Diurnal variation in declarative memory and the involvement of SCOP in cognitive functions in nonhuman primates

Before exploring diurnal variations in recognition memory by a color-taste association task utilizing water-drinking behavior in Japanese macaques, we monitored their daily water consumption from a regular water supply system in a 12-h light/dark cycle (Fig. 1). Two randomly-selected monkeys were fed with a diet three times a day: dawn (ZT [zeitgeber time] 1.5), midday (ZT5.5), and dusk (ZT10.0), where ZT0 is defined as light-on and ZT12 as light-off. We identified that these monkeys drank water mainly when fed. Therefore, we decided that the animals should be fed at the time of the task to let them motivate to drink water during the task.

We explored diurnal variations in hippocampus-dependent recognition memory by examining monkey’s performance in the color-taste association task (Fig. 2a–c). Six monkeys reared in individual cages were subjected to the task at randomly-assigned either three-time points of the day: dawn (ZT 1.5), midday (ZT5.5), and dusk (ZT10). (Fig. 2a, b). The task was performed for five consecutive days consisting of three parts: 3-day “practice,” 1-day “training,” and 1-day “testing” (Fig. 2b). In the “practice,” two bottles of bitter water and normal water equipped with nozzles in different colors were presented to the monkeys to learn that colors and water tastes are associated. The animals were allowed to freely drink water from the bottles for 2 h with their position exchange 1 h after the bottle setting. The bottle position exchange eliminates an association between water taste and bottle position, left or right. “Training” and “testing” were performed by using a set of water bottles with nozzles having a color set different from that used in the “practice”. An association between specific nozzle color and water taste formed during the "training” was evaluated in the” testing”, during which the two bottles of the same color set as was used in the “training” were both filled with normal water. Then, the monkeys were allowed to freely drink water from the bottles for 30 min, during which the position of the bottles was exchanged 15 min after the bottle setting. In the “testing”, we monitored which color of a nozzle the animals had first selected 24 h after “training”, by which they had experienced an association between the bitter taste and the specific nozzle color. The memory of the association between the bitterness and the nozzle color during the “training” would let the animals drink water from the color nozzle that was not bitter in the “testing.” When the animals first chose the color nozzle with normal water both before and after the bottle exchange, then 1-point was given as it was judged that they remembered the association. This scoring method should exclude the case of selecting the right color nozzle by chance. The tasks were carried out 1–4 times for each time point for each monkey. The resulting points of the tasks (a total of 16 tasks for ZT1.5, 13 tasks for ZT5.5, and 16 tasks for ZT10.0 at each time point in six monkeys) were combined. The color-taste association task revealed an apparent diurnal change with the highest level at ZT5.5 midday among the three-time points during the light period (Fig. 2c). The accuracy rate at midday was significantly higher than the 0.25 chance level (dotted line). Since the color-taste association task determined midday as the best time of day for memory reinforcement, we then approached the molecular mechanism to augment declarative memory performance.

SCOP in the hippocampus is a critical molecule for diurnal regulation and memory reinforcement of recognition memory in nocturnal mice [6]. Of interest was to ask whether SCOP in the hippocampus might play a similar role in mice and diurnal mammalian species. To investigate the role of SCOP in the hippocampus in monkeys, we performed knockdown experiments of SCOP by using a lentiviral vector. The lentiviral vector was designed to express both Scop shRNA (short-hairpin RNA) driven by Japanese macaque H1 promoter (mkH1) and enhanced green fluorescent protein (GFP) driven by CMV promoter (Fig. 3a). The COS-7 monkey cell line was used to evaluate the knockdown efficiency of SCOP by the Scop shRNA lentiviral vector. Infection of the vector efficiently reduced the endogenous SCOP protein level as compared to COS-7 cells treated by the control Scramble shRNA lentiviral vector (Fig. 3b).

We randomly selected two monkeys, one for knockdown (Scop shRNA) and another for its control (Scramble shRNA). The shRNA lentiviral vectors were injected into the hippocampus. Fourteen injections were made into the bilateral CA1 regions (seven injections per unilateral CA1) of each monkey to cover the whole CA1 area positioned by an MRI-guided navigation system. After the behavioral analysis, the injection sites of the lentiviral vectors in the hippocampal CA1 region were confirmed by GFP immunostaining (Fig. 4a), indicating that the vectors were transfected at the targeted CA1 region. The efficiency of Scop knockdown was verified by quantitation of Scop mRNA in the hippocampus using RT-qPCR, and Scop mRNA in the hippocampus was reduced significantly to 51% in comparison with the control case delivered with Scramble shRNA (Fig. 4b).

After the lentiviral vector injections, the animals were allowed to recover for eight weeks, and then the color-taste association task was performed at ZT5.5 (midday), which was the highest performance time of day in the non-injected animals (Fig. 2c). Six trials of the color-taste association task were conducted over time for each animal. The Scop knockdown suppressed the performance of the color-taste association task, contrasting with the control animal that had been injected with the lentiviral vector expressing Scramble shRNA (Fig. 4c). The Scop-knocked-down monkey did not earn even 1-point over six trials (Fig. 4c Scop shRNA), whereas the control monkey correctly chose the colored nozzle before and after three over six trials (Fig. 4c Scramble shRNA). These results suggest that SCOP in the hippocampal CA1 is indispensable for the color-taste association memory in macaques, although our results from the SCOP knocked-down experiments are relatively premature, because of the small sample size. Further studies with a more sample size will confirm the conclusion.

Water consumption was measured by a drinkometer (O’Hara & Co., Tokyo, Japan) placed into a regular water supply system route when not conducting the memory task. The drinkometer records the number of drops consumed per 5 min.

The color-taste association task was conducted with feeding/drinking limitations. In order to let the macaques more motivate the drinking behavior during the behavioral task, the regular water supply was stopped for 3 h before starting the behavioral paradigm for midday and dusk. For tests in the dawn, the regular water supply was stopped at a light-off time (ZT12) the day before the behavioral experiment. The behavioral tests were performed using a two-bottle system approved by the Animal Welfare and Animal Care Committee of the Primate Research Institute, Kyoto University (no. 2011-093). Briefly, two filled bottles (500 mL) were set in front of the monkeys: one contained bitter water containing 20 mM salicin (Sigma-Aldrich, MO, USA), and the other contained normal water in the “practice” and the “training” (see Fig. 2b). For the discrimination of the two bottles, stainless steel nozzles of the bottles were treated with oxidized coloration and stabilized in six different colorings, i.e., magenta, black, blue, brown, gray, and white (processed by Nakano Kagaku, Niigata, Japan).

The task was performed in five consecutive days consisting of three parts: three days of “practice,” 1-day “training,” and 1-day “testing.” In the “practice,” two bottles of bitter water and normal water equipped with nozzles in different colors were presented simultaneously to the macaques (see Fig. 2a). The three days of the “practice” let animals learn that nozzle color is associated with water taste. Monkeys were allowed to freely drink water from the bottles for 2 h with the location exchange at 1-h after the bottle setting. In the “training,” bitter water and normal water were presented to the same animals by bottles with nozzle color sets different from those used in the “practice.” Again, monkeys were allowed to freely drink water from the bottles for 2 h with the location exchange at 1-h after the bottle setting. An association between specific nozzle color and water taste formed during the "training” was evaluated in the” testing,” during which the two bottles of the same color set as was used in the “training” were both filled with normal water and presented at the same time. Then, animals were allowed to freely drink water from the bottles for 30 min, during which the location of the bottles was exchanged 15 min after the bottle setting. In this “testing” process, memory performance was assessed as the degree of correspondence between the nozzle color and water taste the animals had experienced in the “training.” All the “testing” process was recorded by a video camera. When the monkey in the “testing” first chose the nozzle color both before and after the bottle exchange that the animal had experienced normal water in it, then 1-point was given as it was judged that the animal remembered the association between the color and taste. The task (practice, training, and testing) was carried out a maximum of four times at each time point at ZT1.5, ZT5.5, or ZT10.0 (ZT10 means 10 h after the light ON) for each monkey. The total tasks at each time point in six monkeys were 16 for ZT1.5, 13 for ZT5.5, and 16 for ZT10.0. A different color set was used for each task trial at a single time points for each animal. The number of points divided by the trials was used as the accuracy rate. The inter-task interval was at least 1 month.

The plasmids, pENTR4-H1, CS-RfA-CG, pCMV-VSV-G-RSV-Rev and pCAG-HIVgp were provided by Dr. Hiroyuki Miyoshi, RIKEN Bioresource Center, Tsukuba, Japan. shRNA targeting Scop was designed using siDirect (http://​design.​RNAi.​jp/​), and the target sequence (GGATA TTGGC CATAA TCAAA CGTGT GCTGT CCGTT TGATT ATGGC CAATA TCCA) was used for the down-regulation of macaque Scop. A control shRNA with a scrambled sequence (GATAT GGCAC TGATA ATCAA CGTGT GCTGT CCGTT GATTA TCAGT GCCAT ATCA) was designed. The pairs of the complementary oligonucleotides containing these sequences were synthesized (SIGMA), annealed, and cloned into the modified pENTR4-H1, in which human H1promoter was replaced by Japanese macaque H1 promoter (mkH1; see Fig. 3a). Cloning of the mkH1 promotor was performed from the genome of a Japanese macaque with reference to the homologous region of the rhesus H1 sequence and the human H1 sequence at the UCSC genome browser (https://​genome.​ucsc.​edu). The H1-shRNA fragment from pENTR4 H1-shRNA was then inserted into lentiviral vector CS-RfA-CG by the Gateway system (Invitrogen, CA, USA) to obtain CS-H1-shRNA-CMV-GFP.

For measuring RNA titer, viral RNA in 50 nL of the vector stock solution was isolated with a NucleoSpin RNA virus kit (Takara, Shiga, Japan), and the copy number of the RNA genome was determined by quantitative PCR using Taq-Man technology (Thermo Fisher Scientific, Waltham, MA, USA). The viral biological titers were also determined by infection of COS-7 cells with a dilution series and counting colonies of GFP-positive cells.

To determine the efficiency of Scop shRNA expressing lentivirus, we chose monkey cell line COS-7. The COS-7 cells were infected with Scop shRNA or scrambled shRNA (Scramble shRNA) expression lentivirus (2 × 10⁴ ifu / 3.8 cm² dish surface) and cultured for three days before western blot analysis for SCOP protein expression. For western blot analysis, proteins separated by SDS–PAGE were transferred to a polyvinylidene difluoride membrane (Millipore, USA). The blot was blocked in a blocking solution of 3% bovine serum albumin in T-TBS (0.05% Tween20, 50 mM Tris–HCl, 140 mM NaCl and 1 mM MgCl₂; pH 7.4), for 2 h at room temperature. Then the blots were incubated for 4 h at room temperature with anti-SCOP antibody (1:2,000, αCB in Ref. 24) diluted in the blocking solution. The signals were visualized by an enhanced chemiluminescence detection system (PerkinElmer, Boston, MA, USA).

Based on the 3R principle (Replacement, Reduction, and Refinement) for animal experiments, lentivirus was administered to each macaque for Scop shRNA (one male) and Scramble shRNA (one female) in the experiment. The animals were first sedated with ketamine hydrochloride (5 mg/kg, i.m.) and xylazine hydrochloride (0.5 mg/kg, i.m.) and then anesthetized with sodium pentobarbital (20 mg/kg, i.v.). The monkeys were kept hydrated during the surgical operation with a lactated Ringer’s solution (i.v.). An antibiotic (Ceftazidime; 25 mg/kg, i.v.) and an analgesic (Meloxicam; 0.2 mg/kg, s.c.) were administered at the first anesthesia. After partial removal of the skull, multiple injections of each vector were performed into the hippocampal CA1 area with the aid of an MRI-guided navigation system (Brainsight Primate, Rogue Research, Montreal, QC, Canada). A total volume of 70 µL of each vector was injected into multiple sites (5 µL/site, seven sites per side, 14 sites per animal) through a 10 µL Hamilton microsyringe. The injection titer of the viral vector was 2 × 10¹⁰ gc/mL. After the injections were completed, the scalp incision was closed. All experiments were performed in a specific laboratory (biosafety level 2) established at the Primate Research Institute, Kyoto University, designed for in vivo animal infectious experiments.

At the end of experiments in macaques, the animals were deeply anesthetized with sodium pentobarbital (50 mg/kg, i.p.) and perfused transcardially with PBS. The brains were cut in the coronal plane at the 3-mm thickness, and the slices containing the hippocampus were immersed in 4% paraformaldehyde in PBS overnight, followed by 30% sucrose in PBS for two days at 4° C and cut into serial coronal sections (20 µm) by a microtome in a cryostat (Leica, Germany). The sections were washed with 0.1% Triton-X100 in PBS for 15 min × 3 times and then blocked in 1% normal goat serum, 1% BSA, 0.1% Triton-100 in PBS for 1 h at room temperature, and incubated in anti-GFP antibody (1:1000, Invitrogen, G10362). The immunoreactivity was visualized with Alexa 488-conjugated goat anti-rabbit IgG (1:1000; Molecular Probes) and then stained with 1 ng/ml Hoechst 33342 (Sigma) to visualize nuclei. The hippocampal sections were imaged on BZ-9000TS Microscope (Keyence, Osaka, Japan).

Total RNA was isolated from six subregions of the macaque hippocampus using TRIzol reagent (Invitrogen) and was subsequently purified by RNeasy Mini kit (Qiagen) according to the manufacturer’s protocol. RT–qPCR analysis was performed using Go Taq 2-step RT–PCR system (Promega) in a Step One Plus (Applied Biosystems). Data are presented as values normalized to the housekeeping gene Gapdh. PCR primers used are;for Scop FW 5′-CCCCA GCTGT TTGGA GTCAT-3′ and RV 5′-TCAAA CACAC CGTAG AGGGC-3′ for Gapdh FW 5′-ACCGT GGTCA TGAGT CCTTC C-3′ and RV 5′-GCACC ACCAA CTGCT TAGCA-3′.