The results of the present study demonstrated that 24-hour-per-day masticatory deprivation had differential laminar effects on the number of astrocytes in CA1 and on water maze memory test performance in 3-, 6- and 18-month-old female albino Swiss mice. We suggest that masticatory deprivation can induce abnormal cognitive development and may enhance changes in aging astrocytes in CA1.
Hippocampal astrocytes, water maze performances and reduced mastication
In the present report, water maze tests revealed that after 6 or 18 months of masticatory deprivation, experimental SD mice could not learn and remember the position of a hidden platform; in contrast, control 6-month old HD mice could learn and remember the position of a hidden platform, whereas 18-month-old mice from both groups SD and HD were unable to reduce escape latency during the 5-day testing period. There were no significant differences in the swimming speeds of young (3-month-old), mature (6-month-old) and aged (18-month-old) mice in either the SD or the HD group, suggesting that spatial memory impairment, rather than motivation or physical performance, was a more likely explanation of the differences in the water maze performance. In parallel with memory dysfunction, there were significant changes in the laminar distribution of astrocytes in CA1. A continuous SD that started early in life (21st postnatal day) induced astrocytic hypoplasia at 3 months in all layers and in the pyramidal layer at 6 months as compared to age-matched HD mice. Compared to 18-month-old SD mice and 6-month-old HD mice, 18-month-old HD mice showed hypoplasia in the strata lacunosum moleculare. In addition, 18-month-old SD mice showed significantly more astrocytosis than 3-month-old and 6-month-old SD mice. Taken together, the results suggest that masticatory deprivation induced abnormal cognitive development and may have enhanced changes in aging astrocytes in CA1.
Although there were no learning and memory differences in the water maze test  after 3 months of masticatory deprivation, the laminar distribution of CA1 astrocytes started to change when the SD mice were 3 months old and became especially severe after six months or more of masticatory deprivation. Since spatial memory and learning were investigated previously only in adult (6-month-old) but not in aged (18-month-old) albino Swiss mice [29–31], this is the first report to assess the effects of long-term masticatory deprivation (using SD to model deprivation) on astrocytes by stereological analysis of CA1.
The accelerated cognitive decline observed in SD mice was not directly correlated with time-specific changes in astrocyte number in any CA1 layer. Thus, we investigated whether there were morphological changes (e.g. evident shrinkage of astrocyte arbors) that affected the astrocyte network. Long-term potentiation (LTP), considered the neurophysiological basis for learning and memory, is facilitated by brain-derived neurotrophic factor (BDNF) in both young and aged mice. However, in aged mice the effects of exogenous BDNF on LTP does not translate into improved spatial memory . Interestingly, BDNF expression in the hippocampus of C57Bl6 mice is reduced under SD as compared to HD at 3 and 6 months ; in parallel, there is reduced neurogenesis in SD mice of the same ages . We also observed a reduction in the number of astrocytes at 3 and 6 months in SD mice vs. HD mice. Notably, astrocytes synthesize and release BDNF, which is critical for experience-dependent synaptic plasticity in the mature brain , and there is a reduction of hippocampal BDNF levels in aged mice, with deleterious consequences for synaptic plasticity and spatial memory .
In the present report, there were significantly more astrocytes in the stratum oriens at 18 months in SD mice compared to HD mice. The combination of aging and masticatory deprivation affects the stratum oriens in an additive manner, reducing both dendritic spine density  and synaptic density  and significantly reducing cell size in SD and HD groups. These results suggest that astrocytic changes induced by aging and masticatory deprivation are greater in the strata lacunosum-moleculare and radiatum than changes in the strata pyramidale and oriens.
Many studies show that the CA1 is important in signalling pathways that are key for learning and memory (reviewed in ). Specifically, the lacunosum-moleculare layer is the target of the temporoammonic pathway, which originates in the entorhinal cortex and creates excitatory glutamatergic synapses with distal pyramidal dendrites in CA1 in mice  and rats [37, 38]. Temporoammonic synapses are associated with theta oscillations and late-phase LTP and long-term memory consolidation [39–41]. In addition, astrocyte calcium signals at Schaffer collateral to CA1 pyramidal cell synapses correlate with the number of activated synapses , demonstrating the direct participation of astrocytes in the hippocampal circuits that are involved in spatial memory. The presence of synaptic potentiation has been described previously to occur via astrocytic glutamate exocytosis at the entorhinal-to-dentate granular cells (the perforant pathway) [43, 44] and Schaffer collaterals . In this way, astrocytes play a critical role in memory functioning and LTP and contribute to synaptic tuning in the hippocampus .
Astrocytes in the hippocampus seem particularly sensitive to age-related changes [46–48] and can be impaired by structural/functional changes induced by masticatory imbalances (see, for example, [5–7]). Indeed, impairment of spatial memory is reported to occur in aged mice when the molar teeth are extracted or cut when the mice are young  as well as in adult rats that are fed a SD after the weaning period . In these studies, both neuronal density in the CA1 hippocampus  and synaptic formation in the hippocampus and the parietal cortex  were reported to decrease. In agreement with these studies, tooth loss and the resulting masticatory alterations leads to a reduction in the number of ChAT-positive neurons in the Broca diagonal band and medial septum, resulting in a decrease of acetylcholine levels in the hippocampus [49, 50].
Most of the earlier studies of the impact of reduced mastication on the hippocampus focused on neuronal changes. To our knowledge, there are only a few reports that employed GFAP-immunolabeling of the CA1 hippocampal field, and all such studies were restricted to aged SAMP8 mice. These studies revealed that the molarless condition enhances the age-dependent increase in the density and hypertrophy of astrocytes, and that these effects increase the longer the molarless condition persists [8, 15].
The notion that aging and masticatory deprivation may cause similar cellular and molecular alterations in the hippocampus is supported by other studies (reviewed in ). Indeed astrocyte hypertrophy has been reported previously in the CA1 hippocampal field [8, 15], suggesting that these cells may increase their production of pro-inflammatory cytokines in response to inflammation. Similar changes have been described in the senile hippocampus, where a more reactive astrocyte phenotype is expressed during aging, even in the absence of neurological disease, as part of an increased and maintained pro-inflammatory profile that may be associated with cognitive dysfunction (reviewed in ).
It is unclear why the number of astrocytes in the stratum lacunosum-moleculare was reduced in 18-month-old HD mice compared to 6-month-old HD mice. However, since BDNF regulates gliogenesis, it could be that the reduced levels of BDNF induced by aging contributed to the decrease in the number of neurotrophic astrocytes . In addition, we suggest that the astrocytosis observed in the stratum oriens of 18-month-old SD mice is not dependent on BDNF but, rather, is part of the inflammatory aging process . Although the present investigation is related to chronic stress (since there was a long period of masticatory deprivation), previous results regarding the acute stress response during immobilization revealed that biting can restore the BDNF mRNA levels that are reduced by stress . Taken together, the results suggest that a decrease in masticatory activity along with aging may promote the observed differential laminar effects on CA1 astrocytic plasticity.
Astrocytic changes, corticosteroids and masticatory stress
Active mastication can rescue the stress-attenuated hippocampal memory process in animals by suppressing endocrinological and autonomic stress responses . Indeed, previous studies of molarless mice showed plasma corticosterone levels that were significantly greater than those in molar-intact control mice. In addition, pretreatment with suppressors of stress-induced increases in plasma corticosterone levels prevented the molarless-induced increase in plasma corticosterone levels . In addition, elevated corticosterone levels suppress synaptic plasticity in the hippocampus , and mastication suppresses the stress-activated expression of corticotropin-releasing factor (CRF). Astrocytic plasticity in the hippocampus may also be affected by corticosteroids due to changes in the number of astrocytes [64, 65] or to changes in astrocyte morphology and function . In particular, corticosterone increases the number of astrocytes in CA1 in a dose-dependent fashion .
In this study, masticatory deprivation-induced stress may have altered the plasma glucocorticoid levels, and this may in turn have affected astrocytic plasticity. We did not measure plasma corticosteroid levels; therefore, we cannot exclude the possibility that altered levels of corticosteroids might explain the observed effects of masticatory stress. However, there is currently no information about changes in glucocorticoid levels that might be induced by a SD, so it is difficult to evaluate this possibility.