|Home | About | Journals | Submit | Contact Us | Français|
In vitro hippocampal studies by Gay et al. (2008) demonstrated that a myristoylated alanin-rich C kinase substrate (MARCKS) peptide comprising the phosophorylation site or effector domain of the protein acts as a powerful inhibitor of α7 nicotinic acetylcholine receptors (nAChRs), which are known to be critically involved in memory function. However, behavioral consequences of hippocampal MARCKS peptide infusions have not been investigated. The purpose of the current study was to determine if local infusions in the rat ventral hippocampus of long (comprising amino acids 151 to 175) and short (amino acids 159 to 165) forms of MARCKS peptides could affect memory performance in the 16-arm radial maze. Our results demonstrated a dramatic impairment of both working (changing) and reference (constant) memory with MARCKS151-175 only. The shorter MARCKS peptide did not affect memory performance. This is in line with in vitro results reported by Gay et al. (2008) that long, but not short, MARCKS peptides inhibit α7 nAChRs. We also found that the effect of the MARCKS151-175 peptide was dose-dependent, with a robust memory impairment at 10 μg/side, and smaller inconsistent effects at lower doses. Our present behavioral study, together with the earlier in vitro study by Gay et al. (2008), suggest that effector domain MARCKS peptides could play a significant role in memory regulation and impairment.
A diverse group of small peptides, such as naturally existing snail and snake toxins and brain-derived protein fragments (peptides from β-amyloid, acetylcholinesterase, apoliprotein E and myristoylated alanin-rich C kinase substrate (MARCKS) protein) have been shown to interact with the cholinergic system (Di Angelantonio et al., 2003; Eddins et al., 2009; Gay et al., 2008), particularly with nicotinic acetylcholine receptors (nAChRs). These findings are of significant importance because numerous investigations demonstrated a critical involvement of nAChRs in cognitive functions, such as learning, attention and memory (Levin et al., 2006). Moreover, apoliprotein E4 has been identified as a major risk factor for developing Alzheimer's disease (AD) (Corder et al., 1993; Saunders et al., 1993). AD is characterized by a decrease in the number of hippocampal cholinergic neurons and hippocampal expression of nAChRs. Recent in vitro studies have shown that peptides derived from apolipoprotein E block hippocampal nAChRs and receptors expressed in Xenopus oocytes (Gay et al., 2006; Klein and Yakel, 2004), and that hippocampal infusions of this peptide cause long-lasting cognitive impairment in rats (Eddins et al., 2009).
The brain-derived peptides could represent a novel tool for investigating mechanisms of regulation and impairments of cognitive functions. Recent studies demonstrated that heterozygous MARCKS knockout mice show a decrease in long-term potentiation, but not in short-term potentiation in the mossy fiber-CA3 pathway (Hussain et al., 2006), and that both MARCKS heterozygous mutant mice and transgenic mice overexpressing MARCKS have impairment in spatial learning acquisition (McNamara et al., 2005). We recently found that knockout of either α7 or β2 subunits of nicotinic receptors also caused significant deficits in spatial discrimination in mice (Levin et al., 2009). Further investigations (Gay et al., 2008) discovered that a 25-amino acid peptide derived from the effector domain (ED) of MARCKS acted as a potent α7 nAChR antagonist, and that truncated forms of the MARCKS ED were unable to significantly inhibit α7 nAChRs. This study also suggested that MARCKS peptides act rather as allosteric (and not competitive) antagonists than as an open channel blockers. In addition, they inhibit α4β2 nAChRs, but to a significantly lesser extent. The α7 and α4β2 nAChRs are densely expressed in the hippocampus and cortex, the structures critically involved in cognition. Importantly, they are affected in humans with cognitive impairment, such as seen in AD and schizophrenia. However, effects of acute hippocampal infusions of MARCKS peptides on cognition, particularly memory performance, have not yet been investigated. The purpose of the current study was to investigate whether acute infusions of the MARCKS ED peptide, or its truncated form, into the rat ventral hippocampus would impair memory performance in the 16-arm radial maze task.
Histological examinations showed that the cannulae were all contained within the ventral hippocampus. As shown in Figure 1, all of the points of infusion were located within the ventral aspect of the hippocampus. There was very little damage seen resulting from the infusions, at most 0.5 mm in diameter at the point of infusion on one side in two animals.
As shown in Figure 2, the MARCKS peptide151-175, at the dose of 10 μg/side, caused a highly significant (p<0.0001) working memory impairment relative to both the vehicle control and the equimolar dose of control peptide (MARCKS159-165). The rats acutely treated with this dose of MARCKS151-175 demonstrated difficulties in making decisions of which arm they should enter. Typically, they rapidly moved in the central arena of the maze, passing by the arm entrances, and stopping eventually at one of them, hesitating to enter. The rats heavily marked the entered arms with saliva and urine. Such behavior, which helps the animals to keep track of the entered arms, typically occurred at the beginning of training in the 16-arm radial maze and disappeared on its completion. It was not seen after the treatments with vehicle or control inactive peptide. When the rats entered the baited arms, they ate the food and returned to the central arena to continue the search for further rewards. Thus, the animals' food motivation did not seem to be affected by the treatment. We did not observe any changes in their coordination or motor activity. The increase in working memory errors did not seem to be associated either with reduced food motivation or with disturbances in motor activity. The exposure to the lowest dose of 1 μg/side, but not 3 μg/side, produced a smaller but significant (p<0.05) increase in working memory errors. This effect of the low dose reached significance only in comparison to the effects of vehicle, but not the equimolar dose of MARCKS159-165.
With regard to reference memory, Figure 3 shows a more modest, but significant (p<0.05) effect of the 10 μg/side dose of MARCKS151-175 when compared to both the vehicle control and equimolar dose of inactive MARCKS peptide. Lower MARCKS peptide doses did not significantly change reference memory.
Response latency reflects the speed with which the rat chooses arms. Latency was significantly affected at the 10μg/side MARCKS peptide dose (Figure 4), which caused a substantial increase in response latency compared with both the equimolar dose of inactive MARCKS peptide (p<0.001) and control vehicle (p<0.05). Although working memory was significantly impaired at the lowest dose (1 μg/side, p<0.05 v. control), the response latency was not affected. At the dose of 3 μg/side, working memory impairment did not reach a statistically significant level, while response latency was modestly but significantly decreased (p<0.01 v. control). Thus, the highest studied dose (10 μg/side) of MARCKS151-175 caused a robust and consistent impairment of memory in the 16-arm radial maze, accompanied by significantly increased response latency. The effects were significant when compared with responses produced by both the control vehicle and the inactive MARCKS control peptide. Our observations suggest that difficulties in making decisions of which arm to enter might be behind the increased response latency. The effects of lower doses were smaller in magnitude and inconsistent. MARCKS151-175 at doses lower than 10 μg/side caused a certain increase in working memory errors when compared to control vehicle and did not affect reference memory. Unlike MARCKS151-175, truncated MARCKS peptide159-165 produced no significant effect on any of the studied parameters.
The present study demonstrates for the first time that hippocampal infusions of the MARCKS peptide, consisting of amino acids 151 to 175, significantly impairs memory. This effect in rats tested in the 16-arm radial maze is dose-dependent with robust impairment at the dose of local hippocampal infusions of 10 μg/0.5μl/side (6.4 mM), and smaller, inconsistent impairments at lower doses. The impairment was also reversible: at least 48 h after the local infusion of the high dose of MARCKS151-175, the animals' performance in the maze returned to normal. The MARCKS151-175 peptide, at the dose of 10 μg/site (6.4 mM), impaired both working and reference memories with larger working memory deficits. Although the dramatic impairment of working memory was accompanied by a significant increase in the response latency, the effect does not seem to be associated with the disturbances in the motor activity or food motivation. Our observations did not detect any sedative effect of the drug or slowness in behavior of the treated rats, nor did we observe in the rats reduced interest in the food reward. The animals moved quickly in the central maze arena and although they heavily marked the entered arms with saliva and urine, the rats hesitated at the arm entrances. Typically, animals mark the entered arm at the beginning of the training in the 16-arm radial maze. The markings help them to keep track of the entered and un-entered arms. Trained animals usually quickly collect food in the radial maze, not marking the arms, a behavior which relies on the spatial working memory. This observation implies that working memory impairment produced by MARCKS151-175 is rather a result of disruption of hippocampal mechanisms, which are critically involved in the spatial working memory regulation. In this regard, an in vitro hippocampal investigation by Gay et al. (2008) is of significant importance. The authors demonstrated that MARCKS151-175 acts as a powerful α7 nAChR antagonist and binds directly to the extracellular domain of nAChRs. They also found that a truncated form of the peptide, MARCKS159-165, was unable to inhibit acetylcholine responses of the α7 nAChR. Consistent with this report, our current investigation found no effect of the truncated MARCKS peptide infused in the ventral hippocampus on the rat memory performance in the radial maze. Gay et al. (2008) suggested that truncated, short forms of the MARCKS peptide could act in a competitive manner with the MARCKS ED peptide for binding to the nAChR, and thus could regulate nAChR signaling. They discovered that the MARCKS ED peptide binds to the α7 nAChR at a location distinct from the α-bungarotoxin binding site and acts as an allosteric antagonist rather than an open channel blocker. The authors proposed that MARCKS ED peptide could function as a reversible inhibitor under physiological conditions.
Nicotinic receptors are highly expressed in the brain with α7 nAChRs concentrated in the hippocampus (Kawai et al., 2002; Lendvai and Vizi, 2008; Placzek et al., 2009; Tribollet et al., 2004; Wada et al., 1989), the structure critically involved in memory function. The critical role of α7 nAChRs in working memory has been also well documented in pharmacological studies with receptor antagonists (Bettany and Levin, 2001; Hashimoto et al., 2008; Levin et al., 2002; Nott and Levin, 2006; Pichat et al., 2007; Thomsen et al., 2009) as well as genetic studies with α7 nAChR knockout mice (Levin et al. 2009). Our previous studies (Felix and Levin, 1997; Levin et al., 2002) demonstrated that acute infusions of α7 or α4β2 nicotinic receptor antagonists (methyllycaconitine, MLA or dihydro-β-erythroidine, DHβE, respectively) in the rat ventral hippocampus resulted in significant impairment of working and reference memory, with working memory affected more severely. The first investigation was done in the less challenging 8-arm radial maze. Such setting required higher doses of the nicotinic receptor antagonists (78.7 μg/side of MLA and 106.9 μg/side of DHβE) to affect memory performance. However, these doses also elicited convulsive manifestations. Therefore we repeated investigation in the more challenging 16-arm radial maze, which allowed to use lower sub-convulsant concentrations of the antagonists. We found that MLA at dose of 27 and DHβE at dose 6.75 μg/side caused a significant deficit in working and to a lesser extent in reference memory. To our surprise combined infusions of the two antagonists did not produce an additive effect with respect to memory impairment. Infusions of MARCKS151-175 at dose 10 μg/side produced even more powerful memory impairing effect in the similar settings (the effect was more consistent and higher in magnitude). Although MARCKS151-175 could antagonize both α7 and α4β2 nicotinic receptors, we suggest that MARCKS151-175 exerts its action through additional mechanisms different from inhibition of nicotinic receptors, given the previous work showing that combined α7 and α4β2 nicotinic receptor blockade did not cause additive effects.
MARCKS is a primary PKC substrate that binds plasma membrane via n-terminus myristoylation and electrostatic interaction with the phosphorylation site domain, binds calmodulin, binds and cross-links filamentous actin, all in PKC phosphorylation-reversible manner (Arbuzova et al., 2002; Blackshear, 1993). Based on these properties, MARCKS has been proposed to be an important intermediate transducer of receptor generated PKC activity into lasting alterations in membrane-actin cytoskeletal morphology (McNamara and Lenox, 2004). In their recent work McNamara et al. (McNamara et al., 2005) demonstrated that MARCKS overexpression was detrimental for cognitive function such as spatial learning. They suggested that elevation in the MARCKS:PKC ratio could produce a negative effect on specific aspects hippocampal function.
In neurons MARCKS is located to dendritic spines and to axonal terminals in association with synaptic vesicles (Ouimet et al., 1990), and co-localizes with synaptophysin (Lu et al., 1998). Notably, MARCKS binds calmodulin at low calcium concentrations and may sequester free calmodulin at the cell membrane (Verghese et al., 1994). The principal role of calmodulin is to bind calcium. MARCKS, thus, could impair calcium-signaling processes associated with neurotransmitter release and memory function. In addition, an overexpression of nonphosphorylated MARCKS in Schaffer collateral terminals, which normally express MARCKS at low levels in the mature brain (McNamara and Lenox, 1997) was suggested to enhance cross-linking of F-actin at the plasma membrane and thus diminish F-actin cytoskeletol plasticity necessary for maintenance of hippocampal long-term potentiation (McNamara et al., 2005). Thus, memory impairing effect caused by acute infusions of MARCKS151-175 peptide in the rat ventral hippocampus could be a combined result of α7 nAChR inhibition and alteration of different cellular signaling mechanisms in the hippocampus.
The current study provides further evidence at behavioral level that effector domain MARCKS peptides could be involved in the regulation and impairment of cognitive functions. Future studies should explore the issue of a possible involvement of MARCKS peptides in human diseases of cognitive impairment.
Adult female Sprague-Dawley rats (n=10) were housed in plastic cages with corncob shavings. The rats lived in the vivarium immediately adjacent to the behavioral test facility, and were maintained on a reverse 12:12 light-dark cycle with testing during the behaviorally active, dark phase. Before implantation of brain infusion cannulae, animals were housed in groups of three; after the implantation they were singly housed. All rats had ad libitum access to water with one daily meal administered after the behavioral testing. The animals were kept on a restricted diet, which maintained their weight at approximately 85% of ad libitum levels, adjusted for growth. The experimental protocol was approved by Duke University Institutional Review Committee for the use of animals.
The maze was located within a room with a variety of environmental cues: wall pictures, an experimental table, and the experimenter who always sat in the same place. The maze was made of painted black wood, and consisted of a central platform 50 cm in diameter and 16 arms (10 × 60 cm). Each arm was flanked by transparent Plexiglas walls along the proximal part of the arm to insure that the rats went back into the center platform between arm choices. Twelve out of sixteen arms were baited with reward (1/2 piece of Kellogg's Froot Loops®) placed in food cups located in the end of each arm. Each rat had a unique pattern of baited and unbaited arms that remained constant throughout the training and testing periods. The rats underwent at least 18 training sessions, which brought them to a stable level of performance so that working memory function could be assessed. Each session began with placing a rat in the central platform in an opaque plastic cylinder. After 10 seconds the cylinder was removed allowing the rat to navigate the maze. Each arm entered was recorded. Working memory, which is memory with changing contents was differentiated from reference memory, which is memory with constant contents. Working memory errors were defined as a number of times the rat re-entered the initially baited arms. Reference memory errors were recorded each time the rat entered an initially unbaited arm. The session was completed either when the rat had entered all 12 baited arms, or when 10 min. had expired. Latency, or the average number of seconds per arm entry, was computed upon completion of the session.
Following at least 18 radial-arm maze training sessions the rats were implanted with cannulae (22-gauge guide cannulae) into ventral hippocampus, bilaterally. The target coordinates from bregma were A/P -3.2, M/L 5.2, and D/V -7.0 based on the rat brain atlas of Pelligrino (Pellegrino, 1979). The rats were anaesthetized with an IP injection of 65 mg/kg of ketamine and 15 mg/kg of dormitor (medetomidine HCL), shaved and placed in a stereotaxic frame (David Kopf Instruments, Tuiunga, CA, USA) with the bite bar held 5 mm above the intra-aural line. The scull was exposed and appropriate cannula holes were drilled. Two screws were fixed and tied together with wire in the posterior part of the scull to anchor the cannulae and protective cranioplastic cap. The cannulae were lowered and cranioplastic cement was applied to secure their position. Dummy cannulae were placed in the guide cannulae to prevent infection or obstruction. The rats were allowed to recover for 4-5 days after surgery. After the recovery, three additional sessions in the 16-arm radial maze were conducted to stabilize the rat memory performance.
The two MARCKS proteins used in the current study consisted of amino acids 151 -175 (H - Lys - Lys - Lys - Lys - Lys - Arg - Phe - Ser - Phe - Lys - Lys - Ser - Phe - Lys - Leu - Ser - Gly - Phe - Ser - Phe - Lys - Lys - Asn - Lys - Lys - OH and amino acids 159-165 (H - Phe - Lys - Lys - Ser - Phe - Lys - Leu - OH) and were purchased from AnaSpec, Inc. (San Jose, CA). These peptides were used because recent in vitro investigations by Gay et al. (2008) demonstrated that MARCKS peptide151-175 is a potent nAChR antagonist, while MARCKS peptide159-165 (truncated MARCKS) has no significant effect on nAChR responses. Our pilot studies found no significant effect of MARCKS peptide151-175 infused in the ventral hippocampus at doses lower than 1μg/0.5μl/side. Based on these data, three acute doses (1, 3, 10 μg/0.5μl/side) of MARCKS peptide151-175 were chosen. The drugs were dissolved in aCSF, which served also as a control vehicle. Molar concentrations of MARCKS peptide151-175 were calculated (0.64 mM, 1.92 mM and 6.4 mM in 0.5μl/side, for 1, 3 and 10 μg/0.5μl/side) and used for preparation of equimolar doses of the control MARCKS peptide159-165. The drugs were infused with an infusion pump (Harvard Apparatus Model 22) in a repeated measures counterbalanced design. The peptides were infused at the rate of 0.170 μl/min over a period of 3 minutes, finishing 10 minutes before testing. There were at least 48 hours between treatments; the intervals served to minimize possible carryover effects of the previous infusion.
After completion of the behavioral testing the rats were deeply anesthetized with Nembutal and transcardially perfused with a 0.9% phosphate buffered saline followed by 4% paraformaldehyde solution. The brains were removed and preserved in 4% formaldehyde. Before being sliced on a cryostat microtome the brains were frozen on dry ice. Histological slides were then made and studied under a microscope for placement verification.
The dependent measures were working memory errors, reference memory errors and response latency. The data were assessed by analysis of variance for repeated measures with p<0.05 as a threshold for significance.
Supported in part by the Intramural Research Program of the NIH-NIEHS.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.