Folate and vitamin B-12 deficiencies additively impaire memory function and disturb the gut microbiota in amyloid-β infused rats
Abstract
Abstract. Folate and vitamin B12(V-B12) deficiencies are associated with metabolic diseases that may impair memory function. We hypothesized that folate and V-B12 may differently alter mild cognitive impairment, glucose metabolism, and inflammation by modulating the gut microbiome in rats with Alzheimer’s disease (AD)-like dementia. The hypothesis was examined in hippocampal amyloid-β infused rats, and its mechanism was explored. Rats that received an amyloid-β(25–35) infusion into the CA1 region of the hippocampus were fed either control(2.5 mg folate plus 25 μg V-B12/kg diet; AD-CON, n = 10), no folate(0 folate plus 25 μg V-B12/kg diet; AD-FA, n = 10), no V-B12(2.5 mg folate plus 0 μg V-B12/kg diet; AD-V-B12, n = 10), or no folate plus no V-B12(0 mg folate plus 0 μg V-B12/kg diet; AD-FAB12, n = 10) in high-fat diets for 8 weeks. AD-FA and AD-VB12 exacerbated bone mineral loss in the lumbar spine and femur whereas AD-FA lowered lean body mass in the hip compared to AD-CON(P < 0.05). Only AD-FAB12 exacerbated memory impairment by 1.3 and 1.4 folds, respectively, as measured by passive avoidance and water maze tests, compared to AD-CON(P < 0.01). Hippocampal insulin signaling and neuroinflammation were attenuated in AD-CON compared to Non-AD-CON. AD-FAB12 impaired the signaling (pAkt→pGSK-3β) and serum TNF-α and IL-1β levels the most among all groups. AD-CON decreased glucose tolerance by increasing insulin resistance compared to Non-AD-CON. AD-VB12 and AD-FAB12 increased insulin resistance by 1.2 and 1.3 folds, respectively, compared to the AD-CON. AD-CON and Non-AD-CON had a separate communities of gut microbiota. The relative counts of Bacteroidia were lower and those of Clostridia were higher in AD-CON than Non-AD-CON. AD-FA, but not V-B12, separated the gut microbiome community compared to AD-CON and AD-VB12(P = 0.009). In conclusion, folate and B-12 deficiencies impaired memory function by impairing hippocampal insulin signaling and gut microbiota in AD rats.
Introduction
Alzheimer’s disease is a progressive neurodegenerative disease that finally develops into dementia. Alzheimer’s disease exhibits mild cognitive impairment in the early stage. Mild cognitive impairment is defined as the deterioration of learning, memory, perception, and central executive function. However, cognitive dysfunction from brain disease and cerebral dysfunction is not included [1]. Mild cognitive impairment is often the result of aging, and it is accelerated in the elderly and people with Alzheimer’s disease [1]. However, the cause of mild cognitive impairment remains unknown. The common pathway for Alzheimer’s disease and type 2 diabetes is known to include increased insulin resistance and inflammation [2]. Increased insulin resistance accelerates mild cognitive impairment leading to its progression to dementia [3]. Increased brain insulin resistance reduces glucose uptake into the neuronal cells that are linked to Alzheimer’s disease [2]. Thus, type 2 diabetes impairs energy metabolism in neurons which exacerbates mild cognitive impairment, leading to dementia including Alzheimer’s disease.
Some modifiable predictors of progression from mild cognitive impairment into dementia are depression, neuropsychiatric symptoms, insulin resistance, and impaired glucose metabolism [4]. Nutrient intakes, including folate, also acts as modifiable predictors [5]. Folate and vitamin B12 (V-B12) are involved in methionine metabolism and the dysregulation of methionine metabolism increases serum homocysteine [6]. Homocysteine is an intermediate in the catabolism of methionine, and elevated serum homocysteine levels are a risk factor for metabolic diseases associated with insulin resistance. A recent meta-analysis of randomized clinical trials confirmed that folate supplementation decreases fasting serum glucose and insulin concentrations and homeostasis model assessment for insulin resistance index (HOMA-IR) in type 2 diabetic patients [5]. Furthermore, elevated serum homocysteine concentration is a potential risk factor for mild cognitive impairment. Another recent meta-analysis [7] demonstrated that vitamin B supplementation (mainly folic acid, vitamin B6 and B12) lowers serum homocysteine levels in elderly patients with cognitive impairment, but there was no significant difference in cognitive function determined by Mini-Mental State Examination. However, another systematic review and meta-analysis [6] found a positive relationship between serum homocysteine levels and cognitive dysfunction. People with elevated serum homocysteine concentrations exhibit cognitive decline, but folate supplementation fails to reverse cognitive dysfunction once it appears [8]. Thus, it remains unclear that folate intake alters the course of mild cognitive impairment.
V-B12 along with folate is associated with homocysteine-methionine recycling by transferring one-carbon [9]. Both folate and V-B12 may be associated with glucose metabolism and mild cognitive impairment [9]. However, the balance of folate and V-B12 status may be more crucial. Indians who are mostly vegetarians (most Indian Hindus eat dairy products) easily become insulin resistance and have V-B12 deficiency, but with sufficient folate, which may increase the prevalence of type 2 diabetes and mild cognitive impairment. Moreover, excessive folate intake was associated with low plasma levels of V-B12 and higher homocysteine in the elderly in the Sacramento Area Latino Study on Aging [10]. Although the previous studies suggest that the ratio of folate and V-B12 plays a crucial role in insulin resistance and mild cognitive impairment, the results are inferred from cross-sectional studies [10].
Inflammation in various tissues is associated with insulin resistance and contributes to type 2 diabetes and dementia. Consumption of pro-inflammatory nutrients is associated with mild cognitive impairment in the elderly population [10, 11]. However, some nutrients, including folate, are considered to be anti-inflammatory nutrients and folate deficiency is associated with mild cognitive impairment [12], the mechanism has not been demonstrated. Recent studies explain that intestinal bacteria are important modulators of host immunity and systemic inflammation that is involved in insulin resistance-related diseases, including obesity [13]. However, few studies have been conducted to show that the deficiency of micronutrients including folate and V-B12 results in changes to the microbial community [14]. Therefore, we hypothesized that folate and V-B12 may alter mild cognitive impairment and glucose metabolism differently by changing inflammation through modulating the gut microbiome in amyloid-β infused rats. We also examined the hypothesis and its mechanism was explored.
Materials and methods
Animal care, and surgical procedures
Male Sprague Dawley rats that weighed an average of 202 ± 11 g were purchased from Daehan Bio, Inc. (Eum-Sung, Korea) and were acclimated in the animal facility. All rats were raised in individual stainless steel cages in a controlled environment (23 °C, 12-h light/dark cycle) and they had ad libitum access to food and water. All study procedures were adopted based on the Guide for the Care and Use of Laboratory Animals from the National Institutes of Health and were approved by the Institutional Animal Care and Use Committee of Hoseo University (HSIAUC-2014-06). After 1-week acclimation in the animal facility, the rats were anesthetized with an intraperitoneal injection of a mixture of ketamine and xylazine (100 and 10 mg/kg, respectively) and placed in a stereotaxic device. The cannula was inserted into the bilateral CA1 subregions of the hippocampus as previously described. The cannula was connected to 22-gauge tubing filled with amyloid-β(25–35) for the four AD groups or amyloid-β(35–25) for the Non-AD-CON group; amyloid-β(35–25) was used for the Non-AD-CON group as a control because it has the reverse sequence of amyloid-β(25–35) and is not toxic and does not aggregate in the brain [15]. Both types of amyloid-β were dissolved in sterile saline and infused into the bilateral CA1 subregions using an osmotic pump (Alzet Osmotic Pump Company; Cupertino, CA, USA) at a rate of 3.6 nmol/day for 14 days. At the initial day of amyloid-β infusion, the assigned diet was provided in each group for 8 weeks.
Diet preparation and experimental groups
All rats had free access to a modified polyphenol-free AIN-93 semi-purified high-fat diet, which has been shown to induce obesity and insulin resistance [16]. The vitamin mixture was made without folate and/or V-B12. Control diet supplemented 2.5 mg folate plus 25 μg V-B12/kg diet and no folate diet added 25 μg V-B12/kg diet. No V-B12 diet included 2.5 mg folate/kg diet and no folate plus no V-B12 did not contain any folate and V-B12. Each vitamin mixture was mixed with the proper amounts of starch, casein, lard and mineral mixture, sifted again, and stored at 4 °C. The respective diet was provided by pressing the weighed powder tightly into the food container every other day and the remaining feed was weighed and discarded. The composition of the diet was 35% carbohydrate (cornstarch and sucrose), 22% protein (casein), and 43% fat (lard). The major carbohydrate, protein and fat sources were starch plus sugar, casein (milk protein) and lard (CJ Co., Seoul, Korea), respectively. Thus, the difference among the groups was the contents of folate and V-B12 in the diets.
The rats were divided into five treatment groups of 10 rats each: AD model rats that received a high-fat diet (AD-CON), no folate diet (AD-FA), no V-B12 diet (AD-VB12), and no folate plus V-B12 diet (AD-FAB12). The sham-operated rats received the same diet as AD-CON (Non-AD-CON; a high-fat diet).
Body composition
Body composition was examined before hippocampal infusion and before the rats were sacrificed. Prior to euthanizing the rats, they were laid in a prone position with their hind legs maintained in external rotation and hip, knee and ankle articulations in 90° flexion with tape. The body was scanned by the dual-energy X-ray absorptiometry method using an absorptiometer (pDEXA Sabre; Norland Medical Systems Inc., Fort Atkinson, WI, USA). The facility was equipped with the appropriate software for assessment in small animals. Bone mineral density was measured in the lumbar spine and femur, and lean body mass and fat mass were determined in the abdomen and leg [16].
Metabolic analyses
At 7 week of the experiment, an oral glucose tolerance test was conducted after an overnight fast using an oral administration of glucose (2 g/kg). During the oral glucose tolerance test, serum glucose levels were measured every 10 min until 90 min and again at 120 min, and serum insulin levels were measured at 0, 20, 40, 60, 90, and 120 min. The serum glucose and insulin levels were measured using a Glucometer (Accuchek, Roche Diagnostics; Basel, Switzerland) and radioimmunoassay kit (Linco Research, Inc.; St. Charles, MO, USA), respectively.
At the end of the experiment, the rats were anesthetized with ketamine and xylazine (100 and 10 mg/kg body weight, respectively). Blood was collected by cardiac puncture and the serum was separated by centrifugation at 3,000 rpm for 20 min. After blood collection, human insulin (5 U/kg body weight) was injected through the inferior vena cava to determine insulin signaling. Serum and tissue samples were stored at –70 °C for biochemical analysis. The epididymal and retroperitoneal fat masses and uteri were then removed and weighed. HOMA-IR was calculated as follows: serum insulin (μU) × serum glucose (mmol/L)/22.5. Serum tumor necrosis factor-α (TNF-α) and homocysteine levels were measured using enzyme-linked immunosorbent assay (ELISA) kits from eBioscience (San Diego, CA, USA) and Crystal Chem (Elk Grove Village, IL, USA), respectively.
Energy expenditure
Energy expenditure was conducted by indirect calorimetry during the dark phase following 6 h of fasting at 3 days after intraperitoneal insulin tolerance test. The rats were placed in a computer-controlled metabolic chamber with the respiratory chamber with a computer-monitored O2 and CO2 system and 800 mL/min airflow (Biopac Systems Inc., Goleta, CA, USA). The CO2 production (VCO2) and O2 consumption (VO2) of the rat were measured every minute for 30 min and the average VO2 and average VCO2 were integrated. The respiratory quotient (RQ) was calculated as VCO2/VO2 and VO2 and VCO2 values were adjusted for metabolic body size (kg0.75) [16]. Resting energy expenditure and oxidations of fat and carbohydrate calculated as previously described [16].
Memory function by passive avoidance and water maze tests
At the 8th week, the rats were tested for memory deficits using a passive avoidance apparatus consisting of a two-compartment dark/light shuttle box [16, 17]. Rats were put into the light-box and they immediately moved into the dark-room. In the acquisition trial, electroshocks (75 V, 0.2 mA, 50 Hz) were delivered for 5 s, when the rats entered into the dark chamber. Five seconds later, the rat was removed from the dark chamber and returned to its home cage. After 24 and 48 h from the 1st trial, 2nd and 3rd trials were conducted. The retention latency time to enter the dark chamber was measured in the same way as in the acquisition trial but electric foot shock was not delivered. The rats remembered the electric shock from the first trial then they did not enter into the dark-room. Shorter latencies to enter the dark-room indicate memory deficit. The latency time to enter the dark-room was recorded to a maximum of 600 s.
Spatial memory function was assessed with a Morris water maze test, as previously described [15, 17], at day 49. The Morris water maze tests hippocampal-dependent learning, including the acquisition of spatial memory. At the 1st trial, the platform was located in zone 5 of the water pool and rats started to swim from zone 1 that was the opposite side of zone 5. Water maze test was conducted on day 1, 2 3 and 5. The point for entry of the rat into the pool and the location of the platform for the escape remained unchanged for each trial. On day 5, the platform was removed and the time to go to zone 5 and staying in zone 5 were measured. Each test was conducted with a cut-off time of 600 s. The longer period to find the platform represented memory deficient.
Realtime quantitative reverse transcription-polymerase chain reaction (RT-PCR)
Hippocampal tissues were randomly excised from 4 rats from each group and total RNA was isolated using a monophasic solution of phenol and guanidine isothiocyanate (Trizol reagent, Invitrogen, Rockville, MD, USA). The cDNA was synthesized from total RNA with superscript III reverse transcriptase and high fidelity Taq DNA polymerase (1:1:1, v:v:v) by reverse transcription reaction in PCR. The synthesized cDNA was mixed with the primers of the interest genes and sybergreen mix and their expressions were analyzed using a real-time PCR machine (BioRad Laboratories, Hercules, CA, USA). The primers of TNF-α, interleukin (IL)-1β, and β-actin were given in previous studies [15, 16]. The expression levels of genes of interest were quantitated using the cycle of threshold method [18].
Immunoblot analysis
Hippocampal tissues from four rats from each group were dissected as previously described [15]. Each tissue was lysed with RIPA lysis buffer adding protease inhibitors, and their protein contents were measured using a Bio-Rad protein assay kit (Hercules, CA, USA). The lysates with equivalent amounts of protein (30–50 μg) were resolved into sodium dodecyl sulfate-polyacrylamide gel electrophoresis and the amount of the interested protein was examined by immunoblotting with the specific antibodies as follows: protein kinase B (PKB or Akt), phosphorylated PKBSer473, glycogen synthase kinase (GSK)-3β, phosphorylated GSK-3βser9, phosphorylated tauser396 and tau (Cell Signaling Technology, Danvers, MA, USA) and β-actin (Santa Cruz Biotech, Dallas, TX, USA). The intensity of the proteins of interest was measured using Imagequant TL (Amersham Biosciences, Piscataway, NJ).
Next-Generation Sequencing gut microbiome
The gut microbiome community was examined from the feces by analyzing metagenome sequencing using next-generation sequencing procedures [19]. Bacterial DNA was extracted from the feces of each rat using a Power Water DNA Isolation Kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions. The DNA was amplified with 16S amplicon primers using PCR and each library was prepared using the PCR products according to the GS FLX plus library prep guide. The emPCR, corresponding to clonal amplification of the purified library, was carried out using the GS-FLX plus emPCR Kit (454 Life Sciences, Branford, CT). Libraries were immobilized onto DNA capture beads. The library-beads were added to the amplification mix and oil and the mixture was vigorously shaken on a Tissue Lyser II (Qiagen, Valencia, CA) to create “micro-reactors” containing both amplification mix and a single bead. The emulsion was dispensed into a 96-well plate and the PCR amplification program was run with 16S universal primers in the FastStart High Fidelity PCR System (Roche, Basel, Switzerland) according to the manufacturer’s recommendations. Sequencing of bacterial DNA in the feces was measured by Illumina MiSeq standard operating procedure by a Genome Sequencer FLX plus (454 Life Sciences) in Macrogen Ltd. (Seoul, Korea).
Bacterial sequence processing
The 16S amplicon sequences were processed using Mothur v.1.36. We followed the Miseq SOP to identify the taxonomy and counts of the bacteria in each fecal sample. We aligned the sequences using Silva reference alignment v.12350. In a preclustering step, the sequences with identity 99% were merged. The chimeric sequences were detected and discarded by UCHIME. All the sequences were assigned with taxonomic classifications using Greengenes 13_8_99 and the sequences classified as mitochondria, Eukaryota or unknown were removed. We conducted the picking of operational taxonomic units (OTUs) delimited at 98% identity, which was taxonomically classified by consensus using Greengenes 13_8_99. A relaxed neighbor-joining tree with one representative sequence per OTU was obtained with Clearcut after calculating uncorrected pairwise distances between aligned reads. The relative amount of bacteria was calculated in the taxonomic assignments [order] level of each sample. Principle coordinate analysis (PCoA) in gut bacteria were visualized by R package.
Statistical analysis
All statistical analyses were conducted using SAS version 7 (SAS Institute; Cary, NC, USA). All results are expressed as a mean ± standard deviation (SD). All variables were checked for having a normal distribution in the univariate analysis. One-way analysis of variance was used to compare the groups when the results were measured only once at the end of the experiment. Multiple comparisons among the groups were conducted using Tukey’s test. The statistical differences of PCoA in gut bacteria were analyzed by analysis of molecular variance in Mothur. A p-value < 0.05 was considered as statistically significant.
Results
Body-weight, food intakes, and energy expenditure
Body-weight at the 7th week was lower in AD-FA and AD-FAB12 than AD-CON, Non-AD-CON, and AD-B12. Body-weight gain during the experimental periods was much lower in AD-FA than the other groups (Table 1). However, epididymal fat and retroperitoneal fat mass, representing visceral fat, was significantly higher in AD-CON than Non-AD-CON; AD-B12 increased visceral fat compared to AD-CON but AD-FA and AD-FAB12 did not alter from AD-CON (Table 1).
The food intakes did not significantly differ among all groups although AD-FA tended to decrease food intake (Table 1). Interestingly, energy expenditure was higher in the AD-CON than the Non-AD-CON and it increased more in AD-FA than the AD-CON (Table 2). Fat oxidation was lower in the AD-CON than the Non-AD-CON and it was lowered in AD-VB12 and AD-FAB12 than AD-CON (Table 2). However, fat oxidation was much higher in AD-FA than AD-CON. CHO oxidation was opposite to fat oxidation (Table 2).
Serum concentrations of TNF-α, an index of inflammation, were much higher in AD-CON than Non-AD-CON and AD-FA, AD-VB12 and AD-FB12 increased serum TNF-α concentrations compared to AD-CON (Table 3). Serum homocysteine concentrations were much higher in AD-CON than Non-AD-CON and it increased in ascending order of AD-CON, AD-FA, AD-VB12, and AD-FAB12 (Table 3). Therefore, V-B12 and folate deficiency suggested the exacerbation of insulin resistance in AD rats.
Body composition
Bone mineral density was decreased in AD-CON compared to the Non-AD-CON in the lumbar spine and femur whereas it was not changed by the deficiency of folate and V-B12 in AD rats (Figure 1A). Unlike bone mineral density, lean body mass was not different in the hip and leg between AD-CON and Non-AD-CON but lean body mass was lower in the hip in the AD-FA, AD-VB12, and AD-FAB12 and AD-FA lowered the hip lean body mass the most among all groups (Figure 1B).
Memory impairment
Memory loss was determined by passive avoidance and water maze tests. The rats had experienced an electric shock upon entering the dark-room in the 1st and 2nd trial and the shorter retention time to enter into the dark-room indicated the rat remembered the electric shock, causing it not to enter the dark-room. Shorter retention time represented the short-term memory in 2nd and 3rd trials during the passive avoidance test. The AD-CON group showed a shorter retention time to enter the dark-room than the Non-AD-CON group (Fig 2A). The retention times of the AD-FA and AD-VB12 were similar to the AD-CON, but it was shorter in the AD-FAB12 group in the 3rd trial (Fig 2A). Thus, the deficiency of both folate and vitamin-B12 lowered short-term memory. Furthermore, latency to the first visit into the zone 5 where the platform existed was higher in the AD-CON then the Non-AD-CON in the water maze test representing spatial memory, and it was similar between AD-CON and AD-FA (Fig, 2B). The latency in the water maze test was shorter in the AD-VB12 and longer in AD-FAB12 than AD-CON. Duration of the retention times and frequencies to zone 5 were the opposite to the latency of the first visit to zone 5. They were lower in the AD-CON then the Non-AD-CON (Figure 2B). The duration of the retention time was lower in AD-FA, AD-VB12, and AD-FAB12 than AD-CON whereas the frequencies were lower in AD-FAB12 the AD-CON (Figure 2B). Therefore, the combination of folate and V-B12 deficiency worsened short-term memory and spatial memory in comparison to AD-CON.
Insulin signaling and neuroinflammation in the hippocampus
AD-FA and AD-VB12 diets decreased the phosphorylation of Akt and GSK-3β in the hippocampus and these genes are downstream from insulin receptors and AD-FAB12 additively reduced their phosphorylation (P < 0.05; Figure 3A). Due to the attenuation of hippocampal insulin signaling, the phosphorylation of tau was potentiated in the AD-FA, AD-VB12 and AD-FAB12 groups and AD-FA increased the tau phosphorylation the most (Figure 3A). Thus, both folate and V-B12 deficiency additively attenuated hippocampal insulin signaling and it might stimulate amyloid-β accumulation in the hippocampus.
The expressions of TNF-α and IL-1β in the hippocampus increased in AD-CON compared to Non-AD-CON and their expressions increased in AD-FA and AD-VB12 more than the AD-CON (Figure 3B). Their expression was additively elevated in AD-FAB12 in comparison to AD-CON (P < 0.05; Figure 3B). Thus, folate and V-B12 deficiency additively increased neuroinflammation.
Glucose tolerance
Serum glucose levels at the overnight fasting state were higher in AD-CON than Non-AD-CON. The levels in AD-FA, AD-VB12 and AD-FAB12 were similar to the AD-CON (Table 3). Serum insulin levels at fasting state were significantly higher in the AD-CON than the Non-AD-CON whereas they increased only in the AD-FAB12 compared to the AD-CON (Table 3). HOMA-IR calculated from serum glucose and insulin levels at fasting state was much higher in AD-CON than Non-AD-CON and it was elevated in the ascending order of AD-CON, AD-FA, AD-VB12, and AD-FAB12 (Table 3).
After a glucose challenge, serum glucose levels increased until 30–40 min and then decreased from the peak level (Figure 4A). Serum glucose levels at the peak were much higher in the AD-CON than the Non-AD-CON whereas the folate and V-B12 deficiency did not alter the peak level of serum glucose (Figure 4A). After the peak, serum glucose levels decreased and they were not significantly different among AD-CON, AD-FA, AD-VB12, and AD-FAB12 (Figure 4A). The area under the curve (AUC) of serum glucose levels in the first and second parts was much higher in AD-CON than Non-AD-CON (Figure 4B). It was not changed by folate and V-B12 deficiency in AD rats (Figure 4).
Serum insulin concentrations increased until 20 min in AD-CON, AD-FA, and Non-AD-CON, but AD-FA, AD-B12, and AD-FAB12 increased serum insulin levels until 40 min (Figure 4C), after which the serum insulin concentrations decreased in all groups. Serum insulin concentrations during oral glucose tolerance test were lower in AD-FA than the other group but the levels were higher in AD-FA than Non-AD-CON. The peak levels at 40 min increased similar to AD-CON (Figure 4C). The concentrations at 40 and 60 mins were much higher in AD-VB12 and AD-FAB12 than the AD-CON (Figure 4C). AUC of the 1st part of serum insulin concentrations was much higher in AD-CON than Non-AD-CON (Figure 4D). The AUC was higher in AD-FAB12 than the AD-CON. AUC of the 2nd part of serum insulin concentrations was higher in AD-CON than Non-AD-CON (Figure 4D). The AUC in AD-CON was similar to that in AD-FA and it was lower in AD-VB12 and AD-FAB12 (Figure 4D). Thus, V-B12 deficiency increased insulin resistance in AD rats.
Gut microbiome
Principal coordinate analysis (PCoA) shows the clustering of the gut bacterial community in the groups (Figure 5A). The AD-CON group exhibited a separation from the Non-AD-CON and only AD-FA group had a clear separation of PcoA with AD-CON (Figure 5A). In addition, the gut microbial community was modulated in AD-FA compared to AD-CON and Non-AD-CON. AD-FAB12 did not make a separate cluster from the other groups. Thus, AD-CON somewhat changed the gut bacterial community from Non-AD-CON and folate deficiency influenced the gut microbiome the most in AD rats.
The significant differences of gut microbiota community among all groups were examined by analysis of molecular variance (P = 0.009). The gut microbiome community was significantly separated between AD-CON and Non-AD-CON groups (P = 0.013) (Figure 5B). The relative counts of Bacteroidia were significantly higher in Non-AD-CON than AD-CON and they were significantly different between AD-CON and AD-FA (Figure 5B). Those of Clostridia showed an opposite pattern of Bacteroidia, but they were not significantly different among the groups. Some proteobacteria such as Erysipelotrichi and Epsilonproteobacteria were higher in AD-CON than Non-AD-CON (Figure 5B). V-B12 deficiency increased Gammaproteobacteria and Alphaproteobacteria but decreased Deferribacteres and Actinobacteria compared to AD-CON (Figure 5B).
Discussion
Folate and V-B12 play important roles in methionine metabolism and their deficiency is associated with the increase of serum homocysteine concentrations, an index of insulin resistance [20]. However, folate is involved in thymidine synthesis without the requirement of V-B12 [21] and V-B12 is associated with the metabolism of odd-chain fatty acids without folate requirement. Folate deficiency without V-B12 deficiency during pregnancy is known to increased risk of neural tube defects in offspring, indicating that folate is involved in the development and growth of neuronal cells [22]. Furthermore, folate and V-B12 deficiencies are associated with increased risk of cognitive impairment, depression, Alzheimer’s disease and stroke in elderly adults [12]. However, a direct effect of their deficiency on energy, glucose metabolism, memory function, and the gut microbiome has not been investigated. The present study hypothesized that folate and V-B12 would influence mild cognitive impairment and energy and glucose metabolism by modulating inflammatory status and the gut microbiome in amyloid-β infused rats. We tested the hypothesis in hippocampal amyloid-β (25–35) infused rats. The study suggested that the combination of folate and V-B12 deficiencies additively impaired memory function by attenuating hippocampal insulin signaling but their deficiencies differently influenced systemic insulin resistance in amyloid-β infused rats. Both folate and V-B12 deficiency increased serum homocysteine levels by modulating methionine metabolism. However, they differently affected body composition and V-B12 deficiency increased adiposity additively increasing insulin resistance.
Folate and V-B6 and V-B12 are prominently involved in homocysteine metabolism [23–25]. Deficiency of these vitamins causes hyperhomocysteinemia, which is associated with the disturbance of lipid and glucose metabolism and cognitive dysfunction [25]. A high homocysteine level is associated with low V-B12 and folate levels in Alzheimer’s diseases not due to their deficient intake [26]. In the present study, V-B12 deficiency and folate deficiency increased serum homocysteine levels and co-deficiency further elevated the levels. Serum homocysteine levels were positively correlated with insulin resistance. Memory function was impaired with deficiency of both folate and V-B12 individually, and co-deficiency additively deteriorated memory function in rats with Alzheimer’s disease-like dementia. This might be due to the elevated serum homocysteine levels, which would be consistent with the results that disturbed homocysteine metabolism contributes to amyloidogenesis by modifying the amyloid precursor protein production and processing [27]. It is explained by the results that high plasma folate levels are associated with the exacerbation of both biochemical and clinical states under V-B12 deficiency [28]. Therefore, folate and V-B12 deficiency additively exacerbated memory dysfunction in the hippocampal amyloid-β infused rat.
Interestingly, folate and V-B12 differently influenced body composition and modulated energy metabolism. Folate deficiency decreased body weight gain but the absolute fat mass was not significantly different from AD-CON and V-B12 deficiency increased the fat mass in AD-CON rats in the present study. Co-deficiency of folate and V-B12 increased adiposity. It indicated that body composition was altered by folate and V-B12 deficiency in an independent mechanism. V-B12 deficiency is known to cause methylmalonic acid accumulation and increase of odd-chain fatty acid synthesis. V-B12 deficiency inhibits branched-chain amino acid to elevate adipogenesis [29]. It explained the increase of adiposity with V-B12 deficiencies in the present study. In addition, V-B12 deficiency had a positive association with body weight and adiposity during pregnancy in a non-diabetic White British Population [25]. However, high intakes of folic acid are associated with a lower BMI at the age of 6 years, opposite to the present study and early higher methionine intakes are negatively associated with body weight at the age of 6 y [30]. This may be associated with something other than just folate deficiency in humans. However, Fan et al. have demonstrated that NADPH production is dependent on folate by the cells with knockdown of methylene tetrahydrofolate dehydrogenase (MTHFD) genes [31]. Knock-down MTHFD genes increased oxidative stress by decreasing the reduction of oxidative glutathione [31]. Since NADPH is also related to fatty acid and cholesterol synthesis, it may be associated with reducing growth and decreasing fat contents. Since folate is associated with thymidine synthesis [21] and it may be associated with muscle synthesis. Folate deficiency decreased lean body mass possibly due to the impairment of thymidine synthesis in the present study. Therefore, deficiency of both folate and V-B12 showed the summation of each deficiency and it indicated that the deficiency of the two vitamins may not be involved in a common pathway to regulate energy metabolism.
Alzheimer’s disease has a positive relationship to osteoporosis [32], and it also increases the risk of hip fracture [33]. In addition, folate and V-B12 are related to a risk factor of bone mineral density and elevated serum homocysteine levels are known to be a risk factor for osteoporosis. However, the relationship remains controversial [34, 35]. The present study was consistent with previous studies: the AD rats showed a lower bone mineral density than the non-AD rats. In addition, folate and V-B12 deficiency increased serum homocysteine concentrations and decreased bone mineral density more than AD-CON.
Alzheimer’s disease and type 2 diabetes have a common etiological pathway of insulin resistance, and the insulin signaling pathway is attenuated in both diseases [2]. Glucose uptake is reduced in neuronal cells to modulate glucose homeostasis in the peripheries and vice versa [36]. Hippocampal insulin resistance induces amyloid-β deposition to initiate the death of neuronal cells, which exacerbates insulin resistance, thus creating a vicious cycle to exacerbate insulin resistance and amyloid-β deposition [37]. The present study also showed that amyloid-β deposition impaired glucose intolerance and folate and V-B12 deficiency further exacerbated glucose intolerance. Glucose tolerance is determined by the balance of insulin resistance and insulin secretion. Insulin secretion was higher in the AD-Con than the Non-AD-CON whereas folate and V-B12 deficiency differently impaired glucose-stimulated insulin secretion. Folate deficiency impaired the insulin secretion pattern and insulin secretion was delayed when serum glucose concentrations were elevated. However, V-B12 deficiency highly elevated insulin secretion due to increased insulin resistance. Thus, V-B12 deficiency increased insulin resistance to exacerbate memory dysfunction in rats with Alzheimer’s disease. However, folate deficiency increased phosphorylation of tau in the hippocampus more than V-B12 deficiency. Thus, each deficiency of folate and V-B12 differently exacerbated memory function although both deficiencies deteriorated hippocampal insulin resistance.
The gut microbiome is known to be closely related to neuroinflammation and brain insulin resistance which lead to the development of Alzheimer’s disease, although the specific mechanism remains unclear [3]. Gut microbiota has bidirectional interactions between the gut and the brain that can alter the development and progression of neurodegenerative and metabolic diseases [38]. Obesity and type 2 diabetes are associated with reduced bacterial diversity and bacterial composition and they may develop as a result of changes in the composition of gut microbiota. Type 2 diabetes has a direct association with the pathogenesis of Alzheimer’s disease [39]. Their etiology is related to a state of chronic and low-grade inflammation that contributes to type 2 diabetes and Alzheimer’s disease [39, 40]. Hence, understanding the microbiota of the gut important to understanding the pathogenesis of chronic inflammation, as it is a key contributor to diabetes which has a direct relationship to the pathogenesis of Alzheimer’s disease. Hoffman et al. [41] presented results that were consistent with the present study, that old mice aged 18–20 months increased the ratio of Firmicutes to Bacteroidetes and diversity of the microbiome more than young mice aged 5–6 months. Old mice had impaired cognitive functions with higher body weight. Thus, the changes in the gut microbiome might be associated with body weight and cognitive function additively. However, in the present study, the relative amount of Clostridia was higher in AD-CON than Non-AD-CON whereas AD-FA lowered Clostridia and increased Bacteroidia than AD-CON. Rats in the AD-CON and Non-AD-CON showed similar body weights, but AD-CON exhibited impaired cognitive function compared to Non-AD-CON due to induced dementia. AD-FA decreased body-weight but worsened memory function. Thus, the changes in gut microbiome were additively influenced by body weight and cognitive function.
The strength of the study was to demonstrate that folate plus V-B12 deficiency additively impaired memory function and attenuated hippocampal insulin signaling in an Alzheimer’s-type dementia rat model. V-B12 deficiency might be damaging to both glucose metabolism and cognition with developing gut microbiome dysbiosis in rats with Alzheimer’s-like dementia, and that folate deficiency can exacerbate the effects. In the limitation of this study, rats had sufficient single carbon sources (methionine and choline) and it might mitigate the exacerbation of memory function by folate and V-B12 deficiency. In conclusions, folate deficiency itself impaired memory function with no additional increase of serum insulin levels and it differently modulated gut microbiota compared to the AD-CON. However, V-B12 deficiency also deteriorated memory loss with exacerbated insulin resistance in comparison to AD-CON. Both folate and V-B12 deficiency additively impaired memory function by attenuating hippocampal insulin signaling. It might be involved in the results that folate and V-B12 deficiency differently influenced systemic insulin resistance and gut microbiota to modulate memory function in amyloid-β infused rats.
References
1 (2017) Mild cognitive impairment in type 2 diabetes mellitus and related risk factors: a review. Rev Neurosci. 28, 715–723.
2 (2018) Brain insulin resistance in type 2 diabetes and Alzheimer disease: concepts and conundrums. Nat Rev Neurol. 14, 168–181.
3 (2018) Neuroinflammation, Gut Microbiome, and Alzheimer’s Disease. Mol Neurobiol.
4 (2015) Modifiable predictors of dementia in mild cognitive impairment: a systematic review and meta-analysis. Am J Psychiatry. 172, 323–334.
5 (2018) The effects of folate supplementation on glucose metabolism and risk of type 2 diabetes: a systematic review and meta-analysis of randomized controlled trials. Ann Epidemiol. 28 (249–257): e241.
6 (2016) Homocysteine and cognition: a systematic review of 111 studies. Neurosci Biobehav Rev. 69, 280–298.
7 (2017) Efficacy of vitamin B supplementation on cognition in elderly patients with cognitive-related diseases. J Geriatr Psychiatry Neurol. 30, 50–59.
8 (2017) Interaction between excess folate and low vitamin B12 status. Mol Aspects Med. 53, 43–47.
9 (2014) The role of the one-carbon cycle in the developmental origins of Type 2 diabetes and obesity. Diabet Med. 31, 263–272.
10 (2018) An inflammation-related nutrient pattern is associated with both brain and cognitive measures in a multiethnic elderly population. Curr Alzheimer Res. 15, 493–501.
11 (2014) Inflammation, defective insulin signaling, and mitochondrial dysfunction as common molecular denominators connecting type 2 diabetes to Alzheimer disease. Diabetes. 63, 2262–2272.
12 (2017) Vitamin D, homocysteine, and folate in subcortical vascular dementia and Alzheimer dementia. Front Aging Neurosci. 9, 169.
13 (2018) Gut microbiome may contribute to insulin resistance and systemic inflammation in obese rodents: a meta-analysis. Physiol Genomics. 50, 244–254.
14 (2017) The effects of micronutrient deficiencies on bacterial species from the human gut microbiota. Sci Transl Med. 9.
15 (2013) Yuzu extract prevents cognitive decline and impaired glucose homeostasis in beta-amyloid-infused rats. J Nutr. 143, 1093–1099.
16 (2017) Intermittent fasting reduces body fat but exacerbates hepatic insulin resistance in young rats regardless of high protein and fat diets. J Nutr Biochem. 40, 14–22.
17 (2017) Agrimonia pilosa Ledeb., Cinnamomum cassia Blume, and Lonicera japonica Thunb. protect against cognitive dysfunction and energy and glucose dysregulation by reducing neuroinflammation and hippocampal insulin resistance in beta-amyloid-infused rats. Nutr Neurosci. 20, 77–88.
18 (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 25, 402–408.
19 (2018) Low dose brain estrogen prevents menopausal syndrome while maintaining the diversity of the gut microbiomes in estrogen-deficient rats. Am J Physiol Endocrinol Metab.
20 (2009) Cobalamin, folic acid, and homocysteine. Nutr Rev. 67 (Suppl 1): 69–72.
21 (2015) Maternal dietary uridine causes, and deoxyuridine prevents, neural tube closure defects in a mouse model of folate-responsive neural tube defects. Am J Clin Nutr. 101, 860–869.
22 (2017) Associations of maternal folic acid supplementation and folate concentrations during pregnancy with foetal and child head growth: the Generation R Study. Eur J Nutr. 56, 65–75.
23 (2012) Cognitive impairment and vitamin B12: a review. Int Psychogeriatr. 24, 541–556.
24 (2015) Associations between homocysteine, folic acid, vitamin b12 and Alzheimer’s disease: insights from meta-analyses. J Alzheimers Dis. 46, 777–790.
25 (2015) Folate, vitamin B-6, and vitamin B-12 intake and mild cognitive impairment and probable dementia in the Women’s Health Initiative Memory Study. J Acad Nutr Diet. 115, 231–241.
26 (2011) Homocysteine in Alzheimer’s disease: role of dietary folate, vitamin B6 and B12. Int J Geriatr Psychiatry. 26, 876–877.
27 (2015) Folic acid deficiency enhances abeta accumulation in APP/PS1 mice brain and decreases amyloid-associated miRNAs expression. J Nutr Biochem. 26, 1502–1508.
28 (2011) Folic acid fortification: why not vitamin B12 also? Biofactors. 37, 269–271.
29 (2016) Branched-chain amino acid catabolism fuels adipocyte differentiation and lipogenesis. Nat Chem Biol. 12, 15–21.
30 (2015) Dietary intakes of folic acid and methionine in early childhood are associated with body composition at school age. J Nutr. 145, 2123–2129.
31 (2006) Folate deficiency results in alteration in intestinal brush border membrane composition and enzyme activities in weanling rats. Journal of Nutritional Science and Vitaminology. 52, 163–167.
32 (2011) Association between bone mineral density and the risk of Alzheimer’s disease. J Alzheimers Dis. 24, 101–108.
33 (2012) Alzheimer’s disease and risk of hip fracture: a meta-analysis study. Sci World J. 2012, 872173.
34 (2012) Bone mineral density is not associated with homocysteine level, folate and vitamin B12 status. Arch Gynecol Obstet. 285, 991–1000.
35 (2012) Relation of plasma total homocysteine, folate and vitamin B12 levels to bone mineral density in Moroccan healthy postmenopausal women. Rheumatol Int. 32, 123–128.
36 (2016) P21-activated kinase 2 (PAK2) regulates glucose uptake and insulin sensitivity in neuronal cells. Mol Cell Endocrinol. 429, 50–61.
37 (2017) Amyloid-beta and islet amyloid pathologies link Alzheimer’s disease and type 2 diabetes in a transgenic model. Faseb j. 31, 5409–5418.
38 (2018) Mind-altering with the gut: modulation of the gut-brain axis with probiotics. J Microbiol. 56, 172–182.
39 (2014) Role of gut microbiota in obesity, type 2 diabetes and Alzheimer’s disease. CNS Neurol Disord Drug Targets. 13, 305–311.
40 (2017) The gut microbiota and Alzheimer’s disease. J Alzheimers Dis. 58, 1–15.
41 (2017) Age drives distortion of brain metabolic, vascular and cognitive functions, and the gut microbiome. Front Aging Neurosci. 9, 298.
