. Fixed cells have been stained with 0.04% cresyl violet in 20% methanol for 30 minutes. Staining has been extracted with a 10% citric acid and 50% ethanol solution; absorbance has been determined at 600 nm with a Microplate Autoreader EL 13 BIO-TEK. The results obtained as means of 4 experiments have been converted into percentages; optical density values (OD) relative to the cultures seeded into non-conditioned wells have been assumed as 100%. Results: The following peptides have been tested with the endothelial cells: (GRGDSP)4K, GRGDSPK, and (351-359)HVP. The results are illustrated in Table 3. Table 3 : adhesion of adult rat endothelial cells peptide OD average OD % 0.171 0.190 no peptide 0.182 100 0.185 0.183 0.189 0.187 (GRGDSP)4K 0.188 103 0.190 0.186 0.187 0.199 GRGDSPK 0.192 105 0.192 0.189 0.179 0.187 (351-359)HVP 0.182 100 0.185 0.178

The synthesis of the (351-359)HVP (FRHRNRKGY) of sequence 2 allowed to obtain a compound that has been tested to be more active in the promotion of osteoblast adhesion by means of a specific mechanism probably meditated by proteoglycans for which the sequence of basic amino acids RHRNRK can be relevant. Assays with adult rat bone marrow osteoblasts demonstrated that the (351-359)HVP (FRHRNRKGY) of sequence 2 is able to promote osteoblast adhesion on polystyrene, much more than the entire fibronectin and than the other RGD containing peptides: the (351-364) fragment of the human vitronectin does not exhibit any activity. Moreover, the adhesion assays on acellular matrix showed that the (351-359)HVP (FRHRNRKGY) of sequence 2 is able to promote a complete covering with long-shaped osteoblast (spreading), while samples treated with fibronectin exhibited a less appealing result: few round-shaped osteoblasts. Furthermore, the nonapeptide sequence of this invention is not effective in the promotion of endothelial cell adhesion In order to get information on the secondary structure of the peptides under investigation NMR (nuclear magnetic resonance) and CD (circular dichroism) studies have been performed. These analyses allowed appreciating that different fragments of the vitronectin (sequence 1 and sequence 2) and different shapes of sequence 2 (linear and cyclic) present different conformations.

Example 7: conformational analysis of (351-359)HVP, (351-364)HVP, cyclic (351-359)HVP by NMR NMR analysis has been carried out in TFEZH7O 90:10 (v/v) at 11.7 T. For the DG/AMBER structural calculations a set of experimental constraints has been determined from NOE data. 100 conformers have been calculated by DYANA software via REDAC strategy (Guntert, P., et al. J. MoI. Biol. 1997, 273, 283-298) and the first 30 best structures have been processed by energetic minimization with AMBER 6.0 software (Case D.A., et al. AMBER 6, University of California, San Francisco, 1999). The first 10 best structures after AMBER refining have been selected to represent peptide conformations. This procedure has been followed for all the peptides investigated with the exception of cyclic (351-359)HVP: for this peptide molecular dynamic simulations under vacuum at 300 K for 210 ps have been performed with time steps of 0.5 fs using DISCOVER/INSIGHT software. The data registered during the last 50 ps of the simulations have been used for the statistical analysis. Model peptides show flexible conformations with segments 5-6 residues long bearing defined structural motifs. These segments exhibit extended and/or curved structures. The areas showing better structural definition implies central parts of the peptides where most of the basic residues are located (Figure X).

Example 8: CD conformational studies of the (351-359)HVP, (351-364)HVP, cyclic (351- 359)HVP into three different solvents: 10 mM phosphate buffer pH 7; 14 mM sodium dodecylsulphate in 10 mM phosphate buffer pH 7; 2,2,2-trifluoroethanoL CD spectra have been registered at room temperature by using a Jasco J710 apparatus. Quartz cylindrical cell with 0.1 optical length has been utilized. Spectra are expressed in molar ellipticity ([Θ ]R in deg x cm2 x dMol"1). Spectra have been registered between 250 and 180 nm with a band width of 2 nm and a resolution of 0.2 nm; then, the signal due to the solvent, measured under the same conditions, has been subtracted. Peptide solutions have been prepared by dissolving weighted amounts of each peptide in the minimum water volume and diluting the obtained solution with a solvent until reaching a final amount as of 98% (v/v). These conformational analyses (NMR and CD) demonstrated that the nonapeptide of this invention of sequence 2 (FRHKNRKGY) corresponding to the (351-359)HVP fragment has a secondary structure that is characterized by extended conformation close to the N-terminus and by a type I β-turn involving the 6th residue. The nonapeptide of sequence 2 (FRHRNRKGY) shows a different spatial organization of the basic residue side chains in comparison with the (351-364)HVP peptide of sequence 1. Two characteristics are appreciably identified: basic residues clustering on one side of the peptide and a higher spacing among charged side chains. This latter is probably necessary for the interaction with glycosaniinoglycans with a lower solvatation degree, such as heparin-sulphate. With regard to the specific adhesive properties of adult rat bone marrow osteoblasts, the nonapeptide (351-359) from vitronectin (human) of sequence 2 (FRHRNRKGY) containing the RHRNRK sequence and peptides and polypeptides from mammalian vitronectin containing the core RHRNRK sequence, can be usefully utilized for pathologies that therapeutically require osteoblast adhesion promotion such as in odontoiatric and orthopaedic field, in particular in implantology. For example, pathologies that could be treated with the investigated peptides and polypeptides are: cranio-maxillo-facial bone defects (odontoiatric implantology and reconstructive surgery), congenital defects, fractures of skeletal system (pre-implant and prosthetic surgery). For the above mentioned therapeutic aims, peptides and polypeptides of this invention can be included into pharmaceutical composition with eccipients, diluents, support materials pharmaceutically suitable even for the controlled release, and pharmaceutical compounds already known for this purpose. Without excluding any other eccipient, diluent, support material for the purpose, peptides and polypeptides of the invention can be usefully utilized as included or adsorbed in/or bound to polymeric materials such as nylon, polycaprolactames, polylactic acid, polylactic-polyglycolic acid, polyurethanes, polyacrylates, polypirrole, poly(DTE) carbonate, polysaccharide hydrogel, such as sodium alginate and or hyaluronic acid, silica-based and hydroxyapatite coatings, metals and composites to obtain biomaterials suitable for the therapeutic application (i.e., semi-solid gel, film, layer, tapes, sponge, tissues, and bioactive surfaces).