- Email: [email protected]
In vitro cultured autologous pre-confluent oral keratinocytes for experimental prefabrication of oral mucosa
Int. J. Oral Maxillofac. Surg. 2004; 33: 476–485 doi:10.1016/j.ijom.2003.12.005, available online at http://www.sciencedirect.com
Leading Research Paper Tissue Engineering
In vitro cultured autologous pre-confluent oral keratinocytes for experimental prefabrication of oral mucosa
S. Schultze-Mosgau1, B. -K. Lee1,2, J. Ries1, K. Amann3, J. Wiltfang2 1 Department of Oral and Maxillofacial Surgery, University of Erlangen-Nuremberg, Erlangen, Germany; 2Department of Oral and Maxillofacial Surgery, Seoul National University, 28-2 Yon-gon dong, Jong-no ku, Seoul, South Korea; 3Institute of Pathology, University of Erlangen-Nuremberg, Erlangen, Germany
S. Schultze-Mosgau, B. -K. Lee, J. Ries, K. Amann, J. Wiltfang:In vitro cultured autologous pre-confluent oral keratinocytes for experimental prefabrication of oral mucosa. Int. J. Oral Maxillofac. Surg. 2004; 33: 476–485. # 2004 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved. Abstract. The reconstruction of large defects after head and neck cancer resection often requires composite tissue transfer to replace a combination of bone, muscle and mucosa. Thus, tissue engineering techniques may be useful for oral mucosal reconstructive surgery to prefabricate mucosal tissue on the muscle flap in vivo, instead of using conventional skin-bearing composite flaps. The aim of this study was to investigate whether autogenous pre-confluent oral keratinocytes (PCOK) cultured in vitro can create mucosal coverage on muscle in vivo, in a single grafting procedure. In 30 Wistar rats, with a small piece of oral mucosa (2 mm 5 mm), oral keratinocytes were isolated and then seeded on a hydrophilic PTFE membrane (n ¼ 50) in serum-free culture condition. After 48 h, the membrane, together with the PCOK, was transplanted onto the gracilis muscle to fabricate a mucosal flap in vivo. The wound bed was closed primarily until the time of examination. Biopsies were carried out 1, 2, 3, and 4 weeks, respectively, after transplantation and were evaluated immunohistochemically (AE1/AE3 anti-pancytokeratin, cytokeratin 5/6, collagen IV, laminin, lectin-specific labeling of N-acetylglucosamine oligomeres of endothelial cells) with relation to the following criteria: (1) graft acceptance; (2) inflammatory signs; (3) structural changes and keratinocyte lining; (4) expression of basement membrane components; and (5) vascularization. Ninety-one percent of the grafts showed uniform epithelial layers. The mean number of reconstructed epithelial cell layers was 1.7, 2.0, 1.85 and 2.7 at 1, 2, 3 and 4 weeks, respectively after transplantation (P ¼ 0:342). Collagen IV, laminin and lectin-specific capillaries developed between the neoepithelium and the underlying muscular layer. Only two specimens showed signs of infection 2 weeks after transplantation. In conclusion, this experiment demonstrated that PCOK grafts on muscle in vivo can achieve uniform multi-layered oral epithelial coverage in a short period of time. This technique may be a useful alternative tool for oropharyngeal reconstructive surgery and is also worth considering for further clinical studies.
0901-5027/050476 + 10 $30.00/0
Key words: keratinocyte; pre-confluent; cell culture; prefabrication; oral mucosal flap; tissue engineering; laminin; collagen IV; cytokeratin 5/ 6; lectin; rats; free flaps. Accepted for publication 5 December 2003 Available online 28 March 2004
# 2004 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
In vitro cultured pre-confluent oral keratinocytes Ablative tumor surgery in the oropharyngeal region frequently results in large composite tissue defects, which include mucosa. Skin-bearing composite flaps are usually chosen for reconstruction of the defects. Various types of skin-bearing flaps have been described for oropharyngeal composite defects3,6,7,24,31. In some patients, however, the use of these skinbearing flaps frequently causes complications, such as unwanted hair growth and excessive keratinization in the oropharyngeal recipient site42 which may result in compromised functional or aesthetic impairment after reconstruction. Furthermore, skin harvesting in the donor site frequently results in donor site morbidity or limited availability of skin, or may require additional skin grafting4,10,18,45. Recent tissue engineering techniques may provide a new tool in solving these problems. Since RHEINWALD & GREEN34 first described the technique of transforming epidermal sheets from a minimal skin biopsy using transformed 3T3 feeder cell layers, numerous research groups have studied cultured epidermal sheets in clinical and experimental aspects26,32,43,44,51. However, several disadvantages, such as long culture periods (3–4 weeks), xenotypic gene transfection, low intake rates due to enzymatic treatment and other technical obstacles, have prevented it from becoming more widespread14. In order to overcome such disadvantages of cultured epidermal sheet grafts, pre-confluent keratinocyte grafting has recently been addressed in the field of epidermal tissue engineering13. This concept is based on the fact that relatively undifferentiated keratinocytes, which may include keratinocyte stem cells, are sufficient to achieve subsequent epidermal reconstruction in vivo. Although the process of establishing a standard protocol for the clinical use of this technique has not yet been completed, this could simplify the method and shorten the time periods by which keratinocytes are transplanted to the wound bed. It could also solve most other obstacles of the conventional technique mentioned above. So far, in the field of oral mucosal tissue engineering, few studies have been performed21,23,44. Recently, LAUER & SCHIMMING21,22 reported that tissueengineered mucosal grafts produced according to the technique described by LAUER21 were used as an alternative method to cover the resulting intraoral wound created by freeing of the tongue which was fixed to the adjacent tissue after the resection of small tumors.
However, in the case of large tumors, it may be difficult to primarily close the defect or to wait for secondary reconstruction. The same applies to graft tissue-engineered mucosa directly on the fresh large defect. Therefore, the use of tissue-engineered mucosal grafts in oropharyngeal large defects after cancer ablation may be limited. Moreover, it is quite clear that primary reconstruction for oral composite defects after cancer ablation is more advantageous than secondary reconstruction50. Therefore, in order to overcome both complications of conventional skin-bearing flaps in the reconstruction of oral mucosal composite defects and the limited use of tissue-engineered mucosa grafts for large oropharyngeal cancer defects, we have designed a new concept of composite flaps that replaces the skin portion of conventional skin-bearing composite flaps by autogenous mucosa which is pre-fabricated by grafting preconfluent oral keratinocytes (PCOK) cultured in vitro. Furthermore, to reduce the overall treatment period and to simplify the prefabrication method, we have established a keratinocyte graft technique based on the concept of pre-confluent grafts as mentioned above via a hydrophilic PTFE membrane. In contrast to the clinical reports of an in vitro cultivation of oral keratinocytes for oral mucosa replacement over a period of 3 weeks, the concept of a 48-h in vitro cultivation of PCOK may result in a reduction of the treatment period. Clinical reports have not yet clarified to what extent an ingrowth on the membrane and reepithelialization with pre-existing keratinocytes from the intraoral graft bed takes place. The progression of vascularization over time and the synthesis of the basement membrane components have also only been partially clarified. An in vivo model for transplanting PCOK on a gracilis muscle of the groin was chosen to exclusively examine the in vivo reepithelialization of transplanted PCOKs and to exclude a peripheral ingrowth on the membrane through preexisting oral keratinocytes. A further objective was to establish prefabricated vascular composite grafts consisting of a vascular muscle graft with oral mucosa. The following questions should be clarified: 1. Can grafted autogenous pre-confluent oral keratinocytes (PCOK) cultured in vitro create a uniform mucosal coverage over a muscle flap in vivo in a single grafting procedure?
477
2. In what way does epithelialization progress? To what extent is there a formation of keratinocyte layers? 3. Does vascularization occur? 4. Is it possible to detect components of a basement membrane? 5. Is it possible to prefabricate vascular composite grafts with PCOK? Material and methods Animal model
A total of 30 male Wistar rats (Charles River, Sulzfeld, Germany) with a mean body weight of 400–500 g were used in this study. The animals were first acclimatized in the laboratory under standard conditions approved by the local authorities. The rats received a pelleted standard rodent diet (No. 1320, Altromin) and tap water ad libitum. Antibiotics comprising 10,000 IU of broad-spectrum penicillin i.p. (Tardomyocel, Bayer, Leverkusen, Germany) were administered preoperatively. During the first two postoperative days, the analgesic Buprenorphine HCl (0.216 mg) (Temgesic, Boeringer, Mannheim, Germany) was administered subcutaneously with a dose of 0.2 mg/kg body weight. Local authorities approved this animal experiment (No. 621-2531.31-3/99). Technique of mucosal tissue engineering
The procedure of mucosal tissue engineering on the muscle flap is schematically illustrated in Fig. 1. Isolation of keratinocytes from oral rat mucosa and cell culture on the carrier was carried out step by step as shown in Fig. 1. Three days prior to transplanting the membrane comprising PCOK onto the muscle flap, a split-thickness graft of oral mucosa with a size of 2 mm 5 mm was obtained from the lateral surface of the tongue in each rat. Five additional samples were taken in the biopsy to determine the total number of keratinocytes. The piece of tissue was then washed three times with phosphate-buffered saline (PBS) containing antibiotics, and suspended in a serum-free keratinocyte growth medium (KGM) (Clonetics, San Diego, CA, USA) containing Dispase II (Gibco, Mannheim, Germany) in a concentration of 100 U/ml for 16 h at 4 8C. The epithelial layer was then peeled from the stromal layer in sterile condition. It was then dissociated with 0.25% trypsin (Gibco) for 20 min at room temperature. The trypsin activity was then neutralized by 10% fetal calf serum
478
Schultze-Mosgau et al.
Fig. 1. Schematic illustration of tissue engineering of a mucosal composite flap: After taking a mucosal biopsy (2 mm 5 mm), the keratinocytes were isolated and attached to a PTFE membrane. After in vitro cultivation of 48 h, the pre-confluent keratinocytes were transplanted on the gracilis muscle. The membrane was rotated by 1808, so that the pre-confluent keratinocytes were placed directly on the muscle tissue.
(FCS) in PBS and centrifuged for 5 min at 1500 rpm. The cell pellet was resuspended in 10% FCS in PBS for 30 min. The suspension was then filtered through a 70 mm nylon cell strainer (Falkon, Beckton Dickinson, USA) and again centrifuged for 5 min at 1500 rpm. The cellular pellet was resuspended in 1 ml of KGM and the total vital cells were then counted using a hemocytometer (CASY1, Schaefer, Germany). Approximately 20,000/cm2 keratinocytes were seeded on KGM soaked hydrophilic PTFE membrane (2 cm 2 cm) (Millicell, Millipore, MA, USA) launched in culture plate inserts. Before seeding the cells, the cell culture plate insert was prepared in 6-cell culture plates (Falkon) according to the manufacturer’s protocol. The effective area of the membrane for seeding cells was 4.2 cm2. The cells were cultured for 48 h until transplantation in KGM with a low level (0.15 mM) of Ca2þ concentration and supplemented with growth factors (Clonetics) containing 0.05 mg hEGF (human recombinant epidermal growth factor), 2.5 mg insulin, 0.25 mg hydrocortisone, 25 mg gentamicin, 25 mg amphotericin-B and 15 mg bovine pituitary extract (BPE) in 500 ml KGM, respectively.
Transplantation of the membrane including oral keratinocytes onto the muscle flap
After 48 h of culturing the oral keratinocytes, the recipient bed for the membrane was prepared in Wistar rats. Based on the previous report35, a free gracilis muscle flap model in Wistar rats was chosen as the recipient muscle flap for grafting PCOK. The model was suitable to establish a prefabricated composite flap. Moreover, positioning of the PCOK between the muscle and the subcutaneous fatty tissue ensured the exclusive study of the in vivo growth of the transplanted PCOK and excluded ingrowth through pre-existing keratinocytes. The rats were anesthetized with a mixture of ketamine and rompun at a ratio of 2:1 (2.5 ml/kg body weight). A skin incision was made 2 cm parallel to the leg axis on the inner thigh over the groin muscle. The gracilis muscle was exposed and the muscle fascia was then removed under a microscope (Zeiss, Oberkochen, Germany). Then the membrane including PCOK was excised with a sterilized surgical knife (No. 11) (Schmidt, Nuremberg, Germany) from the cell culture insert. The membrane was placed upside down onto the pre-
pared recipient bed8. This ensured clear identification of the healing areas in question. Once the membrane was positioned, it was not moved to avoid potential cellular damage. The membrane was then fixed with four sutures (10-0 nylon) (Ethicon, Norderstedt, Germany) under a microscope. The wound bed was primarily closed. In total, 50 hydrophilic PTFE membranes including PCOK cultured in vitro were grafted on both gracilis muscles in 25 Wistar rats. Finally, biopsies of the specimen were carried out at 1, 2, 3 and 4 weeks after transplantation. According to the time of the biopsies, these were divided up into Group 1 (n ¼ 10 membranes), Group 2 (n ¼ 14 membranes), Group 3 (n ¼ 16 membranes), Group 4 (n ¼ 10 membranes), respectively. Five rats (n ¼ 10 membranes) served as a control group. All surgical procedures were performed as described above, but without oral keratinocytes on the membrane. Immunohistochemical analysis
The tissue specimens were fixed with 4% paraformaldehyde-neutral buffer solution. After being embedded in paraffin, the biopsy specimens were cut vertically
In vitro cultured pre-confluent oral keratinocytes (4 mm) and then H&E stained. In order to confirm the presence of epithelial cells, indirect immunofluorescent staining specific for epithelial cells was also performed on the paraffin-embedded tissue sections as mentioned above. The immunofluorescent staining was carried out as follows: the specimens were collected on gelatin-potassium-chrome-sulfate-coated slides. The slides were incubated with 1% bovine serum albumin (BSA) in 0.1 M PBS, pH 7.6, for 30 min to eliminate nonspecific binding of primary antibodies. AE1/AE3 anticytokeratin monoclonal antibodies (DAKO, Glostrup, Denmark) in a dilution of 1:100 and anti-pan cytokeratin monoclonal antibodies (Santa Cruz, CA, USA) in a dilution of 1:50 were applied to the slides for 4 h at room temperature. Slides were then incubated in the dark with fluorescein-isothiocyanate (FITC)labeled secondary antibodies (goat-antimouse) (Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany) in a dilution of 1:800 for 1 h at room temperature. The avidin–biotin–peroxidase complex (ABC-POX) method was used to analyze the expression of the epitopes of cytokeratin 5/6, collagen IV and laminin as components of the basement membrane, and lectin as an assay for vascularization36–38. After rehydration in TBS (0.05 M Tris–HCl; 0.15 M NaCl; pH 7.6) (5 min, room temperature (RT)), H2O2 (3%, 20 min, RT) was used to suppress the endogenous peroxidase activity. The specific epitopes were demasked with protease solution (0.1% trypsin, 0.1% CaCl2 in TBS; pH > 7:8; 15 min 37 8C). Incubation with the primary antibody was carried out with cytokeratin 5/6-specific mouse-IgG (#180267; Zymed Laboratories Inc., South San Francisco, USA) at RT over a period of 12 h in TBS/2.5% BSA in a dilution of 1:50, with collagen-IV-specific goat-IgG (#1340-01, Southern Biotechnology Associates Inc., Birmingham, AL, USA) at 4 8C over a period of 12 h in TBS/2.5% BSA in a dilution of 1:500, with laminin-specific rabbit-IgG (#2233PLA, Trinova Biochem GmbH, Giessen, Germany) at 4 8C over a period of 12 h in TBS/2.5% BSA in a dilution of 1:750. As a secondary antibody, biotinylated IgG-rabbit-anti-mouse-antibody (Dako, Glostrup, Denmark) was used in a dilution of 1:50 (30 min, RT) for cytokeratin 5/6, biotinylated IgG-goat-antirabbit-antibody (Vector Laboratories, Inc., Burlingame, USA) in a dilution of 1:200 (30 min, RT) for laminin, and biotinylated IgG-rabbit-anti-goat-antibody
(Vector Laboratories, Inc.) in a dilution of 1:200 (30 min, RT) for collagen IV. For marking with the avidin–biotin–peroxidase complex (Dako), incubation was carried out for 30 min at RT. Lectin-specific marking was used as an endothelial marker to acquire the extent of vascularization. The sensitivity and specificity of a selective lectin-binding to N-acetyl-glucosamine-oligomeres of the cell structure of endothelium was proven2. For this, incubation was carried out with biotinylated lectin (l-0651, Sigma-Aldrich Chemie GmbH) at 37 8C over a period of 1 h in TBS/2.5% BSA in a dilution of 1:200. The avidin was coupled directly to the biotinylated lectin (1:2000). Three sections were taken per sample with one negative control (incubation with BSA without primary antibody). A preparation known to be positive was also stained. The chromogenic coupling was carried out with AEC (0.02% 3-amino-9-ethylcarbazole in 50 mM acetate buffer pH 5; 5.5% dimethylformanide) (Dako) and H2O2 (conc. 0.005%). The sections were examined qualitatively under a bright-field microscope (Axioskop, Zeiss) at a magnification of 10–40 according to the following criteria: (1) graft acceptance; (2) inflammatory signs; (3) structural changes and keratinocyte lining; (4) expression of basement membrane components; and (5) vascularization. A criterion for successful grafting was a continuous, uniform lining of mucosal keratinocytes over the muscle flap. Discontinuous or undetectable epithelial lining was regarded as a graft failure. The number of keratinocyte layers was determined for histomorphometric quantification. For this, the method of randomized systematic subsampling was used1,47. The following methodical procedure was used to determine the number of keratinocyte layers below the membrane: Four visual fields were selected per sample per rat/week and evaluated under a bright-field microscope (Aristoplan, Leitz, Wetzlar, Germany) at a magnification of 40. The analysis was carried out independently by three examiners, and mean values were formed. The criterion for a cellular layer of positively expressing keratinocytes was the existence of a clear cell structure with specific cytoplasmatic staining. By counting the number of nuclei of epithelial cells placed on an imaginary vertical line to the membrane, the epithelial thickness was measured in 40 representative H&E cross-sections at
479
the membrane margins and four points equally spaced across the membrane. Statistics
The differences between the groups were tested non-parametrically using the Kruskal–Wallis test and the Mann–Whitney U-test. The data from three sections per sample/rat were aggregated. The data were shown as boxplots as median values and interquartile ranges (IQR). An ANOVA test was carried out to check the influence of the healing time after in vivo transplantation on the increase of keratinocyte cell layers. Two-sided Pvalues of 0.05 were considered to be significant. All calculations were carried out using the SPSS V11 program under Windows (SPSS Inc., Chicago, USA). Results Graft acceptance
Three rats (n ¼ 6) died by the 1st postoperative day due to surgical trauma or anesthesia complications. These were excluded from this study. A total of 44 membranes with in vitro cultivated PCOK were examined. Forty-two of 44 membranes showed irritation-free healing at the time of taking the samples. Two membranes showed florid purulent inflammation at the time of taking the samples. One of them was of Group 2, the other (n ¼ 3) was of Group 3. Epithelial structure—expression of cytokeratin 5/6
A uniform lining of epithelial cells was observed directly below the membrane in 40 membranes. The immunofluorescence assay of AE1/AE3 anti-cytokeratin and pancytokeratin (Fig. 2a and b) and the immunohistochemical assay of an expression of cytokeratin 5/6 (Fig. 2c and d) prove that this was transplanted PCOK. In 40 of 44 membranes it was possible to show a uniform epithelial layer with cytoplasmatic expression of cytokeratin 5/6 of the epithelial cells attached to the membrane. This corresponds to a success rate of 91%. The epithelial structure at each point during the study is shown in Fig. 3a–d. One week after transplantation of the PCOK, one- to two-layer section of keratinocytes was found below the membrane (Fig. 3a). The keratinocytes had a small square cell structure. A layer of connective tissue with fibroblasts and extracellular matrix components were
480
Schultze-Mosgau et al.
Fig. 2. (a) Left: Single-layer keratinocytes below the membrane 1 week after in vivo grafting (H&E; original magnification 40). Right: Immunofluorescence-optical assay of AE1/AE3 anti-cytokeratin and pancytokeratin. In the region of the single keratinocyte layer a demarcation of deeper layers fluorescence-optical illustration of a pancytokeratin marking (indirect immunofluorescence; original magnification 40). (b) Left: Multi-layer keratinocytes below the membrane 3 weeks after in vivo transplantation of PCOKs (H&E; original magnification 15). Right: It was possible to verify keratinocytes through the immunofluorescence-optical assay of AE1/AE3 anti-cytokeratin and pancytokeratin (indirect immunofluorescence; original magnification 15). (c) The specific expression of cytokeratin 5/6 in the cell layers confirms the presence of a uniform, two- to three-layer keratinocyte layer 2 weeks after the transplantation of PCOKs (cytokeratin 5/6; original magnification 25). (d) Expression of cytokeratin 5/6 in the keratinocyte cell layers (two to five layers) 4 weeks after transplantation (cytokeratin 5/6; original magnification 25).
found between the epithelial layer and the muscles without showing signs of extensive leucocytic infiltrates (Fig. 3a). At this point in time there were hardly any capillaries in the connective tissue as an indication of vascularization. Figure 3b shows an increase of keratinocyte lining 2 weeks after transplantation with simultaneous formation of individual
cylindrical cell structures. In the adjacent layer of connective tissue, individual capillaries with small lumina were seen. In some areas the epithelial layer detached from the PTFE membrane (Fig. 3b). Three weeks after transplantation (Fig. 3c), a clear increase in the number of capillaries and capillary lumina was observed. The adjacent connective
tissue showed numerous fibroblasts without signs of infiltrates of inflamed cells. Four weeks after transplantation (Fig. 3d), a multilayer of keratinocytes with cylindrical cells was observed. The histomorphometric analysis of the keratinocyte layers is shown in Fig. 4. A median value of 1.7 (1.0–3.0) was measured for the number of keratinocyte
In vitro cultured pre-confluent oral keratinocytes
481
Fig. 3. (a) One week after transplantation of the PCOK, one- to two-layer section of keratinocytes was found below the membrane. The keratinocytes had a small square cell structure. A layer of connective tissue with fibroblasts and extracellular matrix components was found between the epithelial layer and the muscles without showing signs of extensive leucocytic infiltrates (H&E; original magnification 25). (b) Two weeks after transplantation of the PCOK, an increase of keratinocyte lining with simultaneous formation of individual cylindrical cell structures was seen. In the adjacent layer of connective tissue, individual capillaries with small lumina were seen. In some areas the epithelial layer detached from the PTFE membrane (H&E; original magnification 25). (c) Three weeks after transplantation of the PCOK, a clear increase of the number of capillaries and capillary lumina was observed. The adjacent connective tissue showed numerous fibroblasts without signs of infiltrates of inflamed cells (H&E; original magnification 25). (d) Four weeks after transplantation a multilayer of keratinocytes with cylindrical cells was observed (H&E; original magnification 25).
layers 1 week after transplantation, of 2.0 (0.0–4.0) after 2 weeks, of 1.85 (0.0–4.0) after 3 weeks, and of 2.7 (1.0–5.0) after 4 weeks. Significant differences were not observed at the various points of time during the study (P ¼ 0:342, ANOVA, F ¼ 1:148: df ¼ 3). Basement membrane components: expression of collagen IV and laminin
Fig. 4. The histomorphometric analysis of the keratinocyte layers. A median value of 1.7 (1.0– 3.0) was measured for the number of keratinocyte layers 1 week after transplantation, of 2.0 (0.0–4.0) after 2 weeks, of 1.85 (0.0–4.0) after 3 weeks, and of 2.7 (1.0–5.0) after 4 weeks. Significant differences were not observed at the various points of time during the study (P ¼ 0:342, ANOVA, F ¼ 1:148: df ¼ 3).
The positive expression of collagen IV and laminin served as an indication for the structure of a basement membrane (Figs 5a–c and 6). Positive expression of collagen IV was found at all points of time during the study. Collagen IV was expressed below the keratinocyte cell layers. One week after transplantation (Fig. 5a), parallel collagen IV structures were seen in a broad zone below the
keratinocytes. Overall, vascularization was low. Individual capillaries were seen below the relaxed layer of connective tissue. Three weeks (Fig. 5b) and 4 weeks (Fig. 5c) after transplantation, solidification of the collagen IV structures was observed directly below the keratinocyte cell layers. Increased capillaries were found below the collagen IV fibrous layer as an indication for neovascularization. Laminin was proven below the keratinocyte cell layers (Fig. 6). Vascularization: lectin-specific endothelial marking
Lectin-specific expression of N-acetylglucosamine-oligomeres in the cell membrane structures of endothelial cells of capillaries is shown in Fig. 7a and b. In addition to an increase of collagen IV
482
Schultze-Mosgau et al.
Fig. 5. (a) One week after transplantation, collagen IV was expressed below the keratinocyte cell layers. Parallel collagen IV structures were seen in a broad zone below the keratinocytes. Overall, vascularization was low. Individual capillaries were seen below the relaxed layer of connective tissue (collagen IV; original magnification 25). (b) Three weeks after transplantation, solidification of the collagen IV structures was observed directly below the keratinocyte cell layers. Increased capillaries were found below the collagen IV fibrous layer as an indication for neovascularization (collagen IV; original magnification 25). (c) Four weeks after transplantation, collagen IV structures were found below the keratinocyte cell layers (collagen IV; original magnification 25).
fiber, capillaries were seen from the 2nd week onwards in the direct vicinity of the transplanted PCOKs as an indication for an associated vascularization (Fig. 7a).
The capillaries had small lumina and were mainly localized in groups. After 4 weeks, an increase of the capillary lumina was seen (Fig. 7b). The capillaries were observed below the collagen IV fiber layer (Fig. 7b). Discussion
Fig. 6. Laminin 1 week after transplantation: laminin was proven in the basement membrane like structures from the 1st week onwards after transplantation (laminin; original magnification 25).
It was possible to show that pre-confluent keratinocytes cultivated over 48 h in vitro can be transplanted on a non-resorbent membrane onto a muscle. It was possible to prove that during the in vivo healing period of 4 weeks, a further growth of keratinocyte cell structures and the formation of neoepithelium consisting of two- to three-layer keratinocyte layers occur. The highly-prismatic and cylindrical cell structures and the assay of cytokeratin 5/6 and pancytokeratin proved the specificity of keratinocytes.
Fig. 7. (a) Lectin-specific expression of N-acetyl-glucosamine-oligomeres in the cell membrane structures of endothelial cells of capillaries: In addition to an increase of collagen IV fiber, capillaries were seen from the 2nd week onwards in the direct vicinity of the transplanted PCOKs as an indication for an associated vascularization (lectin; original magnification 25). (b) Lectin-specific expression of N-acetyl-glucosamine-oligomeres in the cell membrane structures of endothelial cells of capillaries: The capillaries had small lumina and were mainly localized in groups. After 4 weeks, an increase of the capillary lumina was seen. The capillaries were observed below the collagen IV fiber layer (lectin; original magnification 25).
With the model of subcutaneous transplantation onto the gracilis muscle it was possible to show that these were in vitro cultivated and transplanted keratinocytes, and that an in vivo construction of oral neoepithelium on muscle tissue is possible in regard to prefabrication of vascular flaps with oral mucosa. Basement membrane like structures, particularly collagen IV and laminin, were proven below the neoepithelium. These extracellular matrix (ECM) structures of the basement membrane are a prerequisite for cell adhesion, migration and differentiation of keratinocytes. The adherence of the keratinocytes with ECM structures of the basement membrane takes place via integrin receptors, e.g., a5b128. The basal keratinocyte layer is highly proliferative, but does not express any markers of terminal differentiation. Keratinocytes that migrated into the upper cell layers and that lost contact with the basal membrane expressed differentiation markers28. An existing environment that permits the formation of a keratinocyte basal membrane contact and cell matrix interactions is an advantage of early transplantation of pre-confluent keratinocytes under in vivo conditions. Here, this model is superior to the in vitro cultivation of keratinocytes on feeder layers of 3T3cells over several weeks11. Since collagen IV fibers were proven in the relaxed connective tissue 1 week after in vivo transplantation, it can be assumed that basement membrane like structures develop on the basis of pre-existing connective tissue. Studies on cell-ECM interactions and sequential deposition of ECM molecules after wounding showed that de novo deposition of fibronectin occurs 24 h after wounding and is followed by deposition of collagen IV and laminin16.
In vitro cultured pre-confluent oral keratinocytes Lectin-specific vascularization was also proven directly below the basement membrane like structures. The increase in the number of capillaries and the individual capillary lumina at the various points in time during the study suggests neoangiogenesis or the formation of capillaries from pre-existing vessels. The observed topographical proximity of basement membrane like structures and capillaries indicates a connection between the existing basal membrane and neoangiogenesis19,20. The significance of laminin in addition to other ECM structures for endothelial cell interactions has already been reported5,12. Laminin is responsible as a disulfide-linked complex for the attachment of keratinocytes to the basement membrane via a6b4 integrins in hemidesmosomes29,39. Furthermore, laminin influences angiogenesis via interaction with the avb3 integrin9. Here, collagen IV and laminin form the basis for the development and arrangement of endothelial cells in addition to proteoglycans. Since endothelial cells migrate and penetrate adjacent tissues, and matrixdegrading enzymes are active, it is assumed that angiogenetic factors are released from the matrix or through appropriate cleavage of ECM molecules such as laminin49. The detachment of the neoepithelium from the PTFE membrane in some areas during the shrinkage shows that the neoepithelium, the basement membrane like structures, the newly formed capillaries and the underlying connective and muscle tissue form a stable structure. It was demonstrated that PCOK prepared with a small piece of autogenous oral mucosa (2 mm 5 mm) can successfully reconstruct uniform and continuous epithelial layers on the muscle flap in vivo. By seeding keratinocytes at a density of 20,000 cells/cm2, the mucosa could be expanded to 2 cm 2 cm. The success rate (91%) of graft acceptance is significantly higher than that of conventional cultured epithelial sheets in previous studies (15–75%)52. It has been reported that serum-free medium with a low calcium concentration may serve to decrease the amount of differentiated cells and increase the number of cells able to adhere and proliferate8,41. If keratinocytes can be transferred to the patient whilst still hyperproliferative, before terminal differentiation is induced, then the wound bed may act as a culture system, allowing the cells to attach and proliferate, rapidly reforming the complete epidermis in vivo. Furthermore, several growth factors that stimulate
wound repair25,53 and certain proteins such as fibrin15 and hyaluronic acid48 that involve the wound healing process may offer favorable conditions for cell proliferation. In addition to these advantages, the short period of in vitro cultivation of 48 h is seen as an advantage. The use of bovine pituitary extract with the residual risk of contamination of the cell culture is a disadvantage. Here the objective is to substitute the bovine content with a non-bovine content in further experiments. So far, in order to deliver pre-confluent keratinocytes to the wound bed, several carriers such as fibrin glue17, polyurethane film33 and collagen14, have been introduced in various models. Regarding the issue of mucosal engineering on the muscle in vivo, WECHSEL46 recently reported that a BERGER et al. muscle can be prelaminated by simple injection onto the muscle with autogenously cultured urothelial mucosal cells via fibrin glue as a carrier. Although WECHSELBERGER et al. showed some successful results, they could not reconstruct uniform and continuous cellular lining along the muscle, which is important for reconstructive surgery. With respect to this result, when pre-confluent mucosal keratinocytes are transplanted, it may be necessary to use a matrix on which transplanted keratinocytes can migrate, forming uniform mucosal coverage along the muscle. With the hydrophilic PTFE membrane used in this study, PCOK were successfully delivered, and continuous neo-epithelium was reconstructed on the muscle flap along the membrane. In addition to the function of a carrier, the objective was to ensure a demarkation of keratinocytes from the covering soft tissue. The spontaneous detachment of the neoepithelium showed that a future removal of the PTFE membrane is possible. The additional use of a collagen membrane or of collagen gel plus SPARC cells would certainly result in an improvement of the cell culture system27. This would facilitate the construction of basement like structures and the early attachment of keratinocytes to the ECM components of the basement membrane. As mentioned in previous reports14,40, cultured epidermal sheets or epithelial cells show some shrinkage after grafting due to natural wound contraction. Likewise, the transplanted membrane also shows some shrinkage. It has been reported that differentiation and maturation of epithelial cells require air exposure30, but it is currently unclear whether
483
these epithelial cells can differentiate and mature after being exposed to the oral environment. From a clinical point of view, we are currently conducting subsequent research to determine whether this prefabricated mucosal flap model can be adapted to the recipient site after transplantation. Furthermore, we are also considering several cytokine treatments to facilitate the epithelialization of grafted keratinocytes in vivo. In summary, this experiment demonstrated that PCOK grafts on the muscle in vivo could achieve uniform multi-layered oral epithelial coverage on the muscle flap in a short period of time. Although the process of establishing a standard protocol for the clinical use of this technique has not yet been completed, it may be a useful tool for future mucosal composite reconstructive surgery in the oral and maxillofacial region that is worth considering for further studies. Acknowledgments. This Research Project was supported by the grants of the Wilhelm-Sander-Foundation No. 2002.017.1 and the Interdisciplinary Center for Clinical Research (IZKF) Project No. B34, financed by the Federal Ministry of Education and Research (BMBF). We would also like to thank Dr. Mai, Department of Anatomy II, Friedrich-Alexander University ErlangenNuremberg, for his kind support with this study. References 1. Amann K. New parameters in kidney biopsy diagnostic—morphometry. Kidney Blood Press Res 2000: 23: 181–182. 2. Amann K, Breitbach M, Ritz E, Mall G. Myocyte/capillary mismatch in the heart of uremic patients. J Am Soc Nephrol 1998: 9: 1018–1022. 3. Alcalde J, Garcia-Tapia R, Espinosa JM, Perez N. Clinical application of free microvascular flaps in reconstructive surgery of the oral cavity. Acta Otorrinolaringol Esp 1994: 45: 457–460. 4. Avery CM, Pereira J, Brown AE. Suprafascial dissection of the radial forearm flap and donor site morbidity. Int J Oral Maxillofac Surg 2001: 30: 37–41. 5. Baldwin HS. Early embryonic development. Cardiovasc Res 1996: 31: 34–45. 6. Baumann I, Greschniok A, Bootz F, Kaiserling E. Free transplanted, microvascular reanastomosed forearm flap for reconstruction of the mouth cavity and oropharynx. Clinical and morphologic findings with special reference to reinnervation. HNO 1996: 44: 616–623. 7. Bortolani A, Barisoni D, Chiamenti C, Lorenzini M, Pasqualini M. Intra-
484
8.
9.
10.
11.
12. 13.
14.
15.
16.
17.
18.
19. 20.
21.
22.
Schultze-Mosgau et al.
oral cancer: immediate reconstruction using a free radial forearm flap. Acta Otorhinolaryngol Ital 1994: 14: 29–40. Boyce ST, Ham RG. Calcium-regulated differentiation of normal human epidermal keratinocytes in chemically defined clonal culture and serum-free serial culture. J Invest Dermatol 1983: 81: 33–40. Brooks PC, Montgomery AM, Rosenfeld M. Integrin alpha v beta 3 antagonists promote tumor regression by inducing apoptosis of angiogenetic blood vessels. Cell 1994: 79: 1157–1164. Graham B, Adkins P, Scheker LR. Complications and morbidity of the donor and recipient sites in 123 lateral arm flaps. J Hand Surg Br 1992: 17: 189–192. Gross J. Getting to mammalian wound repair and amphibian limb regeneration: a mechanistic link in the early events. Wound Repair Regener 1996: 4: 190–202. Hanahan D. Signaling vascular morphogenesis and maintenance. Science 1997: 277: 48–50. Harris PA, Leigh IM, Navsaria HA. Pre-confluent keratinocyte grafting: the future for cultured skin replacements? Burns 1998: 24: 591–593. Horch RE, Debus M, Wagner G, Stark GB. Cultured human keratinocytes on type I collagen membranes to reconstitute the epidermis. Tissue Eng 2000: 6: 53–67. Isogai N, Ueda Y, Kurozumi N, Kamiishi H. Wound healing at the site of microvascular anastomosis: fibrin-stabilizing factor XIII administration and its effects. Microsurgery 1990: 11: 40–46. Kamei M, Kawasaki A, Tano Y. Analysis of extracellular matrix synthesis during wound healing of retinal pigment epithelial cells. Microsc Tes Tech 1998: 42: 311–316. Kaiser HW, Stark GB, Kopp J, Balcerkiewicz A, Spilker G, Kreysel HW. Cultured autologous keratinocytes in fibrin glue suspension, exclusively and combined with STS-allograft (preliminary clinical and histological report of a new technique). Burns 1994: 20: 23–29. Kimata Y, Uchiyama K, Ebihara S, Saikawa M, Hayashi R, Haneda T. Postoperative complications and functional results after total glossectomy with microvascular reconstruction. Plast Reconstr Surg 2000: 106: 1028–1035. Kumar CC. Signaling by integrin receptors. Oncogene 1998: 17: 1355–1363. Kumar CC, Yoneda J, Bucana CD, Fidler IJ. Regulation of distinct steps of angiogenesis by different angiogenic molecules. Int J Oncol 1998: 12: 749– 757. Lauer G. Autografting of feeder-cell free cultured gingival epithelium. Method and clinical application. J Craniomaxillofac Surg 1994: 22: 18–22. Lauer G, Schimming R. Tissue-engineered mucosa graft for reconstruction of the intraoral lining after freeing of
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
the tongue: a clinical and immunohistologic study. J Oral Maxillofac Surg 2001: 59: 169–175. Lauer G, Schimming R, Frankenschmidt A. Intraoral wound closure with tissue-engineered mucosa: new perspectives for urethra reconstruction with buccal mucosa grafts. Plast Reconstr Surg 2001: 107: 25–33. Lind MG, Arnander C, Gylbert L, Heden P, Jurell G. Reconstruction in the head and neck regions with free radial forearm flaps and split-rib bone grafts. Am J Surg 1987: 154: 459–462. Martin P, Hopkinson-Woolley J, McCluskey J. Growth factors and cutaneous wound repair. Prog Growth Factor Res 1992: 4: 25–44. McKay I, Woodward B, Wood K, Navsaria HA, Hoekstra H, Green C. Reconstruction of human skin from glycerol-preserved allodermis and cultured keratinocyte sheets. Burns 1994: 20(Suppl 1): 19–22. Mishima H, Hibino T, Hara H, Murakami J, Otori T. SPARC from corneal epithelial cells modulates collagen contraction by keratocytes. Invest Opthalmol Vis Sci 1998: 39: 2547–2553. Nicholson LJ, Watt FM. Decreased expression of fibronectin and the a5b1 integrin during terminal differentiation of human keratinocytes. J Cell Sci 1991: 98: 225–232. Niessen CM, Hogervorst F, Jaspars LH. The alpha 6 beta 4 integrin is a receptor for both laminin and kalinin. Exp Cell Res 1994: 211: 360–367. Ootani A, Toda S, Fujimoto K, Sugihara H. An air-liquid interface promotes the differentiation of gastric surface mucous cells (GSM06) in culture. Biochem Biophys Res Commun 2000: 271: 741–746. Pickford MA, Soutar DS. Intraoral reconstruction using a second free flap for recurrent or metachronous carcinoma. Br J Plast Surg 1995: 48: 559–563. Regauer S, Compton CC. Cultured keratinocyte sheets enhance spontaneous reepithelialization in a dermal explant model of partial-thickness wound healing. J Invest Dermatol 1990: 95: 341–346. Rennekampff HO, Hansbrough JF, Kiessig V, Abiezzi S, Woods Jr V. Wound closure with human keratinocytes cultured on a polyurethane dressing overlaid on a cultured human dermal replacement. Surgery 1996: 120: 16–22. Rheinwald JG, Green H. Serial cultivation of strains of human epidermal keratinocytes: the formation of keratinizing colonies from single cells. Cell 1975: 6: 331–343. Schultze-Mosgau S, Keilholz L, Rodel F, Labahn D, Neukam FW. Experimental model for transplantation of a modified free myocutaneous gracilis flap to an irradiated neck region in rats. Int J Oral Maxillofac Surg 2001: 30: 63–69.
36. Schultze-Mosgau S, Grabenbauer GG, Radespiel-Tro¨ger M, Wiltfang J, Ries J, Neukam FW, Ro¨del F. Vascularisation in the transition area between free grafted soft tissues and pre-irradiated graft bed tissues following preoperative radiotherapy in the head and neck region. Head Neck 2002: 24: 42–51. 37. Schultze-Mosgau S, Ro¨del F, Wehrhan F, Grabenbauer GG, Amann K, Radespiel-Tro¨ger M, Neukam FW. Transforming growth factor b1 and b2 (TGFb1/TGFb2) profile changes in previously irradiated free flap beds. Head Neck 2002: 24: 33–41. 38. Schultze-Mosgau S, Wehrhan F, Amann K, Radespiel-Tro¨ger M, Ro¨del F, Grabenbauer GG. In vivo TGFb3 expression in wound healing in irradiated tissues—an experimental study. Strahlenther Onkol 2003: 179: 410–416. 39. Sonnenberg A, Calafat J, Janssen H. Integrin alpha 6/beta 4 complex is located in hemidesmosomes, suggesting a major role in epidermal cell-basement membrane adhesion. J Cell Biol 1991: 113: 907–917. 40. Sugimura Y, Hata K, Torii S, Ueda M. Transplantation of cultured mucosal epithelium: an experimental study. J Craniomaxillofac Surg 1996: 24: 352–359. 41. Svensjo T, Yao F, Pomahac B, Eriksson E. Autologous keratinocyte suspensions accelerate epidermal wound healing in pigs. J Surg Res 2001: 99: 211– 221. 42. Toft K, Keller GS, Blackwell KE. Ectopic hair growth after flap reconstruction of the head and neck. Arch Fac Plast Surg 2000: 2: 148–150. 43. Tomson AM, Scholma J, Blaauw EH. The effect of detachment of cultured epithelial sheets on cellular ultrastructure and organisation. Epithelial Cell Biol 1995: 4: 43–51. 44. Ueda M. Formation of epithelial sheets by serially cultivated human mucosal cells and their applications as a graft material. Nagoya J Med Sci 1995: 58: 13–28. 45. vander LB, Spronk CA, de Visscher JG. Closure of radial forearm free flap donor site with local full-thickness skin graft. Br J Oral Maxillofac Surg 1999: 37: 119–122. 46. Wechselberger G, Bauer T, Meirer R, Piza-Katzer H, Lille S, Russell RC. Muscle prelamination with urothelial cell cultures via fibrin glue in rats. Tissue Eng 2001: 7: 153–159. 47. Weibel ER. Measuring through the microscope: development and evolution of stereological methods. J Microsc 1989: 155: 393–403. 48. Weigel PH, Fuller GM, LeBoeuf RD. A model for the role of hyaluronic acid and fibrin in the early events during the inflammatory response and wound healing. J Theor Biol 1986: 119: 219–234.
In vitro cultured pre-confluent oral keratinocytes 49. Werb Z, Vu TH, Rinkenberger JL, Coussens LM. Matrix-degrading proteases and angiogenesis during development and tumor formation. Apmis 1999: 107: 11–18. 50. Wynne JM, Apostolis CG, Perl TZ, McQuaide JR, Graham T, Levin W. Function-preserving procedures in primary reconstruction following surgery for advanced oral cancer—an appraisal. S Afr J Surg 1978: 16: 191–197. 51. Xu W, Li H, Brodniewicz T, Auger FA, Germain L. Cultured epidermal
sheet grafting with Hemaseel HMN fibrin sealant on nude mice. Burns 1996: 22: 191–196. 52. Yanaga H, Tai Y, Yamauchi T, Mori S, Udoh Y, Yamamoto M. Take percentage and conditions of cultured epithelium were improved when basement membrane of the graft was maintained and anchoring mesh was added. Plast Reconstr Surg 2001: 107: 105– 115. 53. Yates RA, Nanney LB, Gates RE, King Jr LE. Epidermal growth factor
485
and related growth factors. Int J Dermatol 1991: 30: 687–694. Address: Stefan Schultze-Mosgau Glueckstrasse 11 91054 Erlangen Germany. Tel: þ49 9131 853 3601 Fax: þ49 9131 853 4219 E-mail: [email protected]. uni-erlangen.de
Leading Research Paper Tissue Engineering
In vitro cultured autologous pre-confluent oral keratinocytes for experimental prefabrication of oral mucosa
S. Schultze-Mosgau1, B. -K. Lee1,2, J. Ries1, K. Amann3, J. Wiltfang2 1 Department of Oral and Maxillofacial Surgery, University of Erlangen-Nuremberg, Erlangen, Germany; 2Department of Oral and Maxillofacial Surgery, Seoul National University, 28-2 Yon-gon dong, Jong-no ku, Seoul, South Korea; 3Institute of Pathology, University of Erlangen-Nuremberg, Erlangen, Germany
S. Schultze-Mosgau, B. -K. Lee, J. Ries, K. Amann, J. Wiltfang:In vitro cultured autologous pre-confluent oral keratinocytes for experimental prefabrication of oral mucosa. Int. J. Oral Maxillofac. Surg. 2004; 33: 476–485. # 2004 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved. Abstract. The reconstruction of large defects after head and neck cancer resection often requires composite tissue transfer to replace a combination of bone, muscle and mucosa. Thus, tissue engineering techniques may be useful for oral mucosal reconstructive surgery to prefabricate mucosal tissue on the muscle flap in vivo, instead of using conventional skin-bearing composite flaps. The aim of this study was to investigate whether autogenous pre-confluent oral keratinocytes (PCOK) cultured in vitro can create mucosal coverage on muscle in vivo, in a single grafting procedure. In 30 Wistar rats, with a small piece of oral mucosa (2 mm 5 mm), oral keratinocytes were isolated and then seeded on a hydrophilic PTFE membrane (n ¼ 50) in serum-free culture condition. After 48 h, the membrane, together with the PCOK, was transplanted onto the gracilis muscle to fabricate a mucosal flap in vivo. The wound bed was closed primarily until the time of examination. Biopsies were carried out 1, 2, 3, and 4 weeks, respectively, after transplantation and were evaluated immunohistochemically (AE1/AE3 anti-pancytokeratin, cytokeratin 5/6, collagen IV, laminin, lectin-specific labeling of N-acetylglucosamine oligomeres of endothelial cells) with relation to the following criteria: (1) graft acceptance; (2) inflammatory signs; (3) structural changes and keratinocyte lining; (4) expression of basement membrane components; and (5) vascularization. Ninety-one percent of the grafts showed uniform epithelial layers. The mean number of reconstructed epithelial cell layers was 1.7, 2.0, 1.85 and 2.7 at 1, 2, 3 and 4 weeks, respectively after transplantation (P ¼ 0:342). Collagen IV, laminin and lectin-specific capillaries developed between the neoepithelium and the underlying muscular layer. Only two specimens showed signs of infection 2 weeks after transplantation. In conclusion, this experiment demonstrated that PCOK grafts on muscle in vivo can achieve uniform multi-layered oral epithelial coverage in a short period of time. This technique may be a useful alternative tool for oropharyngeal reconstructive surgery and is also worth considering for further clinical studies.
0901-5027/050476 + 10 $30.00/0
Key words: keratinocyte; pre-confluent; cell culture; prefabrication; oral mucosal flap; tissue engineering; laminin; collagen IV; cytokeratin 5/ 6; lectin; rats; free flaps. Accepted for publication 5 December 2003 Available online 28 March 2004
# 2004 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
In vitro cultured pre-confluent oral keratinocytes Ablative tumor surgery in the oropharyngeal region frequently results in large composite tissue defects, which include mucosa. Skin-bearing composite flaps are usually chosen for reconstruction of the defects. Various types of skin-bearing flaps have been described for oropharyngeal composite defects3,6,7,24,31. In some patients, however, the use of these skinbearing flaps frequently causes complications, such as unwanted hair growth and excessive keratinization in the oropharyngeal recipient site42 which may result in compromised functional or aesthetic impairment after reconstruction. Furthermore, skin harvesting in the donor site frequently results in donor site morbidity or limited availability of skin, or may require additional skin grafting4,10,18,45. Recent tissue engineering techniques may provide a new tool in solving these problems. Since RHEINWALD & GREEN34 first described the technique of transforming epidermal sheets from a minimal skin biopsy using transformed 3T3 feeder cell layers, numerous research groups have studied cultured epidermal sheets in clinical and experimental aspects26,32,43,44,51. However, several disadvantages, such as long culture periods (3–4 weeks), xenotypic gene transfection, low intake rates due to enzymatic treatment and other technical obstacles, have prevented it from becoming more widespread14. In order to overcome such disadvantages of cultured epidermal sheet grafts, pre-confluent keratinocyte grafting has recently been addressed in the field of epidermal tissue engineering13. This concept is based on the fact that relatively undifferentiated keratinocytes, which may include keratinocyte stem cells, are sufficient to achieve subsequent epidermal reconstruction in vivo. Although the process of establishing a standard protocol for the clinical use of this technique has not yet been completed, this could simplify the method and shorten the time periods by which keratinocytes are transplanted to the wound bed. It could also solve most other obstacles of the conventional technique mentioned above. So far, in the field of oral mucosal tissue engineering, few studies have been performed21,23,44. Recently, LAUER & SCHIMMING21,22 reported that tissueengineered mucosal grafts produced according to the technique described by LAUER21 were used as an alternative method to cover the resulting intraoral wound created by freeing of the tongue which was fixed to the adjacent tissue after the resection of small tumors.
However, in the case of large tumors, it may be difficult to primarily close the defect or to wait for secondary reconstruction. The same applies to graft tissue-engineered mucosa directly on the fresh large defect. Therefore, the use of tissue-engineered mucosal grafts in oropharyngeal large defects after cancer ablation may be limited. Moreover, it is quite clear that primary reconstruction for oral composite defects after cancer ablation is more advantageous than secondary reconstruction50. Therefore, in order to overcome both complications of conventional skin-bearing flaps in the reconstruction of oral mucosal composite defects and the limited use of tissue-engineered mucosa grafts for large oropharyngeal cancer defects, we have designed a new concept of composite flaps that replaces the skin portion of conventional skin-bearing composite flaps by autogenous mucosa which is pre-fabricated by grafting preconfluent oral keratinocytes (PCOK) cultured in vitro. Furthermore, to reduce the overall treatment period and to simplify the prefabrication method, we have established a keratinocyte graft technique based on the concept of pre-confluent grafts as mentioned above via a hydrophilic PTFE membrane. In contrast to the clinical reports of an in vitro cultivation of oral keratinocytes for oral mucosa replacement over a period of 3 weeks, the concept of a 48-h in vitro cultivation of PCOK may result in a reduction of the treatment period. Clinical reports have not yet clarified to what extent an ingrowth on the membrane and reepithelialization with pre-existing keratinocytes from the intraoral graft bed takes place. The progression of vascularization over time and the synthesis of the basement membrane components have also only been partially clarified. An in vivo model for transplanting PCOK on a gracilis muscle of the groin was chosen to exclusively examine the in vivo reepithelialization of transplanted PCOKs and to exclude a peripheral ingrowth on the membrane through preexisting oral keratinocytes. A further objective was to establish prefabricated vascular composite grafts consisting of a vascular muscle graft with oral mucosa. The following questions should be clarified: 1. Can grafted autogenous pre-confluent oral keratinocytes (PCOK) cultured in vitro create a uniform mucosal coverage over a muscle flap in vivo in a single grafting procedure?
477
2. In what way does epithelialization progress? To what extent is there a formation of keratinocyte layers? 3. Does vascularization occur? 4. Is it possible to detect components of a basement membrane? 5. Is it possible to prefabricate vascular composite grafts with PCOK? Material and methods Animal model
A total of 30 male Wistar rats (Charles River, Sulzfeld, Germany) with a mean body weight of 400–500 g were used in this study. The animals were first acclimatized in the laboratory under standard conditions approved by the local authorities. The rats received a pelleted standard rodent diet (No. 1320, Altromin) and tap water ad libitum. Antibiotics comprising 10,000 IU of broad-spectrum penicillin i.p. (Tardomyocel, Bayer, Leverkusen, Germany) were administered preoperatively. During the first two postoperative days, the analgesic Buprenorphine HCl (0.216 mg) (Temgesic, Boeringer, Mannheim, Germany) was administered subcutaneously with a dose of 0.2 mg/kg body weight. Local authorities approved this animal experiment (No. 621-2531.31-3/99). Technique of mucosal tissue engineering
The procedure of mucosal tissue engineering on the muscle flap is schematically illustrated in Fig. 1. Isolation of keratinocytes from oral rat mucosa and cell culture on the carrier was carried out step by step as shown in Fig. 1. Three days prior to transplanting the membrane comprising PCOK onto the muscle flap, a split-thickness graft of oral mucosa with a size of 2 mm 5 mm was obtained from the lateral surface of the tongue in each rat. Five additional samples were taken in the biopsy to determine the total number of keratinocytes. The piece of tissue was then washed three times with phosphate-buffered saline (PBS) containing antibiotics, and suspended in a serum-free keratinocyte growth medium (KGM) (Clonetics, San Diego, CA, USA) containing Dispase II (Gibco, Mannheim, Germany) in a concentration of 100 U/ml for 16 h at 4 8C. The epithelial layer was then peeled from the stromal layer in sterile condition. It was then dissociated with 0.25% trypsin (Gibco) for 20 min at room temperature. The trypsin activity was then neutralized by 10% fetal calf serum
478
Schultze-Mosgau et al.
Fig. 1. Schematic illustration of tissue engineering of a mucosal composite flap: After taking a mucosal biopsy (2 mm 5 mm), the keratinocytes were isolated and attached to a PTFE membrane. After in vitro cultivation of 48 h, the pre-confluent keratinocytes were transplanted on the gracilis muscle. The membrane was rotated by 1808, so that the pre-confluent keratinocytes were placed directly on the muscle tissue.
(FCS) in PBS and centrifuged for 5 min at 1500 rpm. The cell pellet was resuspended in 10% FCS in PBS for 30 min. The suspension was then filtered through a 70 mm nylon cell strainer (Falkon, Beckton Dickinson, USA) and again centrifuged for 5 min at 1500 rpm. The cellular pellet was resuspended in 1 ml of KGM and the total vital cells were then counted using a hemocytometer (CASY1, Schaefer, Germany). Approximately 20,000/cm2 keratinocytes were seeded on KGM soaked hydrophilic PTFE membrane (2 cm 2 cm) (Millicell, Millipore, MA, USA) launched in culture plate inserts. Before seeding the cells, the cell culture plate insert was prepared in 6-cell culture plates (Falkon) according to the manufacturer’s protocol. The effective area of the membrane for seeding cells was 4.2 cm2. The cells were cultured for 48 h until transplantation in KGM with a low level (0.15 mM) of Ca2þ concentration and supplemented with growth factors (Clonetics) containing 0.05 mg hEGF (human recombinant epidermal growth factor), 2.5 mg insulin, 0.25 mg hydrocortisone, 25 mg gentamicin, 25 mg amphotericin-B and 15 mg bovine pituitary extract (BPE) in 500 ml KGM, respectively.
Transplantation of the membrane including oral keratinocytes onto the muscle flap
After 48 h of culturing the oral keratinocytes, the recipient bed for the membrane was prepared in Wistar rats. Based on the previous report35, a free gracilis muscle flap model in Wistar rats was chosen as the recipient muscle flap for grafting PCOK. The model was suitable to establish a prefabricated composite flap. Moreover, positioning of the PCOK between the muscle and the subcutaneous fatty tissue ensured the exclusive study of the in vivo growth of the transplanted PCOK and excluded ingrowth through pre-existing keratinocytes. The rats were anesthetized with a mixture of ketamine and rompun at a ratio of 2:1 (2.5 ml/kg body weight). A skin incision was made 2 cm parallel to the leg axis on the inner thigh over the groin muscle. The gracilis muscle was exposed and the muscle fascia was then removed under a microscope (Zeiss, Oberkochen, Germany). Then the membrane including PCOK was excised with a sterilized surgical knife (No. 11) (Schmidt, Nuremberg, Germany) from the cell culture insert. The membrane was placed upside down onto the pre-
pared recipient bed8. This ensured clear identification of the healing areas in question. Once the membrane was positioned, it was not moved to avoid potential cellular damage. The membrane was then fixed with four sutures (10-0 nylon) (Ethicon, Norderstedt, Germany) under a microscope. The wound bed was primarily closed. In total, 50 hydrophilic PTFE membranes including PCOK cultured in vitro were grafted on both gracilis muscles in 25 Wistar rats. Finally, biopsies of the specimen were carried out at 1, 2, 3 and 4 weeks after transplantation. According to the time of the biopsies, these were divided up into Group 1 (n ¼ 10 membranes), Group 2 (n ¼ 14 membranes), Group 3 (n ¼ 16 membranes), Group 4 (n ¼ 10 membranes), respectively. Five rats (n ¼ 10 membranes) served as a control group. All surgical procedures were performed as described above, but without oral keratinocytes on the membrane. Immunohistochemical analysis
The tissue specimens were fixed with 4% paraformaldehyde-neutral buffer solution. After being embedded in paraffin, the biopsy specimens were cut vertically
In vitro cultured pre-confluent oral keratinocytes (4 mm) and then H&E stained. In order to confirm the presence of epithelial cells, indirect immunofluorescent staining specific for epithelial cells was also performed on the paraffin-embedded tissue sections as mentioned above. The immunofluorescent staining was carried out as follows: the specimens were collected on gelatin-potassium-chrome-sulfate-coated slides. The slides were incubated with 1% bovine serum albumin (BSA) in 0.1 M PBS, pH 7.6, for 30 min to eliminate nonspecific binding of primary antibodies. AE1/AE3 anticytokeratin monoclonal antibodies (DAKO, Glostrup, Denmark) in a dilution of 1:100 and anti-pan cytokeratin monoclonal antibodies (Santa Cruz, CA, USA) in a dilution of 1:50 were applied to the slides for 4 h at room temperature. Slides were then incubated in the dark with fluorescein-isothiocyanate (FITC)labeled secondary antibodies (goat-antimouse) (Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany) in a dilution of 1:800 for 1 h at room temperature. The avidin–biotin–peroxidase complex (ABC-POX) method was used to analyze the expression of the epitopes of cytokeratin 5/6, collagen IV and laminin as components of the basement membrane, and lectin as an assay for vascularization36–38. After rehydration in TBS (0.05 M Tris–HCl; 0.15 M NaCl; pH 7.6) (5 min, room temperature (RT)), H2O2 (3%, 20 min, RT) was used to suppress the endogenous peroxidase activity. The specific epitopes were demasked with protease solution (0.1% trypsin, 0.1% CaCl2 in TBS; pH > 7:8; 15 min 37 8C). Incubation with the primary antibody was carried out with cytokeratin 5/6-specific mouse-IgG (#180267; Zymed Laboratories Inc., South San Francisco, USA) at RT over a period of 12 h in TBS/2.5% BSA in a dilution of 1:50, with collagen-IV-specific goat-IgG (#1340-01, Southern Biotechnology Associates Inc., Birmingham, AL, USA) at 4 8C over a period of 12 h in TBS/2.5% BSA in a dilution of 1:500, with laminin-specific rabbit-IgG (#2233PLA, Trinova Biochem GmbH, Giessen, Germany) at 4 8C over a period of 12 h in TBS/2.5% BSA in a dilution of 1:750. As a secondary antibody, biotinylated IgG-rabbit-anti-mouse-antibody (Dako, Glostrup, Denmark) was used in a dilution of 1:50 (30 min, RT) for cytokeratin 5/6, biotinylated IgG-goat-antirabbit-antibody (Vector Laboratories, Inc., Burlingame, USA) in a dilution of 1:200 (30 min, RT) for laminin, and biotinylated IgG-rabbit-anti-goat-antibody
(Vector Laboratories, Inc.) in a dilution of 1:200 (30 min, RT) for collagen IV. For marking with the avidin–biotin–peroxidase complex (Dako), incubation was carried out for 30 min at RT. Lectin-specific marking was used as an endothelial marker to acquire the extent of vascularization. The sensitivity and specificity of a selective lectin-binding to N-acetyl-glucosamine-oligomeres of the cell structure of endothelium was proven2. For this, incubation was carried out with biotinylated lectin (l-0651, Sigma-Aldrich Chemie GmbH) at 37 8C over a period of 1 h in TBS/2.5% BSA in a dilution of 1:200. The avidin was coupled directly to the biotinylated lectin (1:2000). Three sections were taken per sample with one negative control (incubation with BSA without primary antibody). A preparation known to be positive was also stained. The chromogenic coupling was carried out with AEC (0.02% 3-amino-9-ethylcarbazole in 50 mM acetate buffer pH 5; 5.5% dimethylformanide) (Dako) and H2O2 (conc. 0.005%). The sections were examined qualitatively under a bright-field microscope (Axioskop, Zeiss) at a magnification of 10–40 according to the following criteria: (1) graft acceptance; (2) inflammatory signs; (3) structural changes and keratinocyte lining; (4) expression of basement membrane components; and (5) vascularization. A criterion for successful grafting was a continuous, uniform lining of mucosal keratinocytes over the muscle flap. Discontinuous or undetectable epithelial lining was regarded as a graft failure. The number of keratinocyte layers was determined for histomorphometric quantification. For this, the method of randomized systematic subsampling was used1,47. The following methodical procedure was used to determine the number of keratinocyte layers below the membrane: Four visual fields were selected per sample per rat/week and evaluated under a bright-field microscope (Aristoplan, Leitz, Wetzlar, Germany) at a magnification of 40. The analysis was carried out independently by three examiners, and mean values were formed. The criterion for a cellular layer of positively expressing keratinocytes was the existence of a clear cell structure with specific cytoplasmatic staining. By counting the number of nuclei of epithelial cells placed on an imaginary vertical line to the membrane, the epithelial thickness was measured in 40 representative H&E cross-sections at
479
the membrane margins and four points equally spaced across the membrane. Statistics
The differences between the groups were tested non-parametrically using the Kruskal–Wallis test and the Mann–Whitney U-test. The data from three sections per sample/rat were aggregated. The data were shown as boxplots as median values and interquartile ranges (IQR). An ANOVA test was carried out to check the influence of the healing time after in vivo transplantation on the increase of keratinocyte cell layers. Two-sided Pvalues of 0.05 were considered to be significant. All calculations were carried out using the SPSS V11 program under Windows (SPSS Inc., Chicago, USA). Results Graft acceptance
Three rats (n ¼ 6) died by the 1st postoperative day due to surgical trauma or anesthesia complications. These were excluded from this study. A total of 44 membranes with in vitro cultivated PCOK were examined. Forty-two of 44 membranes showed irritation-free healing at the time of taking the samples. Two membranes showed florid purulent inflammation at the time of taking the samples. One of them was of Group 2, the other (n ¼ 3) was of Group 3. Epithelial structure—expression of cytokeratin 5/6
A uniform lining of epithelial cells was observed directly below the membrane in 40 membranes. The immunofluorescence assay of AE1/AE3 anti-cytokeratin and pancytokeratin (Fig. 2a and b) and the immunohistochemical assay of an expression of cytokeratin 5/6 (Fig. 2c and d) prove that this was transplanted PCOK. In 40 of 44 membranes it was possible to show a uniform epithelial layer with cytoplasmatic expression of cytokeratin 5/6 of the epithelial cells attached to the membrane. This corresponds to a success rate of 91%. The epithelial structure at each point during the study is shown in Fig. 3a–d. One week after transplantation of the PCOK, one- to two-layer section of keratinocytes was found below the membrane (Fig. 3a). The keratinocytes had a small square cell structure. A layer of connective tissue with fibroblasts and extracellular matrix components were
480
Schultze-Mosgau et al.
Fig. 2. (a) Left: Single-layer keratinocytes below the membrane 1 week after in vivo grafting (H&E; original magnification 40). Right: Immunofluorescence-optical assay of AE1/AE3 anti-cytokeratin and pancytokeratin. In the region of the single keratinocyte layer a demarcation of deeper layers fluorescence-optical illustration of a pancytokeratin marking (indirect immunofluorescence; original magnification 40). (b) Left: Multi-layer keratinocytes below the membrane 3 weeks after in vivo transplantation of PCOKs (H&E; original magnification 15). Right: It was possible to verify keratinocytes through the immunofluorescence-optical assay of AE1/AE3 anti-cytokeratin and pancytokeratin (indirect immunofluorescence; original magnification 15). (c) The specific expression of cytokeratin 5/6 in the cell layers confirms the presence of a uniform, two- to three-layer keratinocyte layer 2 weeks after the transplantation of PCOKs (cytokeratin 5/6; original magnification 25). (d) Expression of cytokeratin 5/6 in the keratinocyte cell layers (two to five layers) 4 weeks after transplantation (cytokeratin 5/6; original magnification 25).
found between the epithelial layer and the muscles without showing signs of extensive leucocytic infiltrates (Fig. 3a). At this point in time there were hardly any capillaries in the connective tissue as an indication of vascularization. Figure 3b shows an increase of keratinocyte lining 2 weeks after transplantation with simultaneous formation of individual
cylindrical cell structures. In the adjacent layer of connective tissue, individual capillaries with small lumina were seen. In some areas the epithelial layer detached from the PTFE membrane (Fig. 3b). Three weeks after transplantation (Fig. 3c), a clear increase in the number of capillaries and capillary lumina was observed. The adjacent connective
tissue showed numerous fibroblasts without signs of infiltrates of inflamed cells. Four weeks after transplantation (Fig. 3d), a multilayer of keratinocytes with cylindrical cells was observed. The histomorphometric analysis of the keratinocyte layers is shown in Fig. 4. A median value of 1.7 (1.0–3.0) was measured for the number of keratinocyte
In vitro cultured pre-confluent oral keratinocytes
481
Fig. 3. (a) One week after transplantation of the PCOK, one- to two-layer section of keratinocytes was found below the membrane. The keratinocytes had a small square cell structure. A layer of connective tissue with fibroblasts and extracellular matrix components was found between the epithelial layer and the muscles without showing signs of extensive leucocytic infiltrates (H&E; original magnification 25). (b) Two weeks after transplantation of the PCOK, an increase of keratinocyte lining with simultaneous formation of individual cylindrical cell structures was seen. In the adjacent layer of connective tissue, individual capillaries with small lumina were seen. In some areas the epithelial layer detached from the PTFE membrane (H&E; original magnification 25). (c) Three weeks after transplantation of the PCOK, a clear increase of the number of capillaries and capillary lumina was observed. The adjacent connective tissue showed numerous fibroblasts without signs of infiltrates of inflamed cells (H&E; original magnification 25). (d) Four weeks after transplantation a multilayer of keratinocytes with cylindrical cells was observed (H&E; original magnification 25).
layers 1 week after transplantation, of 2.0 (0.0–4.0) after 2 weeks, of 1.85 (0.0–4.0) after 3 weeks, and of 2.7 (1.0–5.0) after 4 weeks. Significant differences were not observed at the various points of time during the study (P ¼ 0:342, ANOVA, F ¼ 1:148: df ¼ 3). Basement membrane components: expression of collagen IV and laminin
Fig. 4. The histomorphometric analysis of the keratinocyte layers. A median value of 1.7 (1.0– 3.0) was measured for the number of keratinocyte layers 1 week after transplantation, of 2.0 (0.0–4.0) after 2 weeks, of 1.85 (0.0–4.0) after 3 weeks, and of 2.7 (1.0–5.0) after 4 weeks. Significant differences were not observed at the various points of time during the study (P ¼ 0:342, ANOVA, F ¼ 1:148: df ¼ 3).
The positive expression of collagen IV and laminin served as an indication for the structure of a basement membrane (Figs 5a–c and 6). Positive expression of collagen IV was found at all points of time during the study. Collagen IV was expressed below the keratinocyte cell layers. One week after transplantation (Fig. 5a), parallel collagen IV structures were seen in a broad zone below the
keratinocytes. Overall, vascularization was low. Individual capillaries were seen below the relaxed layer of connective tissue. Three weeks (Fig. 5b) and 4 weeks (Fig. 5c) after transplantation, solidification of the collagen IV structures was observed directly below the keratinocyte cell layers. Increased capillaries were found below the collagen IV fibrous layer as an indication for neovascularization. Laminin was proven below the keratinocyte cell layers (Fig. 6). Vascularization: lectin-specific endothelial marking
Lectin-specific expression of N-acetylglucosamine-oligomeres in the cell membrane structures of endothelial cells of capillaries is shown in Fig. 7a and b. In addition to an increase of collagen IV
482
Schultze-Mosgau et al.
Fig. 5. (a) One week after transplantation, collagen IV was expressed below the keratinocyte cell layers. Parallel collagen IV structures were seen in a broad zone below the keratinocytes. Overall, vascularization was low. Individual capillaries were seen below the relaxed layer of connective tissue (collagen IV; original magnification 25). (b) Three weeks after transplantation, solidification of the collagen IV structures was observed directly below the keratinocyte cell layers. Increased capillaries were found below the collagen IV fibrous layer as an indication for neovascularization (collagen IV; original magnification 25). (c) Four weeks after transplantation, collagen IV structures were found below the keratinocyte cell layers (collagen IV; original magnification 25).
fiber, capillaries were seen from the 2nd week onwards in the direct vicinity of the transplanted PCOKs as an indication for an associated vascularization (Fig. 7a).
The capillaries had small lumina and were mainly localized in groups. After 4 weeks, an increase of the capillary lumina was seen (Fig. 7b). The capillaries were observed below the collagen IV fiber layer (Fig. 7b). Discussion
Fig. 6. Laminin 1 week after transplantation: laminin was proven in the basement membrane like structures from the 1st week onwards after transplantation (laminin; original magnification 25).
It was possible to show that pre-confluent keratinocytes cultivated over 48 h in vitro can be transplanted on a non-resorbent membrane onto a muscle. It was possible to prove that during the in vivo healing period of 4 weeks, a further growth of keratinocyte cell structures and the formation of neoepithelium consisting of two- to three-layer keratinocyte layers occur. The highly-prismatic and cylindrical cell structures and the assay of cytokeratin 5/6 and pancytokeratin proved the specificity of keratinocytes.
Fig. 7. (a) Lectin-specific expression of N-acetyl-glucosamine-oligomeres in the cell membrane structures of endothelial cells of capillaries: In addition to an increase of collagen IV fiber, capillaries were seen from the 2nd week onwards in the direct vicinity of the transplanted PCOKs as an indication for an associated vascularization (lectin; original magnification 25). (b) Lectin-specific expression of N-acetyl-glucosamine-oligomeres in the cell membrane structures of endothelial cells of capillaries: The capillaries had small lumina and were mainly localized in groups. After 4 weeks, an increase of the capillary lumina was seen. The capillaries were observed below the collagen IV fiber layer (lectin; original magnification 25).
With the model of subcutaneous transplantation onto the gracilis muscle it was possible to show that these were in vitro cultivated and transplanted keratinocytes, and that an in vivo construction of oral neoepithelium on muscle tissue is possible in regard to prefabrication of vascular flaps with oral mucosa. Basement membrane like structures, particularly collagen IV and laminin, were proven below the neoepithelium. These extracellular matrix (ECM) structures of the basement membrane are a prerequisite for cell adhesion, migration and differentiation of keratinocytes. The adherence of the keratinocytes with ECM structures of the basement membrane takes place via integrin receptors, e.g., a5b128. The basal keratinocyte layer is highly proliferative, but does not express any markers of terminal differentiation. Keratinocytes that migrated into the upper cell layers and that lost contact with the basal membrane expressed differentiation markers28. An existing environment that permits the formation of a keratinocyte basal membrane contact and cell matrix interactions is an advantage of early transplantation of pre-confluent keratinocytes under in vivo conditions. Here, this model is superior to the in vitro cultivation of keratinocytes on feeder layers of 3T3cells over several weeks11. Since collagen IV fibers were proven in the relaxed connective tissue 1 week after in vivo transplantation, it can be assumed that basement membrane like structures develop on the basis of pre-existing connective tissue. Studies on cell-ECM interactions and sequential deposition of ECM molecules after wounding showed that de novo deposition of fibronectin occurs 24 h after wounding and is followed by deposition of collagen IV and laminin16.
In vitro cultured pre-confluent oral keratinocytes Lectin-specific vascularization was also proven directly below the basement membrane like structures. The increase in the number of capillaries and the individual capillary lumina at the various points in time during the study suggests neoangiogenesis or the formation of capillaries from pre-existing vessels. The observed topographical proximity of basement membrane like structures and capillaries indicates a connection between the existing basal membrane and neoangiogenesis19,20. The significance of laminin in addition to other ECM structures for endothelial cell interactions has already been reported5,12. Laminin is responsible as a disulfide-linked complex for the attachment of keratinocytes to the basement membrane via a6b4 integrins in hemidesmosomes29,39. Furthermore, laminin influences angiogenesis via interaction with the avb3 integrin9. Here, collagen IV and laminin form the basis for the development and arrangement of endothelial cells in addition to proteoglycans. Since endothelial cells migrate and penetrate adjacent tissues, and matrixdegrading enzymes are active, it is assumed that angiogenetic factors are released from the matrix or through appropriate cleavage of ECM molecules such as laminin49. The detachment of the neoepithelium from the PTFE membrane in some areas during the shrinkage shows that the neoepithelium, the basement membrane like structures, the newly formed capillaries and the underlying connective and muscle tissue form a stable structure. It was demonstrated that PCOK prepared with a small piece of autogenous oral mucosa (2 mm 5 mm) can successfully reconstruct uniform and continuous epithelial layers on the muscle flap in vivo. By seeding keratinocytes at a density of 20,000 cells/cm2, the mucosa could be expanded to 2 cm 2 cm. The success rate (91%) of graft acceptance is significantly higher than that of conventional cultured epithelial sheets in previous studies (15–75%)52. It has been reported that serum-free medium with a low calcium concentration may serve to decrease the amount of differentiated cells and increase the number of cells able to adhere and proliferate8,41. If keratinocytes can be transferred to the patient whilst still hyperproliferative, before terminal differentiation is induced, then the wound bed may act as a culture system, allowing the cells to attach and proliferate, rapidly reforming the complete epidermis in vivo. Furthermore, several growth factors that stimulate
wound repair25,53 and certain proteins such as fibrin15 and hyaluronic acid48 that involve the wound healing process may offer favorable conditions for cell proliferation. In addition to these advantages, the short period of in vitro cultivation of 48 h is seen as an advantage. The use of bovine pituitary extract with the residual risk of contamination of the cell culture is a disadvantage. Here the objective is to substitute the bovine content with a non-bovine content in further experiments. So far, in order to deliver pre-confluent keratinocytes to the wound bed, several carriers such as fibrin glue17, polyurethane film33 and collagen14, have been introduced in various models. Regarding the issue of mucosal engineering on the muscle in vivo, WECHSEL46 recently reported that a BERGER et al. muscle can be prelaminated by simple injection onto the muscle with autogenously cultured urothelial mucosal cells via fibrin glue as a carrier. Although WECHSELBERGER et al. showed some successful results, they could not reconstruct uniform and continuous cellular lining along the muscle, which is important for reconstructive surgery. With respect to this result, when pre-confluent mucosal keratinocytes are transplanted, it may be necessary to use a matrix on which transplanted keratinocytes can migrate, forming uniform mucosal coverage along the muscle. With the hydrophilic PTFE membrane used in this study, PCOK were successfully delivered, and continuous neo-epithelium was reconstructed on the muscle flap along the membrane. In addition to the function of a carrier, the objective was to ensure a demarkation of keratinocytes from the covering soft tissue. The spontaneous detachment of the neoepithelium showed that a future removal of the PTFE membrane is possible. The additional use of a collagen membrane or of collagen gel plus SPARC cells would certainly result in an improvement of the cell culture system27. This would facilitate the construction of basement like structures and the early attachment of keratinocytes to the ECM components of the basement membrane. As mentioned in previous reports14,40, cultured epidermal sheets or epithelial cells show some shrinkage after grafting due to natural wound contraction. Likewise, the transplanted membrane also shows some shrinkage. It has been reported that differentiation and maturation of epithelial cells require air exposure30, but it is currently unclear whether
483
these epithelial cells can differentiate and mature after being exposed to the oral environment. From a clinical point of view, we are currently conducting subsequent research to determine whether this prefabricated mucosal flap model can be adapted to the recipient site after transplantation. Furthermore, we are also considering several cytokine treatments to facilitate the epithelialization of grafted keratinocytes in vivo. In summary, this experiment demonstrated that PCOK grafts on the muscle in vivo could achieve uniform multi-layered oral epithelial coverage on the muscle flap in a short period of time. Although the process of establishing a standard protocol for the clinical use of this technique has not yet been completed, it may be a useful tool for future mucosal composite reconstructive surgery in the oral and maxillofacial region that is worth considering for further studies. Acknowledgments. This Research Project was supported by the grants of the Wilhelm-Sander-Foundation No. 2002.017.1 and the Interdisciplinary Center for Clinical Research (IZKF) Project No. B34, financed by the Federal Ministry of Education and Research (BMBF). We would also like to thank Dr. Mai, Department of Anatomy II, Friedrich-Alexander University ErlangenNuremberg, for his kind support with this study. References 1. Amann K. New parameters in kidney biopsy diagnostic—morphometry. Kidney Blood Press Res 2000: 23: 181–182. 2. Amann K, Breitbach M, Ritz E, Mall G. Myocyte/capillary mismatch in the heart of uremic patients. J Am Soc Nephrol 1998: 9: 1018–1022. 3. Alcalde J, Garcia-Tapia R, Espinosa JM, Perez N. Clinical application of free microvascular flaps in reconstructive surgery of the oral cavity. Acta Otorrinolaringol Esp 1994: 45: 457–460. 4. Avery CM, Pereira J, Brown AE. Suprafascial dissection of the radial forearm flap and donor site morbidity. Int J Oral Maxillofac Surg 2001: 30: 37–41. 5. Baldwin HS. Early embryonic development. Cardiovasc Res 1996: 31: 34–45. 6. Baumann I, Greschniok A, Bootz F, Kaiserling E. Free transplanted, microvascular reanastomosed forearm flap for reconstruction of the mouth cavity and oropharynx. Clinical and morphologic findings with special reference to reinnervation. HNO 1996: 44: 616–623. 7. Bortolani A, Barisoni D, Chiamenti C, Lorenzini M, Pasqualini M. Intra-
484
8.
9.
10.
11.
12. 13.
14.
15.
16.
17.
18.
19. 20.
21.
22.
Schultze-Mosgau et al.
oral cancer: immediate reconstruction using a free radial forearm flap. Acta Otorhinolaryngol Ital 1994: 14: 29–40. Boyce ST, Ham RG. Calcium-regulated differentiation of normal human epidermal keratinocytes in chemically defined clonal culture and serum-free serial culture. J Invest Dermatol 1983: 81: 33–40. Brooks PC, Montgomery AM, Rosenfeld M. Integrin alpha v beta 3 antagonists promote tumor regression by inducing apoptosis of angiogenetic blood vessels. Cell 1994: 79: 1157–1164. Graham B, Adkins P, Scheker LR. Complications and morbidity of the donor and recipient sites in 123 lateral arm flaps. J Hand Surg Br 1992: 17: 189–192. Gross J. Getting to mammalian wound repair and amphibian limb regeneration: a mechanistic link in the early events. Wound Repair Regener 1996: 4: 190–202. Hanahan D. Signaling vascular morphogenesis and maintenance. Science 1997: 277: 48–50. Harris PA, Leigh IM, Navsaria HA. Pre-confluent keratinocyte grafting: the future for cultured skin replacements? Burns 1998: 24: 591–593. Horch RE, Debus M, Wagner G, Stark GB. Cultured human keratinocytes on type I collagen membranes to reconstitute the epidermis. Tissue Eng 2000: 6: 53–67. Isogai N, Ueda Y, Kurozumi N, Kamiishi H. Wound healing at the site of microvascular anastomosis: fibrin-stabilizing factor XIII administration and its effects. Microsurgery 1990: 11: 40–46. Kamei M, Kawasaki A, Tano Y. Analysis of extracellular matrix synthesis during wound healing of retinal pigment epithelial cells. Microsc Tes Tech 1998: 42: 311–316. Kaiser HW, Stark GB, Kopp J, Balcerkiewicz A, Spilker G, Kreysel HW. Cultured autologous keratinocytes in fibrin glue suspension, exclusively and combined with STS-allograft (preliminary clinical and histological report of a new technique). Burns 1994: 20: 23–29. Kimata Y, Uchiyama K, Ebihara S, Saikawa M, Hayashi R, Haneda T. Postoperative complications and functional results after total glossectomy with microvascular reconstruction. Plast Reconstr Surg 2000: 106: 1028–1035. Kumar CC. Signaling by integrin receptors. Oncogene 1998: 17: 1355–1363. Kumar CC, Yoneda J, Bucana CD, Fidler IJ. Regulation of distinct steps of angiogenesis by different angiogenic molecules. Int J Oncol 1998: 12: 749– 757. Lauer G. Autografting of feeder-cell free cultured gingival epithelium. Method and clinical application. J Craniomaxillofac Surg 1994: 22: 18–22. Lauer G, Schimming R. Tissue-engineered mucosa graft for reconstruction of the intraoral lining after freeing of
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
the tongue: a clinical and immunohistologic study. J Oral Maxillofac Surg 2001: 59: 169–175. Lauer G, Schimming R, Frankenschmidt A. Intraoral wound closure with tissue-engineered mucosa: new perspectives for urethra reconstruction with buccal mucosa grafts. Plast Reconstr Surg 2001: 107: 25–33. Lind MG, Arnander C, Gylbert L, Heden P, Jurell G. Reconstruction in the head and neck regions with free radial forearm flaps and split-rib bone grafts. Am J Surg 1987: 154: 459–462. Martin P, Hopkinson-Woolley J, McCluskey J. Growth factors and cutaneous wound repair. Prog Growth Factor Res 1992: 4: 25–44. McKay I, Woodward B, Wood K, Navsaria HA, Hoekstra H, Green C. Reconstruction of human skin from glycerol-preserved allodermis and cultured keratinocyte sheets. Burns 1994: 20(Suppl 1): 19–22. Mishima H, Hibino T, Hara H, Murakami J, Otori T. SPARC from corneal epithelial cells modulates collagen contraction by keratocytes. Invest Opthalmol Vis Sci 1998: 39: 2547–2553. Nicholson LJ, Watt FM. Decreased expression of fibronectin and the a5b1 integrin during terminal differentiation of human keratinocytes. J Cell Sci 1991: 98: 225–232. Niessen CM, Hogervorst F, Jaspars LH. The alpha 6 beta 4 integrin is a receptor for both laminin and kalinin. Exp Cell Res 1994: 211: 360–367. Ootani A, Toda S, Fujimoto K, Sugihara H. An air-liquid interface promotes the differentiation of gastric surface mucous cells (GSM06) in culture. Biochem Biophys Res Commun 2000: 271: 741–746. Pickford MA, Soutar DS. Intraoral reconstruction using a second free flap for recurrent or metachronous carcinoma. Br J Plast Surg 1995: 48: 559–563. Regauer S, Compton CC. Cultured keratinocyte sheets enhance spontaneous reepithelialization in a dermal explant model of partial-thickness wound healing. J Invest Dermatol 1990: 95: 341–346. Rennekampff HO, Hansbrough JF, Kiessig V, Abiezzi S, Woods Jr V. Wound closure with human keratinocytes cultured on a polyurethane dressing overlaid on a cultured human dermal replacement. Surgery 1996: 120: 16–22. Rheinwald JG, Green H. Serial cultivation of strains of human epidermal keratinocytes: the formation of keratinizing colonies from single cells. Cell 1975: 6: 331–343. Schultze-Mosgau S, Keilholz L, Rodel F, Labahn D, Neukam FW. Experimental model for transplantation of a modified free myocutaneous gracilis flap to an irradiated neck region in rats. Int J Oral Maxillofac Surg 2001: 30: 63–69.
36. Schultze-Mosgau S, Grabenbauer GG, Radespiel-Tro¨ger M, Wiltfang J, Ries J, Neukam FW, Ro¨del F. Vascularisation in the transition area between free grafted soft tissues and pre-irradiated graft bed tissues following preoperative radiotherapy in the head and neck region. Head Neck 2002: 24: 42–51. 37. Schultze-Mosgau S, Ro¨del F, Wehrhan F, Grabenbauer GG, Amann K, Radespiel-Tro¨ger M, Neukam FW. Transforming growth factor b1 and b2 (TGFb1/TGFb2) profile changes in previously irradiated free flap beds. Head Neck 2002: 24: 33–41. 38. Schultze-Mosgau S, Wehrhan F, Amann K, Radespiel-Tro¨ger M, Ro¨del F, Grabenbauer GG. In vivo TGFb3 expression in wound healing in irradiated tissues—an experimental study. Strahlenther Onkol 2003: 179: 410–416. 39. Sonnenberg A, Calafat J, Janssen H. Integrin alpha 6/beta 4 complex is located in hemidesmosomes, suggesting a major role in epidermal cell-basement membrane adhesion. J Cell Biol 1991: 113: 907–917. 40. Sugimura Y, Hata K, Torii S, Ueda M. Transplantation of cultured mucosal epithelium: an experimental study. J Craniomaxillofac Surg 1996: 24: 352–359. 41. Svensjo T, Yao F, Pomahac B, Eriksson E. Autologous keratinocyte suspensions accelerate epidermal wound healing in pigs. J Surg Res 2001: 99: 211– 221. 42. Toft K, Keller GS, Blackwell KE. Ectopic hair growth after flap reconstruction of the head and neck. Arch Fac Plast Surg 2000: 2: 148–150. 43. Tomson AM, Scholma J, Blaauw EH. The effect of detachment of cultured epithelial sheets on cellular ultrastructure and organisation. Epithelial Cell Biol 1995: 4: 43–51. 44. Ueda M. Formation of epithelial sheets by serially cultivated human mucosal cells and their applications as a graft material. Nagoya J Med Sci 1995: 58: 13–28. 45. vander LB, Spronk CA, de Visscher JG. Closure of radial forearm free flap donor site with local full-thickness skin graft. Br J Oral Maxillofac Surg 1999: 37: 119–122. 46. Wechselberger G, Bauer T, Meirer R, Piza-Katzer H, Lille S, Russell RC. Muscle prelamination with urothelial cell cultures via fibrin glue in rats. Tissue Eng 2001: 7: 153–159. 47. Weibel ER. Measuring through the microscope: development and evolution of stereological methods. J Microsc 1989: 155: 393–403. 48. Weigel PH, Fuller GM, LeBoeuf RD. A model for the role of hyaluronic acid and fibrin in the early events during the inflammatory response and wound healing. J Theor Biol 1986: 119: 219–234.
In vitro cultured pre-confluent oral keratinocytes 49. Werb Z, Vu TH, Rinkenberger JL, Coussens LM. Matrix-degrading proteases and angiogenesis during development and tumor formation. Apmis 1999: 107: 11–18. 50. Wynne JM, Apostolis CG, Perl TZ, McQuaide JR, Graham T, Levin W. Function-preserving procedures in primary reconstruction following surgery for advanced oral cancer—an appraisal. S Afr J Surg 1978: 16: 191–197. 51. Xu W, Li H, Brodniewicz T, Auger FA, Germain L. Cultured epidermal
sheet grafting with Hemaseel HMN fibrin sealant on nude mice. Burns 1996: 22: 191–196. 52. Yanaga H, Tai Y, Yamauchi T, Mori S, Udoh Y, Yamamoto M. Take percentage and conditions of cultured epithelium were improved when basement membrane of the graft was maintained and anchoring mesh was added. Plast Reconstr Surg 2001: 107: 105– 115. 53. Yates RA, Nanney LB, Gates RE, King Jr LE. Epidermal growth factor
485
and related growth factors. Int J Dermatol 1991: 30: 687–694. Address: Stefan Schultze-Mosgau Glueckstrasse 11 91054 Erlangen Germany. Tel: þ49 9131 853 3601 Fax: þ49 9131 853 4219 E-mail: [email protected]. uni-erlangen.de