Effect of Unloaded and Curcumin-Loaded Solid Lipid Nanoparticles on Tissue Transglutaminase Isoforms Expression Levels in an Experimental Model of Alzheimer’s Disease

Alzheimer’s disease (AD) is a neurodegenerative disease representing the most prevalent cause of dementia. It is also related to the aberrant amyloid-beta (Aβ) protein deposition in the brain. Since oxidative stress is involved in AD, there is a possible role of antioxidants present in the effected person’s diet. Thus, we assessed the effect of the systemic administration of solid lipid nanoparticles (SLNs) to facilitate curcumin (CUR) delivery on TG2 isoform expression levels in Wild Type (WT) and in TgCRND8 (Tg) mice. An experimental model of AD, which expresses two mutated human amyloid precursor protein (APP) genes, was used. Behavioral studies were also performed to evaluate the improvement of cognitive performance and memory function induced by all treatments. The expression levels of Bcl-2, Cyclin-D1, and caspase-3 cleavage were evaluated as well. In this research, for the first time, we demonstrated that the systemic administration of SLNs-CUR, both in WT and in Tg mice, allows one to differently modulate TG2 isoforms, which act either on apoptotic pathway activation or on the ability of the protein to repair cellular damage in the brains of Tg mice. In this study, we also suggest that SLNs-CUR could be an innovative tool for the treatment of AD.


Introduction
Alzheimer's disease (AD) is a neurodegenerative disease that represents the most prevalent cause of dementia, affecting about 50 million people in the world. It is characterized by cognitive dysfunction, progressive memory loss, language disturbance, and mood-behavioral changes [1]. AD is associated with synaptic failure, oxidative stress, neuroinflammation, and mitochondrial dysfunction [2]. It is also related to the aberrant amyloid-beta (Aβ) protein deposition in the brain, which is a gradual process involving the post-translational modification responsible for fibrillar inclusions and aggregates in the brain [3]. Aβ causes oxidative stress, inflammation, neurotoxicity and the deterioration of the neurotransmission system [4,5]. In addition, it is a substrate of tissue transglutaminase (TG2), a ubiquitarian calcium-dependent protein, which is involved in physiological and pathological processes, including AD [6,7]. TG2 is distributed in many types of tissues, including brain, and is mainly localized in the cytoplasmic compartment of neurons, in 2 of 12 nuclei [8], and the extracellular matrix [9]. Depending on its intracellular localization, TG2 plays different functions: in the cytosol, the protein is involved in the apoptotic processes in a stimuli-dependent manner [10]; when it is localized in the nucleus, it phosphorylates some proteins essential for its TG2 kinase activity [11]. In addition, TG2 shows two isoforms, short TG2 (TG2-S) and long TG2 (TG2-L), which have different cellular localizations [12][13][14]. In particular, TG2-S is localized both in the cytosol and in the mitochondria, and its expression levels increase during apoptosis [12,13]. TG2-L is present in the nucleus, playing a protective role during cellular injury and/or apoptosis [12,13].
Our previous findings highlighted that the exposure of olfactory ensheathing cells (OECs)-a particular glial cell type present in the olfactory system-to Aβ(1-42) full native peptide, and its fragments, induced an increase in TG2 expression levels and a different expression pattern of its isoforms [15]. Furthermore, we found that indicaxanthin pretreatment, a phytochemical with antiproliferative and anti-inflammatory activities [16], was able to reduce TG2 overexpression. This would also modulate the TG2 isoform expression and stimulate the reparative role exerted by the protein. We further reported that Aβ induced TG2 conformation change from "open" to "close" status in OECs, thus reversing the effect induced by Aβ. These data show that indicaxanthin pretreatment might stimulate neural regeneration in AD.
Therefore, the involvement of oxidative stress in AD suggests a possible role of antioxidants present in the diet of the effected person [17]. In particular, curcumin (CUR), a natural antioxidant phenolic compound extracted from the rhizome of Curcuma longa, and its derivatives, represent promising active compounds for the treatment of AD [18][19][20][21][22].
Several findings demonstrated that CUR shows strong neuro-protective effects improving cognitive function; promoting axonal regeneration [23][24][25][26][27][28]; reducing apoptotic proteins; upregulating the anti-apoptotic protein Bcl-2; and also stimulating neurogenesis and neuroplasticity. In addition, recent data demonstrated that CUR is able to downregulate TG2 overexpression in mouse microglial cells (exposed to lipopolysaccharide or Aβ 1-42 full peptide), thus suggesting its possible use for neuroprotection in neuroinflammatory pathologies, including AD [29]. In vitro and in vivo studies on AD models also demonstrated that CUR was able to inhibit the formation of Aβ oligomers [30,31]. However, in vivo use of CUR is limited because of its low bioavailability in the brain, poor absorption, and rapid systemic clearance [27]. To overcome the unfavorable features of this interesting natural compound, and allow its application for AD treatment, innovative strategies are required [32,33]. An emerging approach to this is the use of nanotechnologies. In particular, solid lipid nanoparticles (SLNs) possess unique physicochemical properties and are able to cross the blood-brain barrier (BBB), representing a promising strategy for the treatment of nervous system disorders, including AD [34]. Notably, in our previous studies, it was demonstrated that encapsulated CUR in SLNs increased its bioavailability and also its residence time in the brain. Furthermore, since the use of SLNs through parenteral administration is limited by the opsonization process (which has a short half-life of approximately 3-5 min), stealth nanoparticles have been realized to avoid the defense line represented by the macrophages. Specifically, stealth SLNs are colloidal carriers made from solid lipids with a modified surface in order to prolong their systemic residence time, and achieve a therapeutic drug concentration at the level of the central nervous system (CNS). This stealth strategy was obtained using a flexible hydrophilic polymer, such as polyethylene glycol (PEG) [35]. Many studies have proved that PEGylation not only stabilizes the SLNs, but also modulates their kinetics of cellular absorption [36,37].
Herein, we assessed the effect of the systemic administration of unloaded and CURloaded SLNs on the expression levels of TG2-L and TG2-S in transgenic, hemizygous, CRND8 (Tg) mice (which harbor a double-mutant gene of APP695 with a (C57)/(C57/C3H) genetic background), and in non-Tg hybrid (C57)/(C57/C3H) wild type (WT), control, littermate mice [38]. Empty and SLNs-CUR were formulated and characterized according to the method of Puglia et al. [34]. Behavioral studies were also performed to evaluate the improvement of cognitive performance and memory functions, which were induced by all of the treatments. Furthermore, in order to verify if the treatment with unloaded and CUR-loaded SLNs were able to counteract neuronal death, and the role played by TG2 isoforms, the expression levels of Bcl-2 (a mitochondrial protein with anti-apoptotic activity) was detected. Cyclin-D 1 , a protein involved in the cell progression cycle, and caspase-3 cleavage (the executioner caspase in apoptosis), were also tested.

SLNs Characterization
Dynamic light scattering (DLS) and electrophoretic light scattering (ELS) were used for measuring average size (Z-Ave), polydispersity index (PDI), and zeta potential (ZP, ξ), respectively. Analyses were performed at 20 • C using a Zeta Sizer Nano-ZS90 (Malvern Instrument Ltd., Worcs, UK), while using a solid-state laser with a nominal power of 4.5 mW and a maximum output of 5 mW. Each measurement was performed, at least, in triplicate.

Animals and Treatment
WT and Tg mice (20-30 g) were maintained at 12 h light/dark room at 21-23 • C with tap water and standard chow provided ad libitum. Behavioral experiments and treatments were performed in accordance with the European (2010/63/EC) and National (DL 26/2014, Permit number 931/2020-PR) guidelines in order to minimize the suffering of animals. Four-month-old WT and Tg mice (n = 5/group/genotype) were used for the treatment. The animals were divided into four groups and intraperitoneally (i.p.) injected for 3 consecutive weeks, three times a week: (i) WT mice treated with SLNs (n = 5 WT, 0.1 mL/10 g); (ii) WT treated with SLNs-CUR (n = 5, WT 150 mg/kg); (iii) Tg mice treated with SLNs (n = 5 Tg, 0.1 mL/10 g); (iv) Tg treated with SLNs-CUR (n = 5 Tg, 150 mg/kg). Additionally, the CTR groups for both WT and Tg mice, treated with saline solution, were used; these results were not reported in Figure 1 because they did not differ from the SLNs CTR groups. The animals were weighed once a day, the body weight recorded, and the dose to be administered i.p. was recalculated according to the body weight.

Behavioral Experiment: Step down Inhibitory Passive Avoidance Task
At the end of the treatment, the memory performance was investigated via a frequently used behavioral "Step down test". This is based on the animal learning regarding the exploration of a compartment with a foot shock delivered through the floor grid [40]. After exposure to the same environment, the mice will avoid going down, or increase the latency, before going down onto the grid. The word "passive avoidance" indicates in vivo experiments in which the animal learns to avoid a specific event, typically an innate behavior, which is dangerous due to a linked punishment [41]. An animal, when it is located in an open field, spontaneously employs most of its time in dark places, particularly near walls and corners. When it is placed on an elevated platform in the center of a field, the animal will try to leave the platform, explore the environment, and move toward walls.
Based on this natural behavior, 60 different variants of the step-down inhibitory avoidance task have been used [42,43]. The apparatus consists of a 50 × 25 × 25 cm Plexiglas box with a steel grid floor through which electric shocks (20 mA, 50 Hz) were delivered. A Plexiglas platform (5 × 8 × 25 cm) was also fixed in the center of the grid floor. For the training test (TT; day 1), each mouse was placed on the platform. When a mouse stepped down with all four paws on the grid floor, intermittent electric shocks were delivered continuously for 3 s. Shock parameters, such as animal vocalization, were recorded in order to select the mice to be used in the next test, i.e., the retention test (RT). Twenty-four hours after training, each mouse was placed on the platform again and the step-down latencies were compared between the TT and RT. An upper cut-off time of 180 s was set. If RT latency was longer than that of the TT it meant that the mouse learnt and remembered the punishment received the day before; on the contrary, however, the mouse showed a memory deficit. At the end of the assay, all of the mice were sacrificed by cervical dislocation and the brains were quickly removed and stored at −80 • C for in vitro experiments. Step-down inhibitory avoidance test, the training test (black bars) showed no significant differences between groups. The 24 h retention test (red bars) showed increased latencies in control (WT-SLNs, * p < 0.05) and SLNs-CUR-treated WT and Tg mice, when compared to their respective training latencies and to the retention latencies of SLNs Tg mice (*** p < 0.001). In untreated Tg mice, retention latencies were not significantly different from training latencies. (B) Animal weight during 3 weeks experimental time: SLNs and SLNs-CUR did not affect body weight of WT and Tg mice.

Behavioral Experiment: Step down Inhibitory Passive Avoidance Task
At the end of the treatment, the memory performance was investigated via a frequently used behavioral "Step down test". This is based on the animal learning regarding the exploration of a compartment with a foot shock delivered through the floor grid [40]. After exposure to the same environment, the mice will avoid going down, or increase the latency, before going down onto the grid. The word "passive avoidance" indicates in vivo experiments in which the animal learns to avoid a specific event, typically an innate behavior, which is dangerous due to a linked punishment [41]. An animal, when it is located in an open field, spontaneously employs most of its time in dark places, particularly near walls and corners. When it is placed on an elevated platform in the center of a field, the animal will try to leave the platform, explore the environment, and move toward walls. Based on this natural behavior, 60 different variants of the step-down inhibitory avoidance task have been used [42,43]. The apparatus consists of a 50 × 25 × 25 cm Plexiglas box with a steel grid floor through which electric shocks (20 mA, 50 Hz) were delivered. A Plexiglas platform (5 × 8 × 25 cm) was also fixed in the center of the grid floor. For the training test (TT; day 1), each mouse was placed on the platform. When a mouse stepped down with all four paws on the grid floor, intermittent electric shocks were delivered continuously for 3 s. Shock parameters, such as animal vocalization, were recorded in order to select the mice to be used in the next test, i.e., the retention test (RT). Twenty-four hours after training, each mouse was placed on the platform again and the step-down Step-down inhibitory avoidance test, the training test (black bars) showed no significant differences between groups. The 24 h retention test (red bars) showed increased latencies in control (WT-SLNs, * p < 0.05) and SLNs-CUR-treated WT and Tg mice, when compared to their respective training latencies and to the retention latencies of SLNs Tg mice (*** p < 0.001). In untreated Tg mice, retention latencies were not significantly different from training latencies. (B) Animal weight during 3 weeks experimental time: SLNs and SLNs-CUR did not affect body weight of WT and Tg mice.

Statistical Analysis
One-way ANOVA with Bonferroni's post hoc test was used to analyze step down inhibitory avoidance data, and Holm-Sidak's post hoc test was used to analyze in vitro experiments. Data are reported as mean ± SD of four separated experiments in duplicate, and differences between groups were considered to be significant at * p < 0.05.

SLNs Characterization
SLNs-CUR were formulated following a validated method and all components used for the formulation were found to be safe [39]. DLS data showed technological parameters suitable for parenteral administration: polydispersity index value (PDI) was 0.26; mean diameter (Z-Ave) ranging from 150 to 180 nm; and zeta potential (ZP) values were in the range of −29 to −24 mV ( Table 1). As previously reported [39], this latter value predicts a good long-term stability. Table 1. The values of Z-Ave, PDI, and Z-potential for unloaded and CUR-SLN recorded at 20 • C.

Behavioral Experiments
WT and Tg mice were systemically i.p. administered for 3 weeks, three times a week, with unloaded SLNs or SLNs-CUR (150 mg/kg).
We found that both empty SLNs and SLNs-CUR were well tolerated, no evident side effects were revealed and no animal died. To evaluate potential effects of SLNs-CUR on cognitive functions at the end of treatment, the animals were tested for behavioral performance in the step-down test and the data are reported in Figure 1A. To exclude potential SLNs and SLNs-CUR toxicity, the body weight of the animals were recorded during the period of treatment shown in Figure 1B. In the step-down inhibitory avoidance test, no significant differences were observed among all groups during the training test. In the 24 h RT, the step-down latencies recorded with SLNs-Tg mice were significantly reduced in respect to WT groups (*** p < 0.001). SLNs-CUR treatment to Tg mice significantly improved their performance (*** p < 0.001, vs. SLNs-Tg mice) to levels comparable to those displayed by WT-SLNs and WT-SLN-CUR mice.
This set of experiments highlighted that the systemic administration in Tg mice of SLNs-CUR induced a significant improvement of cognitive performance and memory function, showing that CUR loaded into SLNs was able to explicate a protective and/or reparative function in the brain in an AD mouse model. Figure 2 shows a representative immunoblot and densitometric analysis related to TG2 isoform expression levels, these were performed in the total cell lysates of WT-SLNs, WT-SLNs-CUR, Tg-SLNs mice, and Tg-SLNs-CUR-treated mice after 3 weeks of systemic administration [12,13,15]. No significant changes in the TG2 isoform expressions, both in WT-SLNs and Tg-SLNs-CUR mice, were observed. In Tg-SLNs mice, TG2-L expression levels (molecular weight 77-78 kDa) appeared at lower levels compared to those observed in WT-SLNs ones. The systemic administration of Tg-SLNs-CUR mice induced a significant increase in TG2-L levels. In contrast, TG2-S (molecular weight 62-63 kDa) was found at low levels in WT-SLNs mice, whereas it was expressed at high levels in Tg-SLNs. Systemic administration of Tg mice with SLNs-CUR enabled an increase in TG2-L expression which appeared at similar values to those found in WT-SLNs-CUR. In parallel, it reduced TG2-S expression at the same levels to those observed in the WT-SLNs-CUR-treated mice. Densit- TG2-S expression at the same levels to those observed in the WT-SLNs-CUR-treated mice. Densitometric analysis, performed after normalization with α-tubulin, confirmed all the results.

Effect of SLNs and SLNs-CUR on TG2 Isoform Expressions
These data highlighted that systemic administration of SLNs-CUR, both in the WT and in the Tg mice, differently modulates TG2 isoforms acting on apoptotic pathway activation or on the ability of the protein to repair cellular damage in the brain of Tg mice.

Cyclin-D1 Expression Levels
To assess the role played by TG2 in the cellular brain repair, induced by 3 weeks of systemic administration to WT-SLNs, WT-SLNs-CUR, Tg-SLNs mice, and Tg-SLNs-CURtreated mice, Cyclin-D1 protein levels were analyzed through Western Blot (Figure 3). These data highlighted that systemic administration of SLNs-CUR, both in the WT and in the Tg mice, differently modulates TG2 isoforms acting on apoptotic pathway activation or on the ability of the protein to repair cellular damage in the brain of Tg mice.

Cyclin-D 1 Expression Levels
To assess the role played by TG2 in the cellular brain repair, induced by 3 weeks of systemic administration to WT-SLNs, WT-SLNs-CUR, Tg-SLNs mice, and Tg-SLNs-CURtreated mice, Cyclin-D 1 protein levels were analyzed through Western Blot (Figure 3).
A slight, but significant, difference between WT-SLNs and WT-SLNs-CUR mice in Cyclin-D 1 expression levels was observed. This result highlighted that the systemic administration of the animals with SLNs did not induce any toxic effect and that the presence of CUR, stimulating cellular proliferation, might ameliorate brain functions in control mice. In Tg-SLNs mice, Cyclin-D 1 expression appeared at low levels. SLNs-CUR systemic administration to Tg mice induced a strong increase in Cyclin-D 1 expression levels, when compared with untreated ones, that were also at similar levels to those found in SLN-WT mice. Data were confirmed by densitometric analysis performed after normalization with α-tubulin.
This set of experiments demonstrated that the systemic administration of CUR-loaded SLNs was able to stimulate TG2 repair activity, activating the stem self-renewal in the brain through the increase in Cyclin-D 1 expression levels. compared with untreated ones, that were also at similar levels to those found in S mice. Data were confirmed by densitometric analysis performed after normalizati α-tubulin.
This set of experiments demonstrated that the systemic administration o loaded SLNs was able to stimulate TG2 repair activity, activating the stem self-ren the brain through the increase in Cyclin-D1 expression levels.

Bcl-2 Expression Levels
To verify if SLNs-CUR systemic administration in Tg mice was able to res aberrant death stimulus, the impairment of the mitochondria functions and the cor with TG2 over expression [46], Bcl-2 expression levels were assessed. Figure 4 shows Western blot and densitometric analysis for Bcl-2, performed cellular lysates from WT-SLNs, WT-SLNs-CUR, Tg-SLNs, and SLNs-CUR-Tg mice. Notably, Bcl-2 expression levels appeared at low levels in Tg mice.
No significant changes between WT-SLNs and WT-SLNs-CUR mice in Bcl-2 sion levels were found. This data showed that the systemic administration of SL

Bcl-2 Expression Levels
To verify if SLNs-CUR systemic administration in Tg mice was able to restore the aberrant death stimulus, the impairment of the mitochondria functions and the correlation with TG2 over expression [46], Bcl-2 expression levels were assessed. Figure 4 shows Western blot and densitometric analysis for Bcl-2, performed on total cellular lysates from WT-SLNs, WT-SLNs-CUR, Tg-SLNs, and SLNs-CUR-Tg-treated mice. Notably, Bcl-2 expression levels appeared at low levels in Tg mice.
No significant changes between WT-SLNs and WT-SLNs-CUR mice in Bcl-2 expression levels were found. This data showed that the systemic administration of SLNs did not induce mitochondrial damage. In Tg-SLNs mice, Bcl-2 expression appeared at low levels. The administration of SLNs-CUR in Tg mice induced a strong increase in Bcl-2 expression levels that, when compared with untreated ones, were also similar to those found in WT-SLNs mice. Data were confirmed by densitometric analysis, which was performed after normalization with α-tubulin.
This set of experiments highlighted that treatment with SLNs-CUR was able to restore the mitochondrial functions modified in Tg mice. expression levels that, when compared with untreated ones, were also similar found in WT-SLNs mice. Data were confirmed by densitometric analysis, which w formed after normalization with α-tubulin.
This set of experiments highlighted that treatment with SLNs-CUR was ab store the mitochondrial functions modified in Tg mice.

Caspase-3 Cleavage
To verify the TG2-mediated apoptotic pathway in Tg-SLNs and in Tg-SL mice, caspase-3 cleavage through Western Blot analysis was performed and eval total cell lysates from WT-SLNs, Tg-SLNs, and Tg-SLNs-CUR mice ( Figure 5). No cant change in caspase-3 cleavage, both in WT-SLNs-treated and WT-SLNs-CU was observed. This result highlighted that the systemic administration of SLNs activate the apoptotic pathway, thus confirming that they are not toxic. High l caspase-3 cleavage in Tg-SLNs mice were found, however. Moreover, systemic istration of Tg-SLNs-CUR induced a strong decrease in caspase-3 cleavage wh pared with those observed in the Tg-SLNs mice, even if its levels appeared slightl than those found in the WT-SLNs-CUR-treated mice.
These findings highlighted the role played by TG2 in the control of apopto way activation in the brain of mice treated with unloaded-SLNs and loaded-SLN

Caspase-3 Cleavage
To verify the TG2-mediated apoptotic pathway in Tg-SLNs and in Tg-SLNs-CUR mice, caspase-3 cleavage through Western Blot analysis was performed and evaluated to total cell lysates from WT-SLNs, Tg-SLNs, and Tg-SLNs-CUR mice ( Figure 5). No significant change in caspase-3 cleavage, both in WT-SLNs-treated and WT-SLNs-CUR mice, was observed. This result highlighted that the systemic administration of SLNs did not activate the apoptotic pathway, thus confirming that they are not toxic. High levels of caspase-3 cleavage in Tg-SLNs mice were found, however. Moreover, systemic administration of Tg-SLNs-CUR induced a strong decrease in caspase-3 cleavage when compared with those observed in the Tg-SLNs mice, even if its levels appeared slightly higher than those found in the WT-SLNs-CUR-treated mice.
These findings highlighted the role played by TG2 in the control of apoptotic pathway activation in the brain of mice treated with unloaded-SLNs and loaded-SLNs-CUR. Antioxidants 2022, 11, x FOR PEER REVIEW 9 of 12

Discussion
The aim of this study was to assess the effect of the systemic administration of unloaded and CUR-loaded SLNs on TG2-L and TG2-S expression in WT and Tg mice. In particular, Tg mice used for these studies expressed two mutated human APP genes, which are implicated in AD, thus showing a very rapid development of amyloid pathology due to the deposition of Aβ, which are responsible for amyloid plaques in the cerebral cortex and hippocampus, within three months of age [38]. To verify the improvement of cognitive performance and memory function induced by the treatments, behavioral studies were performed. In addition, the expression levels of Bcl-2, a mitochondrial protein with anti-apoptotic activity, were evaluated in order to verify if the treatment with SLNs-CUR was able to counteract neuronal death and the role played by TG2 isoforms. Cyclin-D1, a protein involved in the cell progression cycle, and caspase-3 cleavage, the executioner caspase in apoptosis, were also tested.
In this research, for the first time, we demonstrate that the systemic administration of SLNs-CUR, both in WT and in Tg mice, can differently modulate TG2 isoforms acting

Discussion
The aim of this study was to assess the effect of the systemic administration of unloaded and CUR-loaded SLNs on TG2-L and TG2-S expression in WT and Tg mice. In particular, Tg mice used for these studies expressed two mutated human APP genes, which are implicated in AD, thus showing a very rapid development of amyloid pathology due to the deposition of Aβ, which are responsible for amyloid plaques in the cerebral cortex and hippocampus, within three months of age [38]. To verify the improvement of cognitive performance and memory function induced by the treatments, behavioral studies were performed. In addition, the expression levels of Bcl-2, a mitochondrial protein with anti-apoptotic activity, were evaluated in order to verify if the treatment with SLNs-CUR was able to counteract neuronal death and the role played by TG2 isoforms. Cyclin-D 1 , a protein involved in the cell progression cycle, and caspase-3 cleavage, the executioner caspase in apoptosis, were also tested.
In this research, for the first time, we demonstrate that the systemic administration of SLNs-CUR, both in WT and in Tg mice, can differently modulate TG2 isoforms acting on apoptotic pathway activation or on the ability of the protein to repair cellular damage in the brains of Tg mice. In particular, systemic administration of Tg-SLNs-CUR mice induced a significant increase in TG2-L levels. TG2 isoforms play a role in the control of cell proliferation, the regulation gene expression, cell survival, and cellular differentiation [14,44,48]. In contrast, TG2-S, exerting transamidating activity and acting as an apoptotic factor [10,44], was found at low levels in WT-SLNs mice, whereas it was expressed at high levels in Tg-SLNs. To verify the protective role played by TG2-L in Tg mice (when administrated with SLNs-CUR Cyclin-D 1 ), a marker of cellular proliferation was evaluated [49]. In parallel, a significant reduction in TG2-S expression levels in Tg-SLNs-CUR-treated mice was observed. A significant increase in Bcl-2 expression accompanied by a decrease in caspase-3 cleavage was also found. These last results demonstrate that the treatment of Tg mice with SLNs-CUR for three weeks can reduce the formation of an alternative splice variant of TG2, which is responsible for apoptotic activation and cell death. Furthermore, in vivo data support our previous observations, performed on OEC cultures exposed to Aβ, both in the absence and in the presence of indicaxanthin [15]. These results also suggest that SLNs-CUR treatment in an AD experimental model could achieve TG2 conformation change from "open state" during the damage to "close state" which is involved in the stimulating of neural repair.
Behavioral results confirmed that the systemic administration of SNLs-CUR in Tg mice induced a significant improvement of cognitive performance and memory function, contributing also to the repair of cellular damage in transgenic mice.

Conclusions
Our results provide clear evidence that the systemic administration of curcuminloaded SLNs in TgCRND8 mice can reduce TG2-S expression, therefore decreasing apoptotic pathway activation. At the same time, it was able to increase TG2-L levels, which play the reparative role in an AD experimental model. Therefore, further studies are underway to better understand the effect of SLNs-CUR on AD. Moreover, the administration of this nanoformulation could represent a promising and innovative tool for the treatment of this pathology. Funding: This research was supported by the University of Catania, Progetto PIA.CE.RI 2020-2022, Dotazione Ordinaria, Linea di intervento 2, grant number 57722172121 "Effetto di antiossidanti naturali sull'overespressione della transglutaminase tissutale indotta da beta-amiloide: studi biologico-molecolari e computazionali" and grant number 57722172128 "NanoRET-Patologie neurodegenerative retiniche: nuovi approcci farmacologici e nanotecnologici".

Institutional Review Board Statement:
The study was conducted in accordance with the European (2010/63/EC) and the National Ministry of Health guidelines (DL 26/2014), and approved by the animal welfare body of the University of Florence (approval date: 21 July 2020) and National Ministry of Health (Permit number 931/2020-PR, approval date: 5 October 2020).

Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are available in the article.

Conflicts of Interest:
The authors declare no conflict of interest.