AU1283000A - Skin composites - Google Patents

Description

WO 00/32252 PCT/GB99/03889 1 SKIN COMPOSITES The invention relates to a method to prepare skin tissue for use in, particularly 5 but not exclusively, the repair of wounds and/or damaged tissue and/or cosmetic reconstruction. Skin is a highly complex organ covering the external surface of the body. Skin functions, amongst other things, to prevent water loss from the body and 10 to act as a protective barrier against the action of physical, chemical or infectious agents. Skin has an elastic property and varies in thickness from 0.5 mm, on for example the eyelids, to 4 mm on for example the palms and soles. Skin is composed of two layers. The outer layer, which is comparatively thin 15 is called the epidermis. It is several cells thick and has an external layer of dead cells that are constantly shed from the surface and replaced from below by a basal layer of cells, called the stratum germinativum. The epidermis is composed predominantly of keratinocytes which make up over 95% of the cell population, the rest include dendritic cells such as Langerhans and pigmented 20 cells called melanocytes. It is essentially cellular and non vascular, there being relatively little extra cellular matrix except for the layer of collagen and other proteins beneath the basal layer of keratinocytes. Keratinocytes of the basal layer are constantly dividing, and daughter cells subsequently move outwards, during which they undergo a period of differentiation and are 25 eventually sloughed off from the surface. The inner layer of the skin is called the dermis and is composed of a network of collagenous extracellular material, elastic fibres, blood vessels and nerves. Contained within it are hair follicles and associated sebaceous glands, WO 00/32252 PCT/GB99/03889 2 collectively known as the pilosebaceous unit. The interface between the epidermis and dermis is extremely irregular and consists of a succession of papillae, or finger like projections. 5 Skin is therefore a highly complex organ the cells of which are periodically replaced. Although skin provides a hard wearing and resilient outer surface skin can be physically stressed with relative ease thus leading to irreparable damage. Damage of this sort can result from: high impact contusions, diseases or burns. 10 Diseases that result in irreparable damage include gastric ulcers; ischemic and/or arterial ulcers; and/or diabetic ulcers. Of these, so called ambulatory ulcers are the most common and result from 15 poor venous blood supply to the lower limbs. Venous insufficiency results in severe necrosis leading to deep open wounds and scarring. The scarring results in a condition referred to as lipodermatosclerosis which is a hardened and leathery skin overlying sites of fibrin cuffs. This is particularly evident in diabetes sufferers where foot ulcers can result in amputation of toes and/or 20 feet. There are a number of treatments of diabetes that alleviate many of these symptoms. However the ulcers can be deep and do not heal well. In these situations skin grafts may be appropriate to remedy the condition. Gastric ulcers are a common ailment of the major industrialised nations. There 25 is a strong correlation between the presence of Helicobacter pylori and the incidence of gastric ulcers and this correlation has changed how patients prone to gastric ulcers are treated. However in those individuals suffering from established gastric ulcers the only remedial action is to surgically repair the WO 00/32252 PCT/GB99/03889 3 stomach wall. This may require replacing the tissue with an autograft from a region of the stomach that is unaffected. In addition to the above, skin is particularly susceptible to heat damage which 5 can be of varying severity. For example, first degree burns affect the outer layer of the skin, causing pain redness and swelling. Second degree burns affect both the outer and underlying layer of skin causing pain, redness, swelling and blistering. Third degree bums extend into deeper tissue causing brown or blackened skin that may result in irreparable damage; this requires 10 surgery to remove dead skin and replacement, via grafting of skin, to repair the wound. Currently three types of graft are available to a surgeon: i) an autograft, in which a piece of tissue is removed from one area of a patient's body and placed in another location; 15 ii) an allograft, in which a section of tissue from one human, for example a cadaver, is grafted onto another human; and iii) a xenograft, where tissue is harvested from another animal species, for 20 example a pig, and placed over the wound area. Autografts can be problematic due to the availability of suitable tissue and the added trauma to the patient during the removal of the tissue from another part of the body to the wound area. Allografts are limited by immunological 25 reactivity of the host and the availability of donor tissue. Xenografts are even more problematic due to the severe immunological reactivity. Currently a number of approaches are being used to address the problems identified above. One approach is to provide human skin composites or WO 00/32252 PCT/GB99/03889 4 reconstructed skin that have reduced immune reactivity leading to a decrease in tissue rejection by the patient. Alternatively, research has been directed to the development of artificial skin composites which provide many of the features of natural skin. In yet a further alternative strategy the combination of 5 cultured epithelium in combination with either an allograft dermis or synthetic dermis have been used to decrease rejection. Culturing epithelial cells by the technique described by Green et al (PNAS USA 1979; 76: 5565 - 5568 (35)) made it possible to prepare large numbers of 10 epithelial sheets reliably within about four weeks post burn. These epithelial sheets are used in isolation of other replacement tissue. Typically epidermal grafts tend to be fragile leading to late graft loss, so that long term "take" rates are often much lower than initial take rates. The results of early and late graft take were reported to be improved by the technique of excising the burn tissue, 15 covering it with an allograft, allowing the allograft to take, followed by shaving off the epidermal layer and replacing it with autogenous cultured epithelial autograft sheets (Cuono et al PNAS, 1987; 80; 826 - 835, (37)). However problems still remain with this technique and include unreliable allograft take, early rejection, infection, necrosis of the allograft, and difficulty 20 in shaving the allograft to the required depth. One method of avoiding these problems is to prepare epidermal-dermal composites in vitro, before application to the burn wound. This approach has been explored, and it has been demonstrated that epidermal-dermal composites 25 can be successfully prepared in vitro. However, there are several issues that need to be addressed before these techniques become clinically applicable on a routine basis. It is important that the allogenic dermal tissue used should be sterile before clinical use to eliminate the risk of transmission of disease. The donor tissue should have a minimum of potentially antigenic allogeneic WO 00/32252 PCT/GB99/03889 5 cellular material. Both the sterilisation procedures and methodology for removal of donor cells could affect the nature of the allograft tissue. It is therefore very important to provide a collagen based matrix which retains many structural features of human skin. This will facilitate the infiltration of 5 the host keratinocytes and accessory cells, to reconstitute skin at the wound site and to provide a resilient and visually appealing skin graft. It is important that the allograft is free of the donor's host cells, thereby reducing immunogenecity, and is free from infectious agents that may 10 compromise skin graft acceptance by the patient. Currently methods of sterilisation and removal of cellular elements from allogeneic skin result in disruption of the collagen matrix leading to reduced colonisation of the graft by the hosts keratinocytes and early graft rejection. 15 It is therefore an object of the invention to provide a method for producing a skin composite that is both sterile and retains its natural conformation; and the skin composite produced by the method. It is a further object of the invention to provide a skin composite that is both 20 functionally and aesthetically equivalent to natural skin for use in surgical repair of damaged tissue and cosmetic applications. According to a first aspect of the invention there is provided a method for providing a skin composite comprising; 25 i) harvesting a collagen based tissue from a human or animal; ii) removing the cellular elements from said collagen based tissue; WO 00/32252 PCT/GB99/03889 6 iii) dehydrating said collagen based tissue in the presence of a sterilising agent; and iv) removing said dehydrated and sterilised collagen based tissue from said 5 dehydrating and sterilising agents. We have found that the above methodology and in particular step (iii) thereof provides a skin composite that is unique in that it is both sterile and importantly, it retains its integrity or natural conformation. 10 In a preferred embodiment of the invention said collagen based tissue is derived from skin, preferably human skin and more preferably still is an allograft. 15 It will be apparent that the collagen based tissue may originate from alternate sources, for example and not by way of limitation, gastric tissue, venous/arterial vessels, heart valves, bone, oral mucosa. In addition these alternative tissues may serve as autografts, allografts, or xenografts. 20 It will be apparent from the above description that an allograft is the preferred source of collagen based tissue. This reduces the physical and psychological stress on the patient as graft material may be provided from a skin bank and abrogates the need to remove skin from a region of the patient's body as would be required if the graft was an autograft. 25 It will be apparent to one skilled in the art that methods exist to treat collagen based tissues to remove cellular elements from said tissue. For example and not by way of limitation, dermal samples derived from skin can be incubated long term in phosphate buffered saline or hyperosmotic saline solutions or WO 00/32252 PCT/GB99/03889 7 proteolytic enzymes ( e.g. trypsin, dispase) or combinations thereof. Samples are monitored periodically to assess the cellularity of the sample until such time as the dermis is reduced to a collagen matrix which is devoid of cellular elements. Acellular dermal samples can then be stored at -20 0 C without 5 appreciable loss of structure. As mentioned, it is desirable to provide a collagen matrix that is essentially free of cellular material to reduce graft rejection due to immune recognition by the patient. In yet a further preferred method of the invention said dehydration of collagen 10 based tissue is through incubation with a liquid dehydration agent. In yet a further preferred method of the invention said liquid dehydration agent is glycerol. 15 In yet still a further preferred method of the invention said dehydration of collagen based tissue is through a sequential and incremental increase in glycerol concentration. In yet still a further preferred method of the invention said glycerol 20 concentration is increased from at least 50% glycerol to at least 98% glycerol. In yet a further preferred method of the invention said increase in glycerol concentration is in increments of at least 10% glycerol. 25 In yet a further preferred method of the invention said dehydration agent is isopropanol WO 00/32252 PCT/GB99/03889 8 In yet still a further preferred method of the invention said dehydration of collagen based tissue is through the sequential and incremental increase in isopropanol concentration. 5 In yet still a further preferred method of the invention said isopropanol concentration is increased from at least 50% isopropanol to at least 98% isopropanol. In yet a further preferred method of the invention said increase in isopropanol 10 concentration is in increments of at least 10% isopropanol. In yet still a further preferred method of the invention said sterilising agent is a fluid. 15 It will be apparent to one skilled in the art that the sterilising agent may be provided as either a liquid or gas. In yet still a further preferred method of the invention said sterilising agent is a liquid. 20 In yet still a further preferred method of the invention said sterilising agent is gaseous ethylene oxide. In yet a further preferred method of the invention said ethylene oxide is 25 provided at at least 15% ethylene oxide, at least 85% carbon dioxide at an atmospheric pressure of at least 5.5 atmospheres at at least 55 0 C for at least 30 mins. Under these conditions a concentration of 1200 mg ethylene oxide per litre is maintained during the sterilising cycle.

WO 00/32252 PCT/GB99/03889 9 It will be apparent to one skilled in the art that the combination of dehydration in the presence of a sterilising agent has a twofold beneficial effect on the collagen based tissue. Firstly, the dehydration step maintains the integrity of the tissue and facilitates its long term storage. Secondly, the severe effects of 5 for example ethylene oxide, on the collagen based tissue is ameliorated by the presence of, for example glycerol but retains its sterilising effect on infectious agents. According to a second aspect of the invention there is provided a skin 10 composite produced by the aforedescribed method. According to a fourth aspect of the invention there is provided a therapeutic composition comprising the skin composite prepared by the method of the invention. 15 In a preferred embodiment of the invention said therapeutic compostion is provided with storage means which facilitates and/or maintains the natural structural conformation of the skin composite. 20 According to a fifth aspect of the invention there is provided a method according to any previous aspect or embodiment of the invention for the preparation of collagen based tissues for use in cosmetic reconstructive surgery. 25 It will be apparent that for example burn victims, both require remedial tissue reconstruction to provide a functional skin replacement and cosmetic replacement to provide an aesthetically pleasing appearance to the grafted skin. In many examples it is also necessary to provide cosmetic surgery in an attempt to return the appearance of the damaged skin to its former state.

WO 00/32252 PCT/GB99/03889 10 However the collagen based skin composites prepared by the method of the invention also has applications in cosmetic surgery, for example and not by way of limitation, to remove birth marks, skin blemishes or scars. Collagen based tissue composites prepared by the method of the invention would be 5 equally applicable to cosmetic uses as to major tissue reconstruction as a consequence of, for example third degree burns. An embodiment of the invention will now be described, by example only, and with reference to the following figures and tables; 10 Figure 1 represents collagen IV in normal skin (b) and residual collagen IV in glycerol sterilised de-epidermised dermis(d), ethylene oxide sterilised de epidermised dermis (f) and glycerol pretreated ethylene oxide sterilised de epidermised dermis (h); 15 Figure 2 represents the appearance of a skin composite prepared by glycerol treatment; Figure 3 represents the appearance of a skin composite prepared by ethylene 20 oxide treatment; Figure 4 represents the appearance of a skin composite prepared by glycerol and ethylene oxide treatment; 25 Figure 5 represents the appearance of a skin composite prepared using keratinocytes expanded in culture and added either as a suspension or as an autograft sheet; WO 00/32252 PCT/GB99/03889 11 Table 1 represents the production of de-epidermised dermis using ethylene oxide and 1M NaCI under various conditions; Table 2 represents the influence of sterilisation protocols on the physical 5 properties of dermis; Table 3 represents the scoring system used to assess skin composites; Table 4 represents a comparison of the in vitro appearance of skin composites 10 prepared using ethylene oxide and glycerol/ ethylene oxide sterilised skin; Table 5 represents the influence of sterilisation methodology on the appearance of composites grafted onto nude mice; 15 Table 6 represents the influence of sterilisation methodology on the presence of collagen IV in composites on nude mice; and Table 7 represents a comparison of the in vitro appearance of skin composites prepared using freshly isolated and cultured keratinocytes using ethylene 20 oxide sterilised skin. MATERIALS AND METHODS Materials 25 Phosphate buffered saline (PBS) tablets were obtained from Oxoid, Unipath, Hampshire, UK. Dulbecco's modified Eagle's medium was from ICN Flow, Thame, Oxfordshire, UK. Ham's F 12, glutamine, amphotericin B, penicillin and streptomycin were from Gibco Europe, Life Technologies, Paisley, UK. Foetal calf serum (FCS) was from Advanced Protein Products, Brierley Hill, WO 00/32252 PCT/GB99/03889 12 West Midlands, UK. Hydrocortisone was from Nova Biochem, Nottingham. Trypsin was from Difco Laboratories, Detroit, USA. Dispase and Collagenase A were from Boehringer-Mannheim, Lewes, East Sussex, UK. Glycerol and isopropanol were from BDH Laboratory Supplies, Lutterworth, Leicestershire, 5 UK. MTT, cholera toxin, epidermal growth factor (EGF), adenine, insulin, sodium chloride, transferrin, triiodothyronine and primary antibody to laminin, transferrin, triiodothyronine, ethylenediaminetetra acetic acid (EDTA), trypan blue, anti-human laminin antibody, anti-human collagen IV antibody were from Sigma Chemical Co, Poole, Dorset. ABC immunostaining kits was 10 bought from Vector Laboratories, Peterborough, UK. Primary antibody to collagen IV was obtained from DAKO Laboratories, High Wycombe. ; Stainless steel clips, Ligaclips were from Ethicon, Ltd, Edinburgh, UK. Stainless steel grids and rings were supplied by the Department of Medical Physics, Royal Hallamshire Hospital, Sheffield, UK. All pre-sterilised plastic 15 containers for use in tissue culture were from Costar, High Wycombe Buckinghamshire, UK. All tissue was handled in class II laminar flow hoods, provided by Walker Safety Cabinets, Glossop, Derbyshire, UK. Nude mice were obtained and investigated in the Research Department of Northwick Park Hospital, Middlesex, UK. Silicone wound chambers of lcm diameter used on 20 the nude mice were purchased from Renner GMBH, Dannstadt, Germany. Freshly isolated keratocytes Human adult keratinocytes were obtained from harvesting of split thickness skin grafts (STSG) after routine plastic surgical procedures (breast reductions 25 and abdominoplasty), following Goberdhan et al.'s (1993) (17) modification of the Rheinwald and Green (1975) (18) methodology. In brief, small pieces (-0.5cm 2 ) of skin were incubated in 0.1% (w/v) trypsin solution for 16 hours at 40C allowing the epidermal and dermal layers of individual skin samples to be separated and the cells between to be collected, washed, re-suspended and WO 00/32252 PCT/GB99/03889 13 counted. These cells were now ready for use as primary suspensions in skin equivalents, production of secondary keratinocytes and keratinocyte sheets. Keratinocytes expanded in culture 5 Primary keratinocytes in suspension were seeded onto a pre-coated feeder layer of lethally irradiated 3T3 cells (i.3T3). These were cultured and refed every 3 days until approximately 80% confluent, usually by 7-9 days. Residual i.3T3 were removed with 0.02% EDTA before the secondary keratinocytes were detached using trypsin. The suspension of cells was centrifuged at 200g 10 for 5 minutes and the pellet of cells re-suspended in a known volume of keratinocyte medium (KM). A cell count was then performed and these cells were now ready for use as secondary suspensions of keratinocytes. Keratinocyte Sheets 15 2 x 106 primary keratinocytes in suspension were seeded onto a petri dish that had been pre-coated with a feeder layer i3T3 cells (2 x 106 per petri dish) and cultured as previously described until a confluent multilayer sheet of cells. The cultured epithelial sheet was then detached by incubating for approximately 20-30 minutes in a 0.25% dispase solution. Once detached, a tegapore backing 20 dressing was applied to the sheet of cells and secured to the dressing using stainless steel ligaclips, so that the basal cells were facing up, on the backing dressing. The keratinocyte sheets were now ready for application. Human Fibroblasts 25 Fibroblasts were obtained as described by Ralston et al. (In Press) (16). Briefly the dermis left after trypsinisation of split thickness skin graft for keratinocyte extraction was washed in PBS, finely minced and placed in 0.5% collagenase A at 37 0 C for 16 hours.

WO 00/32252 PCT/GB99/03889 14 The resulting digest was collected by centrifugation at 200g for 10 minutes and re-suspended in fibroblast media (FM). At this point a cell count could be performed. This was then placed in T75cm 2 flasks and incubated under the standard culture conditions described. 5 Passaging of Human Fibroblasts 80-90% confluent fibroblasts were enzymatically detached using 0.1% w/v trypsin for 5-10 minutes. The detached cells were aspirated and the trypsin was neutralised by the addition of FCS or FM. The suspension of cells was centrifuged at 200g for 5 minutes. The pellet of cells was re-suspended in a 10 known volume of FM solution. A viable cell count was performed cells were now ready for use in composites, or for further passaging. Only cells between passage 4 and 9 were used in the composites. Preparation of Dermal Substrates 15 Several stages were involved in preparation of dermis. These were sterilisation, de-epidermisation, acellularisation and finally preparation of the DED for experiments. All procedures were carried out using standard aseptic techniques. 20 Sterilization Techniques The 2 sterilisation techniques used for the preparation of dermis were glycerol and ethylene oxide. We also explored pre-treatment of skin prior to ethylene oxide sterilisation by dehydrating skin using either glycerol or isopropanol. 25 Sterilization Using Glycerol STSG was placed sequentially in an excess of 50%, then 85% glycerol in PBS for 24 hours each at room temperature. Finally the STSG was transferred into 98% glycerol, where it was stored at room temperature, for 6 weeks. This methodology enabled the slow dehydration of skin. Previous work from this WO 00/32252 PCT/GB99/03889 15 laboratory showed that 3 weeks in 98% glycerol at room temperature was sufficient to inactivate intracellular viruses in normal human fibroblasts ( Marshall et al 1995)(19). The skin samples were ready to go on to the next stage , that of acellularisation. 5 Sterilization Using Ethylene Oxide Skin was transferred from theatre to the laboratories in normal saline. The first step was to wash the skin in distilled water to eradicate the saline and thereby prevent any future chlorhydrin production during ethylene oxide sterilisation. 10 STSG were washed in 6 changes of and then immersed in fresh distilled water in a sealed container and placed in a shaking water bath for a minimum of 10 minutes. This was repeated a minimum of 3-4 times more. Excess water was allowed to drip off and then the skin was placed in a 90mm petri dish base with fenestrated aluminium foil on top and frozen to -20 0 C at 1oC/minute, 15 before being freeze-dried overnight. Freeze-dried skin was then double bagged before ethylene oxide sterilisation, using a standard SterivitTM process, using a mixture of 15% ethylene oxide and 85% carbon dioxide at a pressure of 5.5 atmospheres at 55 0 C. We also explored an alternative method of ethylene oxide sterilisation using a protocol which did not exceed 37 0 C. This produced 20 similar reticular damage to that observed with our own conditions. A minimum of 3 days was allowed for sample aeration to eradicate residual ethylene oxide gas. Confirmation of ethylene oxide sterilisation was by spore strip results, which took 3-5 days and samples were only released after these 25 results were obtained. Once returned from CSSD, sterilised dry skin could be kept at room temperature in their vacuum-sealed packets, until required for use. This provided a convenient method of storing skin.

WO 00/32252 PCT/GB99/03889 16 Dehydration of Skin prior to Ethylene Oxide Sterilization In order to overcome some problems due to ethylene oxide damage to skin, we explored 2 methods of dehydrating the skin. These were pre-treatment with: 5 1. Glycerol STSG were placed in a large volume of sterilised 50% glycerol in PBS followed by 85% glycerol in PBS, each for 4 hours at room temperature. Finally the STSG was transferred into 98% glycerol, and stored at room 10 temperature for 40 hours. The skin was then firmly padded dry to remove any excess glycerol and placed in autoclave bags and seal with autoclave tape. These were then sent to CSSD for vacuum sealing of bags and ethylene oxide sterilisation as already 15 discussed above. Once returned from CSSD, skin can be kept at room temperature until required for use. 2. Isopropanol. STSG were placed in a large volume of sterilised 50% isopropanol in PBS 20 followed by 75% isopropanol in PBS, each for 4 hours at room temperature. Finally the STSG was transferred into 98% isopropanol , and stored at room temperature for 40 hours. The skin was then firmly padded dry to remove any excess glycerol and 25 placed in autoclave bags and seal with autoclave tape. These were then sent to CSSD for vacuum sealing of bags and ethylene oxide sterilisation as already discussed above. Once returned from CSSD, skin can be kept at room temperature until required for use.

WO 00/32252 PCT/GB99/03889 17 De-epidermisation Following previous work in this laboratory we used two methods for de epidermisation. PBS was used for glycerol sterilised skin (Moore et al 1993)(17), whilst 1M sodium chloride was used for skin sterilised by ethylene 5 oxide. De-epidermisation of Glycerol Treated Skin Glycerolised sterilized skin was rehydrated for at least 4 hours with several changes of sterile PBS containing 625pg/ml amphotericin B, 1000iu/ml 10 penicillin and 1000pg/ml streptomycin. The graft was then placed in an excess of PBS for 4-14 days at 37 0 C to produce a dermal/epidermal split using the modified technique of Krejci et al (1991)(20). Such separation techniques have shown collagen IV and laminin retention on the dermal surface (Woodley et al 1983 and Moore et al 1993)( 21,22). 15 Grafts were transferred into sterile 90mm petri dishes, the epidermis was gently peeled from the dermis and discarded. The remaining dermis was washed in PBS with 625 pg/ml amphotericin B, 1000iu/ml penicillin and 1000pg/ml streptomycin for the rest of the acellularisation duration. 20 De-epidermisation of Ethylene Oxide Treated Skin All skin sterilised with ethylene oxide, irrespective of any prior preparation, underwent this process. 25 Freeze-dried sterilised skin was removed from the vacuum sealed packets and hydrated in a large volume of sterile PBS for 24 hours at 370C to eradicate any residual ethylene glycol. Skin was then immersed in IM sodium chloride solution for 6-8 hours at 37 0 C and the epidermis was removed using sterile forceps. Skin was then rehydrated in an excess of KM for 48 hours at 37 0

C.

WO 00/32252 PCT/GB99/03889 18 Acellularisation Following ethylene oxide sterilisation, the split thickness skin contained anucleate epithelium and no obvious fibroblasts in the dermis. After de 5 epidermisation the dermis was completely acellular and no further processing was required. For glycerolised skin, however, it was necessary to remove cells as below: The de-epidermised dermis was washed in sterile PBS with 0.625tg/ml 10 amphotericin B, 1000iu/ml penicillin and 1000pg/ml streptomycin and then immersed in PBS with 0.625 pg/ml amphotericin B, 1000iu/ml penicillin and 1000ptg/ml streptomycin at 37oC. Twice weekly changes of PBS with 0.625 gg/ml amphotericin B, 1 000iu/ml penicillin and 1000pg/ml streptomycin were performed. 15 After 6 weeks, biopsies of the dermis were taken for confirmation of acellularity by H&E stain. Once grafts were free of cells, a biopsy was taken for culture to confirm sterility. Once this was confirmed to be negative, the DED was ready for use in composite manufacture. 20 Assessment of Stress and Strain of Different Dermal Preparations The stress-strain characteristics and tensile strength were measured by the method of Vogel (1971). Dumb-bell shaped skin strips of 12mm length and 4mm width were punched out in the same orientation from all the dermal 25 preparations. These strips were immersed and stored in phosphate buffer solution (pH 7.2) at 4oC prior to testing. At room temperature, each wet strip was tested uniaxially in tension on an Instron Tensile Tester (Model 1112) in a liquid cell with the deformation applied along the strip length. The thickness of the samples was measured using a screw gauge. The broadened ends were WO 00/32252 PCT/GB99/03889 19 gripped between the Instron grips and the stress-strain curves were recorded at an extension rate of 0.5cm/min. The tensile strength (kg/cm 2 ) was obtained by dividing the alternate load by the area of original cross section (thickness X width). 5 Preparation of de-epidermised dermis (DED) for experiments For each experiment, DED from a single donor was used. The DED was cut into approximately 2 x 2cm pieces for the preparation of skin composites. These pieces of DED were placed in FM in a sterile petri dish for 48 hours in a 10 37 0 C incubator to confirm sterility prior to culture with cells. Preparation of De-Epidermised Dermis (DED) for Experiments For each experiment, DED from a single donor was used. The DED was cut 15 into approximately 20 x 20mm pieces for the preparation of skin composites. These pieces of DED were placed in FM in a sterile petri dish for 48 hours in a 37 0 C incubator to confirm sterility prior to culture with cells. Skin Composite Production 20 In a six well cell culture plate DED was placed with the reticular surface facing uppermost. A stainless steel ring with a lcm internal diameter, was placed in the centre of the DED and pressed down with sterile forceps to effect a seal. A suspension of 1 x 105 fibroblasts in FM was placed in the ring and the surrounding DED was bathed in FM. The FM within the ring was replaced 25, after 24 hours with fresh media. After 48 hours incubation, the medium was aspirated from the well and the stainless steel ring was removed. The DED was turned so that the papillary surface was uppermost and the stainless steel ring was applied onto the corresponding papillary surface. A suspension of 1 x 106 primary keratinocytes (or 3 x 105 in the case of secondary keratinocytes WO 00/32252 PCT/GB99/03889 20 unless otherwise stated) in KM was applied in the stainless steel ring and the remainder of the well was filled with KCM. 24 hours later the KM within the ring was aspirated and the keratinocytes were re-fed fresh KM. The following day, 48 hours after application of keratinocytes, the KM was aspirated from 5 inside the rings and from the wells and the skin equivalents were raised onto stainless steel grids and fresh KM was added to the level of the base of the skin equivalent so that they were at an air liquid interface. Skin composites were cultured for the required period, usually 10 days. 10 Medium was changed every 2 - 3 days and at the end of the time period, they were assessed for their histological morphology and basement membrane and some of these were assessed for their in-vivo performance on nude mice. To analyze the contribution of the keratinocytes and fibroblasts, composites 15 were made using different combinations of cells at different seeding densities and different DED. These combinations included DED with no cells, keratinocytes alone and keratinocytes and fibroblasts together. If cells were not to be added at a particular step, then 0.5ml of the appropriate medium was added in the same way into the ring to avoid any variation in experimental 20 conditions. Grafting of Nude Mice Anaesthetised nude mice were placed on a sterile towel. Graft sites were prepared on the anterolateral back with routine aseptic preparation using 70% 25 alcohol and this was then washed off with sterile normal saline. Two graft sites were applied, one on either side. The back was chosen, as it was inaccessible for the animal to preen this area. A small circular full thickness skin wound was made just big enough to allow the insertion of the WO 00/32252 PCT/GB99/03889 21 silicone wound chambers (Renner GMBH, Dannstadt, Germany) down to the panniculus carnosus, the vascular layer of fascia. The adjacent skin was dissected free of subcutaneous tissue to make room for the flanges of the silicone wound chamber. If the hole was made too large then it was tightened 5 up with a 6/0 prolene suture tied as a purse string. Care was taken to avoid haematoma formation as this can interfere with graft vascularisation. Once the wound chamber was in place, the skin equivalent was trimmed and applied as a lay-on graft. The skin equivalent was then dressed with an overlying non adherent dressing and saline soaked damp gauze, which was cut to size to fill 10 the chamber. Once both sides had chambers and dressings in situ then 2.5cm Elastoplasttm was wrapped around the chambers and the trunk of the mouse tightly enough to secure the positions and not compromise the breathing of the mouse. The mice were then placed in individual cages for 2 weeks, after which they were sacrificed by overdose of intraperitoneal barbiturate. Once 15 the mouse was dead, the wound chambers were removed and the wound and its underlying base were excised en bloc. The specimen was cut into pieces for frozen section and light microscopy. RESULTS Production of sterile acellular allodermis suitable for clinical use 20 Our requirements were that skin should be sterilized and that this sterilization technique should give allodermis with physical properties resembling that of fresh dermis. This sterilization should allow removal of the epidermis on a relatively large area of skin readily and simultaneously provide good retention of basement membrane antigens and a cell-free allodermis from this material. 25 Hence, in the first step of this work, we examined the impact that the method of removal of the epidermis had on retention of basement membrane and loss of cells in ethylene oxide sterilized skin. This followed previous work from our laboratory ( Ghosh et al 1997)(23) in which we showed that ethylene oxide WO 00/32252 PCT/GB99/03889 22 could be used to produce allodermis suitable for composite production although the process of ethylene oxide sterilization led to allodermis with varying degrees of dermal damage. Nevertheless ethylene oxide is a relatively rapid and well accepted method of sterilization. Hence, it was used in this 5 study in preference to glycerol sterilization, which requires a minimum of 3 weeks. Table 1 documents removal of epidermis using IM sodium chloride at 4oC and 37 0 C. We compared removal of epidermis from fresh skin and from ethylene 10 oxide sterilized skin varying the temperature and duration of exposure to 1M sodium chloride. Confirmation of the presence of basement membrane antigens was achieved by immunostaining for collagen IV. The results summarized in Table 1 show that at 37oC we were able to remove the epidermis reliably within 8 hours both from fresh skin and from ethylene 15 oxide skin whereas it requires 72 hours at 4oC to obtain equivalent removal of the epidermis from fresh skin at 4oC. In the case of skin at 4oC, we only lost basement membrane after de-epidermisation at 48 hours. To check for survival of viable fibroblasts, we attempted to culture fibroblasts 20 at different stages of processing from the ethylene oxide sterilized split thickness skin. Up to 4 hours, there were anucleate cells and cellular debris in the ethylene oxide sterilized skin that had been treated with IM sodium chloride at 37 0 C. By 8 hours, there was no evidence of any cellular material. We were unable to culture any viable fibroblasts from ethylene oxide 25 sterilized dermis, which had been subjected to sodium chloride removal of epidermis at any timepoint. Figure 1 illustrates the morphology of (a) normal skin, of (b) glycerol sterilized de-epidermised dermis, of (c) ethylene oxide sterilized de- WO 00/32252 PCT/GB99/03889 23 epidermised dermis and of (d) glycerol pre-treated ethylene oxide sterilized de-epidermised dermis. As can be seen, glycerol sterilization gives a dermis, which is indistinguishable from that of normal skin at the light microscopy level. In contrast, ethylene oxide sterilization often leads to some degree of 5 damage to the reticular dermis. However, glycerol pre-treatment of ethylene oxide sterilized dermis gave a dermal architecture which was indistinguishable from that of normal skin. Ethylene oxide sterilized and glycerol pre-treated ethylene oxide sterilized 10 dermal preparations were routinely produced in 5 days and, by this time, were totally acellular. Figure 2 illustrates the presence of residual collagen IV in (a) normal skin, (b) glycerol sterilized de-epidermised dermis, (c) ethylene oxide sterilized de 15 epidermised dermis and (d) glycerol pre-treated ethylene oxide sterilized de epidermised dermis. As can be seen, there is good retention of collagen IV in the ethylene oxide treated dermis whether pre-treated with glycerol or not. This contrasts strongly with the extent of collagen IV retention in the glycerol sterilized dermis (which required soaking in PBS for a minimum of 4 weeks 20 for removal of cellular debris). Figure 2e also illustrates that allodermis prepared by ethylene oxide sterilization (whether pre-treated with glycerol or not) retain the pliability of normal skin following 2 days soaking in phosphate buffered saline or in medium. This photograph shows a piece of de epidermised acellular dermis draped over a needle to illustrate its pliability. 25 Effect of dehydration of skin prior to ethylene oxide sterilization We were aware from previous studies that ethylene oxide sterilization appears to have the one disadvantage of causing some damage particularly to the reticular dermis. Hence, we explored a number of approaches to preventing WO 00/32252 PCT/GB99/03889 24 this damage while retaining the beneficial effects of ethylene oxide sterilization. We noted that with glycerol (which gives a slow dehydration of the skin), there was no apparent damage to the dermis. Accordingly, we explored initial dehydration of the skin with either glycerol or isopropanol 5 prior to ethylene oxide sterilization. Table 2 shows the effects of the dehydration and sterilisation techniques on the physical strength of the dermal preparations. Five different donor preparations were tested and we examined the effect of glycerol and isopropanol and ethylene oxide on their own as well as in appropriate combinations. All results are expressed as a percentage of 10 the values obtained for fresh and treated dermis. Tensile strength and strain were measured for all samples. Tensile strength or stress is the maximum load (kg/unit area mm 2 ) which the sample could bear before breaking. Strain is the increase in length of the sample/unit of the original length of the sample until the sample broke. Of the various treatments, isopropanol did not affect the 15 tensile strength whilst all others significantly increased it. Paradoxically isopropanol was the only treatment that significantly improved resistance to strain. Although standard ethylene oxide treatment resulted in a high tensile strength, 20 the percentage strain was less than normal skin reflecting the damage to the dermis that we had noted. This skin would break without much elongation reflecting the brittle nature of the skin. Glycerol treated skin showed higher tensile strength than isopropanol treated skin. Both glycerol and isopropanol pre-treatment with subsequent ethylene oxide sterilization resulted in good 25 strength and good elongation compared to fresh dermis. On the basis of physical properties alone, both techniques of dehydration, isopropanol and glycerol, followed by ethylene oxide sterilization would give suitable dermal replacements.

WO 00/32252 PCT/GB99/03889 25 However, in examining the appearance of the isopropanol/ethylene oxide dermis, we found the papillary dermis to have a more tightly packed conformation than in normal skin and there was some discolouration. Hence, this preparation was not carried forward for in vitro or in vivo composite 5 production. Following this investigation of physical properties it was decided to compare composites based on dermis produced using the standard ethylene oxide method or dermis produced using glycerol pre-treatment followed by ethylene 10 oxide sterilization Use of sterile allodenrmis in composite production To compare and standardise results for composite performance both in vitro and in vivo (on nude mice), a scoring system was devised. As detailed in 15 Table 3, this scoring system takes into account the nature of the dermis (whether it had normal architecture or not), the extent of fibroblast penetration into the dermis for those composites where fibroblasts were added, the nature of the keratinocyte layer, the nature of the keratin layer and the quality of the dermal epidermal junction. A maximal score of 15 was possible where both 20 cell types were included in the composites with features resembling normal skin. All composites were scored blind by 3 independent observers throughout the study and the median of these scores was taken. 25 We prepared composites both with and without fibroblasts and we examined the performance of these composites both in vitro and on the nude mice. All composites were cultured for 10 days at an air-liquid interface in vitro and either examined at this stage or 2 weeks post-grafting onto nude mice.

WO 00/32252 PCT/GB99/03889 26 Figures 3, 4 and 5 show composites based on glycerol, ethylene oxide, and glycerol/ethylene oxide prepared dermis respectively. In each Figure, the in vitro and in vivo performance of the composites in the presence and absence of 5 fibroblasts is depicted. Table 4 compares the performance of composites based on ethylene oxide and glycerol/ethylene oxide sterilized skin. Four experiments are compared, each using skin from a separate donor dermis. (Table 5 compares the in vivo 10 performance of the composites prepared using glycerol, ethylene oxide and glycerol/ethylene oxide sterilized composites on the nude mice.) Figure 3 illustrates that glycerol sterilized dermis gave rise to composites which showed a poor epidermal attachment in the absence of fibroblasts once 15 composites had undergone the shearing forces of cryosectioning. It is unlikely that such composites will provide secure epidermal attachment on a wound bed when subjected to any manipulation. Similarly, in vivo, composites prepared with fibroblasts performed relatively well on the nude mice (Figure 3c) but showed poor epidermal attachment on the mice in the absence of 20 fibroblasts (Figure 3d). In contrast, Figures 4 and 5 show that the composites prepared using ethylene oxide sterilization and sodium chloride removal of the epidermis performed much better than the glycerol based composites. Epidermal cells remained 25 attached following cryosectioning both in vitro and ex vivo. Clearly composites could be prepared in the absence of fibroblasts and yet maintain secure epidermal attachment for at least 7 days in culture and after 2 weeks of grafting on the mouse.

WO 00/32252 PCT/GB99/03889 27 Examining Table 4 for the semi-quantitative scoring of the ethylene oxide and glycerol/ethylene oxide composites in vitro, it was clear that inclusion of the glycerol dehydration step prior to ethylene oxide significantly improved the quality of the dermis. In contrast glycerol/ethylene oxide sterilized dermis 5 gave composites with flatter epidermal/dermal junctions than ethylene oxide sterilized composites. However, both protocols of ethylene oxide sterilization gave composites with a similar overall score in vitro (achieving an average of 11 out of a maximum of 15 points in comparison to normal skin). Thus, the addition of the glycerol dehydration step prior to the ethylene oxide 10 sterilization did not lead to any loss of composite quality, while slightly reducing the damage to dermis. Table 5 summarizes the in vivo performance of composites prepared using glycerol, ethylene oxide and glycerol/ethylene oxide sterilized skin focussing 15 on the quality of the collagen IV basement membrane antigen layer in these composites both in the presence and absence of fibroblasts. As can be seen, the quality of the collagen IV in the glycerol composites was relatively poor irrespective of the presence of fibroblasts, whereas a good layer of collagen IV was seen in composites prepared with ethylene oxide or glycerol/ethylene 20 oxide. Thus it appears that the ethylene oxide sterilization protocol is clearly better than the glycerol based sterilization protocol for composite performance in vivo. Collagen IV staining was equally good in ethylene oxide sterilized composites and in those pre-treated with glycerol dehydration prior to ethylene oxide sterilization. 25 Also, we noted that keratinocytes expanded in culture could also be used to produce composites, which performed well (as will be discussed in the following section of the application).

WO 00/32252 PCT/GB99/03889 28 Scaling up and speeding up of production of composites for clinical use To address the issue of speeding up composite production, we then examined the use of keratinocytes expanded in culture in composite production. Our preliminary data in this area show that keratinocytes initially expanded in 5 culture will give reasonable composite performance both in vitro ( data not shown) and in vivo (please see table 5) even in the absence of fibroblasts.

WO 00/32252 PCT/GB99/03889 29 Table 1: Production of DED using ethylene oxide and 1M NaCl solution De-Epidermisation Acellularity BM Retention Hours Skin Skin EOS Skin Skin EOS Skin Skin EOS 4 0 C 37 0 C 37 0 C 4 0 C 37 0 C 37 0 C 4 0 C 37 0 C 37 0 C 0 - - - +++ +++ + +++ +++ +++ 2 - - - +++ +++ + +- +++ +++ 4 - + + +++ +++ + +++ +++ +++ 8 - ++ +++ +++ ++ +++ ++ ++ 16 - +++ +++ +++ ++ - +++ + + 24 + +++ +++ +++ ++ - +++ + + 48 ++ +++ +++ ++ + - ++ + + 72 +++ +++ +++ ++ + - ++ + + 96 +++ + + +++ ++ + - ++ + + 5 EOS = Ethylene oxide sterilized skin De-epidermisation: Acellularity: BM Retention: +++ = complete de-epidermisation +++ = normal cellularity +++ = thick and continuous staining of BM 10 ++ = epidermis comes away in patches but is incomplete ++= abnormal cells ± cellular debris ++ =less intense continuous staining of BM + = epidermis comes away in only a few places with difficulty += cellular debris += discontinuous/patchy staining of BM - = epidermis remains fully attached -= no cellular material -= no BM staining WO 00/32252 PCT/GB99/03889 30 Table 2: The influence of sterilisation protocols on the physical properties of dermis Dermal Preparation % Tensile Strength/Stress % Strain mean ± SEM (p value) mean ± SEM (p value) Fresh dermis 100 100 Glycerol 134 ± 11 (< 0.005) 116 ± 8 Isopropanol 97 ± 13 151 ± 9 (<0.05) Freeze-dried and 153 ± 19 (<0.01) 85 ± 10 ethylene oxide Glycerol and ethylene 175 ± 27 (<0.01) 124 ± 6 oxide Isopropanol and 179 ± 33 (<0.05) 119 ± 6 ethylene oxide All preparations were compared to the equivalent untreated control dermis by 5 Student's t- test WO 00/32252 PCT/GB99/03889 31 Table 3: Skin Composite Scoring System Regio Findings/Description Score Keratin Layer Normal keratin 3 Parakeratosis 2 Thin, "nucleated" layer 1 Non-existent 0 Keratinocyte Layer Organised, thick (> 2-3 cells), differentiated 3 Organised, thin (< 2-3 cells), continuous 2 Monolayer, discontinuous 1 Non-existent 0 Dermo-Epidermal Junction Attached epidermis with rate ridges present 3 (DEJ) Flat DEJ with attached epidermis 2 Partial attachment of epidermis 1 No attachment of epidermal layer 0 Dermis Normal regular reticular and papillary 3 regions Small gaps/holes, disorganised 2 Large gaps/holes, disorganised 1 Disintegrating 0 Fibroblasts Good penetration and presence of cells 3 Good penetration, but few cells present 2 No penetration, cells at reticular surface 1 No cells present 0 WO 00/32252 PCT/GB99/03889 32 Table 4: Comparison of in vitro appearance of composites prepared using ethylene oxide and glycerol/ethylene oxide sterilised skin Ethylene Oxide Glycerol/Ethylene Oxide Significance mean ± SEM mean ± SEM p value Keratin layer: K 1.8 ± 0.06 2.0 ± 0.20 n.s. K & F 1.8 ± 0.15 1.9 ± 0.10 n.s. All together 1.8 ± 0.10 1.9 ± 0.10 n.s. Keratinocyte layer: K 2.1 ± 0.13 2.0 ± 0.04 n.s. K & F 2.6 ± 0.15 2.6 ± 0.18 n.s. All together 2.3 ± 0.12 2.3 ± 0.18 n.s. DEJ: K 2.8 ± 0.14 2.4 ± 0.19 n.s. K & F 2.8 ± 0.10 2.0 ± 0.20 <0.05 All together 2.8 ± 0.09 2.2 ± 0.15 <0.025 Dermal quality: No cells 1.75 ± 0.24 2.4 ± 0.15 n.s. F 2.2 ± 0.20 2.5 ± 0.16 n.s. K 2.2 ± 0.15 2.7 ± 0.09 <0.1 K & F 2.4 ± 0.06 2.6 ± 0.11 n.s. All together 2.1 ± 0.21 2.6 ± 0.12 <0.005 Fibroblast penetration: F 2.1 ± 0.4 1.5 ± 0.4 n.s. K & F 1.4 ± 0.3 2.0 ± 0.2 n.s. All together 1.8 ± 0.3 1.8 ± 0.2 n.s. Total score for composites: K 8.8 ± 0.36 9.1 ± 0.36 n.s. K&F 11.0 ± 0.70 11.1 ± 0.36 n.s. n.s.=notsignificant WO 00/32252 PCT/GB99/03889 33 Table 5: Influence of sterilization methodology on performance of composites on nude mice Sterilization Presence of collagen IV in composite Significance All together (n) protocol No fibroblasts (n) Fibroblasts(n) Glycerol 1 ± 0.3 (7) 0.9 ± 0.5 (7) n.s. 0.9 ± 0.25 (14) Ethylene oxide a 2.3 ± 0.1(15) 2.3 ± 0.2 (15) n.s. 2.3 ± 0.1 (30)*** Ethylene oxide b 2.0 ± 0.3 (5) 2.0 + 0.4 (6) n.s. 2.0 ± 0.2 (11)** Glycerol/ 2.0 ± 0 (5) 2.5 ± 0.2 (4) n.s. 2.2 ± 0.13 (9)** Ethylene oxide 5 Scoring for collagen IV: 0 = none, 1 = patchy, 2= thin and continuous, 3 = thick and continuous All composites were cultured for 10 days in vitro with or without the addition of fibroblasts as indicated and then grafted onto nude mice and assessed at 2 10 weeks. Collagen IV was scored blind on an arbitrary scale of 0-3, where 3 was based on the findings in normal skin. Additionally, keratinocytes initially expanded in culture were also used on ethylene oxide sterilised dermis (indicated by ethylene oxide (b), where (a) denotes freshly isolated cells). 15 The influence of fibroblasts on composite performance was assessed using Student's paired t-test, as each mouse received 2 composites, one with and one without fibroblasts. When it was ascertained that fibroblast addition did not influence collagen IV presence, then the data for composites with and without fibroblasts were combined. Student's non-paired t-test was used to compare 20 between groups and values differing significantly from glycerol sterilized composites are indicated as **p< 0.005 or ***p< 0.001.

WO 00/32252 PCT/GB99/03889 34 DISCUSSION The aim of this study was to progress the development of a human skin replacement based on a sterile human allodermis with cultured autologous 5 keratinocytes for clinical use. We report methodologies which allow composites to be produced and ready for grafting on a small scale within 2 weeks. Confirmation of the in vivo performance of these composites was achieved using the nude mouse model. Many studies of epidermal/dermal composites, based either on human allodermis or artificial dermis show that 10 composites can be produced to an experimental level and can perform well on nude mice ( please see refs 1-12). To the best of our knowledge, this is the first study to demonstrate that such composites can be successfully produced using a sterilised human dermis. We 15 suggest that this will prove a significant advance in the clinical usefulness of this material. The main issues we considered in advancing the development of composites from the laboratory to clinical use were how sterilisation of the skin might 20 affect the morphology of the skin and the residual basement membrane antigens which we had previously established were essential for secure keratinocyte attachment in vitro ( Ralston et al 1998)(16). In the absence of basement membrane antigens, keratinocytes will attach initially but (from our experience) this attachment deteriorates with time in vitro. The most important 25 single factor associated with de-epidermisation is the retention of as much basement membrane as possible at the dermo-epidermal junction. Other groups have reported the importance of the basement membrane in determining secure and preferential attachment of keratinocytes ( Murray et al 1979 and Terranova et al 1980)(26,27).

WO 00/32252 PCT/GB99/03889 35 Our previous work using glycerol sterilised dermis has shown that even when basement membrane antigens are present initially, there is a requirement for continued fibroblasts presence in vitro to avoid deterioration of this basement 5 membrane zone ( Ralston et al 1998)(16). Focusing on the requirement for good basement membrane retention, our objective was firstly to establish a sterilization protocol which allowed retention of a good basement membrane zone following removal of the 10 epithelium and removal of donor fibroblasts. This acellular allodermis had to be suitable for production of composites with a secure epidermal/dermal junction when assembled in the laboratory and when assessed after two weeks of grafting on nude mice. Our second objective related to speeding up composite production and here we wished to see if we could provide such 15 epidermal/dermal attachment without the need for donor fibroblasts. Following on from this, we addressed the problem of scaling up the provision of epidermal/dermal composites for clinical use, making use of this allodermis using keratinocytes either expanded in culture, or freshly isolated and seeded at a lower density to go further. 20 We explored a combination of sterilisation and de-epidermisation techniques, which would allow good retention of basement membrane antigens and good retention of dermal morphology. We found that optimisation of conditions for removal of the epidermis (using IM sodium chloride at 37oC for 8 hours) and 25 introducing a preparative gentle dehydration step prior to ethylene oxide sterilisation of skin gave a de-epidermised acellular sterile dermis with good dermal morphology and good retention of basement membrane antigens. This material would also be suitable for use as a dermal replacement on its own.

WO 00/32252 PCT/GB99/03889 36 We also found that it performed well as a dermal substrate for the production of reconstructed skin using keratinocytes both with and without fibroblasts. With respect to dermal morphology, we have previously shown that glycerol 5 sterilisation of skin is almost certainly based on the hygroscopic properties of glycerol which allow it to slowly dehydrate the skin providing the basis of its bacteriacidal and viricidal activity ( Marshall et al 1995)(19). Thus we have shown that exposure to 98% glycerol for 3 weeks at room temperature is sufficient to prevent infectivity of herpes simplex and polioviruses. However, 10 although glycerol has been shown to be effective in the elimination of viruses, bacteria and fungi ( Marshall et al 1995, Basile et al 1982, Van Baare et al 1994)(19,24,25) it is not a standard method of sterilisation. Dermis produced by this method maintains good histological morphology and remains pliable with good handling properties and retains expression of some basement 15 membrane antigens. Following on from glycerol sterilisation of dermis, we then explored ethylene oxide sterilisation of dermis and showed that successful composites could be produced using ethylene oxide sterilised skin. Ethylene oxide sterilization is a universally recognised method of sterilization used for surgical equipment ( Brigden et al 1980)(28). It is readily available 20 and is a widely used sterilization process, being present in the majority of hospitals. However, there was a variable degree of ethylene oxide induced damage to the dermis which was most evident in the reticular dermis. Using a standard 25 protocol of ethylene oxide sterilisation ie. lyophilisation of material prior to ethylene oxide sterilisation, we were unable to prevent this dermal damage (Ghosh et al 1997)(23). However, we accept that other protocols of ethylene oxide sterilisation do exist which may result in less damage although this is not evident from the literature. The current study shows that by combining the WO 00/32252 PCT/GB99/03889 37 hygroscopic properties of glycerol to achieve a more gentle dehydration over the first 2 days with ethylene oxide sterilisation, we achieve sterilised dermis with very little ethylene oxide induced damage to the reticular dermis. A similar result could be achieved using isopropanol as a preparative 5 dehydration step but we noted that there was some compaction of the papillary dermis with isopropanol and some discolouration of the skin. We did not explore isopropanol/ethylene oxide sterilised skin further in composite production in this study. 10 The issue of remodelling of the basement membrane zone is clearly different in vitro and in vivo. In vivo, basement membrane formation has been demonstrated as early as 6 days on mice grafted with bovine collagen I lattice preseeded with fibroblasts in cultured keratinocytes ( Hansbrough et al 1994)(5). However, in vitro, we found that with glycerol sterilised dermis, 15 both fibroblasts and keratinocytes were required for retention of (or remodelling of) the basement membrane zone when composites were maintained in culture throughout 3 weeks. Other studies have found that remodelling of the basement membrane zone is greatly enhanced if both fibroblasts and keratinocytes are present in composites than only one cell type 20 ( Konig et al 1991 and Ralston et al 1997)(30,16). Up to 10 days, keratinocytes appeared to attach well to composites in the absence of fibroblasts. However, sectioning at this point revealed not only poor epithelial cell attachment to the underlying dermis but also very poor organisation of the epidermal layer. This is consistent with a previous report based on collagen 25 sponge ( Maruguchi et al 1994)(32). Thus, in providing a one-stage material, we decided that it was important to address the issue of secure epidermal cell attachment to the dermis as any grafted material would be expected to undergo some degree of shear forces following grafting onto the patient.

WO 00/32252 PCT/GB99/03889 38 The role of fibroblasts in maintaining/remodelling the basement membrane is not surprising. However, we asked whether good keratinocyte attachment could be achieved with a good basement membrane in the absence of fibroblasts for one practical reason. While it is feasible to culture large 5 numbers of keratinocytes within 2 - 3 weeks, we find it takes much longer to obtain the necessary expansion of autologous fibroblasts for composite production. Thus, if one wanted to use autologous keratinocytes and fibroblasts in composite production, this would have implications for the speed with which reasonable areas of composite material could be produced. 10 Other studies have shown culture and application of composites without fibroblasts onto nude mice ( Medalie et al 1996 and Demarches et al 1992)(11,12) to be successful, in both cases using a non-sterile human acellular de-epidermised dermis. Thus, the question appeared to be that if a good basement membrane region were present, could composites be prepared 15 for clinical use in the absence of fibroblasts? The data presented in this study clearly shows that by 2 weeks there is infiltration of host (mouse) fibroblasts into the composites such that there was no significant advantage to adding human fibroblasts to the composites prior 20 to grafting. Composites engrafted successfully and were stable with secure epidermal/dermal junctions, which resisted the shear forces of sectioning at 2 weeks. We also began to address the final issue of scaling up and speeding up this 25 methodology in this study to make such composites clinically useful. We confirmed that keratinocytes expanded in culture can be used to produce successful composites on nude mice and, further, that there are two options to the expansion either using the application of cultured epithelial autografts onto the sterilised dermal preparations or using passaged keratinocyte suspensions.

WO 00/32252 PCT/GB99/03889 39 Cultured keratinocytes have been used in composites on nude mice (please see refs 3-6 and 8-12) and have been used in conjunction with a human allodermis ( Medalie et al 1996 and Demanches et al 1992)(11,12). Work is now continuing in this laboratory to scale up and speed up the production of 5 such composites based on this information. In conclusion, this study shows that by paying attention to the details of sterilisation of the dermis and removal of the epidermis, a sterilised human acellular DED can be produced with essentially normal dermal morphology 10 and retention of basement membrane antigens. This acellular DED would be appropriate for clinical use as a dermal replacement material. We have also shown that it can be used to produce composites which perform well on nude mice at 2 weeks and that there is the potential to speed up and 15 scale up the production of this material as keratinocytes initially expanded in culture also perform well on this dermis. Expediting and expanding composite production should now be feasible for clinical use. The sterilised allodermis can be stored on the shelf requiring only 20 the addition of the expanded patient keratinocytes. We suggest that this will represent a significant advance on other methodologies currently available for treating patients with extensive full thickness skin loss.

WO 00/32252 PCT/GB99/03889 40 REFERENCES 1. Manning D., Reed N. & Shaffer C., (1973). Maintenance of skin xenografts of widely divergent phylogenetic origin on congenitally 5 athymic (nude) mice. J Exp. Med., 138, 488-494. 2. Demarchez M., Desbas C. & Prunieras M., (1985). Wound healing of human skin transplanted on to the nude mouse. Brit J Dermatol 113, Supplement 28, 177-182. 10 3. Bosca A.R., Tinois E., Faure M., Kanitakis J., Roche P. & Thivolet J., (1988). Epithelial Differentiation of human skin equivalents after grafting onto nude mice. J Invest Dermatol Vol. 91: 136-141, 1988. 15 4. Cooper M.L & Hansbrough J.F., (1991). Use of a composite skin graft composed of cultured human keratinocytes and fibroblasts and a collagen-GAG matrix to cover full-thickness wounds on athymic mice. Surgery 109: 198-206. 20 5. Hansbrough J.F., Morgan J., Greenleaf G., Parikh M., Nolte C. & Wilkins L., (1994). Evaluation of Graftskin Composite Grafts on Full Thickness Wounds on Athymic Mice. J Burn Care Rehabil Vol. 15, No. 4, 346-353. 25 6. Boyce S.T., Foreman T.J., English K.B., Stayner N., Cooper M.L., Sakabu S. & Hansbrough J.F., (1991). Skin wound closure in athymic mice with cultured human cells, biopolymers, and growth factors. Surgery 110 (5): 866-876.

WO 00/32252 PCT/GB99/03889 41 7. Hansbrough J.F., Cooper M.L., Cohen R., Spielvogel R.L., Greenleaf G., Bartel R.L. & Naughton G., (1992). Evaluation of a biodegradable matrix containing cutured human fibroblasts as a dermal replacement 5 beneath meshed skin grafts on athymic mice. Surgery 111: 438-446. 8. Boyce S.T., Medrano E.E., Abdel-Malek Z., Supp A.P., Dodick J.A.M., Nordlund J.J. & Warden G.D., (1993). Pigmentation and Inhibition of Wound Contraction by Cultured Skin Substitutes with Adult 10 Melanocytes After Transplantation to Athymic Mice. J Invest Dermatol Vol.100, No. 4, 360-365. 9. Cooper M.L., Andree C., Hansbrough J.F., Zapata-Sirvent R.L., & Spielvogel R.L., (1993). Direct Comparison of a Cultured Composite 15 Skin Substitute Containing Human Keratinocytes and Fibroblasts to an Epidermal Sheet Graft Containing Human Keratinocytes on Athymic Mice. J Invest Dermatol Vol.101, No. 6, 811-819. 10. Boyce S.T., Supp A.P., Harriger D., Greenhalgh D.G. & Warden G.D., 20 (1995). Topical Nutrients Promote Engraftment and Inhibit Wound Contraction of Cultured Skin Substitutes in Athymic Mice. J Invest Dermatol Vol.104, No. 3, 345-349. 11. Medalie D.A., Erning S.A., Tompkins R.G., Yarmush M.L., Krueger 25 G.G. & Morgan J.R., (1996). Evaluation of Human Skin Reconstituted from Composite Grafts of Cultured Keratinocytes and Human Acellular Dermis Transplanted to Athymic Mice. J Invest Dermatol Vol.107, No. 1,121-127.

WO 00/32252 PCT/GB99/03889 42 12. Demarchez M, Hartmann DJ, Regnier M, Asselineau D (1992). The role of fibroblasts in dermal vascularisation and remodelling of reconstructed human skin after transplantation onto the nude mouse. Transplantation 54,317-326. 5 13. Compton C, Gill JM, Bradford DA, Regauer S, Gallico GG, O'Connor NE, (1989). Skin regenerated from cultured epithelial autografts on full thickness wounds from 6 days to 5 years after grafting: a light, electron microscopic and immunohistochemical study. Lab. Invest, 60, 600-612. 10 14. Desai MH, Mlakar JM, McCauley RL, Abdullah KM, Rutan RL, Wymack JP et al., (1992). Lack of long term durability of cultured keratinocyte burn-wound coverage: a case report. J Burn Care Rehab. 13: 174-180. 15 15. Nanchahal J, Ward CM, (1992). New grafts for old? A review of alternatives to autologous skin. Brit J Plast Surg, 45:354-363. 16. Ralston DR, Layton C, Dalley AJ, Boyce SG, Freedlander E, MacNeil S, 20 (1998). Investigation of basement membrane formation in vitro in human epidermal/dermal skin equivalents. 17. Goberdhan NJ, Dawson Ra, Freedlander E, MacNeil S, (1993). A calmodulin-like protein as an extracellular mitogen for the keratinocyte. 25 Br J Dermatol, 129:678-688. 18. Rheinwald JG, Green H (1975). Serial cultivation of strains of epidermal keratinocytes: the formation of keratinising colonies for single cells. Cell, 6: 331-344.

WO 00/32252 PCT/GB99/03889 43 19. Marshall L, Ghosh MM, Boyce S, Freedlander E, Kudesia G, MacNeil S, (1995). The effect of glycerol on intracellular viruses; implications for the clinical use of glycerol preserved cadaver skin. Bums, 21:356-361. 5 20. Krejci NC, Cuono CB, Langdon RC, McGuire J,(1991). In vitro reconstitution of skin: fibroblasts facilitate keratinocyte growth and differentiation on acellular reticular dermis. J Invest Dermatol 97, 843 848. 10 21. Woodley, D., Sauder, D., Talley, M. J., Silver, M., Grotendorst, G., and Qwarnstrom, E.(1983). Localisation of basement membrane components after dermal-epidermal junction separation. J Invest Dermatol 81, 149 153. 15 22. Moore JC, Ray AK, and Shakespeare, P. ,G.(1993). Surface ultrastructure of human dermis and wounds. Br J Plast Surg 46, 460-465. 23. Ghosh MM, Boyce S, Layton C, Freedlander E, MacNeil S, (1997). A 20 comparison of methodologies for the preparation of human epidermal dermal composites. Ann Plast Surg 39, 390-404. 24. Basile ARD (1982). A comparative study of glycerinised and lyophilised porcine skin in dressings for third degree bums. Plast Reconstr Surg. 69: 25 969-972. 25. Van Baare J, Buitenwerf J, Hoekstra MJ, du Pont JS, (1994). Virucidal effect of glycerol as used in donor skin preservation. Bums 20: S77-S80.

WO 00/32252 PCT/GB99/03889 44 26. Murray JC, Stingl G, Kleinman HK, Martin GR, Katz SI, (1979). Epidermal cells adhere preferentially to Type IV (basement membrane) collagen. J Cell Biol. 80:197-202. 5 27. Terranova VP, Rohrbach DH, Martin GR, (1980). Role of laminin in the attachment of PAM 212 (epithelial) cells to basement membrane collagen Cell, 22:719-726. 28. Brigden R, (1980). Operating Theatre Technique. 4th Ed. Edinburgh: 10 Churchill Livingstone. 29. Kearney JN, Franklin UC, Aguirregocoa V, Holland KT, (1989). Evaluation of ethylene oxide sterilisation of tissue implants. J Hosp Infection, 13: 71-80. 15 30. Konig A, Bruckner-Tuderman L, (1991). Epithelial-mesenchymal interactions enhance expression of collagen VII in vitro. J Invest Dermatol, 96,803-808. 20 31. Ralston DR, Layton C, Dalley AJ, Boyce SG, Freedlander E, MacNeil S. (1997). Keratinocytes contract human dermal extrracellular matrix and reduce soluble fibronectin production by fibroblasts in a skin composite model. Br J Plast Surg 25 32. Maruguchi T, Maruguchi Y, Suzuki S, Matsuda K, Toda K, Isshiki N, (1994). A new skin equivalent: keratinocytes proliferated and differentiated on collagen sponge containing fibroblasts. Plast Reconst Surg, 93, 537-544.

WO 00/32252 PCT/GB99/03889 45 33. Herzog SR, Meyer A, Woodley D, Peterson HD, (1988). Wound coverage with cultured autologous keratinocytes: use after burn wound excision, including biopsy followup. J Trauma, 28: 195-198. 5 34. Woodley DT, Peterson HD, Herzog SR, Stricklin GP, Burgeson RE, Briggaman RE, Cronce DJ, O'Keefe EJ, (1988). Burn wounds resurfaced by cultured epidermal autografts show abnormal reconstitution of anchoring fibrils. JAMA, 259: 2566-2571. 10 35. Green H, Kehinde O, Thomas J, (1979). Growth of cultured human epidermal cells into multiple epithelia suitable for grafting. Proc Natl Acad Sci. USA, 76:5665-5668. 36. Banks-Schlegel S & Green H (1980). Formation of epidermis by serially 15 cultivated human epidermal cells transplanted as an epithelium to athymic mice. Transplantation, 29: 308-313. 37. Cuono et a. Proc.Natl.Acad.Sci. USA, 1987;80;826-835. 20

Claims (23)

1. A method for providing a collagen based tissue for use in tissue replacement comprising: 5 (i) harvesting a collagen based tissue from a human or animal; (ii) removing the cellular elements from said collagen based tissue; (iii) dehydrating said collagen based tissue in the presence of a sterilising agent; and (iv) removing said dehydrated and sterilised collagen based tissue from 10 said dehydrating and sterilising agents.

2. A method according to Claim 1 characterised in that said collagen based tissue is selected from at least one of gastric tissue; venous vessels; arterial vessels; heart valves; bone; oral mucosa; or skin. 15

3. A method according to Claim 1 or 2 characterised in that said collagen based tissue is derived from human skin.

4. A method according to any of Claims 1 - 3 characterised in that said human 20 skin is an allograft.

5. A method according to any of Claims 1 - 4 characterised in that said dehydration of collagen based tissue is through incubation with a liquid dehydration agent. 25

6. A method according to Claim 5 characterised in that said dehydration of collagen based tissue is through a sequential and incremental increase in the concentration of said dehydration agent. 30

7. A method according to Claims 1-6 characterised in that said dehydration agent concentration is increased from at least 50% to at least 98%. WO 00/32252 PCT/GB99/03889 47

8. A method according to Claims 6 or 7 characterised in that said increase in said dehydration agent concentration is in increments of at least 10%. 5

9. A method according to Claims 5 - 8 characterised in that said dehydration agent is glycerol.

10. A method according to Claims 5 - 8 characterised in that said dehydration agent is isopropanol. 10

11. A method according to Claims 1-10 characterised in that said sterilising agent is a fluid.

12. A method according to Claim 11 characterised in that said sterilising agent is 15 a liquid.

13. A method according to Claim 11 characterised in that said sterilising agent is gaseous. 20

14. A method according to Claim 13 characterised in that said sterilising agent is gaseous ethylene oxide.

15. A method according to Claim 14 characterised in that said ethylene oxide is provided at 15% ethylene oxide, carbon dioxide is provided at 85%, at an 25 atmospheric pressure of at least 5.5 atmospheres at between 45-55 0 C for at least 30 minutes.

16. A composition comprising a collagen based tissue produced by the method to any of Claims 1 - 15. 30 WO 00/32252 PCT/GB99/03889 48

17. A composition according to Claim 16 characterised in that said tissue is a human skin composite.

18. A composition according to Claims 18 or 19 for use as a therapeutic 5 composition.

19. A product comprising a therapeutic composition according to Claim 18 and provided with storage means which facilitate and/or maintain the natural structural conformation of the skin composite. 10

20. A method according to Claims 1-15 characterised in that said method is for the preparation of collagen based tissues for use in cosmetic reconstructive surgery. 15

21. A method of treatment comprising a collagen based tissue for use in tissue replacement therapy comprising; i) preparation of a collagen based tissue according to any of Claims 1 15; 20 ii) surgical insertion of said collagen based tissue to a patient to be treated; and, optionally iii) monitoring the status of said inserted collagen based tissue.

22. A method according to Claim 21 characterised in that said collagen based 25 tissue is human skin.

23. An isolated collagen based tissue characterised in that said tissue substantially retains its natural conformation. 30