FIBROBLAST GROWTH FACTOR-7 REGULATES STRATIFICATION OF THE BLADDER UROTHELIUM

0022-5347/01/1666-2536/0 THE JOURNAL OF UROLOGY® Copyright © 2001 by AMERICAN UROLOGICAL ASSOCIATION, INC.®

Vol. 166, 2536 –2541, December 2001 Printed in U.S.A.

FIBROBLAST GROWTH FACTOR-7 REGULATES STRATIFICATION OF THE BLADDER UROTHELIUM JENNIFER A. TASH, SCOTT G. DAVID, E. DARRACOTT VAUGHAN, JR. AND DORIS A. HERZLINGER From the Department of Urology, James Buchanan Brady Urology Foundation, and Department of Physiology and Biophysics, New York Presbyterian Hospital-Weill Medical College of Cornell University, New York, New York

ABSTRACT

Purpose: The cellular and molecular mechanisms that regulate the organization of bladder urothelium into basal, intermediate and superficial cell layers remain poorly understood. We tested the hypothesis that fibroblast growth factor (FGF)-7 is essential for generating a multilayered stratified bladder epithelium. Materials and Methods: The morphological and molecular characteristics of bladder urothelium in age and sex matched FGF-7 ⫹/⫹ wild-type and ⫺/⫺ null mice were evaluated. In addition, the effect of exogenous FGF-7 on the growth and differentiation of primary murine urothelial cells was assessed. Results: Morphometric analyses demonstrate that FGF-7 null urothelium is markedly thinned compared with wild-type urothelium. Electron microscopy revealed that null urothelium lacks the intermediate cell layers and molecular marker analyses confirmed this observation. In vitro cell culture experiments indicated that FGF-7 regulates urothelial cell growth, differentiation and stratification. Primary urothelial cultures maintained without FGF-7 ceased to divide and expressed proteins characteristic of terminally differentiated umbrella cells. In contrast, cultures maintained with exogenous FGF-7 contained proliferating epithelial cells with protein expression patterns consistent with those of intermediate cells in addition to terminally differentiated, post-mitotic umbrella cells. Importantly, isolated urothelial cells maintained with exogenous FGF-7 formed a multilayered epithelium in vitro. Conclusions: Collectively these data indicate that FGF-7 is essential for normal bladder urothelial stratification, specifically the formation of the intermediate cell layers. Fibroblast growth factor-7 stimulates urothelial proliferation and delays the differentiation of these cells into post-mitotic umbrella cells. KEY WORDS: bladder, fibroblast growth factor, epithelium, cell differentiation, urothelium

The bladder urothelium accommodates large changes in bladder volume, while maintaining an impermeable barrier between urine and blood. This important functional characteristic of the urothelium depends on its unique multilayered or stratified structure. The stratified layers of the urothelium can be subdivided into 3 distinct zones with distinct morphological and growth properties. The basal zone lies immediately adjacent to the basement membrane is characterized by a single layer of 5 to 10 ␮m. cells. Previous studies indicate that basal cells have a steady but low level of proliferation and give rise to cells of the suprabasal zones, including the intermediate and superficial zones in that order.1 The intermediate zone consists of 3 to 4 layers of 20 ␮m. cells containing lysosomes and numerous cytoplasmic vesicles.2 Although intermediate cells are larger than basal cells, they do not have the differentiated properties of cells comprising the terminally differentiated superficial zone. Intermediate cells can be induced to proliferate when the urothelium is wounded.3 Current hypotheses suggest that intermediate cells fuse to form the most differentiated zone of the urothelium, namely the superficial zone.1 This zone is characterized by a single layer of large multinucleated post-mitotic cells that cover relatively extensive areas of the underlying intermediate zone, giving rise to the term umbrella cells.4 Depending

on the degree of bladder stretch, umbrella cells are 50 to 120 ␮m. wide and are joined by tight junctions that are crucial for maintaining mucosal impermeability.5 Although the distinct properties of the basal, intermediate and superficial layers of the bladder urothelium have been extensively characterized, the cellular and molecular mechanisms that regulate the formation and maintenance of these urothelial cell layers remain poorly understood. We tested the hypothesis that growth factors secreted by stromal cells underlying the bladder urothelium are essential in regulating the urothelial architecture. One candidate growth factor expressed by stromal cells underlying the bladder urothelium is keratinocyte growth factor, or fibroblast growth factor (FGF)-7 (fig. 1, A). FGF-7 has been shown to modulate the growth and differentiation of epithelium expressing FGF receptor-2b, and abundant FGF receptor-2b messenger (m)RNA has been detected in bladder urothelium.6, 7 We used mice expressing normal FGF-7 levels and those in which the FGF-7 gene was deleted8 to determine whether this factor secreted by the bladder stroma has a role in regulating bladder epithelial growth and urothelial formation. METHODS

Bladder tissues. The bladders of 4-week-old C57/BL6 and C57/BL6 FGF-7 ⫺/⫺ null male mice were used for morphological analysis. Null mice were constructed by targeted deletion of exon 1, which is a portion of the FGF-7 gene required for heparin association and subsequent receptor binding.8 In

Accepted for publication June 15, 2001. Supported by National Institutes of Health Grant RO1 DK-45218 (DH) and the James Buchanan Brady Urology Foundation, Department of Urology, New York Presbyterian Hospital-Weill Medical College of Cornell University. 2536

FIBROBLAST GROWTH FACTOR-7 REGULATES BLADDER UROTHELIUM STRATIFICATION

FIG. 1. Mice lacking expression of FGF-7, factor secreted by bladder stroma, exhibit thinned urothelium as compared to wild type mice. Diagram of bladder urothelium indicating expression patterns of FGF-7 and its high affinity receptor, FGFR-2b based on data from Finch et al (A).7 FGF-7 mRNA is expressed by stromal cells underlying FGFR-2b-expressing urothelium. Light microscopic examination of bladders obtained from age and sex matched C57/BL6 FGF7⫹/⫹ (B, wild type) and FGF-7⫺/⫺ (C, null) mice. At low magnification (top panels) relaxed state of bladder tissues is confirmed by presence of numerous mucosal folds extending into lumen. Reduced from ⫻100. At high magnification (bottom panels) it is evident that wild type urothelium (B) is thick and comprised of multiple cell layers. Urothelium of FGF-7 null mice is markedly thinned and appears to lack multiple cell layers observed in wild type tissue (C). Reduced from ⫻400.

these mice FGF-7 mRNA is not detectable in any tissues, as determined by reverse transcriptase-polymerase chain reaction. No significant differences in total body mass or crownto-rump length were observed in null and wild-type animals. However, null mice have 30% fewer nephrons and smaller kidneys compared to the wild-type C57/BL6 mice,9 and they exhibit a greasy and matted-appearing hair coat.8 No other gross abnormalities have been detected in FGF -7 null mice. Preparation of tissues for morphological and molecular marker analyses. Bladders were cut in half before fixation to achieve complete decompression. This maneuver avoids variability in epithelial cell layer number due to bladder contractile status.10, 11 After removal the bladders were fixed, embedded in plastic, and sectioned for light and electron microscopy by conventional techniques.12 For immunocytochemical analyses frozen sections of bladder tissues were prepared.13 Molecular marker analyses. Proteins expressed by bladder epithelium were detected by indirect immunofluorescence microscopy using frozen bladder sections and primary cultures of bladder urothelium. A rabbit polyclonal antibody directed against uroplakin14, 15 was used to identify terminally differentiated umbrella cells. RGE53 (ICN Biomedicals, Inc., Costa Mesa, California), a murine mAb to cytokeratin 18 that has restricted activity with superficial cells,16 was also used as an umbrella cell marker. Epithelial cells below the superficial umbrella cell layer, that is the basal and intermediate cells, were identified by the broadly reacting

2537

monoclonal Ab AE1/AE3 (ICN), which stains acidic and basic cytokeratins.17 The basement membrane separating the urothelium from the underlying lamina propria was detected by a rabbit polyclonal antibody specific for collagen type IV (ICN). Conjugated anti-rabbit and anti-mouse IgGs were used as secondary fluorochromes (Molecular Probes, Eugene, Oregon). Primary urothelial cell cultures. The protocol described by Fujiyama et al18 with certain modifications was used to isolate bladder epithelium from male Swiss Webster mice. Briefly, with the aid of a dissecting microscope and fine-tooth forceps, the muscularis propria was separated from the lamina propria and attached urothelium. The lamina propria/urothelium was then cut into 2 ⫻ 2 mm. pieces and incubated with 0.7 units per mg. Dispase (Sigma Chemical Co., St. Louis, Missouri) in 35 mg./ml. Dulbecco’s phosphate buffered saline for 1 hour on ice. At this point the epithelial sheets were easily separated from the lamina propria. The sheets were collected in cold Dulbecco’s phosphate buffered saline and dissociated into single cell suspensions by incubation with 0.25% trypsin/ethylenediaminetetraacetic acid at 37C for 5 to 10 minutes. Trypsinized cells were mechanically dissociated by rigorous pipetting and then suspended in Dulbecco’s modified Eagle’s medium/10% fetal calf serum. Cell suspensions were counted with a hemocytometer and plated onto 24 mm. collagen Transwell type I coated wells (Becton Dickinson Labware, Bedford, Massachusetts) at 1 ⫻ 105 cells per well with or without 100 ng./ml. recombinant FGF-7 (R & D Systems, Inc., Minneapolis, Minnesota). The medium was changed every 2 to 3 days. Cell proliferation assays. The number of cells present in FGF-7 treated and control urothelial cultures was determined by measuring DNA content with the CyQUANT Cell Proliferation Assay kit (Molecular Probes) according to manufacturer protocols. Paired t tests were done to determine the significance of the difference in DNA content in the 2 groups using Excel software (Microsoft, Redmond, Washington). Proliferating cells in the S phase of the cell cycle were identified by the detection of incorporated 5-bromo-2deoxyuridine (BrdU) using indirect immunofluorescence microscopic techniques, as described by the protocols of the BrdU Labeling and Detection Kit I (Roche Molecular Biochemical, Indianapolis, Indiana). RESULTS

Fibroblast growth factor-7 null urothelium was thinned compared with wild-type urothelium. Examination of 1 ␮m. toluidine blue stained methacrylate sections demonstrated that FGF-7 null bladder urothelium was markedly thinned compared with wild-type urothelium (fig. 1, B and C). Extensive morphometric analyses confirmed this observation. The number of epithelial layers in 12 randomly chosen areas of the urothelium from 3 FGF -7 ⫹/⫹ and 3 FGF-7 ⫺/⫺ mice was documented. To ensure that the areas chosen for counting represented a true cross section of bladder urothelium, only those areas containing nuclei of basal, intermediate and superficial layers were evaluated. The results of these experiments revealed that normal bladder epithelium in wild-type mice consists of a mean plus or minus standard deviation of 5.25 ⫾ 1.14 cell layers. This value is in agreement with previous documentation of the bladder urothelial thickness.4 In contrast, bladder urothelium of the null mice consisted of a mean maximum of 1.67 ⫾ 0.49 cells. Fibroblast growth factor-7 null urothelium lacked intermediate cell layers. Electron microscopy showed that the decreased number of cell layers in null urothelium was due to absent intermediate cell layers. In wild-type urothelium, a contiguous layer of small basal cells was observed immediately contacting the basement membrane and separated from the superficial umbrella cell layer by 3 to 4 layers of larger

2538

FIBROBLAST GROWTH FACTOR-7 REGULATES BLADDER UROTHELIUM STRATIFICATION

intermediate cells with characteristic lysosomes and cytoplasmic vesicles (fig. 2, A). In contrast, null urothelium lacked well organized, stratified layers. Cells with the ultrastructural characteristics of basal cells were noted sporadically in null urothelium. However, large areas of the urothelium were comprised solely of a single cell layer (fig. 2, B). Cells within these portions of the null urothelium exhibited the ultrastructural characteristics of both intermediate and umbrella cells (fig. 2, B). In addition, cells that appeared indistinguishable from wild-type terminally differentiated umbrella cells were present in the null urothelium (fig. 2, C). The basal surface of these cells was characterized by extensive plasma membrane interdigitations. Moreover, it appeared that large areas of the basal surface of these umbrella cells directly contacted the basement membrane. However, we cannot rule out the possibility that thin cytoplasmic extensions of neighboring basal cells underlay this monolayer of umbrella cells. Together these morphological data indicate that FGF-7 null urothelium lacks a well-defined intermediate cell zone. Since many cells in the null urothelium exhibited the ultrastructural characteristics of intermediate and umbrella cells, we assayed the differentiated phenotype of urothelium

in wild-type and FGF-7 null bladders using molecular markers. Frozen sections were stained with antibodies specific for proteins expressed by superficial umbrella cells, all urothelium, and the urothelial basement membrane (fig. 3). Fluorescence signals were not detected when primary antibody incubation was omitted (data not shown). Wild-type urothelium was characterized by a collagen IV rich basement membrane, at least 4 layers of urothelial cells identified by binding a broadly reacting antibody directed against cytokeratins, and a superficial cell layer that bound to antibodies directed against uroplakin and cytokeratin 18, markers of terminally differentiated umbrella cells (fig. 3, A and C). Basal cells cannot be discriminated from intermediate cells by this technique. In contrast, in the null urothelium, uroplakin and cytokeratin 18 expressing umbrella cells were located immediately adjacent to the basement membrane (fig. 3, B and D). Alexa 488 and 586 fluorescence signals were not detected when primary antibody incubations were omitted (data not shown). Together these electron microscopy and immunohistochemical data show that basal, intermediate and terminally differentiated umbrella cells are present in FGF-7 null urothelium. Importantly, these data indicate that FGF-7 is essential for normal urothelial stratification, spe-

FIG. 2. Electron microscopy of FGF-7 ⫹/⫹ wild-type and ⫺/⫺ null bladder urothelium. Superficial surface of wild-type urothelium consists of electron dense umbrella (umb) cell layer characterized by large flattened cells with scalloped apical border and numerous intracellular membrane vesicles (A). Beneath umbrella cell layer were intermediate cells pleomorphic in shape and containing many lysosomes. Three layers of intermediate cells are evident. Beneath intermediate cells and directly opposed to basement membrane (bs mem) were basal cells, identifiable by their location and small size. Reduced from ⫻5,000. FGF-7 null urothelium lacking intermediate cell layers (B). Cell with ultrastructural characteristics of basal cell is adjacent to cell with scalloped apical membrane, consistent with that of umbrella cells, as well as cytoplasmic vesicles more consistent with those in wild-type intermediate cells. Many cells in FGF-7 null urothelium had ultrastructural characteristics of intermediate and umbrella cells. Stromal fibroblast (s fib) is visible underlying basement membrane. Reduced from ⫻5,000. Cells with ultrastructural characteristics identical to those of wild-type urothelium umbrella cells in FGF-7 null bladder (C). Reduced from ⫻7,000. Top inset shows characteristic scalloped apical border and numerous fusiform membrane vesicles of umbrella cells. Reduced from ⫻9,000. Bottom inset shows basal surface of these umbrella cells, characterized by extensive plasma membrane interdigitations, which appears to be in direct contact with basement membrane. Reduced from ⫻9,000.

FIBROBLAST GROWTH FACTOR-7 REGULATES BLADDER UROTHELIUM STRATIFICATION

FIG. 3. Molecular marker analyses of wild-type and FGF-7 null urothelium. Expression of several proteins that discriminate umbrella and intermediate or basal cells was assayed by indirect immunofluorescence microscopy in frozen of bladder tissue sections. RGE 53, mAb directed against cytokeratin-18 followed by Alexa 488 conjugated antimouse IgG was used to identify umbrella cells (green areas) (A and B). Polyclonal Ab directed against collagen type IV followed by Alexa 568 conjugated anti-rabbit IgG was used to detect basement membrane (red areas) separating urothelium from underlying lamina propria (lp). Cell nuclei (blue areas) were visualized by DAPI staining. In wild-type urothelium umbrella cell layer was separated from basement membrane by several layers of cells identified by DAPI stained nuclei. In contrast, thinned FGF-7 null urothelium lacked intermediate cell layers. Umbrella cells were directly opposed to basement membrane and only single layer of DAPI stained nuclei was observed. Umbrella cell layer (red areas) were identified by staining with polyclonal Abs directed against uroplakin followed by Alexa 568 anti-rabbit IgG (C and D). All urothelial cells (green areas) were identified by broadly reacting anti-cytokeratin mAb AE1/AE3 followed by Alexa 488 anti-mouse IgG. White line represents basement membrane location on phase microscopy. In wild-type urothelium several layers of DAPI stained AE1/AE3 positive cells were present beneath superficial umbrella cell layer (red and yellow areas). Yellow color was due to co-localization of antiuroplakin and AE1/AE3 staining. Fibroblast growth factor-7 null urothelium was composed of thin layer of AE1/AE3 and uroplakin positive umbrella cells, and single layer of DAPI stained nuclei. Scale bar represents 10 ␮m.

cifically the formation of the intermediate cell layers of the bladder urothelium in vivo. Fibroblast growth factor-7 modulated urothelial growth and differentiation in vitro. Fibroblast growth factor-7 may regulate urothelial stratification and intermediate cell layer formation by stimulating basal cell proliferation and/or delaying the differentiation of intermediate cells into terminally differentiated umbrella cells. To test these possibilities we cultured primary urothelial cells with and without recombinant FGF-7. The purity of urothelial cultures was assessed by immunofluorescent detection of cytokeratins, which are expressed by epithelium but not by stromal fibroblasts or smooth muscle cells. Cultures were examined after 1 and 7 days of growth using the nuclear stain 4⬘-6-diamidino-2phenylindole (DAPI) to detect all cells and the broadly reacting AE1/AE3 cytokeratin mAb to identify DAPI stained cells exhibiting an epithelial phenotype. All DAPI-stained cells showed abundant AE1/AE3 anti-cytokeratin staining at 1 and 7 days of culture (data not shown). We further analyzed the differentiated phenotype and proliferative state of urothelium in primary cultures maintained with and without 100 ng./ml. FGF-7. Cultures maintained for 7 days with Dulbecco’s modified Eagle’s medium/10% fetal calf serum (control medium) consisted entirely of a homogeneous population of uroplakin positive, AE1/AE3 positive epithelium characteristic of umbrella cells (fig. 4, A, C and E). 5-Bromo-2-deoxyuridine incorporation studies showed that few, if any, cells in control cultures were proliferating after 7 days of culture (fig. 4, G). In contrast, cultures maintained with exogenous FGF -7 consisted of a mixed population of uroplakin positive, AE1/AE3 positive and uroplakin negative, AE1/AE3 positive cells (fig. 4, B, D and F). Protein expression patterns of the latter cell type were consistent with a basal or intermediate cell phenotype. Fluorescence

2539

staining was not detected when primary antibody incubation was omitted. Furthermore, BrdU incorporation studies revealed that a large percent of AE1/AE3 positive, uroplakin negative cells in FGF-7 treated cultures were proliferating (fig. 4, H). Anti-BrdU staining was not observed in cultures that were not incubated with BrdU (data not shown). In addition, fluorescent staining was not detected in BrdU cultures when primary antibodies were omitted. Most proliferating BrdU incorporating cells were uroplakin negative. Thus, uroplakin negative cells in treated cultures showed growth and protein expression characteristics consistent with an intermediate cell phenotype. Furthermore, many BrdU incorporating, uroplakin negative cells appeared to be beneath large uroplakin positive cells, suggesting that a multilayered urothelium may have been present. Thus, only treated cultures contained cells with the protein expression

FIG. 4. Urothelium was isolated and cultured for 7 days with Dulbecco’s modified Eagle’s medium (DMEM)/10% fetal calf serum (FCS) as controls (A, C, E and G) or 10% Dulbecco’s modified Eagle’s medium/10% fetal calf serum supplemented with 100 ng./ml. recombinant FGF-7 as treated specimens (B, D, F and H). Cultured cells with some characteristics of terminally differentiated umbrella cells were identified by uroplakin expression using anti-uroplakin polyclonal Abs followed by Alexa 568 anti-rabbit IgG (red areas) (A and B). AE1/AE3 anti-cytokeratin mAbs followed by Alexa 488 antimouse IgG were used to detect all urothelial cell types (green areas) (C and D). Both antibodies were also used simultaneously (E and F). Cultures were incubated with BrdU 12 hours before fixation and cells that progressed through S phase of cell cycle were detected by indirect immunofluorescence microscopy using anti-BrdU mAbs followed by fluorescein isothiocyanate anti-mouse IgG, showing nuclei (green areas) (G and H). Control cultures were homogeneous population of cells labeled with Abs against uroplakin and AE1/AE3 anti-cytokeratin mAb (A, C and E). When Abs were visualized together, cells in these cultures appeared yellow due to co-localization of green and red fluorochromes (E). In contrast, FGF-7 treated cultures (B, D and F) were characterized by uroplakin positive, AE1/ AE3 positive cells and uroplakin negative, AE1/AE3 positive cells (green areas). This latter staining pattern was consistent with basal or intermediate cell phenotype. Overlapping yellow and green areas suggested that multilayered epithelium formed in some treated culture areas (F). Few proliferating cells were present in homogeneously stained, uroplakin positive control cultures, evidenced by paucity of BrdU incorporating nuclei (green areas) (G). In contrast, many BrdU incorporating nuclei were detected in treated cultures (H). Scale bar indicates 10 ␮m. Reduced from ⫻20.

2540

FIBROBLAST GROWTH FACTOR-7 REGULATES BLADDER UROTHELIUM STRATIFICATION

patterns and growth properties expected of basal or intermediate cells. In addition, FGF-7 treated cultures exhibited a statistically significant 39% increase in cell number, as determined by quantitative DNA measurements (p ⫽ 0.01). We also observed that cultures with added FGF -7 often formed overlapping areas of umbrella and intermediate/basal cells, indicating formation of a multilayered epithelium (fig. 4, F). This layering occurred in proportion to local cell concentration and, therefore, was evident much less often in control cultures. DISCUSSION

Although the proliferative state and differentiated phenotype of epithelium comprising the basal, intermediate and superficial urothelial zones have been extensively characterized, the molecular mechanisms regulating urothelial stratification remain poorly understood. Substantial progress has been made in characterizing the tissue source and identity of growth factors essential for urothelial cell viability and proliferation in vitro. However, few groups have successfully established primary stratified urothelial cultures containing basal, intermediate and superficial cell layers.18, 19, 20, 21 Truschel et al recently reported that a multilayer urothelium can be generated in vitro when isolated bladder epithelium are plated at a density approximating the cellularity of urothelium in vivo.21 These cells are initially cultured with medium that inhibits cell differentiation in other systems. They are then exposed in the short term to medium that supports terminal epithelial cell differentiation. Using these protocols primary urothelial cultures show most if not all of the differentiated properties of native urothelium and provide an elegant model system for studying urothelial function in vitro. However, these urothelial cultures become senescent, suggesting that the factors required for maintaining regenerative, stratified urothelium are not present. The pioneering study of Donjacour and Cunha showed that the structure and differentiated phenotype of epithelium in the prostate, seminal vesicle and bladder can be modulated by stromal cells from various organs.22 Furthermore, Fujiyama et al observed that isolated bladder epithelium cultured on top of synthetic lamina propria including stromal fibroblasts formed a multilayered epithelium with the morphological, molecular and functional properties of stratified urothelium.18 Most importantly, cells in the basal layer of urothelium co-cultured with stromal fibroblasts continued to proliferate in vitro. Collectively these studies led us to hypothesize that stromal factors are essential for the formation and maintenance of stratified bladder urothelium. One of the growth factors secreted by bladder stroma that is likely to regulate urothelial growth, differentiation, and stratification is FGF-7. The high affinity FGF-7 receptor, fibroblast growth factor-R2b, is abundantly expressed in the bladder urothelium, and members of the fibroblast growth factor gene family initiate a variety of cellular responses in fibroblast growth factor-R-expressing epithelium including cell division, modulations in cell differentiation, and migration. Furthermore, FGF -7 is an important mediator of the epithelial-mesenchymal interactions required for the development of several organs in the gastrointestinal tract,23 and it plays a role in the androgen-dependent branching of the seminal vesicles24 and prostate.25 Finally, FGF-7 secreted by stromal cells of the developing kidney regulates the growth and differentiation of the embryonic renal collecting sytem.9 In this study we showed that FGF-7 is essential for urothelial stratification in vivo. Bladder urothelium of FGF-7 null mice lacked the 3 to 4 layers of intermediate cells characteristic of wild-type urothelium. It has previously been demonstrated that an increased level of exogenous FGF-7 causes urothelial proliferation in rats, as evidenced by the histolog-

ical appearance of increased epithelial cellularity, increased mitoses, and increased proliferating cell nuclear antigen and BrdU expression in basal and intermediate cell nuclei in vivo.10 Injection of FGF-7 in neonatal mice has resulted in a substantial increase in the cellularity of the intermediate layers of urothelium compared with the basal and luminal layers.3 Collectively these data indicate that FGF-7 regulates urothelial stratification in vivo. Elevated levels of this factor induce an increased number of intermediate layers, while its absence results in a bladder urothelium that lacks the intermediate cell layers. We also investigated how FGF-7 induces the generation of the intermediate cell layers in vivo. It is likely that intermediate zone thickness is regulated by the rate of basal and intermediate cell proliferation as well as the rate at which these cells mature into terminally differentiated, post-mitotic umbrella cells. We prepared pure primary bladder epithelial cell cultures and observed that these cells ceased to divide after 7 days of culture with Dulbecco’s modified Eagle’s medium/10% fetal calf serum. In addition, they showed certain but not all properties of terminally differentiated umbrella cells. In contrast, cells maintained with recombinant FGF-7 remained proliferative after 7 days of culture and contained a mixed population of cells with the molecular characteristics of terminally differentiated umbrella cells and intermediate or basal cells. Furthermore, exogenous FGF-7 induced a significant increase in the number of cells in isolated urothelial cell cultures. Collectively, these data suggest that FGF-7 mediates formation of the intermediate cell layers by stimulating the proliferation and delaying the maturation of bladder epithelial cells into the terminally differentiated, postmitotic cell phenotype. CONCLUSIONS

Our data demonstrate that FGF-7 secreted by bladder stroma is essential in the formation and maintenance of the intermediate urothelial cell layers. Strikingly, FGF-7 mRNA increase after experimentally induced wounding in mice.3 These data raise the possibility that wound healing depends on both stimulating basal and intermediate cell proliferation as well as prolonging the time that these cells remain in the cell cycle. Although FGF-7 does not appear to be essential for skin wound healing in mice, FGF-7 null mice had no structural changes in the number of layers comprising the epidermis. Thus, it is likely that other factors or members of the fibroblast growth factor gene family can substitute for FGF-7 in regulating keratinocyte proliferation, differentiation and epidermal thickness. In contrast, bladder urothelium thickness was markedly altered in null mice. Thus, it is possible that this member of the fibroblast growth factor gene family has a crucial role in bladder wound healing. Studies to assess the requirement of FGF-7 in bladder wound healing using FGF-7 null mice are currently under way. Drs. Elaine Fuchs and Linda Degenstein, Departments of Molecular Genetics and Cell Biology, Howard Hughes Medical Institute, University of Chicago, Chicago, Illinois provided FGF-7 -/- mice and Dr. T.-T. Sun, Epithelial Biology Group, Department of Dermatology, New York University Medical School, New York, New York provided polyclonal antibodies against uroplakin. REFERENCES

1. Martin, B. F.: Cell replacement and differentiation in transitional epithelium: a histological and autoradiographic study of the guinea-pig bladder and ureter. J Anat, 112: 433, 1972 2. Steers, W. D.: Physiology and Pharmacology of the Bladder and Urethra. In: Campbell’s Urology, 7th ed. Edited by P. C. Walsh, A. B. Retik, E. D. Vaughan, Jr. et al. Philadelphia: W. B. Saunders, chapt. 26, p. 870, 1998 3. Baskin, L. S., Sutherland, R. S., Thomson, A. A. et al: Growth

FIBROBLAST GROWTH FACTOR-7 REGULATES BLADDER UROTHELIUM STRATIFICATION factors in bladder wound healing. J Urol, 157: 2388, 1997 4. Cotran, R. S., Kumar, V. and Robbins, S. L.: The Lower Urinary Tract. In: S. L. Robbins Pathologic Basis of Disease, 4th ed. Philadelphia: W. B. Saunders, p. 1083, 1989 5. Lewis, S. A.: Everything you wanted to know about the bladder epithelium but were afraid to ask. Am J Physiol Renal Physiol, 278: F867, 2000 6. Rubin, J. S., Bottaro, D. P., Chedid, M. et al.: Keratinocyte growth factor. Cell Biol Int, 19: 399, 1995 7. Finch, P. W., Cunha, G. R., Rubin, J. S. et al: Pattern of keratinocyte growth factor and keratinocyte growth factor receptor expression during mouse fetal development suggests a role in mediating morphogenetic mesenchymal-epithelial interactions. Dev Dyn, 203: 223, 1995 8. Guo, L., Degenstein, L. and Fuchs, E.: Keratinocyte growth factor is required for hair development but not for wound healing. Genes Dev, 10: 165, 1996 9. Qiao, J., Uzzo, R., Obara-Ishihara, T. et al: FGF-7 modulates ureteric bud growth and nephron number in the developing kidney. Development, 126: 547, 1999 10. Yi, E. S., Shabaik, A. S., Lacey, D. L. et al.: Keratinocyte growth factor causes proliferation of urothelium in vivo. J Urol, 154: 1566, 1995 11. Kissane, J. M.: Development and structure of the urogenital system. In: Urological Pathology. Edited by W. M. Murphy. Philadelphia: W. B. Saunders, p. 14, 1989 12. Herzlinger, D. A. and Ojakian, G. K.: Studies on the development and maintenance of epithelial cell surface polarity with monoclonal antibodies. J Cell Biol, 98: 1777, 1984 13. Herzlinger, D. A., Easton, T. G. and Ojakian, G. K.: The MDCK epithelial cell line expresses a cell surface antigen of the kidney distal tubule. J Cell Biol, 90: 269, 1982 14. Moll, R., Wu, X.-R., Lin, J.-H. et al: Uroplakins, specific membrane proteins of urothelial umbrella cells, as histological markers of metastatic transitional cell carcinomas. Am J Pathol, 147: 1383, 1995 15. Wu, R.-L., Osman, I., Wu, X.-R. et al: Uroplakin II gene is

16.

17. 18.

19. 20. 21. 22. 23. 24. 25.

2541

expressed in transitional cell carcinoma but not in bilharzial bladder squamous cell carcinoma: alternative pathways of bladder epithelial differentiation and tumor formation. Cancer Res, 58: 1291, 1998 Southgate, J., Harnden, P. and Trejdosiewicz, L. K.: Cytokeratin expression patterns in normal and malignant urothelium: a review of the biological and diagnostic implications. Histol Histopathol, 14: 657, 1999 Cooper, D. and Sun, T.-T.: Monoclonal antibody analysis of bovine epithelial keratins. Specific pairs as defined by coexpression. J Biol Chem, 261: 4646, 1986 Fujiyama, C., Masaki, Z. and Sugihara, H.: Reconstruction of the urinary bladder mucosa in three-dimensional collagen gel culture: fibroblast-extracellular matrix interactions on the differentiation of transitional epithelial cells. J Urol, 153: 2060, 1995 De Boer, W. I., Rebel, J. M., Vermey, M. et al: Multiparameter analysis of primary epithelial cultures grown on cyclopore membranes. J Histochem Cytochem, 42: 277, 1994 Cilento, B. G., Freeman, M. R., Schneck, F. X. et al: Phenotypic and cytogenetic characterization of human bladder urothelia expanded in vitro. J Urol, part 2, 152: 665, 1994 Truschel, S. T., Ruiz, W. G., Shulman, T. et al: Primary uroepithelial cultures: a model system to analyze umbrella cell barrier function. J Biol Chem, 274: 15020, 1999 Donjacour, A. A. and Cunha, G. R.: Induction of prostatic morphology in urothelium by seminal vesicle mesenchyme. Development, 121: 2199, 1995 Mason, I. J.: The ins and outs of fibroblast growth factors. Cell, 78: 547, 1994 Alarid, E. T., Rubin, J. S., Young, P. et al: Keratinocyte growth factor functions in epithelial induction during seminal vesicle development. Proc Natl Acad Sci USA, 91: 1074, 1994 Sugimura, Y., Foster, B. A., Hom, Y. K. et al: Keratinocyte growth factor can replace testosterone in the ductal branching morphogenesis of the rat ventral prostate. Int J Dev Biol 40: 941, 1996