17β-estradiol ameliorates delirium-like phenotypes in a murine model of urinary tract infection

Abstract

Urinary tract infections (UTIs) are common and frequently precipitate delirium-like states. Advanced age coincident with the postmenopausal period is a risk factor for delirium following UTIs. We previously demonstrated a pathological role for interleukin-6 (IL-6) in mediating delirium-like phenotypes in a murine model of UTI. Estrogen has been implicated in reducing peripheral IL-6 expression, but it is unknown whether the increased susceptibility of postmenopausal females to developing delirium concomitant with UTIs reflects diminished effects of circulating estrogen. Here, we tested this hypothesis in a mouse model of UTI. Female C57BL/6J mice were oophorectomized, UTIs induced by transurethral inoculation of E. coli, and treated with 17β-estradiol. Delirium-like behaviors were evaluated prior to and following UTI and 17β-estradiol treatment. Compared to controls, mice treated with 17β-estradiol had less neuronal injury, improved delirium-like behaviors, and less plasma and frontal cortex IL-6. In vitro studies further showed that 17β-estradiol may also directly mediate neuronal protection, suggesting pleiotropic mechanisms of 17β-estradiol-mediated neuroprotection. In summary, we demonstrate a beneficial role for 17β-estradiol in ameliorating acute UTI-induced structural and functional delirium-like phenotypes. These findings provide pre-clinical justification for 17β-estradiol as a therapeutic target to ameliorate delirium following UTI.

Figure 1

Oophorectomy (OVX) and urinary tract infection (UTI)-mediated changes in cleaved caspase-3 (CC3). (A) A schematic of design of a preliminary experiment examining the effects of OVX, UTI, and OVX + UTI on levels of cortical and hippocampal CC3. (B) Representative regions of interest (ROIs) in the frontal cortex and hippocampus. (C) Representative staining of CC3 and neuronal nuclear protein (NeuN) in each group (D,E) Quantification of CC3 in the frontal cortex and hippocampus demonstrates a statistically significant increase in CC3 in mice with UTIs alone [sham (green), n = 6] compared to mice with OVX alone (blue, n = 6). Additionally, among mice with UTIs, OVX animals (red, n = 6) had significantly higher levels of CC3 compared to the sham group (green). (F) Regression analysis across groups demonstrates a statistically significant relationship between frontal cortex and hippocampal CC3 levels. Data are expressed as mean ± SD. Dotted lines represent 95% confidence intervals. *p < 0.05, ***p < 0.001, ****p < 0.0001.

Figure 2

17β-estradiol (E2) significantly ameliorates UTI-induced delirium-like behavior. (A) Schematic of primary experimental design and treatment schedule of E2. Note that to prevent any habituation, novel object recognition (NOR) and Y-maze behavioral tests were only utilized post UTI induction and E2 (or vehicle) treatment. At a given time point behavioral tests were performed sequentially. NOR-T represents training the training phase of the behavioral test. (B) Distance (m) and time immobile (sec) as a percent of baseline (initial 5 min). Prior to UTI induction and E2 treatment (OVX + Pre-UTI), there were no treatment effects but a statistically significant effect of time over the course of the 45-min open field test (OF); decreased distance travelled (F_3,110 = 22.2; p < 0.0001) and increased time spent immobile (F_5,171 = 14.8; p < 0.0001). Following UTI (OVX + Post-UTI) there was statistically significant time-by-E2 treatment interactions for distance traveled (F_24,496 = 1.9; p = 0.0052) and time spent immobile (F_24,448 = 3.0; p < 0.0001). Results of post-hoc tests, comparing treatment effects (vehicle vs. E2) at individual time points, are indicated by asterixis. (C) Among oophorectomized and UTI inoculated mice (OVX + Post-UTI), mice treated with E2 spent statistically significantly less time in the periphery of the open field. (D–F) Mice (OVX + Post-UTI) treated with E2 spent statistically significantly more time in the open arm (elevated plus maze, EPM), more time and oriented towards a novel object (novel object recognition), and more % spontaneous alterations (Y-maze) compared to the vehicle treated group. Following oophorectomy, but prior to UTI induction and E2 treatment (OVX + Pre-UTI), there were no statistically significant behavioral differences between groups. Group size was n = 15. Data are expressed as mean ± SD. *p < 0.05, **p < 0.01, ****p < 0.0001.

Figure 3

17β-estradiol (E2) decreases UTI-induced cleaved caspase-3 (CC3) in the frontal cortex and hippocampus. (A,B) Among oophorectomized mice with UTIs (OVX + UTI), those treated with E2 (n = 24) show a statistically significant reduction in CC3 in both the frontal cortex and hippocampus as compared to vehicle (sesame oil) treated (n = 24). (C) Representative staining of CC3 and neuronal nuclear protein (NeuN) in both vehicle and E2 treated mice. For (A), a single statistical outlier was identified and removed from the E2 group. Exclusion of the data point yielded p = 0.0072 (t = 2.814, df = 45), while inclusion resulted in p = 0.0352 (t = 2.170, df = 46) (see Fig. S7a) *p < 0.05, **p < 0.01.

Figure 4

Cleaved caspase-3 (CC3) variability explains delirium-like behaviors. (A) Regression analysis across groups (n = 21–30) demonstrates a statistically significant positive relationship between frontal cortex cleaved caspase-3 (CC3) and time spent in the periphery during an open field test (OF). (B) Regression analysis reveals a significant negative relationship between hippocampal CC3 and time oriented towards a novel object during a novel object recognition test (NOR). Veh = vehicle treated; E2 = 17β-estradiol treated. Dotted lines represent 95% confidence intervals.

Figure 5

17β-estradiol (E2) attenuates plasma and frontal cortex concentrations of IL-6 but not soluble IL-6 receptor alpha (sIL-6R). (A–C) Among ovariectomized and UTI inoculated mice (OVX + UTI), individuals treated with E2 (7 µg) have significantly lower IL-6 plasma concentrations in the plasma and frontal cortex with no significant change in sIL-6R concentrations as measured via ELISA. Plasma samples (n = 15) could reliably be obtained from all but 3 vehicle and 2 E2 animals. For (B), plasma IL-6R was measured in a randomly chosen subset of animals (n = 10/group) to confirm differences in delirium-like phenotypes were not driven by receptor density. For (A), a single statistical outlier was identified and removed from the E2 group – the abnormally high value may have resulted from hemolysis of the sample. Exclusion of the data point yielded p = 0.0495 (t = 2.102, df = 18.45), while inclusion resulted in p = 0.1654 (t = 1.430, df = 24.66) (Fig. S7). (D,E) Regression analyses of plasma IL-6 (pg/ml) on cleaved caspase-3 (CC3) and IL-6 levels in the frontal cortex. Dotted lines represent 95% confidence intervals. Group size was n = 10–18. Data are expressed as mean ± SD. *p < 0.05, **p < 0.01.

Figure 6

17β-estradiol (E2) reverses UTI-mediated neuronal changes. (A) Representative staining of primary rat cortical neurons 4 days in vitro (DIV) showing expression of two neuron-specific markers, neuronal nuclear protein (NeuN, red) and neurofilament-light chain (NF-L, green). (B) At 7 DIV, cortical neurons were treated for 1 h with either vehicle (growth medium), a combination of IL-6/sIL-6R, or combination of IL-6/sIL-6R plus E2. Representative western blot images of relevant signaling molecules that contribute to apoptosis, neurogenesis, and synaptic activity. Full unedited gel images are provided in supplementary information (Fig. S6). (C) Summary plot for fold changes relative to vehicle are presented for the indicated signaling molecules. Neurons treated with IL-6/sIL-6R had statistically significantly more cleaved caspase-3 (CC3) and p-p38 MAP kinase (MAPK) compared to the vehicle and those treated with combination of IL-6/sIL-6R plus E2 (n = 3). There was no significant difference in CC3 or p-p38 MAPK levels between the vehicle and neurons treated with IL-6/sIL-6R plus E2 (n = 3). Neurons treated with treated with combination of IL-6/sIL-6R plus E2 had significantly more p-AKT than the vehicle or those treated with only IL-6/sIL-6R (n = 3). IL-6 was always administered with sIL-6R since neurons do not express the IL-6 receptor. As an additional control we quantified CC3 in the presence of IL-6 alone and found no difference compared to vehicle (Fig. S8). Data are expressed as mean ± SD. *p < 0.05.

Figure 7

Hypothesized schematic of the IL-6 trans-signaling pathway and protective effects of 17β-estradiol. UTI-induced increases in IL-6 and soluble IL-6R (sIL-6R) complex in the periphery and directly induce neuronal dysfunction (i.e., increased frontal cortex and hippocampal CC3) via the gp130 transmembrane protein. 17β-estradiol reduces systemic and central inflammation and provides protective effects via reduction in neuronal CC3 mediated through neuronal G-protein estrogen receptor. Figure is adapted from Hogg et al.. Created with BioRender.com.