The Effect of Choline Alphoscerate on Non spatial memory and Neurogenesis in a Rat Model of Dual Stress

Choline alphoscerate (α-GPC) is a choline-based compound and acetylcholine precursor commonly found in the brain; it has been known to be effective in treating neuronal injury and increasing the levels of acetylcholine (Ach) and brain-derived neurotrophic factor (BDNF) which in turn enhances memory and cognitive function. This study was designed to establish rat models of dual stress using noise and restraint in order to investigate the effect of α-GPC on cognitive function and neurogenesis after dual stress. The rats were randomly divided into four groups as follows: a control group (CG), a control with α-GPC group (CDG), a noise-restraint stress group (NRSG), and a noise-restraint stress with α-GPC group (NRSDG). Two experimental groups were exposed to the double stress stimuli of noise and restraint, which involved 110dB sound pressure level (SPL) white band noise and restraint at the same time for 3 hours/day for 7 days. While the CG and NRSG received saline, the CDG and NRSDG received α-GPC (400mg/kg) orally after stress exposure. The α-GPC–treated group showed increased memory function compared to the dual stress group in the novel object recognition test. In analysis of the hippocampus, the α-GPC–treated group showed greater Choline acetyltransferase (ChAT) and BDNF expression compared to the dual stress group. The α-GPC–treated group showed significantly increased neuroblast expression compared to the dual stress group, which suggests that α-GPC enhances BDNF expression and protects the activity of the immature cells at the dentate gyrus. Our results suggest that α-GPC treatment can protect cognitive function and neurogenesis in a dual stress model.


56
Certain physical or psychological stressors disrupt the homeostasis of animals (1, 2). Noise 57 stress can cause damage to cochlear hair cells, which induces hearing loss, and it also impairs 58 non-auditory systems that caused cognitive dysfunction, sleep disturbance as well as physical 59 changes such as neurotransmitters and immune system. Restraint stress in rodents impairs 60 their feeding behavior and emotions, and it is also known to be the most extreme stressor to 61 occur behavioral, neurochemical and immunological changes in response to various types of 62 stress (3, 4).

63
The brain plays the main role in recognizing the intensity of sound exposure, and is 64 involved in interpreting and responding to potential stressors (5). In the adult brain, the 65 subgranular zone (SGZ) of the hippocampus plays an important role in memory and learning 66 as a major site of neurogenesis (6). The hippocampus is vulnerable to neurotoxic conditions 67 and factors such as stress and depression, which have harmful effects on neurogenesis (7).

68
Studies have been shown chronic restraint stress caused loss of hippocampal cell and 69 reduction of hippocampus neurogenesis (8). 70 Brain-derived neurotrophic factor (BDNF) is a member of the family of nerve growth 71 factors, which are otherwise known as neurotrophins. BDNF is involved in various functions, 72 ranging from food intake to behavior, spatial and non-spatial memory associated with central 73 nervous system structures such as the hippocampus, cerebral cortex, and hypothalamus (9, 74 10). BDNF in the hippocampus is known to induce neuronal development by promoting the 75 normal development, survival and plasticity of neurons, and differentiation of SGZ 76 progenitor cells (11). However, experimental studies showed that stress can decrease BDNF 77 expression in the hippocampus (12). 78 Choline alphoscerate (α-GPC) is a choline-based compound commonly found in the brain 79 and is an important intermediate in the synthesis of both acetylcholine (Ach) and cell 80 membrane phospholipids as a cholinergic precursor (13). Increasing the Ach levels in the 81 hippocampus is known to enhance cholinergic neurotransmission and to increase learning and 82 memory by enhancing BDNF expression (7,(14)(15)(16). Studies have been shown that α-GPC is 83 effective in enhancing cognition in animal models of Alzheimer's disease or dementia (17). 84 However, the effect of α-GPC on the severely stressed brain has, to the best of our knowledge, 85 never been studied before. The purpose of this study was to induce dual stresses in rats using 86 noise and restraint, and to investigate the effect of α-GPC on memory function as well as 87 neurogenesis and neuronal protection in the hippocampus after severe stress.  90 Eight-week-old male Wistar rats (240-320g) were purchased from Orient Bio (Sungnam,91 Korea). The rats were randomly divided into 4 age-matched groups as follows: a control 92 group (CG); a control with α-GPC drug administered group (CDG); a noise and restraint 93 stress group (NRSG); a noise and restraint stress with α-GPC drug administered group 94 (NRSDG).

95
All procedures for animal research were performed in accordance with the Laboratory  103 Two of the experimental groups were exposed to the dual stress stimuli of noise and 104 restraint at the same time. Rats in a pie-shaped wire cage were exposed to 110 dB sound 105 pressure level (SPL) white band noise, and restraint stress was administered by putting the 106 rats in cylindrical plastic films, DecapiCones, with rubber bands fixed at the tails for 3 107 hours/day for 7 days. While the CG and the NRSG received saline, the other 2 groups 108 received α-GPC (400mg/kg) orally immediately after the dual stress exposure (Fig 1). 110 All hearing tests were performed with the rats under anesthesia, using a mixture of 111 Rompune (0.4 ml/kg) and Zoletil (0.6 ml/kg). The auditory brainstem response was recorded 112 using an Intelligent Hearing System (IHS) Smart EP fitted with high-frequency transducers 113 (HFT9911-20-0035) and running IHS high-frequency software version 2.33 (IHS, Miami, 114 FL). The details of the hearing test were described in our previous study (18). ABR 115 thresholds and DPOAE levels were compared among the 4 groups in this study. The dissected cochleae were perfused through the round and oval windows with 0.1 M 119 phosphate buffer containing 2 % paraformaldehyde and 2 % glutaraldehyde, and incubated in 120 the same fixatives overnight at 4 °C. They were then rinsed with 0.1 M PBS, perfused in 1 % 121 osmium tetroxide, and incubated overnight, followed by immersion in 0.5 M EDTA, which 122 was changed every day for 2 weeks. The decalcified cochleae were dehydrated in 50, 70, 90, and normal cytoarchitecture of the OC with intact hair cells 5). The averaged regional scores 137 for the OC in the base, middle, and apical turns were compared between the 4 groups.  139 Non-spatial memory was assessed with the novel object recognition (NOR) test, based on 140 the experimental protocol described earlier with a slight modification (2). Simply, a 60 × 60 141 cm, grey-colored, polyvinyl chloride plastic box was used. The whole test period consisted of 142 4 sessions: handling, habituation, adaptation, and testing. After each rat was handled for 20 143 min, it was placed in the apparatus to freely explore the environment (15 min/day, 3 days).

144
On the next day, each rat was placed in the apparatus facing the wall opposite to the segment 145 in which 2 identical sample objects were placed. The rat was immediately put back into its 146 home cage after freely exploring these objects for 10 min. After 1h, the test was performed.

147
One of the sample objects was replaced by a novel object, and the rat was placed in the same 148 way. Exploration and recognition of the objects were defined as the nose being in contact 149 with an object or directed to the object within a defined distance (<2 cm). Sitting or leaning    The ABR threshold and the DPOAE levels were measured to confirm that they were 217 properly exposed to noise stress. After dual stress exposure, the NRSG and NRSDG showed 218 hearing loss, with mean ABR thresholds of 56.25 ± 4.41, 61.25 ± 5.77, 71.25 ± 11.66, and 219 81.25 ± 3.33 dB sound pressure levels (SPL) and of 63.75 ± 2.88, 63.75 ± 5.00, 80 ± 10. 40,220 and 83.75 ± 6.00 dB SPL for click, 8kHz, 16kHz, and 32kHz respectively, which were 221 significantly higher hearing threshold levels in the stress group than in the control group (p< 222 0.001 for 8, 16, click kHz; p< 0.05 for 32 kHz) (Fig 2A). The DPOAE levels were also 223 significantly different between the experimental group and the control group at the 6 to 14 224 kHz geometric mean (GM) frequencies (kHz) (Fig 2B). At the higher GM frequencies,

225
DPOAE responses were small in all groups and there was significance at only 26 kHz GM.  227 The grades for OC degeneration in the base, middle and apex in the control group were 228 4.61 ± 0.05, 4.61 ± 0.03, and 4.71 ± 0.04, respectively. Those in the CDG were 4.55 ± 0.04, 229 4.48 ± 0.04, and 4.53 ± 0.07, respectively. Those in the NRSG were 3.46 ± 0.17, 3.54 ± 0.17, 230 and 3.41± 0.22, respectively. Those in the NRSDG were 3.23 ± 0.13, 3.55 ± 0.10, and 3.55 ± 231 0.17, respectively (Figs 3A and 3B). The grading scores for the organ of Corti for NRSG and 232 NRSDG were significantly decreased compared to the CG at all cochlear turns (p< 0.01). To investigate the effect of dual stress on feeding behavior, we measured the body weight 236 of each rat on the next day after 7 days of dual stress. The CG, CDG, NRSG, and NRSDG 237 mean body weights were 270 ± 2.66 g, 272.5 ± 6.84 g, 230 ± 4.61 g, and 240.3 ± 5.74 g, 238 respectively. The dual-stressed groups showed a greater than 10 % reduction in body weight 239 compared to the CG and CDG (p< 0.05) (Fig 4A).  241 Corticosterone level has been used as a representative stress marker. The plasma 242 corticosterone concentration was measured by ELISA in this study. The levels of 243 corticosterone in the CG, CDG, NRSG, and NRSDG groups were 540 ± 66.84 ng/ml, 526 ± 244 83.32 ng/ml, 980 ± 141.73 ng/ml, and 915 ± 57.47 ng/ml, respectively. The plasma 245 corticosterone level in the dual stressed NRSG group was significantly higher than those of 246 the CG and CDG (p< 0.05). However, there were no significant differences in corticosterone 247 levels among the NRSDG and the CG and CDG (p< 0.05) (Fig 4B).   250 H&E staining was performed to confirm the number of neuronal cells in the hippocampus 251 (Fig 5A). The number of neuronal cells in the CG, CDG, NRSG, and NRSDG were 1661 ± 252 21. 8, 1712.33 ± 88.3, 1343.66 ± 20.3, and 1712.33 ± 68.693, respectively. The total neuronal 253 cell number in the DG significantly decreased in the NRSG compared to all the other groups 254 (p< 0.05 for NRSG vs CG, NRSDG). However, the NRSDG showed a significant increase in 255 total neuronal cell numbers compared to the CG and the CDG (p< 0.05) (Fig 5B).  257 Staining for IL-1β was performed to confirm neuro-immune response in the hippocampus 258 (Fig 5C). The number of IL-1β positive cells in the CG, CDG, NRSG, and NRSDG were 259 12.86 ± 0.13, 12.73 ± 1.63, 18.86 ± 0.85 and 11.66 ± 0.85, respectively. The positive cell 260 numbers in the DG significantly increased in the NRSG compared to all the other groups (p< 261 0.001) (Fig 5D).  263 All throughout the adaptation phase, all groups of the rats consumed equal time exploring 264 identical items (left and right). They showed no significant differences in time spent 265 exploring both in the adaptation phase (Fig 6A). However, they spent significantly different 266 amounts of time exploring a novel object except for the NRSG group in the test phase (t-test, 267 p< 0.05) (Fig 6B). Based on test phase exploration time, the discrimination index (DI) was 268 calculated. The DI of the CG, CDG, NRSG, and NRSDG were 47.35 ± 4.07, 49 ± 6.09, 17.06 269 ± 7.22, and 38.85 ± 3.68, respectively. The NRSG showed a significantly decreased DI 270 compared to the other groups, and the NRSDG showed an increased DI compared to the 271 NRSG (p< 0.05) (Fig 6C).  (Fig 7B).  281 Immunofluorescence was performed to evaluate the effect of α-GPC on BDNF expression 282 in the hippocampus, especially in the dentate gyrus (DG) area (Fig 8A). The intensity levels 283 of BDNF expression in CG, CDG, NRSG, and NRSDG were 5222.93 ± 752, 3966.45 ± 783, 284 2090.859 ± 624, and 4941.28 ± 463, respectively. BDNF expression in the DG was 285 significantly lower in the NRSG than in the CG (p< 0.05), and the NRSDG showed a 286 significant increase of BDNF in the DG compared to the NRSG (p< 0.05) (Fig 8B).  288 Immunofluorescence measurements in neuroblasts were performed to confirm the increase of 289 immature neurons, using the DCX marker in DG at SGZ (Fig 9A). The number of DCX-290 positive cells in the CG, CDG, NRSG, and NRSDG were 126.04 ± 10.7, 107.30 ± 9.5, 85.65 291 ± 6.3, and 122.73 ± 8.9, respectively, which indicates that neurogenesis in the SGZ 292 significantly decreased in the NRSG compared to the CG and the NRSDG (p< 0.05 for 293 NRSG vs CG and NRSDG). Interestingly, the NRSDG showed a significant increase in 294 DCX-positive cells compared to the CG (p< 0.05) (Fig 9B).

296
Here, we report that dual stress induced severe deficits in the behavioral performance and 297 cholinergic activity along with neuronal degeneration in the hippocampus. While, treatment 298 of α-GPC significantly recovered non-spatial memory impairments and it had protect to 299 damage expression of neurotransmitters in the brain caused by dual stress in male rats.

300
However, studies have shown that female rats are more resistant to the stress stimuli than 301 males, demonstrating decreased behavioral performance that males showed cognitive 302 dysfunction (19). Moreover, the estrogen that hormone affected neurogenesis to female (20).

303
Based on these findings, we performed experiment using male rats that would be more 304 suitable for induced model.

327
Stressor including restraint stress is often used to confirm molecular and behavioral effects 328 in experimental studies and has been helped to understand the stress-related brain pathology 329 and the changes in cognition and severe neuronal cell loss (23,25,26). Other experimental 330 studies showed that neuronal histological changes and damages in the hippocampus as well 331 as behavioral alterations after noise or restraint stress (2). In this study, we also observed that   346 The NOR test is a well-known assessment of non-spatial memory function in rats (28).

347
Several studies have been shown that hippocampus is essential brain for NOR test that 348 involved in memory processing and retrieval of object memory by interacting with perirhinal 349 cortex (29, 30). Moreover, in the rats with hippocampal lesions, impaired novelty 350 performance in NOR test with longer interval has been reported (31). Experimental study also 351 showed the increased firing rates and glutamate efflux in the hippocampal pyramidal neurons 352 which means activation of hippocampal function during NOR test (30). However, other study 353 demonstrated that rodents with stress stimuli decreased the non-spatial memory in NOR test

365
In this study, we hypothesized that α-GPC maintains acetylcholine release in the 366 hippocampus in stressed rats and may protect cognitive function after the severe stress. In 367 order to demonstrate our hypothesis, we performed NOR test to investigate the changes in 368 cognitive function after stress in rats and to evaluate the effect of α-GPC in this study. Our 369 study results showed that the DI in the NRSG significantly decreased after the stress, which 370 indicates that cognitive dysfunction occurred after dual stress. While, the NRSDG showed 371 increases in DI and time spent searching for new objects, which indicates increased memory 372 function in dual stressed rats after α-GPC administration.  374 Decrease of cholinergic neurotransmitter from basal forebrain to hippocampus is thought 375 to be one of the factors involved in determining memory impairment in AD disease and 376 normal aging (37). ChAT is a most suitable marker that evaluate the activity of cholinergic neurons, which is biosynthetic enzyme of acetylcholine and it has been known to regulate 378 Ach synthesis that affected Ach levels (38). Other study showed that decreased the ChAT 379 activity in hippocampus after repeated restraint stress and impaired the cholinergic function 380 in patients with memory-impaired dementia (32,39). In this study, we demonstrated that the 381 increase of ChAT expression in NRSDG, suggesting effects of α-GPC on cholinergic 382 neurotransmission by increasing hippocampal ChAT activity and it seems to affect Ach level.

383
Moreover, the neurogenesis caused by cholinesterase inhibitor and Ach is possibly due to 384 activation of muscarinic and nicotinic receptor, and inhibition of inflammation, thereby 385 increasing BDNF production (16).

386
Based on our behavioral and neurotransmitter results, we moved on to evaluate BDNF 387 expression and neurogenesis in the hippocampus of the rats after dual stress.

Alpha-GPC increased BDNF expression, which may be
389 essential for memory enhancement 390 BDNF is a well-known neurotrophic factor that binds to the TrkB receptor at the surface of 391 neuronal cells, and it has been shown to interact with the signaling pathway (40). In 392 experimental studies, knockout rodents for any neurotrophins or their receptors developed 393 severe neuronal diseases. Moreover, the BDNF signaling pathway has been shown to play a 394 critical role in neuronal differentiation, survival, plasticity, and cognition (12,(41)(42)(43). In this 395 study, our results showed that the expression of BDNF in the hippocampus of the NRSDG 396 significantly increased compared to that of the NRSG, which suggests that α-GPC 397 significantly increases BDNF expression in the hippocampus after severe stress in rats, 398 suggesting its contribution to neurogenesis in the hippocampus after severe stress exposure.  400 In the adult brain, neurogenesis occurs in the SGZ of the hippocampus at DG, and this 401 neuronal development has been known to play an important role in the learning and memory 402 functions of the hippocampus (6). It has been reported that immature neurons could be used 403 to detect or process novel stimuli and new neurons at adulthood may be related to temporary 404 storage of information in hippocampal function (20). However, studies have shown that 405 rodents exposed to various types of stress inhibited development of neuronal progenitor cells, 406 resulting in decreased neuroblast production in the hippocampus (28,44,45). Our study 407 results also showed that the number of DCX-positive cells, a marker for neuroblasts, was 408 significantly increased in NRSDG compared to the NRSG. Our study also supported α-GPC 409 increased new neuronal memory function in NRSDG and these results support our hypothesis 410 that α-GPC activates or promotes neurogenesis in the stressed hippocampus.

412
In this study, we showed that α-GPC increased ChAT and BDNF expression in the 413 hippocampus of rats after dual stress, resulting in increased neurogenesis and cognitive 414 function. Therefore, the mechanism by which α-GPC improves cognitive function after 415 severe stress might involve neurogenesis and inhibit neuronal inflammation in the 416 hippocampus. Our basic study results support a clinical role of α-GPC for patients with 417 memory impairment after severe stress, although further studies will be needed to test its role 418 in this area.

419
Author disclosure statements 420 The authors state no conflict of interest.      Figure 6. The novel object recognition (NOR) test for cognitive function, measuring the time 577 spent in object exploration during the adaptation and test phases. There were no significant 578 differences among the groups in the adaptation phase, but the rats in the control groups and 579 the NRSDG explored novel objects more frequently than familiar ones in the test phase. The 580 rats in the NRSG did not show a difference in exploration time for novel objects, which 581 indicates decreased cognitive function after dual stress (Student's t-test) (A&B). The 582 discrimination index (DI) results demonstrated that the rats exposed to dual stress failed to 583 distinguish familiar objects from novel ones. The NRSG showed significantly decreased DI