Autophagy is Activated In Vivo during Trimethyltin-Induced Apoptotic Neurodegeneration: A Study in the Rat Hippocampus

Trimethyltin (TMT) is an organotin compound known to produce significant and selective neuronal degeneration and reactive astrogliosis in the rodent central nervous system. Autophagy is the main cellular mechanism for degrading and recycling protein aggregates and damaged organelles, which in different stress conditions, such as starvation, generally improves cell survival. Autophagy is documented in several pathologic conditions, including neurodegenerative diseases. This study aimed to investigate the autophagy and apoptosis signaling pathways in hippocampal neurons of TMT-treated (Wistar) rats to explore molecular mechanisms involved in toxicant-induced neuronal injury. The microtubule-associated protein light chain (LC3, autophagosome marker) and sequestosome1 (SQSTM1/p62) (substrate of autophagy-mediated degradation) expressions were examined by Western blotting at different time points after intoxication. The results demonstrate that the LC3 II/I ratio significantly increased at 3 and 5 days, and that p62 levels significantly decreased at 7 and 14 days. Immunofluorescence images of LC3/neuronal nuclear antigen (NeuN) showed numerous strongly positive LC3 neurons throughout the hippocampus at 3 and 5 days. The terminal deoxynucleotidyltransferase dUTP nick end labeling (TUNEL) assay indicated an increase in apoptotic cells starting from 5 days after treatment. In order to clarify apoptotic pathway, immunofluorescence images of apoptosis-inducing factor (AIF)/NeuN did not show nuclear translocation of AIF in neurons. Increased expression of cleaved Caspase-3 was revealed at 5–14 days in all hippocampal regions by Western blotting and immunohistochemistry analyses. These data clearly demonstrate that TMT intoxication induces a marked increase in both autophagy and caspase-dependent apoptosis, and that autophagy occurring just before apoptosis could have a potential role in neuronal loss in this experimental model of neurodegeneration.


Introduction
TMT is an organotin compound known to produce significant and selective neuronal degeneration in the rodent central nervous system. In particular, it induces a selective loss of pyramidal neurons, especially in the CA1 and CA3/CA4 hippocampal subfields [1,2]. TMT is considered a useful tool to study the molecular

Autophagy is Activated in the Hippocampus of TMT-Treated Rats
To determine the effects of TMT on autophagy activation in the rat hippocampus, we analyzed the expression of adaptor proteins: LC3 (I and II, respectively the constitutive and active forms) and p62, which conjugates autophagy substrates with LC3 II.
The conversion of LC3-I to LC3-II (LC3-II/LC3-I) was calculated in order to estimate autophagy flux after Western blot analysis of the single forms (I and II) at 3, 5, 7, and 14 days after treatment. The ratio of LC3II/I increased significantly already from 3 days (p < 0.05) after treatment and the peak was observed at 5 days (p < 0.001), while from Day 7 it gradually decreased and at 14 days almost returned to the control value ( Figure 1a). Trimethyltin (TMT) increased LC3II/LC3I, reduced SQSTM1/p62 and Beclin-1 expression levels, and did not affect ATG5 protein in the rat hippocampus after treatment. Graphic presentations and Western blotting images of LC3II/LC3I (a), SQSTM1/p62 (b), Beclin-1 (c), and ATG5 (d) proteins are shown. Values are presented as mean ± standard error of the mean (SEM) for each group: the control rats (n = 3) and rats treated for 3, 5, 7, and 14 days (n = 3/group), * p < 0.05, ** p < 0.01 and *** p < 0.001 compared with controls, Dunnett's test. Ctrl: control sample.
In order to visualize the distribution and cellular localization of LC3 in the hippocampus of control and TMT-treated rats, double-labeling immunofluorescence experiments using an antibody to LC3 (which does not distinguish LC3-I from LC3-II) in combination with NeuN marker were performed. We observed a diffuse and marked LC3 immunoreactivity (red) at 3 and 5 days after TMT-intoxication in the CA1 (Figure 2e (3,5,7, and 14 days) are labeled for Neu N (green, a,d,g,j,m), LC3 (red, b,e,h,k,n), and Neu N/LC3 (merge, c,f,i,l,o). Control sections (a-c). At 3 days after treatment, marked LC3 immunoreactivity was observed (e) and LC3 labeling strongly increased at 5 days (h). LC3 staining decreased at 7 (k) and 14 days (n) post intoxication. Note the progressive loss of Neu N-positive neurons in TMT-treated rats (d,g,j,m) compared with control rats (a). Arrows and inserts show details. Scale bar: 100 µm. Ctrl: control sample, d: days. For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article. Control sections (a-c). At 3 days after treatment, marked LC3 immunoreactivity was observed (e) and LC3 labeling strongly increased at 5 days (h). LC3 staining decreased at 7 (k) and 14 days (n) post intoxication. Note the progressive loss of Neu N-positive neurons in TMT-treated rats (d,g,j,m) compared with control rats (a). Arrows and inserts show details. Scale bar: 100 µm. Ctrl: control sample, d: days. For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.  (3,5,7, and 14 days) are labeled for Neu N (green, a,d,g,j,m), LC3 (red, b,e,h,k,n), and Neu N/LC3 (merge, c,f,i,l,o). Control sections (a-c). At 3 days after treatment, marked LC3 immunoreactivity was observed (e) and LC3 labeling strongly increased at 5 days (h). LC3 staining decreased at 7 (k) and 14 days (n) post intoxication. Non-relevant LC3 reactivity in the Dentate Gyrus (*) was observed at any time of treatment (e,h,k,n). Note the progressive loss of Neu N-positive neurons in TMT-treated rats (d,g,j,m) compared with control rats (a). Arrows and inserts show details. Scale bar: 100 µm. Ctrl: control sample, d: days. For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.
Interestingly, a progressive loss of NeuN-positive cells in TMT-treated rats (Figures 2, 3 and 4d,g,j,m) compared with control animals (Figures 2, 3 and 4a) was observed.
Western blotting analysis showed levels of p62 protein that gradually decreased starting from 3 days after TMT treatment, with differences that became statistically significant at 7 (p < 0.01) and 14 days (p < 0.001) compared with the control value ( Figure 1b).

TMT Reduces Beclin-1 Levels and Has No Effect on Atg5 Expression
In order to understand better the results of LC3 and p62 expression levels obtained previously and examine how the autophagy system is regulated in the hippocampus in response to TMT-toxicity, we targeted upstream regulatory protein markers Beclin-1 and Atg5. Our results revealed that the levels of Beclin-1 gradually decreased from 3 to 14 days following intoxication, becoming statistically significant at 7 (p < 0.001) and 14 days (p < 0.01) compared with the control group ( Figure 1c). Regarding Atg5, we showed that this protein was not significantly affected by TMT, as its expression levels were similar to those of the controls throughout ( Figure 1d).

Apoptotic Cells Increase after TMT-Treatment
TUNEL assay was performed to stain DNA strand breaks in the nuclei and examine apoptosis [46] in the hippocampus of the TMT-treated rats. TUNEL-positive nuclei were stained light green, in the CA1, CA3, and CA4 areas, in the control animals and 3, 5, 7, and 14 days after treatment (Figure 5a  Values are presented as mean±SEM for each group: the control rats (n = 3) and 3, 5, 7, and 14 days TMT-treated rats (n = 3/group), * p < 0.05, ** p < 0.01, *** p < 0.001 compared with controls, Dunnett's test.

TMT Did Not Induce AIF Nuclear Translocation but Activated Caspase-3 Expression
In order to examine whether TMT induced apoptosis in a Caspase-3-independent manner, we investigated AIF protein translocation from the cytoplasm to the nucleus in rat hippocampal neurons. Double-label immunofluorescence experiments were performed, using an anti-AIF antibody in combination with NeuN neuronal nuclear marker on sections of control rats and rats treated for 3, 5, 7, and 14 days. We observed no localization of AIF in the neuronal nuclei at any time of treatment, while intense immunoreactivity of apoptotic factor AIF (red) co-localized with the cytoplasm of numerous neurons (green) of all hippocampal regions, such as the CA1 (Figure 6), for the different treatment times (3, 5, and 7 days), as shown (Figure 6f,i,l).
No AIF translocation to the nuclei was observed in control sections (Figure 6c). In order to verify Caspase-3-dependent apoptosis, cleaved Caspase-3 expression was then assessed by investigating the presence of a 17 kDa band corresponding to the active form of the protease at 3, 5, 7, and 14 days after TMT treatment, using Western blot analysis. The cleaved Caspase-3 expression levels were significantly higher than the control value at 5 (p < 0.01), 7 (p < 0.01), and 14 (p < 0.001) days after intoxication (Figure 7a). In parallel, we analyzed cleaved Caspase-3 localization in the hippocampus, using DABimmunohistochemistry experiments. A marked immunoreactivity of cleaved Caspase-3 was observed in the cytoplasm of some cells localized in the CA1, CA3, and CA4 areas at 5 and 7 days (Figure 8d-i) and of numerous cells at 14 days after TMT exposure in the same regions (Figure 8j-l) in accordance with blotting data. Caspase-3-positive cells not only displayed different immunoreactivity levels but also showed evidence of degeneration. We observed no staining for cleaved Caspase-3 at 3 days in any subfields analyzed, probably because our experimental method is not sufficiently sensitive enough to detect low levels (as shows Western blotting analysis) of this protease.
No cleaved Caspase-3 labeling was observed in any of the analyzed hippocampal areas of control rat sections (Figure 8a-c).

Cytoplasmic Expression of Cytochrome c Increases after TMT-Treatment
We detected expression of cytoplasmic Cytochrome c in the rat hippocampus by Western blotting analysis during TMT-induced neurodegeneration because the release of Cytochrome c from the mitochondrial membrane space to the cytosol represents a critical event in activation of the intrinsic apoptotic pathway.
Cytochrome c immunoreactivity was evident as a single 14-kDa band on the cytoplasmic fractions from the hippocampus homogenates at 3, 5, 7, and 14 days after treatment. Cytoplasmic Cytochrome c expression levels showed a gradual and significant increase compared with the control value, starting from 3 (p < 0.01) up to 14 days (p < 0.001) after exposure to TMT (Figure 7b).

Discussion
The present study demonstrates, for the first time, that TMT treatment, in vivo, in the rat hippocampus, induces activation of autophagy. The neurotoxicant initially activates autophagy, followed immediately by Caspase-3-dependent apoptosis, as evidenced by the analysis of dynamic changes in the expression of the main molecular markers.
Among the proteins involved in regulation of the autophagic process, LC3 is the most known and studied. LC3 is a soluble protein distributed ubiquitously in mammalian tissues and cultured cells and is considered the major marker of the autophagy pathway. It is cleaved into the LC3-I cytosolic constitutive form and is then conjugated to PE to form active LC3-II. LC3-II is recruited to autophagosomal membranes [47], indicating the number of autophagosome vacuoles formed. In particular, LC3-I is more abundant and readily detected in brain tissue than LC3-II [48]. It is noteworthy in this respect that our double labeling immunofluorescence (LC3/NeuN) images evidenced numerous LC3-positive neurons, likely characterized by an elevated autophagosome number, at 3 and 5 days after TMT treatment, in the CA1, CA3, and CA4 areas (Figures 2-4).
We demonstrated that the levels of the LC3-II/LC3-I ratio were significantly increased in the rat hippocampus at 3 and 5 days following intoxication compared with the basal level of the control animals ( Figure 1a). Notably, the LC3-II/LC3-I ratio is considered a hallmark of the degree of autophagy activation [49]. These results indicate that TMT action triggers the conversion of LC3-I to LC3-II in the rat hippocampal neurons. Decreased LC3-II/LC3-I values, observed at later treatment times (7 and 14 days) (Figure 1a), suggested both delipidation and recycling of LC3-II to LC3-I (with consequent decrease of LC3-II levels) in the autolysosomal lumen by lysosomal hydrolases [50] and, partly, the effect of a progressive TMT-induced neuronal loss in all the hippocampal regions, especially detected at 14 days. The LC3-II/LC3-I expression levels are in accordance with LC3 reactivity, localized in the hippocampal neurons, which the immunofluorescence images show (Figures 2-4).
Changes in the levels of the p62 protein are also used as an indicator of autophagy flux, since p62 is a substrate that is degraded via autophagic pathway [51]. Increased p62 levels that indicate autophagosome-lysosome fusion step impairment (and consequent absence of the cargo degradation) have been observed in some neurodegenerative disorders characterized by accumulation of protein aggregates, including Lewy bodies (in Parkinson's disease), neurofibrillary tangles (in Alzheimer's disease), Huntingtin aggregates (in Huntington's disease), and Mallory bodies (in alcoholic and nonalcoholic steatohepatitis) [52][53][54]. On the contrary, in our model, the results showed a decrease in the p62 values starting from 3 days, which became statistically significant at 7 and 14 compared to the value of the control (Figure 1b), indicating that active autophagosome-lysosome fusion (forming autolysosome) occurs with the consequent degradation of the protein. These data suggest that autophagy is activated by TMT in the hippocampal neurons already at early treatment times.
In order to understand the molecular pathway that regulates autophagy, we also examined the upstream protein Beclin-1. We found that TMT does not stimulate an increase in Beclin-1 expression, but on the contrary induces a gradual decrease of this protein with significant levels at 7 and 14 days (Figure 1c). This result is unexpected. Numerous studies have demonstrated that autophagy activation is generally related to increase of Beclin-1 expression during neurodegeneration process [55][56][57][58]. The decrease of the Beclin-1 levels that we found, however, may merely reflect the involvement, and consequent degradation, of the protein during increased autophagy in our neurodegeneration model. Beclin-1 has also be reported not to be required for the induction of autophagy through the formation of new vacuoles in some experimental models. For example, staurosporine, resveratrol, and H 2 O 2 have been shown to induce a Beclin-1-independent non-canonical autophagy pathway in different experimental models [59][60][61].
Since examples of Beclin-1-independent but Atg5-dependent autophagy activation have been discussed [59,62], we subsequently explored an autophagy regulation pathway alternative to that of Beclin-1, involving Atg5 protein. Our results showed no significant difference in the levels of Atg5 during the treatment in comparison with the control value (Figure 1d). A change in steady-state levels of Atg5 may therefore not necessarily be required for the induction of autophagy in our model of neurodegeneration.
In this study, we also wanted to further investigate apoptosis triggered by TMT, in order to relate temporally autophagic activity with cellular death, since these processes are often both involved in a wide variety of neurodegenerative diseases. It should be noted that the TMT neurodegeneration shares a pathophysiological characteristic common to other neurodegenerative diseases such as the Alzheimer's and temporal lobe epilepsy. Thus, TMT-induced neurotoxicity models might be useful tools in researching hippocampal lesions, recovery after brain injury, and specific mechanisms of human neurodegenerative diseases [63].
Much evidence suggests an association between TMT neurodegeneration models and apoptotic neuronal death both in vitro and in vivo. Based on in vitro experiments, Gunasekar and coworkers [41] concluded that low concentrations of TMT (0.01-0.1 µmol/l) cause apoptotic death in cerebellar granule cells. It has also been shown that the incubation with TMT triggers time-/dose-dependent apoptosis in an immortalized hippocampal neuronal cell line (HT-22 cells), which is Caspase-3 mediated [64]. Geloso and collaborators [42] demonstrated, in vivo, that TMT induces apoptotic cell death in adult mice, accompanied by the expression both of cleaved Caspase-3 and of COX-2 in the degenerating hippocampal granular cells.
We showed TUNEL-stained neuronal nuclei to be significantly increased in the CA1, CA3, and CA4 areas at 5 days and, gradually, to peak at 14 days after exposure to the neurotoxicant ( Figure 5). In contrast, other studies on rats treated with TMT showed a number of TUNEL-positive neurons in the CA1 and CA3 subfields that peaked in one case at 5 days after TMT gavage (9 mg/kg body weight) [65] and, in another case, at 7 days in the CA3 region after intraperitoneal treatment (7 mg/kg) [66]. This discrepancy could be attributed to the different methods of administration and TMT doses in different experimental settings.
The analysis in time course of the LC3 and p62 expression levels with TUNEL labeling shows a temporal correlation between autophagy and apoptosis activation in TMT-treated rat hippocampal neurons. An autophagic process was activated just before apoptotic neuronal death in early treatment times, and continued to remain activated in late times. On the other hand, we were unable to identify a possible co-localization of the two processes through LC3/TUNEL (or Caspase-3) immunofluorescence double labeling experiments. Further studies, possibly using autophagy inhibitors or inducers, in the TMT-incubated hippocampal cultures, will be needed in order to clarify this aspect.
We also investigated, for the first time, the caspase-independent apoptosis pathway in the TMT-induced neurodegeneration process, where the involvement of the AIF protein has been well established [67,68]. In normal conditions this protein is generally localized in the intermembrane space of mitochondria, where it is involved in oxidative phosphorylation and energy production [69]. In the damaged neurons, it is released to the cytoplasm in response to specific death signals and translocates to the nucleus, where it induces apoptosis [70,71]. The double-label immunofluorescence images for AIF and NeuN showed no nuclear localization of AIF in hippocampal neurons in any of the areas analyzed at the different time points after treatment ( Figure 6 for the CA1 area), thus excluding the possible involvement and activation of a caspase-independent apoptotic pathway in the TMT-injured neurons.
We therefore subsequently investigated whether Caspase-3 activation and apoptosis are associated in our experimental system, since this protease is known to be one of the major key regulators of apoptotic mechanisms [72]. Caspases exist as inactive proenzymes and are activated, upon apoptotic signals, after cleavage that can occur via both intrinsic and extrinsic pathways [73,74]. Interestingly, our Western blotting analysis demonstrated increased and significant levels of cleaved Caspase-3 at 5, 7, and 14 days after treatment (Figure 7a). These findings are in accordance with the immunohistochemistry images, which showed numerous cleaved Caspase-3-positive cells ( Figure 8) and with TUNEL assay data related to the number of apoptotic nuclei ( Figure 5) scattered in all hippocampal subfields at the same treatment times. Importantly, our data showing the expression and localization of Caspase-3 integrate those obtained in a study performed in vivo on TMT-exposed rat hippocampus at early (1-3 days) and late (22-24 days) post-treatment time points, where a marked cleaved Caspase-3 positivity only during the late stages of neuronal damage has been reported using immunohistochemistry [75].
To demonstrate the mitochondrial apoptotic pathway in the present in vivo model, we found the levels of cytoplasmic Cytochrome c to be over-expressed before (3 days) and during (5,7, and 14 days) Caspase-3 activation (Figure 7b). Cytochrome c is important to maintain cell viability when it is located in the intermembrane space of mitochondria, but when it is released from the damaged mitochondria it activates Caspase-3-dependent cell death [76]. A previous study, in vitro, demonstrated the release of Cytochrome c in the cytosol from damaged mitochondria in primary cultures of hippocampal neurons incubated with TMT [77]. In addition, another earlier study reported that mitochondrial functionality is impaired in PC12 cells after TMT incubation [78]. Evidently, the mitochondrial dysfunction is a prominent phenomenon following TMT treatment. The selective removal of damaged mitochondria through autophagy (indicated as mitophagy) would probably have a great impact in our neurodegeneration model for the protective role against the hippocampal neuron death. Increased accumulation of cytoplasmic Cytochrome c (Figure 7) indicates that, in the hippocampal neurons, there is a mitochondrial dysfunction and that, although there is increased autophagy (Figures 1-4), the mitophagy might be defective.
In summary, our study shows for the first time that, in vivo, TMT causes an increase in the LC3 II/LC3 I ratio and a decrease in p62 expression levels in the rat hippocampus, indicating, interestingly, the activation of autophagy during toxic-induced neurodegeneration. We also show that the mitochondrial signaling pathway is involved in TMT-induced apoptosis, including the expression of the Cytochrome c and cleaved Caspase-3. Furthermore, our results document the chronology of the dynamic changes of the main autophagic and apoptotic proteins, evidencing that autophagy is activated just before apoptosis and suggesting a potential role in TMT-induced neuronal loss. One might speculate that, in the early stages of treatment, increased autophagy can represent a protective response of the neurons to the stressful action of TMT. In the later stages, when cellular loss is increased, autophagy can be involved directly in neuronal death or may trigger apoptosis, in the same and/or in different neuron subpopulations. Finally, this study also offers information useful to design new therapeutic strategies based on autophagy modulation in order to prevent or delay neuronal death in the human neurodegenerative disease.

Animal Model
Adult female Wistar rats (200-250 g body weight, age 2 months) were used in this study. The rats were housed in an air-conditioned environment with a constant temperature (25 • C) and had food and water ad libitum on a standardized light/dark cycle. Hippocampal injury was induced by injecting TMT. The rats were divided: in a control group and in groups treated with TMT for 3, 5, 7, and 14 days. The treated rats received a single intraperitoneal injection of TMT (Sigma-Aldrich, St. Louis, MO, USA) (8 mg/kg body weight in a volume of 1 ml/kg body weight, dissolved in saline 0.9% NaCl). The control rats were administered an equal volume of saline solution. Successively, the animals were anesthetized intraperitoneally with 87 mg/kg body weight of Ketavet 100 (Merck Sharp & Dohme, New Jersey, USA) and 12.5 mg/kg body weight of Rompum (Xilazina) (Bayer, Leverkusen, Germany) before being sacrificed.
The animal protocol (No. 04-01) was reviewed and approved (October 2016) by the Animal Experimentation Committee of Catholic University, Rome. In particular, animal experiments were carried out in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC). Every effort was made to minimize the number of animals used and their suffering.

Western Blotting Experiments
Rats (control rats and rats treated for 3, 5, 7, and 14 days, 3 animals/group) were sacrificed by decapitation at various time points of treatment for the Western blotting analyses. The brains were removed, and the hippocampus was rapidly isolated on ice from the ipsilateral hemisphere and stored at −80 • C before being processed. Subsequently, the samples were mechanically homogenized in an ice-cold lysis buffer containing 50 mM Tris-HCL, pH 7.4, 150 mM NaCl, 1% Triton X-100, 0.1% sodium dodecyl sulfate (SDS), 0.5% sodium deoxycholate, and with a protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO, USA) added just before use. The lysates were incubated on ice for 30 min, sonicated, and then centrifuged at 14,000 g for 30 min at 4 • C. The supernatants were used to detect the expression of Beclin-1, Atg5, LC3, p62, and cleaved Caspase-3 proteins.
For Cytochrome c analysis, protein extraction of the cytosolic fractions was performed using a multiple centrifugation method as described elsewhere [65,79]. The hippocampal tissue was mechanically homogenized in a cold suspension buffer [20 mM hydroxyethyl piperazineethanesulfonic acid (HEPES)-KOH, pH 7.5, 250 mM sucrose, 10 mM KCl, 1.5 mM MgCl 2 , 1 mM ethylenediaminetetraacetic acid (EDTA), and 1 mM ethyleneglycotetraacetic acid (EGTA) plus 0.7% protease and phosphatase inhibitor cocktails (Sigma-Aldrich, St. Louis, MO, USA)]. The homogenate was centrifuged at 700 g for 10 min at 4 • C and then at 10,000 g for 20 min at 4 • C. The supernatant was further centrifuged at 100,000 g for 60 min at 4 • C and was then used for cytosolic analysis.
All Western blotting analyses were performed in at least three independent experiments.

Neuronal Death Assay: TUNEL
The TUNEL assay was performed using a fluorescein in situ cell death detection kit according to the manufacturer's protocol (In Situ Cell Death Detection Kit, Fluorescein, Roche, Germany).
The rats (control rats and rats treated for 3, 5, 7, and 14 days, 3 animals/group), after anesthesia, were perfused through the heart with 0.9% saline followed by 4% paraformaldehyde (PFA). The brains were removed, post-fixed in 4% PFA overnight at 4 • C, and cryoprotected in two steps overnight of 20 and 30% sucrose in phosphate buffer saline (PBS) pH 7.4 at 4 • C. Coronal sections were cut on a cryostat (Leica CM 1850) at a thickness of 10 µm. Before the TUNEL assay, the sections were incubated in acetone for 30 min at 4 • C, washed with PBS, placed in 0.5% Triton X-100 for 30 min at RT, and washed three times with PBS. Subsequently, sections were treated with 0.1 M citrate buffer pH 6.0, microwaved for epitope retrieval, and air-cooled. Briefly, 5 µL of enzyme solution (terminal deoxynucleotidyl transferase in storage buffer) were mixed with 45 µL of labeling solution (fluorescein-nucleotide mixture in reaction buffer). Each section was incubated with 50 µL of this TUNEL reaction solution at 37 • C for 60 min.
After further washing with PBS, sections were cover-slipped with a fluoromount aqueous mounting medium (Sigma-Aldrich, St. Louis, MO, USA), examined, and photographed under a Zeiss LSM 510 confocal laser scanning microscopy system.
Finally, we selected the entire hippocampus from each of three non-consecutive brain sections/time point and counted the number of TUNEL-positive neurons.

Immunofluorescence Localization of LC3 and AIF
The rats were sacrificed after anesthesia, and perfused transcardially with 4% PFA, at different time intervals (control and 3 Immunofluorescence images were analyzed and photographed under a Zeiss LSM 510 confocal laser scanning microscopy system.
To verify and confirm the specificity of the immunolabeling, primary antisera were omitted, and the sections were incubated only with secondary antibodies. No immunoreactivity was detected.
Five non-consecutive sections were processed for double label immunofluorescence from each animal/time point.

Immunohistochemistry for Cleaved Caspase-3
For cleaved Caspase-3 immunohistochemistry analysis, the brain tissues of rats (control rats and rats treated for 3, 5, 7, and 14 days, 3 animals/group) were processed as described above. Serial coronal sections of 10 µm were cut on a cryostat, mounted on slides, and air-dried. The slides, first, were incubated with acetone for 10 min at −20 • C to enhance antigenicity, and then permeabilized with 0.5% Triton X-100 in PBS overnight at 4 • C. After several rinses in PBS, nonspecific binding sites were blocked for 60 min in PBS containing 5% normal goat serum 0.5% Triton X-100 at RT. Subsequently, the sections were incubated with anti-cleaved Caspase-3 primary antibody (Cell Signaling Technology, Inc., Leiden, Netherlands), diluted 1:250 in PBS-T (PBS with 0.1% Tween 20) containing 5% normal goat serum and 0.5% Triton X-100, on a rotating stand overnight at 4 • C. After washing in PBS, biotinylated goat anti rabbit IgG secondary antibody (Vector Laboratories, Burlingame, CA, USA), diluted 1:200 in PBS and containing 0.3% Triton X-100, was added for 60 min at RT. The sections were then placed in an avidin-biotin peroxidase complex (1:100, Vector Laboratories, Burlingame, CA, USA), rinsed once in PBS, and incubated in a 3,3'-diaminobenzidine (DAB) peroxidase substrate kit as chromogen (Vector Laboratories, Burlingame, CA, USA). Finally, the sections were dehydrated and cover-slipped with a Eukitt mounting medium (Sigma-Aldrich, St. Louis, MO, USA). Cleaved Caspase-3 analysis was performed using a Zeiss Axiophot microscope, and the images were captured with an AxioCam 503 color camera coupled to the microscope.
The specificity of the labeling was confirmed by incubation of sections with only the secondary antibody, omitting the primary antibody. No immunoreactivity was observed under these conditions. Five non-consecutive sections were immunostained from each animal/time point.

Statistical Analysis
All statistical analyses were performed using the GraphPad software (GraphPad Prism version 6.01 for Windows, GraphPad Software, La Jolla California USA). The quantitative data were expressed as the mean±SEM. Differences between control groups and TMT-treated rats were assessed using one-way analysis of variance (ANOVA) followed by Dunnett's multiple comparison post-hoc test, assuming the levels of probability * p < 0.05, ** p < 0.01, and *** p < 0.001 as statistically significant.

Acknowledgments:
We would like to thank Enrico Guadagni and Roberto Passalacqua for their skillful technical support.

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