Long-term D -allose administration ameliorates age-related cognitive impairment and loss of bone strength in male mice

.


Introduction
The world's elderly population is growing rapidly (United Nations, 2020).Cognitive decline is common with advancing age, as is physical dysfunction like muscle weakness and decreased bone strength (Amarya et al., 2015).The age-related cognitive decline, especially, memory decline is associated with early phase of dementia.However, it is difficult to distinguish between pathological and non-pathological cognitive decline in the elderly (Deary et al., 2009).Dementia is predicted to become an increasingly important public health issue because of the continued explosion of population aging (World Health Organization, 2021b).The number of people with dementia now exceeds 55 million globally (World Health Organization, 2021b), and in 2019, it was the seventh leading cause of death (World Health Organization, 2021a).Besides cognitive decline, muscle and bone aging have been reported to result in significant morbidity and mortality in older people, all of which leads to a gradual erosion of quality of life (Dennison et al., 2010;Rizzoli et al., 2013).Therefore, it is imperative to find a way to prevent, or at least slow down, these age-related mental and functional declines.
Control of diet and food consumption throughout one's life is important to promote healthier aging (Wickramasinghe et al., 2020).Some reports indicate that improving healthy eating behaviors in midto later-life is likely to lead to a healthier aging phenotype (Kiefte-De Jong et al., 2014;Wickramasinghe et al., 2020).Dietary components such as vitamins and Ω-3 fatty acids have been reported to attenuate age-related cognitive decline through anti-inflammatory, antioxidant, and neuroprotective effects (Deary et al., 2009).Rare sugars are monosaccharides that are relatively rare in common foods.Certain rare sugars have been shown to have beneficial effects on lifestyle-related diseases such as diet-induced diabetes and obesity (Smith et al., 2022).For example, D-allulose, a well-studied rare sugar, is reported to suppress the postprandial blood glucose elevation (Hayashi et al., 2010;Iwasaki et al., 2018), and alter lipid metabolism such as inhibiting fatty acid synthesis and promoting fatty acid oxidation (Nagata et al., 2015;Ochiai et al., 2014).Another rare sugar, D-tagatose, also has several beneficial effects such as reducing hemoglobin A1c levels (Ensor et al., 2016), reducing body weight, and increasing highdensity lipoprotein cholesterol (Donner et al., 2010).D-Allose, a C-3 epimer of D-glucose, is also one of rare sugars and being studied as a novel functional food and medicine.Approximately 70 % of ingested D-allose is absorbed in the small intestine through glucose transporters and excreted in urine, while the remaining 30 % is fermented slightly in the intestine and then excreted in feces (Kitagawa et al., 2018).In the safety evaluation study in rats, the LD 50 of D-allose is 20.5 g/kg body weight.It has been observed that D-allose is safe under condition of continuous intake of 0.3-3 % D-allose containing diet for 6 months (Iga et al., 2010).D-Allose has previously been shown to modulate several diseases or tissue pathologies such as tumors (Sui et al., 2005), hypertension (Kimura et al., 2005), ischemic injury (Huang et al., 2016), and osteoclast differentiation (Yamada et al., 2012).These effects may be modulated in part by suppressing reactive oxygen species (ROS) generation (Ishihara et al., 2011), protecting blood-brain barrier (Huang et al., 2016), and favorably altering of intestinal bacterial species and microflora in aged mice (Shintani et al., 2022).However, the effects of D-allose on age-related phenotypes are not known.
In the present study, we tested whether long-term oral D-allose administration in mice during middle age to the early stage of aging slows negative, age-related changes in phenotypes, especially cognitive impairment.We used a battery of standard behavioral tests on 18month-old mice after they were fed 3 % D-allose-containing diet for 6 months.In the behavioral test battery, cognitive function was assessed using the Morris water maze task and the fear-conditioning task.To characterize the behavioral phenotype that might influence cognitive functions, the open field test and analgesia tests were used to assess locomotor activity/anxiety-like behavior and pain sensitivity, respectively.Knowledge on multi-domain behavioral phenotype offers a better way to understand disease mechanisms and aging and the influence by D-allose administration.DNA microarray analysis of hippocampal tissue was also performed to determine changes in gene expression.Performance on the wire-hanging test and bone strength were assessed and measured to identify how D-allose influences the aging phenotypes of muscle and bone, respectively.

Ethics statement
All experiments were approved by the Animal Experiment Committee of the Tokyo Metropolitan Institute for Geriatrics and Gerontology (Animal Protocol Approval nos. 17012, 20018 and 23001).They were carried out according to its guidelines and in accordance with the Guide for the Care and Use of Laboratory Animals.In this paper, we describe and report the details of the animal experiments according to the Animals in Research: Reporting In Vivo Experiments (ARRIVE) guidelines 2.0 (Kilkenny et al., 2010;Percie Du Sert et al., 2020).

Animals and experimental design
Experimentally naive C57BL/6J mice (male, 1 month old) were obtained from CLEA Japan Inc. (Tokyo, Japan).Mice were housed (5 per cage) in individual ventilated cages (Tecniplast S.p.A., Buguggiate (VA), Italy) containing paper chip bedding (Japan SLC, Inc., Shizuoka, Japan).The specific pathogen free (SPF) vivarium was maintained at 22 ± 1 • C and 55 ± 5 % humidity under a 12-h light-dark cycle (light on at 7:00 A. M.).Periodic monitoring of microbial state guaranteed that the mice were maintained under SPF conditions throughout the entire experimental period.
Mice had free access to commercial chow (CRF-1, Oriental Yeast Ltd., Tokyo, Japan) and water.A schematic diagram of the experimental design is shown in Fig. 1.The number of mice used in this study is presented in Table S1.To mitigate the possible influence of experimenter bias, prejudice, and unintentionally presented physical cues, all experiments were conducted in a blinded manner; the experimenter did not know which mice were administered D-allose.Japan), as previously reported (Bhuiyan et al., 1998).For dietary administration, D-allose (3 % [w/w]) was mixed into the chow (CRF-1).Given the established safety in rats (Iga et al., 2010), we calculated the dose of D-allose (3 %) using a dose conversion formula based on body surface area (Reagan-Shaw et al., 2008).At 12 months of age, mice were assigned to either the control (15 mice) or D-allose group (15 mice) in such a way that the mean body weights in the two groups were similar (Fig. 2).Mice assigned to the D-allose group were fed the D-allose-chow mixture ad libitum until the end of the experiments.The control group was fed the same chow, but without D-allose.

Behavioral experiments
At 18 months of age (i.e., had received 6 months of daily D-allose treatment or control diet), mice were handled for 3 consecutive days, and then they underwent a battery of behavioral tests (below).Mice were tested sequentially on the wire-hanging test, open-field test, Morris water maze task, and Pavlovian fear-conditioning task.After completion of the Pavlovian fear-conditioning task, analgesia tests (electrical foot shock sensitivity test and hotplate test) were performed.Mice were transferred to each experiment room and habituated for 15 min before the beginning of the test.Procedures for the behavioral experiments were described in detail elsewhere (Yanai and Endo, 2021).
All behavioral apparatuses were obtained from O'Hara & Co., Ltd.(Tokyo, Japan), unless specified otherwise.Experiments were conducted between 9 a.m. and 5 p.m.An automated software program called Time® (open field, Morris water maze, and fear conditioning; O'Hara & Co., Ltd.) was used to control the apparatus and collect and analyze the data.

Wire hanging test
To assess grip strength in the mice, we used the wire-hanging test, a well-characterized test for mice (Kojima et al., 2008).Briefly, mice were forced to hang onto a suspended stainless-steel wire grid (2 mm in diameter, wires spaced 1 cm apart) located 30 cm above the floor over soft paper chip bedding.The mouse was placed in the center of the top lid lined with a wire grid, and then the lid was immediately closed and flipped upside down so that the mouse was inverted inside of the apparatus.The latency from start of the inversion to drop off the wire grid was recorded; a cutoff time of 300 s was used.Each mouse was tested twice with a 30-min intertrial interval.The longest latency of the two trials was used for analysis.

Open-field test
To assess general locomotor activity and anxiety-like behavior in a novel environment (Walsh and Cummins, 1976), the open-field test was used as previously described (Kojima et al., 2008;Yanai et al., 2014).To begin, each mouse was placed onto the floor of a square arena (50 cm × 50 cm) having transparent walls (height: 50 cm).It was allowed to explore for 15 min.Mouse behaviors in the arena were assessed in relatively dark ambient illumination (10 lx) on the first day of testing and under bright illumination (300 lx) on the second day.Throughout the open-field test, the distance traveled and number of rearing were measured as indicators of locomotor activity.Immobile time and time spent at the perimeter of the arena were taken as proxy measures of anxiety (Walsh and Cummins, 1976).

Morris water maze task
To assess spatial memory, the Morris water maze task (Morris, 1981;Morris et al., 1982) was used as previously described (Yanai and Endo, 2016).Briefly, a circular pool (100 cm in diameter) surrounded by distinctive distal visual cues was filled with tap water (22 ± 1 • C) and was rendered opaque by adding titanium dioxide.Thus, anything below the water surface was not visible.
For each mouse, an escape platform, submerged 1 cm below the water level, was placed in the center of one pool quadrant and remained in that quadrant throughout task acquisition.The mice underwent multiple training trials for acquiring the spatial location of the escape platform.During this training, mice were allowed to swim for a maximum of 60 s to find the platform.Four acquisition training trials were carried out each day for 10 days.One day after completion of acquisition training, a 30-s spatial-memory probe test was conducted in which the platform was removed from the pool.During the probe test, the number of times the mouse crossed the exact location where the platform was located during training and the time the mouse searched within the training pool quadrant were tallied.These measures assessed the mouse's spatial memory for the platform location obtained during acquisition training.
On the next day of the probe test, mice were subjected to cued training in the pool for 4 consecutive days.In this phase of the task, a prominent black flag was attached to the platform, making it visible.

Pavlovian fear-conditioning memory task
To evaluate conditioned fear memory to both tone and context (LeDoux, 1997), a fear-conditioning task was used as previously described (Takahashi et al., 2019;Yanai et al., 2018;Yanai and Endo, 2021).Briefly, on the first day, mice were placed individually in the conditioning chamber.After an exploratory period of 60 s, a tone stimulus (10 kHz, 70 dB) was presented for 3 s.The tone stimulus coterminated with a scrambled electrical foot shock (0.12 mA for 0.5 s).During the cue-dependent fear memory test (1 h and 24 h after conditioning), the mouse was placed in a novel chamber, and the tone stimulation was presented for 60 s.The chamber size was the same as the conditioning chamber, but the brightness, smell, and texture differed from those of the conditioning chamber.During the context-dependent fear memory test (48 h after conditioning), the mouse was placed in the original conditioning chamber.No electrical foot shock was delivered during this test.Throughout the experiment, freezing behavior, which was defined as complete immobility (except for movement required for heartbeat and respiration), was taken as an index of conditioned fear memory (Fanselow, 1990;Phillips and LeDoux, 1992).Total freezing time was scored as a percentage of freezing within a specified period (60 s).

Electrical footshock sensitivity test
The sensitivity to electrical footshock was examined as previously described (Yanai et al., 2018;Yanai and Endo, 2021).Mice were placed individually in the conditioning chamber, and electrical footshock was delivered at a starting intensity of 0.01 mA; the intensity was increased by increments of 0.01 mA.During this test, the shock intensity that evoked a paw flick and vocalization was recorded.

Hotplate test
After the completion of the electrical footshock sensitivity test, a hotplate test was performed as previously described (Yanai et al., 2018;Yanai and Endo, 2021).Pain sensitivity to heat was assessed by measuring the latency to lick a forepaw after the mouse was placed on a 55 • C hotplate (Muromachi Kikai Co., Ltd., Tokyo, Japan).

Preparation of bone and hippocampus tissue
After all the behavioral assessments were done and based on their performance on the Morris water maze, eight mice were selected from each group.The mice were prepared for bone and hippocampus extraction.Mice were deeply anesthetized with isoflurane and transcardially perfused with 0.01 M phosphate buffer (pH 7.4) containing 0.15 M NaCl.The hippocampus was bilaterally dissected from the brain.Femurs were also dissected at this time from the surrounding tissue collected.All samples were stored at − 80 • C (hippocampal tissue) or 4 • C (femurs) until analysis.

Analysis of bone strength
Bone strength measurements were made in postmortem femurs.To evaluate bone strength, the maximum load (max load), displacement, and stiffness were measured by the three-point bending test (Hellings et al., 2019).Orientation of the femurs in the measuring devices is described below.Max load is the maximum load that is applied until bone fracture; stronger femurs can withstand greater applied forces.Displacement is the distance the bone is displaced under loading up to the point the femur fractures.Displacement is a measure of bone rigidity; smaller values indicate that the bone is brittle.Stiffness is calculated as the slope of the load versus displacement curve and is measured as a ratio of bone resistance to displacement under loading.The degree of stiffness reveals, at least in part, certain properties of the liquid and fibrous tissue in the bone (Prodinger et al., 2018).Increased stiffness reflects a bone's decreased ability to deform when stressed or its ductility.
These tests were performed using a bone strength measuring instruments (MZ-500D; Malto Instrument Co., Ltd., Tokyo, Japan) connected to a strength-loading device (MZ-500S; Malto Instrument Co., Ltd.).Femurs were individually placed with the anterior surface down on two supports equidistant at a fulcrum distance of 6 mm.They were centrally loaded at a constant rate; the loading speed was 2 mm/min until fracture.Load displacement curves were used to calculate max load (N), displacement (mm), and stiffness (N/mm) (Wergedal et al., 2005).
Bone density, one of the most important determinants of bone strength (Weinstein, 2000), was measured by the dual-energy, X-ray absorptiometry (DXA) method using a bone densitometer (DCS-600EX-IIIR; Aloka Co. Ltd., Tokyo, Japan).The long axis of the femur was scanned at 1 mm-thick slices and at a rate of 12.5 mm/s.Bone mineral content, bone area, and bone density were measured using DXA.Bone density was calculated by dividing bone mineral content by bone area.This calculated value was then expressed as bone mineral mass per unit area (g × cm 2 ) (Saraiva et al., 2021).

DNA microarray
Total RNA was extracted postmortem from the hippocampus by using TRIzol reagent (ThermoFisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's instructions, and purified with an RNeasy Mini Kit (Qiagen GmbH, Hilden, Germany).The quality of the isolated RNA was assessed with an Agilent TapeStation 4200 RNA ScreenTape Assay Kit (5067-5576; Agilent, Santa Clara, CA, USA), which determines the sample's RNA integrity number (RIN) and RNA concentration.Next, cRNA was fragmented, and the RNA ScreenTape assay of the biotin-labeled targets was performed prior to sample hybridizations.The biotin-labeled targets were then hybridized to an Affymetrix Clariom™ S Array (mouse; Affymetrix; ThermoFisher Scientific, Inc.).All microarrays were scanned using the Affymetrix Gene Chip Command Console, which was installed in the Gene Chip Scanner 3000 7G.All microarray datasets were analyzed using Affymetrix default analysis settings and global scaling as a normalization method; datasets were uploaded onto the microarray data management system (MDMS) for storage.
Microarray datasets were expressed as fold-change values, and differentially expressed genes were identified using Affymetrix's Transcriptome Analysis Console software applying the Clariom™ S mouse library for group comparisons.Gene values were considered differentially organized between the control and D-allose-treated groups when there were larger than 1.5-fold differences in absolute expression.The expression of genes in the positive or negative direction was considered as up-or downregulated, respectively.The functional pathway analysis by WikiPathways was conducted using TAC software.

Statistical analysis
All data were expressed as means ± SEM.Statistical analysis was performed on the results obtained from 13 mice in the control group and 15 mice in the D-allose group; two mice in the control group died during the study period.Statistical differences between groups were determined by unpaired Student's t-test or mixed design two-way analysis of variance (ANOVA) using IBM SPSS statistics software version 27 (IBM Corp., Armonk, NY, USA).Statistical significance was set at p < 0.05.When the interaction was statistically significant, a test for the simple main effect was done.Details of the statistical analysis of all experiments are provided in the supplementary materials (Table S2).

Body weight
Both groups showed a gradual increase in body weight, with D-allose- administered mice slightly lighter than control mice (Fig. 2).However, there were no statistical difference between the body weight of both groups (F (1, 26) = 1.53, n.s.).

Wire-hanging test
We tested both groups of mice on the wire-hanging test, a measure of grip strength.Both the experimental and control groups of mice performed similarly (latency to drop: 44.2 ± 13.4 s control; 40.2 ± 4.0 s D- allose, t (26) = 0.31, n.s.).The results suggest that long-term dietary consumption of D-allose has no detectable effect on grip strength in mice.

Open-field test
In the open-field test, distance traveled and number of rearings are indices of general locomotor activity, whereas immobility time and time spent away from the center are indices of anxiety level (Walsh and Cummins, 1976).These behaviors were assessed under relatively dark conditions on the first day and under bright-light conditions on the second day of testing.Both the experimental and control groups of mice performed similarly in the open-field test (Fig. 3).Mean distance traveled (Fig. 3A) and mean number of rearings (Fig. 3B) in the bright-light condition decreased significantly compared to the dark condition on day 1 (distance traveled: F (1, 26) = 15.16,p < 0.001; number of rearings: F (1, 26) = 4.70, p < 0.05).Because two groups performed comparably in these indices, long-term administration of D-allose had no apparent effect on locomotor activity.T. Shintani et al. Experimental Gerontology 196 (2024) 112555 A similar pattern of results was found for the measures of anxiety level.Both groups of mice behaved similarly; they both showed significantly longer mean immobility time in the bright-light condition compared to the dark condition (Fig. 3C; F (1, 26) = 29.85,p < 0.001).We found no significant group difference in the time spent in the center of the field (Fig. 3D; F (1, 26) = 0.20, n.s.).However, we noticed that sometimes older mice seem to be unaffected by the light-on condition.Our previous studies reported that the time spent in the center is not replicable when using aged mice (Yanai et al., 2018(Yanai et al., , 2022;;Yanai and Endo, 2021).Since mice tend to be anxious in well-lit environments, the behavior we observed is not consistent with this tendency.This might be related to their more advanced age and/or intervention used; further examination is required to validate time spent in the center in our protocol.Taken together, these results suggest that D-allose, at least 6 months of consumption, had no detectable influence on locomotor activity or anxiety-like behaviors in a novel environment.

Morris water maze task
The Morris water maze task assesses hippocampus-dependent spatial memory (Morris et al., 1982).Performance in memory tasks in animals partially depends on physical factors like muscle strength and endurance that typically decline with aging (Doherty, 2003;Yanai and Endo, 2021).Regardless, swimming speed in the task appeared to be unaffected by D-allose administration, as the average swimming speed was statistically indistinguishable during all stages of the Morris water maze task (Table 1).Therefore, it is reasonable to exclude possible loss of muscle strength and endurance as confounding factors when interpreting results of the memory tasks below.
Trends in the results differed for Day 1-5 training period compared to the Day 6-10 training period.Thus, we analyzed the results separately for each period.Visual inspection of the graphical results (Fig. 4A, B) appeared to show that both groups acquired the task similarly in the first half of acquisition training (latency to find platform: F (1, 26) = 0.05, n. s.; Distance traveled: F (1, 26) = 0.41, n.s.).However, over Days 6-10, the mean escape latency for the D-allose group continued to decline.By contrast, performance of the control group became asymptotic.This result suggests that D-allose may improve spatial memory, as escape latency in the latter half of training was significantly less (F (1, 26) = 7.37, p < 0.05).Mean swim distance in the Day 6-10 period was also shorter for the D-allose group but fell short of statistical significance (F (1, 26) = 3.32, p = 0.080).These results suggest that oral D-allose administration affects learning in the latter phase of spatial learning, rather than initial acquisition.
In the probe test, performance of the D-allose group was significantly better than that of the control group.Compared to the control group, the D-allose group had significantly more crossings over the specific location where the platform was during training (Fig. 4C; t (25) = 2.08, p < 0.05).However, the two groups spent a similar amount of time in the quadrant where the platform was located during training (Fig. 4D; t (25) = 0.27, n. s.).
In the probe test of the Morris water maze task, learning of and memory for precise location and rough location are often distinguished (Yanai et al., 2022;Yanai and Endo, 2021).This difference is sometimes described as a rough spatial representation versus a fine-grained representation about the locations of objects (Evensmoen et al., 2013).The number of platform crossings during a probe trial in the Morris water maze, then, reflects the memory of this more precise fine-grained spatial memory (Yanai and Endo, 2021).According to this interpretation, the present study suggest that D-allose helped aged mice better encoding on the exact platform location during the training, which was manifested by more precise traverses over the training location of the platform location during the probe test (Fig. 4C).Whereas precise spatial memory normally declines with advancing age in human (Nilakantan et al., 2018;Segen et al., 2021), D-allose may be involved in refining and strengthening precise spatial memory, or alternatively, it may be involved in suppressing the decline in spatial memory for a precise location that occurs with aging.
For cued training in the Morris water maze task, D-allose had no detectable effect on escape latency (Fig. 4E; F (1, 26) = 2.15, n.s.) or swim distance (Fig. 4F; F (1, 26) = 0.22, n.s.).Cued training assesses whether performance factors unrelated to spatial memory are present (Vorhees and Williams, 2006), like visual acuity.Since performance of the D-allose group on cued training was indistinguishable from that of the control group, it suggests that the D-allose-related improvements in spatial learning and memory described above is not confounded by an apparent improvement in visual ability.

Pavlovian fear-conditioning task
After conditioning with tone and footshock, mice were sequentially tested on short-term (1 h) and long-term (24 h) cue-dependent fear memory tests.These were followed by assessment of long-term contextual fear memory (48 h) (Fig. 5).The control and D-allose- treated mice exhibited a similar level of freezing in the cue-dependent fear memory test conducted at 1 h (t (26) = 0.88, n.s.) and 24 h (t (26) = 0.69, n.s.) after conditioning.The two groups of mice showed comparative level of freezing in the context-dependent fear memory test 48 h after conditioning (t (26) = 0.73, n.s.).Cue-dependent fear memory and context-dependent fear memory depend on the amygdala and hippocampus, respectively (Phillips and LeDoux, 1992).These results suggest that D-allose has no effects on conditioned-fear memory.Data are means ± SEM.

Analgesia tests
In the analgesia tests, the D-allose and control groups performed similarly (Table 2; hotplate test: t (26) = 1.77, p = 0.088 n.s.; paw flick latency in the electrical foot shock sensitivity test: t (26) = 0.58, p = n.s.; vocalization in the electrical foot shock sensitivity test: t (26) = 1.29, n. s.).These results from two kinds of analgesia tests show that long-term dietary D-allose consumption has little or no effect on pain sensitivity in mice.

DNA microarray
Bioinformatic analysis using DNA microarray technology detected a total number of 22,206 genes.D-Allose altered the expression of several genes, however, the functional pathway analysis revealed no significant crosstalk nor correlation among the genes.Fig. 7 shows 15 genes that   Values are means ± SEM. had a >1.5-fold change in expression.These 15 genes included functionally disparate and unannotated genes.Among the 11 upregulated genes, predicted gene 11096 (Gm11096) showed the highest fold change (fold change: 5.69, p < 0.05); its function has yet to be identified.The remaining 10 upregulated genes had similar fold changes.Of the 4 downregulated genes, 2 were very large inducible GTPase 1 pseudogenes (Gm8979, Gm8989) (fold change: − 1.52, p < 0.001, respectively) and 1 was regucalcin (Rgn) (fold change: − 1.52, p < 0.05); these three genes had the same fold changes.We register the data in Gene Expression Omnibus and the number is GSE270340.

Discussion
Functional foods like D-allose might mitigate age-related cognitive and physical decline often seen with aging (Vlachos and Scarmeas, 2019).However, the effects of dietary D-allose for this purpose have been understudied.In the present study, we demonstrated for the first time some limited beneficial effects of long-term dietary consumption of D-allose in aging male mice.As measured in the number of escape platform crossings in the Morris water maze (Fig. 4C), D-allose improved spatial memory.This aspect of memory typically declines in aging mice (Yanai and Endo, 2021).Also, the decline of bone strength typically observed in aging mice (Amarya et al., 2015) was lessened (or reduced) by D-allose consumption (Fig. 6A).Age-related loss of grip strength and general locomotor activity were not improved by long-term D-allose consumption (Fig. 3A, B).The present results indicate that D-allose may help prevent, or slow-down the loss, of some aspects of age-related cognitive function and bone strength.
Grip strength is reported to well-predict future age-related declines in physical function in humans (Rantanen et al., 1999).In this context, we tested the D-allose and control groups on the wire-hanging test, a well-characterized test of grip strength for mice (Kojima et al., 2008).There was no detectable effect of dietary D-allose consumption on grip strength in aging mice, as evaluated by the wire-hanging test.A previous study showed that male C57BL/6J mice exhibit a gradual decline in grip strength starting at 6 months of age, with a significant decline emerging at 12 months of age (Yanai and Endo, 2021).Thus, in the present study, grip strength in our mice likely had already decreased significantly when D-allose administration was started (i.e., at 12 months of age).This suggests that D-allose may fail to improve grip strength in mice that already have a large declined in strength.Alternatively, the effects of dietary D-allose in aging may be smaller than anticipated.Intervention with D-allose in younger mice may be necessary to assess any agemoderating effect of dietary D-allose on the decline of grip strength.
Decreased locomotor activity (Tveter et al., 2014) and increased anxiety (Tomitaka et al., 2015) are often present in older people.Low activity level is a diagnostic criterion for human frailty (Fried et al., 2001).This parameter is reported to correspond to spontaneous activity in mice, which can be measured by distance traveled in an open-field test and number of rearings (Parks et al., 2012;Yanai and Endo, 2021).Locomotor activity and anxiety states gradually decline with age in mice, beginning around 12 months of age (Yanai and Endo, 2021).Our results appear to show that D-allose administration does not affect locomotor activity (Fig. 3A, B) and anxiety states (Fig. 3C, D).These results provide experimental evidence that D-allose has little or no effect on age-dependent phenotypic/emotional changes.This is considered to be a major advantage of functional foods, which are supposed to be administered (consumed) over a long period of time in human.
Aging primarily degrades hippocampal-dependent memory (Burke and Barnes, 2006;Rosenzweig and Barnes, 2003;Squire, 2004;Squire and Zola, 1996), including memory for names and shapes of objects and places.Age-relate decline of memory can also negatively affect the quality of life of older adults (Li et al., 2022).Once dementia develops, it is very difficult to completely reverse or temper, and treatment currently consists primarily of reducing symptoms (Dey et al., 2017).Therefore, it is very important to address the emergence of cognitive decline, especially memory problems, early on before it fully develops.Thus, in the present study, we administered D-allose to 12-month-old mice, an age when cognitive decline begins in aging (Yanai and Endo, 2021).We used the Morris water maze task to evaluate hippocampus-dependent spatial memory.
In general, spatial learning in the water maze becomes gentle in the  second half of training, however dietary D-allose maintained the effect in the latter half of training.In classical theory of information processing, memory is considered to consist of three processes: acquisition, consolidation, and retrieval (Abel and Lattal, 2001).Recently, a phenomenon called reconsolidation has been advocated as an alternative mechanism in memory formation (Rossato et al., 2006).Reconsolidation is defined as stabilization of memory through a further round of protein synthesis, which makes the memory endure after retrieval.It has been suggested that performance improvements in the late phase of training in the Morris water maze may reflect memory reconsolidation (Kimura et al., 2008).Accordingly, D-allose may play a supportive role in reconsolidation for spatial memory and learning, enhancing memory stabilization in late stages, but playing little or no role in initial stages of information processing underlying early memory acquisition and subsequent consolidation.Thus, these results suggest that the intriguing possibility that oral D-allose administration suppresses the normal agerelated decline in memory reconsolidation but may have little effect on early stages of memory acquisition.Improved memory reconsolidation can be expected to improve older people's ability to learn new things and new objects in actual social life, thereby favorably altering a quality of life that requires learning (Li et al., 2022).The result on the probe test (Fig. 4C) also supports the idea that D-allose administration suppress age-related decline of spatial memory having high attention demands.Taking into account that Flurkey et al. (2007) consider mice aged 18 to 24 months as "aged mice", 18-month-old mice used in this study can be considered as in the early stage of aging.Taken together, then, D-allose consumed during middle-age to early stage of aging could be expected to help maintain cognitive functions necessary for daily life in older people.
To explore the mechanism of learning and memory enhancement after long-term D-allose consumption, we evaluated whether any changes in gene expression may contribute to D-allose-related cognitive and physical improvements.Our DNA analysis revealed that D-allose increased the expression of Slc17a6, a gene that highly expressed in neurons (Aihara et al., 2001).Intriguingly, the protein of Slc17a6 is reported to function as a vesicular glutamate transporter, to control synaptic phosphate levels (Cheret et al., 2021), and to be associated with improving spatial learning and memory (He et al., 2012).This result suggests that D-allose administration may modulate glutamate levels in synaptic vesicles and excitatory transmission at synapses in the hippocampus.Such modulation may contribute to staving off cognitive decline in aging.D-Allose downregulated two very large inducible GTPase 1 pseudogenes (Gm8979, Gm8989; Fig. 7).The expression of these two genes is upregulated by interferon gamma, an indicator of inflammation, and by microglial activation in a mouse model of Alzheimer's disease (Landel et al., 2014).Microglial activation is a prominent characteristic of neuroinflammation (Akiyama et al., 2000).Thus, D-allose may prevent cognitive decline by suppressing neuroinflammation through the downregulation of GTPase 1 pseudogene.Regucalcin, also known as senescence marker protein-30 (SMP30) (Arun et al., 2011), was decreased by D-allose administration.Regucalcin inhibits the activity of calcium/calmodulin-dependent protein kinase by binding to calmodulin (Yamaguchi, 2012).Calcium/calmodulindependent protein kinase II (CaMKII), which is highly abundant in the brain, also enhances the efficacy of synaptic transmission; mutation of this enzyme blocks long-term potentiation, experience-dependent plasticity, and memory (Lisman et al., 2002).These reports support the hypothesis that the decrease in regucalcin observed in our study may partly contribute to maintenance of cognitive function during aging by increasing CaMKII activity.
Previous studies showed that D-allose suppressed the generation of ROS, which cause oxidative stress (Ishihara et al., 2011), and increases Sirtuin gene expression in Caenorhabditis elegans (Shintani et al., 2019) and in mice fed 5 % D-allose (as we observed in our unpublished experiments).Oxidative stress (Kandlur et al., 2020) and changes in Sirtuin genes (Fagerli et al., 2022) are known to affect cognitive function.
We hypothesized that D-allose may maintain cognitive function at a higher level in aging mice through these factors.However, our microarray analysis showed that our protocol of D-allose administration did not change the expression of genes related to oxidative stress and sirtuin (Fig. 7).Brain-derived neurotrophic factor (BDNF) is known to regulate cognition and other complex behaviors through interactions between neuronal activity and synaptic plasticity (Lu et al., 2015).In our preliminary experiments, however, D-allose did not affect BDNF gene expression in aged-mice.Taken together, oral intake of D-allose might have a beneficial effect on cognitive functions via several pathways associated with the above-mentioned gene changes.However, it is necessary to explore the contribution of other factors in the future.
In this study, we also observed that D-allose might have a favorable influence on max load, a parameter of bone strength, without significantly affecting bone density.Bone strength is determined not only by bone density but also by bone quality, which includes several factors such as microarchitecture, mineralization, ability to repair damage, collagen structure, crystal size, and marrow composition (Ott, 2016).The present study did not measure bone quality, therefore, further investigation is required to reveal the roles of D-allose on bone health improvement and its underlying mechanisms.Taken together with findings that the maximum load of femurs is reported to peak at 13 months of age and diminish thereafter (Ferguson et al., 2003), and that D-allose prevents osteoporosis by suppressing osteoclast differentiation in vitro (Yamada et al., 2012), our findings may suggest that D-allose mitigate the decline in bone strength with aging.
In the present study, male mice were used in accordance with a tradition in scientific research (Beery and Zucker, 2011), however, such a male bias has to be changed.As evident in human studies, prevalence of dementia is higher in females than in males (Zhu et al., 2021), and therapeutic efficacy of some antidementia drug is stronger and more beneficial for males (Davis and Barrett, 2009).In animal model of aging, recent studies reveal the substantial sex-based differences in behavioral phenotype (Matsuda et al., 2015;Tucker et al., 2016) and therapeutic efficacy of antidementia drug (Alves-Amaral et al., 2010).As balanced population of male and female mice are advocated (Shansky, 2019;Wald and Wu, 2010), elucidating sex differences in behavioral phenotype and in therapeutic efficacy will be useful for the development of sex-sensitive strategies in the context of functional foods development.Additionally, considering that aging is thought of as occurring in an asynchronous and non-linear fashion in which several physical functions decline at the various rate (López-Otín et al., 2013;Rando and Wyss-Coray, 2021), further experiments are required to fully understand the effects and mechanisms of D-allose on age-related physical functions.

Conclusions
We demonstrated for the first time that long-term D-allose administration staves off age-related memory decline, to a degree, in 18-monthold male mice.We also found that D-allose administration reduces the loss of bone strength that is common in aging.These results suggested that D-allose may have the potential to be a multi-dimensional functional food.Accumulation of more experimental evidence is necessary to bolster the idea that functional foods may represent a relatively simple way for older people to maintain their independent lifestyle longer.
Committee of the Tokyo Metropolitan Institute for Geriatrics and Gerontology (Animal Protocol Approval nos. 17012, 20018 and 23001).

Declaration of competing interest
The authors declare the following financial interests/personal relationships that may be considered as potential competing interests.Tomoya Shintani, Akane Kanasaki, and Tetsuo Iida are employees of Matsutani Chemical Industry Co., Ltd. that provided D-allose for this research.These authors from Matsutani Chemical Industry carried out behavioral/bone strength experiments in a blinded manner and positively involved in preparing manuscript, however, they were not involved in the interpretation of data.

D
Fig. 1.Schematic diagram of the experimental design.C57BL/6J mice (12 months old) were fed a control (no D-allose) diet or a 3 % D-allose-containing diet for 6 months.Six months later, all of the mice underwent behavioral testings, which included the wire-hanging test, open-field test, Morris water maze task, Pavlovian fear-conditioning memory task, electrical footshock sensitivity test, and hotplate test.After behavioral testing, the mice were euthanized and femur and hippocampus samples were collected for bone analysis and DNA microarray analysis, respectively.

Fig. 2 .
Fig. 2. Temporal changes in body weight after mixed-in-feed administration of D-allose.D-Allose administered mice were constantly lighter than control mice, however, these differences were not statistically significant.Error bars indicate ± SEM.Significant differences were determined by two-way ANOVA.

Fig. 3 .
Fig. 3. Locomotor activity and anxiety-like behavior in the open-field test.(A, B) Loco-motor activity: distance traveled (A) and number of rearings (B).(C, D) Anxiety-like behavior in a novel environment.Immobile time (C) and time spent in the center of the field (D) were not affected by D-allose treatment.Error bars indicate SEM.Significant differences were determined by two-way ANOVA.*p < 0.05, **p < 0.001.

Fig. 4 .
Fig. 4. Spatial memory as assessed in the Morris water maze test.(A) Mean escape latency in the D-allose group was shorter than that of the control group during the latter half of training.(B) Mean swim distance to the submerged platform during spatial acquisition training was not affected by D-allose treatment.(C) Mean number of platform crossings during the probe test was significantly higher in the D-allose group than in the control group.(D) Mean time spent in the trained quadrant during the probe test was not affected by D-allose treatment.(E) Mean escape latency and (F) mean swim distance to the visible platform during cued training were not affected by D-allose treatment.Error bars indicate ± SEM.Significant differences were determined by two-way ANOVA (A, B and E, F) or Student's t-test (C, D). *p < 0.05.

Fig. 5 .
Fig. 5. Conditioned freezing in the Pavlovian fear-conditioning task.D-Allose did not affect conditioned freezing in short-and long-term cue-dependent fearmemory tasks nor in a context-dependent fear-memory task.Error bars indicate SEM.Significant differences were determined by Student's t-test.

Fig. 6 .
Fig. 6.Bone density and strength.(A-C) Bone strength of the left femurs was measured in terms of max load (A), displacement (B), and stiffness (C) using three-point bending tests.The max load in the D-allose group was significantly higher than that in the control group.(D) Bone density was not affected by D- allose.Bone density of the right femur was determined by dual energy X-ray absorptiometry (DXA).Error bars indicate SEM.Significant differences were determined by Student's t-test.*p < 0.05.

Table 1
Mean swim speed in the Morris water maze task.