Klaus W.Lange,Ewelina Stollberg,Yukiko Nakamura,Joachim Hauser

Department of Experimental Psychology,University of Regensburg,93040 Regensburg,Germany

ABSTRACT Pattern separation keeps items distinct in memory and is mediated by the hippocampus.A relationship between hippocampal function and diet quality has been suggested by findings in both humans and animals.In the present study,rats were fed over seven generations a diet containing increased amounts of sugar and saturated fatty acids,reduced levels of polyunsaturated fatty acids and an increased ratio of omega-6/omega-3 fatty acids(“Western”diet).Spatial pattern separation(or local discrimination)performance of these animals was compared with that of rats fed a standard diet.A separation-dependent difference between the standard and Western diet groups was found in the number of discriminations performed in the pattern separation task,with rats of the“Western”group performing fewer discriminations.The present results suggest that behavioral assessment of spatial pattern separation can detect effects of dietary interventions in rats and that pattern separation can be impaired by transgenerational administration of a“Western”diet.Future studies should determine which components of this diet induce the memory impairments related to the hippocampus.The translational relevance of these findings in regard to mental disorders such as dementia and depression needs to be investigated.

Keywords:Memory Hippocampus Western diet Fatty acids Sugar

1.Introduction

Nutrition and diet play an important role in the promotion of long-term wellbeing and health as well as in the prevention of chronic diseases [1].While the association between dietary factors and physical health is well established,the relationship between mental disorders and quality of diet has only recently been acknowledged[2].A wide range of nutrients have been linked to brain development,neuronal functioning,and the prevalence of psychiatric disorders.For example,early studies exploring a nutritional basis of psychiatric symptoms concerned the dietary availability of amino acids such as tryptophan,a precursor of the neurotransmitter serotonin,which may play a role in the occurrence of symptoms of depression [3,4].Furthermore,associations between dietary patterns and the risk for depression have been shown in epidemiological studies [5].In addition,mental health outcomes in childhood appear to be influenced by maternal and early postnatal nutrition[6].

The currently available psychopharmacologicaltreatmentshave so far provided modest benefits in the management of mental disorders.Diet offers a modifiable variable in the prevention of psychiatric disorders,and the recognition of nutrition as a central factor in regard to mental health has therefore been advocated[2].Further studies using scientifically rigorous methods are needed to investigate the causal etiological role of nutrition and to assess the efficacy of dietary interventions in psychiatry.

While dietary therapeutic approaches in humans can be tested using randomized controlled trials,behavioral changes following experimental (long-term) variation of nutritional patterns cannot be investigated for ethical reasons.In this context,animal studies offer an opportunity to systematically vary nutrients and dietary factors and to assess their influence on brain functioning and behavior.

The hippocampus plays an essential role in memory(for review see [7,8]) and its function is involved in various mental disorders including major depression and Alzheimer’s disease [9].Severalspecific memory processes of the hippocampus have been investigated.One such process is pattern separation which entails the ability to separate memory components into distinct representations which cannot be easily confused[10].Pattern separation has been shown to be related to hippocampal function[11,12].A novel automated method for assessing spatial pattern separation(or location discrimination)in rats has recently been developed[13].Using this test,rats with lesions of the dorsal hippocampus were shown to be impaired when the stimuli to be discriminated were located closely together but not when they were far apart[14].

Table1 Composition of the two diets used.

Pattern separation has not,as yet,been used in studies investigating the association between diet and behavior.The present experiment has therefore introduced substantial changes in the experimental diet compared to the diet commonly used in laboratory rats in order to assess the ability of a spatial pattern separation paradigm to detect behavioral alterations following a dietary intervention.For this purpose,we have chosen an experimental diet with elevated levels of sugar,an increase in saturated fatty acids,a reduction in polyunsaturated fatty acids (especially omega-3 fatty acids)and an increased ratio of omega-6/omega-3 fatty acids.These dietary changes have been described as characteristic of the so-called “Western” diet [14].In addition,the experimental diet was administered transgenerationally,since it has previously been shown that,in rats,interventional diets may need to be administered over several generations in order to induce behavioral effects[15,16].

The aim of the present study was twofold:to investigate(1)the utility of spatial pattern separation,a behavioral process associated with the hippocampus,in the assessment of dietary interventions,and(2)the behavioral effects of the transgenerational administration of a“Western”diet on pattern separation.

2.Materials and methods

2.1.Animals

Eight generations of Wistar rats were bred in our laboratory and fed an experimental(“Western”)diet(Ssniff,Soest,Germany;based on American Institute of Nutrition – 93 G (AIN93 G),for details see Table1).The first generation male and female Wistar rats were delivered by Charles River Laboratories(Sulzbach,Germany).The study was performed with three-month-old male rats of the eighth generation of Western diet fed dams and three-month-old male Wistar rats(delivered by Charles River Laboratories,Sulzbach,Germany)fed a standard diet.Fifteen males from the Western diet group and 15 standard diet fed control males were tested.The rats were housed in groups of five animals per cage and kept on a 12:12 h light-dark cycle(room temperature 22°C,humidity 50%).They were given free access to water and their respective diets,except for the behavioral testing period.At the beginning of the behavioral experiments,water was provided ad libitum while food was restricted in order to prepare the animals for the experiments and to increase their motivation.Body weight was monitored daily.A weight reduction of more than 10–15% was avoided in order to prevent stress[17,18]and subsequent changes in the dopaminergic system[19].

2.2.Experimental(Western)and control(standard)diets

The Western diet was based on the AIN93 G form.The standard and Western diets were provided by Ssniff,Soest,Germany.Both diets met all current nutrition standards for rat growth [20]and had similar caloric content.The total amount of energy (Atwater)was 17.1 MJ/kg in the Western diet and 13.6 MJ/kg in the standard diet.The percentage of n-6 PUFAs as linoleic(LA,18:2 n-6)and n-3 PUFAs as alpha-linolenic acids(ALA,18:3 n-3)was 1.58%,0.01%in the Western diet and 1.90%,0.24% in the standard diet.Both diet compositions are shown in Table1.The Western diet was stored at a consistent－20°C.Fresh food was supplied daily.2

.3.Apparatus

The pattern separation task was based on the Bussey-Saksida Touch Screen Chambers(B-STSC;Campden Instruments,Loughborough,Leicestershire,England).A detailed description of the touch screen system as well as the training and test procedures can be found elsewhere[13,21,22].The experiment was performed using four ventilated wooden chambers (Campden Instruments,Loughborough,Leicestershire,England),each of which contained a box of stainless steel and black Perspex(trapezoid area 13–24.5 x 33.5 cm horizontally and approximately 30 cm high).The task was performed with a house light(3-Watt light bulb),a touchscreen system(24 x 28 cm)covered by a mask of Perspex,which included two horizontal rows of seven white squares(2 x 2 cm;distance:1 cm from each other)for the stimulus presentation,and an illuminated food tray on the wall opposite to the touch screen system.The two rows of squares were arranged horizontally on the rear wall and placed about 17 cm above the floor (stainless steel grid).In the present experiment,only the lower row of squares was used to display one of three different stimulus arrays,each of which consisted of two white squares (2 x 2 cm) denoting two locations.Three conditions were created by the separation of two locations by either one (small),three (medium) or five (large) empty squares.Nose pokes to touch screen and food tray were registered by photocells.A response consisting of a nose poke to the touchscreen was rewarded with a 45 mg dustless sucrose pellet (Bio-Serv,Frenchtown,New Jersey,USA),which was delivered into the food tray and accompanied by an acoustic signal.A false reaction was followed by a time-out period during which the houselights were turned on.

2.4.Procedure

The test assessed the pattern-separation ability using a twochoice spatial discrimination learning task.Two locations are presented in each trial,and the rat learns across trials to respond to the correct location.Touching the same location on the screen is rewarded in consecutive trials,which is then followed by a reversal of location-reward contingencies.During spatial discrimination training,locations 2 and 6 (see Fig.2,medium separation) were illuminated and one was designated as correct.The animals were rewarded for touching the correct location,while a 5-s time-out period followed when the wrong location was touched.If the rats touched the correct location in nine of 10 trials,the correct and incorrect locations were reversed.The pattern separation task consisted of six phases (schedules),the habituation,the Initial Touch schedule,the Must Touch schedule,the Must Initiate schedule,the Punish Incorrect Schedule and the pattern separation task (with three levels:large,medium and small,see Fig.1).All schedules were managed and executed by the computer programs Whisker and ABET II;all schedules including the criteria to pass one schedule were consistent with the procedures described in the instruction manual of the pattern separation task[23].

Fig.1.Schematic illustration presenting the three possible stimulus arrays on the touchscreen.Rats were trained initially using the“medium”stimulus array,followed by probe sessions on the“small”and“l(fā)arge”conditions.

2.5.Statistics

The statistical analysis of differences between the two groups(standard diet and Western diet) was performed with the Mann-WhitneyUtest using the Statistical Package for Social Sciences 20(SPSS)for Windows.An alpha level of 0.05 was applied.The number of discriminations performed(the number of correct switches between left and right side or vice versa)was analysed.

2.6.Ethics

All experiments were performed in accordance with the national laws(German law on Protection of Animals)and the principles of laboratory animal care(NIH publication No.86-23,revised 1996).The rats were handled according to the guidelines of the Federation for European Laboratory Animal Science Associations(FELASA) and were monitored daily for health concerns and body weight.Body weight was assessed in order to avoid weight loss as a consequence of restricted food access.In case of a reduction of body weight,the rats were fed individually and were given free access to food for more than three hours per day.

3.Results

The comparison between the standard and Western diet groups revealed differences for the number of discriminations performed in a session in the “medium” and “l(fā)arge” separation conditions(see Fig.2); the comparison between the groups was statistically significant for“the medium”sessions(Z=－2.195,p=0.028).

4.Discussion

Pattern separation is a mnemonic process that reduces overlap between similar inputs and allows memories to be kept distinct and resistant to confusion[10].Neuronal circuits within the hippocampus mediate pattern separation by detecting and storing similar input patterns as distinct representations [24].In the spatial pattern separation task used in the present study,rats were trained to discriminate locations presented on a touchscreen.The locations varied in their similarity,i.e.their distance from one another on the screen.

Fig.2.Number of separations performed in the pattern separation task (large,medium and small separations) for the standard and Western diet groups (means and standard errors).

The results showed that pattern separation in rats,as assessed using a touchscreen test [13],was impaired following transgenerational administration of a diet containing increased sugar and saturated fatty acid levels as well as reduced concentrations of polyunsaturated fatty acids compared to a standard diet commonly administered to laboratory rats.Rats with lesions of the dorsal hippocampus showed impaired pattern separation when the locations to be discriminated were close together but not when they were farther apart from each other [13].Other studies using a delayed nonmatching-to-sample task have also demonstrated separationdependent impairments following both a complete destruction of the hippocampus [25]and a selective lesion of the dentate gyrus[26].A separation-dependent deficit was also found in the present study,with a decrease in the number of discriminations performed at the medium and small separation levels(see Fig.2,statistically significant for“medium”)following transgenerational administration of a“Western”diet.This impairment is likely to be associated with hippocampal dysfunctioning [13,25,26].These findings suggest that the pattern separation paradigm used may be useful in the assessment of dietary effects on this specific memory function and that a“Western”diet can impair pattern separation.

In order to induce behavioral effects as a result of dietary changes in experimental animals,it may be necessary to maintain these changes over several generations.The question of whether a dietary reduction of omega-3 polyunsaturated fatty acids causes changes in attention and impulsive behavior in rats has recently been investigated.While a reduction over four generations showed no significant behavioral consequences [27],alterations in attention and impulsivity were observed following a transgenerational omega-3 polyunsaturated fatty acid reduction over more than seven generations [16].Future studies should assess pattern separation performance following alterations in single nutrients such as sugar or omega-3 fatty acids,including dose-response relations over fewer generations.In addition,neurobiological changes in the hippocampus and other brain areas should be investigated.

A relationship between diet quality and hippocampal function is suggested by findings in both animals and humans.Western diets high in saturated fats and refined carbohydrates appear to impair cognitive functions,damage brain regions associated with these functions,and contribute to the occurrence of neurodegenerative diseases (for review see [28]).However,the mechanisms involved in the association between the consumption of a Western diet and cognitive functions dependent on the hippocampus are unclear.Several physiological processes may be involved.Neurogenesis in the hippocampus is influenced by neurotrophins such as brain-derived neurotrophic factor,which are affected in animals by diets high in fat and refined sugar[29].Pattern-separated memories have been demonstrated to require brain-derived neurotrophic factor in the dentate gyrus[30].While pro-inflammatory processes in the brain have been shown to be up-regulated by high fat/sugar foods,thereby causing an increase in neurodegeneration and impairments in learning and memory[31],they can be reduced by the intake of omega-3 fatty acids [32].In addition,increased dietary consumption of omega-3 fatty acids may ameliorate or prevent inflammatory and autoimmune diseases[33].The findings of a study investigating the effects of Western diet maintenance in rats suggested that this diet reduces glucose and monocarboxylate transport in the hippocampus,which may lead to impaired learning and memory [34].Furthermore,a putative mechanism linking intake of Western diet,degradation of the blood brain barrier,hippocampal damage and dementia neuropathology has been suggested[35].

While Western diet intake in animals appears to be associated with impaired learning and memory functions related to the integrity of the hippocampus [36],the translational relevance of these findings remains to be investigated.In humans,a recent cohort study in participants without dementia aged ≥60 years reported that high adherence to a Western diet may increase cognitive decline[37],and a higher consumption of a Western dietary pattern was found to be associated with a decrease in hippocampal volume[38].

In summary,the present results suggest that spatial pattern separation(location discrimination),a memory function related to the hippocampus,may be able to detect the effects of dietary interventions and that pattern separation can be impaired by a diet containing increased levels of sugar and saturated fatty acids as well as reduced amounts of polyunsaturated fatty acids and an increase in the ratio of omega-6/omega-3 fatty acids (“Western” diet).Further studies could attempt to determine which components of this diet are responsible for the impairments in hippocampal function.Furthermore,the physiological and neurochemical mechanisms underlying the dietary effects on the hippocampus should be examined.Long-term consumption of Western diet has been demonstrated to produce oxidative stress,insulin resistance,low-grade inflammation,and cognitive impairment due to the generation of high levels of lipid peroxidation products [39].High concentrations of lipid peroxidation products,proinflammatory cytokines,and inflammatory mediators are associated with the pathogenesis of age-related brain diseases [39].Finally,the relevance of these findings for human conditions such as dementia and depression needs to be investigated.

Acknowledgement

The authors are grateful to Elisabeth Windhager for technical support.