Molecular and Cellular Biochemistry 283: 107–114, 2006.DOI: 10.1007/s11010-006-2388-1 c�Springer 2006

Functional changes in bladder tissue from typeIII collagen-deficient mice

Karen Stevenson,1 Umberto Kucich,1 Catherine Whitbeck,2

Robert M. Levin2 and Pamela S. Howard1

1School of Dental Medicine, Department of Anatomy & Cell Biology, University of Pennsylvania, Philadelphia, PA, USA;2Albany College of Pharmacy, Albany, NY, USA

Received 28 July 2005; accepted 23 August 2005

Abstract

Objective: Collagen fibers impart tensile strength and transfer tension from bladder smooth muscle cells. We have previouslyshown that fibrotic bladders are characterized by an increased type III:type I collagen ratio. To determine the effect of decreasedtype III collagen on bladder function, type III collagen-deficient mice (COL3A1) were studied physiologically. Methods:Bladders from wild-type (+/+) and heterozygous (+/−) COL3A1 mice were biochemically characterized to determine totalcollagen (hydroxyproline analysis) and collagen subtype concentration (cyanogen bromide digestion and ELISA). Alterationsin collagen fiber diameter were assessed by electron microscopy. Bladder muscle strips were used to assess physiologic function.Results: Hydroxyproline content decreased in heterozygous bladders, which had 50% less type III collagen. Wild-type bladdershad a biphasic distribution of collagen fiber sizes, whereas heterozygous bladder collagen fibers spanned a broad range.Physiologically, there were no differences in contractile responses between wild-type and heterozygotes when stimulatedwith ATP, carbachol or KCl, indicating normal contraction via purinergic and muscarinic receptors, and in response to directmembrane depolarization. In contrast, tension generation in heterozygotes was decreased after field stimulation (FS), indicatingdecreased synaptic transmission. Length–tension studies showed that the heterozygote muscle strips generated less tension perunit length, indicating that they were more compliant than wild-type controls. Conclusions: Critical levels of type III collagenappear to be a requirement for normal bladder tension development and contraction. Our data show that a decrease in the typeIII:type I collagen ratio, and altered fiber size, results in a more compliant bladder with altered neurotransmitter function. (MolCell Biochem 283: 107–114, 2006)

Key words: bladder, COL3A1, compliance, neurotransmitters, type III collagen-deficient mice

Introduction

The interstitial collagens, types I and III, are structuralproteins that function in imparting tensile strength to nu-merous tissues and organs. Type III collagen is expresseddevelopmentally and during wound healing, and is asso-ciated with distensible organs and tissues. Studies haveshown that type III collagen may be co-expressed with type

Address for offprints: P. S. Howard, Department of Anatomy & Cell Biology, School of Dental Medicine, University of Pennsylvania, 240 South 40th Street,Philadelphia, PA 19104-6030, USA (E-mail: howard@biochem.dental.upenn.edu)

I collagen, forming heterotypic fibers, and may functionin regulating fibril diameter [1]. It is likely that changesin collagen fiber composition and diameter could resultin altered mechanical properties and functional parame-ters of the tissue or organ. This has been shown in astudy by Derwin and Soslowsky, whereby analysis of ten-dons from the type I collagen heterozygous (+/−) mutantmice (COL1A1-Mov13) showed that tendons with 50% less

108

type I collagen were characterized by altered mechanicalproperties [2].

Mutant mice have been generated by gene targetingin which the type III collagen gene (COL3A1) has beenknocked-out [3]. The knock-out animals exhibited irregu-larly sized collagen fibers in the dermis of the skin as wellas the aortic adventitia, indicating the importance of type IIIcollagen in regulating collagen fiber size and diameter. Thecollagen III null mice die perinatally, usually owing to bloodvessel or intestinal rupture. Hence, critical levels of type IIIcollagen appear to be necessary for proper organ and tissuefunction, especially in distensible organs.

Collagen types I and III are the major collagenous proteinsof the urinary bladder and are localized to: (1) the connectivetissue-rich lamina propria beneath the urothelial layer, (2) theperimysial regions surrounding the detrusor smooth musclefascicles, and (3) the serosal layer. The urinary bladder isan organ that slowly expands upon filling and rapidly con-tracts at micturition. As the bladder slowly fills, the coiledcollagen fibers in the lamina propria gradually expand as theintravesical pressure remains low, resulting in thinning of thelamina propria and vesical expansion [4–6]. Once volumecapacity has been reached, the collagen fibers between mus-cle fascicles become taut and intravesical pressure increasesrapidly and the micturition response is then initiated. Stud-ies from our laboratory have shown that there is a shift inthe type III:type I collagen ratio in obstructed human blad-ders, with an increase in type III collagen compared to typeI collagen [7]. As well, type III collagen has been shown tobe localized not only in a perimysial location, but also infil-trated within the smooth muscle fascicles in an endomysiallocation [8]. Therefore, the abnormal arrangement and con-tent of collagen fibers can alter organ function, resulting ina stiff-walled bladder with decreased functional complianceproperties.

In the present studies, heterozygous COL3A1 mutant micewere utilized to evaluate the role of decreased type III colla-gen in bladder function compared to wild-type control blad-ders. Bladder function can be characterized by the in vitroresponse to various forms of stimulation. Field stimulation(FS) mediates a contraction through the release of neuro-transmitters (acetylcholine and ATP) from synapses, diffu-sion across the synaptic cleft and activation of post-synapticmuscarinic cholinergic and purinergic receptors. Carbacholand ATP are used to directly activate muscarinic and puriner-gic receptors, respectively, resulting in contractions. KCl de-polarizes the cell membrane directly and mediates contrac-tility. Thus, FS requires functional pre-synaptic nerves andsynapses, post-synaptic receptors and contractile filaments.Carbachol and ATP require post-synaptic receptors and con-tractile filaments, whereas KCl only requires the contractilefilaments. We hypothesized that urinary bladders with lesstype III collagen would be more compliant due to altered

collagen fiber diameters and distribution, but contractile func-tion would not be affected.

Materials and methods

COL3A1 mice

Breeder pairs of COL3A1 heterozygous (+/−) mice werepurchased from Jackson Laboratories. The mice were orig-inally generated by homologous recombination with a dele-tion of the promoter region and first exon that encoded for thesignal peptide and replacement of this region with a 1.8-kbPGKneo cassette [3]. Mice were bred and care was main-tained in accordance with the guidelines of the InstitutionalAnimal Care and Use Committee (IACUC) of the Universityof Pennsylvania. Animals were genotyped by PCR analysisof DNA extracted from tail snips. At 8 weeks of age, animalswere sacrificed by CO2 inhalation. Animals were weighedand body weights recorded. The urinary bladders were mi-crodissected at the level of the bladder neck, any residualurine was expelled and the organs were weighed. Bladdersfrom both male and female mice were utilized in all assays.

Mouse genotyping

Mouse tail snips were digested in proteinase K overnightat 55 ◦C. DNA was amplified by a touchdown PCR pro-tocol utilizing the primer pairs, Neo primers OIMR158:5′-CTGAATGAACTGCAGGACCA-3′ and OIMR159: 5′-ATACTTTCTCGGCAGGAGCA-3′, generating a 172 bpfragment from the neo cassette in the knock-out al-lele, and COL3A1 exon 1 forward primer OIMR714: 5′-CTTCTCACCCTTCTTCATCCC-3′ and intron 1 reverseprimer OIMR715: 5′-AGCCTGTTCAAATCGGTACC-3′,generating a 79 bp fragment from the wild-type allele. Con-ditions of the cycling protocol were: denaturation at 94 ◦C for3 min, denaturation at 94 ◦C for 20 s, annealing at 64 ◦C for30 s, extension at 72 ◦C for 35 s. These parameters were uti-lized for 12 cycles, with the annealing temperature decreasedby 1 ◦C every two cycles. This was followed by 26 cycles ofdenaturation at 94 ◦C for 20 s, annealing at 58 ◦C for 30 s andextension at 72 ◦C for 35 s. After the final cycle, the reactionwas extended at 72 ◦C for 2 min and held at 4 ◦C. Sampleswere loaded onto a 2.5% agarose gel and electrophoresedat 58 V.

Electron microscopy

Following dissection, bladders were fixed in a solution of 5%glutaraldehyde, 4% paraformaldehyde in 0.1 M cacodylate

109

buffer, pH 7.4 containing 0.5 mg/ml CaCl2 and post-fixed in1% osmium tetroxide. Tissues were dehydrated in a gradedethanol series and embedded in polyBed 812 Epon. Thinsections were stained with uranyl acetate and lead citrate, andexamined in a JEOL CXII transmission electron microscope.Fiber diameters were determined by measurement of bothcross sectional and longitudinal sections of collagen fibersby image analysis, using NIH software.

Collagen biochemical analyses

Freshly dissected bladders (n = 18, 9+/+, 9+/−) werefrozen and stored at −70 ◦C. Collagen was extracted by treat-ment of pulverized tissue with cyanogen bromide (CNBr), aspreviously described [7, 9]. The CNBr-solubilized bladdertissue was analysed by a colorimetric microtiter plate assayto quantify hydroxyproline residues as a measure of totalcollagen content [10]. Soluble CNBr-peptides were acid hy-drolysed in 6N HCl at 108 ◦C for 16–24 h, neutralized with1 M NaOH, centrifuged, and the clear solutions were di-luted 1:4–1:20 in dilute acid and applied in 50 ul triplicatealiquots to microtiter wells. Oxidizing agent, chloramine-T(100 ul) was added to each well and mixed prior to the ad-dition of 100 ul Ehrlich’s reagent. Hydroxyproline standards(1–12 ug/ml) were treated similarly. Colour development wasallowed to proceed at 60 ◦C for 45 min. Absorbancy at 570 nmwas determined with an automatic plate reader (DynatechLaboratories). Unknown values were calculated directly fromthe hydroxyproline calibration curve. At least two differentassays for each sample were carried out with minimum vari-ance (<5%). The percentage of total collagen was then calcu-lated from the hydroxyproline concentration by multiplyingby a factor of 7.46 [11].

The CNBr-solubilized bladder tissue (n = 18, 7+/+,11+/−) was further analysed by enzyme linked immuno-sorbent assay (ELISA) to determine type I and type III colla-gen concentration, as previously described [7, 9]. Monospe-cific antibodies against the murine carboxyl telopeptide se-quence of α2(I) collagen and the amino telopeptide sequenceof α1(III) were generated in rabbits (Table 1) and utilized inthe ELISA.

Physiologic assays

Following dissection, bladders (n = 16, 8+/+, 8+/−) wereplaced in oxygenated, sterile Hank’s buffered saline on ice.The muscle strips were microdissected from the dome ofthe bladder and the urothelium/lamina propria was separatedfrom the underlying detrusor smooth muscle layer. One mus-cle strip/bladder was evaluated. Each strip was mounted in aseparate bath containing Tyrode’s solution (124.9 mM NaCl,

Table 1. Type-specific collagen amino acid sequences used to generateantibodies

Antibody Mouse Collagen Residue #identification collagen type peptide sequence (mouse sequence)

MC3 α1(III) Amino QFDSYDVC 155–161telopeptide

MC1 α2(I) Carboxy CYDFGFEGDFYRA 1114–1125telopeptide

Note: An italicized and underlined “C” indicates a cysteine, not in the pri-mary sequence that was added to the synthetic immunizing peptide to facil-itate coupling to KLH.

2.6 mM KCl, 23.8 mM NaHCO3, 0.5 mM MgCl2, 0.4 mMNaH2PO4, 1.8 mM CaCl2 and 5.5 mM dextrose) at 37 ◦C.The strips were equilibrated with a mixture of 95% O2 and5% CO2 for 30 min, and the end of each strip was connectedto a force displacement transducer. A Model D Polygraph(Grass Instruments, Quincy, MA) was used to record con-tractile responses. Signals were digitized using a PolyviewA/D computer analytical system.

Each strip was set to 0 tension. The strips were stretchedin 2 mm units and allowed to equilibrate until the tension be-came stable. The strip was then stimulated with 32 Hz FS andthe active tension recorded. The length that gave the maximalactive tension (100%) was used for all further contractile ex-periments. To complete the length–tension curves, the stripswere stretched to 200% and length–tension curves recorded.

Following equilibration for 30 min at the resting tensionthat gave maximal active tension, the responses to field stim-ulation at 2, 8 and 32 Hz (80 V square wave pulses for 1 msduration and 20 s trains) were recorded. Following field stim-ulation, the strips were allowed to rest for 30 min and thensequentially treated with 20 μM carbachol, 1 mM adenosinetriphosphate (ATP) and 120 mM KCl. Following each phar-macological treatment, the strips were washed three timeswith Tyrode’s solution for 15 min.

Data analysis

All values are reported as mean ± S.E.M. For hydroxyprolineand ELISA analyses, a Student’s t-test was used to comparedata of wild-type and heterozygotes. An analysis of variance(ANOVA) with Bonferonni’s post-test was used for com-parisons among groups in all physiologic analyses. p-Values<0.05 considered statistically significant.

Results

DNA from tail snips of 10-day-old pups was extracted usingproteinase K and the genotypes were determined by PCR and

110

Fig. 1. PCR analysis of the genotypes of COL3A1 mice. DNA from mouse tail snips was extracted and PCR amplified using primer pairs recognizing a 79 bpsequence of the wild-type (+/+) allele and a 172 bp neo sequence of the knock-out (−/−) allele. Agarose gel electrophoresis of amplified DNA from wild-type(+/+) and heterozygous (+/−) animals shows the presence of appropriate-sized DNA fragments for wild-type and knock-out alleles.

Fig. 2. Effect of COL3A1 gene manipulation on 8-week-old wild-type (WT) and heterozygous (HT) mouse body, bladder and strip weights. Each bar is themean ± S.E.M. of eight individual mice. There were no statistically significant differences between the groups.

agarose gel electrophoresis (Fig. 1). No homozygous mutantanimals survived in these litters. At 8 weeks of age, there wereno gross phenotypic differences between COL3A1 wild-typeand heterozygous animals. Since the animals were not housedin metabolic cages, micturition rates, volumes and frequencywere not assessed in this study. Body and bladder weightswere comparable between 8-week-old wild-type and mutant(+/−) animals (Fig. 2), with the bladder weights approxi-mately 1000-fold less than body weight.

Hydroxyproline analysis of CNBr-solubilized bladder tis-sue showed that the COL3A1 mutant heterozygotes contained14.2% total collagen compared to 16% total collagen in thewild-type bladders. Hence, total collagen in the mutant blad-ders was decreased by 11% (Table 2). Competitive ELISAassay showed that the wild-type bladders contained 80% typeI collagen and 20% type III collagen. In the mutant heterozy-gous bladders, the type III collagen decreased to 10.4% witha concomitant increase in type I collagen to 89.6% (Table 2).

Table 2. Total bladder collagen and subtype concentration

Mouse Hydroxyproline analysis ELISA ELISAgenotype (% total collagen) (% collagen I) (% collagen III)

Wild-type 16.32 ± 0.96 80.08 ± 2.24 19.92 ± 2.24

Heterozygote 14.17 ± 0.68∗ 89.61 ± 2.36∗ 10.39 ± 2.36∗

Values were compared using a Student’s t-test. ∗ p < 0.05, statisticallysignificant.

This 50% decrease in type III collagen reflects the expectedloss of type III collagen by the gene ablation.

Ultrastructural analysis of the collagen fibers in the blad-der walls revealed that there was a shift, or broadening of thecollagen fiber diameter in the mutant heterozygous bladders.The collagen fibers of the wild-type bladders were character-ized by a biphasic diameter pattern, from 20 to 34 nm fibers

111

Fig. 3. Mean collagen fiber diameters from wild-type (WT) and heterozy-gous (HET) COL3A1 bladders. Note the biphasic distribution of the colla-gen fibers in the wild-type bladders compared to the broad distribution ofthe fibers in the heterozygous bladders.

and from 40 to 49 nm. The mutant fibers ranged from 20to 59 nm in size, indicating a change in the pattern of fib-rillogenesis (Fig. 3). As well, there were increased areas ofspace separating collagen fiber bundles in the heterozygouslamina propria region (Fig. 4B), whereas the wild-type blad-ders showed abundant packing of the collagen bundles amonglamina propria fibroblasts (Fig. 4D).

Length–tension studies showed that the heterozygote blad-der tissue generated significantly less tension when stretchedto both 100 and 200% of optimal length, indicating that themutant bladders were less stiff, or more compliant, than thewild-type control bladder strips (Fig. 5).

Surprisingly, the mutant heterozygous bladders respondedto different frequencies of field stimulation by generating de-creased contractile force compared to wild-type control blad-ders (Fig. 6), indicating that bladders with 50% less type IIIcollagen have significantly altered neurotransmitter releaseor diffusion properties. Interestingly, there were no differ-ences in the responses of wild-type and heterozygous blad-ders to stimulation with ATP, carbachol and KCl, indicatingthat purinergic and muscarinic receptor stimulated contrac-tion and contractile responses to direct membrane depolar-ization, respectively, were normal (Fig. 6).

Discussion

Collagenous proteins are the major tension-bearing structuralelements in the urinary bladder. These proteins and their ar-rangement within the extracellular matrix mediate bladderfunction by allowing expansion of the bladder wall at lowpressures upon filling and, at the same time, preventing overexpansion at volume capacity. It has been previously shownthat collagen fibers can be heterotypic, i.e. fibers can be anadmixture of different collagen types, and that within theseheterotypic fibers, the presence of molecules of one specificcollagen subtype can control the overall diameter and size ofthe heterotypic fiber [1]. If the molecular composition of the

heterotypic fibers is different, then the functional, mechani-cal properties may also be altered. By analogy, metal alloyscontaining different ratios of constituent elements have dif-ferent tensile strengths and different mechanical properties.The study of genetically manipulated mice is a tool usefulfor relating changes in functional properties of tissues andorgans with alterations in their molecular composition.

Using the COL3A1 heterozygous and wild-type mice, wesought to determine whether alteration of the ratio of type III:Icollagen could affect the compliance properties of the blad-der. Two-month-old COL3A1 heterozygous (+/−) mice hadbody and bladder weights comparable to those of wild-type(+/+) control mice. Biochemically, COL3A1 heterozygousbladders were characterized as having 11% less total colla-gen, and 50% less type III collagen compared to wild-typecontrol bladders. This reduction in type III collagen contentresulted in resistance to deformation, i.e. increased compli-ance, as determined by length–tension functional analyses.

Phenotypically, the 50% decrease in type III collagen re-sulted in collagen fibers that had a broader distribution offiber diameters. Fibers ranged from 20 to 59 nm in diam-eter, compared to a biphasic distribution of 20–34 nm and40–49 nm in the wild-type bladders. As well, the packing ofthe collagen fibers varied in the heterozygous bladders com-pared to wild-type controls, with increased spacing amongfiber bundles in the heterozygous lamina propria comparedto highly packed bundles between fibroblasts in the laminapropria of the wild-type bladders. This alteration in pack-ing and size may be related to the functional alterations, inthat the length–tension studies showed that the heterozygousbladders with decreased type III collagen and altered fibersize were more compliant than the wild-type bladders. Thesefindings of changes in compliance are in agreement with de-velopmental studies reported by Baskin et al. [12]. Theseinvestigators showed that during fetal bladder development,type III collagen content decreased during the third trimesterwith a concomitant increase in bladder compliance. Hence,less type III collagen results in a more compliant bladderwall. However, it remains to be determined how this reduc-tion actually mediates an increase in compliance.

The complexity of composition of heterotypic collagenfibers, the arrangement of the different molecules within thefibers and their relationship to functional tissue properties un-derscores the importance of both quantitative and qualitativeassessment of structure–function studies. It has been shownthat heterotypic fibers in the dermis of the skin are composedof a core of type I collagen molecules coated with an exteriormolecular layer of type III collagen [1]. In contrast, in thecorneal stroma, where the collagen fibers form an orthogonalarray that results in corneal transparency, the type I collagenfibers have an interior core of type V collagen molecules,which helps to regulate the heterotypic fiber diameter [13].Alterations in the amount and type of collagens (and other

112

Fig. 4. Electron micrograph of the lamina propria of the COL3A1 heterozygous (+/−) murine bladders (A, B) and wild-type (+/+) bladders (C, D). Note theincreased spacing among the fiber bundles in the heterozygous bladders (B) compared to wild-type (D). Bar (A, C): 1 μm; Bar (B, D): 500 nm.

Fig. 5. Effect of COL3A1 gene manipulation on compliance properties ofwild-type (diagonal-hatched bars) and heterozygous (cross-hatched bars)murine bladders. Length–tension measurements were determined and ex-pressed as tension in grams as a function of percent of optimum length.Each bar is the mean ± S.E.M. of eight bladder strips. ∗, significantly dif-ferent from wild-type. p < 0.05.

matrix proteins, such as beta-ig-h3) in the corneal stroma canlead to a less than optimal arrangement of fiber organiza-tion and increased corneal opacity, resulting in compromisedvision.

Physiologically, at 2 months of age, contractile responsesto ATP, carbachol and KCl were unchanged in heterozygous(+/−) bladders compared to wild-type (+/+), indicating nor-mal contraction via purinergic and muscarinic receptors andin response to direct membrane depolarization, respectively.However, the bladders with 50% less type III collagen showedsignificantly decreased force generation in response to elec-tric field stimulation. These data suggest that COL3A1 het-erozygous mice may have either reduced neurotransmitterrelease from the nerve terminals, or normal neurotransmitterrelease but decreased response as a consequence of an in-crease in diffusion distance for the neurotransmitters to travelto their cognate receptors.

Disruption of neurotransmitter release has been docu-mented in the urinary bladders of caveolin-1 (cav-1) knock-out mice [14]. Caveolin-1 is an integral membrane pro-tein found in invaginating membrane cholesterol-rich mi-crodomains in detrusor smooth muscle cells of the urinary

113

Fig. 6. Contractile responses of COL3A1 wild-type (diagonal-hatched bars) and heterozygous (cross-hatched bars) bladder strips to 2, 8 and 32 Hz electricfield stimulation, 1 mM ATP, 20 μM carbachol and 120 mM KCl: maximal contractile response. Each bar is the mean ± S.E.M. of eight bladder strips. ∗,significantly different from wild-type. p < 0.05.

bladder. These proteins are important regulators of cellularsignal transduction pathways. These investigators showedthat bladders, in which cav-1 was knocked-out, were char-acterized by a decreased contractile response to electric fieldstimulation, owing to a 70% reduction in the release of acetyl-choline from nerve terminals. As well, post-junctional mus-carinic receptor, M3, signaling was also impaired. Exper-iments in our study did not discriminate between releaseof muscarinic cholinergic and purinergic neurotransmitters.However, the post-junctional muscarinic and purinergic re-ceptors evoked normal contractile responses when stimu-lated with carbachol and ATP, respectively in the 8-week-oldCOL3A1 heterozygous bladders. Future studies will deter-mine whether specific neurotransmitters are involved, as wellas possible mechanisms for release impairment.

The diffusion distance between pre- and post-synapticjunctions in skeletal muscle is approximately 50-nm wide[15]. In the bladder, the synaptic distance between pre-junctional neuroeffectors and the detrusor smooth muscle cellmembrane can range between 10 and 100 nm, with the av-erage of approximately 30–50 nm [16]. While diffusion dis-tance may be responsible for altered neurotransmitter func-tion, the chemical composition of the pre- and post-synapticbasal laminas may also play a role in both normal and aberrantneurotransmitter signaling. The synaptic region (in skeletalmuscle) is composed of a basal lamina consisting of typeIV collagen, laminins, proteoglycans and agrin. The pre-synaptic basal laminas associated with efferent nerves andpost-synaptic basal laminas of the detrusor smooth musclecells in the urinary bladder have not been chemically charac-

terized. A recent study of neuroeffector junctions in rat uri-nary bladder showed that a specific splice form of agrin, (z-),is localized at neuroeffector junctions as well as associatedwith detrusor smooth muscle cells [17]. There is no informa-tion in the literature concerning type III collagen in relationto nerve terminal synaptic regions or control of neurotrans-mitter release. However, one study has shown that inhibitionof collagen synthesis decreased acetylcholine receptor clus-tering in muscle cells [18]. As well, other studies have shownthat the extracellular matrix protein, fibronectin, may play arole in regulating transmitter release in embryonic Xenopusneuromuscular synapses [19, 20]. Future studies will exam-ine the questions of altered diffusion distances and matrixcomposition of the smooth muscle neuromuscular junctionalcomplexes in COL3A1 mutant mice.

Conclusion

Critical levels of type III collagen appear to be a requirementfor normal bladder tension development and contraction. Adecrease in the type III:type I collagen ratio, and altered fibersize distribution, results in a more compliant bladder withdecreased neurotransmitter function.

Acknowledgments

The authors thank Ms. Sylvia Decker for her assistance withthe electron microscopy and Mr. Zuozhen Tian for help with

114

the tables and figures. This study was supported by NationalInstitutes of Health Grants DK-48215 (PSH) and DK-067114(RML).

References

1. Fleischmajer R, MacDonald ED, Perlish JS, Burgeson RE, Fisher LW:Dermal collagen fibrils are hybrids of type I and type III collagenmolecules. J Struct Biol 105: 162–169, 1990

2. Derwin KA, Soslowsky LJ: A quantitative investigation of structure–function relationships in a tendon fascicle model. J Biomech Eng 121:598–604, 1999

3. Liu X, We H, Byrne M, Krane S, Jaenisch R: Type III collagen iscrucial for collagen I fibrillogenesis and for normal cardiovascular de-velopment. Proc Natl Acad Sci USA 94: 1852–1856, 1997

4. Chang SL, Chung JS, Yeung MK, Howard PS, Macarak EJ: Roles ofthe lamina propria and the detrusor in tension transfer during bladderfilling. Scand J Urol Nephrol 201S: 38–45, 1999

5. Chang SL, Howard PS, Koo HP, Macarak EJ: Role of type III collagenin bladder filling. Neurol Urodynam 17: 135–145, 1998

6. Macarak EJ, Howard PS: The collagens and their urologic significance.Scand J Urol Nephrol 184S: 25–33, 1997

7. Deveaud C, Macarak EJ, Kucich U, Ewalt DH, Abrams WR, HowardPS: Molecular analysis of collagen in non-compliant bladders. J Urol160: 1518–1527, 1998

8. Ewalt DH, Howard PS, Blyth B, Snyder H, Duckett J, Levin R, MacarakE: Is lamina propria matrix responsible for normal bladder compliance?J Urol 148: 544–549, 1992

9. Howard PS, Renfrow D, Schechter NM, Kucich U: Mast cell chymase isa possible mediator of neurogenic bladder fibrosis. Neurourol Urodynam23: 374–382, 2004

10. Brown S, Worsfold M, Sharp C: Microplate assay for the measurementof hydroxyproline in acid-hydrolyzed tissue samples. Biotechniques 30:38–42, 2001

11. Neuman R, Logan M: The determination of collagen and elastin intissues. J Biol Chem 186: 549–556, 1950

12. Baskin LS, Constantinescu S, Duckett JW, Snyder HM, Macarak E:Type III collagen decreases in normal fetal bovine bladder development.J Urol 152: 688–691, 1994

13. Linsenmayer TF, Fitch JM, Gross J, Mayne R: Are collagen fibrils inthe developing avian cornea composed of two different collagen types?Evidence from monoclonal antibody studies. Ann NY Acad Sci 460:232–245, 1985

14. Lai HH, Boone TB, Yang G, Smith CP, Kiss S, Thompson TC, Somo-gyi GT: Loss of caveolin-1 expression is associated with disruption ofmuscarinic cholinergic activities in the urinary bladder. Neurochem Int45: 1185–1193, 2004

15. Sanes J: The basement membrane/basal lamina of skeletal muscle. JBiol Chem 278: 12601–12604, 2003

16. Gabella G: Cells and cell junctions in the muscle coat of the bladder.Scand J Urol Nephrol 184: 3–6, 1997

17. Gingras J, Spicer J, Altares M, Zhu Q, Kuchel GA, Ferns M: Agrinbecomes concentrated at neuroeffector junctions in developing rodenturinary bladder. Cell Tissue Res 320: 115–125, 2005

18. Bixby JL: Collagen synthesis inhibition reduces clustering of heparinsulfate proteoglycan and acetylcholine receptors but not agrin or p65,at neuromuscular contacts in vitro. J Neurobiol 26: 262–272, 1995

19. Fu W-M, Liou H-H, Wang C-L: Collaboration of fibronectin matrix andneurotrophin in regulating spontaneous transmitter release at developingneuromuscular synapses in Xenopus cell cultures. Neurosci Lett 300:115–119, 2001

20. Liou H-H, Lin W, Liou H-C, Huang T-F, Fu W-M: Modulation of proteinkinase A activation by fibronectin matrix proteins at developing neuro-muscular synapses in Xenopus laevis cell cultures. Mol Pharmacol 60:348–354, 2001