Hypothesis / aims of study
Organ bath experiments with isolated tissue strips are a cornerstone for the in vitro assessment of urinary bladder contractility and its possible alterations in disease and/or upon treatment. It appears obvious that a larger strip can generate a greater force of contraction. As it is close to impossible to cut bladder strips that have an identical size, investigators typically normalize force of contraction based on strip size. However, there is disagreement in the community whether size should be defined by strip weight, length or cross-sectional area, and all three approaches are used (1-3). Comprehensive comparisons of normalization approaches have not been reported. Moreover, it remains unclear whether a given normalization approach is equally applicable for contraction induced by different stimuli. Therefore, we have analyzed data from an ongoing study to compare these three normalization approaches for three different contraction parameters, peak and plateau contraction by a muscarinic receptor agonist and peak contraction by the receptor-independent stimulus KCl.
Study design, materials and methods
The present study has used animals from an ongoing study related to type 2 diabetes and its treatment. Male Sprague-Dawley (4 weeks of age) rats were obtained from an external animal breeding facility. At 6 weeks of age, some rats were put on a high-fat diet, whereas others continued on a standard diet. Four weeks later, the rats on the high-fat diet received an intra-peritoneal injection of streptozotocin (25 mg/kg); rats on standard chow received a vehicle injection. At 12 weeks of age, one group each on high-fat and standard diet started receiving the SGLT2 inhibitor dapagliflozin (1 mg/kg once daily by oral gavage). At 24 weeks of age, animals were sacrificed under isoflurane anesthesia, and urinary bladder and other tissues were harvested. In total, 4 groups of animals were studied: A) control, B) control + dapagliflozin, C) high-fat diet + streptozotocin and D) high-fat diet + streptozotocin + dapagliflozin.
The bladder was cleaned from surrounding adipose and connective tissue, the trigone and he uppermost part of the dome, and 4 strips were cut from each bladder body. They were mounted in 10 ml organ bath in Krebs-Henseleit buffer and connected to a force transducer for isometric tension recording under a resting tension of 10 mN. Following 75 min of equilibration with regular freshening of buffer, strips were exposed twice to 50 mM KCl. The peak tension recorded after the second KCl addition was used to define “KCl peak”. After washing and an additional 45 min of equilibration, a cumulative carbachol concentration-response curve was generated. Based on peak contraction response to all concentrations, the maximum response (Emax of peak carbachol response) was calculated by fitting a sigmoidal curve to the experimental data. After another wash and 45 min of equilibration, 1 µM carbachol was added and the plateau tension was recorded (plateau carbachol response).
At the end of the experiment, strips were taken from the organ bath, strip weight and length were measured, and cross-sectional area calculated (weight/(length*1.05)). For each of the three contractile responses (KCl peak, carbachol peak, carbachol plateau), a correlation analysis against weight, length and cross-sectional area was performed. A greater r2 of the correlation analysis was defined as a more suitable form of normalization. The investigators performing the organ bath experiments and those doing the data analysis were blinded to group allocation of experimental animals.
Interpretation of results
Our data are based on a considerably larger number of strips than previous validation attempts. In contrast to previously reported smaller data sets (2), they suggest that contraction is only poorly correlated with strip length; therefore, normalization for length does not appear useful. In confirmation of a previously reported small data set (1), contraction was correlated to strip weight, but the tightness of correlation was only moderate (r2 ≈ 0.3-0.4). Cross-sectional area exhibited a tighter correlation than strip length, but a poorer than strip weight. However, all three parameters explained only up to about 40% of variability. Accordingly, comparison of contraction across the 4 experimental groups was similar regardless which, if any, normalization was applied.