Moderate-intensity exercise training improves long-term memory in fructose-fed rats

Aim: Animal disease model studies are widely used to show the effectiveness of physical exercise to improve cognitive function. Thus far, few studies are investigating the effects of exercise training on memory performance in fructose feed animals. Method: The present study investigated the effects of physical exercise protocol carried out with three weekly sessions, on the short and long-term memory performance of animals fed with fructose. Male Wistar rats were divided into sedentary (SD); sedentary+fructose (SDF); trained (TR); trained+fructose (TRF). Treadmill running sessions consisted of a five-minute warm-up at 20% maximum speed (MS) followed by 40 minutes at 40% MS and a 5-minute cool-down at 20% MS. Sessions were carried out three days a week (Monday, Wednesday, and Friday) for six weeks. Object Recognition Test was used to evaluate short and long-term memory. Results: The access to fructose altered food intake and drinking volume, as fructose-fed animals had lower food intake (SDF: -27% and TRF: -24%) and higher drinking volume (SDF: +49% and TRF: +45%) than an animal which drank water. Trained groups had lower epididymal fat pad compared to their sedentary counterparts (TR: -30% and TRF: -11%). In addition, TR and TRF had an improvement in glucose tolerance. Regarding memory performance, neither fructose intake nor exercise training influenced short-term memory. On the other hand, long-term memory was enhanced by exercise training. An improvement of about 39% was observed for TR and the largest effect was seen for TRF, which improved long-term memory in 76%. Conclusion: In conclusion, moderate-intensity exercise training, carried out three days a week, for six weeks was effective to improve long-term memory in fructose-fed rats. This result was related neither to the visceral fat amount nor to the glucose metabolism. Further studies should considerer the investigation regarding cerebral areas, associated with memory that might be adapted facing physical exercise.


Introduction
The increase in processed food intake, which has high fructose content, has been linked to a higher prevalence of cardiovascular diseases, non-alcoholic fatty liver disease, type 2 diabetes, and obesity 1,2 . Excessive fructose consumption induces oxidative stress, insulin resistance, high blood pressure, and impaired appetite hormone signaling 3 , features that could be associated with changes in cognitive function, which comprises the phases of the information process, perception, learning, memory, attention, surveillance, reasoning and problem solving, psychomotor functioning, reaction time, movement time and performance speed [4][5][6] .
The mechanisms in which excessive fructose intake might lead to cognitive impairment are not yet fully elucidated 7 . The hypotheses include insulin resistance that would lead to the functional decline of hippocampus 8,9 , increased blood pressure, oxidative stress, and inflammation that impairs short and longterm working memory 10 .
On the other hand, the beneficial effects of exercise training to prevent or manage cardiovascular 11 and metabolic diseases 12 is well recognized. Animal model studies carried out in our laboratory have shown that moderate-intensity aerobic exercise, performed five days a week, preserved endothelial function in rats fed a highly palatable diet rich in sugar and fat 13 , prevented weight gain and the accumulation of visceral fat 14 .
Regarding the impact of exercise training on cognitive performance, the majority of studies have investigated the effects of aerobic exercise training to revert memory impairment. Aerobic exercise, performed five days a week, was effective to counteract neurodegeneration, improve spatial learning and long-term memory in a rat model of Alzheimer 15 . Improvement in short-term memory and spatial learning was seen in rats that suffered cortex injury and were submitted to daily exercise sessions 16 , a similar result was seen in an intracerebral hemorrhage model 17 . Type 1 diabetic rats have also performed better in spatial memory testing after physical training 18 . One of the few studies carried out with healthy animals showed that Wistar rats submitted to aerobic exercise training had working memory and spatial memory improved 19 .
Nevertheless, an emerging question in sports science is related to the establishment of a minimal amount of physical exercise, which might cause improvement in different physiological systems function. It is already known that physiological adaptations, including cerebrovascular, are dependent on physical exercise intensity 20 .
Animal disease model studies are widely used to show the effectiveness of physical exercise to counteract neurodegeneration and improve cognitive function. These studies were carried out using exercise training protocols with five weekly sessions that take up to 60 minutes per session. Taking into account the prevalence of insufficient physical activity in adults this training protocol might not be feasible, as most people do not accumulate this amount of exercise per week 21 . Considering the importance of translational research, the present study investigated the effects of physical exercise protocol carried out with three weekly sessions, on the short and long-term memory performance of animals fed with fructose.

Animals and procedures
Forty male Wistar rats (444 ± 46 g; 70 days old) were divided into 4 groups according to the protocol of exercise training (sedentary or trained) and drink offered (water or fructose solution) as follows: sedentary (SD, n=9), sedentary + fructose (SDF, n=9), training (TR, n= 11), training + fructose (TRF, n=11). The animals were randomized into groups after a one-week treadmill familiarization. Only the rats, which were able to maintain a running behavior, were randomly allocated in trained groups.
Animals were housed in collective polypropylene cages (41×34×30 cm) (five animals/cage) and kept under a 12h light/ dark cycle. For the duration of the experiment (7 weeks), animals had free access to standard chow and drinking tap water or fructose solution (10%). Food intake was measured daily by weighing the remaining chow and subtracting the chow amount placed the day before on the cage top. The values were divided by the number of rats maintained in the cage. Water/fructose intake was also measured subtracting the remaining amount of liquid from the amount placed into the drinker. Body weight was also measured once a week. After 48h of the last exercise session, animals were submitted to overnight fasting followed by decapitation under anesthesia with Urethane (25%, 1.25 g/ kg). Epidydimal fat was excised and weighted. During fasting, the fructose solution was replaced by tap water.

Incremental maximal treadmill test and exercise training protocol
Before the incremental maximal test and training protocol, animals were submitted to a familiarization period with increments in running speed and minutes along the week (Table 1). After this period, rats underwent an incremental treadmill test (adapted from 22 ), beginning at 11.6 m/min followed by progressive increases of 1.6 m/min every 2 minutes until 20 m/ min. Subsequently, the speed was increased by 3.2 m/min until exhaustion (determined when the animal touched the bottom of the bay five times within one minute). The exhaustion speed was used to determined maximal speed (MS) using the following equation: MS= W1 + (W2 x t/120), where W1= exhaustion speed, W2= speed increase (1.6 m/min or 3.2 m/min), t= duration of the incomplete test stage 23 .

Exercise training protocol
Physical exercise sessions were carried out on Mondays, Wednesdays, and Fridays, in the same period of the day (11 a.m to 2 p.m.) for six weeks. Running sessions consisted of a five-minute warm-up at 20% MS followed by running at 40% MS for 40 minutes and a cool-down at 20% MS for 5 minutes. At the final of the 3 rd week, trained rats underwent Incremental Maximal Test, as previously described, to adjust speed training and guarantee the maintenance of exercise intensity during the six weeks.

Short and Long-term memory evaluation
Object Recognition Test was used to evaluate memory performance as previously described 24 . Briefly, on the first day, rats were submitted to a habituation session in an open field (45 × 40 × 60 cm), during which they were placed in the empty arena for 5 min. On the following day, rats were given one 5-min training trial in which they were exposed to two identical objects (A1 and A2). The objects were positioned in two adjacent corners, 9 cm from the walls. Ninety minutes after the training trial, rats were allowed to explore the open field for 5 min in the presence of two objects: the familiar object A and a novel object B, which are placed in the same locations as in the training trial. In this trial, short-term memory was evaluated. On the long-term memory testing trial (24 h after the training trial), rats were allowed to explore the open field for 5 min in the presence of two objects: the familiar object A and a third novel object C. All objects presented similar textures, colors, and sizes, but distinctive shapes. Object exploration was measured using two stopwatches to record the time spent exploring the objects during the experimental sessions. Sniffing or touching the object with the nose was considered as exploration behavior. The recognition index (RI) for each animal was determined using the following equation: RI = TB/(TA + TB), as TA: time spent exploring the knowing object; TB: time spent exploring the novel object.:

Glucose Tolerance Test
Being fasting for 8 hours, glycaemia was determined before and 30, 60, 90, and 120 minutes after intraperitoneal glucose injection (2g/kg) (25). Blood samples were collected from the animal tail to determine blood glucose using test strips (Accu-CheckAdvantage-Roche). The area under the curve was calculated using GraphPad Prism software.

Statistical procedures
Data are presented as means ± standard error of the mean (SEM). All data were assessed for normality using Kolmogorov and Smirnov test. Analysis of variance (two-way ANOVA) followed by Bonferroni post-hoc test was used to identify differences between groups. Factors analyzed were exercise (yes/no) and diet (fructose or water). The strength of association between independent variables (memory performance, glucose tolerance, and epididymal fat mass) was determined by the Pearson correlation coefficient. Data analysis was done using GraphPad Prism software. Statistical significance was considered at 5% (P<0.05).

Results
Body weight gain was not altered either by fructose (F (1;38) =0. 43 1C) and higher drinking volume (F (1;38) =24.01, p<0.001, fig 1D) than an animal which drank water. Interaction between physical exercise and diet was observed for drinking volume (F (1;38) =7,07, p=0.02), highlighting the different effects of physical exercise on drinking volume between trained groups. While SD and TR had also the same drinking volume in the 6 th week, TRF tends to drink less than SDF ( fig 1D).
Regarding memory performance, neither fructose intake (F (1;38) =0.22, p=0.64) nor exercise training (F (1;38) =0.00, p=0.99) influenced short-term memory. On the other hand, long-term memory is enhanced by exercise training. An improvement of about 39% was observed for TR and the largest effect was seen for TRF, which improved long-term memory by 76% ( figure 3).
A weak correlation was observed between long-term memory performance and glucose tolerance in trained animals (r=0.39, p=0.08). A similar result was verified between long-term memory performance and epididymal fat mass (r=-0.11, p=0.63) (figure 4).

Discussion
Modern humans live in an environment that demands very low daily physical activity, as a result, chronic diseases prevalence raises as sedentary behavior is increased 26 . In addition, our environment is full of food cues that may promote overeating of high-sugar, high-fat palatable foods which in turn increases cardiometabolic disease 27 . The negative impact of physical inactivity and poor nutritional habit on cognitive performance are poorly studied 5,28 .
Our results have shown that fructose intake, per se, did not provoke major body weight or epidydimal fat gain. Similar results were seen in Fisher rats fed with a 60% fructose diet 29 and C57BL/6 mice fed with a 40% fructose diet 30 . As expected, trained groups had lower epidydimal fat amounts compared to their sedentary counterparts. The effects of exercise training on the prevention of fat gain have already been demonstrated in rats fed with either standard chow or a high-fat diet 14,31 . Eating behavior was modified by fructose intake as SDF and TRF had lower food intake and higher drinking than SD and TR, which had water to drink. A similar result was observed in a previous study carried out with the same animal lineage and fructose solution concentration 32 . Exercise training carried out three days a week, did not affect eating behavior. However, other studies that submitted Wistar rats to five days a week treadmill training had demonstrated exercise effect to lower food intake 33,34 .
Regarding fructose-fed effects on cognitive function, Sprague-Dawley rats fed with fructose (10%) during eight months had impaired performance in spatial memory test 9 . A similar result was seen in C57BL/6J mice fed with fructose (15%) during eight weeks 35 . However, this study failed to demonstrate the detrimental effect of fructose-fed on memory performance. It might be related to the lower fructose concentration offered (10%) and/or the short-term duration of the experimental protocol (6 weeks).
On the other hand, the exercise training protocol proposed in the present study, which was carried out during six weeks with three days per week sessions was sufficient to enhance working memory performance in both trained groups. Improvement in spatial memory was seen in rats fed with a high-fat diet and submitted to treadmill running five days per week during eight weeks 36 . Mice fed with high fat or high fructose diets and submitted to treadmill running five days per week, during 23 weeks, did not have impairment on spatial memory and working memory 37 .
Our results demonstrated that trained animals had improvement in memory performance as well as better glucose metabolism and lower epididymal fat. Therefore, we investigate if peripheral factors influenced memory performance. It is well known that impaired peripheral glucose uptake and metabolism, due to insulin resistance, extends to impairment of insulin signaling within the hippocampus and impairs learning and memory 38 . A recent study demonstrates that insulin resistance, induced by LPS intracerebral injection, promotes memory impairment, which is nullified by intracerebral injection of insulin 39 . In addition, obesity appears to have a deleterious effect on cognition as low-grade systemic inflammation precipitates local inflammation within the hypothalamus that alters synaptic plasticity, contributes to neurodegeneration, and even initiates brain atrophy 40 .
However, a non-significant association between the area under the curve (from GTT) and the long-term memory recognition index was observed. A similar result was detected between epididymal fat, which is the indirect measure of visceral fat, and memory performance. Thus, in the present study, neither systemic glucose metabolism nor fat amount was factors that affect memory performance.
Probably, local mechanisms would have a greater association with this response. The impairment of cognitive function has been attributed to cerebral microvascular complications 41 as it was observed an inverse association between cerebral blood flow and cognitive performance in Alzheimer`s patients 42 and hypertensive middle-aged adults 43 . Endothelial dysfunction in brain capillaries would lead either to an increase in tau protein phosphorylation or its aggregation 44 . In addition, the increase of blood-brain barrier permeability, observed in high-fat diet mice, is related to inflammation and leucocyte recruitment precedes neurodegeneration and the memory performance decline 45 .
Physical exercise has long been recognized as a non-pharmacological approach that improves vascular function 31,34 , insulin signaling, and glucose metabolism 14,46 . The anti-inflammatory effect of physical exercise has also been described 47 . Recently, the existence of muscle-brain crosstalk was proposed. Myokines such as cathepsin B and irisin have been recognized as promoters of brain neurotrophic factor (BDNF) 48 . Physical exercise induces astrocytic plasticity and enhances cognitive performance in either healthy or diabetic rats 18 . The type and intensity of exercise training are a factor that might modulate brain adaptation response. In older rats, both moderate-intensity aerobic training and strength training improved spatial memory through distinct molecular mechanisms of neuroplasticity, showing that the effects of physical exercise on brain plasticity and spatial memory occur in an exercise type-dependent manner 49 . A recent study shows that Wistar rats, with cognitive impairment induced by scopolamine, submitted to treadmill running at moderate intensity (70% of maximal speed) exhibit better memory performance and higher expression of BDNF and synaptophysin in the hippocampus than rats trained at a lower intensity (40% of maximal speed) (CEFIS et al., 2019). However, this dose-response effect appears to be limited. Female mice submitted to treadmill running at high intensity had no improvement in spatial discrimination and had a lower response in terms of the increase in neurotrophic factors compared with mice that ran at moderate intensity 50 .
In conclusion, moderate-intensity exercise training carried out three days a week, for six weeks was effective to improve long-term memory in fructose-fed rats. This result was neither related to the visceral fat amount nor glucose metabolism. Further studies should considerer the investigation regarding cerebral areas, associated with memory that might be adapted facing physical exercise.