Urine Stasis Predisposes to Urinary Tract Infection by an Opportunistic Uropathogen in the Megabladder (Mgb) Mouse

Purpose Urinary stasis is a risk factor for recurrent urinary tract infection (UTI). Homozygous mutant Megabladder (Mgb-/-) mice exhibit incomplete bladder emptying as a consequence of congenital detrusor aplasia. We hypothesize that this predisposes Mgb-/- mice to spontaneous and experimental UTI. Methods Mgb-/-, Mgb+/-, and wild-type female mice underwent serial ultrasound and urine cultures at 4, 6, and 8 weeks to detect spontaneous UTI. Urine bacterial isolates were analyzed by Gram stain and speciated. Bladder stones were analyzed by x-ray diffractometry. Bladders and kidneys were subject to histologic analysis. The pathogenicity of coagulase-negative Staphylococcus (CONS) isolated from Mgb-/- urine was tested by transurethral administration to culture-negative Mgb-/- or wild-type animals. The contribution of urinary stasis to CONS susceptibility was evaluated by cutaneous vesicostomy in Mgb-/- mice. Results Mgb-/- mice develop spontaneous bacteriuria (42%) and struvite bladder stones (31%) by 8 weeks, findings absent in Mgb+/- and wild-type controls. CONS was cultured as a solitary isolate from Mgb-/- bladder stones. Bladders and kidneys from mice with struvite stones exhibit mucosal injury, inflammation, and fibrosis. These pathologic features of cystitis and pyelonephritis are replicated by transurethral inoculation of CONS in culture-negative Mgb-/- females, whereas wild-type animals are less susceptible to CONS colonization and organ injury. Cutaneous vesicostomy prior to CONS inoculation significantly reduces the quantity of CONS recovered from Mgb-/- urine, bladders, and kidneys. Conclusions CONS is an opportunistic uropathogen in the setting of urinary stasis, leading to enhanced UTI incidence and severity in Mgb-/- mice.


Introduction
Congenital obstructive uropathy (COU) ranks among the most common causes of pediatric chronic kidney disease (CKD) and end stage renal disease (ESRD). [1] Recurrent urinary tract infections (UTIs) are common in patients with COU, including posterior urethral valves (PUV) and prune belly syndrome. UTI prolongs hospitalization in patients with COU, resulting in 30% increase in hospital costs. [2] UTI severity increases in the setting of urinary tract obstruction, placing patients at risk for severe pyelonephritis and urosepsis. Pyelonephritis and urosepsis are risk factors for CKD progression in COU patients with PUV and prune belly syndrome. [3,4,5] These observations justify efforts to aggressively prevent, detect, and treat UTIs in COU patients. However, the lack of animal models of spontaneous UTI has limited research in this arena.
The Megabladder (Mgb) mouse is a unique genetic model of COU. [6,7] Mgb arose as a consequence of a complex genomic rearrangement during transgenesis. Whereas mice with a single Mgb allele (Mgb+/-) are phenotypically normal, Mgb-/-embryos exhibit impaired mesenchymal compaction during detrusor muscle development. [7,8] Detrusor agenesis results in poor contractility and impaired urine outflow. The Mgb-/-mouse has been previously characterized as a model of a congenital functional obstruction at the level of the bladder, leading to hydroureteronephrosis and renal injury beginning in utero. [1,6,7,8] The atonic Mgb-/-bladder parallels that seen in children with prune belly syndrome, Hinman syndrome, and certain cases of myelomeningocele. This contrasts with detrusor hypertrophy associated with anatomic bladder obstruction, such as PUV. [1] Mgb-/-males with severe hydroureteronephrosis require urinary diversion by cutaneous vesicostomy (CV) to prevent death by 5-6 weeks. [6,9] Bladder stones and UTI can be induced in up to 30% of Mgb-/-males by CV with polypropylene suture. [10,11,12] Compared to their male counterparts, Mgb-/-females exhibit reduced severity of hydronephrosis, prolonged lifespan, decreased incidence of ESRD, and rarely require surgical rescue. Hence, female Mgb-/mice can serve as a model of urinary stasis and long-term CKD secondary to a functional obstruction. In routine care of Mgb-/-animals, we identified UTI associated with struvite urolithiasis in moribund females, none of whom had undergone surgical intervention (unpublished observations). We undertook this study to test the hypothesis that urine stasis predisposes to spontaneous UTI in Mgb-/-females.

Study approval
Use of the Mgb-/-mouse and experimental UTI models was approved by The Research Institute at Nationwide Children's Hospital Institutional Laboratory Animal Care and Use Committee (Welfare Assurance Number A3544-01), protocols AR13-00057 (BB) and AR13-00001 (SEI).
transgenic barrier facility on a 12-hour light-dark cycle and fed standard chow (Harlan Laboratories, Indianapolis, IN).

Microbiology
Urinary tract organs were sterilely dissected from euthanized mice and homogenized in phosphate-buffered saline. Organ and stone homogenates were serially plated on Luria Broth (LB) agar without antibiotic selection. Colonies were speciated (ChildLab, Nationwide Children's Hospital) and frozen as glycerol stocks. CONS292 is a urease(+) CONS strain recovered from the bladder stone of a spontaneously infected Mgb-/-mouse.
Experimental UTI model 6-week-old FVB/N female mice were purchased (Jackson Laboratories, Bar Harbor, ME) and rested for 1 week. Immediately prior to introduction of inoculum, urine was tested for pH and absence of bacteria. One day prior to inoculation, CONS292 was seeded into LB medium and cultured statically at 37°C for 16 hours. On the day of inoculation, bacteria were harvested by centrifugation and resuspended in phosphate buffered saline (PBS). Under isoflurane anesthesia, mice underwent urethral catheterization. [13] 10 8 colony-forming units (CFU) of CONS292 were transurethrally introduced in 50 μl saline. The inoculum was confirmed by plating on LB agar. Urine was collected from the urethral meatus at the specified intervals following inoculation. 14 days post inoculation (dpi), animals were re-anesthetized for ultrasound, urine collection, and then euthanized. Bladders and kidney homogenates were plated on LB agar to enumerate bacterial burden.

Ultrasound
Anesthetized Mgb-/-mice underwent ultrasonography at 4, 6, and 8 weeks of age. [6,14] All animals undergoing experimental UTI had ultrasound prior to inoculation and again 14 dpi. Although we observed excellent agreement between ultrasound and necropsy observation of stones, ultrasound was used as a screening modality to determine if the stone burden endpoint criterion for euthanasia was reached at early time points. Necropsy was considered definitive for the diagnosis of stones.

Urine and stone evaluation
Urine was obtained from the urethral meatus (WT or Mgb+/-animals) or by direct aspiration from the bladder (Mgb-/-). Urine pH was determined by indicator strips (pH range 4.5-10, Cardinal Health, Waukegan, IL). Urine was centrifuged at 3000 x g for 5 minutes, and the sediment was evaluated by light microscopy using a DMI4000 microscope equipped with DFC450C camera (Leica Microsystems, Wetzler, Germany). Bladder stone composition was evaluated by x-ray diffractometry (Louis C. Herring and Company, Orlando, FL).

Data analysis
Based on our initial observations, we determined that a cohort of 17 animals would provide >80% power to detect spontaneous urolithiasis and/or bacteriuria at a frequency of 5% and 90% power to detect an incidence of 10% during the study period. To allow for animal loss or other unforeseen circumstances, we established a cohort of 19 consecutive mgb-/-female mice.
Due to the small sample size, Fisher's exact test was used for the analysis of categorical data. To determine significance of quantitative data, Mann-Whitney U test was used. Except as noted, p < 0.05 was considered statistically significant. All statistical tests were performed using Stata v.11.1 (StataCorp, College Station TX).

Results
Coagulase-negative Staphylococcus species are associated with struvite urolithiasis and urinary tract colonization in Mgb-/-females To investigate the incidence and etiology of bladder stones in Mgb-/-females, we performed serial ultrasounds and urine cultures in a cohort of 19 consecutive 4 week old Mgb-/-female mice. If stones were absent and urine remained sterile, we repeated ultrasound and cultures at 6 and 8 weeks of age. At 8 weeks, all animals were euthanized and tissues were collected for culture and histopathology. We identified bladder stones in 6/19 Mgb-/-mice (31.6%; Fig  1A). In two instances, we observed concurrent kidney and bladder stones (Fig 1B). At necropsy, we enucleated multiple stones from dissected Mgb-/-bladders ( Fig 1C). All bladder stones were composed of 90% struvite with up to 10% apatite, consistent with an infectious origin.
We identified bacteriuria in 8/19 (42%) of Mgb-/-mice. Urine cultures from six animals with bladder stones grew coagulase negative Staphylococcus (CONS) non-saprophyticus species as solitary isolates. Urine from two animals lacking bladder stones contained Enterococcus species as solitary isolates. Cultures from homogenized stones grew coagulase negative Staphylococcus (CONS) non-saprophyticus species as a solitary isolate in all six urolithiasis cases.
We confirmed that CONS from Mgb-/-stones expresses urease, the enzyme responsible for hydrolyzing urea to ammonia, a required step in infection stone formation. [15] As expected in urine colonized by urea-splitting organisms, urine pH was significantly more alkaline in Mgb-/-females with stones (p = 0.0003, Mann-Whitney U test) or bacteriuria (p = 0.0004), compared to Mgb-/-females with sterile urine and absent stones ( Fig 1D). Urinary sediment contained Gram-positive cocci in clusters (Fig 1E), as well as struvite crystals (Fig 1F and 1G) associated with polymorphonuclear leukocytes and urothelial cells ( Fig 1G).
By eight weeks of age, urine from 8/19 (42.1%) Mgb-/-females contained > 100,000 colony forming units (cfu) bacteria/ml, identified as CONS in hosts with alkaline urine or a bladder stone. Animals with bladder stones also displayed evidence of lower and upper urinary tract colonization, based on recovery of CONS from bladder and kidney homogenates (Fig 2). Rare positive urine cultures from Mgb-/-females with acidic acid and absence of bladder stones grew Escherichia coli or Enterococcus species. We analyzed a cohort of consecutive Mgb+/-and wild-type (WT) mice from the parental FVB/N strain, to determine if the incidence and magnitude of urinary tract colonization were distinct from Mgb-/-animals. Whereas WT urine was consistently sterile, we detected low levels of spontaneous bacteriuria (1000-8000 cfu/ml) in 3/ 14 urine samples from Mgb+/-mice. The magnitude of CONS recovered from the urinary tract of Mgb-/-mice was significantly higher than that observed in the Mgb+/-and WT cohorts ( Fig  2). We observed no alkaline urine or bladder stones in Mgb+/-or WT mice (Fisher exact test, p = 0.027).

Cystitis and Pyelonephritis in Mgb-/-mice
We next examined the histology of bladders and kidneys of Mgb-/-mice with bladder stones. In the absence of bladder stones and bacteriuria, the Mgb-/-bladder exhibits distinct features compared to WT controls: detrusor hypoplasia, urothelial thickening, and luminal debris (Fig 3A  and 3B), consistent with previous reports. [7,10] Mgb-/-bladders with stones displayed increased urothelial thickening, mucosal detachment, and histologic evidence of luminal stones (Fig 3C). A mononuclear and neutrophilic infiltrate occupied the urothelium and submucosa, with submucosal fibrosis evident on Trichrome stain (Fig 3D). Cytoplasmic vacuoles within the urothelium indicated acute cellular degeneration (Fig 3D inset). Infected Mgb-/-kidneys exhibited interstitial nephritis in the cortex and medulla (Fig 3E). Medullary collecting ducts were markedly dilated with cytoplasmic flattening, reactive nuclei, and luminal cellular degeneration consistent with acute tubular necrosis ( Fig 3F). As in the bladder, trichrome staining identified collagen deposition immediately beneath the renal urothelium ( Fig 3F). These features of cystitis or pyelonephritis observed in Mgb-/-urinary tracts with bladder stones were not evident in tissues from culture-negative Mgb-/-, Mgb+/-or WT mice (Fig 3A and 3B; data not shown).

Mgb-/-hosts are susceptible to CONS experimental UTI
Based on the preponderance of CONS with urease activity among our stone bacterial isolates, along with the reported urease production by 55% of Staphylococcus species in humans with complicated UTIs [15], we reasoned that CONS was most likely responsible for bladder stones and UTI in Mgb-/-mice. We therefore investigated whether CONS isolated from spontaneously infected Mgb-/-mice could establish UTIs in uninfected Mgb-/-animals or WT controls. We isolated CONS strain 292 (CONS292) from a Mgb-/-bladder stone and confirmed urease activity. Next, after confirming urine sterility, we transurethrally injected 10 8 colony forming units (CFU) of CONS292 into the bladders of 6 week old Mgb-/-or WT females. These mice were serially monitored for signs of morbidity as well as bacteriuria, crystalluria, urine pH and ultrasound changes.
CONS292 initially colonized the urinary tracts of both Mgb-/-and WT mice (Fig 4A). However, we observed significantly less CONS burden in WT urine and bladders 1 day post infection (dpi), whereas renal colonization was comparable between genotypes (Fig 4A). CONS bacteriuria persisted for at least 2 weeks in Mgb-/-mice, whereas wild-type urine was sterile by 10 days post infection (dpi; Fig 4B). Mgb-/-urine uniquely contained Gram-positive cocci in clusters 14 dpi (Fig 4C). Mgb-/-urine pH was significantly more alkaline than WT urine as early as 3 dpi (p = 0.0001, data not shown) and at 10 dpi (p = 0.0002, Fig 4D). Mgb-/-mice infected with CONS292 demonstrated struvite crystalluria and bladder stones 14 dpi, whereas WT mice did not (Fig 4E).   Upon histological analysis 14 dpi, CONS-inoculated WT bladders exhibited prominent umbrella cells and a localized inflammatory infiltrate (Fig 5A). In contrast, CONS-infected Mgb-/-bladders demonstrated struvite crystals, urothelial thickening, and submucosal edema (Fig 5B). Inflammation in WT kidneys was absent in the renal parenchyma and restricted to the perihilar fat 14 dpi (Fig 5C). CONS-infected Mgb-/-kidneys manifested widespread interstitial and tubular inflammation with neutrophil-rich leukocyte casts in tubules and a vacuolated urothelium, as observed during spontaneous infection (Fig 5D).

Urinary diversion reduces susceptibility of Mgb-/-mice to CONS experimental UTI
We hypothesized that the heightened susceptibility of the Mgb-/-urinary tract to CONS can be attributed at least in part to urine stasis. We therefore performed CV using absorbable polydioxanone suture in a cohort of uninfected Mgb-/-females, and confirmed the absence of urolithiasis, acidic pH, and sterility of the urine immediately prior to inoculation with 10 8 CFU CONS292. [6,9,10] Following a 14 day period of observation, animals were euthanized and CONS extracted from urinary tracts to determine bacterial burden. As expected from our previous results, we recovered significantly more CONS from undiverted Mgb-/-than WT urine, bladders, and kidneys (Fig 6). WT bladders and kidneys consistently showed reduced bacterial burden relative to undiverted Mgb-/-, with 37.5% of WT organs demonstrating complete clearance of CONS compared to 0% in undiverted Mgb-/-mice. We also recovered significantly fewer CONS from urine, bladder, and kidneys of diverted Mgb-/-than from undiverted Mgb-/-animals. In the kidneys, bacterial burden was comparable between Mgb-/with CV and WT (p = 0.79 [right kidney] and p = 0.88 [left kidney]). However, bladder (p = 0.0009) and urine (p = 0.02) CONS levels remained increased in Mgb-/-versus WT regardless of CV status (Fig 6).

Discussion
COU is associated with increased risk and severity of UTI, and UTIs prolong hospitalization in patients with COU undergoing surgical repair. [2] The lack of animal models of spontaneous infection in the setting of COU has hampered the investigation of UTI risk factors in this patient population. We undertook this study in response to the anecdotal observation that Mgb-/-females develop spontaneous UTIs and struvite stones. Mgb-/-embryos demonstrate severely impaired detrusor muscle development, resulting in a highly compliant, atonic bladder in adults regardless of gender. [7] Adult Mgb-/-females dribble urine continuously, and gentle abdominal compression results in voids of one ml or more. Mgb-/-mice possess a large bladder volume, and the presence of urine at all times in the bladder argues that the animals never void to completion. Bladder ultrasound reveals nonspecific debris, as observed in human patients with urine stasis. The atonic, compliant Mgb-/-bladder parallels that found in patients with prune belly syndrome and non-neurogenic neurogenic bladder (Hinman syndrome). Such patients often develop complex UTIs, which are associated with unconventional uropathogens, prolonged hospitalization, and renal failure. [2,3,4] Spontaneous and experimental CONS UTI in Mgb-/-mice provides novel opportunities to model UTI and nephrolithiasis in the context of bladder dysfunction and urinary stasis. In previous efforts to model UTI in patients with urinary tract obstruction, investigators have transurethrally administered human uropathogenic bacterial isolates to rodent models of acute acquired urinary tract obstruction, such as spinal cord transection, partial unilateral ureteral obstruction, or transient bladder outlet obstruction. These studies have shown that acutely obstructed urinary tracts exhibit increased magnitude and duration of bacterial burden and tissue injury following experimental UTI, compared to unobstructed units. [16,17,18,19] Rats with spinal cord injury exhibit exquisite susceptibility to UTI following transurethral administration of low (10 2 CFU) bacterial inocula of uropathogenic E. coli, as well as increased bladder and renal inflammation. [16] In addition, acutely obstructed urinary tracts are uniquely susceptible to pyelonephritis following transurethral inoculation of bacterial strains that are non-pathogenic in unobstructed hosts, such as the laboratory K-12 Escherichia (E.) coli strain. [18,19]  Green dotted line (inset) demarcates the urothelial basement membrane. H&E, 5x objective; inset 20x objective (D) Mgb-/-kidney 14 dpi with CONS292 exhibits widespread changes including areas of dense mixed cellular infiltrate in the interstitium and tubules (inset1) as well as urothelial thickening, suburothelial inflammatory cells, and fibrosis (inset2). Nests of inflammatory cells are found scattered throughout the parenchyma (yellow ovals). Urothelial vacuolization is also present (orange circle, inset2). H&E, 5x objective; inset1 40x objective; inset2 20x objective. Proposed mechanisms for heightened sensitivity of acutely obstructed urinary tracts to infection include urine stasis and incomplete bladder emptying; bacteremia as a consequence of bacterial vesicoureteral reflux at high pressures; and altered mucosal immune response to uropathogenic bacteria. [16,17,18,19] Spinal cord injury attenuates the induction of bladder transcripts with roles in the innate immune response typically seen following E. coli experimental UTI, such as Il1b, Ltf, Cxcl3, and Tnf, raising important questions about the regulation of gene expression by local bladder dynamics. [17] The chronicity of urine stasis, congenital inhibition of urine flow due to detrusor maldevelopment, and susceptibility to spontaneous CONS UTI are unique features that set the Mgb-/-model apart from existing models of experimental UTI in the setting of acute, postnatal urinary tract obstruction.
We identified CONS as the bacteria responsible for struvite stone formation and UTIs in this model system. This conclusion is supported by the role of urease in CONS pathogenicity [20]; isolation of CONS from stone cultures; visualization of Gram-positive cocci in clusters within stone matrix; and ability of CONS to generate struvite stones and UTIs in previously sterile urinary tracts of Mgb-/-mice following transurethral inoculation. CONS replicates slowly in urine compared to Enterobacteriaceae, so urine stasis likely accounts in part for the ability of CONS to accumulate in Mgb-/-urine. [21] Accordingly, Mgb-/-mice exhibit unique susceptibility to transurethral inoculation of CONS292, developing persistent bacteriuria and struvite urolithiasis compared to WT animals.
Urinary stasis in Mgb-/-may lead to increased luminal CONS and elevated local NH 4 + concentrations via urease activity. The resulting alkaline pH favors precipitation of supersaturated calcium phosphate, together with magnesium. [15] While we previously identified non-absorbable suture as the nidus for struvite stones in male Mgb-/-mice following CV, [10] this study investigated spontaneous urolithiasis in Mgb-/-females without an antecedent surgery. The unique susceptibility of Mgb-/-hosts to CONS292 will enable studies to identify the nidus for stone formation in this setting, as well as to distinguish the roles of urease and other Staphylococcus virulence factors in UTI. [20,22,23] Since we postulate a central role for urine stasis in the ability of CONS to act as a uropathogen, we expected that relief of stasis by cutaneous vesicostomy would reduce susceptibility of Mgb-/-animals to experimental UTI. We found that CONS-inoculated, vesicostomized Mgb-/mice exhibited reduced renal bacterial burden, mirroring the clinical observation that urinary diversion protects the upper tract from infection. In contrast to the beneficial effect of CV on renal bacterial burden, bacterial burden in the Mgb-/-bladder remained significantly higher than WT controls despite CV. The impaired bladder clearance of CONS despite CV argues that local antimicrobial defenses are impaired in Mgb-/-bladders in a manner not entirely attributable to urinary stasis. Possible explanations for impaired bladder clearance of CONS in Mgb-/-following CV include: residual urinary stasis or altered bladder dynamics despite urinary tract decompression; increased bacterial attachment to urothelium; aberrant terminal differentiation of Mgb-/-urothelium leading to barrier disruption or epithelial dysfunction; impaired leukocyte recruitment; and/or poor production of antimicrobial peptides which rely upon regular bladder filling and emptying for their expression.

Conclusions
Mgb-/-female mice exhibit susceptibility to spontaneous and experimental UTI, due at least in part to urinary stasis. Furthermore, CONS acts as an opportunistic murine uropathogen in the setting of stasis. The Mgb-/-model of spontaneous and experimental CONS infection complements studies of superimposed infection in acquired obstructed uropathy such as spinal cord injury, transient bladder obstruction, and partial ureteral obstruction, allowing the dissection of host and bacterial factors that govern infection susceptibility.