Impedance measurements and connexin expression in human detrusor muscle from stable and unstable bladders



Three of this month's Scientific Discovery papers highlight the importance of collaboration in delivering high quality scientific research. As scientific technology increases in power and cost, and specific areas of interest become more specialized, it is becoming more difficult to cover all aspects of a completeresearch story. Collaborating with other experts in the field or other fields, including industry, allows strong scientific proof to be generated for the hypothesis and aims. Building strong collaborative,inter-disciplinary, multi-institutional, international groups with academic and industrial partners is the way forward for all discovery. We look forward to publishing more of these collaborative papersin future issues of the BJU International.


To test the hypothesis that intercellular electrical coupling is altered in human detrusor smooth muscle from patients with unstable bladders.


Human detrusor biopsy samples were obtained from patients with stable and unstable bladders. Intracellular electrical impedance was measured with alternating current (20 Hz−300 kHz) across the ends of detrusor strips in an oil-gap, after correcting for extracellular space resistance. Gap junctions were identified by localization of connexins (Cx), specifically Cx45, Cx43 and Cx40 transcripts, using immunoconfocal microscopy.


Total intracellular resistivity was greater in strips from unstable than from stable bladders (median 1246 vs 817 Ω.cm). The increase was attributed to an increase in junctional resistance; cytoplasmic resistance was unchanged. Cx43 was localized to a submucosal layer and to connective tissue; Cx40 label was confined to endothelial cells of blood vessels. Cx45 labelling was localized to detrusor bundles and appeared to be less marked in samples from unstable bladders. Semi-quantitative analysis of Northern blots showed that Cx45 expression in unstable was less than that in stable bladders.


These data suggest that intercellular coupling is reduced in detrusor from unstable bladders. Cx45 was localized to the detrusor layer, with Cx 43 more evident in the suburothelial mucosa. Cx45 labelling was less intense in detrusor samples from unstable bladders. These results are consistent with reduced gap junction coupling in detrusor from unstable bladders.


Unstable contractions of the bladder are thought to contribute to the symptoms of the overactive bladder and there has been considerable debate as to whether unstable contractions of the bladder originate within the smooth muscle or are generated in response to neuronal or other humoral influences. A myogenic cause of bladder instability implies a fundamental change to the properties of detrusor smooth muscle. Proposed mechanisms include supersensitivity to neuromuscular transmitters [1], the appearance of novel transmitters [2] or enhanced spontaneous activity [3]. Morphological changes between normal and unstable human bladder, including the appearance of ‘protrusion junctions’ between cells, were shown in electron microscopy studies [4,5], leading to speculation that electrical conductivity may be enhanced in the unstable detrusor. If such junctions represent enhanced intercellular electrical coupling these may coordinate electrical activity more efficiently and possibly generate greater spontaneous activity. However, only gap junctions can mediate electrical coupling and there is sparse morphological evidence for these junctions in human detrusor.

The role of electrophysiological changes in generating contractions in human detrusor remains equivocal as the primary neurotransmitter acetylcholine acts independently of changes to membrane potential [6]. However, the appearance of ATP as a secondary transmitter in the unstable human bladder [7], which depolarizes the detrusor cell [6], and the presence of T-type Ca2+ channels in guinea-pig detrusor which may underlie oscillatory, subthreshold changes to membrane potential [8], indicate that electrical currents occur in this tissue. A further question arises as to whether contiguous detrusor smooth muscle cells are electrically connected.

Microelectrode measurements using small current sources found little evidence of coupling [9]; however, using larger currents, significant intercellular current coupling was measured [10,11], at least in guinea-pig detrusor. Whether increased electrical coupling occurs in detrusor from unstable bladders and contributes to mechanical overactivity has not been quantitatively addressed. In obstructed guinea-pig detrusor, measurements of intercellular current flow suggest that there is indeed a diminution of electrical coupling [11].

In the present study we aimed to investigate changes in the electrical conductivity between normal and human detrusor muscle, and how this relates to the presence and identity of gap-junctional connexins in these tissue samples. We measured the electrical resistance of the intercellular pathway, in particular the resistance between adjacent cells, in samples of human detrusor from stable and unstable bladders. We correlated these electrical measurements with the location of connexins, the protein components of gap junctions, in these detrusor samples. Connexin43 is the principal gap-junctional connexin in well-coupled cells such as myocardium, but other isotypes are also expressed. In seeking to identify the key connexin(s) of detrusor, we therefore considered existing findings that in both cardiac and smooth muscle, connexin40 and connexin45 may be expressed with or instead of connexin43 [12–15]. Gap junction channels made from different connexin types show distinct biophysical properties, including unitary conductance values, facilitating the fine-tuning of intercellular communication requirements to the precise cellular setting [16].


Human bladder biopsy samples were obtained at open surgery, placed in Ca-free HEPES buffer and used immediately for experiments. Local ethical committee approval and patient consent were obtained for each specimen. For immunoconfocal microscopy of cryosections, full-thickness specimens (3–5 mm2) were placed in a mould containing Cryo-M-Bed (Bright Instrument Co. Ltd, UK) and frozen directly in liquid nitrogen. For electrophysiological measurements and Northern blotting, mucosa was dissected from detrusor in Ca-free HEPES buffer. Samples of both portions were frozen for Northern blots and detrusor strips cut for impedance measurements. Samples were classified into two categories: (i) stable, i.e. from patients with bladder cancer and no symptomatic evidence of detrusor instability, taken distant from the malignant sites; (ii) patients with urodynamically confirmed idiopathic detrusor instability. Three samples were also obtained from patients with hyper-reflexia and their data are reported separately.

Impedance was measured at 37°C in Tyrode's solution (mmol/L): NaCl, 118; KCl, 4.0; NaHCO3, 24; NaH2PO4, 0.4; MgCl2, 1.0; CaCl2, 1.8; glucose, 6.0; Na pyruvate, 5.0; pH 7.4 with 5% CO2/95% O2. Tissues were transported in a Ca-free HEPES-buffered solution with NaHCO3 replaced by 10 mmol/L HEPES and 14 mmol/L NaCl, pH 7.1 with NaOH.

For impedance measurements, a three-compartment oil-gap chamber was used, previously validated with guinea-pig detrusor muscle and myocardium [10,17]. A detrusor strip (diameter ≈ 1 mm, length 4–7 mm) was pulled through rubber partitions separating the compartments, with at least 1 mm in the outer chambers containing Tyrode solution; the middle chamber was filled with mineral oil. Constant amplitude alternating current (frequency, f 20 Hz−300 kHz; ω = 2 πf) flowed between platinum-black electrodes in the outer chambers and thus was constrained to flow via the preparation intracellular space, with a proportion through an extracellular shunt. Resistance, r, and capacitance, c, of the system were recorded with a balanced Wien bridge (Wayne-Kerr 6425, UK) assuming a parallel rc configuration [18]. Two sets of recordings were made 10 min apart; values were always within 5% and the average used. Platinum-black electrode resistance, re, and capacitance, ce were measured separately in a large volume of Tyrode solution. The resistivity of Tyrode solution (RT, 49 ± 1 Ω.cm; 37°C, n = 3) was measured in a conductivity cell (length 1.0 cm, cross-sectional area 0.070 cm2).


Electrode polarization resistance, rp, and capacitance, cp, were calculated from re and ce[10,19]:


Preparation impedance, zs, was initially calculated as resistive, rs, and reactive, xs, components inline image, using r and c values and electrode polarization properties:


Extracellular shunt correction, rec, to yield detrusor impedance zd, was obtained from the relation:


Detrusor impedance, zd, was finally separated into resistive, rd (= zs.cosφ) and reactive, xd (=zs.sinφ) components, where φ= tan−1(xs/rs). Values of zd, rd and xd (Ω.cm−1) were converted to specific values Z, Rd and Xd (Ω.cm) on multiplication by muscle cross-sectional area.

Measurement of extracellular shunt resistance. An electrical model of the preparation assumed an extracellular shunt, rec, in parallel with the detrusor impedance. Two needle platinum-black electrodes were placed in the preparation within the oil-gap at different distances apart; rec (Ω.cm−1) was determined from the slope of the resistance vs distance plot. The equivalent cross-sectional area of strip contributing to rec, assuming it was filled with Tyrode soultion, is RT/rec (Ω.cm/(cm−1) = cm2) and is expressed as a proportion of total preparation cross-sectional area.

To detect connexin43 by immunoconfocal microscopy a commercially available mouse monoclonal antibody of proven specificity was used [12,20] (Chemicon International Ltd, UK; dilution 1 : 1000). Connexin40 and connexin45 were detected with polyclonal antibodies [12]. For connexin40, rabbit polyclonal antibody S15C(R83) was used at 1 : 250; for connexin45, guinea-pig polyclonal antibodies Q14E(GP42) or Q14E(GP42B) were used at 1 : 100 and 1 : 500, respectively. All polyclonal anticonnexin antibodies raised in the National Heart & Lung Institute laboratories were confirmed be isotype and gap junction-specific by immunofluorescence and Western blotting of connexin transfectants, and by electron-microscopic immunogold labelling [12]. Q14E(GP42B) is a newly purified batch of the original antiserum Q14E(GP42) raised in the same guinea-pig [12], which was characterized by Western blot and immunofluorescence against the original Q14E(GP42) for comparison.

The fluorescent secondary detection systems were: donkey antimouse-Cy3 (1 : 250; Chemicon) or Texas Red (1 : 500; Jackson ImmunoResearch Laboratory, Inc, USA) for connexin43; antirabbit TRITC (1 : 25, DAKO Ltd, UK) for connexin40; and donkey antiguinea-pig-Cy3 (1 : 250; Chemicon) for connexin45. The use of different fluorochromes ensured clear visualization against the natural autofluorescence of the tissue. Both primary and secondary antibodies were diluted in PBS containing 1% BSA and 5% human serum.

Sections (12 µm thick) of directly frozen (not chemically fixed) specimens were prepared using a cryomicrotome, mounted on poly l-lysine-coated slides and stored at − 80°C until use. The unfixed sections were immersed in methanol at −20°C for 10 min, rinsed three times in PBS for 10 min and blocked in PBS-buffered 1% BSA for 1 h before primary antibody incubation for 2 h at room temperature. The sections were then washed three times in PBS and incubated with the matching fluorochrome-conjugated secondary antibody for 1 h at room temperature. Negative immunolabelling controls omitted the primary antibody; positive controls used sections of rat heart in which the distribution of the three connexin isotypes is established [21].

Immunolabelled sections were examined by confocal laser scanning microscopy, using a system equipped with argon, krypton and helium-neon lasers. To distinguish autofluorescence (e.g. that from elastic laminae) from specific immunolabel signal, the fluorescein isothiocyanate channel detector was used. Series of images were sequentially recorded for each channel to avoid signal crossover.

For thin-section electron microscopy, samples were fixed in 2.5% glutaraldehyde, followed by 2% osmium tetroxide in cacodylate buffer. After dehydration through an ethanol series, the samples were embedded in epoxy resin (Araldite CY212, Agar Scientific Ltd, UK). Thin sections were stained with uranyl acetate and lead citrate, and examined in an electron microscope.

For Northern blot analysis, total cellular RNA was purified from frozen, pulverized tissues using a modified guanidinium isothiocyanate/acid phenol extraction [22,23]. Equal amounts (5 µg/lane) of each sample were run in formaldehyde-agarose gels and capillary-transferred onto nylon membrane (Hybond N, Amersham Int., UK). High stringency hybridization was used (65 °C, 5 × saline-citrate) with random primed probes generated from gel-purified inserts (connexin43, connexin40, connexin45) radiolabelled using 32P [23]. Northern blots were quantified by densitometric scanning of autoradiograms, using 10 samples in each group; multiple exposures were obtained to ensure linearity of the film response.


Impedance data are shown as the median (interquartile range) as sets were not normally distributed; other data are expressed as the mean (sd). Statistical tests used parametric or nonparametric unpaired Student's t-tests, with the null hypothesis rejected when P < 0.05. Plots of Rd vs –Xd were fitted by a least-squares method to the equation of a circle (radius ρ): (Rd − a)2 + (–Xd − b)2 = ρ2 (a and b are constants).


In the control group 15 samples were assessed from patients with stable bladders who had bladder, or in one case ureteric, carcinoma; the data were combined to form a control group. Four remaining specimens were obtained from patients with small (two), atonic or painful bladders but were excluded from the analysis.

For the unstable group, 10 samples were from patients with idiopathic detrusor instability (nine) or instability secondary to obstruction (one). Three further samples were from patients with hyper-reflexia and these data formed a separate group. Previous investigations measuring in vitro contractile properties of human detrusor have shown that hyper-reflexic tissue is indistinguishable from that obtained from stable bladders, but different from tissue obtained from patients with idiopathic instability [7].


The dimensions of the tissue samples were not significantly different in the two groups. The length of tissue in the central oil gap was 4.5 (1.6) and 3.9 (0.8) mm for preparations from stable and unstable bladders, and their cross-sectional areas were 1.9 (0.6) and 2.2 (1.0) mm2, respectively.

The effective electrical extracellular space, estimated from the extracellular resistance, rec, was 5.6 (2.7)% of total cross-sectional area in the control and 4.4 (2.9)% in the unstable group (not significantly different). The corresponding values in the three hyper-reflexic samples were 4.4 (0.1) mm, 1.4 (0.4) mm2 and 7.2 (1.8)%.


Figure 1a shows values of specific detrusor impedance, Z, for two samples of tissue from the control and unstable groups, respectively. The impedance is greatest at low measuring frequencies and declines to a lower steady-state finite value at higher frequencies. The frequency dependent decline of Z is characteristic of capacitative elements in the sample. A more accurate, quantitative characterization of the circuit elements contributing to the total tissue impedance can be obtained by dividing the Z-values at each frequency into their resistance (R) and reactance (–X) components and plotting –X as a function of R for all frequencies (Fig. 1b). Such plots generate a separate locus for each distinct time-constant (parallel resistance and capacitance) in the system and these plots suggest that there is a single time constant in these specimens. The sample can thus be modelled in terms of a resistance, R1, in series with a second parallel resistance, R2, and capacitance, C. The intercept of the locus with the R-axis at high frequencies, R8, gives the value for R1 and the intercept at low frequencies, R0, gives the value of R1 + R2, from which R2 may be calculated. Previous work using guinea-pig detrusor showed that R1 is equivalent to the sarcoplasmic resistivity, Rc, and that R2 is equivalent to the junctional resistance, Rj. The value of R1 + R2 (= Rc + Rj) is the total intracellular resistivity, Ri.

Figure 1.

Impedance measurements in detrusor. a, The variation in specific impedance, Z, as a function of frequency in detrusor samples from a stable (open green symbols) and unstable (closed red symbols) bladder. b, A plot of reactance (– X) as a function of resistance (R) for the data in a, the lines are a fit of a circle function to the data.


The total intracellular resistivity of the intracellular space, Ri, after correcting for extracellular space resistance was 817  (572–1173) Ω.cm (15 estimates) in specimens from 14 patients with stable bladders. With samples from patients in the unstable group Ri was significantly greater, at 1426 (1264–1787) Ω.cm (10 estimates, P < 0.05). These data excluded three samples from patients with hyper-reflexia that had a median impedance of 767 (652–826) Ω.cm, and similar to that for samples from patients with stable bladders. The Ri values were divided into their component fractions of Rc and Rj; the former were not significantly different in the two groups, at 294 (253–411) and 374 (306–412) Ω.cm (P > 0.05), whereas the latter were significantly greater in the unstable group, at 511 (368–778) and 1052 (895–1368) Ω.cm (P = 0.01). Thus the increase in Ri in the unstable groups can be attributed solely to greater junctional impedance. The corresponding values for the three samples from hyper-reflexic patients were: Rc 290 (233–314) Ω.cm and Rj 377 (335–452) Ω.cm.


The three connexins investigated showed distinctive, cell type-specific localization patterns (Fig. 2). Punctate connexin43 labelling (red fluorescent spots in all parts of Fig. 2) was particularly intense in a suburothelial band of interstitial cells (Fig. 2A) from a stable bladder sample. There was no unequivocal detrusor-related connexin43 labelling in the detrusor smooth muscle layer. Figure 2B shows that the few immunolabelled spots observed in the detrusor region were associated with elastic tissue autofluorescence, indicating that the label originated from interstitial cells within patches of extracellular matrix between detrusor bundles. The arrowheads show connexin43 label close to, but external to the detrusor, and identified with interstitial cells in the extracellular matrix. Where connexin43 label occurred within detrusor regions (faint red background fluorescence, arrows), it occurred within or adjacent to small patches of extracellular elastic fibres (green autofluorescence).

Figure 2.

Figure 2.

Immunoconfocal microscopy illustrating the localization of connexins (red fluorescent spots) in human bladder. A , connexin43 is located predominantly in a layer of interstitial cells beneath the urothelium (uro). B , Where connexin43 is located close to or apparently in the detrusor, the labelling is attributable to interstitial cells. Images in A and B are single optical sections. A low-power haematoxylin and eosin section through a separate preparation is shown, to indicate the regions from which A and B are taken. C , Connexin40 is confined to endothelial cells of the vascular system (L, vessel lumen). D, No connexin40 is detected in or between the detrusor bundles. E and F, The localization of connexin45 to the detrusor of stable ( E ) and unstable ( F ) bladders. The images in E and F were created as projections from stacks of serial optical sections to enhance visualization of connexin45 labelling. Scale bars = 25 µm in all figures.

Figure 2.

Figure 2.

Immunoconfocal microscopy illustrating the localization of connexins (red fluorescent spots) in human bladder. A , connexin43 is located predominantly in a layer of interstitial cells beneath the urothelium (uro). B , Where connexin43 is located close to or apparently in the detrusor, the labelling is attributable to interstitial cells. Images in A and B are single optical sections. A low-power haematoxylin and eosin section through a separate preparation is shown, to indicate the regions from which A and B are taken. C , Connexin40 is confined to endothelial cells of the vascular system (L, vessel lumen). D, No connexin40 is detected in or between the detrusor bundles. E and F, The localization of connexin45 to the detrusor of stable ( E ) and unstable ( F ) bladders. The images in E and F were created as projections from stacks of serial optical sections to enhance visualization of connexin45 labelling. Scale bars = 25 µm in all figures.

Connexin40 labelling was prominent in the endothelial cells of arteries and larger arterioles (Fig. 2C), but no convincing connexin40 labelling was apparent elsewhere in the tissue, including the detrusor (Fig. 2D). However, Fig. 2E,F show that connexin45 labelling, in contrast to connexin43 and connexin40, consistently revealed small but sharply defined fluorescent spots specifically in the detrusor. This connexin45 label was localized to the boundaries between interacting smooth muscle cells, and was markedly heterogeneous. Projection views obtained by combining stacks of serial optical sections allowed an appreciation of the consistent presence, extent and heterogeneity of the distribution of connexin45 labelling in the detrusor samples from both stable (Fig. 2E) and unstable (Fig. 2F) bladders. In keeping with the small, sparse connexin45 spots between detrusor cells detected by immunofluorescence confocal microscopy, thin-section electron microscopy revealed occasional small gap junctions of characteristic morphological structure (Fig. 3).

Figure 3.

Examples of gap junctions (gj) between detrusor cells as revealed by thin-section electron microscopy. In favourable section planes, the gap junctions show a characteristic pentalaminar structure formed from the two closely apposed unit membranes. Note that the gap junctions are small, as also assessed by immunoconfocal microscopy after labelling for connexin45. Scale bar = 100 nm.


Northern blots of the separated mucosal and detrusor layers (Fig. 4) showed a tissue distribution of the connexin transcripts mirroring that of the corresponding proteins localized by confocal microscopy. Abundant signal for connexin43 mRNA was apparent in the mucosal layer (which includes the connexin43-positive suburothelial layer of interstitial cells) but only low levels of connexin43 signal were detected in the detrusor layer (attributable to the connexin43-expressing interstitial cells visualized close to detrusor bundles, i.e. Fig. 3B). Connexin40 mRNA was only barely detectable in both layers, consistent with the overall very low levels of connexin40 arising from the restriction of this connexin to vascular endothelium. Connexin45 mRNA signal was marked in the detrusor layer, in keeping with the location of connexin45 in the smooth muscle, but only barely visible in the mucosal layer. The connexin45 transcript was detected as a main ≈ 8 kB band and a smear of lower molecular weight down to ≈ 2 kB (the molecular weight of the rodent orthologue), as has been reported previously in human uterus [23] and heart [15].

Figure 4.

Northern blot analysis comparing the levels of connexin43, -40 and -45 transcripts in mucosa and in detrusor layers dissected from human bladder. As comparisons, RNAs isolated from human right and left atria are also shown.

For comparison, RNA was isolated from left and right human atrial samples. Connexin43 was abundant, even compared with bladder mucosal samples, and connexin40 was also expressed at high levels [15]. However connexin45 was less abundant than in the detrusor layer.

Quantitative Northern blotting was used to compare the relative levels of transcript for each connexin type in detrusor from stable and unstable bladders. Figure 5 shows examples of blots from seven each of stable and unstable bladder samples. There was no significant difference in the levels of connexin43 or -40 transcripts, but the mean connexin45 transcript levels were reduced to 65% in the unstable compared with the stable bladders. The final quantitative analysis presented in the histograms contains data from 10 samples for each stable and unstable group. This difference in connexin45 mRNA levels was a statistically significant reduction in connexin45 transcript quantity.

Figure 5.

Examples of Northern blot analysis of connexin43, -40 and -45, from stable and unstable bladders. The quantification is shown as the histograms. Connexin45 level was significantly higher (*P < 0.05) by ≈ 45% in the control (stable) than in the unstable group. In the Northern blot examples, seven samples are shown in each group; 10 samples were used for the final quantitative analysis in the histograms.


Two key conclusions can be drawn from this combined electrophysiological and connexin study of human detrusor. First, the impedance analysis shows that the intracellular resistance is finite, with a component attributable to gap junction impedance [10,24]. This correlates with our report that the gap-junctional protein connexin45 is localized specifically to the smooth muscle component. Second, the electrophysiological analysis shows that, in tissue obtained from patients with unstable bladders, gap junction impedance is increased and this correlates with a distribution of connexin isotypes in the bladder.

The present study identified connexin45, rather than -40 or -43, as the principal connexin of human detrusor. The smallness of the connexin45 immunofluorescence spots is consistent with the smallness of the gap junctions visualized between detrusor cells by thin-section electron microscopy, and together with the sparsity of the junctions apparent by both techniques, explains the difficulty in detecting gap junctions by electron microscopy. Importantly, the anticonnexin45 antibody used here has been characterized rigorously by Western blot and immunofluorescence of cells transfected to express a range of connexin types, and by immunogold labelling and electron microscopy showing specific binding to connexin45-containing gap junctions and no binding to connexin43 or connexin40 [12,21]. There was a notable lack of connexin43 signal in the human detrusor smooth muscle; as reported elsewhere, connexin43 localizes most prominently to a suburothelial zone, where it forms the principal gap-junctional protein of interstitial cells [25]. A sparse distribution of connexin43 in the detrusor musculature was between, rather than within, muscle bundles, and its association with autofluorescent elastic fibres indicates that in this region too connexin43 labelling is attributable to interstitial cells. Connexin43 has previously also been described in the urothelium, where its down-regulation is implicated in the development of TCC [26]. Connexin40 was confined to the endothelium of the vasculature, in accord with previous observations [27,28].

The Northern blot comparison on mucosal/submucosal vs detrusor layers showed that the distribution of transcripts for the connexin isotypes paralleled that observed by immunocytochemistry. These findings emphasize the need for defining the precise cellular location of the various connexin isotypes when describing changes in bladder connexins. The description here of connexins in detrusor is at variance with electron microscopic work that has not in general identified gap junctions in this tissue.


The present electrophysiological and connexin findings provide evidence that gap junction coupling is reduced in tissue from patients with unstable bladders. The increase in intracellular impedance was attributed solely to greater gap junction resistivity, as sarcoplasmic resistivity was similar in the two groups. The ability of the impedance experiments to measure differentially a change to one component of the intracellular impedance pathway also implies that the method is precise enough to record accurately such changes. The observation that connexin45 transcript levels by Northern blot analysis were lower suggests a mechanism whereby gap junction resistivity is increased. Quantitative comparison of connexin45 protein levels between stable and unstable bladders was impracticable both by confocal microscopy, because of the heterogeneity of connexin45 signal distribution, and by Western blotting, as the levels of connexin45 were too low to be detectable with available antibodies. Although Northern blots only provide data on transcript levels, given the correspondence between the transcript distributions observed in mucosal vs detrusor layers with immunolocalization data on the corresponding protein, and the frequently reported positive correlation between connexin mRNA and protein levels in other studies [23], the reduction in connexin45 transcript level in the unstable bladder is likely to correspond with a reduction of connexin45 protein amounts.

Although there is a decrease in muscle mass in detrusor from unstable bladders, this would not account for the decreased connexin45 transcript level, as equal amounts of RNA (and therefore equivalent total cell volumes) were loaded per Northern blot lane. The smallness of spots on labelling for connexin45 suggests minute gap-junctional plaques, consistent with the difficulty in observing gap junctions by electron microscopy [29]. Previous work showed that isolated detrusor myocytes from unstable human bladders are hypertrophied compared with myocytes from control bladders [30]. Thus it cannot be stated at present whether the decrease of connexin45 label in the unstable bladder samples is a result solely of a decrease per cell, but the electrophysiological implications would not be altered, as intercellular coupling on a macroscopic scale would be affected only by the number of gap junctions per unit volume of tissue.

The impedance data have been interpreted using an electrical equivalent model of a parallel resistance and capacitance for gap junctions in series with a cytoplasmic resistance; these elements in turn are shunted by an extracellular resistance. Several equivalent (canonical) circuits are possible but this is preferred for several reasons, i.e. manoeuvres expected to change fairly selectively extracellular resistance (sucrose addition to Tyrode's solution) and gap junction resistance (low heptanol concentrations) generate impedance changes that are compatible with this model [10,17].


These data do not support the postulate that detrusor muscle cells from the unstable bladder are electrically better coupled. This was originally proposed by electron microscopy studies showing more so-called ‘protrusion junctions’ in tissue from a group of patients with overactive bladders [4,5]. However, the only junction type relevant to electrical coupling is the gap junction, and those studies provided no specific data on these [4,5]. Moreover, in an animal model of outlet obstruction, which may have generated some bladder instability, the electrical space constant was reduced [11], consistent with an increase of intracellular resistivity, as reported here.

The comparatively sparse connexin45 signal in detrusor does not preclude reasonable electrical coupling between contiguous muscle cells. In detrusor the membrane resistance is large [31] so that current flow is more readily confined to the intracellular space, despite a relatively high gap junction resistance. Indeed the space constant, which reflects the distance that electric currents can flow in the intracellular compartment, is ≈ 1 mm [11], similar to that in conventionally well-coupled tissues such as myocardium. The important conclusion of a relatively high intracellular resistivity, as recorded in detrusor, compared with myocardium, is that the propagation velocity of any regenerative electrical signal is slowed. Approximate calculations show that an action potential would propagate at ≈ 5 cm/s in normal detrusor [10] and ≈ 25% less in tissue from unstable bladders. These values are unlikely to coordinate contraction of the entire bladder but could easily coordinate small, localized contractions of the bladder wall.

Modelling experiments in myocardium show a more significant implication of a modest increase of intracellular resistivity as recorded here, i.e. an augmented generation of re-entrant arrhythmogenic circuits [32]. This is because local circuits that precede propagated action potentials are confined and so allow adjacent tissue to reach threshold more quickly, i.e. excitability is increased. Whether such a paradoxical effect is present in detrusor muscle is unknown. However by analogy, the present electrophysiological changes in human detrusor should generate a substrate for the development of local electrical, and thereby mechanical activity in the unstable bladder. The importance of developing accurate electrical equivalent models of detrusor tissue is thus evident to determine the role of such phenomena in generating abnormal bladder function.


This work was supported by a Link grant between the Medical Research Council and Pfizer Global Research & Development. Development of anticonnexin probes was aided by support from the European Commission (QLG1-CT-1999-00516).