Triiodothyronine Treatment reverses Depression-Like Behavior in a triple-transgenic animal model of Alzheimer’s Disease

Alzheimer disease’s (AD) is a neurodegenerative disorder characterized by cognitive and behavioral impairment. The central nervous system is an important target of thyroid hormones (TH). An inverse association between serum triiodothyronine (T3) levels and the risk of AD symptoms and progression has been reported. We investigated the effects of T3 treatment on the depression-like behavior in male transgenic 3xTg-AD mice. Animals were divided into 2 groups treated with daily intraperitoneal injections of 20 ng/g of body weight (b.w.) L-T3 (T3 group) or saline (vehicle, control group). The experimental protocol lasted 21 days, and behavioral tests were conducted on days 18–20. At the end of the experiment, the TH profile and hippocampal gene expression were evaluated. The T3-treated group significantly increased serum T3 and decreased thyroxine (T4) levels. When compared to control hippocampal samples, the T3 group exhibited attenuated glycogen synthase kinase-3 (GSK3), metalloproteinase 10 (ADAM10), amyloid-beta precursor-protein (APP), serotonin transporter (SERT), 5HT1A receptor, monocarboxylate transporter 8 (MCT8) and bone morphogenetic protein 7 (BMP-7) gene expression, whereas augmented superoxide dismutase 2 (SOD2) and Hairless gene expression. T3-treated animals also displayed reduced immobility time in both the tail suspension and forced swim tests, and in the latter presented a higher latency time compared to the control group. Therefore, our findings suggest that in an AD mouse model, T3 supplementation promotes improvements in depression-like behavior, through the modulation of the serotonergic related genes involved in the transmission mediated by 5HT1A receptors and serotonin reuptake, and attenuated disease progression.


Introduction
Alzheimer's disease (AD) is the most common cause of dementia in the aging population. It has been reported that among AD cases, less than 1% of the patients are under 60 years of age and more than 40% are over the age of 85 (Reitz et al., 2011). It is an irreversible and severe neurological disorder, which can slowly progress for years (Stelzmann et al., 1995). Pathologically, AD is characterized by extensive neuronal loss accompanied by accumulated neurofibrillary tangles and amyloid plaques in the brain (Karch and Goate, 2015). In addition to the well-documented cognitive decline, natural AD progression includes functional losses and behavioral/psychological changes that are present in neuropathogenic pathway. Nevertheless, genetic alterations cannot account for and do not explain all AD cases (Iacono et al. ,2014;Xia et al., 2015). Recently, evidence has underpinned the concept that AD is a metabolic and degenerative disease (Procaccini et al., 2016). Studies have also shown that, at some level, insulin, growth hormone (GH), insulinlike growth factor (IGF-1), cholesterol, adipokine and thyroid hormone (TH) are involved in AD pathogenesis (Fu et al., 2010;Manuel Gomez Saez, 2012;Bavarsad et al., 2019).
It is well known that the central nervous system is an important target for TH action (Oliveira et al., 2015;Zendedel et al., 2016). During brain maturation, THs play roles in development, differentiation, myelination, cycle stabilization, and participate in neural and glial signaling (Manzano et al., 2007;Ahmed et al., 2008;Mendes-de-Aguiar et al., 2008). Previous reports investigating hypothyroidism in adult rats showed reductions in features such as, differentiation, survival, and hippocampal progenitor cell neurogenesis, which probably contribute to the observed cognitive and behavioral deficits as these symptoms were reversed following hormone replacement (Desouza et al., 2005;Moog et al., 2017;Zhang et al., 2006). Furthermore, in clinical practice, TH therapy has shown to significantly improve cognition and emotions in patients with AD (Bauer et al., 2008). Studies demonstrate that subclinical hypothyroidism could result in depression-like behavior and the reduced hippocampal T3 prior to the reduction of thyroid hormone in plasma might be taken as an early sign of hippocampus impairment (Ge et al., 2014). A study with a type 3 deiodinase-deficient mouse showed that increased thyroid hormone in the brain correlates with hyperactivity and with decreases in anxiety and depression-like behaviors (Stohn et al., 2016).
In vivo studies have demonstrated that APP gene and protein expression, as well as APP secretase cleavage products, are influenced by thyroid status (O'Barr et al., 2006). In addition, other studies have shown that a negative correlation exists between rT3 or rT3/T3 levels in the cerebral spinal fluid (CSF) and cognitive performance in AD patients (Accorroni et al., 2017). It has been also demonstrated that in euthyroid patients, attenuated T4 concentrations and augmented serum TSH levels are associated with AD-related encephalic changes (Choi et al., 2017).
An inverse association between serum free T3 and the risk of AD has been demonstrated by Quinlan et al. (2019). Although this study is very relevant, the mechanisms by which the decreased serum free T3 levels increase the risk of developing this neurodegenerative disease have not been elucidated. Moreover, peripheral and central administration of T3 improved memory deficits in AD animal models (Farbood et al., 2017). However, the clinical significance of those findings remains unclear. Notably, behavioral alterations, including depression, are commonly observed at different stages of AD, as well as in other neurodegenerative diseases, reaching 80-90% of prevalence in some studies (Steinberg et al., 2004;Cortés et al., 2018). It should be pointed out that T3 alone or in combination with other antidepressants has been successfully used for the treatment of major depression (Cooper-Kazaz and Lerer, 2008;Kelly and Lieberman, 2009;Touma et al., 2017). Previous studies shown that HT 1A receptors are mainly concentrated in the limbic system, particularly the hippocampus (dentate gyrus and CA1), lateral septum, and amygdala, in cingulate and entorhinal cortices, and in the dorsal and median raphe nuclei, which are regions implicated in spatial learning and memory (Barnes and Sharp, 1999;Lanfumey and Hamon, 2000;Glikmann-Johnston et al,. 2015). Gamma aminobutyric acid (GABA) plays an important role in the regulation of neuronal excitability, acting as the main inhibitory neurotransmitter in the brain (Roberts and Kuriyama, 1968;Owens and Kriegstein, 2002;Mody and Pearce, 2004). GABA is converted primarily from glutamic acid by the action of glutamic acid decarboxylase (GAD), binds to GABA A and GABA B receptors, and undergoes reuptake to presynaptic nerve terminals with the aid of transporter proteins called GAT-1, GAT-2 GAT-3, and BGT-1 (Fuhrer et al., 2017). Even though earlier reports have shown that GABAergic neurons are relatively resistant in AD, further research has suggested the involvement of GABAergic terminal atrophy in the progression of the disease (Hardy et al., 1987a, b). In addition, Furher et al. (2017) have demonstrated an impaired expression of GABA transporter in the hipoccampus in human Alzheimer disease (Fuhrer et al., 2017). We hypothesized that T3 treatment could reverse the depression-like behavior through molecular changes in the hippocampus of transgenic AD mice. To understand the molecular mechanism involved, we analyzed genes related to GABA and Serotonergic systems, neuroplasticity and AD progression in control and T3 treated 3xTg-AD mice. In addition, we evaluated genes that can be positively or negatively regulated by TH to verify the hippocampal thyroid status.

Animals
Six-months-old male transgenic 3xTg-AD (APPswe, PS1m146v, tauP301L) mice, weighing between 20 and 35 g, were employed for the study. This animal model of AD expresses three dementia-related transgenes, APPSWE, PS1M146V, and tauP301L, and the progression of the pathophysiological alterations is age-dependent. This model has a genetic background that allows the study of the interaction between Aβ peptide and neurofibrillary tangles (Oddo et al., 2003). Furthermore, this animal model has behavioral and cognitive impairments characterized, allowing the study of the link between behavioral impairments and Alzheimer's disease (Sterniczuk et al., 2010). A previous study showed that at six months-old, these animals already developed associative learning and spatial working memory deficits, as well as an impairment in contextual fear conditioning (Webster et al., 2014). The animals were supplied by The Jackson Laboratory and were maintained in the vivarium at the Universidade Presbiteriana Mackenzie. The mice were housed in groups of up to 5 animals, in plastic boxes, and maintained in a temperature-controlled room (20-24°C), with a 12-hour light-dark cycle. The mice had free access to food and water ad libitum throughout the experimental protocol. All procedures were performed in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and were approved by the Animals Ethics Committee of the Universidade Federal de São Paulo (UNIFESP), under protocol # 1880171017. The 3xTg-AD animals were divided into two groups, with one group receiving T3 (T3 group) and the other saline (vehicle, control group).

Experimental protocol
Hypertriiodothyronemia was induced with intraperitoneal (IP) injections of triiodothyronine (T3) (L-T3, Sigma-Aldrich, St. Louis, MO, USA) at the dose of 20 ng/g of b.w., which is a replacement dose as previous described by Bianco et al. (2014). Animals were weighed 4 days per week for dose adjustment according to the body weight. This dose was considered ten folds higher than the physiological doses calculated considering a daily T3 production of roughly 2.0 ng/g of b.w. (Bianco et al., 2014). Stock solutions of T3 (1 µg/µL) in 40 mM sodium hydroxide, pH = 10 were prepared and frozen until the day of use (Bianco et al., 2014). The solutions were adjusted to pH = 7.7 and IP injections were administered using a 1.0 ml syringe. Control animals received injections of 0.9% NaCl solution (saline, vehicle) that had the same volume as the injections received by the T3 group. The treatment consists in chronic administration of T3 (21 days) since previous studies demonstrate a modulation of serotonergic system in this period (Lifschytz et al., 2011). On days 18-20, behavioral tests were performed ( Fig. 1).

Behavioral tests
The behavioral tests included open field, tail suspension and forced swimming, with a random double-blind distribution. All tests were performed at 24-hour intervals and the testing order was determined according to the degree of invasiveness, starting with the open field, followed by the tail suspension and lastly forced swimming test. Each experimental group contained 4 or 6 mice. All tests were carried out between 9:00 AM and 12:00 PM, and the researcher present during each test remained outside the experimental area, only entering the room between trials. Each test was recorded with an overhead camera and behavioral parameters were later analyzed by at least two researchers.

Open-Field Test
The open-field test arena was constructed in acrylic, with square, rectangular or circular enclosures (25 to 250ccm²), a white floor labeled with black symmetrical squares and walls to prevent animal escape. The animals were placed in the center of the arena for 5 min, and the total number of squares crossed, center ratio, time in center zones and rearing were assessed. Exploratory activity is related to the total distance traveled during the test, vertical activity is associated with the number of rearing, and anxiety-like responses are linked to the center ratio (Roth and Katz, 1979).

Tail suspension test
This tail suspension test is based on the observation that when rodents are placed in an inescapable situation, they initially exhibit escape-oriented movements and then assume an immobile posture. In this test, the plight involves hemodynamic stress applied by hanging the animals by the tail (Thierry et al., 1986). In our protocol, the mice were suspended in a rectangular box with a suspension bar of sufficient size, in a location along the bar that was away from the walls of the box (Can et al. 2011). The test was videotaped Fig. 1 Representation of experimental design. Six-months-old (180 days) male transgenic 3xTg-AD were divided into 2 groups treated with daily intraperitoneal injections of 20 ng/g of body weight (b.w.) L-T3 (T3 group) or saline (vehicle, control group). The experimental protocol lasted 21 days, and behavioral tests were performed on days 18-20. At the end of the experiment, the serum hormones (TH profile) and hippocampal mRNA expression were evaluated evaluate the GABAergic system; TPH2, HT1RA e SLC6A4 genes to evaluate the serotonergic system; BDNF, NTF3 and NGF genes to evaluate neuroplasticity; APP, GSK3, ADAM10 and BACE1 genes to evaluate the Alzheimer's disease progression and SOD2, MCT8, SLCO1C1, BMP-7 and HR genes to evaluate the neuronal influence of T3 in the hippocampus.

Serum T3 and T4 measurements
Serum T3 and total T4 levels were measured using the Elecsys® Kit (Roche Diagnostics, Germany), following the instructions of the manufacturer. Negative controls with Cal set were used before making the measurements and the obtained results were used for calibration. The values were measured automatically and based on a standard curve. For T3, the lower limit of detection for the kit is 1.25 nmol/l and the upper limit is 8.50 nmol/l. For T4, the lower limit of detection is 5.40 nmol/l and the upper limit is 320.0 nmol/l. None of the samples used in the data analyses had values that were outside of these ranges.

Data Analysis
The RT-PCR data were analyzed using Microsoft Excel and the samples were normalized based on the respective controls. The difference between samples was normalized according to the variation in the mean of CT (ΔCT), which was subsequently used to calculate the relative expression levels of genes of interest. The results from these experiments are expressed in arbitrary units (Livak and Schmittgen, 2001).
All results are presented as mean ± SEM. The assumption of normal data distribution was determined using the Shapiro-Wilk test. Parametric comparisons were performed with data that passed the normality test. In these cases, between-group comparisons were analyzed with the Student's unpaired t-test. Significance level was set at p < 0.05 and the t value is presented to measure the size of the difference relative to the variation in the sample data. On the other hand, data that did not have a normal distribution were compared using the Mann-Whitney test. The Grubbs' test was employed for detecting outliers. Cohen's d analysis was used to evaluate the effect sizes between the groups, which is represented by the difference between means divided by the standard deviation. Effect sizes were interpreted as: small (0.2 < d < 0.5), moderate (0.5 < d < 0.8) and large (d > 0.8). Relationships between variables were identified with Pearson correlation analyses. Significance level was set at p < 0.05. All statistical analyses were performed with GraphPad Prism 6 statistical software (La Jolla, CA, USA).
for 5 min, and the immobility time and latency to the first immobility episode were recorded.

Forced swim test
This forced swim test is based on the development of an immobile posture immediately following a stressful situation (Porsolt et al., 1977) and involves individually placing of each mouse in a water-filled polypropylene cylinder (radius = 30 cm; depth = 50 cm) at 25°C. The animals are unable to escape or touch the bottom of the cylinder. Animals were submitted to the test for 5 min and immobility time and latency to the first immobility episode were recorded. It is known that immobility time is indicative of low resilience and highly associated with depression-like behavior (Porsolt et al., 1979).

Euthanasia process
Two days after the last behavior test, animals were anesthetized with ketamine (100 mg/ml) and xylazine (20 mg/ml) administered by IP injection in a dose of 0.2 ml of ketamine and 0.1 mL of xylazine per 100 g of b.w.. Blood samples were collected from the left ventricle, centrifuged at 3500 rpm at 4 ° C (Excelsa II 206 BL, Sao Paulo, SP, Brazil) for 15 min and stored in a freezer at − 20°C. These samples were later used for determining serum T3 and total T4 levels. Hippocampal samples were collected, immediately frozen in liquid nitrogen and stored at -80°C for use later on in the gene expression experiments by Real-Time quantitative PCR (qPCR).

Total RNA isolation and real-time PCR
Total RNA from brain tissue corresponding to the murine hippocampus was isolated using the TRIzol® reagent (Invitrogen, Carlsbad, CA, USA), according to the RNA extraction protocol provided by Invitrogen. RNA concentrations were determined spectrophotometrically by recording the absorbance readings at 260 and 280 nm (NanoDrop® ND-1000 UV-Vis, Delaware, USA). Approximately 2 µg of total RNA from each sample was used to perform the reverse transcriptase reactions. First-strand cDNAs were synthesized using the MML-V reverse transcriptase (Invitrogen, Carlsbad, CA, USA). From the obtained cDNA, cycle curves were performed with each primer set. Real-Time qPCR was performed using the EVA Green RT-PCR assay (Solis Biodyne, European Union) and run on an ABI PRISM 7700 Sequence Detector (ABI Applied Biosystems). The primers sequences were manufactured by Integrated DNA Technologies (IDT, Sao Paulo, SP, Brazil) and are listed in Table 1. We used GABA1 and GABA2 genes to and a significant decrease in total T4 (14.9 ± 1.07 µg/dl vs. 80.9 ± 3.02 µg/dl, p < 0.0001, t = 20) levels when compared to control animals.

Tail suspension test
In the tail suspension test (Fig. 2E), the T3 treated animals had a latency time that was similar to control mice (36.25 ± 12.76 s vs. 63.75 ± 29.04 s, p = 0.42, t = 0.87). In contrast, the T3-treated group exhibited a significant reduction in immobility time when compared to control mice

2.08
Immobility time 3.59 Numbers in italics represent a moderate magnitude of effect, whereas numbers in bold represent a large magnitude of effect  Table 4 Pearson correlation matrix between serum hormones measurement, hippocampal gene expression and behavioral parameters assessed in forced swimming test and tail suspension test.

Discussion
The findings of this study showed that T3 supplementation in the transgenic AD mouse model (APPswe, PS1m146v, tauP301L, 3xTg-AD) promotes improvements in depression-like behavior. These positive effects appear to be mediated through the modulation of the serotonergic pathway, since the immobility time on Tail Suspension Test presented a strong correlation with SERT and HTR1a expression, whereas immobility time on Forced Swim Test presented a strong correlation with SERT expression. The Open Field Hippocampal gene expression Fig. 3 shows the effects of T3 treatment on the expression of 17 hippocampal genes. A significant increase in SOD2 (2.35 ± 0.43 vs. 1.00 ± 0.06, p = 0.01, t = 3.0) and Hairless (2.71 ± 0.66 vs. 0.84 ± 0.16, p = 0.01, t = 2.97) and a decrease in MCT8 (0.60 ± 0.08 vs. 1.00 ± 0.08, p = 0.01, t = 3.41) and BMP-7 (0.55 ± 0.06 vs. 1.00 ± 0.08, p = 0.002, t = 4.30) gene expression levels were observed in animals treated with T3 compared to control mice, respectively. When compared with the control group, the T3 animals displayed significant differences in the expression of genes associated with AD, as evidenced by reductions in GSK3β (1.00 ± 0.10 control vs. 0.59 ± 0.11 T3 group, p = 0.02, t = 2.61), ADAM 10 (1.00 ± 0.12 control vs. 0.51 ± 0.11 T3 group, p = 0.01, t = 2.90) and APP (1.18 ± 0.40 control vs. 0.28 ± 0.04 T3 group, p = 0.04, t = 2.24) levels. It is worth pointing out that we did not observe any significant changes in the gene expression levels of the neurotrophins. Additionally, the expression of behavior-related genes analyses showed that SERT (0.50 ± 0.13 vs. 1.13 ± 0.15, p = 0.01, t = 3.20) and HTR1a gene expression levels (0.78 ± 0.08 vs. 1.00 ± 0.05, p = 0.04, t = 2.29) were significantly reduced compared to the control group. SERT and HTR1a are related to serotonergic transport and transmission, respectively, The w., Black). n = 6. Notably, there was a significant downregulation in the expression of serotonin-related genes in the T3-treated group. Data represent mean (± SEM) analyzed by Student's unpaired t-test. The significance between groups was * p < 0.05 and ** p < 0.01 compared to control group As it is known that thyroid hormones, including T3, have antidepressant effects, it is plausible that enhancement of monoaminergic neurotransmission could account for the observed T3-mediated behavioral effect, likely through direct modulation of serotonergic system gene expression. According to Bauer et al. (2002), thyroid hormones exert a modulatory effect on depressive behavior by decreasing 5HT1a receptor sensitivity in some areas of the brain (Bauer et al., 2002). More recently, higher hippocampal 5HT1a receptor densities were found in rats with hypothyroidism, possibly representing an early response to the local synaptic loss of serotonin receptors (Lee et al., 2018). In addition, a substantial body of evidence indicates that central 5-HT metabolism is altered in response to changes in thyroid status (Newman et al., 2000;Lifschytz et al., 2006). In fact, in thyroid intact rats, T3-releasing implants suppressed 5-hydroxyindoleacetic acid (5-HIAA) levels, the principal metabolite of serotonin, within 7 days, in a dose-dependent manner (Henley and Koehnle, 1997). Consistent with those findings, our study shows that thyroid hormones modulate the serotonergic pathway, as shown by the strong correlation with the gene expression of proteins acting at pre-and postsynaptic levels mediated by a serotonin transporter protein (SERT), which transports serotonin from the synaptic cleft back to the presynaptic neuron, and the 5-HT1a receptor in the postsynaptic neuron, consequently influencing the depression-like behavior commonly associated with AD.
Additionally, it has been shown that 5HT1a autoreceptors and SERT play an important role in the regulation of serotonergic neurotransmission, through negative feedback and neurotransmitter reuptake mechanisms, respectively (Quentin et al., 2018). In the present study, T3 treatment attenuated 5-HT1A and SERT gene expression in this AD mouse model, suggesting that serotonergic neurotransmission is influenced by T3, which could contribute to the observed antidepressant action. This proposal is further reinforced by the strong correlation between serum T3 levels and the expression of genes involved in these processes, and the strong correlation between the serotonergic systemrelated genes and the antidepressant parameters associated with each test, especially the tail suspension test.
Interestingly, the beneficial aspects of serotonergic neurotransmission modulation are not restricted to the improvement of AD-affected behaviors, but are also involved in delaying disease progression. According to Noristani et al. (2012), a high tryptophan diet reduces CA1 intraneuronal β-amyloid in the 3xTg-AD mouse model, suggesting that enhanced tryptophan intake and consequent increase in 5-HT neurotransmission could effectively reduce plaque formation in AD (Noristani et al., 2012). In addition, fluoxetine stimulates type 2A protein phosphatases (PP2A), consequently activating Wnt/β-catenin signaling and inhibiting Test does not show a correlation with genes related to serotonergic pathway in Pearson analysis, although the Cohen's d analysis showed that T3 had a strong effect on the total number of squares crossed, center ratio and rearing. This result of Open Field can be explained by freezing and limited exploratory behavior, characteristics of this animal model as related symptom of Alzheimer's Disease. Other studies demonstrated that 3xTg-AD mice showed hypoactivity in Open Field Test (Sterniczuk et al., 2010;Filali et al., 2012).
Previous studies have shown that thyroid hormones play important roles in a variety of organs. An association between the central nervous system and AD as well as hormonal disturbances, such as in hypo-and hyperthyroidism, has been previously reported (O'Barr et al., 2006;Gessl et al., 2012;Chaker et al., 2016). In the current study, we showed that the IP injections of T3 increased serum T3 and decreased T4 levels. We also showed that T3-treated animals presented changes in gene expression levels that were suggestive of a hyperthyroid status in the hippocampus. These results do not only demonstrate the clinical accuracy of our hypertriiodothyronemia model but are also in line with previous studies (Hamidi et al., 2010;Denereaz and Medicine, 2014). Notably, T3-treated animals displayed signs of improved AD-related depression-like behavior in this model, which was accompanied by the downregulation of genes related to the hippocampal serotonergic system. Nevertheless, there were no detectable alterations in expression patterns of genes related to the GABAergic system and neurotrophins. in the brain of patients with AD (Sathya et al., 2012). The T3-treated animals had also reduced ADAM10 gene expression levels. Once translated, this metalloproteinase plays an important role in glutamatergic synapse modulation (Marcello et al., 2017). Indeed, several studies have shown that ADAM10 expression is reduced in patients with moderate to severe AD (Colciaghi et al., 2004;Kim et al., 2009) and it has been speculated that the attenuation ADAM10 expression would consequently increase N-cadherin levels, ultimately resulting in more AMPA receptors and larger, more stable synapses (Malinverno et al., 2010). Moreover, a recent review showed that T3 increases N-cadherin levels in stem-cells repressing EGF-EGFR and cAMP-PKA signaling (Izaguirre and Casco, 2016). The Pearson correlation analyses detected a positive association between ADAM10 gene expression and the behavioral parameters of the tail suspension test, as well as with serum T3 and T4 levels. Thus, the results from the present study indicate that T3-mediated ADAM10 attenuation, increases N-cadherin levels, and likely, resulting in better glutamatergic synapses and improved depressive behavior.
Another important aspect that needs to be discussed is mitochondrial dysfunction in AD. Extensive studies have demonstrated mitochondrial degradation by autophagy and impaired oxidative phosphorylation in the brains of AD patients that is induced by reactive APP and β-amyloid (Perez Ortiz and Swerdlow, 2019). These deleterious effects are also accompanied by impaired mitochondrial metabolism and increased reactive oxygen species production (Zhao and Zhao, 2013). Further, mitochondrial SOD1 deficiency has been shown to induce amyloid β protein oligomerization and memory loss, whereas SOD2 overexpression has decreased hippocampal oxidative damage (Murakami et al., 2011). Increased SOD2 expression also prevented memory deficits in the Tg2576 AD mouse model (Massaad et al., 2009). Notably, in the present study, T3 treatment significantly increased the hippocampal SOD2 expression, which could slow AD progression. GSK3β also plays a key role in the regulation of mitochondrial function and the observed downregulation of GSK3β induced by T3 treatment could represent another mechanism for reducing the impacts of mitochondrial dysfunction in AD.
A recent study revealed that T3 supplementation was capable of reversing mnemonic deficits and the progression of neuroinflammation in hypothyroid rats. In that study, T3 replacement reduced the expression of Aβ42 peptide and proinflammatory markers, including TNFα (Chaalal et al., 2019). Furthermore, GSKβ activation induces NF-κB signaling, consequently increasing the transcription of TNFα and other pro-inflammatory cytokines (Maixner and Weng, 2013). These factors promote neuroinflammation by activating JNK and p38 MAPK signaling cascades. Taking together the hippocampal GSK3β activity. Fluoxetine treatment also reduces APP cleavage and Aβ generation, thus suggesting that serotonergic targets could be exploited in both AD prevention and treatment (Huang et al., 2018).
In this context, the modulation of 5-HT receptors, especially 5-HT2A, 5-HT2C, and 5-HT4 receptors, may be beneficial for AD management (Thathiah and De Strooper, 2011). The beneficial effects of 5-HT receptor modulation in promoting the non-amyloidogenic pathway appears to be mediated by a direct interaction with α-secretase ADAM10 or β-secretase BACE1 (Cochet et al., 2013). Additionally, a recent review proposed that the role of 5HT1A, in terms of synaptic plasticity, changes with aging, transitioning from a stimulated to inhibited state (Duda et al., 2018). This same review also provided evidence for serotonin-mediated inhibition of GSK3β expression, via 5HT1 receptors (Duda et al., 2018). Moreover, 5-HT1A receptor activation was shown to increase the inhibitory serine-phosphorylation of GSK3β (Li et al., 2004). On the other hand, the dopaminergic system can reduce GSK3β serine-phosphorylation through D2 receptor activation, which is due to the inactivation of Akt by protein phosphatase-2 in a β-arrestin2-driven protein complex (Beaulieu et al., 2005). In a mechanism similar to β-arrestin2 recruitment, 5-HT2A receptor activation reduces serine-phosphorylation of GSK3, thereby increasing its activity (Li et al., 2004). Consistent with these observations, GSK3β inhibitors have been shown to reduce depressive-like behavior in animal models of depression (Duda et al., 2018).
In our study, the T3-treated group exhibited a significant reduction in GSK3β and ADAM10 gene expression. This is particularly relevant to AD as GSK3 is a kinase that contributes to the abnormal phosphorylation during tau proteinmediated microtubule stabilization, a process that leads to neurofibrillary tangle formation (Beurel et al., 2015). Studies have shown that β-amyloid peptide formation results in abnormal APP processing and leads to the induction of tau phosphorylation and GSK3 activation (Kuruva and Reddy, 2017). Based on these assumptions, GK3β inhibitors have great potential for treating AD. Indeed, in different AD animal models, different GSKβ inhibitors promoted lower levels of tau phosphorylation and amyloid deposition, rescuing neuronal loss and restoring memory deficits (Serenó et al., 2009;Griebel et al., 2019). Notably, our results identified a strong correlation between 5HT1A, GSK3β and ADAM10 gene expression, as well as with serum T3 and T4 concentrations. Thus, the T3-induced GSK3 and ADAM 10 downregulation may be indirectly mediated by the serotonergic system.
Herein, the expression of BACE1 is decreased in T3-treated animals. This aspartyl protease is responsible for the amyloidogenic cleavage of APP and it is increased presubiculum, are the least resistant to the effects of several stressors (Grieve et al. 2005;Kalpouzos et al., 2009;La Joie et al., 2010). In the current study, we have performed the gene expression in the whole hippocampus and we did not have additional data to provide the information about what cells in the sub-regions of the hippocampus were affected by T3 treatment, which is a limitation of this study. In our study, we have performed the expression of AD-related and behavior-related genes, however we did not carry out the protein analysis by Western Blotting and immunohistochemistry to evaluate the alteration in the target genes, which is also a limitation of this study and that will still require further investigation.
In conclusion, our data suggest that T3 therapy improves the depression-like behavior observed in an AD transgenic mouse model and that these effects are mediated through the modulation of the serotonergic related genes involved in the transmission mediated by 5HT1A receptors and serotonin reuptake The benefits of T3 treatment are not only restricted with the results from the present study, the T3-induced ADrelated gene downregulation may be indirectly modulated by low levels of GSK3β and TNFα. In Fig. 3, we summarize all possible T3 actions that could slow the AD progression (Fig. 4).
The hippocampus is constituted of several histologically defined and interconnected subfields including parasubiculum, presubiculum, subiculum, Cornu Ammonis (CA1, 3, and 4) (Boccardi et al. 2019). The CA1 sub-region is particularly sensitive to hypoxia, ischemia, and hyperglycemia while CA3 and dentate gyrus are usually spared (de Flores et al. 2015b, a). In vivo MRI studies demonstrated that subiculum has been consistently found to be the most vulnerable substructure to age and neurodegeneration in the hippocampus (Frisoni et al. 2008). This vulnerability is supported by the developmental theory, according to which the last cerebral structures to develop, such as subiculum and Fig. 4 Graphic representation of possible thyroid hormone actions in the hippocampus neurons and its modulation in the expression of AD-related and behavior-related genes. In the presynaptic neuron, the interaction between T3 and thyroid receptors results in binding to TREs and culminates in the negative effect in SERT and 5-HT1A expression. The downregulation of these genes implies in low somatodendritic regulation and low 5-HT reuptake, inducing a higher serotonergic transmission. It is possible that other presynaptic 5-HT auto-receptors could be modulated by T3, such as: 5-HT1B and 5-HT1D. In the postsynaptic neuron, it is possible that the 5-HT4 receptor activation and the inactivation of 5-HT6 receptors would promote a decreased neuroinflammation and better cognitive performance, respectively. Moreover, 5-HT1a receptor activation increase the inhibitory serine-phosphorylation of GSK3β by PI3K/AKT activation and AMPc reduction. The GSK3β inhibition has a negative effect in AD-related genes. Further studies will be needed to determine whether the genomic actions of T3 could directly reduce AD-related genes expression. In addition, the modulation of mTOR by non-genomic actions of thyroid hormones through the integrin receptor αvβ3 increases mitochondrial autophagy. Concomitantly, increase in SOD2 expression may decrease oxidative stress activity and apoptosis. to improving AD-related depressive behavior but are also applicable to disease progression. Further studies will be necessary to assess the impact of T3 treatment in order to reduce neuroinflammation and how T3 reduces depression and downregulates AD-related genes.