The animals were treated in accordance with the guidelines published by the NIH (Guide for the Care and Use of Laboratory Animals), and all experimental procedures were approved prior to study initiation by the Ethics Committee on Experimental Animal Research (Institute of Biological Sciences, Federal University of Pará, Brazil).
Animals and experimental groups
In this study, 66 mice were housed in groups of 6 in standard plastic cages (32 × 39 × 16.5 cm) covered by metal grids until they were sacrificed. All behavioural assays were performed using 3-, 6- and 18-month old mice. The reduced mastication condition began on the 21st post-natal day by the introduction of a powder diet (SD mice); control mice (HD mice) were fed a pellet diet. Mice had free access to food and water and were raised under controlled room temperature (23 ± 1°C) and 12-h light-dark cycles. There were no significant differences in body weight in the SD and HD experimental groups (measured the day of sacrifice), suggesting that the nutritional value of SD and HD was the same.
Compared to the groups of young mice (3- and 6-month-old mice), aging was associated with significant number of deaths. Although we have not found any significant differences in body weight or behaviour before deaths there was greater mortality in the 18-month-old SD group compared to the 18-month-old HD group.
All tests were recorded with a webcam, and images were analysed with a computer program to score mouse performance in open field and water maze tests (ANY-maze Tracking System, Stöelting). Computer analysis was performed off-line. Behavioural tests were performed during the light cycle (08 h-12 h).
Water maze test
3-month-old (HD3M, n = 12; SD3M, n = 12), 6-month-old (HD6M, n = 12; SD6M, n = 12) and 18-month-old (HD18M, n = 13; SD18M, n = 7) mice were trained in a water maze adapted for mouse dimensions. The circular pool and platform were 94 and 14 cm in diameter, respectively, and the platform was 1 cm below the water surface. To hide the black platform, the pool was filled with black water (22 ± 2°C) coloured with a non-toxic dye. In each trial, the mice were allowed 60 s to find the hidden platform; trials were separated by intervals of 30 s. The task was considered complete when animals found and remained on the platform for 10 s. The first day of water maze training allowed the animals to adapt to the aquatic labyrinth. On the remaining four days, animals were tested in three trials once per day. We recorded the escape latency, distance travelled, average swimming speed and trajectories for each mouse. The visual cues outside the water maze were stable on all training days. The learning rate for the water maze was calculated by comparing results on the 1st and 5th test days. All groups were compared using one-way ANOVA, the Bonferroni a priori test or two-way ANOVA followed by Bonferroni post-tests, with intra- or between-group differences considered significant at a 95% confidence level (p < 0.05).
Perfusion and histology procedures
After the 5-day water maze behavioural tests, the mice were weighed and sacrificed with an overdose of ketamine (100 mg/kg) and xylazine (10 mg/kg) (Konig Laboratories). They were then perfused transcardially with heparinized saline for 10 min, followed by an aldehyde fixative (4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.2-7.4) for 30 min. All chemicals were purchased from Sigma (São Paulo, Brazil). After perfusion and craniotomy, the brains were removed and cut on a vibratome at a 70-μm thickness. One of each 5 sections was used for GFAP detection using free-floating immunohistochemistry. Free-floating sections were rinsed once in 0.1 M phosphate buffer, transferred to 0.2 M boric acid pH 9.0, heated to 65-70°C for 1 h and then washed 3 × 5 min in PBST (5%). The sections were incubated in a 1% hydrogen peroxide solution in methanol under constant and gentle shaking for 10 min and rinsed 2 × 2 min in 0.1 M PBS. The sections were then blocked with immunoglobulin for 1 h, following the instructions for the Mouse-on-Mouse Immunodetection kit (M.O.M. kit, Vector Laboratories, USA). Blocking was followed by washing (3 × 2 min) in PBS. Sections were incubated in a working solution of protein concentrate for 5 min, then incubated with monoclonal mouse anti-GFAP primary antibody (MAB360, Chemicon International, USA) diluted 1:800 in protein concentrate solution (M.O.M. kit) at 4°C for 3 days with continuous gentle agitation. Next, the sections were washed (3 × 2 min) in PBS and incubated for 20 h with biotinylated horse anti-mouse secondary antibody (M.O.M. kit), diluted 1:100 in PBS. After washing (3 × 2 min) in PBS, sections were transferred to an avidin-biotin-peroxidase complex solution (ABC, Vector Laboratories, USA, 1:200) for 1.5 h, washed (3 × 2 min) in 0.1 M PB, and processed using the glucose oxidase-DAB-nickel method and peroxidase histochemistry (Shu et al. 1988).
The reaction was interrupted when fine astrocytic branches were detected under the microscope. Sections were rinsed (4 × 5 min) in 0.1 M PBS, mounted on gelatinized slides, dehydrated in alcohol and xylene, and topped by a coverslip with Enthelan (Merck).
Photomicrograph documentation and processing
Digital photomicrographs were taken with a digital camera (Microfire, Optronics, CA, USA) coupled to a Nikon microscope (Optiphot-2, NY, USA). The digital photomicrographs were processed with Adobe Photoshop 7.0.1 C.S.2 software (San Jose, CA, USA) for scaling; adjustments to the levels of brightness and contrast were applied to the entire image. The selected micrographs display representative sections from each experimental group i.e. the astrocyte number in each region of interest was close to the mean value for that region.
Microscopy and optical fractionator
Details of the optical fractionator methodology and experimental parameters are described in online supplementary material (Additional file 5
: Table S5, Additional file 6
: Table S6, Additional file 7
: Table S7 and Additional file 8
: Table S8). In brief, we delineated the region and layers of CA1 at all levels in the histological sections, digitizing directly from sections using a low power 3.2× objective on a Optiphot-2 microscope (Nikon, Japan) equipped with a motorized stage (MAC200, Ludl Electronic Products, Hawthorne, NY, USA). This system was coupled to a computer running Stereoinvestigator software (MicroBrightField, Williston, VT, USA), which was used to store and analyse the x, y and z coordinates of digitized points. To detect and count unambiguously the objects of interest in the dissector probe, the low power objective was replaced by a 60× oil immersion Plan apochromatic objective (NIKON, NA 1.4) to count astrocytes.
All groups of animals were tested for statistical normality. Possible outliers identified based on standard deviations were eliminated from the data set. Parametric statistical analysis was used to assess the level of significance of the results of behavioural tests, one-way ANOVA, and the Bonferroni a priori test. The significance level for statistical differences was set at alpha < 0.05 (i.e. at a 95% confidence level). Results from the optical fractionator were analysed using two-way ANOVA. A two-tailed t-test was used to compare age-matched groups with different diets (SD vs. HD). Statistical analyses were performed using BioEstat® 5.0, Statistica for Windows® version 5.0 A and GraphPad Prism 5 for Windows®.