Effect of ceftriaxone on the outcome of murine pyelonephritis caused by extended-spectrum-β-lactamase-producing Escherichia coli

Abstract

Urinary tract infections (UTIs) due to extended-spectrum-β-lactamase (ESBL)-producing Enterobacteriaceae in children are becoming more frequent, and they are commonly treated initially with a second- or third-generation cephalosporin. We developed a murine model of ascending UTI caused by ESBL-producing Escherichia coli. Using this model, we investigated the renal bacterial burden, interleukin-6 (IL-6) expression, and histopathological alterations caused by ESBL- and non-ESBL-producing bacteria after 1, 2, or 6 days with or without ceftriaxone therapy. The renal bacterial burden, IL-6 concentration, and histological inflammatory lesions were not significantly different between mice infected with ESBL- and non-ESBL-producing bacteria without treatment at any of the time points examined. Following ceftriaxone administration, the bacterial burden was eliminated in the kidneys of mice infected with ESBL- and non-ESBL-producing bacteria on the 6th postinfection day. The histological analysis demonstrated that among mice treated with ceftriaxone, those infected with ESBL-producing bacteria had more profound renal alterations than those infected with non-ESBL-producing bacteria on the 6th day (P < 0.001). In comparison, microbiological outcomes did not differ significantly between mice infected with ESBL- and non-ESBL-producing bacteria at any of the time points examined. The effectiveness of ceftriaxone in mice with UTIs due to ESBL-producing E. coli may have therapeutic implications; it is, however, hampered by limited activity on the histopathological lesions, a finding that needs further investigation.

FIG 1

Impact of ceftriaxone (CRO) treatment on the renal bacterial burden of mice infected with ESBL- and non-ESBL-producing bacteria. Mice inoculated with ESBL-producing (E94, n = 22) and non-ESBL-producing (BRA, n = 22) bacteria were treated with 2.5 mg CRO at 9 h postinfection and every 12 h thereafter until the end of the study period. Viable cell counts in the kidneys (CFU/g) are shown for CRO-treated mice infected with non-ESBL- and ESBL-producing bacteria for day 1 (d1), day 2 (d2), and day 6 (d6). Values are expressed as means ± standard errors. The statistical significance between columns is denoted by a bar and an asterisk (P < 0.01).

FIG 2

Comparison of the effect of ceftriaxone (CRO) treatment versus no antibiotic on the renal bacterial burden of mice infected with ESBL- and non-ESBL-producing bacteria. Mice inoculated with ESBL-producing (E94, n = 22) and non-ESBL-producing (BRA, n = 22) bacteria were treated with ceftriaxone, and 21 mice received no antibiotic. Viable cell counts in the kidneys (CFU/g) are shown for untreated and CRO-treated mice infected with non-ESBL- and ESBL-producing bacteria for day 1, day 2, and day 6. Values are expressed as means ± standard errors. The statistical significance between columns is denoted by a bar and an asterisk (P < 0.01).

FIG 3

Impact of ceftriaxone (CRO) treatment on renal IL-6 concentration in mice infected with ESBL- and non-ESBL-producing bacteria. The supernatants of homogenized kidneys explanted from CRO-treated mice infected with ESBL-producing (E94, n = 22) and non-ESBL-producing (BRA, n = 22) bacteria were assessed for the production of IL-6 using ELISAs. IL-6 production in CRO-treated mice infected with non-ESBL- and ESBL-producing bacteria for day 1, day 2, and day 6 are shown. Levels of IL-6 release are expressed as means ± standard errors in pg/g. The statistical significance between columns is denoted by a bar and an asterisk (P < 0.01).

FIG 4

Photomicrographs of H&E-stained renal sections of mice infected with ESBL-producing (A1, B1, and C1) and non-ESBL-producing (A2, B2, and C2) bacteria, at different time points, without antibiotic treatment. (A1 and A2) First day postinfection. (A1) Inflammatory lesion score 1. Neutrophils within the renal pelvic cavity (left arrow) and medulla (right arrow) are shown. (A2) Inflammatory lesion score 3. Numerous neutrophils within the pelvic cavity (right arrow) and extensive inflammatory cellular infiltration of the medullar parenchyma (left arrow) are shown. (B1 and B2) Second day postinfection. (B1) Inflammatory lesion score 1. Hemorrhage is shown with moderate numbers of neutrophils within the renal pelvic cavity (arrow). (B2) Inflammatory lesion score 2. Numerous neutrophils within the renal pelvic cavity (arrow) are shown. (C1 and C2) Sixth day postinfection. (C1) Inflammatory lesion score 3. Congestion of neutrophils within the renal pelvic cavity (left arrow) is shown. Extensive inflammatory cellular infiltration and tubular destruction in the medulla (right arrow) are shown. (C2) Inflammatory lesion score 1. Hemorrhage is shown with neutrophils within the renal pelvic cavity (arrow) and disrupted uroepithelium.

FIG 5

Linear correlation between the microbial burden and the histopathology score of ESBL- and non-ESBL-producing strains with and without ceftriaxone (CRO).

FIG 6

Histopathological alterations defined as inflammatory lesion scores induced by ESBL-producing or non-ESBL-producing E. coli strains at the 1st, 2nd, and 6th postinfection day. All mice infected with ESBL- and non-ESBL-producing strains were treated with ceftriaxone. Lesion scores of histopathological alterations are defined in Materials and Methods. An asterisk denotes a significant difference between non-ESBL and ESBL producers at day 6.

FIG 7

Box plots of histopathology scores for the 2nd and 6th postinfection day of ESBL-producing strain E94 and non-ESBL-producing strain BRA with and without ceftriaxone (CRO). By Mann-Whitney U test significant difference between the scores of mice infected with ESBL- and non-ESBL-producing bacteria and treated with CRO on the 6th postinfection day is indicated by an asterisk.

FIG 8

Photomicrographs from pairs of consecutive renal sections of mice infected with ESBL-producing bacteria with (B1 and B2) or without (A1 and A2) ceftriaxone (CRO) treatment. In the left panels, sections are stained with H&E (A1 and B1). In the right panels, sections are immunohistochemically labeled for IL-6 (A2 and B2). Numerous inflammatory cells (mainly neutrophils) infiltrating the tubules of renal medulla (arrows in A1) are positive for IL-6 (arrows in A2). Inflammatory cells located within the renal medulla and uroepithelium, adjacent to the pelvis (arrows in B1), are immunoreactive for IL-6 (arrows in B2).