sign in sign up
<<
Home , Dermal fibroblast, Keratinocyte

Journal of Science 107, 2285-2289 (1994) 2285 Printed in Great Britain © The Company of Biologists Limited 1994

Dermal fibroblasts activate keratinocyte outgrowth on gels

Tai-Lan Tuan*, Lia C. Keller, David Sun, Marcel E. Nimni and David Cheung Department of Surgery, Childrens Hospital Los Angeles and University of Southern California School of Medicine, 4650 Sunset Blvd, Los Angeles, California 90027, USA *Author for correspondence

SUMMARY

The effects of dermal fibroblasts on keratinocyte outgrowth was further studied with a co-culture system using Millicell on collagen substrata was studied using an in vitro ker- inserts. It was found that the co-culture of fibroblasts with atinocyte-collagen gel composite model. fibroblasts the composite enhanced keratinocyte outgrowth on were seeded inside collagen gels, which remained attached collagen gels that had previously contained viable fibrob- to the plastic substratum. incorpo- lasts for 5 days. Among all, however, the keratinocyte rated in collagen gels were either kept viable throughout outgrowth was far better on gels containing viable fibrob- the study, or were lysed hypotonically with water at lasts. Addition of keratinocyte growth factor or its neu- different time intervals (2 hours and 5 days). Results show tralizing did not affect keratinocyte outgrowth. that very little keratinocyte outgrowth occurred on either These results suggest that dermal fibroblasts can activate plain collagen gels or gels that had previously contained keratinocyte outgrowth on collagen matrices through some viable fibroblasts for 2 hours. A 3- to 4-fold increase in ker- diffusible factors other than keratinocyte growth factor, atinocyte outgrowth occurred on collagen gels that had pre- and epithelial-mesenchymal interactions exert some special viously contained viable fibroblasts for 5 days. A striking effects on keratinocyte outgrowth on collagen gels. increase (20-fold) in keratinocyte outgrowth was observed on collagen gels that contain viable fibroblasts. The effect of fibroblast diffusible factors on keratinocyte outgrowth Key words: dermal fibroblast, keratinocyte, collagen

INTRODUCTION wound repair. In addition, they are capable of secreting factors that regulate the growth and differentiation of other cells Keratinocyte migration is essential for skin coverage and ker- (Ehrlich, 1988; Marks et al., 1991; Smola et al., 1993). atinocyte growth/differentiation during wound repair (reviews: Therefore, we have initiated this study to investigate the effect Peacock, 1984; Barrandon and Green, 1987; Stenn and of dermal fibroblasts on keratinocyte outgrowth using an in Dealma, 1988). The process is influenced by factors present in vitro skin composite model modified from Bell et al. (1981) the local wound environment. For example, extracellular and Guidry and Grinnell (1985). The dermal equivalent was matrix (ECM) components such as laminin, collagen type I and made by seeding human dermal fibroblasts inside reconstituted type IV, fibronectin, vitronectin and thrombospondin all show type I collagen gels, which were attached to the plastic tissue differential effects on keratinocyte migration (Takashima and culture surfaces (attached gels). Unlike floating collagen gels, Grinnell, 1985; review: Donaldson and Mahan, 1987; fibroblasts in attached gels resembled those in tissues in Nickoloff et al., 1988; Woodley et al., 1988; Guo et al., 1990; terms of cell growth and growth factor responsiveness Brown et al., 1991). Meanwhile, soluble growth factors also (Nakagawa et al., 1989). Keratinocyte outgrowth was studied regulate keratinocyte migration through upregulation of by placing dispase-treated fresh keratinocyte sheets on top of cellular receptors for ECM components or through secretion of collagen gels. Results of the study are reported herein. proteolytic and their inhibitors in modulating ECM degradation (Grinnell et al., 1981; Clark et al., 1982; Barrandon and Green, 1987; Hebda, 1988; Nickoloff et al., MATERIALS AND METHODS 1988; Kaur and Carter, 1992). In this regard, that directly underlies has been shown to play an Cells essential role in keratinocyte migration (Guo et al., 1990) and fibroblasts were established from newborn foreskins in growth/differentiation (Bell et al., 1981, 1983; Mackenzie obtained from circumcisions. Cells were cultured in 75 cm2 tissue and Fusenig, 1983; Sengel, 1983; Leipziger et al., 1985; Saiag, culture flasks (Falcon Labware, Oxnard, CA) using DMEM (Gibco, et al., 1985; Boyce et al., 1988; Yannas et al., 1989; Boukamp Grand Island, NY) plus 10% fetal calf serum (FCS, Gibco), 50 µg/ml et al., 1990). Fibroblasts are the major cell type in the dermis gentamicin, and 20 µg/ml Fungizone (Gibco). Cultures were for synthesis and reorganization of ECM components during incubated in a 37¡C humidified incubator with 5% CO2. Cells were 2286 T.-L. Tuan and others harvested using a 0.25% trypsin/1 mM EDTA solution (Gibco). Early atinocytes in the outgrowth area was estimated. Briefly, keratinocyte passages of cells (3 to 10) were used in the experiments. layers were lifted from the underlying collagen gels after dispase Human skin keratinocyte sheets were prepared from freshly treatment, and this was followed by trypsinization (Takashima and obtained surgical skin samples. Samples were treated with dispase II Grinnell, 1985). Cell counts were performed using a hemocytometer. (Sigma) for 90 minutes, and then rinsed in DMEM containing 10% The number of cells in the outgrowth area was calculated by sub- FCS. Keratinocyte layers were separated from the dermis using a pair tracting the number of cells in the original explant from the total cell of fine forceps and cut into 4 mm2 square pieces. Samples were stored number. at 4¡C in DMEM until use. Recombinant human keratinocyte growth factor (KGF) and 2B8 neutralizing murine monoclonal antibody against KGF were kindly Preparation of collagen gels provided by Dr Jeffrey Rubin of the National Cancer Institute, Pepsin-solubilized bovine tendon type I collagen was prepared Bethesda, Maryland, USA. Their effects on keratinocyte outgrowth according to the method of Cheung et al. (1990). Collagen substrata were tested by adding KGF (25 ng/ml) or anti-KGF antibody (5 were prepared according to the methods described by Guidry and µg/ml) to the composite or co-culture samples every other day. Grinnell (1987) and Nakagawa et al. (1989). Briefly, collagen was adjusted to physiological ionic strength and pH with 10× PBS (1.5 M Scanning electron microscopy NaCl, 0.1 M sodium phosphate, pH 7.2) and 0.1 M NaOH while being Briefly, samples for scanning electron microscopy were fixed for 2 maintained at 4¡C. The final collagen concentration was 1.5 mg/ml. hours at 22¡C with 2% glutaraldehyde in 0.1 M sodium cacodylate Fibroblasts were incorporated into the reconstituted collagen to a final (pH 7.4) and post-fixed for 30 minutes at 22¡C with 1% aqueous 6 concentration of 0.5×10 cells/ml. Samples (200 µl or 400 µl) of the OsO4, followed by dehydration in an ethanol series. Critical point reconstituted collagen or collagen-fibroblast suspension were placed drying was performed with liquid CO2. Finally, specimens were in 24-well or 6-well culture plates (Costar). Each sample filled an area coated with 20 nm of platinum using a sputter coater, examined and outlined by a 12 mm (24-well plate) or 20 mm (6 well) diameter photographed with a Zeiss Nova Scan 30 scanning electron micro- circular score within the well. The plates were placed in a 37¡C scope. incubator for 45 minutes to allow collagen to form fibrils. Samples were then covered by DMEM containing 10% FCS and placed in culture. At different time intervals, fibroblasts in some wells were RESULTS lysed hypotonically using sterile distilled water. These samples were rinsed three times over a 24 hour period to remove fibroblast debris Keratinocyte outgrowth and finally covered with culture medium. The area of keratinocyte outgrowth on various collagen gels Measurement of collagen gel thickness was measured at day 14 and the result is shown in Fig. 1. There The method of measuring gel thickness using digital dial micrometer was very little keratinocyte outgrowth on plain collagen gels is similar to that described previously (Tuan and Grinnell, 1989). (Fig. 1A) or gels that had previously contained viable fibrob- Briefly, gel thickness was measured on an Olympus CK2 inverted lasts for 2 hours (Fig. 1B). An increase in keratinocyte microscope equipped with a Mitutoyo digital test indicator (0.01-10 outgrowth was observed on collagen gels that had previously mm). The plane of the focus was adjusted from the bottom to the top contained viable fibroblasts for 5 days (Fig. 1C). A striking of the gel, and the span of focal adjustment was recorded by the increase (20-fold) in keratinocyte outgrowth occurred on indicator. collagen gels containing viable fibroblasts (Fig. 1D). Pho- Preparation of keratinocyte-collagen gel composites tographs of two typical samples are also shown in Fig. 1. A piece of freshly prepared keratinocyte layer (as described above) Extended keratinocyte outgrowth was observed on collagen about 4 mm2 was placed on the top of each collagen matrix and immo- gels containing viable fibroblasts (D) and, in this sample, ker- bilized with 3 µl of neutralized collagen. Samples were cultured in atinocytes have covered the entire surface of the gel and keratinocyte culture medium (10% FBS, 20 mM HEPES buffer, 100 migrated further onto the plastic culture dish surface. units/ml penicillin, 100 µg/ml streptomycin, 0.25 µg/ml Fungizone, 10 ng/ml , 10−9 M , and 0.4 Morphology of collagen gels µg/ml hydrocortisone) (Toda et al., 1987), which was changed every The morphology of collagen gel matrices was studied under 3 days. scanning electron microscopy. Plain collagen gels appeared as Co-cultures using Millicell inserts a network of fibrils that were interconnected with each other, although some intertwined in pairs giving a thicker appearance Co-cultures of fibroblasts with keratinocyte-collagen gel composites were set up using Millipore Millicell inserts (Millipore). Ker- (data not shown). Similar fibrilar morphology was observed in atinocyte-collagen composites were made as described above in 24- collagen gels that contained viable fibroblasts for 2 hours, well tissue culture plates. cultures (in collagen gels) were except that the space between collagen fibrils was much made in Millicell inserts similarly to those of the collagen gels used smaller, and thus gave a denser appearance. Collagen gel con- for the composites. These inserts were placed in tissue culture wells traction occurred when fibroblasts were incorporated into the that contained keratinocyte-collagen gel composites. Controls gel (Bell et al., 1981; Guidry and Grinnell, 1985). In this study, included inserts with plain collagen gels. Co-cultures were covered collagen gels were attached to the culture plastic surface; with keratinocyte culture medium and incubated at 37¡C in a humid- therefore, collagen matrices responded to fibroblast contraction ified incubator supplemented with 5% CO2. force by reducing gel thickness without changing gel diameter. Keratinocyte outgrowth Typically, there was about a 55 to 60% reduction in gel After 14 days in culture, samples were fixed in formalin and stained thickness after 2 hours. Lysis of fibroblasts with water did not with Nile Blue Sulfate (Sigma). Subsequently, samples were pho- cause gels to swell; instead, gel thickness remained constant tographed and the area of keratinocyte outgrowth was calculated using throughout the entire study. In addition to the reduction Sigma Scan (Jandel Scientific). Each experiment involved 4 sample between fibril space that caused by the fibroblast traction force, replicas and was repeated 3 times. In some groups, the number of ker- it was noticed that there were areas in which collagen fibrils Mesenchyme and keratinocyte outgrowth 2287

Fig. 1. Areas of keratinocyte outgrowth on various collagen gel matrices. (A) Plain collagen gels. (B) Collagen gels that had previously contained viable fibroblasts for 2 hours. (C) Collagen gels that had previously contained viable fibroblasts for 5 days. (D) Collagen gels containing viable fibroblasts. Bar, 1 cm.

Fig. 2. Scanning electron micrographs of collagen gels that had appeared distorted and more concentrated, which may be areas previously contained viable fibroblasts for 5 days (A) or collagen previously occupied by fibroblasts. gels with viable fibroblasts (B and C). (A) Fibroblasts in the collagen The morphology of a collagen gel that had previously gel were lysed hypotonically with water on day 5. Arrowheads contained viable fibroblasts for 5 days is shown in Fig. 2A. In indicate some amorphous material attached to collagen fibrils. contrast to the 2 hour sample, collagen fibrils of these samples (B) Viable fibroblasts (X) located in collagen fibrilar network were decorated with amorphous material (arrowheads), which (arrowheads). (C) An enlargement of the outlined area from B showing close connections between fibroblast processes and collagen may be the secretory products or residual cellular structures fibrils. Bars: A,C, 1 µm; B, 10 µm. from fibroblasts after cell lysis. Additional washing of these gels with water did not seem to remove the material or affect the gel thickness. The thickness of these gels was about 5% of material that possibly originated from fibroblasts (Fig. 2A), the the original gel thickness. increase in keratinocyte outgrowth on these gels, compared to An intricate meshwork of fibroblasts with their surrounding that on 2 hour contracted gels, may be attributed to fibroblast collagen fibrils was observed in gels containing viable fibrob- secretory products instead of collagen fibril density. The lasts (Fig. 2B). Fibroblasts (X) were spread and their cellular dramatic increase in keratinocyte outgrowth on collagen gels processes were in close contact with collagen fibrils (arrow- containing viable fibroblasts may reflect the effects of fibrob- heads). Fig. 2C shows a higher magnification of the framed last secretory products and/or fibroblast-keratinocyte interac- area from Fig. 2B. Gel contraction reached the maximum on tions on keratinocyte outgrowth and was further investigated the 5th day. in a co-culture study. Collagen fibril density did not seem to affect keratinocyte outgrowth in this study, since keratinocyte outgrowth occurred Co-culture of fibroblasts with composites at similar rates in both plain collagen gels and 2 hour con- The system consisted of keratinocyte-collagen gel composites tracted gels (about 60% reduction in gel thickness). Because made in tissue culture wells and additional fibroblast cultures the collagen fibrils of 5-day samples were attached with prepared in Millipore Millicell inserts. To be consistent in 2288 T.-L. Tuan and others

attached collagen gels and a co-culture system for the following reasons: (1) fibroblasts cultured in attached, con- tracting collagen gels bear resemblances in growth and growth factor responsiveness to those of scar tissue (Nakagawa et al., 1989). (2) The size of collagen gel surface area was consistent for keratinocyte outgrowth. The variability of collagen fibril density in a given area was also controlled in this way for the co-culture study. (3) The co-culture system is closer to the situation in vivo, in which epidermal and mesenchymal tissues in close proximity exhibit mutual effects during development and tissue regeneration (Sharpe and Ferguson, 1988; Smola et Fig. 3. Areas of keratinocyte outgrowth in co-culture system. Millicell inserts containing fibroblasts (in collagen gels) were added al., 1993). to wells containing either type I (keratinocyte implant on collagen In this study, collagen fibril density did not seem to be an gel that had previously contained viable fibroblasts for 5 days) or influential factor in keratinocyte outgrowth because there was type II (keratinocyte implant on collagen gel containing viable little difference between 2 hour contracted gels (60% gel con- fibroblasts) composites. (A) Co-cultures of control Millicell inserts traction) and non-contracted plain collagen gels. On the surface (plain collagen gels only) and type I composites. (B) Co-cultures of of the 2 hour contracted gels, there were often areas with Millicell inserts containing fibroblasts and type I composites. (C) Co- increased fibril density that might be local foci where collagen cultures of Millicell inserts containing fibroblasts and type II substrata were distorted by fibroblasts due to the traction force, composites. as described by Stopak and Harris (1982). Fibroblasts cultured in attached collagen gels showed high collagen fibril density, 5-day contracted collagen gels were metabolic activity (Nakagawa et al., 1989) and could secrete used (these gels had reached the maximum gel contraction). ECM and other products into the medium or incorporate them Type I composites consisted of collagen gels with lysed fibrob- into matrices in culture (for review: Postlethwaite and Kang, lasts and type II composites consisted of gels with viable 1988). There was a similar increase (1.5 times) in the number fibroblasts. Type I composites with control inserts showed a of fibroblasts cultured in collagen gels from inserts or moderate amount of keratinocyte outgrowth (Fig. 3A). A 3- composite samples after 7 days (data not shown). The fold increase in keratinocyte outgrowth occurred when the co- amorphous material, deposited on the collagen fibrils of 5-day culture inserts contained fibroblasts (Fig. 3B). An additional 2- contracted gels originated from fibroblasts and contributed to fold increase in keratinocyte outgrowth was found on type II the increased keratinocyte outgrowth (Fig. 1C). On the other composites with either control inserts (data not shown) or hand, fibroblast diffusible factors enhanced further ker- inserts containing fibroblasts (Fig. 3C). The number of ker- atinocyte outgrowth on these collagen gels as demonstrated atinocytes in the outgrowth area was also determined. Results in the co-culture study (Fig. 3B). Fibroblasts are known showed that although there was only a 2-fold increase in areas sources of proteolytic enzymes (Hashimoto et al., 1988) and of outgrowth, there was a 6-fold increase in the number of ker- paracrine or autocrine growth regulating factors such as KGF atinocytes in type II co-culture (453±110) compared to type I (Rubin et al., 1989; review: Roberts and Sporn, 1990). co-culture (77±33). Recombinant human KGF or its neutralizing antibody tested Human recombinant KGF and its neutralizing antibody were in this study, however, did not exert any effect on keratinocyte used to determine whether KGF was responsible for ker- outgrowth. atinocyte outgrowth on collagen gels. Addition of KGF did not The most extensive keratinocyte outgrowth was observed on enhance keratinocyte outgrowth on any composite samples, collagen gels containing viable fibroblasts (with or without nor did KGF antibody show any effect on keratinocyte additional fibroblast co-cultures), indicating that cell-cell inter- outgrowth (data not shown). actions may induce the release of additional factors. In the con- ventional (feeder layer) co-cultures, Smola et al. (1993) showed that keratinocytes stimulated mRNA levels of KGF, DISCUSSION -6 (Il-6), and granulocyte-macrophage colony stim- ulating factor (GM-CSF) in fibroblasts, meanwhile the effect Using the in vitro skin composite model and a co-culture of the feeder fibroblasts on keratinocyte growth could not be system, we have demonstrated in this study that human dermal achieved by using conditioned medium or by addition of fibroblasts were able to activate and enhance keratinocyte isolated growth factors. Another example of cellular prolifer- outgrowth on collagen gel matrices. The effect was mediated ation/differentiation that is affected by cell-cell interactions is through fibroblast-secreted diffusible factors or factors illustrated in endothelial cell differentiation. It was shown that deposited in the collagen matrix. Keratinocyte growth factor or interaction between endothelial cells and pericytes or smooth its neutralizing antibody did not affect keratinocyte outgrowth muscle cells is essential for the activation of transforming in the present study. The outgrowth was a result of both ker- growth factor beta, which in turn induces endothelial cell atinocyte proliferation and migration. differentiation (Orlidge and D’Amore, 1987; Sato and Rifkin, Previously, Coulomb et al. (1989) studied the effect of 1989; Sato et al., 1990). Therefore, -mesenchyme dermal fibroblasts on keratinocyte outgrowth using a floating interactions may induce paracrine factors or even such collagen gel model. Their study demonstrated that conditioned as interstitial collagenase (Smola et al., 1993) to release matrix- medium from fibroblast monolayer cultures enhanced ker- bound factors, and they in turn facilitate keratinocyte growth atinocyte outgrowth. In the present study we chose to use and/or migration (review: Schnaper and Kleinman, 1993).

Mesenchyme and keratinocyte outgrowth 2289

We are grateful to Drs F. Grinnell and A. Wysocki for their helpful basic fibroblast growth factors on facilitation of dermal by comments and suggestions. These studies were supported by NIH type I collagen matrices. J. Biomed. Mater. Res. 25, 683-696. grant AR 40409 to Dr Tuan. Nakagawa, S., Pawalek, P. and Grinnell, F. (1989). organization modulates fibroblast growth and growth factor reponsiveness. Exp. Cell Res. 182, 572-582. Nickoloff, B. J., Mitra, R. S., Riser, B. L., Dixit, V. M. and Varani, J. (1988). REFERENCES Modulation of keratinocyte motility. Correlation with production of extracellular matrix molecules in response to growth promoting and Barrandon, Y. and Green, H. (1987). Cell migration is essential for sustained antiproliferative factors. Am. J. Pathol. 132, 543-551. growth of keratinocyte colonies: the roles of transforming growth factor- Orlidge, A. and D’Amore, P. A. (1987). Inhibition of capillary endothelial cell alpha and epidermal growth factor. Cell 50, 1131-1137. growth by pericytes and smooth muscle cells. J. Cell Biol. 105, 1455-1462. Bell, E., Ehrlich, H. P., Buttle, D. and Nakatsuji, T. (1981). Living tissue Peacock, E. E. (1984). Wound Repair, 3rd edn. W.B. Saunders Co., formed in vitro and accepted as a skin equivalent tissue of full thickness. Philadelphia. Science 211, 1052-1054. Postlethwaite, A. E. and Kang, A. H. (1988). Fibroblasts. In Inflammation: Bell, E., Sher, S., Hull, B., Merrill, C. Rosen, S., Chamson, A., Asselineau, Basic Principals and Clinical Correlates (ed. J. I. Gallin, I. M. Goldstein and D., Dubertret, L., Coulomb, B., Lapiere, C., Nusgens, B. and Neveux, Y. R. Snyderman), pp. 577-597. Raven Press, New York. (1983). The reconstitution of living skin. J. Invest. Dermatol. 81 (Suppl.), 2s- Roberts, A. B. and Sporn, M. B. (1990). The transforming growth factor 10s. betas. In Peptide Growth Factors and Their Receptors. Handbook of Boukamp, P., Breitkreutz, D., Stark, H. J. and Fusenig, N. E. (1990). Experimental Pharmacology (ed. A. B. Robert and M. B. Sporn), pp. 419- Mesenchyme-mediated and endogenous regulation of growth and 472. Springer-Verlag, Heidelberg. differentiation of human skin keratinocyte derived from different body sites. Rubin, J. S., Osada, H., Finch, P. W., Taylor, W. G., Rudikoff, S. and Differentiation 44, 150-161. Aaronson, S. A. (1989). Purification and characterization of a newly Boyce, S. T., Christianson, D. J. and Hansbrough, J. F. (1988). Structure of identified growth factor specific for epithelial cells. Proc. Nat. Acad. Sci. a collagen-GAG dermal skin substitute optimized for cultured human USA 86, 802-806. epidermal keratinocytes. J. Biomed. Mater. Res. 22, 939-957. Saiag, P. H., Coulomb, B., Bell, E., Lebreton, C. and Dubertret, L. (1985). Brown, C., Stenn, K.S., Falk, R. J., Woodley, D. T. and O’Keefe, E. J. Psoriatic fibroblasts induce hyperproliferation of normal keratinocytes in a (1991). Vitronectin: Effects on keratinocyte motility and inhibition of skin equivalent model in vitro. Science 230, 669-672. collagen-induced motility. J. Invest. Dermatol. 96, 724-728. Sato, Y. and Rifkin, D. B. (1989). Inhibition of endothelial cell movement by Cheung, D. T., Perelman, N., DiCesare, P. E. and Nimni, M. E. (1990). A pericytes and smooth muscle cells: Activation of a latent transforming highly specific and quantitative method for determining type III/I collagen growth factor-B1-like molecule by plasmin during co-culture. J. Cell Biol. ratios in tissues. Matrix 10, 164-171. 109, 309-315. Clark, R. A. F., Lanigan, J. M., DellaPelle, P., Dvorak, H. F. and Colvin, R. Sato, Y., Tsuboi, R., Lyons, R., Moses, H. and Rifkin, D. B. (1990). B. (1982). Fibronectin and fibrin provide a provisional matrix for epidermal Characterization of the activation of latent TGF-beta by co-cultures of cell migration during wound reepithelialization. J. Invest. Dermatol. 79, 264- endothelial cells and pericytes or smooth muscular cells: A self-regulating 269. system. J. Cell Biol. 111, 757-763. Coulomb, B., Lebreton, C. and Dubertret, L. (1989). Influence of human Schnaper, H. W. and Kleinman, H. K. (1993). Regulation of cell function by dermal fibroblasts on epidermalization. J. Invest. Dermatol. 92, 122-125. extracellular matrix. Pediat. Nephrol. 7, 96-104. Donaldson, D. J. and Mahan, J. T. (1987). Keratinocyte migration and the Sengel, P. (1983). Epidermal-dermal interactions during formation of skin and extracellular matrix (review). J. Invest. Dermatol. 90, 623-628. cutaneous appendages. In Biochemistry and Physiology of the Skin (ed. L. Ehrlich, H. P. (1988). The role of matrix in wound healing. Goldsmith), pp. 102-131. Oxford University Press, Oxford. Prog. Clin. Biol. Res. 226, 243-258. Sharpe, P. M. and Ferguson, M. W. J. (1988). Messenchymal influences on Grinnell, F., Billingham, R. E. and Burgess, L. (1981). Distribution of epithelial differentiation in developing systems. J. Cell Sci. 10, 195-250. fibronectin during wound healing in vivo. J. Invest. Dermatol. 76, 181-189. Smola, H., Thiekotter, G. and Fusenig, N. E. (1993). Mutual induction of Guidry, C. and Grinnell, F. (1985). Studies on the mechanism of hydrated growth factor gene expression by epidermal-dermal cell interaction. J. Cell collagen reorganization by human skin fibroblasts. J. Cell Sci. 79, 67-81. Biol. 122, 417-429. Guidry, C. and Grinnell, F. (1987). Heparin modulates the supramolecular Stenn, K. S. and Dealma, L. (1988). Re-epithelialization. In The Molecular organization of hydrated collagen gels and inhibits gel contraction by and Cellular Biology of Wound Repair (ed. R. A. F. Clark and P. M. Henson), fibroblasts. J. Cell Biol. 104, 1097-1103. pp. 321-335. Plenum Press, New York. Guo, M., Toda, K. I. and Grinnell, F. (1990). Activation of human Stopak, D. and Harris, A. K. (1982). Connective tissue morphogenesis by keratinocyte migration on type I collagen and fibronectin. J. Cell Sci. 96, fibroblasts traction. Dev. Biol. 90, 383-398. 197-205. Takashima, A. and Grinnell, F. (1985). Fibronectin-mediated keratinocyte Hashimoto, K., Prystowsky, J. H., Baird, J., Lazarus, G. and Jensen, P. migration and inhibition of fibronectin receptor function in vitro. J. Invest. (1988). Keratinocyte urokinase-type plasminogen activator is secreted as a Dermatol. 85, 304-308. single chain precursor. J. Invest. Dermatol. 90, 823-828. Toda, K. I., Tuan, T.-L., Brown, P. J. and Grinnell, F. (1987). Fibronectin Hebda, P. A. (1988). Stimulatory effects of transforming growth factor-beta receptors of human keratinocytes and their expression during cell culture. J. and epidermal growth factor on epidermal cell outgrowth from porcine skin Cell Biol. 105, 3097-3104. explain cultures. J. Invest. Dermatol. 91, 440-445. Tuan, T.-L. and Grinnell, F. (1989). Fibronectin and fibrinolysis are not Kaur, P. and Carter, W. G. (1992). Integrin expression and differentiation in required for fibrin gel contraction by human skin fibroblasts. J. Cell. Physiol. transformed human epidermal cells is regulated by fibroblasts. J. Cell Sci. 140, 577-583. 103, 755-763. Woodley, D. T., Bachmann, P. M. and O’Keefe, E. J. (1988). Laminin Leipziger, L. S., Glushko, V., DiBernardo, B., Shafaie, F., Noble, J., inhibits human keratinocyte migration. J. Cell. Physiol. 136, 140-146. Nichols, J. and Alvarez, O. M. (1985). Dermal wound repair: role of Yannas, I. V., Lee, E., Orgill, D. P., Skrabut, E. M. and Murphy, G. F. collagen matrix implants and synthetic polymer dressings. J. Amer. (1989). Synthesis and characterization of a model extracellular matrix that Dermatol. 12, 409-419. induces partial regeneration of adult mammalian skin. Proc. Nat. Acad. Sci. Mackenzie, I. C. and Fusenig, N. E. (1983). Regeneration of organized USA 86, 933-937. epithelial structure. J. Invest. Dermatol. 81, 189s-194s. Marks, M. G., Doillon, C. and Silver, F. H. (1991). Effects of fibroblasts and (Received 5 July 1993 - Accepted 4 May 1994)