Microvascular anatomy of the urinary bladder in the adult African clawed toad, Xenopus laevis: A scanning electron microscope study of vascular casts

Abstract We studied urinary bladders of adult male and female Xenopus laevis using light microscopy of stained tissue sections and scanning electron microscopy (SEM) of vascular corrosion casts (VCCs). Results showed that bilaterally a vesical artery branched off the femoral artery. At the dorso‐lateral serosal surface of the body of the bladder each artery splitted within a short distance into up to five smaller arteries that supplied body and neck regions. Arteries gave off short and long terminal arterioles, which fed the mucosal capillary meshwork. Long terminal arterioles followed dimensional changes of the bladder, while short ones anchored the capillary network to the arterial system. Capillary mesh sizes and shapes varied according to the filling state of the urinary bladder. In the highly to moderately distended (filled) bladder, capillaries were rather straight or undulated only slightly, in the contracted (emptied) bladder they undulated strongly and lay side by side. Postcapillary venules formed by two equally sized capillaries or from capillaries, which serially drained into a small postcapillary venule. Vesical venules formed a large dorsal vesical and a varying number of smaller lateral and ventral vesical veins. The dorsal vesical vein drained either directly or via the posterior hemorrhoidal vein into the common pelvic vein. Lateral and ventral vesical veins also drained into the latter. The vascular patterns found were discussed in respect to the bladder spatial movements during distention (filling) and relaxation (emptying). Furthermore, it was hypothesized that an extensively filled bladder could compress the overlaying abdominal vein forcing part of the blood otherwise drained towards the liver to be detoured via the renal portal veins to the kidneys.


| INTRODUCTION
The urinary bladder plays an important role in amphibian lifestyles. As anuran amphibians inhabit diverse environments, the amount of urine stored in their urinary bladders ranges from 50% of the body weight in the desert-burrowing frogs Notaden nicholsi and Neobatrachus wilsmorei (Main & Bentley, 1964) to as little as 1% in the aquatic Xenopus laevis (Duellman & Trueb, 1994). Different to the mammalian anatomy, where the urinary bladder fills via ureters, which directly empty into the bladder, in anurans ureters empty into the cloaca. From there, urine is backed up to the urinary bladder by cloacal smooth muscle contractions that have to overcome urinary bladder pressure resulting from bladder compliance, lung inflation, and buccal movements (Martin & Hillman, 2009).
Structure and function of the anuran urinary bladder are well described (see e.g. Bentley, 1966;Calamita et al., 1994;Gaupp, 1904;Mochida et al., 2008;Peachey & Rasmussen, 1961;Shibata et al., 2015;Wiechmann & Wirsig-Wiechmann, 2003) as are gross arterial supply and venous drainage (Gaupp, 1904;Millard, 1940;Roth, 1973;Szarski, 1948). Millard (1940) found that in the African clawed toad, Xenopus laevis, a single vesical artery originated bilaterally from the femoral artery and approached the urinary bladder at its dorsal circumference (see her Figures 8 and 9). Her study was based on excellent dissections, but because of the limited depth of focus of the dissecting microscope she could not visualize the bladder microvascular anatomy and so studies on the microvascular anatomy of the anuran urinary bladder are still lacking. This strongly contrasts with the situation in mammals including humans, where we know the three-dimensional microvascular anatomy of the healthy and diseased urinary bladder in detail (Hossler et al., 2013;Hossler & Kao, 2007;Hossler & Monson, 1995, 1998a, 1998bMiodonski et al., 1998;Miodonski et al., 2001;Miodonski & Litwin, 1999).
The muscular layers allow amphibian urinary bladders for a great distention when they are filled with urine (Duellman & Trueb, 1994).
We assumed that in such a situation a sufficiently high vesical blood flow can be only maintained if sufficient lengths of vesical vessels are present, which in the filled state should be rather straight or slightly undulating, but should highly undulate and be narrow spaced in empty bladders. To test this assumption, we analyzed vascular corrosion casts (VCCs) of urinary bladders of adult Xenopus laevis by the scanning electron microscope (SEM). The high depth of focus and a sufficiently high resolution gained at an acceleration voltage of 10 kV in the conventional scanning electron microscope allowed (i) to identify cast capillaries as the blood vessels with the thinnest diameter, (ii) to differentiate arteries and veins by their characteristic endothelial cell nuclei imprint patterns displayed on the surface of vascular casts (Miodonski et al., 1976), and (iii) to localize anatomical structures regulating blood flow in the bladder, i.e. arterial and venous sphincters (Aharinejad et al., 1991;Franz et al., 1990;Schraufnagel & Patel, 1990), flow dividers, intimal cushions (Bond et al., 2010;Casellas et al., 1982;Fourman & Moffat, 1961) and venous valves (Caggiati et al., 2006;Hossler & West, 1988).
Additionally, paraplast embedded, stained urinary bladder tissue sections were studied in order to attribute displayed vascular patterns in the fully digested urinary bladder cast specimens to defined tissue layers.

| Animals
Ten adult individuals of both sexes of the African Clawed Toad, Xenopus laevis (Daudin, 1802) were studied. Adults were housed in aquaria (tap water depth: 15 cm) equipped with aquarium filters and fed twice a week with either dried Gammarus pulex or ground beef heart.   Cologne, DE). Casts of the abdominal area with the urinary bladder insitu were mounted dorsal side down onto specimen stubs using the "conductive bridge-method" (Lametschwandtner et al., 1980).

| Histomorphology
Mounted specimens were either evaporated with carbon and gold or sputter-coated with gold and examined in the scanning electron microscope XL-30 (FEI, Eindhoven, The Netherlands) at an accelerating voltage of 10 kV.
After a first SEM-inspection, course, branching patterns and areas of supply (or drainage) of individual abdominal and urinary bladder vessels were exposed in some specimens by ripping-off overlaying vessels with fine tipped insect pins under binocular control. Specimens then were sputter-coated with gold and analyzed in the SEM.

| RESULTS
The moderately filled urinary bladder of adult Xenopus laevis had an ovoid to spherical body (Figure 1(a)) and a funnel-like neck, which

| Vascular anatomy
In only one (out of eight) specimen an almost perfect filling with few missing cast microvessels was gained. In all other specimens filling of vascular beds ranged from good to poor but allowed the recognition of individual differences between specimens in courses, calibers, branching and merging patterns of cast vesical arteries, veins and capillaries.

| Arterial supply
The

| Capillary bed
In the moderately distended urinary bladder, capillary beds formed mesh works in which capillaries undulated only slightly (Figures 3(d),  (Figures 3(d),(f) and 4(a),(b)). In the contracted urinary bladder mesh capillaries undulated strongly and lay close aside each other (Figure 4(c)). In the strongly contracted bladder, mucosal capillaries lay side by side and formed dome-like structures representing the mucosal foldings (Figure 4(d)).
Occasionally, single capillaries ran freely over a long distance before they drained into postcapillary venules (Figure 3(d),(e); short arrows).

| Venous drainage
The mergings of several small postcapillary venules to a larger venule occurred within a short distance ( Figure 5, inset 1). In some areas, large postcapillary venules joined larger vesical veins in an obtuse angle ( Figure 5, inset 2). Occasionally, also capillary-sized vessels drained into larger veins ( Figure 5, inset 3). Small vesical venules displayed two basic patterns. In most cases they formed by merging of postcapillary venules of different sizes (Figures 3(d),(e), 4(a)-(c), and 5).
In some specimens larger postcapillary venules received regularly spaced capillaries bilaterally ( Figure 5, inset 4). Vesical veins ascended from lateral and ventral serosal surfaces towards the dorsal surface. In the orthogradely filled large valve, the leaflet structure is clearly replicated (long arrows). Note small valves where retrograde resin flow was stopped at the valves (short arrows). Dashed arrows mark direction of blood flow, asterisks mark slit-like opening of bladder into cloaca. av abdominal vein, cb conductive bridge, cl cloaca, rpv renal portal vein, va vesical artery drained into right or left common pelvic veins (Figure 6(a), arrow).

| DISCUSSION
In the present study, no specific pretreatment of animals was performed to gain a defined filling state of the urinary bladder. The filling state was estimated from casted specimens. Large casted bladders with only slightly undulating vesical arteries and wide capillary meshes were considered filled (distended), smaller ones with strongly undulating vesical arteries and narrow capillary meshes were considered empty (con- Our study confirmed the findings of Millard (1940)

that in
Xenopus laevis one vesical artery branched off the femoral artery bilaterally close to its origin from the common iliac artery. This contrasts with the situation in Rana esculenta (Gaupp, 1904), and Pelobates fuscus, Bombina bombina and Bufo bufo (Szarski, 1948) where the vesical artery branches off the epigastric artery, a branch of the femoral artery. Xenopus lacks an epigastric artery.
Of interest was the pronounced flow divider found in several specimens at the origin of the vesical artery from the femoral artery. This flow divider by its orientation efficiently governed blood flow into the rather long vesical artery and thus it was considered an important structure to ensure the blood supply of the urinary bladder under varying filling conditions of the bladder. Of interest was also that vesical arteries ran unbranched for a long distance behaving as conducting arteries before they changed into supplying arteries, which give off branches within much shorter distances (see Figure 3 guinea pig, rabbit, dog, and pig (Hossler & Kao, 2007;Hossler & Monson, 1995, 1998a, 1998b. Arterial sphincters have been described in dog and pig only (Hossler & Kao, 2007). In the urinary bladder of Xenopus, arterial sphincters were also present and most likely participated in the regulation of blood flow towards the areas of need.
In urinary bladders of many mammals, collateral circulations have been described, particularly between the main supply vessels at the serosal surface and within the mucosal vascular bed (Hossler & Kao, 2007). In the urinary bladder of Xenopus few arterio-arterial anastomoses were found. Arterial branching modes and vessel courses were obviously sufficient to ensure blood supply.
In our opinion, the two patterns of terminal vesical arterioles, i.e. short and long, most likely serve the flexibility of blood flow in the mucosa under varying filling conditions. While the long arterioles allow greater spatial movements of the mucosa, the short arterioles were thought (i) to establish short arteriolar-capillary transition times in all states of bladder filling and (ii) to anchor the mucosal capillary bed to the vesical arterial system.
Of interest to mention are the few single, rather straight running capillary-sized vessels found as the most abluminal located vessels In Xenopus laevis, ureters do not open into the urinary bladder but they open at the urogenital papilla (in males) or at the urinary papilla (in females) into the cloaca (Brown, 1970). The mode of filling of the urinary bladder therefore is quite different from that of mammals in which ureters open directly into the bladder and in which a valve at the bladder entrance prevents reflux of urine back into the ureter (Hossler & Kao, 2007). According to Martin and Hillman (2009) who studied the terrestrial toad, Chaunus marinus and the aquatic frog, Lithobates grylio "..bladder filling is a result of pressures generated in the ureters and the cloaca that are greater than internal bladder pressures and that the more aquatic species had a less compliant urinary bladder compared with that of the terrestrial species". Our observations that the urinary bladder of Xenopus laevis was highly distended in some specimens indicated that its bladder is compliant. Obviously, the maximally four layers of blood vessels within the thin wall of the organ by their courses and patterns easily follow the dimensional changes of the bladder during distention by changing from highly coiled patterns of arteries, capillaries and veins in contracted (emptied, slightly filled) bladders to slightly undulating (in moderately filled) or even straight courses in distended (filled) bladders.

ACKNOWLEDGMENTS
We thank Dr. Heidi Bartel for assistance in histomorphology and processing of SEM specimens, Christine Radner, BSc. for assistance in histomorphology and SEM specimen preparation, and Dr. Wolf-Dietrich Krautgartner for excellent technical advice in the SEM facility.

CONFLICT OF INTEREST
The authors have no conflict of interests to declare.

DATA AVAILABILITY STATEMENT
All corrosion cast and histological samples are stored a the Department of Bioscineces, University of Salzburg. The material is available at request from the corresponding author.