BACKGROUND OF THE INVENTION

1. Field of the Invention This invention relates to a periodontal ligament cell-attractant factor useful for periodontal and bone regeneration. The invention also relates to regeneration methods, compositions and materials.

2. Description of the related art Methods for bonding materials, such as a resin layer, to teeth and bone are described in United States Patents 4,382,792 granted May 10, 1983 and 4,600,383 granted July 15, 1986, both issued to D.C. Smith et al. A number of publications with a focus on regeneration of the periodontium have described biological factors and methods involved in the formation of a new connective tissue attachments to periodontally diseased or denuded root surfaces, such as the publications discussed as follows.

"Molecular Factors Determining Gingival Tissue Interaction with Tooth Structure", Victor P. Terranova et al. , Journal of Periodontal Research Vol., 17, 1982, pp. 530-533, describes that the attachment of epithelial cells and fibroblasts to tooth surfaces is stimulated by, respectively, the glycoproteins laminin and fibronectin.

A review of the function of polypeptide

growth factors and description of their potential role in periodontal therapy to promote migration of periodontal ligament cells to dentin, as well as to root surfaces preconditioned by treatment with tetracycline HC1 and/or citric acid, is given in: "Extracellular Matrices and Polypeptide Growth Factors as Mediators of Functions of Cells of the Periodontium", Victor P. Terranova et al., J. Periodontology, Vol. 58, No.6., June 1987, pp. 371-

380, the disclosure of which is incorporated herein by reference.

United States Patent 4,702,734, granted October 27, 1987 to V.P. Terranova et al., the disclosure of which is incorporated herein by refer¬ ence, discloses a method for promoting periodontal regeneration, which includes exposing an area of tooth surface, preconditioning the area with a tetracycline salt, applying fibronectin (FN) and optionally applying endothelial cell growth factor (ECGF) .

One essential biological event in tissue regeneration is specific cell directed movement (chemotaxis) . Chemotaxis is an essential feature of many biological processes in health and disease. A specific example of chemotaxis is the movement of endothelial cells in the process of neovascularization as disclosed in:

"Human Endothelial Cells are Chemotactic to Endothelial Cell Growth Factor and Heparin", Victor P. Terranova et al.. The Journal of Cell Biology, Vol.

101, December 1985, pp. 2330-2334. This publication, the disclosure of which is incorporated herein by reference, describes a chemotaxis assay.

Another example is the description of movement of tumor cells to form metastases in the publication:

"Migration of Tumor Cells of Organ-derived Chemoattractants", Er ki S. Hujanen et al. Cancer

Research, Vol. 45, August 1985, pp. 3517-3521.

The chemotactic properties of laminin, fibronectin and various polypeptides are described, and laminin is shown to promote chemotaxis and growth of human gingival epithelial cells, in:

"Chemotaxis of Human Gingival Epithelial Cells to Laminin", Victor P. Terranova et al, J. Periodontology, Vol. 57, No. 5, May 1986, pp.311-317. In the article:

"Biochemically Mediated Periodontal Regeneration", Victor P. Terranova et al.. Journal of Periodontal Research, Vol. 22, 1987, pp. 248-251, the chemotactic attraction of periodontal ligament cells to fibronectin and endothelial cell growth factor bound to dentin is disclosed. Chemotactic properties of other proteins and growth factors such as platelet derived growth factor, are also discussed. The article: "A Biochemical Approach to Periodontal

Regeneration", Victor P. Terranova et al., J. Periodontology, Vol. 58, No. 4, April 1987, pp. 247- 257, the disclosure of which is incorporated herein by reference, describes a new assay system that tests the ability of a number of proteins and polypeptide growth factors applied on dentin to stimulate a chemotactic and proliferative response from various cell types. The article:

"Use of a Reconstituted Basement Membrane to Measure Cell Invasiveness and Select for Highly

Invasive Tumor Cells", Victor P. Terranova et al., Proc. Natl. Acad. Sci. USA, Vol. 83, January 1986, pp. 465-469, describes the preparation and use of reconstituted basement membranes consisting of laminin and type IV collagen reconstituted onto a disk of type I collagen, which cannot be penetrated by fibroblasts or epidermal cells.

SUMMARY OF THE INVENTION The invention relates to an isolated periodontal ligament cell-attractant factor (PDL-CTX) , which comprises a protein obtainable from periodontal ligament (PDL) cells, said factor having chemo¬ attractant activity to periodontal ligament cells.

The invention also relates to a composition useful for periodontal regeneration which comprises a pharmaceutically acceptable amount of said PDL-CTX factor and a pharmaceutically acceptable medium. The invention further relates to a kit useful for periodontal regeneration, which includes said PDL-CTX factor. Further, the invention relates to improvements in a method of periodontal regeneration or of bone regeneration in which said PDL-CTX factor is applied to the surface to be regenerated.

In another embodiment, the invention relates to improvements in a method for inducing periodontal cell migration on dentin, in which said PDL-CTX factor is applied to the dentin.

In a further embodiment the invention relates to an improvement in a method for periodontal regeneration which comprises the steps of:

(a) obtaining a specimen of a parent population of periodontal ligament (PDL) cells from a patient;

(b) growing the obtained PDL cells in a tissue culture medium to obtain a culture of the patient's PDL cells;

(c) selecting a sub-population of PDL cells of said culture which migrate through a porous type I collagen barrier by chemotaxis directed against said PDL-CTX factor.

Still further, the invention relates to improvements in a method of periodontal regeneration in which periodontal ligament cells are applied to the

root surface of a tooth, which comprises covering the treated surface with an artificial basement membrane comprised of collagen. The invention also relates to a method of periodontal regeneration, which comprises:

(a) exposing a tooth surface to be treated;

(b) applying to said surface a pharmaceutically acceptable amount of said PDL-CTX factor in a pharmaceutically acceptable medium;

(c) applying a suspension of the patient's periodontal ligament cells to the surface; and

(d) covering the treated surface with an artificial basement membrane comprising type I collagen overlayered with type IV collagen and laminin.

BRIEF DESCRIPTION OF THE DRAWINGS Figs. 1 to 5 are graphical representations of experimental data illustrating the chemotactic attraction of PDL cells to the PDL-CTX factor.

Fig. 6 is a graphical representation of experimental data illustrating the chemotactic attraction of bone cells to the PDL-CTX factor. Fig. 7 is a graphical representation of experimental data illustrating the chemotactic attraction of parent and selected PDL cells to the PDL-CTX factor.

Fig. 8 is a graphical representation of experimental data illustrating ³ H-thymidine incorporation into DNA in parent and selected PDL cells.

Fig. 9 is a chromatogram of a protein- containing concentrate of PDL cell conditioned media precipitated with ammonium sulfate.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The repair of injury to the gums of humans

begins as soon as tissue damage occurs, and the release of polypeptide growth factors from injured cells and inflammatory cells is a critical part of this process. Many of the polypeptide growth factors have been shown to be involved in tissue repair. For example, platelet derived growth factor (PDGF) and transforming growth factor-β (TGF-β) are two polypeptides that have been shown to have an important role. PDGF is initially released from the alpha- granules of the platelets and is a potent mitogen for fibroblasts in the presence of either transforming growth factor-α (TGF-α) or epidermal growth factor (EGF) . Furthermore, PDGF stimulates the production of collagenase by fibroblasts and thus contributes to the remodeling of matrix components, an essential feature of tissue repair. TGF-β appears to have a particu¬ larly important role in the repair process. This peptide is found in relatively high concentrations in platelets, in activated T lymphocytes, as well as in macrophages. TGF-β stimulates the formation of collagen in human or rodent fibroblasts and when injected subcutaneously in newborn mice causes rapid fibrotic and angiogenic response at the site of injection. Another recently discovered source of TGF- β is bone. TGF-β is present in bone in amount of almost 100 fold greater than found in other soft tissues. In vitro studies indicate that TGF-β can control the effects of several other polypeptide growth factors including PDGF, TGF-α, EGF, AFGF (acidic fibroblast growth factor) , BFGF (basic fibroblast growth factor) and IL-2 (interleukin-2) . To fully understand the mechanism of all of these peptide growth factors it must be realized that they act in sets or combination in which each peptide modulates the effects of the next.

Although there may be autocrine action of these peptides in injured cells, it would appear that

their paracrine action, driven by their production and release by various inflammatory cells, accounts for the key role of these peptides in the repair process. However, autocrine function may play an important role as has been demonstrated in traumatized cultures of artereal smooth muscle cells which synthesize and release peptides that resemble PDGF. The function of many of these known peptides in relationship to cells of the periodontium is presently under investigation. In addition, other factors isolated from cementum have the potential to be mitogenic for cells of osseous origin.

One of the least understood aspects of tissue repair is the biochemical factors that control specific cell motility. To expect a unique autocrine factor that could be responsible for initiation of this event is not unreasonable. Recently, an "autocrine motility factor" (AMF) has been identified for melanoma cells. The factor was found to be unique, based on a ino acid analysis and has a molecular weight of 55,000. Its action may be associated with membrane changes in phospholipid methylation. Similar membrane changes have been implicated in the motility of leukocytes. In other systems marked increases in phospholipid methylation have been observed in hormonal effects upon hepatocytes and bladder epithelial cells.

A newly developed assay system (described in the above-mentioned publication "A Biochemical Approach to Periodontal Regeneration" of V.P» Terranova et al.) enables testing of the potential activity of various biological response modifiers on a dentin substrate. The assay systems is divided into two parts. ASSAY I allows the measurement of the chemotactic activity of the test substance bound to dentin. Here the cells must actively move through a filter ("Nuclepore" [registered trademark, 100 μm

thick, 8 ym pore diameter) in response to biological response modifiers bound to dentin. In ASSAY II the ability of the dentin-bound factors to stimulate directed movement and proliferation of periodontal tissue cells on dentin surfaces is measured. Using these assays it was demonstrated that PDL cells migrate to FN, ECGF and AFGF. In addition, FN, ECGF and FGF induce a proliferative response in PDL cells grown on surface-demineralized dentin. Gingival epithelial cells were shown to migrate to LM (laminin) and LM fragments. LM was also shown to stimulate gingival epithelial cell proliferation on native dentin surfaces. These in vitro findings suggest that biological conditioning of the root surface (dentin) may enhance mesenchymal cell attachment and proliferation. These events subsequently lead to an improved healing after periodontal reconstructive surgery. There is a need to develop more potent and selective chemoattractant factors for PDL cells and cells of osseous origin (bone cells) , and for improved methods for the treatment of root surfaces to selectively enhance PDL cell repopulation of the previously diseased root surface.

The PDL-CTX factor of the present invention is a potent new PDL cell and bone cell attractant factor which is a protein isolated from, derived from or obtainable from PDL cells. It may also be obtained by recombinant DNA methodology or lay peptide synthesis, by techniques known for such methodology or synthesis.

It has been found that the PDL-CTX factor of the present invention is a protein which is charac- terized by periodontal ligament cell autocrine motility and mitogenic activity. The protein factor may be recovered from a medium conditioned by PDL cells (the medium obtained after being used in the

tissue culture of the PDL cells) . A concentrate of the protein may be recovered as a precipitate by treatment of said medium with a salt such as a ammonium sulfate.

An especially useful PDL-CTX factor was derived from periodontal ligament cells which were selected from a plurality of periodontal ligament cells, the selected periodontal ligament cells having been selected on the basis of an increased chemotactic response to said factor relative to the chemotactic response of the other periodontal ligament cells of said plurality of cells.

The PDL-CTX factor has been found to comprise a protein having a molecular weight of 45,000 to about 55,000 daltons. Upon work-up further the protein was found to have a molecular weight of about 50,000 to about 53,000 daltons. Further, the protein of the PDL-CTX factor is believed to be a tetramer. Thus the useful factor may comprise a monomer which forms said tetramer, said monomer having a molecular weight of about 12,500 daltons.

The PDL-CTX factor of the present invention has been characterized by an investigation the results of which are summarized in the drawings which are explained further in the following discussion.

The graphical representation of data in Fig. 1 shows the directed migration of PDL cells to conditioned media (50%) after incubation at various temperatures, where migration is assayed using ASSAY I and each assay point represents the mean +/- s.d. (standard deviation) of triplicate samples.

The data represented in Fig. 2 shows directional migration of PDL cells to 50% conditioned media incubated with trypsin for various times where migration is assayed using ASSAY I and each assay point represents the mean +/- s.d. of quadruplicate samples.

In Figure 3, the graphical representation of data shows directed migration of PDL cells, human gingival epithelial cells and human gingival fibroblasts to various concentrations (dilutions) of PDL cell conditioned media where migration is assayed using ASSAY I and each point represents the mean +/- s.d. of triplicate samples.

Figure 4 is a graphical representation of data which shows the effect of various concentrations of TGF-α and TGF-β on the chemotaxis of PDL cells where migration is assayed using ASSAY 1, all factors are allowed to bind to TTC conditioned dentin (dentin preconditioned in 100 mg/ml of tetracycline HC1) for 30 minutes at 37 ^e C in 100% humidity, PDL cell conditioned media is used as a control and each assay point represents the mean +/- s.d. of triplicate samples.

Data represented in Fig 5 show the effect of various antibodies against extracellular matrix components on migration of PDL cells to 50% PDL cell conditioned media where antibody dilutions are incubated with the PDL cells above the Nuclepore filter, migration is assayed using ASSAY I and each assay point is the mean +/- s.d. of triplicate samples.

The graphical representation of data in Fig. 6 shows directed migration of human bone cells (ATCC 7009) to 50% PDL cell conditioned media or TGF-β where migration is assayed using ASSAY I, both factors are allowed to bind to previously TTC conditioned dentin for 30 minutes at 37°C in 100% humidity and each assay point represents the mean of duplicate samples not differing by more than 10%. In Fig. 7, the graphical representation of data shows directed migration of parent and selected PDL cells to various dilutions of conditioned media where selection of PDL cells is accomplished by the

methodology established in the above mentioned publication "A Biochemical Approach to Periodontal Regeneration" of V.P. Terranova et al.), migration is assayed using ASSAY I and each assay point represents the mean +/- s.d. of quadruplicate samples.

The data represented in Fig. 8 shows ³ H- thymidine incorporation into DNA in parent and selected PDL cells wherein the cells are cultivated for 10 days and then labelled for 24 hours with ³ H-TdR 50 μCi/ml in media containing FN free serum, the cells are removed with 0.1% ethylenediaminetetraacetic acid (EDTA) , 0.1% ethylene glycol bis(β-amiήoethyl ether)- N,N,N',N'-teteraacetic acid (EGTA) in a divalent cation free PBS (phosphate-buffered saline) , ³ H- radioactivity incorporated into the cell is quantitated using an LKB scintillation counter and the values represent the mean +/- s.d. triplicate assays with background incorporation subtracted. The chromatogram reproduced in Fig. 9 is a profile resulting from PDL cell conditioned media being subjected to (NH ₄ ) ₂ S0 salting out procedure, subjecting a concentrate containing the precipitated protein to TSK55F (molecular sieve column) open column chromatography, and then subjecting the eluted protein having increased activity to reverse-phase high performance liquid chromatography (HPLC) .

Our studies indicate that there is a specific factor synthesized by PDL cells that is chemotactic for these cells. When conditioned media from cultures of human PDL cells was examined for chemoattractant activity, we found a dose dependent relationship for PDL cells. Other data suggested that a specific polypeptide was responsible for the chemotactic activity in the PDL cell conditioned media. Heating the media to 100° for 30 min decreased in chemoattractant activity as did a 60 min incubation with trypsin, indicating a polypeptide nature for the

responsible factor. Gingival fibroblasts and epithelial cells did not respond to various concentrations of the PDL cell conditioned media indicating the response is most likely not due to other polypeptide growth factors acting on populations of these cells that may be present in the PDL preparations. Moreover, antibodies against FN and LM as well as ECGF, FGF, or EGF did not inhibit the chemotactic properties of the conditioned media for both PDL and osseous cells. TGF- enhances the activity of the PDL cell conditioned media but antibodies directed against TGF-beta did not inhibit the PDL factor-mediated chemotactic response in either cell type. HPLC Reverse Phase (C18 followed by C8) chromatography has enabled us to isolate a 12,500 dalton peptide with high chemotactic and mitogenic activity. Additional data from open column chromatography and HPLC support the peptide nature of this material. Amino acid analysis and preliminary

NH ₂ terminal sequence data indicate a unique peptide. Information from the NIH (U.S. National Institutes of Health) data banks indicates that there is no homologous sequence to the peptide. Observations indicate that PDL cell recolonization of the root surface is necessary for regeneration, see Nyman, S. et. at. (1982) , J. Clin Periodontol 9:290, and Nyman, S., et. al. (1983): Textbook of Clinical Periodontolgy: Lindhe, J. (ed.), pp. 410-432, Munksgarrd, Copenhagen. In view of this, it is hypothesized that specific PDL cell generated factors play a significant role in the migration and proliferation of a subpopulation of responsive PDL cells and can contribute to the healing of this hard- soft tissue interface. These results should aid in the development of therapies that could enhance the biological processes involved in the regeneration of the periodontal ligament as well as the healing a

other hard-soft tissue interfaces.

Briefly, our studies have shown the following: (1) A chemotactic factor for PDL cells is present in the conditioned media of cultures of PDL cells. (2) This factor is heat labile; 100° for 30 minutes. (3) This factor is trypsin sensitive.

(4) Gingival fibroblasts are refractive to this factor.

(5) Gingival epithelial cells are refractive to this factor. (6) Antibody against FN or LM does not inhibit the chemotactic properties of this factor. (7) Antibody against ECGF, FGF or EGF does not inhibit the chemotactic properties of this factor. (8) TGF-α and TGF-β do not stimulate chemotaxis of PDL cells to the same degree as does the factor. (9) A subpopulation of PDL cells can be isolated that are more chemotactically responsive to the factor.

Further data indicate the following: (1A) When the 95% (NH ₄ ) ₂ S0 ₄ precipitate is chromatographed using a mono P column with a pH gradient from 9.5 to 7.0 the material elutes at a pH 9.3.

(2A) The active material from the FPLC is rechromatographed on a C ₈ reverse phase column a single peak of high biological activity is observed with a retention time of 12 min.

(3A) SDS-Urea-PAGE of the peak yields, a single band Mr=12,500. (4A) This peptide has high biological activity

for PDL cells. Both chemotaxis and proliferation are stimulated. (5A) A ino Acid analysis using Pico Tag system is given.

(6A) Sequence Analysis for the first eleven NH ₂ terminal amino acids when checked at the NIH data bank indicates a unique peptide. The connective tissue of the periodonitium is a complex structure consisting of fibroblasts (gingival connective tissue fibroblasts and periodontal ligament fibroblasts) , gingival epithelium, vascular endothelial cells, nerve processes, cementum which is comprised of cementoblasts with associated extracellular matrix, alveolar bone and an extensive extracellular matrix comprised of collagen, glycoproteins and proteoglycans. Essential biological processes involved in periodontal regeneration include chemotaxis, proliferation and differentiation of cells and structures such as blood vessels, a new connective tissue attachment of the tooth to the alveolar process and the epithelial seal to the tooth.

PDL-CTX is believed to have an amino acid sequence from the N-terminal as shown in the following Table A which also, indicate sequence data of known polypeptide growth factors for comparison.

Accordingly, the present invention further relates to an isolated periodontal ligament cell- attractant protein obtainable from human periodontal ligament cells, said protein comprising an initial amino acid sequence, from the N-terminal, as follows: Val Pro Asp Ser Ser Ala His Lys Lys Ala ... .

The present invention further relates to an isolated periodontal ligament cell-attractant protein obtainable from human periodontal ligament cells, said protein comprising an amino acid sequence, ending at the C-terminal, as follows:

. . . Pro Val Val Pro Ala Tyr Ala Pro Pro . The present invention further relates to an isolated periodontal ligament cell-attractant protein obtainable from human periodontal ligament cells , said protein having an amino acid sequence, from the N-terminal to the C-terminal , substantially as follows :

Val Pro Asp Ser Ser Ala His Lys lys Ala Tyr Leu Gin Met Val Pro Gly Gly Asn lie Gly Ser Phe Val Asp Try His Cys Thr Asn Lys Gly Gly Gly Trp Phe Ala Lys Asp Pro Gly Pro lys His Cys Asp Pro Gly Tyr Gly Val Ala Phe Trp lie Met Ala His Lys Asn Gly Pro Ser Pro Val Asp Val Gly His Leu Arg Tyr Val Val Leu Pro Ser Trp Val Asp Pro Ala Gly Pro Trp Leu Ala His Lys Ser Pro Ala Gly Val Ala Ala lys Ala Gly Gly Pro Val Val Pro Ala Tyr Ala Pro Pro.

TABLE A

PDIrCTX AttTNO ACID SEQUENCE AND COMPARISON TO KNOWN POLYPEPTIDE GROWTH FACTORS β-NGF NH ₂ Ser Ser Ser His Pro Val Phe His Arg Gly Glu Phe Ser Val Cys Asp Ser He Ser Val (bovine)

Basic FGF NH ₂ Pro Ala Leu Pro Glu Asp Gly Gly Ser Gly Ala Phe Pro Pro Gly His Phe Lys Asp Pro (bovine)

Acidic FGF NH ₂ Phe Asn Leu Pro Leu Gly Asn Tyr Lys lys Pro ys Leu Leu Tyr cys Ser Asn Gly Gly (bovine)

EGF NH ₂ Asn Ser Asp Ser Glu Cys Pro Leu Ser His Asp Gly Tyr Cys Leu His Asp Gly Val Cys (human)

TGF-α NH ₂ Val Val Ser His Phe Asn Asp Cys Pro Asp Ser His Thr Gin Phe Cys Phe His Gly Thr (human)

TGF-β NH Ala Leu Asp Thr Asn Tyr Cys Phe Ser Ser Thr Glu Lys Asn Cys cys Val Arg Gin Leu (human) I

IL-2 NH ₂ Ala Pro Thr Ser Ser Ser Thr Lys Lys Thr Gin Leu Gin Leu Glu His Leu Leu Leu Asp ϊ (human)

PDGF-A NH Ser He Glu Glu Ala Val Pro Ala Val Cys Lys Thr Arg Thr Val He Tyr Glu lie Pro l (human)

PDGF-B NH Ser Leu Sly Ser Leu Thr He Ala Glu Pro Ala Met He Ala Glu Cys Lys Thr Arg Thr (human)

PDL-CTX NH ₂ Val Pro Asp Ser Ser Ala His Lys Lys Ala Tyr Leu Gin Met Val Pro Gly Gly Asn lie (human) Gly Ser Phe Val Asp Try His Cys Thr Asn lys Gly Gly Gly Trp Phe Ala Lys Asp Pro Gly Pro Lys His Cys Asp Pro Gly Tyr Gly Val Ala Phe Trp He Met Ala His Lys Asn

Gly Pro Ser Pro Val Asp Val Gly His Leu Arg Tyr Val Val Leu Pro Ser Trp Val Asp Pro Ala Gly Pro Trp Leu Ala His Lys Ser Pro Ala Gly Val Ala Ala Lys Ala Gly Gly

Pro Val Val Pro Ala Tyr Ala Pro Pro _*

The PDL-CTX factor of the invention is isolated by being separated from a natural or synthetic source, or concentrated or purified. Thus, the PDL-CTX factor of the invention can be in the form of a concentrate of said factor, e.g. a concentrated aqueous solution or dispersion of said factor. The PDL-CTX factor can suitably be obtained by tissue culture of PDL cells in a customary medium, following which the medium contains the PDL-CTX factor. Media containing the factor are referred to as PDL cell conditioned media. Upon separation media may be concentrated, for example by chromatography, precipitation (e.g. by means of salting out with a salt such as ammonium sulfate) , centrifugation, electrophoresis, or combinations of these techniques. A preferred procedure employs a sequence of (1) salting out with ammonium sulfate; (2) centrifugation and discarding the supernatant; repeating (1) and (2) ; (3) resuspending sediment; (4) fractionating by chromatography; (5) fractionating the eluate containing PDL-CTX by chromatography.

The separated or PDL-CTX factor or concen¬ trate thereof is a component of the composition of the invention. A pharmaceutically acceptable amount of the PDL-CTX factor will vary depending upon the intended use, and depending upon the activity of the concentrate. The acceptable amount also may be in an acceptable dosage form for application to teeth or to bone. The acceptable amount of the factor is-combined with a pharmaceutically acceptable medium to form the composition of the invention.

It is desirable to include in the compo¬ sition of the invention a pharmacologically acceptable amount of a polypeptide growth factor. The growth factors may be selected from the ones described where they enhance regeneration in combination with the PDL- CTX factor. The growth factor suitably may consist of

transforming growth factor-β (TGF-β) , platelet derived growth factor (PDGF) , and mixtures thereof.

Preferably the medium of the composition is aqueous. A suitable amount of the PDL-CTX factor is an amount of from about 10 picograms to about 10 micrograms per ml of the composition.

The composition may further comprise TGF-β in an amount of from about 10 picograms to about 10 micrograms per ml of said composition. Alternatively the composition may comprise both TGF-β and PDGF, with TGF-β in an amount of from about 10 picograms to about 10 micrograms per ml of said composition, and PDGF in an amount of about 10 picograms to about 10 micrograms per ml of said composition.

A kit in accordance with the present inven¬ tion comprises as individual components:

(a) a pharmaceuticaly acceptable amount of the PDL-CTX factor of the invention in a pharmaceu- tically acceptable medium; and.

(b) a pharmacologically acceptable amount of a polypeptide growth factor in a pharmacologically acceptable medium.

The polypeptide growth factor of component (b) may be selected form the group consisting of TGF-α, TGF-β, PDGF, and mixtures thereof.

The kit may further include as an individual component:

(c) a pharmacologically acceptable amount of a polypeptide growth factor different from the polypeptide growth factor of component (b) , in a pharmacologically acceptable medium. Component (c) may be selected from the same growth factors as is component (b) , but will differ from the particular growth factor which was selected from component (b) .

The media of components (a) , (b) and (c) preferably is aqueous. The active factors of (a) , (b)

and (c) may be present in the same amounts as speci¬ fied above in relation to the composition of the present invention. Further, the kit may include an agent which demineralizes enamel and/or dentin, to pre-condition the tooth surface and to provide enhanced binding of components (a) to (c) , and improved regeneration. Preferably the further agent included is: (d) an aqueous solution of tetracycline HC1, preferably in a concentration of from 50 mg/ml to 100 mg/ml; or

(d-1) an aqueous solution of citric acid. Most preferably, the further agent is tetracycline HC1.

The kit may also include, if desired, directions for use, and appropriate materials for the use, such as syringes and materials for suturing and dress the treated area. The kit of the invention containing compo¬ nents (a)-(c) is also useful for bone regeneration.

Another embodiment of the present invention resides in the use of an artificial basement membrane comprised of collagen to cover the treated surface of a tooth in a method of periodontal regeneration.

Following therapy which can include treatment of an exposed, periodontally diseased root surface with chemotactic factors, the treated surface is covered with the membrane. This membrane acts as a selective barrier which inhibits cells which do not display a strong chemotactic response to the factor from competing for adhesion and growth on the treated root surface with cells, such as the PDL cells, that are effective in the formation of a new connective tissue attachment to the treated root surface.

A preferred artificial basement membrane comprises type I collagen overlayered with type IV collagen and laminin. In the method of using the

membrane, preferably the root surface is treated with the PDL-CTX factor of the present invention, and desirably said root surface is additionally treated with a polypeptide growth factor selected from the group consisting of TGF-α, TGF-β, PDGF and mixtures thereof.

The artificial basement membrane also may be used in the kit of the present invention as a further component.

The PDL-CTX factor of the present invention is generally useful in a method of periodontal or bone regeneration. It is particularly useful for applying to dentin in a method directed at inducing PDL cell migration to the dentin. In such methods, a particu¬ larly preferred technique involves treatment of the diseased root with PDL cells selected from the patient's cells on the basis of an increased response to the PDL-CTX factor of the invention. Thus, another embodiment of the present invention relates to a technique for selection of PDL cells in connection with a method of periodontal regeneration. In this embodiment, the steps comprise:

(a) obtaining a specimen of a parent population of periodontal ligament (PDL) cells from a patient;

(b) growing the obtained PDL cells in a tissue culture medium to obtain a culture of the patient's PDL cells; and (c) selecting a sub-population of PDL cells of said culture which migrate through a porous type I collagen barrier by chemotaxis directed against the PDL-CTX factor of the present invention.

In this technique of selecting PDL cells, it is preferred that before step (c) , an enriched population of PDL cells of the culture of step (b) is obtained by incubating the PDL cells of said culture in a first compartment of a chamber having a second

compartment containing a solution comprising the PDL- CTX factor, the first and second chambers being separated by a semipermeable membrane, selecting an enriched sub-population of cells which migrate by chemotaxis directed against said PDL-CTX factor, and using said enriched sub-population to further select therefrom cells which migrate through the collagen barrier in step (c) . In a further embodiment of the present invention, a preferred therapeutic procedure for periodontal regeneration comprises:

(a) exposing a tooth surface to be treated;

(b) applying to said surface a pharmaceutically acceptable amount of the PDL-CTX factor of the present invention in a pharmaceutically acceptable medium;

(c) applying a suspension of the patient's periodontal ligament cells to the surface; and (d) covering the treated surface with an artificial basement membrane comprising type I collagen overlayered with type IV collagen and laminin.

It is further preferred in this therapeutic procedure that before or simultaneously with step (b) there is additionally applied to said surface a pharmaceutically acceptable amount of polypeptide growth factor selected from the group consisting of TGF-α, TGF-β, PDGF, and mixtures thereof. Optionally, before step (b) there is applied to said surface an aqueous solution of tetracycline HC1 or an aqueous, saturated solution of citric acid, the application of tetracycline HC1 being the preferred one.

The invention is further illustrated in the following examples.

EXAMPLE 1 When conditioned media from cultures of human PDL cells are examined for chemoattractant

activity, it is found that there is a dose dependent relationship for both PDL cells and cells of osseous origin as shown in the following Table I.

TABLE I

No. of Cells/Hiσh Power Field

% PDL

Conditioned

Media PDL Cells Bone Cells

0 8 +/" 3 7 +/" ²

1 16 +/- 4 19 +/" 3

5 22 +/" 5 24 +/" 5

10 25 +/- 5 28 +/- 5

25 30 +/- V 39 +/- 6

50 29 +/- 6 38 +/- 5

100 25 +/- 4 35 +/- 5

Assay conditions are 37 ^β C, 100% humidity and a 6 hour incubation using the modified Boyden Chamber Assay System. The barrier in this assay is an 8 ym pore Nuclepore filter (polyvinyl parolodine free) . For both PDL cells and bone cells the filters are precoated with gelatin at a concentration of 5 μg/filter. In addition, when ASSAY I (assay for specific cell migration) is utilized, conditioned media adsorbed to dentin produces a dose dependent response for both PDL cells and bone cells, as shown in the following Table II.

TABLE II

No. of Cells/Hiσh Power Field

% PDL

Conditioned

Media PDL Cells Bone Cells

0 3 +/- 3 1 +/- 1

1 10 +/" 2 15 +/" 3

5 20 +/- 4 26 +/- 4

10 34 +/- 5 39 +/- 9

50 35 +/- 6 39 +/" 10

Assaying conditions were TTC conditioned dentin (50 mg/block for 5 minutes at 37 ^β C) with 5 μl

application of conditioned media for 30 minutes at 37°C and 100% humidity. In addition, when the TTC condi¬ tioned dentin blocks are further conditioned by the application of FN (50 ug/block) a small but not insignificant increase in PDL cell motility is observed.

EXAMPLE 2 When 50% conditioned media (100% conditioned media diluted 1 to 1 with MEM [minimal essential media]) is incubated for 30 minutes at various temperatures, loss of the factor's chemoattractant activity is noted at higher temperatures (Fig. 1) .

Since sensitivity to heat is an indicator of protein character, trypsin sensitivity of the condi- tioned media is tested. Fifty percent conditioned media is incubated with purified trypsin 10 ug/ml for various time points. Incubation is stopped by the addition of SBTI (soybean trypsin inhibitor) at a 10 fold excess. Chemoattractant activity is decreased over a 60 minute incubation with trypsin (Fig. 2) . SBTI by itself had no effect (either positive or negative) on PDL cell migration.

The above data indicate that the factor is of a protein (glycoprotein, lipoprotein) nature. Fifty percent conditioned media is then dialyzed against MEM, 6 changes at 4*C. Molecular weight cut off the dialysis tubing is 6,000. The activity is retained and is nondializable above M _r = 6,000.

EXAMPLE 3 The ability of other cells of the periodontium to respond to 50% conditioned media of PDL cells is evaluated. Both human gingival epithelial cells and human gingival fibroblasts are tested for their ability to respond to various dosages of conditioned media (Fig. 3) . As positive controls, LM is tested at 10 μg/block for gingival epithelial cells and FN is tested at 100 g/block for gingival fibroblasts. Neither gingival epithelial cells nor gingival fibroblasts respond to

conditioned media of PDL cells.

EXAMPLE 4 In addition, it was previously shown that anti-ECGF inhibits ECGF induced chemotaxis of PDL cells, As a further indication that PDL-CTX of the present invention is a unique PDL cell synthesized factor, the ability of anti-ECGF, anti-FGF and anti-EGF to inhibit 50% conditioned media in a dose dependent manner is examined. No effect of any antibodies on PDL cell movement is altered. Non-specific IgG (NSIgG) is also evaluated, see the following Table III.

TABLE III Number of Cells/Hiσh Power Field

Antibody

Dilution Anti-ECGF Anti-FGF Anti-EGF NSIσG

0 38 +/" 4 41 +/- 40 +/" 4 39 +/~ 4

10- ⁶ 37 +/- 6 40 +/- 4 41 +/" 3 40 +/" 5

1100--4* 4411 ++//-- 55 39 +/- 3 39 +/" ⁵ 41 +/" 4

10-2 43 +/- 4 39 +/" 40 +/- 4 40 +/" ⁵

10-1 39 +/- 4 39 +/- 4 41 +/- ³ 38 +/" ⁴ In these assays the migratory response of PDL cells is measured in the presence of antibody directed against various polypeptide mitogens to 50% PDL cell conditioned media. Antibody dilutions (initial concentrations all adjusted to start with 100 mg protein per ml media) are incubated with the PDL cells above the Nuclepore filter. Migration is assayed using ASSAY 1. Each assay point is the mean +/- s.d. of triplicate samples.

Since TGF-β is found in high level in bone and since both TGF-α and TGF-β are implicated in both fibroblast proliferation and chemotaxis, the ability of TGF-α and TGF-β and 50% conditioned media to generate a chemotactic response in PDL cells is compared. A dose dependent chemotactic response of PDL cells to 50% PDL cell conditioned media, TGF-α and TGF-β is observed. PDL cell conditioned media is significantly more

effective as a chemoattractant (Fig. 4) .

EXAMPLE 5 Previously it was shown that affinity purified polyclonal onospecific antibodies inhibit specific factor induced chemotaxis. The ability of anti-LM and anti-FN to inhibit conditioned media induced chemotaxis of PDL cells is therefor examined. The addition of antibody (initial concentration of 100 μg protein/ml) in a dose dependent (logarithmic) fashion does not inhibit 50% conditioned media induced chemotaxis (Fig. 5) .

EXAMPLE 6 Since TGF-β is believed to be a "panregulin" the ability of 1 and 10 ng/block TGF/β ti augment 50% PDL cell conditioned media (CM) is evaluated. In this example, dentin blocks are conditioned with 50 mg/ml TTC for 5 minutes at 37 ^e C, rinsed in cold PBS (6 x 1 minute each rinse) then both TGF-β and 50% conditioned media applied in 5 ul aliquotes. For both 1 and 10 ng/block of added TGF-β an increase in PDL cell migration is observed, as shown in the following Table IV.

TABLE IV Number of Cells/Hiσh Power Field 50% CM lnσ TGF-β lOnσ TGF-β CM+lnσ TGF-β CM+ lOnσ TGF-β 38 +/"4 23+/- 3 25 +/~ 49 +/" ^{5 53} +/" ⁶

This data supports the observation that TGF-β acts in conjunction with other polypeptide growth factors. In addition, antibody directed against TGF-β inhibits the TGF-β induced chemotactic response but not the conditioned media induced chemotactic response. .

Similar data with bone cells ATCC (American Type Culture Collection) 7009 and ATCC 7051 has been obtained. It has been found that:

(1) Bone cells respond to 50% PDL cell conditioned media. The response is dose dependent from 1% to 50%. These assays are performed using both ASSAY I and using the Modified Boyden Chamber Assay. The chemotactic response generated by PDL cell conditioned

media on bone cells is somewhat higher than that observed with PDL cells (see above) .

(2) Antibody studies indicated that the factor in PDL cell conditioned media is not ECGF, FGF, EGF,

TGF-β, fragments of LM or FN.

(3) A major difference noted using ASSAY I and the Modified Boyden Chamber Assay between PDL cells and bone cells is that bone cells respond more actively to TGF-β than do PDL cells.

When ATCC 7009 bone cells are evaluated to respond to TGF-β a dose dependent rapid migration is observed (Fig. 6) .

EXAMPLE 7 An assay system that can be utilized to separate subpopulations of cells based on their migratory capacity (invasive propensity) has been described (see the above-mentioned article "Use of a Reconstituted Basement Membrane to Measure Cell Invasiveness and Select for Highly Invasive Tumor Cells" of V.P.

Terranova et al) . In this assay system a collagen barrier separates the upper and lower chambers of a Modified Boyden Chamber. A collagen barrier is placed directly over a type IV collagen coated Nuclepore filter (1 μm pore diameter). Cells that traverse the collagen barrier attach to the coated filter. After various times of incubation the filters are removed and the attached cells subcultured. These observations have been extended to generate populations of human PDL cells which have the phenotypic characteristic of increased chemotaxis toward PDL cell conditioned media. When these subpopulations are examined for their ability to migrate and incorporate ₃ H-TdR into DNA, it is observed that they migrate more rapidly and have a higher rate of ₃ H-thymidine incorporation than do the parent population (Fig. 7 and 8) .

EXAMPLE 8 (NH ) ₂ S0 ₄ precipitation at 60% of the PDL cell

conditioned media results in a 100 fold enrichment of the factor.

Application of 5 ml of the (NH ₄ ) ₂ S0 precipitate to a molecular sieve column TSK55F equilibrated in 0.05 M Tris, 0.15 M NaCl. pH 7.4, results in elution of a pool of protein(molecular weight range 45,000 to 55,000) with 1000 fold increased activity. This material is applied to a reversed phase c-18 HPLC column and eluted with a gradient of water/acetonitrile (AN) starting with 100% water, 0% AN, reaching 0% water, 100% AN in a retention time of 55 minutes) . The active protein is eluted in 24.77 minutes retention time, equivalent to a molecular weight of about 45,000 (see Fig. 9). Also, when the material is applied to a molecular sieve HPLC column (protein pack. 125) , there is obtained a band with a retention time of 11.56 minutes for a protein having a molecular weight of 53,000. This shows on SDS-PAGE (sodium dodecyl sulfate- polyacrylamide gel electrophoresis) analysis with a M _r of 51,000. These analysis characterize the active PDL- CTX material.

EXAMPLE 9 Artificial basement membranes are prepared as described in the above-mentioned article "Use of a Reconstituted Basement Membrane to Measure cell Invasiveness and Select for Highly Invasive Tumor Cells", of V. P. Terranova et al. Type I collagen is cross-linked such that on formation of a 13 mm disc the pore size is no more than 5 ym. This type I collagen matrix is then overlayered with 200 μg of type XV collagen and 200 yg of laminin dispersed in 0.05 M Tris, 0.15 M NaCl, pH 7.4. After lyophilization, the artificial basement membranes are sterilized (gas sterilization) and packaged in sterile air tight plastic wrap for subsequent use.

EXAMPLE 10 A suspension of PDL cells in PBS is applied to

teeth using extracted human teeth as a model. 5 x 10 ⁶ selected PDL cells in PBS are applied by use of a pipet in a drop-wise fashion. The cells are allowed to attach to the root for 30 minutes. After this time period, the teeth are trypsinized and the number of attached cell quantitated by the use of a cell particles counter. Routinely, 50% attachment of dispersed PDL cells to tooth root structure isgenerated (2.5 x 10 ⁶ cells per tooth root) .

EXAMPLE 11 PERIODONTAL REGENERATION PROCEDURE A. Obtaining and culturing patients' PDL cells:

During the initial periodontal visit Widman-type flaps are raised in the gums of a patient and scrapings from around the junction of the tooth and bone are taken. The scrapings contain PDL cells. These scrapings are cultured for growth of PDL cells as follows. After washing in cold PBS, pH 7.4, the scrapings are added to media containing collagenase-dispase at 100 mg/ml in 10 ml of an isotonic salt solution (ISS) containing 100 mM NaCl, 60 mM mannitol, 25 mM Hepes, 10 mM NaHC0 ₃ , 6 mM K ₂ HP0 ₄ , lmM CaCl ₂ , pH 7.4. Following incubation for 90 minutes at 37 'C, the contents are vigorously vortexed for two minutes. The medium is removed and centrifuged at 500 x g at 4 ^β C for 3 minutes. The resulting cell pellet is resuspended and washed 3 times in DMEM Dulbecco modified minimal essential melin) with 500 μg/ml gentamycin. The resulting cells ar then added to type I collagen and FN coated (300 μg/dish respectively) 35mm tissue culture dishes. The culture medium consists of Media NCTC 109 supplemented with 15% fetal bovine serum (FBS) , 1% sodium pyruvate, 1% non- essential amino acid and 25 μg/ml gentamycin. Unattached cells are removed after 24 hours of incubation by decanting the medium and fresh medium is added and changed daily. After confluency is obtained

(14 to 21 days), the cells are removed from tissue culture dishes by incubation with 0.1% EDTA and 0.01% trypsin in divalent cation-free PBS for 5 minutes. The cells are then grown on the type I collagen-FN coated tissue culture dishes in the presence of 1 μg/ml anti-LM antibody. This treatment inhibits all epithelial cell adherence. The medium is changed after 6 hours to NCTC 109 supplemented with 15% FBS. Confluent cultures of fibroblast-like cells (PDL cells) are obtained in 14 to 21 days.

B. Selecting PDL cells having increased chemotactic response to PDL-CTX factor:

After obtaining cultures of the patient's PDL cells, the most responsive cells are isolated utilizing the following selection technique.

Twenty-five Modified Boyden Chambers are seeded with 5 x 10 ⁵ PDL cells in the upper chamber. After 6 hours of incubation with chemotaxis directed against 10" ⁹ M PDL-CTX, the cells that migrate all the way through the Nuclepore filter are collected and sub-cultured. This is accomplished as follows.

The Nuclepore filters are carefully removed and placed top-side down on sterile glass slides. The under-surface is then carefully scraped using a sterile rubber policeman. Ten μl of NCTC 109 with 10% FBS is next applied to the surface and gently aspirated into a pasteur pipette. These 10 μl samples with removed PDL cells are transferred to a 35 mm tissue culture dish which has previously been coated with 10 μl of type I collagen and 10 μg of FN. The media volume is brought up to 5 ml and the dish immediately anchored at 37" C, 5% C0 ₂ for 120 minutes. After this time, the unattracted cells are removed by decanting the medium, fresh media added, and the dish returned to the incubator. The culture dishes are examined daily and cell confluency is expected after 18 to 24 days. When the dishes are confluent, the cells are removed using

0.1% EDTA, 0.01% Trypsin (1:250) in divalent cation free PBS, pH 7.4, and transferred to T-75 tissue culture flasks. In order to generate an enriched population, this selection procedure is repeated four more times, each time using from the previous selection. Cells at each selection are assayed for their chemotactic response (relative to the parent population) to PDL-CTX. Additionally, porous type 1 collagen barriers (100 μm pore size) are overlayed on gelatin-coated 1 μ pore Nuclepore filters and the selection procedure repeated. Here, the cells that migrate through the collagen barriers attach to the collagen-coated Nuclepore filters. The cells are collected and subcultured.

C. PERIODONTAL THERAPY:

After the patient's selected PDL cell population is obtained, the patient is asked to return to the office for conventional periodontal surgical therapy or for any combination of the following procedures:

(1) Application of both TGF-β (in concentration ranges between 10 picograms per ml PBS and 10 micrograms per ml PBS) followed by application of PDL-CTX in concentrations between 10 picograms and 10 micrograms per ml PBS. Application is accomplished by dripping the material onto the teeth by means of a pasteur pipette.

(2) The teeth are previously treated with an aqueous solution of either saturated citric acid or tetracycline HC1 (50 mg/ml to 100 mg/ml) for 5 minutes after which they are rinsed 4 times with PBS.

(3) Next the patient's own highly responsive PDL cells are applied to the tooth surfaces by allowing a suspension of these cells (5 x 10 ⁶ per ml PBS) to flow onto the tooth structure and remain undisturbed for a period of 30 minutes.

(4) The area treated as described above is then overlayed with an artificial basement membrane. This basement membrane is placed such that the type I

collagen side is next to the tooth-bone surface while the type IV collagen-laminin side is next to the soft tissue. The membrane is placed such that it extends 10 mm below the tooth-bone interface, 10 mm to either side (right and left) of the area treated and 10 mm above the crest of the soft tissue flap.

(5) Soft tissue flaps are then sutured such that the artificial basement membrane is folded over the crest of the soft tissue flap and secured to the soft tissue with methylmethacrylate.