Chronic Bumetanide Infusion Alters Young Neuron Morphology in the Dentate Gyrus Without Affecting Contextual Fear Memory

Abstract

Young neurons in the adult brain are key to some types of learning and memory. They integrate in the dentate gyrus (DG) of the hippocampus contributing to such cognitive processes following timely developmental events. While experimentally impairing GABAergic transmission through the blockade or elimination of the ionic cotransporter NKCC1 leads to alterations in the proper maturation of young neurons, it is still unknown if the in vivo administration of common use diuretic drugs that block the cotransporter, alters the development of young hippocampal neurons and affects DG-related functions. In this study, we delivered chronically and intracerebroventricularly the NKCC1 blocker bumetanide to young-adult rats. We analyzed doublecortin density and development parameters (apical dendrite length and angle and dendritic arbor length) in doublecortin positive neurons from different subregions in the DG and evaluated the performance of animals in contextual fear learning and memory. Our results show that in bumetanide-treated subjects, doublecortin density is diminished in the infra and suprapyramidal blades of the DG; the length of primary dendrites is shortened in the infrapyramidal blade and; the growth angle of primary dendrites in the infrapyramidal blade is different from control animals. Behaviorally, treated animals showed the typical learning curve in a contextual fear task, and freezing-time displayed during contextual fear memory was not different from controls. Thus, in vivo icv delivery of bumetanide negatively alters DCX density associated to young neurons and its proper development but not to the extent of affecting a DG dependent task as aversive context learning and memory.

Keywords: GABA; diuretics; neurogenesis; neuronal maturation; pediatric epilepsy; transverse axis.

FIGURE 1

Nissl stained coronal sections from a bumetanide treated animal. (A) The micrograph shows a representative section where the infusion cannula (arrow) can be appreciated reaching the lateral ventricle. (B) A more posterior section corresponding to the region of the DG where the DCX analysis was performed is shown. The anatomy of the hippocampus is bilaterally similar. (C) A 10× magnification of the DG ipsilateral to the infusion shows a regular hippocampal anatomy. (D) A 40× magnification shows a regular granular cell layer with no evident pyknotic nuclei. Scale bars: (C) 800μm; (D) 50 μm.

FIGURE 2

Representative images and immunofluorescence (IF) density analysis of doublecortin in the dentate gyrus of bumetanide and propylene glycol treated animals. (A) The left panel corresponds to the control group treated with propylene glycol and the right panel, to the experimental group treated with bumetanide. Upper: Crest, middle: suprapyramidal blade, lower: infrapyramidal blade. Each image is a z-stack of 8–10 images. (B) The graph shows the doublecortin (DCX) IF density from binarized images in the three different analyzed regions: crest, infra and suprapyramidal blade. A statistical reduction in DCX density IF was found in the bumetanide group compared to the same areas in the control group for the infrapyramidal blade [t(10) = 2.08, p = 0.031] and for the suprapyramidal blade [t(10) = 1.95, p = 0.039] but not for the crest [t(10) = 1.27, p = 0.115); (C–E) Graphs show the doublecortin (DCX) density IF from binarized images in the subgranular zone, granular layer and molecular layer from the crest (C), the infrapyramidal blade (D) and the suprapyramidal blade (E). The bumetanide group shows a statistical reduction in DCX density IF in the granular [t(10) = 2.26, p = 0.023] and molecular layers [t = 2.53, p = 0.014] from the infrapyramidal blade and in the subgranular [t(10) = 1.89, p = 0.043] and molecular layers [t(10) = 2.36, P = 0.019] from the suprapyramidal blade; t-student. Mean ± SEM per group based on data obtained in 3–4 sections per subject; 2–3 z-stacks per field and 8–10 images per stack. Scale bar: 50 μm. *p < 0.05.

FIGURE 3

Dendritic skeletons from doublecortin positive cells in the dentate gyrus of bumetanide and propylene glycol treated animals. (A) Black and white images show z-stacks where observed dendrites were delineated (purple lines). A magnified manual tracing of a set of dendrites displayed in each black and white image is shown to the right of each image; dendritic tracings are shown in white. Images correspond to the crest (top); the suprapyramidal blade (middle) and the infrapyramidal blade (bottom). (B) Mean lengths of the dendritic arbor show no statistical differences between groups in any analyzed region (see text). (C) Mean lengths of the primary dendrites in the bumetanide treated group are significantly decreased in the infrapyramidal blade [t(10) = 2.31, p = 0.021] but not in the crest [t(10) = 1.69, p = 0.060 nor in the suprapyramidal blade t(10) = 1.58, p = 0.072] compared to the propylene glycol treated group. t-student. Mean ± SEM. Scale bars: 50 μm. *p < 0.05.

FIGURE 4

Dendritic growth angles in primary apical dendrites. (A) Representative images from doublecortin positive cells in the infrapyramidal layer. Black and white images show z-stacks from where dendrites were delineated (purple lines). A magnified manual tracing of a set of dendrites displayed in each black and white image is shown below each image; dendritic tracings are shown in gray and degrees of growth of two dendrites with respect to the granular zone (marked with a yellow line) are depicted in red. Notice that growth angles are predominantly straight in the control group and slanted in the bumetanide group. (B) Mean sexagesimal degrees from primary apical dendrites in the bumetanide treated group are significantly different from the control group in the infrapyramidal blade [t(10) = 2.03, p = 0.034], but not in the crest nor the suprapyramidal blade [t(10) = 1.15, p = 0.137; and t(10) = 0.73, p = 0.238, respectively]. t-student; mean ± SEM. Scale bars: 50 μm. *p < 0.05.

FIGURE 5

Performance in the open field and in contextual fear learning and memory. (A) Bars show the mean crossings for central and peripheral squares. There were no differences between groups for the preference to enter central [t(12) = 0.41, p = 0.690] or peripheral areas [t(12) = 0.95, p = 0.360]. Mean ± SEM; t-test. (B) Contextual fear memory learning curve in a 12 min session. Shocks were randomly delivered after two min of introducing the animal in the chamber. During conditioning both groups reached a maximum of 70% immobility time showing that treatment did not affect their learning capacity. No significant differences along the learning process were observed. (C) Contextual fear memory. Time of immobility was similarly high between groups during the recall session showing that animals from both groups remembered the aversive context after 24 h of learning [t(12) = 0.26, p = 0.318). Mean ± SEM; t-test.