Mitochondrial Ferritin Deletion Exacerbates β-Amyloid-Induced Neurotoxicity in Mice

Mitochondrial ferritin (FtMt) is a mitochondrial iron storage protein which protects mitochondria from iron-induced oxidative damage. Our previous studies indicate that FtMt attenuates β-amyloid- and 6-hydroxydopamine-induced neurotoxicity in SH-SY5Y cells. To explore the protective effects of FtMt on β-amyloid-induced memory impairment and neuronal apoptosis and the mechanisms involved, 10-month-old wild-type and Ftmt knockout mice were infused intracerebroventricularly (ICV) with Aβ25–35 to establish an Alzheimer's disease model. Knockout of Ftmt significantly exacerbated Aβ25–35-induced learning and memory impairment. The Bcl-2/Bax ratio in mouse hippocampi was decreased and the levels of cleaved caspase-3 and PARP were increased. The number of neuronal cells undergoing apoptosis in the hippocampus was also increased in Ftmt knockout mice. In addition, the levels of L-ferritin and FPN1 in the hippocampus were raised, and the expression of TfR1 was decreased. Increased MDA levels were also detected in Ftmt knockout mice treated with Aβ25–35. In conclusion, this study demonstrated that the neurological impairment induced by Aβ25–35 was exacerbated in Ftmt knockout mice and that this may relate to increased levels of oxidative stress.

Oxidative stress has been strongly implicated in the pathophysiology of AD [12]. Increased free radicals can damage proteins, lipids, and nucleic acids. The combination of mitochondrial dysfunction and A accumulation generates reactive oxygen species (ROS) which, in the presence of metal ions such as Fe 2+ and Cu 2+ , may contribute to oxidative damage in AD brains [13]. In addition, dysregulated brain iron homeostasis also accelerates AD progression. Excessive iron in the brain can directly lead to the generation of free radicals that eventually cause neurodegenerative disease.
Mitochondrial ferritin (FtMt) is a recently identified ferritin that accumulates specifically in the mitochondria and possesses a high homology to H-ferritin [14]. The functions of FtMt include iron storage, regulating iron distribution between the cytosol and mitochondrial compartments, and preventing the production of ROS generated through the Fenton reaction [15,16]. FtMt has been reported to be present in relatively low abundance in the liver and splenocytes, 2 Oxidative Medicine and Cellular Longevity the major iron storage sites, while FtMt is found at higher levels in the testis, kidney, heart and brain, and tissues with high metabolic activity. Together with the knowledge that Hferritin expression confers an antioxidant effect, the tissue distribution of FtMt is in line with a protective function of FtMt in mitochondria against iron-dependent oxidative damage [17]. Previous results indicated that FtMt was involved in the pathogenesis of neurodegenerative diseases, including AD, Parkinson's disease, and Friedreich's ataxia [18][19][20]. Increased expression of FtMt has been observed in the brains of AD patients and is associated with the antioxidant role of this protein [2]. Furthermore, our previous studies have shown that FtMt exerted a neuroprotective effect against 6-hydroxydopamine-(6-OHDA-) induced dopaminergic cell damage [20] and FtMt overexpression attenuated A -induced neurotoxicity [11].
Although abnormal iron metabolism and oxidative stress have been reported in AD, little information is available about the role of FtMt in the pathogenesis of AD. In the present study, we investigated memory impairment and neuronal cell death in A 25-35 -injected Ftmt knockout mice. In addition, we explored the molecular mechanisms responsible for neuronal damage in this model. Our data indicate that FtMt deficiency exacerbated A 25-35 -induced neuronal cell damage by altering intracellular iron levels in a way that intensifies the oxidative stress caused by A 25-35 .

Materials and Methods
2.1. Animals. C57BL/6 Ftmt-null mice were obtained from The Jackson Laboratory [21]. Mice were housed under conditions controlled for temperature (22 ∘ C) and humidity (40%), using a 12 hr/12 hr light/dark cycle [22]. Mice were fed a standard rodent diet and water ad libitum. Age-matched C57BL/6J wild-type male mice and Ftmt knockout male mice (10 months) were used in this study. All procedures were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee of the Hebei Science and Technical Bureau in China.

Antibodies and Reagents.
The following antibodies and reagents were used: -actin (Alpha Diagnostic International, USA), TfR1 (Sigma-Aldrich, USA), FPN1, DMT1 (+IRE) and DMT1 (−IRE) (Alpha Diagnostic International, USA), Lferritin (Abcam Inc., SF, USA), cleaved PARP, caspase-3, phospho-p38 (p-p38) and p38 (Cell Signaling Technology, USA), Bcl-2 and Bax (Santa Cruz Biotechnology, USA), A 25-35 peptide (Sigma-Aldrich, USA), and TUNEL in situ Cell Death Detection Kit (Roche Diagnostics GmbH, Mannheim, Germany). [25][26][27][28][29][30][31][32][33][34][35] was dissolved in sterile saline and aggregated by incubation at 37 ∘ C for 4 days before use [23]. The aggregated form of A 25-35 (7.5 nmol in 5 L saline per injection) was injected into the right lateral ventricle as previously described [24]. Mice were randomly divided into four groups: wild-type with saline injection (WT + saline), wild-type with A 25-35 injection (WT + A 25-35 ), Ftmt knockout with saline injection (KO + saline), and Ftmt knockout with A 25-35 injection (KO + A [25][26][27][28][29][30][31][32][33][34][35] ). After injection, the mice were housed for 15 days under normal conditions and then trained and tested in a Morris water maze (MWM) as described below. (MWM Test). Spatial learning and memory deficits were assessed using the Morris water maze as described previously [25,26], with minor modification. The experimental apparatus consisted of a circular tank (diameter = 120 cm, height = 50 cm) that was divided into four quadrants, filled with water, and maintained at 22 ± 2 ∘ C. At first, a visible platform test was performed, which confirmed that there were no significant differences in sensory, motor, or motivational activities among these four groups. Then, hidden platform tests were performed in succession. For the hidden platform test, a round platform (diameter = 9 cm) was placed at the midpoint of the fourth quadrant, 1 cm below the water surface. The test was conducted four times a day for four days, with four randomized starting points. The position of the escape platform was kept constant. Each trial lasted for 90 s or ended as soon as the mice reached the submerged platform.

Probe Test.
To assess memory consolidation, a probe test was performed 24 h after the Morris water maze test [25]. For the probe test, the platform was removed and the mice were allowed to swim freely. The swimming pattern of every mouse was recorded for 90 s with a camera. Consolidated spatial memory was estimated by the time spent in the target quadrant area.

Assessment of Apoptosis.
After the behavioral testing, the animals were perfused with 0.9% saline under anesthesia with 0.4% Nembutal. The brains were immediately collected and then postfixed with 4% paraformaldehyde in 0.1 M phosphate buffer. Serial coronal sections were cut at 15 m on a freezing microtome (Leica CM1950, Leica Microsystems, Shanghai, China) and mounted onto slides covered with APES (Beijing Zhongshan Biotechnology, Beijing, China). The presence of apoptosis in the dentate gyrus of mouse hippocampi was assessed by the terminal deoxynucleotidyl transferase-mediated FITC-dUTP nick-end labeling method (TUNEL) following the manufacturer's protocol. Nuclei were counterstained with DAPI. The number of TUNEL-DAPIpositive cells was counted as described previously [27]. The counting area was located in the same position in all groups. For each group, quantification was performed in sections from three different mice.

Western Blot Analysis.
Protein expression was assessed by western blotting as previously described [28], with minor modifications. Briefly, hippocampi were homogenized and sonicated in RIPA buffer containing 1% NP40 and protease inhibitor cocktail tablets (Roche Diagnostics GmbH, Roche Applied Science, 68298 Mannheim, Germany). After centrifugation at 12,000 ×g for 20 min at 4 ∘ C, the supernatant was collected, and the whole cell lysate protein concentration was measured using the BCA Protein Quantification Kit (Yeasen Biotechnology, Shanghai, China). Protein from each sample (40 mg) was resolved by SDS-PAGE on 12% or 10% gels and then transferred to PVDF membranes. The blots were blocked in 5% nonfat milk containing 20 mM Tris-HCl (pH 7.6, 137 mM NaCl, and 0.1% Tween-20; TBS-T) for 1.5 h at room temperature, followed by incubation with primary antibody overnight at 4 ∘ C. After washing three times with TBS-T, the blots were incubated with horseradish peroxide (HRP) conjugated secondary antibody for 1.5 h at room temperature. Immunoreactive proteins were detected using the enhanced chemiluminescence (ECL) method and quantified by transmittance densitometry using volume integration with Multi Gauge ver. 3.1 software (FUJIFILM Corporation, Tokyo, Japan).

Measurement of MDA and SOD.
Malondialdehyde (MDA), a marker of lipid peroxidation, was assessed using the thiobarbituric acid (TBA) method [29] using a kit from the Nanjing Jiancheng Bioengineering Institute (Nanjing, China) according to the manufacturer's instructions. This method is based on the spectrophotometric measurement of the product of the reaction of TBA with MDA. MDA concentrations were then calculated by the absorbance of TBA reactive substances (TBARS) at 532 nm.
Superoxide dismutases (SODs), which catalyze the dismutation of superoxide into oxygen and hydrogen peroxide, were determined according to xanthine oxidase method using a commercial kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer's instructions. The xanthine-xanthine oxidase system produces superoxide ions, which can react with 2-(4-iodophenyl)-3-(4-nitrophenol-5-phenlyltetrazolium chloride) to form a red formazan dye, which can be detected by its absorbance at 550 nm [29].
The levels of MDA and the total SOD (T-SOD) activity were determined in each group. The hippocampi of mice were homogenized in ice-cold saline. The homogenate was centrifuged at 3000 ×g at 4 ∘ C for 15 min, and the supernatant was used to determine T-SOD activity and MDA levels with a spectrophotometer (Synergy H4, BioTek, USA) at wavelengths of 550 nm and 532 nm, respectively. Each group contained five mice for the MDA and SOD tests, with each test repeated three times.
2.9. Statistical Analysis. All data are expressed as the mean ± standard deviation. One-way analysis of variance was used to estimate overall significance and was followed by Tukey's post hoc test corrected for multiple comparisons. A probability level of 95% ( < 0.05) was considered significant. All the tests were performed with SPSS 21.0 (IBM SPSS21.0, Armonk, New York, United States).

Ftmt Ablation Exacerbates A 25-35 -Induced Spatial
Memory Deficits. The MWM test was conducted to assess learning and memory in 10-month-old wild-type mice (WT) and Ftmt knockout mice (KO). All mice were trained with four trials per day for 4 days. "Escape latency" is the time to reach the platform in the water maze and is used as a proxy for mouse memory. Compared to wild-type mice, Ftmt knockout mice took approximately the same time to reach the platform after training (Figure 1(a)). After the water maze test, we performed a probe test using the metric "time spent in quadrant" to investigate the maintenance of memory. The time spent in the target quadrant was also similar in wildtype and Ftmt knockout mice (Figure 1(b)). After treatment with A 25-35 , both the WT + A 25-35 group and the KO + A 25-35 group took a significantly longer time to reach the platform than the groups without A 25-35 injection (Figure 1(c)). Furthermore, A 25-35 -infused Ftmt knockout mice had a significantly greater memory impairment (longer escape latency time) than A 25-35 -infused wild-type mice. In addition, the time spent in the target quadrant in the probe trial was less in both the WT + A 25-35 and the KO + A 25-35 groups than in the control groups. Importantly, the KO + A [25][26][27][28][29][30][31][32][33][34][35] group was in the target quadrant for even less time than the WT + A 25-35 group (Figure 1(d)). Overall, our results show that knockout of Ftmt in mice significantly exacerbates memory deficits in the A 25-35 -induced AD model.

FtMt Deficiency in the AD Mouse Model Elevates Proapoptotic Signals.
We found that the knockout of Ftmt remarkably decreased the ratio of Bcl-2/Bax (Figure 3(a)) and increased the activation of cleaved caspase-3 (Figure 3(b)) after A 25-35 treatment in mice. In the apoptotic cascade, caspase-3 cleaves poly-ADP-ribose-polymerase (PARP), leading to the accumulation of an 89 kDa PARP fragment [30]. Caspase-3mediated PARP cleavage was enhanced in the KO + A 25-35 group compared to the WT + A 25-35 group (Figure 3(c)). These results indicate that the lack of FtMt can affect the Bcl-2/Bax ratio, leading to caspase-3 activation and a concomitant increase in PARP cleavage and, ultimately, apoptosis after A 25-35 injection.

Ftmt Knockout Increases MDA Levels in AD Mice without
Altering SOD. To determine whether increased levels of oxidative stress are responsible for the increased apoptosis in the hippocampus in the AD mouse model, we examined the levels of MDA and the activity of SOD in each group. Free radicals attack polyunsaturated fatty acids, leading to structural damage to membranes and the generation of MDA, which is considered a marker of lipid peroxidation and thus a surrogate for oxidative damage [32]. The level of MDA was increased in AD mice compared with controls, but this increase was significantly greater in Ftmt knockout mice (Figure 4(a)). SOD is a free radical scavenging enzyme that converts superoxide into H 2 O 2 . The content of total SOD was unchanged in the four groups (Figure 4(b)).

The Effects of Ftmt Knockout on the Levels of L-Ferritin, TfR1, DMT1, and FPN1.
Iron is an essential cofactor in many proteins, but excess free iron contributes to enhanced generation of ROS and oxidative stress [33]. When treated with A 25-35 , the levels of L-ferritin were upregulated while those of TfR1 decreased significantly, compared to the control groups. The highest amount of L-ferritin expression was observed in the KO + A 25-35 group (Figures 5(a) and 5(b)).  In addition, the content of L-ferritin was also increased in the KO + saline group when compared to the WT + Saline group (Figure 5(a)). These observations indicated that A 25-35 stimulation may lead to alterations in iron homeostasis and that FtMt deficiency may accelerate this process. In addition, alterations in cellular iron distribution (as detected by Perls' staining) (see Supplementary Figure 1 Figure 5(d)). These results suggest that injection of A [25][26][27][28][29][30][31][32][33][34][35] into the brain disturbed iron homeostasis, possibly leading to oxidative damage, both of which were exacerbated by the lack of FtMt.

Discussion
Iron is an essential trace element for human health. The metal participates in many biological processes. Iron homeostasis is stringently regulated in vivo as excess iron can catalyze the generation of oxidative damage [20]. Importantly, iron is considered a contributing neurotoxic factor in several neurodegenerative disorders, including AD [34]. Cortical iron elevation has been increasingly reported as a feature of AD [35] and may contribute to the oxidative damage observed in AD brains. In addition, abnormalities in ironregulatory proteins occur in the brains of AD sufferers [19]. in the hippocampus of AD patients [36]. In addition, the downregulation of FtMt causes severe neurodegeneration in the Purkinje cells of the cerebellum [20]. In this study, Ftmt gene knockout mice were firstly used to study the effects of FtMt on the behavioral changes and mechanisms of A 25-35induced neurotoxicity.
Previous results indicate that FtMt-deficient mice are healthy and do not show any evident phenotype under baseline feeding conditions [21]. Here we have also found that 10-month-old, wild-type, and Ftmt knockout mice show no behavior or memory differences, as determined by MWM assays. Thus, FtMt deficiency has no obvious effects in the mouse brain under normal physiological conditions. To further elucidate the role of FtMt in AD pathogenesis, we first showed that intracerebroventricular infusion of A [25][26][27][28][29][30][31][32][33][34][35] exacerbates memory impairment in Ftmt knockout mice compared to the A 25-35 -infused controls. The number of apoptotic cells in the hippocampus was also significantly increased in A 25-35 -infused Ftmt knockout mice, which may account for their poorer performance in the MWM. Our data suggest that FtMt is not essential in mice under normal conditions. However, when challenged, such as with amyloid beta treatment, there appears to be a need for FtMt in a neuroprotective role.
Bcl-2 and Bax play important roles in oxidative stressmediated neuronal apoptosis [37]. It has been reported that Bcl-2 protects neurons against oxidant stress and apoptosis in PD [38]. Bcl-2 also maintains mitochondrial integrity by blocking the release of apoptotic factors from mitochondria into cytoplasm [20]. Bax can promote cell death by activating elements of the caspase pathway [39], especially caspase-3 [40]. As previously described, the activation of caspases, a family of cysteine proteases, is a central mechanism in the apoptotic process. Our results show that knockout of Ftmt decreases the ratio of Bcl-2/Bax and increases the activation of caspase-3 and PARP cleavage, which ultimately leads to cell death.
Accumulated evidence demonstrated that A -induced neuronal injury triggers transcriptional and posttranscriptional processes that regulate neuronal fate, including the activation of the MAPK pathway [41]. In this signaling cascade, p38 MAPK is activated by phosphorylation, and a high p-p38/p38 ratio can simultaneously promote Bax expression and decrease Bcl-2 levels. Our results show a significant elevation of p-p38 levels and its downstream factor Bax, strongly suggesting that this apoptotic signal transduction pathway is enhanced in Ftmt knockout mice treated with A [25][26][27][28][29][30][31][32][33][34][35] .
An increasing number of studies have suggested that oxidative stress is associated with AD neurodegeneration and caspase-mediated apoptosis [42]. We detected a marked increase in the level of MDA, an indicator of oxidative damage, in the hippocampi of Ftmt knockout mice, indicating that knockout of Ftmt aggravates oxidative stress. Previous studies indicate that, in certain antioxidant systems, there might be a time lag between the synthesis of protein and the expression of mRNA following neurotoxicity; the activity of SOD is altered in the process of A 25-35 induced injury. Cu, Zn-SOD, and Mn-SOD activity in hippocampi of A 1-42 treated mice returned to near vehicle levels after 10 days [43]. Our data show that Ftmt ablation did not significantly affect the activity of total SOD, although this may be related to the time point at which SOD activity was measured.
Cellular iron homeostasis is maintained by a strict regulation of various proteins that are involved in iron uptake, export, storage, and utilization [44]. Studies from our group and others have demonstrated that aberrant iron homeostasis can generate ROS which can eventuate in AD pathogenesis [2,11]. In the present study, we observed upregulated Lferritin and FPN1 and a simultaneous decrease in TfR1. These changes are likely to be the result of inhibited iron-regulatory protein binding [45] brought about by an increase in the regulatory iron pool in the neuronal cells injected with A . Consistent with this, the absence of FtMt may decrease the cells ability to sequester excess iron under stressed conditions and enhances the degree of changes in the measured proteins of iron metabolism. Our previous data also indicate that "uncommitted" iron levels, commonly referred to as the "labile iron pool" (LIP), significantly increased in SH-SY5Y cells treated with A 25-35 . FtMt overexpression was able to reverse this change [11]. We propose that a larger LIP, resulting from a redistribution of iron from mitochondria to the cytosol, especially in the absence of FtMt, is responsible for the oxidative stress that mediates the damage of cell components in our AD model [46].
In summary, our research indicates that A 25-35 elevates the LIP and causes oxidative stress, which can be exacerbated by the lack of FtMt. The excess iron donates electrons to generate ROS and lipid peroxidation. These changes initiate the programmed cell death through the p38/MAPK pathway, ultimately causing neuronal apoptosis which causes more severe memory impairment. The alteration of iron maybe provides the feedback regulation to the levels of TfR1 and FPN1 ( Figure 6).  FtMt F er ri ti n ↑ FPN1 ↑ Figure 6: A schematic representation of the mechanism leading to neuronal cell apoptosis induced by A [25][26][27][28][29][30][31][32][33][34][35] in mice with a disrupted Ftmt gene. A 25-35 changes the levels (LIP) and distribution of intracellular iron, thus increasing oxidative stress. Without Ftmt to sequester excess mitochondrial iron, lipid peroxidation and the level of LIP are significantly increased. These changes may signal the cell to begin the process of programmed death through the P38 MAPK pathway, resulting in neuronal cell death, which is enhanced in Ftmt knockout mice, leading to worsened memory impairments.

Conclusion
The current study supports the hypothesis that the functionality of FtMt is not essential under normal conditions. But, in cases of neuronal stress, such as A 25-35 accumulation, FtMt offers a profound neuroprotection through regulating cellular iron content and distribution in a way that keeps oxidative stress in check, preventing the activation of apoptosis.