Although bolstered by intuition and confirmation bias, the dogma of "water-tight" bladder lining is undermined by the tritiated water absorption rate of 0.3-1mL/min from human bladder(ref.1) and saline absorption rate of 0.001mL/min from rat bladder (ref.2). This bladder surface area dependent, 300-fold decline in water absorption rate is facilitated by aquaporin channel subtypes expressed on intermediate cell layer of urothelium. However, asymmetric unit membrane of umbrella cells is renowned for restricting transcellular diffusion of urine constituents and the marked absence of aquaporin channels on luminal surface raises the question: what is the physical mechanism of water movement from bladder lumen to intermediate urothelium cells? Since tritiated water absorption rate (ref.1) increases with distended bladder wall expanding gap of tight junctions gap, a plausible mechanism for water movement preceding the facilitated diffusion of water by aquaporins into systemic circulation is the passive, paracellular diffusion of free and bound water through tight junctions.
The premise of paracellular diffusion is consistent with the decline in bladder volume recorded by periodic ultrasound measurement during sleep of healthy adults which mimics an average 500mL decline in 24hour urine volume with volitional extension of awake voiding interval from 3h to 5h (ref.3) for a week by healthy adults. However, instead of diffusion, water movement into human bladder was conjectured to be osmosis (ref.3), which formed our null hypothesis: water moves from urine to aquaporin channels of intermediate urothelium by osmosis or alternatively by paracellular passive diffusion. Since the net movement of free water from low osmolality of glomerular filtrate in the loop of Henle to the high osmolality of renal medulla is emblematic of osmosis,  we relied on a well-established relationship between spin-lattice (T1) and spin-spin relaxation time (T2) constants of water protons with the osmolality to benchmark the osmolality gradient between urine and urothelium to the gradient between renal medulla and cortex for a conceptual evaluation of null hypothesis.

Four months old female B6D2F1 mice (n=3) were anesthetized by 1-2% isoflurane and abdomen was secured to platform for reducing motion artifacts during MRI at 7 Tesla by 30-cm AVIII spectrometer using an 86 mm quadrature RF volume coil with a 4-channel receive surface array. T2-weighted coronal scans were acquired by rapid acquisition with resolution enhancement (RARE) sequence with following parameters: repetition time (TR)/echo time (TE) = 3000/40 ms, field of view (FOV) of 40 mm2, acquisition matrix = 128 × 128, slice thickness of 0.8 mm for 15 slices, 2 signal averages, and a RARE factor = 8. Coronal T1 maps were acquired using a variable TR sequence:  400, 842, 1,410, 2,208, 3,554 and 10,000 ms, echo time (TE) = 7 ms, 9 contiguous 0.8 mm slices,  RARE factor = 2, 2 signal averages, 28 mm2 FOV and matrix = 218 × 218. T1 maps were processed using a 3-parameter single exponential function and unpaired Student's t test assessed the difference between urothelium and urine T1 for correlation with published osmolality values.

Renal medulla and stored urine in bladder lumen appear brighter than the renal cortex and bladder wall, respectively in T2 weighted images acquired at TR/TE of 3000/40ms (Fig.1). Although  T1 contrast of urine at TR/TE 2762/6.5ms is sub-optimal for bladder wall segmentation, color-coded mapping of T1 relaxation time (T1) for each voxel reconstructed from six T1 weighted images acquired at variableTR 400-7500ms and constant TE of 6.5ms visually segmented the ~ 0.6mm thick mouse bladder wall into 0.15-0.2mm thick urothelium layer displayed by at least two-pixel thick yellow band encircling urine marked by red lumen and ~0.4mm thick detrusor layer is displayed by bluish green. Red and yellow color display significantly higher urine T1 of 5000+/-400ms than urothelium T1 of 3500+/- 350ms (p<0.05) and the difference in physical parameter of T1 mirror the authenticated osmolality gradient from >400mOsmoles/L for urine to 280mOsmoles/L of interstitial fluid- proxy for urothelium. A significantly higher T1 of 1800+/- 300ms (yellowish green) than renal cortex T1 of 1200+/- 100ms (bluish green) are compatible with true osmosis symbolized by free water movement into high osmolality of renal medulla from glomerular filtrate (proxied by cortex). In contrast, reports of water movement into urothelium from hypotonic (ref.1), isotonic (ref.2) and hypertonic urine (ref.3) demonstrate an independence from osmolality gradient (Fig.1) which affirms alternative hypothesis on passive paracellular diffusion of water before aquaporins of intermediate layer facilitate water diffusion into systemic circulation(Fig.2).

Since true osmosis is antithetical to free water movement from significantly higher osmolality of urine to lower osmolality of urothelium, T1 mapping of urine and urothelium affirms the alternative hypothesis of passive paracellular diffusion of free (ref.1) and bound water (ref.2) as spheres of Stokes-Einstein radius (>1.37Angstrom) through the tortuous gap of tight junctions assembled at mammalian umbrella cell borders. The compliance of paracellular diffusion with Stokes diffusion principle-inverse size dependence-is self-evident from three times faster diffusion rate of three times smaller hydrated sodium (1.3 Angstrom) than dextrose (~3.28Angstrom), determining the three times higher systemic uptake of saline than dextrose (ref.2). As the radius of water molecule is three times smaller than polar dyes (Fluorescein and Gadolinium chelate), the fluorescence and image contrast of umbrella cell borders visually affirms the paracellular diffusion of water whereas dark apical surface of umbrella cells attests the restricted transcellular diffusion of water. Furthermore, free water reabsorption is inevitable to cause a feed-forward rise in the osmolality of residual urine, which elevates the concentration gradient for Fickian diffusion of water bound to Na+/K+ or urea from lumen. Importantly, local buildup of diffused agents triggers a reflexive, homeostatic acceleration of urothelial blood flow as measured during potassium sensitivity test (PST) and corroborated by the rapid systemic distribution of instilled drugs (DMSO, Formalin and Lidocaine).  While the concentration gradient generated by higher osmolality of residual urine provides the pushing force, the clearance of diffused agent from urothelial blood provides the pulling force in same direction for sustaining  the concentration gradient which is also facilitated by a forty-fold upregulation of aquaporins in the intermediate cell layer of urothelium (ref.2) to accelerate the Fickian diffusion of free as well as bound water into systemic circulation at the rate of 1mL/min(ref.1). Thus, the reduction in 24h urine volume with mere extension of voiding interval by 2h (ref.3) exhibits a homeostatic mechanism of reflexive acceleration of urothelial blood flow to augment Fickian diffusion of water that delays the awakening of healthy adults while they sleep and prevents the buildup of K+ from irritating sensory nerve endings  of urothelium that occurs in PST of cystitis patients but not in healthy adults. While rich vasculature saves dense innervation of urothelium from irritation, sparse innervation tolerates counter-current multiplier mechanism erected by slow blood flow of vasa recta capillaries for osmotic movement of water into renal medulla (Fig.2).

Hence, water movement from significantly higher osmolality of urine into urothelium is consistent with reverse osmosis as faster blood flow and richer innervation than renal medulla ensure aversion to true osmosis. While 73% of the luminal surface lined by apical surface of umbrella cells is water tight, 27% of luminal surface lined by apicolateral umbrella cell borders is amenable to size dependent passive paracellular diffusion of free and bound water, which gets accelerated with distension driven expansion of gap available for diffusion in tight junction.

Hilson AJ, Lewis CA, Harland SJ. The permeability of the human bladder to water assessed using tritiated water. Contrib Nephrol. 1990;79:41-4.
Morizawa Y, Torimoto K, Hori S, Gotoh D, Nakai Y, Miyake M, Tanaka N, Hirayama A, Fujimoto K. Sodium plays an important role in the absorption of intravesical fluid. Low Urin Tract Symptoms. 2021 Jan;13(1):177-182
Fry C, Tasman K, Goodhead L. Salt And Water Transport Across The Human Bladder Wall. Continence13 July 2023 https://doi.org/10.1016/j.cont.2023.100812

Continence 12S (2024) 101427
DOI: 10.1016/j.cont.2024.101427