11 - Scale development to enhance the validity of in-vivo UTI research

Poster Number: 11
Publication Number: 11.1098

Background: The mouse model of urinary tract infection (UTI) is commonly-used in pre-clinical UTI research. However, there are limitations to this model, including variability in the resulting phenotypes and the reliance on bacteriuria as an outcome of interest in some studies. The ability to obtain tissue for histology, while a gold-standard outcome, is not feasible in some study designs. The development of a non-invasive method scale to both validate the presence of UTI and determine its severity would improve the validity of preclinical UTI research.

Objective: To develop a non-invasive scale that correlates with both the presence and severity of UTI in mice.

Design/Methods: We used transurethral catheterization and instillation of 10^8 colony-forming units of E. coli UTI89 in 9 female C57BL/6 mice. We assessed the mice in 3 domains (behavior, reactions to abdominal and flank palpation and patterns, and voiding behaviors) at baseline, 24, 48, and 72 hours post infection. Voiding behavior was assessed by void spot assays. We collected urine at baseline and then for 3 days for bacterial culture and urinary neutrophil gelatinase-associated lipocalin (NGAL) measurement. We sacrificed mice at 72 hours post instillation and harvested kidneys and bladder. Kidneys were bisected, with half of each kidney used for culture and half used for histology. Sections of the bladder and both kidneys were scored on a 6-point Likert scale with regard to severity of infection by a blinded veterinary pathologist.

Results: The pathology scores of the infected mice are as follows: normal tissue, n=2; mild inflammation, n=2; moderate inflammation, n=3; severe inflammation, n=2. The percent of mice with abnormal findings in each domain are shown in table 1. Urinary NGAL and voiding behaviors were the two domains that were abnormal in 100% of mice with severe pathology. On closer analysis, a 48-hour NGAL demonstrating area under the receiver operating curve of 1.0 for severe pathology. When we examined voiding behavior more closely, we discovered that total number of spots greater than 0.5 microliters had a dose-response relationship with severity from normal to moderate, but that this dropped off after moderate, likely due to severity of illness and associated dehydration or acute kidney injury in the severe group. (Figure 1)

Conclusion(s): There is significant variability in resulting pathology following a standardized protocol to induce UTI in C57Bl/6 mice. Urinary NGAL and voiding behavior, specifically the number of total spots, may help identify mice with moderate and severe inflammation on pathology.