To immunohistochemically identify the protein composition of adherens junctions, which couple smooth muscle cells mechanically, and to confirm their decrease in different bladder dysfunctions, as studies in geriatric bladder dysfunction show fewer such junctions in patients with detrusor overactivity (DO) and bladder outlet obstruction (BOO).
MATERIAL AND METHODS
Detrusor biopsies were obtained from video-urodynamically evaluated patients with neurogenic DO (NDO, 31 patients), BOO (six patients) and from six patients with stress urinary incontinence (SUI) with stable, unobstructed detrusors (serving as controls). Specimens were fixed, paraffin-embedded, sectioned, stained with a polyclonal pan-cadherin antibody, monoclonal α-, β- and γ-catenin antibodies, and a monoclonal integrin-β1 antibody. All antibodies were known to react with proteins of adherens junctions. Two examiners unaware of sample origin evaluated the sections qualitatively and using a semiquantitative scale. The results were correlated with the patient groups.
Specific immunohistochemical staining of pan-cadherin, α-, β- and γ-catenin could not be detected in any detrusor smooth muscle compartment, but was present in the urothelium. There was integrin-β1 reactivity in the basement membranes of bladder smooth muscle cells in 38 of 43 detrusor biopsies. There were no differences among the three groups.
The known proteins of cell-cell adherens junctions are not part of the cell-cell junctions of detrusor smooth muscle. The specific staining of integrin-β1 indicates either the presence of cell-matrix junctions or of cell-cell junctions within the human detrusor. Further studies are needed to identify the complete protein composition of adherens junctions within smooth muscle cells.
A key step in the evolution of multicellularity was the ability of cells to contact and interact specifically with other cells. Various integral membrane proteins, collectively termed cell-adhesion molecules, enable many cells to adhere tightly and specifically with cells of the same, or similar type. These interactions allow populations of cells to segregate into distinct tissues .
In many tissues mechanical cell coupling is associated with adherens junctions, which mediate the adhesion of cells to their neighbours (cell-cell junctions) or to the extracellular matrix (cell-matrix junctions). Biochemically these are specialized regions of plasma membranes formed by specific transmembrane adhesion molecules, cytoskeletal filaments and submembrane plaques, composed of various proteins, which interconnect the two .
In epithelial models these junctions are composed of proteins, some of which are preferentially associated with cell-cell junctions (e.g. cadherin and various catenins) or with cell-matrix junctions (e.g. integrins and talin); others are ubiquitously located in both types of adherens junctions (e.g. vinculin and α-actinin) [1,3–5].
As noted, the protein composition of adherens junctions has mainly been evaluated in in vitro studies for epithelial cells. No studies have been published of the complete protein composition of adherens junctions in smooth muscle tissues, especially the detrusor. Several research groups have evaluated the role of adherens junctions in bladder dysfunction; Elbadawi  showed ultrastructurally that adherens cell-cell junctions, introduced as intermediate cell junctions, were the predominating junction in normal detrusor. In patients with geriatric and neurogenic detrusor overactivity (NDO) or BOO there are fewer of these junctions [7–9].
The protein composition of adherens junctions within the human detrusor was evaluated partly by Wilson et al., who detected integrins between adjacent muscle cells of normal human detrusor using immunohistochemistry. Carey et al. immunohistochemically stained for vinculin, and reported that virtually the entire cell membrane of detrusor smooth muscle fibres was occupied by adherens junctions in classic and rudimentary form in samples from women with DO and from controls.
As Carey et al. only stained for vinculin, a protein ubiquitous in adherens junctions, they were unable to discriminate between cell-cell and cell-matrix junctions in their study. The study of Wilson et al. is limited because only cell-matrix protein components and only normal human detrusor tissues were evaluated.
The aim of the present study was to characterize the protein composition of cell-cell and cell-matrix adherens junctions in the human detrusor, on the basis of proteins known from epithelial junction models. In addition, changes of these adherens junctions in different bladder dysfunctions were evaluated.
MATERIALS AND METHODS
This study used material from 14 females and 17 males (mean age 22 years, range 7–65) with urodynamically confirmed neurogenic bladder dysfunction and DO caused by meningomyelocele in 15 or spinal cord injury in 16. Six men with urodynamically confirmed BOO from benign prostatic enlargement (mean age 67.5 years, range 55–78) and six women (mean age 51 years, range 38–63) with stress urinary incontinence (SUI) and urodynamically confirmed stable unobstructed detrusor served as controls.
Detrusor biopsies were obtained during various scheduled open operations, with the written informed consent of all patients. Open biopsy from the lateral bladder wall was excised by scalpel to obviate tissue destruction and artefacts of electrocautery. Each biopsy was placed immediately in fixative (4% formaldehyde). Standardized specimen processing included trimming of each specimen, dehydration in ascending concentrations of ethanol and in xylol and embedding in paraffin-wax.
Paraffin-embedded tissue samples were cut at 3 µm, placed on slides and dried for 24 h at 37°C. Sections were dewaxed, rehydrated using xylol and descending series of ethanol, and immersed in 3% H2O2 for 10 min to block endogenous peroxidase. After washing with Tris buffered saline (50 mm Tris, 150 mm NaCl pH 7.6) and blockage of unspecific antibody-binding sites using swine serum, the sections were incubated overnight at 4°C with the primary antibodies, a polyclonal antihuman pan-cadherin rabbit antibody (Zytomed, Berlin, Germany) at a dilution of 1 : 4000, a monoclonal antihuman α-catenin mouse antibody at a dilution of 1 : 50, a monoclonal antihuman β-catenin mouse antibody at a dilution of 1 : 2000, a monoclonal antihuman γ-catenin mouse antibody (all Zytomed) at a dilution of 1 : 80 and a monoclonal antihuman integrin-β1 mouse antibody (Abcam, Cambridge, Great Britain) at a dilution of 1 : 200. The polyclonal pan-cadherin antibody was used to detect all variants of the cadherin protein. After another buffer washing, sections were immersed in a 1 : 5 dilution of a biotinylated antibody to mouse and rabbit IgG (LSAB-Plus-Kit Link, DAKO, Carpinteria, CA, USA) for 30 min at room temperature, followed by an incubation with a 1 : 5 dilution of streptavidin peroxidase complex reagent (LSAB-Plus-Kit), again for 30 min at room temperature. The sections were exposed to 3-amino-9-ethylcarbazole solution, the chromogen for peroxidase reaction, for 15 min. Muscle cell nuclei were counterstained by immersing the section in haemalaun for 20 s. The sections were thoroughly washed, cover-slipped and analysed by light microscopy (Olympus BX 41, Hamburg, Germany) at magnifications of × 100 and × 400.
Tissue samples from human heart muscle were used as positive controls for pan-cadherin, α-, β- and γ-catenin. A human spleen tissue sample was used as positive control for integrin-β1. Of 43 detrusor biopsies, 31 contained urothelium and detrusor smooth muscle cells. The urothelium served as an additional internal positive control. As negative controls, sections were incubated with nonimmune serum instead of the primary antibody, followed by the detection method as described above.
All sections were processed at the same time to ensure the chromogen reactions were comparable. At least 10 visual fields per detrusor biopsy were evaluated on the cross sections for the presence of all four proteins. Biopsies were examined in random order by two examiners unaware of the clinical and urodynamic data, using a semiquantative scale, i.e. −, no staining; +, limited staining; and ++, intense staining. The results were correlated with the patient groups.
According to the staining results of Wilson et al. we focused on the immunohistochemical staining of the smooth muscle layer with its basement membrane, the cytoplasm and surrounding connective tissue, and the urothelium with its basement membrane and cytoplasm.
No immune reactions were detected in the negative controls (Fig. 1A–5A) but there was intense specific staining in all positive controls (Fig. 1B–5B). Pan-cadherin and the three catenins could not be detected in any detrusor smooth muscle compartment, whether basement membrane, the cytoplasm of smooth muscle cells or the surrounding connective tissue. In contrast, they were detected in the urothelium of the 31 biopsies, if present. In these biopsies the urothelium was intensely reactive, with visible accentuation of the basement membrane; it therefore served as internal positive control (Fig. 2C–5C).
There was specific integrin-β1 immunohistochemical staining of the muscle cell layer in 38 of 43 detrusor biopsies and in positive controls. In all detrusor samples the immunoreactions for integrin-β1 were confined to the cytoplasmic membranes of smooth muscle cells and urothelium. They had a dense punctuate pattern that virtually occupied the entire cell membrane of detrusor smooth muscle cells (Fig. 5C).
There were no differences in detrusor staining of integrin-β1 among the various patient groups. As there was no specific staining for pan-cadherin and the catenins, a comparison among patient groups was not possible.
In the present study we focused on the protein composition of adherens junctions within the smooth muscle cells of the human detrusor, detecting integrin-β1 immunohistochemically in 38 of 43 human detrusor biopsies. In contrast, there was no staining of pan-cadherin and the three catenins in any smooth muscle compartment. There were no detectable differences among the patient groups of NDO, BOO and SUI.
These results, especially the lack of immunohistochemical evidence of cell-cell adherens junctions, normally represented by pan-cadherin and the three catenins, were unexpected, as Elbadawi et al. reported the presence of these junctions ultrastructurally.
Theoretically, the immunohistochemical method can be the reason why there is no evidence of cell-cell adherens junctions within the human detrusor. We think this possibility can be excluded for the following reasons. In addition to our group, other research groups have also reported the utility of immunohistochemistry for staining proteins associated with adherens junctions [10,11]. Immunohistochemistry is broadly used as a diagnostic tool because of its high sensitivity and specificity. The present staining method, the labelled streptavidin-biotin (LSAB) method, is one of the most specific and preferred methods of immunohistochemistry . In direct comparison with other immunohistochemical staining methods the LSAB method exceeds the peroxidase antiperoxidase and avidin-biotin peroxidase complex methods in both sensitivity and detection efficiency .
Besides the immunohistochemical staining method, the frequency of proteins in the detrusor biopsies is essential for their detection. In former ultrastructural studies of various neurogenic dysfunctions of the human detrusor, the frequency of cell-cell adherens junctions was 0–9 per 100 muscle cells . This frequency is adequate for immunohistochemical detection. A similar frequency of axon terminals and preterminals within the human detrusor was easily detected by staining using the LSAB method with a protein gene-product 9.5 antibody .
Sections were stained using polyclonal (pan-cadherin) and monoclonal (catenins and integrin-β1) antibodies. This offered the possibility of very specific staining of protein using the monoclonal antibody and the possibility to stain all variants of a protein using a polyclonal cross-reactive antibody, resulting in the maximum possible detection rate.
In addition, we can exclude technical errors during fixation, embedding and staining, as no immunoreactions were detected in the negative controls and all positive controls had intense specific staining. The internal positive control, the urothelium, present in 31 of 43 detrusor biopsies, was also intensely and specifically reactive.
Another more reasonable possibility to explain the immunohistochemical absence of cell-cell adherens junctions in the human detrusor are the theoretical protein models on which the junctional structures are based. Most studies of the protein composition of cell-cell and cell-matrix junctions were evaluated in epithelial cells, because of the importance and abundance of these junctions in epithelium . The most extensively studied example of adherens junctions are the zonula adherentes of epithelia . Models depicting the organization of the main structural proteins of adherens junctions were developed in cell-culture studies, with the authors stating that the exact position of the various proteins and the spatial relationships between them are largely hypothetical, and the intermolecular interactions shown are based only on in vitro binding data . Therefore it is possible that a different protein composition is present in other tissues than epithelium (e.g. detrusor) or in in vivo studies.
Another possibility to explain the immunohistochemical presence of integrin-β1 and the lack of pan-cadherin and the catenins in the human detrusor might be the participation of integrin-β1 in cell-cell adherens junctions. Integrin-β1, detected immunohistochemically in 38 of the 43 human detrusor biopsies, is considered the classical protein of cell-matrix junctions [15,16], although there are references indicating a possible role of integrins in cell-cell junctions. Stappert and Kemler  described integrins as the major receptors by which cells attach to the extracellular matrices, but some integrins also mediate important cell-cell adhesion events. Hynes  stated that integrins participate in cell-matrix and cell-cell adhesion in many physiologically important processes, including embryological development, haemostasis, thrombosis, wound healing, immune and nonimmune defence mechanisms, and oncogenic transformation. For example, integrin-β2 mediates cell-cell adhesion in leukocyte adhesion. In addition, Danen and Sonnenberg  stated that integrin-β2 binds mainly to Ig-type counter-receptors, e.g. intercellular adhesion molecules (ICAMs).
Robinson et al. evaluated integrin-β1 and showed that it can mediate strong intercellular adhesions when cells are grown as three-dimensional aggregates in cell cultures. This cohesion was independent of cadherin expression and was significantly greater than the cohesiveness of N-cadherin, a more traditional cell-cell cohesion system. It seems therefore possible that integrin-β1 represents the transmembrane part of the protein composition of cell-cell adherens junctions within the human detrusor in addition to its classical role in cell-matrix junctions.
In conclusion, the present results suggest differences in the protein composition of adherens junctions between the human detrusor and epithelial models thought to assemble cell-cell adherens junctions. We were unable to detect cadherin and the three catenins, but integrin-β1 seems to be important in mechanical cell coupling, either between adjacent smooth muscle cells or in cell-matrix junctions of the human detrusor. Further studies with other diagnostic tools (Western blot, immunogold electron microscopy) and antibodies are necessary to differentiate the role of integrin-β1 between cell-cell or cell-matrix junctions, and to identify the complete protein composition of adherens junctions within smooth muscle cells of the human detrusor. This is a basic requirement before changes in the frequency of these cell-cell adherens junctions and their functional role can be evaluated in association with different bladder dysfunctions. So far, changes in adherens junctions of the human detrusor have no clinical consequences, but it seems possible that an increase in adherens junctions may change bladder function.