WO2005118778A2

WO2005118778A2 - Method for ex vivo isolating and expanding human cornea stem cells and uses thereof

Info

Publication number: WO2005118778A2
Application number: PCT/IT2005/000313
Authority: WO
Inventors: Roberto Revoltella
Original assignee: Roberto Revoltella
Priority date: 2004-06-03
Filing date: 2005-06-01
Publication date: 2005-12-15
Also published as: WO2005118778A3; WO2005118778A8; ITRM20040273A1

Abstract

The invention concerns a method for the ex-vivo isolation and expansion of human corneal epithelial stem cells in tridimensional organotypic cultures, in bidimensional monolayer cultures, and in a co-culture with mammal fibroblasts and stromal cells used as feeders, and their relative uses in therapy and as an ophthalmic drug screening model.

Description

METHOD FOR EX-VIVO ISOLATING AND EXPANDING OF HUMAN CORNEA STEM CELLS AND USES THEREOF

DESCRIPTION

The present invention concerns a method for the ex-vivo isolation and expansion of human cornea stem cells in tridimensional (3D) organotypic cultures and their subsequent clone expansion in bidimensional (2D) monolayer cultures, and their relative uses in therapy and as a screening model for ophthalmic drugs. More specifically, the invention refers to a method for the ex-vivo isolation and expansion of epithelial stem cells of the human cornea from tridimensional organotypic cultures of fragments of the corneal limbus region, and their subsequent clone expansion in bidimensional monolayer cultures, and their use in therapy and for screening ophthalmic drugs. The cornea provides the eye with protection and the refractive properties vital for vision. It is composed of three layers consisting of: the epithelium and the basal membrane located in the front part; the central stroma that makes up about 90% of the whole tissue; and the endothelium in the innermost part, as shown in figure 1A. The epithelium is of the multistratified squamous type, devoid of a corneal layer, transparent and extremely specialised, composed of a base layer and several overlapping layers of keratinocytes closely adherent with one another, which are continually renewed starting from a population of stem cells (SC). The corneal stroma is devoid of blood vessels but contains, above all, water and collagen among which are dispersed the fibroblasts and keratinocytes. The stem cells and the more immature progenitor cells of the corneal epithelium are present in the limbus of the cornea. They have the capacity to self-reproduce and at the same time also to give rise to differentiated progenies in response to mediated micro-environmental stimuli, both from intercellular contacts and from soluble factors (cytokine, synergic soluble factors of their action, growth factors, factors maintaining homeostasis of the corneal epithelium, differentiation factors, various factors stimulating re- epithelisation present in lacrimal fluid). A serious depletion of the stem components of corneal epithelium cells means that the cornea is unable to renew itself and to maintain tissue homeostasis. There are both primary and acquired causes of insufficiency of limbus stem cells. More frequently, the deficiency of stem cells occurs after chemical burns or tissue degeneration from heat or of the post-traumatic kind, chronic limbitis, Sjogren's disease, or Steven-Johnson syndrome. This progressive deficiency of the epithelial cell system allows the contiguous conjunctival epithelium to penetrate the limbus and migrate above the corneal stroma, in a process known as conjunctivization. The result is characterized by the forming of a progressive opacity of the cornea and by a reduced visual capacity, since the conjunctival cells are phenotypically very different from the corneal ones and express very specific proteins and keratins and large quantities of mucopolysaccharides. Various therapeutic strategies have recently been adopted in order to treat the insufficiency of limbus stem cells, above all, in monolateral forms. Excluding an autologous transplant, the preferred intervention today consists of the allogenic transplant of limbus fragments. This transplant involves considerable risks connected to the expression by the transplanted keratinocytes of the histocompatibility antigens (particularly of class II) and of Langherans' dendritic and monocyte-macrophagic cells, thus requiring the receiving patient to undergo immunosuppressive treatment and, in case of rejection, to another transplant. There have been many attempts to isolate corneal epithelium stem cells in a monolayer culture and to expand their number while keeping them unchanged. These attempts involved the use of methods using chemically defined liquid mediums with or without serum enrichment and with some specific growth factors, vitamins, hormones and salts. All these bidimensional monolayer culture systems call for the presence in the system of stromal exogenous cells used as feeder cells, necessary to provide the epithelial stem/precursor cells with a suitable extracellular matrix for the anchoring and forming of intercellular connections and the production of autocrine-paracrine and exocrine soluble growth factors necessary to maintain the cells alive and unchanged (the so-called niche). These methods have so far been unsatisfactory because they do not guarantee a good predictability of success and repeatability of results, above all, as regards expanding the number of stem cells ex- o to a large enough quantity and the prevention of their differentiation, in order to use them in a transplant. Moreover, the stem/precursor cells of the corneal keratinocytes in an in vitro culture constitute a very useful system for screening the effectiveness of new drugs or for evaluating their toxicity. Some transformed human keratinocyte lines have recently been used, such as the HCE line, isolated by Prof. Roger W. Buerman (Louisiana State University, Eye Center, New Orleans, USA) or other transformed epidermic epithelial lines able to reconstruct in-vitro models of epitheliums not forming an external corneal layer in imitation of a corneal epithelial tissue (Doucet et al.,1998; Doucet et al., 1999; SkinEthic Tissue Culture Laboratories Workshop, 2003). In all these cases, these lines have been obtained by immortalizing the cells with exogenous genes whose expression is accompanied by cellular self- replication. The aforesaid cells are able to reconstruct a tridimensional epithelium if sowed on a substrate composed of elements of the extracellular matrix, such as a fibrin or collagen matrix or other protein of the normal extracellular matrix (Cuono et al., 1986; Fusenig et al., 1994; Miller et al., 1997; Fusenig et al., 1998; Pellegrini et al., 1999; Papini et al., 2003). In fact, they constitute a first base layer from which, by successive cell divisions, originate progenies forming overlapping layers of keratinocytes sticking to one another, thus forming a useful epithelium model for screening drugs or for evaluating the toxicity of various compounds under analysis. These models, however, have the drawback of the instability of the lines and by the fact that they are immortalized cells, if not downright already transformed cells, which are incapable of facing the normal maturation and differentiation processes, will easily face subsequent further genetic mutations, and are characterized by a marked phenotypic instability. Despite the fact that many studies have been carried out to identify the specific phenotypic markers of corneal stem cells and of the progenitors of the keratinocytes, these cells are difficult to identify and to isolate, and thus to expand in an undifferentiated manner ex vivo. Their phenotypic characterization and the creation of experimental and environmental conditions enabling their expansion without differentiating them is one of the main aims of today's research, given the importance of these cells for transplants and the need to use them maintaining them multipotential and stable in drug screening tests. In view of the above, there is the evident need to have culture methods and systems for the identification, isolation and expansion ex-vivo of normal corneal stem cells enabling the overcoming of the limitations of the hitherto used techniques. According to the present invention, an advantageous method has now been devised for the isolation and selective expansion of human cornea epithelial stem cells maintaining them alive and unchanged, by using tridimensional organotypic cultures that are particularly suitable for the cornea. In particular, as shown in figure 2, these tridimensional organotypic cultures are made up of small fragments of an explant of normal tissue which, maintained in a suitable culture microenvironment, favour SC survival and division keeping them unchanged, thus enabling the availability of a high number of cells that can divide remaining unchanged and form mature and differentiated progenies to be used respectively for therapy and as drug screening models. Moreover, the method according to the present invention allows maintaining intact, as in-vivo, in an initial stage both cell heterogeneity and the natural cytoarchitecture of the original corneal tissue by using corneal explant fragments. In this environment the SC increase in size first in the limbus and then, tending to flow out of the corneal fragment, migrate in the underlying spongy support from which it is possible to easily isolate the SC via enzymatic digestion. Subsequently, the isolated SC are expanded and maintained in bidimensional (2D) monolayer cultures in an optimal liquid medium conditioned by the presence of specific trophic factors for corneal SC provided by post-mitotic cells (i.e. incapable of dividing because they are pre-irradiated or treated with blocker drugs such as mytomicin-C, but still functioning) of corneal stromal fibroblasts used as feeders. The method allows the SC to be isolated in a high number and to remain unchanged: their stem nature is demonstrated by their capacity to reconstruct a polystratified epithelium with identical functional and phenotypic features as those of the in-vivo cornea, if cultivated on a fibrin or collagen matrix that incorporates the appropriate feeder cells. Following the ex-v/Vo expansion of the corneal epithelium SC preventing their changing, it is possible to use them in transplants and as drug screening models, limiting the use of experimental animals (i.e. the Draize Test). The 3D organotypic culture model for ophthalmic drug screening, such as anti-viral, anti-inflammatory or cicatrizing drugs, shows the advantage of being more physiological with respect to the 2D model because it reflects the complexity of the cytoarchitecture of the natural corneal epithelium. Moreover, as mentioned above, in some models some immortalized keratinocyte lines are used, often transformed, that imitate the epithelium but which do not reproduce the physiology of healthy epithelial cells of the cornea. Finally, a further advantage of the 3D organotypic culture model is that it enables the selection of one or more drugs, thus personalizing the therapy. Therefore, the present invention specifically provides a method for the ex-v/Vo isolation and expansion of human cornea epithelial stem cells, consisting of the following phases: a) selectively expanding human cornea epithelial stem cells in a 3D organotypic culture on a support matrix in a liquid culture medium; b) isolating and expanding human cornea epithelial stem cells obtained via the enzymatic digestion of the support matrix of the said 3D organotypic culture and of the culture itself; c) expanding the said stem cells up to reaching the sub-confluence of the epithelial stem cells isolated in -phase b) in 2D co-cultures in a liquid culture medium in slides containing as a substrate at least one post- mitotic line of mammal stromal fibroblasts; d) isolating the stem cells from the said substrate of phase c) by enzymatic digestion; e) selecting clones of corneal stem cells via immunohistochemical analysis of phenotypical markers and/or cytokine expression. In a preferred embodiment of the present invention, the 3D organotypic culture of phase a) can be composed of explant fragments of the corneal limbus, while the support matrix of phase a) can be chosen from the group consisting of gelatin, bovine collagen sponges, rat collagen sponges, and fibrin. As regards the liquid culture medium used in the method according to the present invention, it can conveniently be selected from among the EpilifelO or Epilife20 mediums. Preferably, the post-mitotic line of mammal stromal fibroblasts of phase c) of the method according to the present invention can be of the cornea, and more preferably of human or murin origin. In the former case, the said line of human origin is HPI.PC3 (PD 04002 filed at the CBA of Genoa on 1^st June 2004); while in the latter case, the line of murin origin is NIH-3T3 (American Tissue Type Culture Collection; ATCC CRL-1658). According to a preferred embodiment of the present invention, the phenotypic markers of phase e) of the aforesaid method are CK19 and p63. Moreover, according to an embodiment of the method according to the present invention, a further phase f) may be envisaged of the culture of the expanded stem cells, isolated and selected in phases c), d) and e) on a substrate of post-mitotic fibroblasts immersed in a medium including the following components: i) selected fetal bovine serum; ii) growth and/or differentiation factors selected from the group including beta FGF, EGF, IL-6, SCF, GM-CSF, G-CSF, IGF-1 ; for the reconstruction of the ex-wVo corneal epithelium. A further object of the present invention is a primary cell line of human post-mitotic fibroblasts called HPI.PC3 (PD 04002 filed at the CBA of Genoa on 1^st June 2004). Moreover, a further object of the present invention is the use of the aforesaid primary cell line of human post-mitotic fibroblasts as a substrate for the selective expansion of cornea epithelial stem cells. The sphere of the present invention also includes the use human cornea epithelial stem cells isolated and expanded according to the afore- described method for the ex-vivo reconstruction of the corneal epithelium. A further object of the present invention is the use of a 3D organotypic culture obtainable in phase a) of the method as a model for screening ophthalmic drugs. In a preferred embodiment of the 3D organotypic culture, it may be pre-treated with physical stimuli, preferably UV radiation, or with at least one chemical. Preferably, the said chemical is chosen from the group consisting of bacterial endotoxins, tumoral promoters, cytokines and recombinant growth factors. At the same time, another object of the present invention is the use of human corneal stem cells isolated in phase b) of the method as a model for evaluating the toxicity of ophthalmic drugs on the said stem cells. In particular, the ophthalmic drugs to be tested are chosen from the group consisting of anti-viral, anti-inflammatory and cicatrizing drugs, cytokines and synergic soluble factors of their action, growth factors, soluble factors stimulating the division of cornea epithelial stem/progenitor cells, factors stimulating re- epithelization, differentiation factors, soluble factors present in lacrimal fluid, eye-drops and artificial tears. A further object of the present invention is a screening method for ophthalmic drugs, such as anti-viral, anti-inflammatory and cicatrizing drugs, cytokines and synergic soluble factors of their action, growth factors, soluble factors stimulating the division of cornea epithelial stem/progenitor cells, factors stimulating re-epithelization, differentiation factors, soluble factors present in lacrimal fluid, eye-drops and artificial tears, consisting of the following phases: a. administering an ophthalmic drug to be evaluated to a 3D organotypic culture obtainable in phase a) according to the method of the present invention. b. evaluating the effect of the said drug by revealing the expression profile of the phenotypical markers, preferably CK19, CK3, p63, AE1/AE3, vimentin, CD31 , and Ki-67, and/or cytokine in the corneal stem cells with histological techniques, immunohistochemistry, molecular hybridation, cell count and by analyzing their vitality. c. comparing the profile obtained in phase b) with the expression profile in untreated cornea fragments, preferably explants of the corneal limbus. Finally, the present invention further provides a method for evaluating the toxicity of ophthalmic drugs, such as anti-viral, anti-inflammatory and cicatrizing drugs, growth factors, differentiation factors, eye-drops and artificial tears, consisting of the following phases: a. administering an ophthalmic drug to be evaluated both to the cornea fragments in the 3D organotypic cultures and to the cornea epithelial stem cells obtainable via enzymatic digestion in phase b) of the method according to the present invention and sowed on a support matrix in the presence of a line of mammal post-mitotic fibroblasts used as feeder cells. b. evaluating the effect of the said drug by revealing the expression profile of the phenotypic markers, preferably CK19, CK3, p63, AE1/AE3, vimentin, CD31 , Ki-67, and/or cytokine in the corneal stem cells with histological techniques, immunohistochemistry, molecular hybridation, cell count and by analyzing their vitality. c. comparing the profile obtained in phase b) with the expression profile in untreated cornea fragments, preferably explants of the corneal limbus. In a preferred embodiment, the said post-mitotic line of mammal corneal fibroblasts of phase a) of the aforesaid method is the primary cell line of human fibroblasts HPI.PC3 (PD 04002 filed at the CBA of Genoa on 1^st

June 2004) or the murin cell line NIH-3T3 (American Tissue Type Culture Collection; ATCC CRL-1658). The present invention will now be described for illustrative, but not for limiative purposes, according to some preferred embodiments thereof, with particular reference to the attached figures and drawings, wherein: Figure 1A shows a cross-section of a cornea with its constituent layers; figure B shows a cross-section of the cornea and of the limbus area with an indication of the putative location of origin of the stem cells and of their progeny. Figure 2 schematically shows the procedure used for the long-term 3D organotypic culture to make a marked selection of stem and more immature progenitor cells of the corneal epithelium and the procedure for recovering the stem cells that are the dominant surviving fraction in long-term cultures. Finally, the figure shows the subsequent phase of expansion of the stem cells kept in co-culture with a pre-irradiated monolayer of fibroblasts of the NIH.3T3 line or stromal fibroblasts of the human cornea. In the third phase of the method, the epithelial SC and progenitors, isolated and transferred to the co- cultures on a collagen gel including the feeder fibroblasts, reconstruct the corneal epithelium as is the case in vivo. The left side of figure 3 shows sections of a cornea immediately after removal from the patient, highlighting the epithelium and underlying stroma. The sections show the reactivity of the more undifferentiated fraction of the base cells that express two molecular markers proper of these cells, the antigen p63 and the antigen CK19, and the epibasal cells expressing the cytokeratin CK3, a specific marker of these cells, revealed with immunohistochemistry techniques. The figure shows sections of a newly removed cornea (on the left) and sections of a cornea maintained in a 3D organotypic culture for 14 days (on the right), highlighting the expression of the cytokeratin CK3, the marker of the epibasal layer of the epithelium, and the expression of the cytokeratin CK19, the marker of the undifferentiated basal cells of the corneal epithelium. Figure 4 shows the presence of positive p63 cells that have migrated from the cornea fragment to the sponge. Figure 5 shows the expansion of epithelial clones grown in a monolayer co-culture with pre-irradiated fibroblasts used as feeders. Figure 6 shows cornea sections reconstructed in a 3D culture from isolated stem cells after expansion on a collagen gel containing post-mitotic fibroblasts used as feeders. The immunohistochemical tests reveal an epithelium whose base cells are positive p63. The basal and some of the pibasal cells are positive CK19, while only the cells of the epibasal layers turn out to be positive CK3. Figure 7 shows an example of the constitutive expression of mRNA for the genes of some indicated cytokines (IL-1 alpha, TNF alpha, GM-CSF) by corneal cells expanded in organotypic cultures for 21 days and then isolated by the support sponges. Figure 8 shows an example of the use of cornea epithelia reconstructed in vitro and assayed with 2 ophthalmic drugs: TG (Dropstar TG, 0.4% hyaluronic acid, Farmigea S.p.A. A032040014) and TSP (TS- Polysaccharide, Farmigea S.p.A., A9.07252783). EXAMPLE 1 MATERIALS AND METHODS Human corneal tissue The explants of human tissue were obtained surgically. The informed consent of the patients' families or of the donors was required beforehand. Alternatively, recourse was made to Corneal Tissue Banks for corneal explants that were otherwise unusable due to their damage or insufficient surface. The tissue explants were kept in EUSOL-C (Graft TEC, Alchimia

Transplant SRL, Padova, cat. N° REF-10040) for 2-4 days at 4°C to avoid contamination. 3D organotypic cultures Small corneal explants near the limbus were cut with scissors into fragments of 1-2 mm³ in sterile conditions, removing the conjunctival tissues. The fragments were washed once in a sterile physiological solution and then in Epilife (Cascade Biologies, Portland, OR, USA, cat. N° M-EPIcf- 500), with the addition of 50 μg/ml of penicillin, 100 μg/ml streptomycin sulphate and glutamine in a concentration of 0.29 mg/ml (Euroclone West York, UK), and EpiLife-Defined-Growth-Supplement (EDGF) (Cascade Biologies, cat. N° M-EPIcf-500) a combination of soluble factors promoting epithelial growth, purified bovine serum albumin, purified bovine transferrin, hydrocortisone, recombinant human insulin-like growth factor (rhlGF-1 ), prostaglandin E2 (PGE2), recombinant human epidermis growth factor (rhEGF). Various lots of fetal bovine serum (FBS) were checked for optimal cell growth and only one lot was used in all these experiments. Gelfoam® sterile pigskin gelatin sponges (Pharmacia & Upjohn,

Kalamazoo, Ml, USA; cat. no. NDC 009-031503) (dimensions 20x60x7 mm), were hydrated with Epilife containing 10% FBS (Euroclone) (Epilife-io) and pressed in order to eliminate air pockets before use, as described (Hanto et al., 1982; Chishima et al.,1997 ; Papini et al., 2004). Each sponge was cut into three pieces (dimensions of 20x20x7 mm), and each piece was transferred to a well of a 6-well slide for tissue cultures (Falcon Plastics Inc.; London Ontario, Canada; cat. N° 353046). The fragments were transferred (3-6 fragments per well) and placed on the upper part of the sponge. Epilife-ι₀ was then added until it covered the upper part of the sponge (3,5-4 ml/well). The slides were kept in a humidified incubator at 37°C and 5% C0₂ in atmospheric air. The culture medium was renewed every two days. The fragments and sponges were collected at different times: at the start of the culture (Od, TO), after 4 or 7 days (4d or 7d, T1 ), after 14 days (14d, T2) and after 21 days (21 d, T3). The samples were fixed in Gliofix (Italscientifica Spa, Genoa, Italy, cat. N° 9186764262) for 4 hours at room temperature and kept in a 75% ethanol solution at 4°C until the histological and immunohistochemical analysis.

The use of murin fibroblasts (NIH-3T3 line) and primary cultures of human fibroblasts from corneal stroma The murin fibroblast line NIH-3T3 was obtained from the American

Tissue Type Culture Collection (ATCC no. CRL-1658). Three primary cultures of human fibroblasts (respectively called HPI.PC3, HPI.PC19 and HPI.PC21 , of which the first was filed with the access number PD 04002 on 1^st June 2004 at the CBA of Genoa, Italy) were obtained from surgical explants of three different human cornea obtained with the prior consent of the respective patients. The explant was cut into small fragments (1x1 mm) and then treated with collagenase-1A (1 mg/ml) and hyaluronidase-IVS (0.4 mg/ml) (Sigma Chemicals Co., St. Louis, MO) at 37°C for 45 minutes, and finally washed in a saline-phosphate physiological solution (PBS) of 7.4 pH. The enzymes were then neutralized with medium containing FBS and the dispersed cells were then washed twice in PBS, centrifuged (1200 rpm at 4°C for 3 min) and re-suspended in the Epilife-_to culture medium containing gentamycin (Sigma), penicillin and streptomycin (20 μg/ml each) and transferred to an incubator at 37°C in a 5% C0₂ atmosphere in air for 14 hours. The medium was then removed and Epilifβio was added to the adhering cells, and the medium was changed every 3 days. Each of these explants gave rise to a primary culture of fibroblasts capable of self-replication and the cultures were thus divided (1 :3) at the semi-confluence, with trypsin-EDTA (0.05% trypsin, 0.02% EDTA, weight/volume) for 10 minutes at 37°C, then blocking the enzyme with 10% FBS. The cells were expanded in Epilife₁₀ in T25 plastic bottles (Corning Inc., Corning, NY) or in 6-well slides (Falcon Plastics Inc, Oxnard, CA), dividing them at semi-confluence (1 :3) at an initial density of about 5x10⁵ cells/ml with medium changes every 3 days. The cells were expanded in 150 ml slides in the first 5 operations and frozen at -180°C in aliquots. Individual phials were then defrosted, the cells washed twice in Iscove medium and then expanded in one operation and used. Only primary cultures within the 5^th operation were used as feeders for expanding the corneal stem cells. The preliminary results showed that up to the first 8-12 operations the cells HPI.PC3, HPI.PC19 and HPI.PC21 , with no significant differences between them, constitutively produce a wide range of growth factors, including, in particular, FGFbeta, EGF, IL-6, SCF, GM-CSF, G-CSF, IGF-1 and NGF. Isolation and clone expansion of corneal stem/precursor cells in a 2D monolayer culture The cornea samples were first cultivated in the upper part sponges for 21 days in Epilifeio, as described above and shown in figure 2. The fragments and sponges were then removed from the slide, washed in a physiological solution and digested separately with a trypsin-EDTA 1X solution (Sigma- Aldrich, NY, USA cat. No T3924) at 37°C for 80 minutes. The cells released in suspension were collected every 20 minutes and the enzymatic reaction was neutralized by adding a medium supplemented with 10% FBS. The cells were then placed on 6-well culture slides with or without a monolayer of post-mitotic stromal cells treated with mytomycin-C (myt-C), sowed in the culture 24 hours before, and used as a feeder cell layer (figure 2) and kept in an incubator at 37°C in a 5% C0₂ atmosphere in air. The conditioned cell culture medium was changed every two days. On reaching the sub-confluence (about 70% of well saturation), the epithelial cells of each well were removed, transferred to a chamber of a glass slide for the cultures (Falcon Plastics Inc.; London Ontario, Canada; cat. No 354104), and processed for histological and immunohistochemical analyses. The clones were selected according to their growth capacity and to the phenotypical traits revealed by the immunocytochemical colouring, shared with their cells. Histology and immunohistochemistry At pre-established time intervals (before culture, TO; after 4 or 7 days of culture, T1 ; after 14 days (T2), after 21 days (T3) of culture, the tissue fragments and sponges were fixed in Gliofix or in 4% formalin and included in paraffin, and some sections of 5 μm thickness from each group were marked with hematoxilin (Bio-optica) and eosin (Eosin Y Sigma-Aldrich)(H&E). Immunohistochemical analysis was carried out on consecutive sections using the Dako Envision+™ System HRP (DakoCytomation, Carpinteria, CA, USA code N° K4007). The positive immunoreaction of the primary antibody was detected using a secondary antibody along with peroxidase, using diaminobenzidine (DAB) as a chromogen. Immunocolouring was carried out to identify the epithelial cell types and to evaluate some specific features, such as the proliferation rate. Table 1 shows the monoclonal antibodies used in these analyses to characterize the various stages of maturation/differentiation of the cornea keratinocytes and the cells present in the stroma. TABLE 1

The cells grown in monolayer in the culture slides were first fixed for 6 minutes in absolute ethanol and then manually marked with murin antibodies

(mAb) for the nuclear antigen Ki-67, anti-CK3, anti-CK19, anti-vimentin, anti- pan-cytokeratin and with mAb anti-p63, by using the Envision+™ kit for antibody detection, according to the manufacturer's instructions. The cells marked positively for the p63 antigens were counted in five non-consecutive fields of the same or other preparations with a 25X enlargement using a Leitz-Dialux 20 EB microscope, and the percentage of positive cells was calculated on the total number of epithelial cells counted for each visual field.

Pharmacological tests The cultures, both the 3D organotypic ones with the cornea fragments and the 2D ones composed of stem cells co-cultivated with fibroblasts, were stimulated with physical agents (UV radiation for 15 minutes) and by adding various exogenous chemical compounds to the culture medium (for example, by adding bacterial lipopolysaccharides, LPS, 5 ng/ml), phorbol diesters as tumoral promoters (12-0-tetradecanoyl-phorbol-13-acetate, TPA, 10 ng/ml), various cytokines and recombinant growth factors IL-1 alpha (5 ng/ml), IL-6 (810 ng/ml), TNF alpha or TNF beta (10 ng/ml each), GM-CSF (810 ng/ml), TGF (810 ng/ml), and IFN gamma (8 ng/ml), cytokine recombinant hexogen, as reported in table 2, besides two ophthalmic products, respectively TG (Dropstar TG 0.4% hyaluronic acid, Farmigea S.pA., A032040014) and TSP (TS-Polysaccharide, Farmigea S.p.A., A907252783). Table 2 shows the complete list of cytokines which turned out to be produced by the epithelial cells in the 3D cultures of cornea reconstructed in vitro, starting from stem cells isolated from fragments of tissue cultivated for 14 days. The expression of these cytokines was demonstrated both with immunohistochemical techniques (monoclonal antibodies) applied on half the fragments of each sample under analysis (about 4-6 fragments per case), and with molecular hybridation techniques (in-situ hybridation and northern blot) by extracting the mRNA from the second half of the same tissue fragments analysed. The 3D organotypic cultures used were both whole cornea fragments after 21 days of culture, and new epithelia reconstructed in vitro from stem cells isolated from 21 -day fragments, after a further 14 days in co-culture on a collagen matrix with pre-irradiated fibroblasts of the stromal HPI.PC3 line used as feeders. TABLE 2 Factors inducing the production of Cytokines produced by the keratinocytes cytokines and growth factors stimulated in 3D organotypic cultures UV rays IL-1alfa, IL-1 beta, IL-6, TNFalpha, IL-8 Bacterial endotoxins (LPS) IFNgamma, MGF, G-CSF

Recombinant and natural cytokines GM-CSF, TGFalpha, TGFbeta, FGF, PDGF

These products were tested at various Iscove medium concentrations enriched with only 5% FCS. Preliminary experiments had shown a non-toxic effect at a dilution of 1 :20 of the two commercial products, and this dilution was thus used in the test. The trials included cells incubated with the same culture medium (lscove ), but without the drug. The culture medium was then changed every two days by adding fresh medium containing the same samples. In the experiments in which the cultures were treated with cytokines or exogenous ophthalmic products, the experiment was interrupted at the indicated times. The fragments and sponges were fixed in Gliofix for 4 hours at room temperature and kept in a 75% ethanol solution at 4°C until the histological and immunohistochemical analyses. RESULTS 3D organotypic cultures on sponges The histological analysis of the cornea fragments cultivated in a native state on gelatin supports, without the addition of exogenous factors, reveals a good maintenance of cell heterogeneity and 3D tissue architecture throughout the experiment. As shown in figure 3, before the culture at TO, the corneal epithelium shows a bright colouring for CK3 at the superbasal layers of the central cornea, while CK19 was detected in the limbus, in particular in the basal cell layer. The same expression profile was found with the anti-p63 antibodies, whose expression is exclusively limited to the disseminated epithelial cells of the basal layer. In contrast with TO, after 4-7 days (T1 ) and 14 days (T2) there are some evident changes. In particular, after 4-7 days the upper epibasal layers almost disappear and the remaining epithelium appears reduced to a monolayer, sometimes double, of epithelial cells, which selectively express both the CK19 and p63markers. Subsequently, these cells start to infiltrate the underlying sponge. At times T1 and T2, the expression of CK3 is reduced, with a positivity restricted to only the epithelia of three or more layers. At T2 (14 days) the surface of each fragment is covered by a squamous polystratified epithelium (3-5 layers of closely adherent nucleate keratinocytes) containing a marked dominance of CK19 positive basal cells and an increasing proportion of positive p63 cells, still in the basal cell layers. Starting from T1 , many cells, CK19 and p63 positive were found inside the sponge, as shown in figure 4. Moreover, no presence was found or a slight immunoreaction was seen for CK3. At T3 (21 days of culture), a significant increase of positive p63 and CK19 cells was found in the basal layer of the epithelia and among the cells infiltrating the sponge. Figure 4 shows the presence of positive p63 cells that had migrated from the cornea fragment to the sponge, revealing that this cell population constitutes - along with the positive CK19 cells - almost all the cells present (over 85-90%). Instead, the positive CK3 cells result limited to the epibasal layers of the epithelia lining the fragment at the interface between the air and the liquid medium of the culture, but are not present among the cell population infiltrating the sponge. Before the culture process and at each established time interval, the stroma appears well preserved, with small vessels (positive CD31 and vimentin) in the limbal area that are apparently intact and unchanged if compared to their state before the culture process. The monostratified endothelium under the Descemet's membrane (endothelial basal membrane) is well kept in the culture, as confirmed by the positive expression of vimentin and CD31. Expansion in a monolayer culture of basal cells isolated from long-term organotypic cultures The enzymatic digestion of small corneal explants carried out before the start of the culture process (TO) or after 21 days of 3D organotypic culture (T3) in Epilife-io (figure 5), was carried out successfully in order to isolate, purify and expand the stem cells ex- Vo by starting co-cultures in 2D monolayers formed by adding the cells (over 85% p63 positive and CK19 positive) to NIH.3T3 post-mitotic fibroblasts or to primary cultures of human post-mitotic fibroblasts [HPI.CP3 (PD 04002 filed at the CBA of Genoa on 1^st June 2004), HPI.PC19 or HPI.PC21] used as feeder cells. The culture medium used was the EpiLife medium added with EDGF and FBS10% (EpiLifeio). This system favours the expansion of a high number of immature epithelial cells, which adhere very rapidly to the slide and actively divide to form mosaic clones that enlarge with a typical polygonal epithelium-like morphology (figure 5). After a second operation, these clones can be subcultivated by successive divisions at the subconfluence starting from an initial implant of about 1 ,000 cells per cm². The presence of NIH.3T3 fibroblasts of human ones deriving from the corneal stroma markedly increases the proliferation potential of these epithelial precursor cells which, in the absence of a feeder layer for their replication, should instead be sowed initially in a much higher number (> 10⁴ cells per cm²). Moreover, among the few epithelial cells capable of sticking to plastic, those that do stick often give rise to abortive colonies only capable of a few divisions. The epithelial cells grown in EpiLife₁₀ almost all constitutively express CK19 and most of them are also positive for p63 (about 60%), as shown in figure 5. They do not, however, express the antigen CK3. The expanded clones containing cells with a high cell-division capacity are representative of the limbus stem/progenitor cells present in-vivo. During the initial culture phases, these clones account for a small fraction, about 0.5-1 % of the initial primary mass cultivated, but the proportion of cells that can replicate with a positive CK19 and p63 phenotype significantly increases in the subsequent cell operation, affecting almost the entire population of the mass developing in the culture. These cells maintain a positive immunoreactivity for CK19 and p63. In order to accelerate the isolation and purification of this subgroup of corneal keratinocyte precursors, rapidly expanding individual clones of positive CK19 and p63 cells were isolated after enzymatic digestion with trypsin-EDTA and delicate scraping, and the cells, dispersed using a pipette, were mixed together and the resulting cultures were divided at the sub- confluence. A selection was thus made of the clones exhibiting the aforesaid phenotype and these cells rapidly formed a homogeneous population of corneal precursor cells that can be transmitted for at least 10 successive operations. These cells can easily be expanded ex-vivo even when sowed on an amniotic membrane used as a niche, as described above for these precursor cells. Ascertainment of the stem nature of the positive CK19 and positive p63 cells by measuring their capacity to reconstruct the corneal epithelium in 3D cultures The CK19 and p63 positive cells isolated after 2 or 3 operations in co- culture with human post-mitotic fibroblasts used as feeders, were recovered via enzymatic digestion and delicate scraping, and were dispersed by repeated actions using a pipette. The cells were placed by pipette on a gel layer (mainly type I collagen, but also fibrin) deposited on a plastic support that was placed in a 6-well slide containing a monolayer of HPI.PC3 cells (PD 04002 filed at the CBA on 1^st June 2004) or cells of the HPI.PC19 or HPI.PC21 lines, sowed 24 hours before as a layer of feeder cells and treated with myt-c to render them incapable of dividing. The medium was added in a sufficient quantity to keep the cornea epithelial stem/precursor cells completely immersed. The conditioned medium was replaced every two days with fresh medium in a dilution ratio of 1 :1. When the cultures were subjected to histological analysis and immunohistochemical colouring, after 10 and 21 days, via fixation with formalin and inclusion in paraffin, the formation and development was seen of a corneal polystratified epithelium with keratinocytes expressing the basal phenotype (p63 positive and CK19 positive) and epibasal one (CK3) normally found in native corneal epithelia. Figure 6 shows the immunohistochemical analyses carried out on cornea sections that reveal an epithelium whose basal cells are p63 positive; the basal cells and in part some basal cells of the second parabasal layer are CK19 positive, while only the cells of the epibasal layers turn out to be CK3 positive. This evidence thus confirms the stem nature of the cells recovered through expansion with the feeder cells. Effects of exposing the organotypic cultures to physical agents and exogenous chemical compounds on the production of cytokines and growth factors The aforesaid table 2 shows how stimulation of the initial corneal organotypic cultures and of the 3D ones reconstructed from stem cells sowed on a collagen matrix incorporating human fibroblasts [HPI.PC3 (PD 04002)] as feeders, stimulated both with physical agents (UV radiation for 15 minutes) and by adding various exogenous chemical compounds to the culture medium (such as by adding bacterial lipopόlysaccharides, LPS, 5 ng/ml), phorbol esters (TPA, 10 ng/ml), various cytokines and recombinant growth factors (e.g. IL-1 alpha (5 ng/ml), IL-6 810 ng/ml), TNF alpha or TNF beta (10 ng /ml each), GM-CSF 810 ng/ml), TGF 810ng /ml), and IFN gamma (8 ng/ml), causes significant changes in corneal epithelium development and at the same time stimulates a significant increase in the levels of cytokines, growth factors or their receptors, such aslL-1 alpha and IL-1 beta, IL-6, TNF alpha, IL- 8, IFN gamma, MGF, GM-CSF, G-CSF, TGF alpha and TGF beta, FGF, PDGF, NGF and its receptor (75kDa), whose levels were measured both with the Northern Blot technique (figure 7) and by using immunohistochemical techniques. Variations in the production of these products by the keratinocytes not only in the initial organotypic cultures, but especially in the corneal epithelium reconstructed in 3D from the stem cells on a collagen matrix incorporating HPI.CP3 cells as feeders, demonstrate the validity of these cultures for comparatively assessing - on samples deriving from the same cornea explant and tested in the same experiment and in the same basal culture conditions - the biological effects induced on cornea epithelial cells by physical agents and exogenous chemical compounds. Effect of drug treatment As shown in the histological and immunohistochemical analyses, the cornea fragments cultivated as described above, in Iscoves in the presence of TG and TSP have well maintained the initial cytoarchitecture, revealing after 14 days of culture a phenotypic aspect comparable to that of cornea fragments cultivated in Iscoveio, very different from the one shown by cultures maintained without the addition of the two exogenous drugs to the culture medium. In the cultures maintained in lscove₅, only one monolayer of foundation basal cells, and rarely a two-layer epithelium, were found in the culture at time T1 (after 7 days). Figure 8 shows an example of the use of corneal epithelia reconstructed in vitro and assayed with TG and TS (dilution of 1 :20 of the commercial stock in Iscoves medium) to evaluate their effects by comparing the results with cultures maintained in the same medium, but without the drug (P). The reconstruction of the corneal epithelium occurred by the corneal stem cells (over 89% p63 positive and CK19 positive) after 14 days of culture process (T2) on collagen matrices including primary cultures of post-mitotic fibroblasts of the HPI.PC3 line, revealing an excellent tolerability to drugs. At times T2, these epithelial cells express cytokeratines of high molecular weight, revealed by the antibody AE1/AE3, but the expression of CK19 and p63 is markedly reduced, and at times absent. As figure 8 shows, compared to controls P, in the cultures in lscove₅ in the presence of TG and TSP, the epithelium at T2 and T3 appears well developed, with many cells in the basal layer expressing the antigen p63 and cytokeratines (revealed with the antibody AE/1 and AE3). The P controls instead revealed signs of stress in the new epithelium with a poor proliferation of precursors (p63 positives) and an evident advanced differentiation of positive AE1/AE3 cells. The presence of cells migrating in the sponge is high, and a high percentage of them express the two basal markers CK19 and p63, whose expression is marked. The method according to the present invention facilitates the study of proliferation and differentiation of human corneal stem/precursor cells within their natural 3D microenvironment in a long-term organotypic culture. Indeed, an important contribution is made consisting of the fact that the epithelial stem/progenitor cells normally located in the limbus can be advantageously expanded ex vivo and isolated as a pure cell population with the phenotype CK19 and p63 positive, which can be used for reconstructing the epithelium of the corneal surface in case of damage. Through the aforesaid method, this group of isolated corneal epithelium stem/precursor cells can be expanded ex vivo on appropriate substrata in the presence of suitable growth factors, and remain unchanged but potentially capable of forming progenies which can instead differentiate. This method thus offers the opportunity to isolate and purify - with a high reproducibility - the necessary basic elementary components for tissue engineering of the corneal epithelium. Moreover, the 3D organotypic cultures that reflect the natural cytoarchitecture and maintain cell heterogeneity can be used as drug screening models.

REFERENCES ■ Doucet et al. In vitro and Molecular Toxicology. 1998; 11 :273. » Doucet et al. In vitro and Molecular Toxicology. 1999; 12: 63. « SkinEthic Tissue Culture Laboratories, Workshop October 16-17, 2003, Nice, France. - Cuono et al. Lancet. 1986; 1 :1123-1124. ■ Fusenig NE in: The Keratinocytes Handbook. 1997; pp.331-362. ■ Fusenig NE, Boukamp P. Mol. Carcinogen. 1998; 23: 144-158. ■ Pellegrini G, Golisano O, Paterna P et al. J.Cell.Biol. 1999; 145: 769- 782. ■ Papini S, Cecchetti D, CampaniD, Fitzgerald W, Grivel JC, Chen S, Margolis L, Revoltella RP. Stem Cells. 2003; 21 :481-494. ^■ Hanto DW, Hopt UT, Hoffman R, Simmons RL. J Immunol. 1982; 129:2437-2443. ■ Chishima T, Yang M, Miy Y, Li L, Tan Y, Baranov E, Shimada H, Moossa AR, Nenman S, Hoffman RM, Proc.Natl.Acad.Sci. 1997; US 94:11573-76. ■ Papini S.Rosellini A, Campani D, DeMatteis A, Selli C, Revoltella RP, The Prostate. 2004; 59:383-392.

Claims

CLAIMS 1. A method for the isolation and expansion ex vivo of human corneal epithelial cells, including the following phases: a) selectively expanding human , cornea epithelial stem cells in a 3D organotypic culture on a support matrix in a liquid culture medium; b) isolating and expanding human cornea epithelial stem cells obtained via the enzymatic digestion of the support matrix of the said 3D organotypic culture and of the culture itself; c) expanding the said stem cells up to reaching the sub-confluence of the epithelial stem cells isolated in phase b) in 2D co-cultures in a liquid culture medium in slides containing as a substrate at least one post- mitotic line of mammal stromal fibroblasts; d) isolating the stem cells from the said substrate of phase c) by enzymatic digestion; e) selecting clones of corneal stem cells via immunohistochemical analysis of phenotypic markers and/or cytokine expression. 2. A method according to claim 1 , wherein the said tridimensional organotypic culture of phase a) consists of fragments of the corneal limbus. 3. A method according to claims 1 or 2, wherein the support matrix of phase a) is selected from the group consisting of gelatin, bovine collagen sponges and rat collagen sponges, and fibrin. 4. A method according to any one of the preceding claims, wherein the liquid culture medium is Epilife-io or Epilife₂o. 5. A method according to any one of the preceding claims, wherein the said at least one post-mitotic line of mammal stromal fibroblasts of phase c) is of the cornea. 6. A method according to claim 5, wherein the said line is human. 7. A method according to claim 6, wherein the said line is HPI.PC3 (PD 04002 filed on 1 June 2004 at the CBA of Genoa, Italy). 8. A method according to claim 5, wherein the said line is murin. 9. A method according to claim 8, wherein the said line is NIH-3T3. 10. A method according to any one of the preceding claims, wherein the said phenotypic markers of phase e) are CK19 and p63. 11. A method according to any one of the preceding claims, including a further culture phase f) of the expanded stem cells, isolated and selected in phases c), d) and e) on a substrate of post-mitotic fibroblasts immersed in a medium including the following components: i) selected fetal bovine serum; ii) growth and/or differentiation factors selected from the group including beta

FGF, EGF, IL-6, SCF, GM-CSF, G-CSF, IGF-1 ; for the reconstruction of the ex-wVo corneal epithelium. 12. A primary cell line of human post-mitotic fibroblasts HPI.PC3 (PD

04002 filed on 1 June 2004 at the CBA of Genoa, Italy). 13. Use of the primary cell line of human post-mitotic fibroblasts, as defined in claim 12, as a substrate for the selective expansion of corneal epithelial stem cells. 14. Use of the human corneal epithelial stem cells isolated and expanded according to the method as defined in claims 1 to 11 , for the ex-wVo reconstruction of the corneal epithelium. 15. Use of the tridimensional organotypic culture obtainable in phase a) according to the method as defined in claims 1 to 11 , as a model for the screening of ophthalmic drugs. 16. Use according to claim 15, wherein the said tridimensional organotypic culture is pretreated with UV radiation or at least one chemical. 17. Use according to claim 16, wherein the said chemical is selected from the group consisting of bacterial endotoxins, tumoral promoters, cytochins and recombinant growth factors. 18. Use according to claim 17, wherein the said tumoral promoters are the diesters of phorbol. 19. Use of human corneal stem cells isolated in phase b) according to the method as defined in claims 1 to 11 as a model for evaluating the toxicity of ophthalmic drugs on the said stem cells. 20. Use according to claims 14 to 19, wherein the said drugs are selected from the group consisting of anti-viral, anti-inflammatory and cicatrizing drugs, cytokines and synergic soluble factors of their action, soluble factors stimulating the proliferation of cornea epithelial cells and their homeostasis present in lacrimal fluid, growth factors, differentiation factors, eye-drops and artificial tears. 21. An ophthalmic drug screening method including the following phases: a. administering an ophthalmic drug to be evaluated to the tridimensional organotypic culture obtainable in phase a), according to the method as defined in claims 1 to 11 ; b. evaluating the effect of the said drug by revealing the expression profile of phenotypic markers and/or cytokines in the corneal stem cells with histological techniques, immunohistochemistry, molecular hybridation, cell count and by analyzing their vitality; c. comparing the profile obtained in phase b) with the expression profile in untreated cornea fragments. 22. A method for evaluating the toxicity of ophthalmic drugs including the following phases: a. administering the said drug to cornea epithelial stem cells obtainable via enzymatic digestion in phase b) of the method as defined in claims 1 to 10, and sowed on a support matrix in the presence of a line of murin post-mitotic fibroblasts (N I H.3T3) or of mammal fibroblasts and corneal stroma cells. b. evaluating the effect of the said drug by detection of the expression profile of phenotype markers and/or cytokines in corneal stem cells with histological techniques, immunohistochemistry, molecular hybridation, cell count and by analyzing their vitality. c. Comparing the profile obtained in phase b) with the expression profile in untreated cornea fragments. 23. A method according to claim 22, wherein the said line of mammal corneal post-mitotic fibroblasts of phase a) is HPI.PC3 (PD 04002 filed on 1

June 2004 at the CBA of Genoa, Italy) or NIH-3T3. 24. A method according to claims 21 or 22, wherein the phenotype markers of phase b) are one or more of the following markers: CK19, CK3, p63, AE1/AE3, vimentin, CD31 and Ki-67. 25. A method according to claims 21 or 22, wherein the said drugs are selected from the group consisting of anti-viral, anti-inflammatory and cicatrizing drugs, growth factors, soluble factors stimulating the division of cornea epithelial stem/progenitor cells, factors stimulating re-epithelization, factors maintaining the homeostasis of the corneal epithelium, cytokines, soluble factors present in lacrimal fluid, differentiation factors, eye-drops and artificial tears. 26. A method according to claims 21 or 22, wherein the said untreated cornea fragments of phase c) are corneal limbus explants.