Androgen exposure impairs neutrophil maturation and function within the infected kidney

Abstract

Urinary tract infections (UTIs) in men are uncommon yet carry an increased risk for severe pyelonephritis and other complications. In models of Escherichia coli UTI, C3H/HeN mice develop high-titer pyelonephritis (most with renal abscesses) in a testosterone-dependent manner, but the mechanisms underlying this phenotype are unknown. Here, using female mouse models, we show that androgen exposure impairs neutrophil maturation in the upper and lower urinary tract, compounded by a reduction of neutrophil function within the infected kidney, enabling persistent high-titer infection and promoting abscess formation. Following intravesical inoculation with uropathogenic E. coli (UPEC), kidneys of androgen-exposed C3H mice showed delayed local pro-inflammatory cytokine responses while robustly recruiting neutrophils. These were enriched for an end-organ-specific population of aged but immature neutrophils (CD49d+, CD101-). Compared to their mature counterparts, these aged immature kidney neutrophils exhibited reduced function in vitro, including impaired degranulation and diminished phagocytic activity, while splenic, bone marrow, and bladder neutrophils did not display these alterations. Furthermore, aged immature neutrophils manifested little phagocytic activity within intratubular UPEC communities in vivo. Experiments with B6 conditional androgen receptor (AR)-deficient mice indicated rescue of the maturation defect when AR was deleted in myeloid cells. We conclude that the recognized enhancement of UTI severity by androgens is attributable, at least in part, to local impairment of neutrophil maturation in the urinary tract (largely via cell-intrinsic AR signaling) and a kidney-specific reduction in neutrophil antimicrobial capacity.IMPORTANCEAlthough urinary tract infections (UTIs) predominantly occur in women, male UTIs carry an increased risk of morbidity and mortality. Pyelonephritis in androgen-exposed mice features robust neutrophil recruitment and abscess formation, while bacterial load remains consistently high. Here, we demonstrate that during UTI, neutrophils infiltrating the urinary tract of androgen-exposed mice exhibit reduced maturation, and those that have infiltrated the kidney have reduced phagocytic and degranulation functions, limiting their ability to effectively control infection. This work helps to elucidate mechanisms by which androgens enhance UTI susceptibility and severity, illuminating why male patients may be predisposed to severe outcomes of pyelonephritis.

Keywords: Escherichia coli; immune response; neutrophils; pyelonephritis; sex differences; urinary tract infection.

Fig 1

Androgen exposure enables chronic UTI with continuous neutrophil recruitment. (A and B) Timeline of bladder and kidney bacterial loads after UPEC inoculation in vehicle-treated (open triangles) and androgenized (filled triangles) C3H/HeN mice. Lines indicate geometric mean. (C) CD45+ cell recruitment to the kidney over time as a percentage of live cells, by flow cytometry. (D) Neutrophil (CD45+, Ly6G+) recruitment to the kidney over time as a percentage of CD45+ cells. (E and F) Timeline of the number of CD45+ cells (E) and neutrophils (F) in the kidneys throughout the course of infection, expressed per 1 million events, in vehicle-treated (white boxes) or androgenized mice (gray boxes). Bars indicate the mean with SEM. Each symbol represents a single mouse; n = 5–15 per condition. *P < 0.05, **P < 0.01, ***P < 0.001 by Mann-Whitney U test. CFU, colony-forming units. ns, not significant.

Fig 2

Production of neutrophil recruiting cytokines in the kidney is intact in androgenized mice. Timeline of whole-kidney levels of the indicated cytokines in vehicle-treated (white bars, open triangles) and androgenized mice (gray bars, filled triangles). Bars indicate the mean with SEM. Each symbol represents a single mouse; n = 5 per time point. *P < 0.05, **P < 0.01 by Mann-Whitney U test. ns, not significant.

Fig 3

UPEC-infected kidneys in androgenized mice harbor a distinct population of aged immature neutrophils. Representative gating of neutrophils (CD45+, Ly6G+) for age (CD49d) and maturity (CD101) in kidneys (A and B) or peripheral blood (C and D) from vehicle-treated and androgenized mice 14 dpi. (E and F) Timeline of the proportion of neutrophils of each subtype (young immature [blue bars], young mature [red], aged immature [green], aged mature [purple]) throughout the course of infection in vehicle-treated and androgenized mice. Bars indicate the mean with SEM. n = 5–15 per condition and time point. *P < 0.05, **P < 0.01 by Mann-Whitney U test.

Fig 4

Inhibition of AR signaling in neutrophils substantially restores their maturation in the infected kidney. (A) The proportion of neutrophils that were immature (CD101–) or mature (CD101+) 7 dpi in vehicle-treated Cre^–AR^f/f (open triangles, white bars), androgenized Cre^–AR^f/f (closed triangles, gray bars), androgenized Ksp-Cre × AR^f/f (squares, green bars), or androgenized LysM-Cre × AR^f/f (circles, pink bars) C57BL/6 mice. (B) The proportion of neutrophils that were young immature, young mature, aged immature, and aged mature in vehicle-treated Cre^–AR^f/f, androgenized Cre^–AR^f/f, androgenized Ksp-Cre × AR^f/f, or androgenized LysM-Cre × AR^f/f C57BL/6 mice 7 dpi (symbols as in panel A). Bars indicate the mean with SEM. (C–F) Representative pseudocolor plots of neutrophil age (CD49d) and maturity (CD101) by flow cytometry in vehicle-treated Cre^–AR^f/f, androgenized Cre^–AR^f/f, androgenized Ksp-Cre × AR^f/f, or androgenized LysM-Cre × AR^f/f C57BL/6 mice. n = 15–26 per mouse strain. *P < 0.05, **P < 0.01, ***P < 0.001 by Mann-Whitney U test. Comparisons with P values >0.05 are not indicated.

Fig 5

Kidney neutrophils in androgenized mice exhibit reduced phagocytic capacity. (A) Percentage of neutrophils that phagocytosed GFP+ UPEC after isolation from the spleen, bone marrow, or kidney in the naïve state, or the kidney 7 dpi, in vehicle-treated (open triangles) or androgenized mice (filled triangles). (B and C) Percentage of young immature, young mature, aged immature, and aged mature neutrophils isolated from naïve or 7-dpi kidneys in vehicle-treated or androgenized mice. (D) Representative immunofluorescence localization of neutrophil subtypes near and within kidney bacterial communities of an androgenized mouse 10 dpi. White, E. coli; red, Ly6G (neutrophils); green, CD49d (age); blue, CD101 (maturity). E, epithelium; L, tubular lumen. Arrows indicate different neutrophil subtypes: YM, young mature; AI, aged immature; AM, aged mature. Bars indicate the mean with SEM. Each symbol represents a single mouse; n = 5–10 per group. *P < 0.05, **P < 0.01, ***P < 0.001 by Mann-Whitney U test. ns, not significant.

Fig 6

Degranulation by kidney neutrophils is blunted and delayed in androgenized mice. Release of secretory vesicles (CD18+ CD11b+; A–E) and primary granules (CD63+; F–J), expressed as a proportion of all neutrophils (A, F) or of each neutrophil subtype at the indicated time points post UPEC infection in vehicle-treated (open triangles) or androgenized mice (filled triangles). Bars indicate the mean with SEM. Each symbol represents a single mouse; n = 5–10 per time point. *P < 0.05, **P < 0.01 by Mann-Whitney U test. ns, not significant.

Fig 7

Neutrophils in the UPEC-infected bladder do not share the functional deficits seen in the kidney. (A) Distribution of neutrophil age and maturity subtypes in the bladders of vehicle-treated (open triangles, white bars) or androgenized mice (filled triangles, gray bars) 1 dpi as a proportion of total neutrophils. (B) Absolute count of each neutrophil age and maturity subtype per million events in the bladder 1 dpi. (C) Phagocytosis of GFP+ UPEC by bladder neutrophils harvested 1 dpi. (D) The proportion of bladder neutrophils exhibiting degranulation activity 1 dpi. (E and F) The proportion of bladder neutrophils 1 dpi from each age and maturity subtype that had released secretory vesicles (E) or primary granules (F). Each symbol represents a single mouse; n = 5–10 per group. **P < 0.01, ***P < 0.001 by Mann–Whitney U-test. ns, not significant.