| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1Program in Human Urothelial Biology, Seattle Children's Hospital Research Institute, Seattle; 2Department of Urology, University of Washington School of Medicine, Seattle; 4Stereotome Northwest, Issaquah, Washington; and 3Department of Genome Sciences, Ernst Orlando Lawrence Berkeley Laboratory, Berkeley, California
Submitted 3 January 2007 ; accepted in final form 21 June 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
transitional epithelium; fibroblast growth factor-7; fibroblast growth factor-10; keratinocyte growth factor; stereology
Transitional epithelium that lines the urinary tract exhibits one of the slowest turnover rates among mammalian epithelia, estimated at once every 365 days (19, 20, 30). Although the urothelial cells that comprise this epithelia rest in G0 during normal conditions, they remain active in metabolic processes that accompany shape change during bladder emptying and filling. In response to injury, transitional epithelium exhibits a remarkable ability to turnover within 24–48 h after the initial insult. The identity of the mitogenic signal that gives rise to basal urothelial cell proliferation has been attributable to growth factors. Such factors are classified as paracrine, autocrine, or juxtacrine, depending on the originating and target cell type. Mitogens known to act on human urothelial cells include epidermal growth factor (4, 50), heparin-binding epidermal growth factor-like growth factor (15, 52), activators of the estrogen receptor (54, 55), transforming growth factor-
(14), transforming growth factor-
(13), fibroblast growth factor (FGF)-1 (14), FGF-10 (1), and insulin growth factors 1 and 2 (13, 61). The model most frequently used to study the urothelial response to injury is partial outlet obstruction of the bladder via urethral ligation (7). The wound-healing response appears to be constant from animal model to animal model and mimics the clinical response seen in humans (4, 31, 37). Such events include urothelial cell proliferation within 24–48 h, subsequent hypertrophy of the bladder wall, and alterations in detrusor muscle and the extracellular matrix. Experimental animal models have been exploited to study bladder dysfunctions that are similar to those observed in men with benign prostatic hyperplasia (25, 26). Furthermore, this experimental model has direct implications for other clinical disease states such as congenital (i.e., posterior urethral valves) or acquired (urethral stricture or bladder neck contraction) urinary tract obstruction.
FGF-7 is a 194-amino acid polypeptide implicated in both the induction of basal urothelial cell proliferation and the expansion of transitional epithelium (7). Also known as keratinocyte growth factor (KGF) (45, 46), FGF-7 is one of 24 members of the FGF family of polypeptide growth factors. The crystal structure of FGF-7 has been solved (40, 62). The protein folds into a beta-trifoil motif similar to other members of the FGF family whose structures have been solved. Although FGFs 11–14 exhibit striking structural similarities to FGF-10, FGFs 11–14 have diverged to direct related surfaces toward interaction with protein targets distinct from canonical FGF receptors (FGFR) (39).
FGF-7 and FGF-10 are considered to be paracrine factors, originating in mesenchyme but active only on epithelium (1, 45, 46, 65). The formation of a specific FGF-7-containing FGFR signal transduction complex requires three components: the FGF-7 polypeptide, proteoglycans that contain either heparan (23, 32), dermatin (56), or chondroitin (47) sulfate, and an alternatively spliced tyrosine kinase protein product of the FGFR2 gene (32–34). This receptor, known as the KGF receptor (KGFR), the FGF-7 receptor, or isoform 2, is translated from the FGFR2IIIb mRNA splice variant and is a cell-surface transmembrane protein that undergoes dimerization upon ligand binding and subsequent autophosphorylation of intracellular tyrosine residues (51). Allelic mutations in the FGFR2 exons that encode the isoform 2 splice variant are known to give rise to a number of birth defects that include craniosynostosis syndromes (38). It is believed that the FGFR2 isoform 2 splice variant serves as a receptor for both FGF-7 and FGF-10.
Mice that contain a targeted disruption of the FGF-7 gene exhibit defects in 1) cells that give rise to the hair shaft (18), 2) ureteric bud outgrowth (43), and 3) stratification of bladder urothelium (53). The collective evidence demonstrates an essential role for FGF-7 in development, growth, differentiation, and homeostasis of the mucosal lining of the urinary tract. In contrast to the FGF-7-null mouse, the FGF-10-null mouse lacks limb bud initiation and lung development (35). In addition, the FGF-10-null mouse exhibits gross anatomical defects in the urinary tract including a urothelium that fails to stratify (64). Consequently, FGF-10-null mice die at birth, precluding their use in experimental models of partial bladder obstruction.
This study demonstrated the mitogenic potential of FGF-7 on human and murine urothelial cells. The growth factor was found to stimulate resting cells to exit from G0 and traverse the cell cycle. To determine whether the mitogen was responsible for the induction of urothelial cell proliferation during the initial response of the bladder to injury, we examined whether there was a difference in the urothelial response to injury between wild-type and FGF-7-null mice in a model of partial outlet obstruction. Stereological analysis of obstructed bladders revealed that a statistically significant difference did not exist indicating that FGF-7 is not essential for urothelial cell proliferation during the response to partial outlet obstruction of the urinary bladder. Further stereological analysis revealed a significant 5.5-fold increase of FGF-10 synthesis in obstructed FGF-7-null mice via a compensatory paracrine mechanism.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Cloning of human bladder FGF-7 cDNA into the bacterial expression vector pET21d. The insert coding region from pSTBlue1-FGF7 was recloned into the pET21d expression vector to ultimately form the expression plasmid pET21d-FGF7. Six and one-half micrograms of pSTBlue-FGF7 were restricted with NcoI and XhoI endonucleases and electrophoresed through an agarose gel. The DNA of a 0.52-kbp band was excised from the gel, extracted, and ligated to the expression plasmid pET21d, previously prepared by incubation with NcoI and XhoI and dephosphorylation with shrimp alkaline phosphatase. After transformation into BL21trxB(DE3) E. coli, recombinant colonies were generated by growth on Luria broth-agar plates that contained 50 µg/ml carbenicillin, stored as glycerol stocks, and designated pET21d-FGF7. The resultant open reading frame that included six COOH-terminal His residues generated a fusion protein designated as rFGF7-His.
Recombinant preparations of rFGF7s. Two types of recombinant (r) FGF-7 were used in this study: rFGF7-His, prepared in our laboratory, and Palifermin (rFGF-7 lacking a COOH-terminal His-tag), prepared and provided to us by Amgen (66).
Expression, isolation, and characterization of rFGF7-His in E. coli.
The expression of recombinant FGF-7 (rFGF-7) by transformed E. coli was found to be dependent on the addition of isopropyl-1-thio--D-galactopyranoside, a reagent known to inactivate the lac repressor, thereby permitting synthesis of FGF-7 mRNA.
A loop of recombinant bacteria from a frozen glycerol stock was streaked on a plate of agar that contained media A (Luria broth, 100 µg/ml carbenicillin, and 0.2% glucose) and incubated at 37°C for 16–24 h. A 50-ml starter culture in media A was inoculated with a single colony and propagated until the optical density at 600 nm reached 
0.6. After storage of the culture overnight at 4°C, cells were recovered by sedimentation at 5,000 g, resuspended in 4 ml of media A, and inoculated into 2 l of prewarmed media A. Once the cultures reached an optical density at 600 nm of 0.6, isopropyl-1-thio--D-galactopyranoside was added to a final concentration of 1 mM. After 3–4 h of continued growth at 37°C, cells were collected by sedimentation at 5,000 g at 4°C and frozen at –20°C.
Luria broth used in these experiments contained peptone derived from meat so as to eliminate lactose that would normally be found in peptone derived from milk. We found that lactose in the medium elicits premature activation of the T7lac promoter and of rFGF7-His synthesis.
Thawed pET21d-FGF7-BL21trxB(DE3) cells were lysed with Bug Buster reagent that contained additional complete protease inhibitors. Lysed material was fractionated into soluble and insoluble fractions by centrifugation at 20,000 g for 45 min at 4°C.
Soluble material was dialyzed overnight at 4°C vs. buffer A (0.1 M NaH2PO4, 0.3 M NaCl, and 0.05 M Tris·HCl; pH 8.0). After clarification of the dialyzate by centrifugation at 20,000 g for 20 min at 4°C, imidazole was added to the soluble phase at a final concentration of 0.025 M. This sample mixture was then applied to a 4-ml column of Ni-NTA resin previously equilibrated in buffer A (pH 8.0). The column was developed with buffer A (pH 8.0) until the A280 nm < 0.01, with buffer A (pH 6.0) until the A280 nm < 0.01, with buffer A (pH 5.3) to elute the monomeric rFGF7-His protein fraction, and finally with buffer A (pH 4.5) to elute the multimeric rFGF7-His protein fraction. Column eluates were monitored at 280 nm by a Pharmacia Uvicord spectrophotometer connected to analog-to-digital boards of an Intel-based computer workstation running Rainin Dynamax software.
Alternatively, heparin affinity chromatography was used to isolate rFGF7-His by the following procedure. Soluble fractions were dialyzed against 0.05 M Tris·HCl (pH 7.4) that contained 0.2 M LiCl, and loaded onto a HiTrap Heparin column, equilibrated in the same buffer, at 1 ml/min. After a baseline was established, the column was developed with step gradients of increasing LiCl concentrations: 0.5, 1.0, 1.5, and 2.0 M. rFGF7-His was eluted at 1.0 M LiCl.
NH2-terminal amino acid sequencing of rFGF7-His. A 1-ml fraction eluted from a nickel-chelate affinity chromatography column that contained 25 µg/ml of rFGF7-His was precipitated with trichloroacetic acid, washed, dried, and dissolved in 0.125 M Tris·HCl (pH 6.8), 2% (wt/vol) SDS, 10% glycerol, 0.05% bromphenol blue, 0.05 M dithiothreitol. After heating to 37°C for 20 min, the fraction was electrophoresed through a 15% polyacrylamide gel that contained 0.1% SDS. Thioglycolate (0.1 mM) was present in the upper buffer chamber to scavenge unpolymerized acrylamide that could potentially result in blockage of the NH2 terminus. Following electrophoresis, fractionated proteins were electrotransferred to a polyvinylidene difluoride membrane in 0.01 M MOPS (pH 11.0) that contained 20% methanol, visualized by staining with Coommasie Brilliant Blue R-250, and destained. The rFGF7-His band was excised from the membrane and sequenced by Edman degradation with an Applied Biosystems 477A protein sequencer.
[3H]Thymidine incorporation assays of urothelial cell proliferation. Because of the use of human transitional epithelial tissue for this study, these experiments were reviewed by the Institutional Review Board of Seattle Children's Hospital. Primary cultures of bladder urothelial cells derived from surgical explants were grown as previously described (1). Assays of DNA synthesis of bladder urothelial cells were performed as previously described (1).
Our preparations of rFGF7-His from soluble extracts were biologically active and the engineering of a COOH-terminal hexamer of histidine residues did not affect solubility, folding, or ability to bind to heparin. However, we found that the yields of bacterial rFGF7-His were disappointing, a result in agreement with a prior report that established that the growth factor exhibited limited stability in aqueous media, as it undergoes denaturation followed by aggregation at 37°C (11). Therefore, we used Amgen's preparations of active recombinant FGF-7 for use in assays where milligram quantities of the growth factor were needed, e.g., injection into mice.
Mice. The Institutional Animal Care and Use Committee of Seattle Children's Hospital approved these experiments. Wild-type C57BL/6J mice were obtained from B and K (Kent, WA). FGF-7-null mice (18) of the same parental inbred strain were provided by Dr. E. Fuchs (The Rockefeller University). All mice were 8-wk-old females.
Induction of murine urothelial cell proliferation.
rFGF-7, lacking a COOH-terminal histidine tag and ending at Thr165 (see 
Fig. 2), was obtained from Amgen as the pharmaceutical Palifermin and dissolved in vehicle [120 mM NaCl, 2.7 mM KCl, 4 mM NaH2PO4 (pH 7.4), 5 µg/ml heparin] at a final growth factor concentration of 100 µg/ml. One milliliter was administered daily by subcutaneous injection into the neck folds of wild-type C57BL/6J mice. On day 15, mice were euthanized and their bladders were harvested en bloc by anatomical positioning on a paraffin block. Specimens were fixed with methyl Carnoy's solution for 16–24 h, transferred to 120 mM NaCl, 2.7 mM KCl, 4 mM NaH2PO4 (pH 7.4), embedded in paraffin, cut into 5- to 10-µm-thick sections, and processed by Ki-67 immunohistochemistry (6) or visualized with Masson's trichrome stain.
|
|
Immunostaining with antibodies specific for Ki-67 and FGF-10. Specimens were fixed with methyl Carnoy's solution for 16–24 h, transferred to 120 mM NaCl, 2.7 mM KCl, 4mM NaH2PO4 (pH 7.4), embedded in paraffin, cut into 5- to 10-µm-thick sections, and processed by Ki-67 immunohistochemistry (6) or visualized with Masson's trichrome stain.
Diaminobenzidine colorimetric staining with primary monoclonal mouse anti-human FGF-10 IgG at a concentration of 15 µg/ml and goat anti-mouse horseradish peroxidase-conjugated IgG at 1:500 dilution were performed with 5% goat serum block. Slides were counterstained with a 1:10 dilution of Gill's formula for 1 min and fixed.
Stereology procedures. Estimates of the number of total and proliferating (Ki-67 positive) urothelial cells, and FGF-10-positive fibroblast cells, of obstructed murine bladders were determined by use of the Optical Fractionator (60). The Optical Fractionator method combines a three-dimensional unbiased counting frame (the optical disector) and an unbiased sampling method (the fractionator) to generate estimates of the total number of particles of interest, in this case Ki-67-positive urothelial cells and FGF-10-positive fibroblasts. The ensuing estimate is unbiased and free of volume artifacts commonly present with other counting methods. Implementation of the Optical Fractionator for determining FGF-10-positive cells was performed using a BH-2 Olympus microscope equipped with an Optronics DEI 750 digital video camera and Ludl high precision motorized stage. Data collection and microscope operation were controlled by Stereo Investigator software (version 5.05, MicroBrightField, Williston, VT).
The application of the fractionator in the present sampling design for counting the FGF-10-positive cells consisted of initially systematically collecting every 40th section across the entire bladder starting with a random start (e.g., a random start might be 12, therefore collect the 12, 52, 92... sections). Hence, this set of sections (and the structure of interest contained within, i.e., the obstructed bladder) represented 1/40th of the entire structure. Following staining of this set of sections, Ki-67-positive urothelial cells were counted in a known aerial fraction of the transitional epithelial layer. The aerial fraction was determined by placing a random stepping grid, generated by the Stereo Investigator software program, over the section. An appropriate stepping grid (500 x 500 µm, or an area of 250,000 µm2) was used to place sequential disectors across the surface of the sections. A disector frame (10 x 10 = 100 µm2) was placed in each stepping grid and used to count Ki-67-positive cells according to optical disector counting rules. Immunopositive cells were readily identified by the dark brown coloration of the nuclei characteristic of the staining procedures (see Fig. 5 for example of staining). The overlay of this stepping grid was placed randomly (independently) over the transitional epithelium, thus ensuring that each step was systematic and random. In our experiments, the optical disectors sampled 1/2,500th of the transitional epithelium. Combined with the 1/40th of the structure represented by the section, the optical disectors sampled 1/100,000th of the entire structure. Cell counts were summed from all disectors across all sections collected, and the total number was determined by multiplying the number counted by the inverse of the fraction sampled. Systematic random sampling has been determined to be fair because every part of the original structure had an equal probability of being selected due to the random position set within the first interval, and systematic random sampling offers additional advantages of efficiency (12). The counting procedures for the FGF-10-positive cells were identical to those described above for the Ki-67-positive cells with the exception that the disector frames were placed sequentially next to each other so that the entire surface of the tissue section was tiled with disector frames (i.e., the stepping interval was the same size as the disector frame).
|
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Analysis of human urinary bladder FGF-7 amino acid sequence. Figure 1 displays the deduced amino acid sequence for the 173-amino acid recombinant rFGF7-His protein. The NH2 terminus was engineered to contain a Met residue to provide a codon for initiation of translation in E. coli. The COOH terminus was engineered to contain a His-hexamer to provide for detection during biological assays and for isolation of the recombinant protein by metal-chelate affinity chromatography. The calculated Mr and isoelectric point of rFGF7-His is 20,151 and 9.58, respectively.
Motifs and sites for posttranslational modification that are predicted to occur in the wild-type FGF-7 sequence are also shown in Fig. 1. Naturally occurring FGF-7 proteins are predicted to contain sites for N-linked glycosylation (Asn16), casein kinase II phosphorylation site (Ser18 and Thr150), phosphorylation site by protein kinase C (Ser124), and amidation (Gly147). A 24-amino acid region (Gly96-Tyr119) is the consensus sequence for the FGF family of heparin-binding polypeptides (49) and a potential transmembrane segment (42). The "glycine box" sequence (NQKGIPVRG, residues 139–147) is the major determinant for the specificity of the binding of FGF-7 to heparin sulfate-FGFR complexes (29). Residues that have been implicated in binding to heparin (40, 62) were identified as Arg43, Asn117, Asn139, Gln140, Val145, Lys148, Asn154, Lys155, and Thr156 (Fig. 1). These heparin-binding residues form a positively charged motif that is present on the surface of the macromolecule (Fig. 2). The heparin-binding motif of rFGF7-His is therefore exposed to the solvent and, more importantly, available to bind to and interact with heparan-, chondroitin-, or dermatin-sulfate proteoglycans.
Expression, isolation, and characterization of recombinant FGF-7 in E. coli.
The expression of the bladder FGF-7 sequence in bacteria was subsequently realized as a fusion protein with a hexamer of His residues at the COOH terminus (Fig. 1). This recombinant fusion protein, designated rFGF7-His, was found to comprise 1–2% of the total cellular protein and to partition equally between the soluble and insoluble fractions of lysed BL21trxB(DE3) E. coli (data not shown). We pursued the characterization of the soluble form of rFGF7-His with respect to 1) intactness, 2) ability to interact with heparin, 3) folding, and 4) biological activity in vitro.
rFGF7-His was determined to be full-length because 1) amino acid sequencing of the NH2 terminus (ACNDMTPEQMATNV) of the isolated protein was found to be consistent with the NH2 terminus of the mature form of the NCBI reference sequence NM_002009 [GenBank] .2 and 2) the COOH-terminal His hexamer conferred the ability of rFGF7-His to bind avidly to nickel-chelate affinity resins (Fig. 3, lane 1).
|
24 kDa by nickel-chelate (Fig. 3, lane 1) and heparin affinity chromatographies (Fig. 3, lane 2). The ability of soluble rFGF7-His to interact with nickel-chelate resins through its COOH-terminal His-hexamer (Fig. 3, lane 1) demonstrates that this His- hexamer is localized to the periphery of the folded protein, in agreement with the crystal structure of the human (40) and rat (62) FGF-7 proteins. Concerns that manipulation of the COOH-terminal region of rFGF7-His would influence its folding and biological activity (36) prompted us to ask whether our preparations of rFGF7-His would bind to heparin affinity resins in the same manner as reported in the literature. Figure 3 (lane 2) demonstrates that rFGF7-His elutes from heparin affinity chromatography columns at 1.0 M LiCl, an ionic strength in agreement with that reported for native (46) or rFGF-7 (44). Sequential chromatography on heparin affinity followed by nickel-chelate affinity chromatographies (Fig. 3, lane 3) or the converse (nickel-chelate followed by heparin, Fig. 3, lane 4) clearly demonstrates that the presence of the COOH-terminal His-hexamer does not impact the ability of the recombinant protein to interact with heparin or fold into a conformationally stable state. Stabilization of rFGF7-His was found to be enhanced by inclusion of heparin in the storage buffer solution (not shown). Biological activities of recombinant FGF-7. rFGF7-His was found to exhibit mitogenic activity on human urothelial cells in vitro (Fig. 4B). Cultures of urothelial cells that attained confluence by incubation in growth media reduced the rate of [3H]thymidine incorporation from 37,600 ± 900 to 6,000 ± 360 cpm after switching to starvation medium (Fig. 4B). The addition of increasing concentrations of rFGF7-His to cells in starvation medium resulted in a dose-dependent increase of incorporation of [3H]thymidine into urothelial cell DNA. A 5.6-fold increase in incorporation was observed at a concentration of 0.1 ng/ml relative to no input growth factor. The mitogenic activity of rFGF7-His was dependent on the absence of input heparin. Incorporation of [3H]thymidine into urothelial cell DNA as a function of rFGF7-His was inhibited by 10 µg/ml heparin, in agreement with a prior report (22, 36), presumably because of the influence of extracellular matrix components present in the cultures. This activity was observed only with confluent cultures, whereas nonconfluent cultures resulted in nonsignificant levels of incorporation (data not shown). The culture in Fig. 4A was confluent and displayed the typical cobblestone appearance of urothelial cells in vitro, compared with published photographs (50). Atop Fig. 4A is a "giant" cell with two prominent nuclei; this urothelial cell subtype corresponds to the large, polygonal superficial cell that comprises the lumenal layer of transitional epithelium. Superficial urothelial cells are believed to be the largest epithelial cells in mammals. Secretory granules, lipid inclusions, and lysosomes are prominent in these superficial cells, which are also positive for uroplakin (data not shown).
|
Immunostaining with antibodies specific for the nuclear protein Ki-67, an established marker of epithelial cell proliferation (16, 17), demonstrated that basal urothelial cells were traversing the cell cycle (Fig. 5). The Ki-67 signal persisted throughout the period of rFGF-7 (Palifermin) administration and was prevalent in all animals tested (n = 3). A Ki-67 signal was observed to decrease in intensity as cells detached from the urothelial basement membrane and migrated toward the bladder lumen. Animals receiving vehicle did not exhibit Ki-67 immunoreactivity. The positive correlation between rFGF-7 (Palifermin) administration, expansion of the layers of transitional epithelium, and Ki-67 immunoreactivity confirms that FGF-7 is a mitogen for urothelial cells and further validates FGF-7-null mice as an experimental model.
FGF-7 is a nonessential mitogen during partial bladder outlet obstruction. An experimental mouse model of partial bladder outlet obstruction was developed wherein a partial surgical ligature was placed around the urinary bladder outlet. In this model, animals that underwent partial ligation exhibited bladders that were abnormally distended. The hallmark feature of transitional epithelium over the 8 days of obstruction was the marked increase of stratification of transitional epithelium and of urothelial cells traversing the cell cycle. The mitogenic signal that stimulated the exit of these cells from the GO phase and the progression through the cell cycle were hypothesized to originate in the mesenchyme, and this principal mesenchymal factor was FGF-7.
To test our hypothesis, a comparison of the urothelial response to obstruction between wild-type and FGF-7-null mice was performed. It was assumed that this response would involve individual urothelial cells undergoing proliferation. Accordingly, urothelial cells traversing the cell cycle were detected by immunostaining with polyclonal antibodies specific to the Ki-67 nuclear antigen, an accepted marker of epithelial cells progressing through the cell cycle.
Quantification of total and cycling urothelial cells was achieved by analysis using the principles of unbiased stereology (9, 58a, 59, 60). If mesenchymal FGF-7, produced and secreted by fibroblasts of the lamina propria, was the principal mitogen that stimulated the proliferation of urothelial cells, then lack of this major paracrine factor in FGF-7-null mice should influence the urothelial response to injury by exhibiting a measurable decrease in the percentage of proliferating urothelial cells, compared with wild-type mice.
The following transitional epithelia exhibited two to five layers of urothelial cells that typify the normal phenotype and were nonreactive with antibodies specific for the Ki-67 antigen: wild-type normal, FGF-7-null normal, wild-type sham, and FGF-7-null sham (data not shown). These results indicate that urothelial cells were not removed from the G0 resting phase of the cell cycle by surgical procedures that did not include urethral ligation.
Mice that underwent ligation for 1–2 days exhibited minimally detectable urothelial expansion (data not shown). Ligation for 4 days elicited subtle, but detectable, urothelial expansion (data not shown). The most consistent and reproducible efforts required 8 days of partial obstruction. Partially obstructed bladders after 8 days postligation were prominently larger and more distended than control animals, a result that indicated that partial outlet obstruction of the bladder was successfully accomplished by ligation of the urethra.
In contrast to control and sham-operated animals, transitional epithelia from animals that underwent partial urethral ligation at the bladder neck exhibited marked immunoreactivity with antibodies specific for the Ki-67 nuclear antigen. The total number of Ki-67-positive cells, as well as the total number of urothelial cells, was determined by stereology (Fig. 6). Calculation of mean values is presented in Table 1. The percentage of Ki-67-positive cells for obstructed wild-type and FGF-7-null mice was calculated to be 1.78 ± 0.46 and 1.60 ± 0.94 cells, respectively. A statistical comparison between these two groups with the Student's t-test indicated that this difference was not significant.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Mesenchymal mediators of epithelial cell behavior, proliferation, and differentiation represent an important means of regulating development, homoeostasis, and the response to injury. In the urinary bladder, the paracrine mediator that has been most studied is FGF-7, a polypeptide synthesized and secreted by fibroblasts of the lamina propria (2–6, 27). Our study demonstrates that this growth factor is a mitogen for the urothelial cells of human and murine transitional epithelium, a result in agreement with prior reports that studied the effect of FGF-7 on transitional epithelia from monkey (63), rat (57, 63), and murine (6, 53) sources.
Equivalent biological activities were observed for rFGF7-His (our preparations) and rFGF-7 (provided by Amgen Incorporated) as the pharmaceutical Palifermin in assays that monitored the incorporation of [3H]thymidine into the DNA of human urothelial cells in vitro. The mitogenic activity of rFGF7-His on human urothelial cells was observed only with confluent cultures, a phenomenon previously reported for normal human keratinocytes (21). Our data are also in agreement with prior reports (22, 44) that demonstrated that heparin was inhibitory to the stimulatory activity of rFGF-7 on Balb/MK keratinocytes.
Since the initial response of the bladder to outlet obstruction and to injury in general (7) is to turnover its urothelium via reciprocal mesenchymal-epithelial interactions (4, 7, 27), we asked to what extent mesenchymally derived FGF-7 regulates urothelial cell proliferation in response to obstruction. To answer this question, we quantified the proliferative response to obstruction by stereological analysis of Ki-67 staining of the urothelium. Not only were the total number of Ki-67-positive cells nearly equivalent between the wild-type and FGF-7-null sets of mice, but the difference of the percentage of urothelial cells traversing the cell cycle between these two sets was statistically insignificant. We noted that the urothelial response to obstruction was variable from animal to animal, an observation shared by other laboratories (4, 24, 31, 41, 48) and noted clinically as well.
The collective data prove our hypothesis that redundant mechanisms compensate for the lack of FGF-7 in FGF-7-null mice that continue to exhibit proliferating urothelial cells in response to obstruction. While we hypothesize that the principal mitogenic candidate to compensate for the absence of FGF-7 is FGF-10, a known mitogen for urothelial cells in vitro and in vivo (1, 65), our data do not experimentally address the biological activity of FGF-10 in this study. Since both FGF-7 and FGF-10 can bind to, interact with, and stimulate a mitogenic signaling pathway through the same FGFR2 isoform 2 receptor (28), we conclude that the FGFR2 isoform 2 receptor remains functional in FGF-7-null mice.
Our prior report that the FGFR2 isoform 2 receptor is expressed by the superficial layer of transitional epithelia (65) is clinically significant because intravesical instillation of recombinant FGF-7 and -10 into the bladder lumen is predicted to trigger urothelial repair in response to injury. In addition, clinical conditions involving the urinary tract which cause epithelial damage such as infections and trauma leading to stricture disease, erosive disorders such as interstitial cystitis, and urinary tract disruption from transurethral resection surgery could be treated with mitogenic stimulation. Ongoing work is expected to lead to a better understanding of the steady-state interrelationships that involve urothelial cells, growth factors, matricellular proteins, the extracellular matrix, and the urothelial basement membrane. Advancing the study of FGF-7 and -10 is relevant because of the polypeptides' potential use as a clinical tool to treat, and ultimately cure, a variety of lower urinary tract conditions and diseases.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
Indicates site of putative amidation (Gly147). * Indicates site of N-linked glycosylation (Asn16). 
Indicate sites of casein kinase II phosphorylation (Ser18 and Thr150). 
Indicates sites of phosphorylation by protein kinase C (Ser124). Underlined residues represent the FGF family signature (Gly96-Tyr119; Prosite Database). Bolded residues represent the glycine box (Asn139-Gly147). Black boxes with white letters designate residues implicated in the heparin-binding motif (Arg43, Asn117, Asn139, Gln140, Val145, Lys148, Asn154, Lys155, and Thr156).
