Maternal antibiotic administration during gestation can affect the memory and brain structure in mouse offspring

Abstract

Maternal antibiotics administration (MAA) is among the widely used therapeutic approaches in pregnancy. Although published evidence demonstrates that infants exposed to antibiotics immediately after birth have altered recognition memory responses at one month of age, very little is known about in utero effects of antibiotics on the neuronal function and behavior of children after birth. Therefore, this study aimed to evaluate the impact of MAA at different periods of pregnancy on memory decline and brain structural alterations in young mouse offspring after their first month of life. To study the effects of MAA on 4-week-old offspring, pregnant C57BL/6J mouse dams (2-3-month-old; n = 4/group) were exposed to a cocktail of amoxicillin (205 mg/kg/day) and azithromycin (51 mg/kg/day) in sterile drinking water (daily/1 week) during either the 2nd or 3rd week of pregnancy and stopped after delivery. A control group of pregnant dams was exposed to sterile drinking water alone during all three weeks of pregnancy. Then, the 4-week-old offspring mice were first evaluated for behavioral changes. Using the Morris water maze assay, we revealed that exposure of pregnant mice to antibiotics at the 2nd and 3rd weeks of pregnancy significantly altered spatial reference memory and learning skills in their offspring compared to those delivered from the control group of dams. In contrast, no significant difference in long-term associative memory was detected between offspring groups using the novel object recognition test. Then, we histologically evaluated brain samples from the same offspring individuals using conventional immunofluorescence and electron microscopy assays. To our knowledge, we observed a reduction in the density of the hippocampal CA1 pyramidal neurons and hypomyelination in the corpus callosum in groups of mice in utero exposed to antibiotics at the 2nd and 3rd weeks of gestation. In addition, offspring exposed to antibiotics at the 2nd or 3rd week of gestation demonstrated a decreased astrocyte cell surface area and astrocyte territories or depletion of neurogenesis in the dentate gyrus and hippocampal synaptic loss, respectively. Altogether, this study shows that MAA at different times of pregnancy can pathologically alter cognitive behavior and brain development in offspring at an early age after weaning.

Keywords: antibiotics; hippocampal structure; memory and learning; myelination; neurogenesis; offspring mice.

FIGURE 1

Scheme of the experiment. Pregnant C57BL/6J mice (between 2 and 3-month-old) were divided into three groups (4 females/group) according to the treatment procedure: (I) sterile drinking water over the entire gestation period (control); (II) sterile drinking water contained a cocktail of antibiotics (amoxicillin and azithromycin) during the 2nd week (8-14 days) of gestation (2-w antibiotics); (III) consumed antibiotics from the 15th day of gestation to delivery (3-w antibiotics). In the present study, offspring mice born from those dams were investigated. In doing so, the first cohort of the animals of all three experimental groups performed behavioral tests during the 5th week of life with further morphological analysis of their brains. There also assessed the permeability of the intestinal and blood-brain barriers in mice from the second cohort using Evans Blue extravasation method.

FIGURE 2

Spatial memory and learning in young mice born to antibiotic-treated C57BL/6J females. (A–E) Morris water maze test: (A) latency to find a hidden platform, (B) path length, and (C) mean speed during the acquisition task, as well as (D) time in the target quadrant of the water pool and (E) heat maps with animals’ trajectories in the probe test. Dotted line circle on the heat map corresponds with the position of the hidden platform. Data on panels (A–C) are expressed as the Median with IQR, while data on graph 2D are expressed as the Mean ± SEM and presented in the form of dot plots (n = 13-16/group). Comparisons among groups were performed with either the one-way ANOVA followed by Dunnett’s post hoc test (data with Gaussian distribution) or the Kruskal-Wallis test followed by Dunn’s post hoc test (data with non-parametric distribution). *p < 0.05, **p < 0.01, and ***p < 0.001.

FIGURE 3

Hippocampal CA1 pyramidal neurons of the experimental mice. (A,B) Immunohistochemical images of NeuN⁺ neurons in the CA1 stratum pyramidale at (A) 20× and (B) 40× objective magnifications. (C) Methylene blue-stained semi-thin sections of the stratum pyramidale used for light microscopic analysis (40x objective magnification). (D) The number of NeuN⁺ cells per mm² of the pyramidal layer. (E) The number of intact (normochromic) neurons per mm² of the pyramidal layer on semi-thin sections. (F,G) Correlational analysis between the density of NeuN⁺ cells and (F) average latency, (G) average path length during the acquisition task in MWM. Data on panel (D) are expressed as the Mean ± SEM and presented in the form of dot plots, whereas data on panel (E) are expressed as the Median with IQR and visualized using box and whisker plots (n = 6/group). Comparisons among groups were performed with either the one-way ANOVA followed by Dunnett’s post-hoc test (data with Gaussian distribution) or the Kruskal–Wallis test followed by Dunn’s post hoc-test (data with non-parametric distribution). *p < 0.05 and ***p < 0.001.

FIGURE 4

Glial cells in the hippocampus of young offspring animals. (A) Immunofluorescent-labeled images of the CA1 stratum pyramidale and stratum radiatum at 20× objective magnification: red—NeuN⁺ neurons, green—Iba1⁺ microglia, magenta—GFAP⁺ astrocytes. (B) Images of individual glial cells at higher objective magnification (40×) displaying fine morphology of cellular somas and processes: magenta—GFAP⁺ astrocytes, green—Iba1⁺ microglia. (C–I) Morphometric indicators of hippocampal glia: the number density of (C) astrocytes and (G) microglial cells per mm² of brain tissue, the cell surface area of (D) astrocytes and (H) microglia, (E) astrocyte territory, as well as the mean number of (F) astrocytic and (I) microglial base processes. Data on graphs 4E/F are expressed as the Median with IQR and visualized using box and whisker plots, while data on other graphs are expressed as the Mean ± SEM and presented in the form of dot plots (n = 13–16/group). Comparisons among groups were performed with either the one-way ANOVA followed by Dunnett’s post-hoc test (data with Gaussian distribution) or the Kruskal-Wallis test followed by Dunn’s post-hoc test (data with non-parametric distribution). *p < 0.05, ^**p < 0.01, and ^***p < 0.001.

FIGURE 5

Neurogenesis in the dentate gyrus of C57BL/6J mice born to antibiotic-treated dams. (A) Immunofluorescent-labeled images of DCX⁺ neuroblasts (green) and BrdU⁺ proliferating cells (red) at 20× objective magnification. We used crops of DCX⁺ cells denoted by the yellow dotted line for morphometric analysis. (B) The integrated density of DCX⁺ pixels per mm² of the subgranular zone and stratum granulosum. (C) The relative proliferation rate of neuroblasts (BrdU/DCX ratio). Data are expressed as the Mean ± SEM (n = 13–16/group). Comparisons among groups were performed with the one-way ANOVA followed by Dunnett’s post-hoc test. *p < 0.05 and ^***p < 0.001.

FIGURE 6

Structural synaptic plasticity in the hippocampal CA1 area of young offspring mice. (A) Electron micrographs of the stratum radiatum (×10,000). Synapses are marked in pink. (B–F) Morphometric characteristics of the neuropil: the number density of (B) synaptic contacts and their different morphological types—(C) simple, (D) perforated, and (E) multiple ones per 100 μm² of brain tissue, as well as the (F) postsynaptic density length. Data are expressed as the Mean ± SEM (n = 6/group). Comparisons among groups were performed with the one-way ANOVA followed by Dunnett’s post-hoc test. *p < 0.05.

FIGURE 7

Myelination in the corpus callosum of experimental animals. (A) Electron micrographs of myelinated and unmyelinated axons in the corpus callosum (×4,800). Enlarged crops framed in yellow provide a more detailed morphology of myelin sheaths. (B) The density of myelinated axons per 100 μm² of brain tissue. (C) Mean diameter of myelinated axons. (D) Electron micrographs of the corpus callosum pseudocolorized in MyelTracer software for quantification of the G-ratio: light green—axons, dark green—myelin (×6,400). (E) G-ratio. (F–H) Correlational analysis between G-ratio and (F) average latency during the acquisition task in MWM, (G) average path length in MWM, (H) the density of NeuN⁺ neurons in the hippocampus. Data are expressed as the Mean ± SEM (n = 6/group). Comparisons among groups were performed with the one-way ANOVA followed by Dunnett’s post-hoc test. *p < 0.05, ^**p < 0.01, and ^***p < 0.001.