Investigation of Cyclo-Z Therapeutic Effect on Insulin Pathway in Alzheimer's Rat Model: Biochemical and Electrophysiological Parameters

Cyclo (his-pro-CHP) plus zinc (Zn+2) (Cyclo-Z) is the only known chemical that increases the production of insulin-degrading enzyme (IDE) and decreases the number of inactive insulin fragments in cells. The aim of the present study was to systematically characterize the effects of Cyclo-Z on the insulin pathway, memory functions, and brain oscillations in the Alzheimer's disease (AD) rat model. The rat model of AD was established by bilateral injection of Aβ42 oligomer (2,5nmol/10μl) into the lateral ventricles. Cyclo-Z (10mg Zn+2/kg and 0.2mg CHP/kg) gavage treatment started seven days after Aβ injection and lasted for 21 days. At the end of the experimental period, memory tests and electrophysiological recordings were performed, which were followed by the biochemical analysis. Aβ42 oligomers led to a significant increase in fasting blood glucose, serum insulin, Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) and phospho-tau-Ser356 levels. Moreover, Aβ42 oligomers caused a significant decrement in body weight, hippocampal insulin, brain insulin receptor substrate (IRS-Ser612), and glycogen synthase kinase-3 beta (GSK-3β) levels. Also, Aβ42 oligomers resulted in a significant reduction in memory. The Cyclo-Z treatment prevented the observed alterations in the ADZ group except for phospho-tau levels and attenuated the increased Aβ42 oligomer levels in the ADZ group. We also found that the Aβ42 oligomer decreased the left temporal spindle and delta power during ketamine anesthesia. Cyclo-Z treatment reversed the Aβ42 oligomer-related alterations in the left temporal spindle power. Cyclo-Z prevents Aβ oligomer-induced changes in the insulin pathway and amyloid toxicity, and may contribute to the improvement of memory deficits and neural network dynamics in this rat model.


Introduction
Alzheimer's Disease (AD) is a type of progressive dementia due to the accumulation of extracellular Aβ peptides and intracellular neurofibrillary tangles (NFT) [1]. To date, studies on AD have indicated that Aβ accumulation is the major factor in the pathogenesis of the disease. Oligomeric types of Aβ peptide are thought to trigger synaptic dysfunction, neurodegeneration, and ultimately dementia [2][3][4][5]. The cellular pathology leading to neural network dysfunction has not been fully explained yet. However, it has been shown that Aβ exposure induces changes in presynaptic and postsynaptic synapses, while tau accumulations lead to synaptic conduction disorder and disrupt neural synaptic network dynamics [6,7].
In recent studies, it has been emphasized that brain insulin resistance is an early feature of AD and develops before the onset of AD symptoms. Increasing insulin resistance contributes to cognitive impairment both before and after AD-specific molecular changes [8]. It is known that insulin has an important role in the regulation of energy metabolism and neuronal vitality in neurons and acts as a growth factor in the brain. The insulin signaling mechanism in the brain is essential for the formation of synaptic plasticity that plays a central role in cognitive functions [9]. Aβ oligomers (AβOs) are potent synaptotoxins that accumulate and act as synapsespecific ligands [10][11][12]. Insulin and Aβ are both amyloidogenic peptides that share a common sequence recognition motif. Therefore, AβOs can bind to the insulin receptor and inhibit its autophosphorylation [13]. They also trigger the detachment of the insulin receptor from the plasma membrane, resulting in decreased surface insulin receptors and insulin sensitivity [14,15]. When amyloid-β oligomers bind to insulin receptors, they block the downstream pathway of insulin signaling. This results in NFT formation with overproduction of Aβ due to tau hyperphosphorylation and excessive γ-secretase activation [16]. Insulin signal also provides a physiological defense mechanism against synapse loss caused by oligomers [15,17].
When insulin binds to its receptor, it activates the phosphatidylinositol-3 kinase (PI3K) → Protein kinase B (PKB/ Akt) pathway by phosphorylating the adapter proteins such as IRS-1 and 2. Activation of this pathway phosphorylates GSK-3β and inactivates it [9]. The insulin receptor acquires an intrinsic tyrosine (Thr) kinase activity when activated by insulin. Tyrosine kinase activity triggered by insulin binding downregulates the oligomer binding sites in neurons [15,17]. Thus, insulin signaling protects against the long-term potentiation (LTP) disorder caused by oligomers and the accumulation of hyperphosphorylated tau [18]. Moreover, the observed increment of brain insulin resistance causes a decrement in the degradation of Aβ oligomer by reducing IDE and Neprilysin (NEP) levels in AD. This leads to worsening insulin resistance and the production of more toxic AβOs [19][20][21]. Therefore, potentiation of insulin signaling and Aβ oligomer degradation in AD seem to be an important therapeutic target.
Zinc has been shown to have insulin-like effects by facilitating insulin signaling [22]. Also, Zn +2 stimulates both IDE and NEP synthesis, which are known to cleave Aβ peptides [23]. Besides, Zn +2 deficiency is a common condition in AD [24,25]. Thus, Zn +2 dyshomeostasis may play a critical role in the pathogenesis of AD, and Zn +2 chelation is a potential therapeutic approach [26]. Zn +2 deficiency is mostly caused by impaired intestinal Zn +2 absorption and Zn +2 uptake into the cell. The normal elemental Zn +2 absorption mechanism takes place through facilitated diffusion through Zn transporter (ZnT) proteins. These ZnTs contain a large number of histidines similar to histidyl-proline diketopiperazine [cyclo (His-Pro); CHP]. CHP is a cyclic form of the amino acids L-histidine and proline. Histidine-proline-rich glycoproteins play an important role in Zn +2 transport [27]. Since CHP contains an L-histidine molecule, it has the ability to chelate Zn +2 and stimulate intestinal Zn +2 absorption and Zn +2 uptake [28]. This cyclic form helps Zn +2 transport independent of the normal elemental Zn +2 transport system [29] and is necessary for its passage across the blood brain barrier [30][31][32]. Studies have shown that it is not very effective when CHP or Zn +2 has applied alone [1]. When used together as Cyclo-Z, which is a combination of CHP and Zn +2 , they work better as medicine.
It is not entirely known how the altered insulin pathway plays a role in the early stage of AD. At the same time, the molecular pathways associated with the hyperphosphorylation of the tau protein due to the altered insulin signaling have still not been clearly elucidated. In this study, we demonstrated the molecular changes in the insulin signaling pathway observed in the Aβ oligomers-induced early AD rat model. We examined the effects of changes in the insulin signaling pathway on tau phosphorylation, Aβ accumulation, cognitive functions, and brain oscillations. Besides, Cyclo-Z has been suggested as a potential stimulatory agent for insulin signaling. So, we examined how altered insulin signaling due to the Cyclo-Z agent affects AβOs-induced neuropathology, cognitive functions, and brain oscillations.

Animal Preparation
The rats were obtained from Akdeniz University Animal Care Unit. Male albino Wistar rats aged 3 months, weighing 250-300 g, were housed in stainless steel cages in groups of 4 rats per cage and given food and water ad libitum. The body weights of the animals were recorded twice, at the beginning of the surgical procedure and before the sacrifice of the animals. Weight gain was obtained by subtracting the first weight measurements from the last weight measurements of each rat (n=13 per group). Animals were fasted for 12 hours before fasting blood glucose was measured before the electrode implantation. The animals were kept at a constant temperature of 23 ± 1 °C and were subjected to 12 h light-dark cycles. Animals were divided into four groups (n=15 each group): (1) sham-operated (1,25 μl DMSO + 8,75 μl PBS/10μl) intracerebroventricular (i.c.v) injection plus water gavage treatment (SH); (2) sham-operated (1,25 μl DMSO + 8,75 μl PBS/10μl) i.c.v. injection plus Cyclo-Z treatment (10 mg Zn/kg and 0,2 mg SHP/kg gavage) (SHZ); (3) Aβ42 oligomer (2,5 nmol/10 μl) i.c.v. injection plus water gavage treatment (AD); (4) Aβ42 oligomer (2,5 nmol/10 μl) i.c.v. injection plus Cyclo-Z treatment (10 mg Zn/kg and 0,2 mg SHP/kg gavage) (ADZ). Daily gavage treatment started seven days after Aβ injection and it lasted for 21 days. Zinc chloride was purchased from Molar Kimya (MZK.100310.1000). The purity of Zinc chloride was >98%. Cyclo was purchased from Santa Cruz (Santa Cruz Biotechnology, Europe). Cyclo has a molecular formula of C 11 H 14 N 4 O 2 and a molecular weight of 234,26 g/ mol. 50 mg of Cyclo and 2.5 g of Zn +2 were dissolved in 1 L of water. The dosage was chosen according to the results

3
of an earlier study [33]. Aβ1-42 oligomers were prepared in accordance with the literature [34]. Aβ1-42 (Sigma-Aldrich, USA, product no: A9810) peptides were dissolved in DMSO to a concentration of 2 mmol and then diluted 8 times in sterile PBS, vortexed for 30 minutes at room temperature, then centrifuged at 15000×g at 4°C for 1 hour. Supernatants (250 μl) were aliquoted (25 μl) and frozen at -20°C. When heated to 4 °C, Aβ-42 oligomers were used within 24 hours. Since DMSO is a toxic substance, a 10 μl solution prepared with 1.25 μl DMSO and 8.75 μl PBS was given to the sham groups.
The rat model of AD was established as described previously [35,36]. Rats were anesthetized with a combination of ketamine (80 mg/kg, i.p.) and xylazine (5 mg/kg, i.p.) and then placed in a standard stereotaxic apparatus. A middle sagittal incision was made in the scalp and was sterilized using standard procedures. Bilateral holes were drilled in the skull using a dental drill over the lateral ventricles (AP: −0.8 mm, ML:±1.4 mm, DV: −4.0 mm). Rats in the AD and ADZ groups were injected with 2,5 nmol/10 μl Aβ-42 oligomers at a rate of 0.5 μl/min. The syringe was removed 5 min after the injection. The SH and SHZ groups received 10μl solution prepared with 1.25 μl of DMSO and 8.75 μl of PBS. After surgery, the scalp was sutured, and sulfamethoxazole was sprinkled on the wound to prevent infection. In addition, penicillin (40,000 U) was injected intramuscularly into the gluteus once a day for 3 days. The animals were allowed to recover post-operatively for a week. After the rats recovered, gavage and learning experiments were started. At the end of the gavage application, learning experiments were performed, and then electrode implantation was performed for electrophysiological recordings.

Surgery Protocols
Rats were anesthetized with a combination of ketamine (80 mg/kg, i.p.) and xylazine (5 mg/kg, i.p.) and then placed in a standard stereotaxic apparatus. Before the electrodes were placed, for measurement of fasting blood glucose levels, the tail region was bled and measured by glucometer (plusMED Blood GlucoseMeter, Accuro, pM1-300, Bionime Corporation, Taiwan). During the anesthesia, the skull of the rats was placed in the stereotaxic apparatus and drilled for the implantation of electrodes. Stainless steel screw electrodes were implanted bilaterally over the frontal (AP: 4.5 mm, ML: +2 and -2 mm, DV: +1.8 mm), temporal (AP: -8.0 mm, ML: +6.6 and -6.6 mm, DV: 3.3 mm), the reference and ground electrodes were inserted into the cerebellum (AP: -12.72 mm, ML: 2.5 and -2.5 mm, DV: 3.2 mm). After the electrodes were placed, the rats were removed from the stereotaxic device and electrophysiological recordings were taken.

Biochemical and Western Blot Investigations
Immediately after taking the electrophysiological recordings, blood was drawn from the veins of the rats into EDTA tubes under ketamine (80 mg/kg, i.p.)and xylazine (5 mg/ kg, i.p.) anesthesia. Brain tissues were collected for sandwich ELISA (n=6-8 per group) and western blot (n=6-7 per group) analysis. Afterward, the brain tissues were purified from the blood by cardiac perfusion using isotonic with heparin through the cardiac cannula. For biochemical measurements, the excised brain tissues were frozen in liquid nitrogen and stored at -80°C.

Measurement of Insulin Amount in Plasma and Hippocampus Tissue Homogenates
Insulin levels in the supernatants of plasma and hippocampus tissue homogenates were measured using a commercially available sandwich ELISA kit (Bioassay Technology Laboratory, E0707Ra) in accordance with the manufacturer's instructions. These kits contain 96 well plates coated with rat insulin antibodies. The biotinylated rat insulin antibody is then added and binds to the insulin in the sample. An insulin-specific horseradish peroxidase (HRP) conjugate polyclonal antibody is added. Thus, the immobilization of insulin is ensured and a sandwich model is created. Therefore, streptavidin-HRP is then added and binds to the biotinylated insulin antibody. After incubation, unbound Streptavidin-HRP is washed 5 times during the wash step. After incubation and washing, unbound components are removed. The substrate solution is then added. When the substrate solution is added, the HRP reacts with the enzyme, leading to color formation. In this method, color formation is observed only in the wells containing insulin. The color development is in proportion to the amount of insulin in the rat. The reaction is terminated by adding an acidic stop solution. Optical density (OD) is measured spectrophotometrically at 450 nm. The sample amount of insulin is calculated with a standard curve plot. Peripheral insulin resistance was assessed using the HOMA-IR formula (fasting insulin (mIU/l) x fasting blood glucose (mM)/22,5) [37].

Measurement of Aβ1-42 Oligomer Amount in Total Brain and Hippocampus Tissue Homogenates
Aβ1-42 oligomer levels in the supernatants of total brain and hippocampus tissue homogenates were measured using a commercially available sandwich ELISA kit (YL biont-YLA0372RA) in accordance with the manufacturer's instructions. Aβ1-42 was added to wells pre-coated with Aβ1-42 monoclonal antibody and then incubated. After that, biotin-labeled anti-Aβ1-42 antibodies were added to combine with streptavidin-HRP. After incubation and washing, unbound components are removed. Substrate solutions A and B are then added. When the substrate solution is added, the HRP reacts with the enzyme, leading to color formation. In this method, color formation is observed only in wells containing Aβ1-42. Color development occurs in proportion to the amount of rat Aβ1-42. The reaction is terminated by adding an acidic stop solution. The OD is measured spectrophotometrically at 450 nm. The amount of sample Aβ1-42 is calculated by the standard curve plot.

Measuring Protein Levels with the Western Blot Technique
For the preparation of tissue homogenates, frozen brain samples were crushed with liquid nitrogen, and then the crushed brain tissue samples were suspended in lysis buffer (50mM Tris-HCl, pH 7,4, 100 μM EDTA, 100 μM EGTA, 1% NP-40, 0.1% SDS, 0.1% deoxycholic acid) to which protease inhibitor cocktail tablets were added. The protein concentrations of supernatants after centrifugation (10,000g, 10 min, at 4 C) were measured with the BCA assay kit (Pierce) according to the manufacturer's instructions. After the protein content was determined, equal amounts of proteins (20μg/ml) were loaded into each well and separated by 12% SDS polyacrylamide gel electrophoresis at 100V, 30 mA for approximately 1.5 hours. Proteins separated by gel electrophoresis were transferred to the polyvinylidene difluoride membrane with a transfer system (Trans-Blot Turbo BioRad) at 25V 1.3mA for 10 minutes. After this step, the membranes were placed in TTBS (0.1% Tween-20, 10 mM Tris and 150 mM NaCl solution (containing 3% albumin)) and incubated for 1 hour at room temperature. After the incubation phase, primary antibodies were prepared in TTBS solution containing 3% albumin according to the desired protocol and incubated overnight (+4 o C). These primary antibodies are IDE (Ylbiont, YID2780), NEP (SinoGene-Clon Biotech Co., SG-0527-R), p-IRS1 (Ylbiont, YIG0585), p-GSK-3β (SinoGeneClon Biotech Co., SG-5368R), p-Tau (ser356) (ThermoFisher, 44-752G). The membranes were then washed with TTBS solution 3 times for 10 minutes, and then the secondary antibody (Goat anti-rabbit IgG/HRP-Ylbiont, YIF0010) step, which was suitable for the primary antibodies, was started. Finally, bound antibodies were detected with the chemiluminescence-based HPR Substrate System. Membranes were exposed to Hyperfilm. The results were analyzed with the software Image J, and the expression of the target proteins were normalised to the expression of GAPDH.

Behavioral Testing
The novel object recognition memory and object location memory tests were modified by Roozendaal et al. and adapted for rats [38]. The system in which the experiment was carried out consists of a box measuring 40 cm x 40 cm x 40 cm, evenly illuminated, with a white matte base and sides. The experiment was carried out in a soundproofed room to prevent any external stimulus from interfering with the experimental parameters during the experiment. An object location test was performed for each rat 3 days after the new object recognition test.

Novel Object Recognition Test
The novel object recognition test consists of two phases: the training phase and the testing phase. During the training phase, two identical objects (A1-A2) were placed in the experimental system at two opposite positions at the same distance from the nearest corner. The rats were then released from the center to the field and allowed 5 minutes to explore these two identical objects (A1-A2). The rats that completed the training phase were taken to the test phase 24 hours after they were placed back in their cages. To test the novel object recognition memory during the test phase, a copy of the familiar object (A3) and a new object (B1) were placed in the same place as during the training experiment. Each of the objects used in the experiment was fixed to the floor to prevent its movement. The behavior of the rats during the training and testing phases was recorded with a video camera and used to determine the time the rats spent with the objects. To preclude the existence of olfactory cues, the entire box and objects were always thoroughly cleaned with 70% ethanol after each trial. In the novel object recognition test, discrimination index (%) values showing memory strength were analyzed. The discrimination index of rats (n=11 per group) was calculated by multiplying the difference between the time to examine the new and the old object divided by the total time spent examining both objects by 100.

Object Location Test
Object location testing consists of two phases: the training phase and the testing phase. During the object location test training phase, two identical objects (A1-A2) were placed in the experimental system. The rats were then released from the center to the field and allowed 5 minutes to explore these two identical objects (A1-A2). The rats that completed the training phase were taken to the test phase 24 hours after they were placed back in their cages. To test the object location memory in the test phase, a copy of the familiar object (A3) was placed in its place in the training phase, while the other familiar object (A4) was placed in a different location from the training phase. Each of the objects used in the experiment was fixed to the floor to prevent its movement. The behavior of the rats during the training and testing phases was recorded with a video camera and used to determine the time the rats spent with the objects. To preclude the existence of olfactory cues, the entire box and objects were always thoroughly cleaned with 70% ethanol after each trial. The discrimination index (%) values were analyzed in the object location test. The discrimination index (%) of rats (n=10 per group) is an equation obtained by dividing the difference between the time spent on the changed object and the time spent on the object in the old location by the total time and multiplying by 100.

Statistical Analysis
The statistical analyses of the obtained data were performed using SPSS 23.0 (SPSS, Chicago, IL, USA) software for Windows. Statistical comparisons between groups for object recognition and location tests, biochemical experiments, and western blot analysis were performed by using one-way ANOVA and a post hoc Tukey test. A repeated measure ANOVA (rANOVA) was used to analyze electrophysiological data, including between-subject factor groups and withinsubject factor electrode locations (Bonferroni post hoc test). Greenhouse-Geisser corrected p-values are reported. Weight changes were examined via rANOVA, including betweensubject factor groups and within-subject factor days (Tukey post hoc test). Results are expressed as the mean ± standard error. Significance levels were set at P < 0.05. All experimenters were blinded to animal experimental group membership during data collection and analyses.

Changes in Body Weight
The body weights of the animals were recorded twice, at the beginning of the surgical procedure and before the sacrifice of the animals. Body weight values for sham and experimental groups are given in Table 1. There was a significant difference between groups [F(3,48) =11.39, p<0.001]. The body weight gain of the animals was significantly decreased in the AD (20,38 ± 13,26 g) versus the SH (31,46 ± 2,17 g) (p<0.001). The Cyclo-Z treated ADZ (28,15 ± 2,93 g) group had higher body weight gain levels versus the AD group (p<0.001). Even though it was seen that the SHZ group gained less weight than the SH group, Cyclo-Z treatment alone didn't have a big effect on weight gain.

Plasma Fasting Blood Glucose Levels
In Fig. 1A, mean values of plasma fasting blood glucose levels are given. There was a statistically significant difference in plasma fasting blood glucose levels between groups [F(3,20)= 24.88, p<0.001]. The plasma fasting blood glucose levels were significantly increased in the AD (191.33±5.524 mg/dL) versus the SH (143.00±6.033 mg/ dL) (p<0.01). The ADZ (161.00±5.092 mg/dL) group had lower plasma fasting blood glucose levels versus the AD group (p<0.05). The plasma fasting blood glucose levels were significantly increased in the SHZ (226.00±10.957 mg/ dL) versus the SH (p<0.001).

Serum Insulin Levels
Mean values of serum insulin levels are given in Fig. 1B. There was a statistically significant difference in serum insulin levels between groups [F(3,28)= 7.84, p<0.01]. The serum insulin levels were significantly increased in the AD (2.71±0.074 mIU/L) versus the SH (2.37±0.059 mIU/L). The serum insulin levels were significantly increased in the SHZ (2.69±0.057 mIU/L) versus the SH (p<0.01 for all comparisons). Although a slight decrease was seen in the serum insulin levels of the ADZ group compared to the AD group. This decrement did not reach a significant level.

Hippocampus Insulin Levels
In Fig. 1C 3.64, p<0.05]. The insulin levels in the hippocampus were significantly lower in the AD (0.162±0.021 mIU/L) compared to the SH (0.233±0.016 mIU/L) (p<0.05). Although a slight decrease was seen in the hippocampus insulin levels of the SHZ group versus the SH group, this decrement did not reach a significant level. Although a slight increase was seen in the hippocampus insulin levels of the ADZ group with respect to the AD group, this increment did not reach the significance level.

HOMA-IR levels
Mean values of HOMA-IR levels are given in Fig. 1D

Total Brain AβO Levels
Mean values of total brain AβO levels are given in Fig. 1E. There was a statistically significant difference in total brain AβO levels between groups [F(3,28)= 7.21, p<0.01]. The total brain AβO levels were significantly increased in the AD (15.90±0.913 mg/protein) versus the SH (11.46±0.573 mg/protein). The total brain AβO levels were significantly decreased in the ADZ (11.24±1.154 mg/protein) compared to the AD. The total brain AβO levels were significantly increased in the SHZ (14.91±0.792 mg/protein) with respect to the SH (p<0.05 for all comparisons).

Hippocampus AβO Levels
Mean values of hippocampus AβO levels are given in Fig. 1F

Total Brain p-Tau (Ser356) Levels
Mean values of total brain p-Tau (Ser356) levels are given in Fig. 2A. There was a statistically significant difference in total brain p-Tau (Ser356) levels between groups [F(3,20)= 10.09, p<0.001]. Although a slight increase was seen in the total brain p-Tau (Ser356) levels of the AD group compared to the SH group. This increase was not statistically significant. No statistically significant difference was observed between the SH and SHZ groups. The total brain p-Tau (Ser356) levels were significantly increased in the ADZ (0.405±0.075) versus the AD (0.705±0.068) (p<0.05). The total brain p-Tau (Ser356) levels were significantly increased in the ADZ (1.245±0.075) with respect to the AD (1.015±0.056) (p<0.01). The total brain p-Tau (Ser356) levels were significantly increased in the ADZ versus the SH (0.833±0.038) (p<0.001) and SHZ (0.935±0.041) (p<0.05).

Hippocampus p-Tau (Ser356) Levels
Mean values of hippocampus p-Tau (Ser356) levels are given in Fig. 2B
The total brain p-IRS-1 (Ser612) levels were significantly increased in the ADZ (0.693±0.040) compared to the AD (p<0.05). Although a slight increase was seen in the total brain p-IRS-1 (Ser612) levels of the SHZ (0.777±0.054) group with respect to the SH group. This increment did not reach the significance level. No statistically significant difference was observed between the ADZ and SHZ groups.

Total Brain p-GSK-3β (Ser-21) Levels
Mean values of total brain p-GSK-3β (Ser-21) levels are given in Fig. 2E. There was a statistically significant difference in total brain p-GSK-3β (Ser-21) levels between groups [F(3,24)= 8.31, p<0.01]. The total brain p-GSK-3β (Ser-21) levels were significantly decreased in the AD (0.405±0.075) with respect to the SH (0.705±0.068) (p<0.05). The total brain p-GSK-3β (Ser-21) levels were significantly increased in the ADZ (0.837±0.084) versus the AD (p<0.01). Although a slight increase was seen in the total brain p-GSK-3β (Ser-21) levels of the SHZ (0.850±0.053) group compared to the SH group. This increment did not reach the significance level. No statistically significant difference was observed between the ADZ and SH groups.

Total Brain IDE Levels
Mean values of total brain IDE levels are given in Fig. 3A.
There was not a statistically significant difference in total brain IDE levels between groups. Although a slight decrease was seen in the total brain IDE levels of the AD group compared to the SH group, this decrement did not reach the significance level. Although a slight increase was seen in the total brain IDE levels of the SHZ group versus the SH group, this increment did not reach the significance level. Although a slight increase was seen in the total brain IDE levels of the ADZ group with respect to the AD group, this increment did not reach the significance level.

Hippocampus IDE Levels
The mean values of hippocampus IDE levels are given in Fig. 3B

Total Brain NEP Levels
Mean values of total brain NEP levels are given in Fig. 3C. There was a statistically significant difference in total brain NEP levels between groups [F(3,24)= 12.80, p<0.001]. The total brain NEP levels were significantly decreased in the AD (0.281±0.033) versus the SH (0.450±0.036) (p<0.05). The total brain NEP levels were significantly increased in the SHZ (0.604±0.048) compared to the SH and AD (p<0.001). The total brain NEP levels were significantly increased in the ADZ (0.460±0.024) versus the AD (p<0.05). No statistically significant difference was observed between the ADZ and SH groups.

Hippocampus NEP Levels
Mean values of hippocampus NEP levels are given in Fig. 3D. There was not a statistically significant difference in hippocampus NEP levels between groups. Although a slight decrease was seen in the hippocampus NEP levels of the AD group with respect to the SH group, this decrement did not reach the significance level. No statistically significant difference was observed between the SHZ and SH groups. Although a slight increase was seen in the hippocampus NEP levels of the ADZ group compared to the AD group, this increment did not reach the significance level.

Object Recognition Memory Discrimination Index
The object recognition memory discrimination index is given in Fig. 4A

Object Location Memory Discrimination Index
The object location memory discrimination index is given in Fig. 4B

Spontaneous EEG power spectrum
The only changes observed were in delta (p<0.001) and spindle (p<0.05) powers, but not in alpha or beta powers using the repeated measures of ANOVA for the group and electrode location in spontaneous EEG. In this study, we particularly focused on delta and spindle rhythms because significant changes were observed in these frequency bands between the groups. Representative EEG traces of rats' delta and spindle power spectrums are given in Fig. 5A and B, respectively. We evaluated the difference between the groups as global power values, the mean of spectral power at the four recording electrodes, for these frequency bands.

Spontaneous EEG Delta Power
In this study, there was not a significant difference between the groups in terms of delta power spectrum. However, rANOVA showed a significant interaction effect for group x location [F(3,52)= 2.13, P < 0.001], indicating that delta power differed between the groups according to the electrode locations. In the left temporal region, delta power values in the AD, SHZ, and ADZ groups were found to be significantly decreased compared to the SH group (P<0.001).

Correlations Between Object Recognition Memory Discrimination Index and Left Temporal (T3) Spindle Power Spectrum
Significant correlations were found between changes in the object recognition memory discrimination index and the left temporal (T3) spindle power spectrum (Fig. 5C).
There was a positive correlation between the object recognition memory discrimination index and the left temporal (T3) spindle power spectrum (Pearson r= 0.448, p<0.001, N=53).

Correlations Between Object Location Memory Discrimination Index and Left Temporal (T3) Spindle Power Spectrum
Significant correlations were found between changes in the object location memory discrimination index and the left temporal (T3) spindle power spectrum (Fig. 5D). There was a positive correlation between the object location memory discrimination index and the left temporal (T3) spindle power spectrum (r= 0.433, p<0.01, N=50).

Discussion
The pathological changes of AD begin 20 years before the appearance of clinical symptoms [39]. In this respect, investigating the pathological changes seen at the early stage of the disease gains importance. Therefore, we used AβOs to induce changes similar to those seen in the early stages of AD pathology. Weight loss, which is often seen in Alzheimer's patients, may be a signal of the disease even before the onset of dementia symptoms. A study has shown that weight loss is a feature of the early stage of AD [40]. In line with this, we also found a significant weight loss in AD group rats injected with AβOs compared to the SH group. Thus, together with previous findings, it could be concluded that weight loss is one of the early signs of AD. It is known that insulin resistance or deficiency can alter Aβ accumulation and tau protein phosphorylation in early-onset AD [1]. At the same time, it has been shown that the HOMA-IR index of patients with mild cognitive impairment (MCI) was significantly higher than the control group [41]. At the same time, it has been shown that the HOMA-IR index, which is an indicator of peripheral insulin resistance, is significantly higher in MCI patients compared to the control group [42]. In this context, fasting blood glucose and serum insulin levels, which are common markers of peripheral glucose metabolism, were measured in our study, and peripheral insulin resistance was determined by the HOMA-IR index. In comparison to SH rats, Aβ oligomer injection significantly increased fasting blood glucose and serum insulin levels in AD rats. At the same time, the HOMA-IR index of the AD group was found to be significantly higher than the SH group. Therefore, our findings support that peripheral insulin resistance contributes to the pathogenesis of early AD biomarkers and reveal insights into the pathogenesis of AD. At the same time, it is known that chronic peripheral hyperinsulinemia causes the downregulation of insulin receptors in the blood-brain barrier and reduces the amount of insulin transported to the brain [43]. In parallel, it has been observed that AD patients who have peripheral hyperinsulinemia have lower brain insulin concentrations [44]. In our study, it was shown that the hippocampal insulin level of the AD group was significantly decreased compared to the SH group. Our findings, together with previous observations, strongly suggest that chronic hyperinsulinemia in AD may induce a decrement in insulin levels in the brain. Insulin and Aβ are known to be amyloidogenic peptides that share a common sequence recognition motif. At the same time, AβOs can bind to insulin receptors and inhibit the phosphorylation of the receptor [13]. Therefore, we examined the effect of AβOs on the insulin signaling pathway by the injection of AβOs.
A previous report showed that IRS-1 Ser 612 phosphorylation is decreased along with the reduction in total IRS-1 levels in the early phase of AD disease. Also, it has been shown that IRS-1 Ser612 phosphorylation gradually increases with the progression of the disease. So, it was concluded that this decrease in IRS-1 Ser612 phosphorylation is a defense mechanism that ensures the continuation of the insulin signaling pathway in the early phase of the disease [45]. In our study, we found that total brain IRS-1 Ser612 phosphorylation was significantly decreased in the AD group compared to the SH group. In addition, hippocampus IRS-1 Ser612 phosphorylation was slightly decreased in the AD group, although it did not reach the significance level. Together with the above-mentioned finding, our findings suggest that a decrement in p-IRS1 Ser612 levels in the brain occurs early in the progression of Alzheimer's disease. Therefore, we can conclude that brain IRS-1 activation is impaired at the onset of Alzheimer's disease.
When the IRS-1 complex is activated in the insulin signaling pathway; it activates the PI3K→PKB/Akt pathway, which phosphorylates the Ser9/21 region of GSK3β, and suppresses the GSK-3β activity [46]. It has been shown that insulin resistance or deficiency can change Aβ and tau phosphorylation, thereby contributing to the onset of AD by overactivity of GSK-3β [47,48]. We found a marked decrement in p-GSK-3β Ser21 phosphorylation, which was accompanied by a reduction of IRS-1 phosphorylation in the hippocampus and total brain. GSK-3β regulates the binding of tau protein to microtubules [49]. In particular, increased tau toxicity associated with Aβ depends on the phosphorylation of tau at Ser262/356 sites that decreases the binding of tau protein to microtubules [50]. Consistent with this, we found a significant increase in the p-Tau Ser356 level in the hippocampus of the AD group versus the SH group. In addition, we observed a trend toward an increase in the p-Tau Ser356 level in the AD group compared to the SH group; however, this increase did not reach statistical significance.
In the central nervous system, insulin signaling is required for the expression of genes that modulate memory [45]. Hence, the observed change in the insulin signaling pathway very likely alters the cognitive functions in the current study. We used object recognition and object location memory tests, which are widely used tests to assess memory deficits. Object location memory requires the hippocampus to encode, consolidate, and recall information [51,52]. According to a recent study, AβOs injections resulted in a significant impairment of the object location memory index. In this study, AβOs impair performance in hippocampaldependent associative learning tasks [53]. Also, AβOs injection resulted in significant impairment of the spatial memory index score for the object location memory test [54]. Consistent with these studies, we found a significant decrease in the object location memory discrimination index in the AD group versus the SH group. Several different brain regions are critical for the new object recognition memory, including the insular cortex [55,56], the perirhinal cortex [57,58,56], and the ventromedial prefrontal cortex [59]. According to a previous study, the object recognition memory discrimination index significantly decreased in the Aβ oligomer-treated mice [54]. In parallel, we found a significant decrease in the new object recognition memory discrimination index in the AD group versus the SH group. These findings suggest that an altered insulin signaling pathway mediated AβOs-induced memory dysfunctions at the onset of Alzheimer's disease and ultimately contributed to the pathology of Alzheimer's disease.
Regulation of proteases that degrade Aβ may represent an important therapeutic approach. The two main peptidases that mainly regulate Aβ metabolism in the brain are the enzymes NEP and IDE [60]. It is well known that IDE [61,62] and NEP [63] levels decrease with aging and at the early stage of AD. In the current study, we found a significant reduction in hippocampus IDE levels in the AD group versus the SH group and a tendency for a reduction in the total brain region, although it was not significant. In addition, we discovered a significant reduction in total brain NEP levels in the AD group compared to the SH group, as well as a non-significant reduction in the hippocampus.
It is well known that Zn +2 deficiency commonly takes place in AD pathology [24,25]. Earlier studies indicate that Cyclo-Z, a combination of CHP and Zn +2 , increases Zn +2 absorption [28] and improves weight control [30]. Therefore, first we aimed to investigate how Cyclo-Z affects weight control. We found a significant increase in weight in the ADZ group versus the AD group. Moreover, Cyclo-Z administration decreases fasting blood glucose [64][65][66] and enhances insulin sensitivity and glucose tolerance [66][67][68]. We found a significant reduction in fasting blood glucose levels in the ADZ group versus the AD group. In addition, although not significant, we found a tendency for a reduction in the HOMA-IR index and serum insulin levels in the ADZ group versus the AD group. When the hippocampus insulin levels were examined, no significant difference was found in the ADZ group. Therefore, in our study, Cyclo-Z was found to have a possible positive effect on weight loss, peripheral insulin resistance, and brain insulin levels in the ADZ rats. However, we found a tendency for a reduction in the weight in the SHZ group versus the SH group. This reduction can be explained by Zn +2 toxicity. In a previous study [65], an overdose of Zn +2 has been shown to have a toxic potential in humans. At the same time, fasting blood glucose levels, serum insulin levels, and HOMA-IR index also increased, and hippocampal insulin levels decreased in the SHZ group versus the SH group. These results demonstrate conclusively that Zn +2 toxicity not only affects weight regulation, but also causes peripheral insulin resistance and a decrease in brain insulin levels.
The effect of Cyclo-Z on the brain insulin signaling pathway was also investigated in our study. Ser phosphorylation of IRS-1 might negatively or positively regulate insulin signaling depending on the sites where it occurs. If insulin phosphorylates IRS-1 on Ser residues, insulin signaling is desensitized and adversely affected. However, the insulin signaling pathway is positively regulated if PKB/ Akt phosphorylates IRS-1 over Ser residues. IRS-1 proteins have four Ser residues that act as PKB/Akt phosphorylation sites, and Ser612 is one of them. When insulin signaling is activated for a long time, it overexpresses PKB/Akt and phosphorylates IRS-1 from these Ser sites. Increased IRS-1 Ser612 phosphorylation protects IRS-1 from the rapid action of tyrosine phosphatases and keeps it in its phosphorylated active form [69]. In line, we found a significant increase in hippocampal and brain p-IRS-1 Ser612 levels in the ADZ group compared to the AD group. Therefore, we concluded that p-IRS-1 Ser612 phosphorylation was increased in ADZ group animals due to long-term activation of insulin signaling. Similar results were found in the p-IRS-1 Ser612 level of the SHZ group, which was seen to have improved insulin signal due to Cyclo-Z treatment. We found a significant decrease in the hippocampal p-IRS-1 Ser612 level in the SHZ group versus the SH group. There was also a trend towards an increase in the total brain p-IRS-1 Ser612 level, although this was not significant. The downstream target, p-GSK-3β Ser21, was also examined to reveal the changes in the insulin signaling pathway. We found a significant increase in the hippocampal and total brain p-GSK-3β Ser21 levels in the ADZ group versus the AD group. In addition, when the p-GSK-3β Ser-21 levels of the SHZ group were examined, it was shown that they tended to increase in the hippocampus and total brain regions compared to the SH group. It is well known from previous studies that Zn +2 has insulin-like effects and facilitates insulin signaling [22]. These effects are largely dependent on the activation of PKB/Akt signaling [70]. Therefore, we suggest that Cyclo-Z showed a facilitatory effect on the brain insulin pathway in the SHZ and ADZ groups.
The effects of Cyclo-Z treatment on AβO and tau protein phosphorylation were also investigated. We found a significant decrease in AβO levels in the ADZ group versus the AD group in both regions, whereas we found a significant increase in p-Tau Ser356 levels in both regions of ADZ group rats compared to the AD group. We presume that other possible mechanisms which contribute to the AD pathophysiology may also play a role in the increased p-Tau Ser356 levels. GSK-3β and protein phosphatase 2A (PP2A) are important enzymes that control tau hyperphosphorylation. The relationship between these two enzymes and their effect on tau hyperphosphorylation is not yet fully understood. In a study, it was found that GSK-3β and PP2A regulate each other and tau phosphorylation directly and indirectly through modulation of each other [46]. Inactivation of GSK-3β resulting from PI3K-AKT activation has been shown to lead to PP2A demethylation and inactivation, resulting in tau hyperphosphorylation. Based on these results, we suggest that targeting GSK-3β could lead to an increment of tau phosphorylation in Ser262/356, which is one of the PP2A sensitive regions. In line, it has been suggested that the Ser262/356 region required for tau pathology is specific for PP2A and that PP2A should be addressed as the therapeutic target for tau pathology [46]. So, while Cyclo-Z treatment decreased the GSK-3β activity in the ADZ group, it could not reduce the p-Tau Ser356 levels in the current study.
Although the Cyclo-Z agent was found to decrease the AβOs level in the ADZ group, it was shown that the AβOs level of the SHZ group was significantly increased in all regions compared to the SH group. When the Cyclo-Z agent was administered to healthy rats, it had a negative effect on the AβOs level. It is known that aggregation of Aβ peptides can be rapidly induced in the presence of Zn +2 ions under physiological conditions in vitro. It has been shown that Zn +2 ions coexist with Aβ deposits. Although the Zn +2 concentration required for fibrillation is controversial, it certainly has an important role in Aβ aggregation [26]. As high Zn +2 ion concentrations are also known to induce Aβ aggregation and tau protein modification [68], it is not surprising that Cyclo-Z treatment increases AβOs levels in healthy rats. Although not significant, Cyclo-Z agent tended to increase the p-Tau Ser356 levels of the SHZ group compared to the SH group. Therefore, we suggest that excessive Zn +2 supplementation may cause AD-like toxicities.
Additionally, we used memory tests to examine the effect of Cyclo-Z treatment on cognitive functions. We found a significant increase in the discriminative index of object recognition test in the ADZ group versus the AD group. At the same time, we found a tendency for an increment in the SHZ group versus the SH group but did not reach the significance level. Likewise, we found a significant increment in the discriminative index of the object location test in the ADZ group versus the AD group. The discriminative index of object location test showed a similar increment in the SHZ group versus the SH group.
It is a known fact that IDE and NEP enzymes, which are known to degrade Aβ [71], require Zn +2 for both gene expression and activity [26]. Thus, Zn +2 supplementation can induce the synthesis of these Aβ-clearing enzymes [23]. At the same time, it has been suggested in previous studies that Cyclo-Z is the only agent that can increase intracellular IDE [66,67]. Therefore, in our study, the effect of Cyclo-Z on IDE and NEP enzymes was also investigated. Cyclo-Z treatment significantly increased total brain NEP level in the ADZ group compared to the AD group, according to our data. Also, we found a significant increase in hippocampal IDE and total brain NEP levels in the SHZ group versus the SH group. Our findings, together with previous observations, strongly suggest that Cyclo-Z has a positive effect on Aβ degradation by increasing the amount of IDE and NEP enzymes.
Accumulating evidence indicates that AβOs exposure induces changes in presynaptic and postsynaptic synapses. AβOs are likely to induce changes in neural network dynamics. For this reason, in our study, we investigated the effects of AβOs on neural network dynamics under ketamine anesthesia using the EEG technique. It is well known that spontaneous EEG under ketamine anesthesia mimics sleeping EEG patterns [72], and AD pathophysiology markedly disrupts sleeping physiology [73]. Moreover, it was shown that Alzheimer's patients showed reduced spindle density and spindle activity [74]. An early study also indicated that AD pathology diminished cortical spindle power in APP/ PS1 Tg mice [75]. Consistent with the aforementioned studies, we found that the left temporal region (T3) spindle power was significantly reduced in the AD group compared to the SH group. Hence, we can conclude that the diminished spindle power spectrum in the left temporal region reveals insights into the pathogenesis of AD. In addition, we found a significant increase in temporal spindle spectral power values in the ADZ group versus the SH group. At the same time, the spindle power values of the SHZ group were significantly higher than SH, AD, and ADZ in all electrode regions. Although AβO and tau pathology were observed in the SHZ group, Cyclo-Z treatment increased spindle power values in all electrode areas. In line with this, adequate Zn +2 concentration is known to improve sleep quality [76]. A previous study has also shown that increased spindle power contributes to sleep quality with significant effects on memory consolidation [73]. Consistent with this, we found a positive correlation between left temporal (T3) spindle power and the new object discrimination index (Pearson r= 0,448, p<0.001, N=53), and the location discrimination index (Pearson r= 0,433, p<0.01, N=50). From this point of view, it could be concluded that Cyclo-Z might increase spindle formation in the left temporal region and memory consolidation. It has been noted that slow brain waves are impaired in the presence of AβOs before the deposition of Aβ plaques [77]. Additionally, it was shown that reduced delta power has a relationship with increased tau deposits [78]. In the left temporal region, delta power values in the SHZ, AD, and ADZ groups were found to be significantly decreased compared to the SH group. The decrease in delta power in the SHZ group might be associated with the AβOs toxicity and tau pathology depending on Zn +2 toxicity. In addition, although sleep delta spectral power was higher in the ADZ group compared to the AD group, this difference did not reach statistical significance.

Conclusions
In conclusion, our results are consistent with previous findings that weight loss, peripheral hyperinsulinemia, and peripheral insulin resistance may play a central role in the early stage of AD. In addition, it can be concluded that peripheral hyperinsulinemia seen in the early phase of AD may decrease the hippocampal insulin level. Furthermore, decrement of brain insulin levels may also cause dysregulation of the brain insulin pathway. These changes have also been shown to reduce the levels of IDE and NEP that exacerbate AβOs toxicity and may play a role in increased p-Tau Ser356. Furthermore, AβOs impair memory functions and network dynamics by altering the brain insulin signaling pathway in the early stage of AD, ultimately contributing to AD pathology. The Cyclo-Z treatment was found to have a positive effect on weight loss, peripheral insulin resistance, and brain insulin level. Cyclo-Z treatment of AD rats increased the levels of amyloid-degrading enzymes and reduced AβOs accumulation. At the same time, the Cyclo-Z agent applied to ADZ rats could not prevent the increase of p-Tau Ser 356 levels. However, it was observed that the Cyclo-Z agent applied to healthy animals had a negative effect due to Zn +2 toxicity. Cyclo-Z treatment in healthy animals increased the level of amyloid-degrading enzymes but resulted in the accumulation of AβOs. It is thought that this increase observed in AβOs accumulation may be caused by Zn +2 toxicity. At the same time, it was found that the p-Tau Ser356 level was similarly increased due to toxicity. Cyclo-Z treatment was found to have positive effects on new object recognition and location memory functions in healthy and AD group rats. Besides, Cyclo-Z treatment was found to have a significant positive effect on spindle power and slightly increased the delta power in the left temporal region. In conclusion, agents that enhance insulin delivery, such as Cyclo-Z, may be an important therapeutic strategy for the prevention of various pathological processes in the early stages of AD.