TNFAIP1 contributes to the neurotoxicity induced by Aβ25–35 in Neuro2a cells

Amyloid-beta (Aβ) accumulation is a hallmark of Alzheimer’s disease (AD) that can lead to neuronal dysfunction and apoptosis. Tumor necrosis factor, alpha-induced protein 1 (TNFAIP1) is an apoptotic protein that was robustly induced in the transgenic C. elegans AD brains. However, the roles of TNFAIP1 in AD have not been investigated. We found TNFAIP1 protein and mRNA levels were dramatically elevated in primary mouse cortical neurons and Neuro2a (N2a) cells exposed to Aβ25–35. Knockdown and overexpression of TNFAIP1 significantly attenuated and exacerbated Aβ25–35-induced neurotoxicity in N2a cells, respectively. Further studies showed that TNFAIP1 knockdown significantly blocked Aβ25–35-induced cleaved caspase 3, whereas TNFAIP1 overexpression enhanced Aβ25–35-induced cleaved caspase 3, suggesting that TNFAIP1 plays an important role in Aβ25–35-induced neuronal apoptosis. Moreover, we observed that TNFAIP1 was capable of inhibiting the levels of phosphorylated Akt and CREB, and also anti-apoptotic protein Bcl-2. TNFAIP1 overexpression enhanced the inhibitory effect of Aβ25–35 on the levels of p-CREB and Bcl-2, while TNFAIP1 knockdown reversed Aβ25–35-induced attenuation in the levels of p-CREB and Bcl-2. These results suggested that TNFAIP1 contributes to Aβ25–35-induced neurotoxicity by attenuating Akt/CREB signaling pathway, and Bcl-2 expression.


Background
Alzheimer's disease (AD) is a chronic neurodegenerative disease that is characterized by the accumulation of amyloid-beta (Aβ) plaques, neurofibrillary tangles, and neuronal loss in various brain regions [1][2][3]. The 37-43 amino acid Aβ fragments in the brain are originally derived from the β-amyloid precursor protein (APP) via proteolytic processing by β-and γ-secretase [4]. Aβ  and Aβ 1-42 are two major neurotoxic Aβ fragments, which were mainly generated in the amyloidogenic processing pathway by the actin of β-and γ-secretase [4]. Aβ [25][26][27][28][29][30][31][32][33][34][35] is not naturally generated in the brain, instead it represents the neurotoxic fragment of Aβ  or Aβ  , and therefore is often used to mimic the neurotoxic role of Aβ  or Aβ  in experimental studies [5]. Despite increasing evidence showing that Aβ could cause morphological and biochemical characteristics of apoptosis such as chromatin condensation, DNA fragmentation, caspase activation and subsequent activation of apoptotic signaling pathways [6][7][8], the molecular mechanisms underlying the neurotoxic effect of Aβ have not been fully elucidated. Therefore, identification of key players and mechanisms of Aβ-induced neurotoxicity would benefit the development of novel therapeutic strategies for preventing and treating AD.
Tumor necrosis factor, alpha-induced protein 1 (TNFAIP1) was originally identified as a gene whose expression can be induced by the tumor necrosis factor alpha (TNFα) in umbilical vein endothelial cells [9]. TNFAIP1 gene is found to be an evolutionarily extremely conserved single-copy gene [9], implying that TNFAIP1 has an important physiological role, which is yet to be explored. TNFAIP1 has been demonstrated to interact directly with proliferating cell nuclear antigen (PCNA) and the small subunit (p50) of DNA polymerase δ, implying that it may be involved in DNA synthesis or DNA repair [10,11]. Kim et al. [12] found that RhoB induces apoptosis by interacting with TNFAIP1 via a JNK-mediated signaling mechanism, suggesting that TNFAIP1 is an apoptosis-related protein. In addition, the transcription levels of TNFAIP1 had been found to be robustly induced in the transgenic C. elegans AD brains and post-mortem AD brain [13,14], suggesting TNFAIP1 may also involve in the process of AD development. Moreover, a recent study implied that estrogen may affect hippocampalrelated diseases by regulating TNFAIP1 [15]. However, the role of TNFAIP1 in AD has not been demonstrated.
In the present study, we examined the roles of TNFAIP1 in Aβ 25-35 -induced apoptosis in neuronal cell line by testing whether the neuronal apoptosis induced by Aβ [25][26][27][28][29][30][31][32][33][34][35] is associated with the expression of TNFAIP1 protein, and if so, whether apoptosis can be blocked by inhibition of TNFAIP1 expression using TNFAIP1 siRNA. In addition, to further clarify the signal transduction pathways involved in the neurotoxicity induced by Aβ, we also examined the potential signal transduction pathways involved in the apoptosis induced by TNFAIP1.

Cell culture and transfection
Animal experiments were following protocols approved by the Ethic Committee of Hunan Normal University, and the Institutional Animal Care and Use Committee of Massachusetts General Hospital in compliance with the NIH Guide for the Care and Use of Laboratory Animals. Primary mouse cortical neurons were isolated form 15 day embryonic cortex obtained from pregnant C57BL/6 female mouse as described before [16]. The cell pellets were resuspended in neuron basal medium, supplemented with 2 % B27 supplement. Cells were seeded at a density of 3 × 10 5 cells/mL into 6 cm wells plate precoated with poly-d-lysine. Medium was half changed every 4 days. All experiments were performed on cultures at days 7-9 in vitro.

Cell viability assay
Cell viability was measured by MTT assay according to manufacturer's protocol. N2a cells were grown on 24-wells plates to approximately 70 % confluence and then transfected with TNFAIP1 siRNA or Myc-TNFAIP1 for 24 h, followed by treatment of Aβ 25-35 for another 24 h. After treatments, cells were rinsed with PBS and replaced with fresh DMEM containing MTT. The cells were then incubated at 37 °C for 4 h under 5 % CO 2 /95 % air. After the medium was removed, DMSO was added into each well, the absorbance at 570 nm was measured on a microplate reader. Percentages of live cell counts were used for assay normalization.

Real-time PCR
The RT-PCR primers for mouse TNFAIP1 and 18sRNA were purchased from SABiosciences. Total RNA from N2a cells was isolated with RNeasy Lipid Tissue Mini Kit (Qiagen) according to the manufacturer's instructions. Then cDNA was synthesized with SuperScript system (Invitrogen). The mRNA levels of TNFAIP1 and 18sRNA were measured by quantitative RT-PCR using SYBR green kit (Applied Biosystems) in an ABI 7000 real-time PCR system (Applied Biosystems). The PCR conditions were as follows: Initial denaturation of DNA, 95 °C for 5 min; denaturation, 32 cycles of 95 °C for 35 s; annealing, 60 °C for 35 s; extension, 72 °C for 35 s; and final extension, 72 °C for 5 min. Data were analyzed according to the comparative threshold cycle method with expression for sample normalization. RT-PCR assay was performed in triplicate for each sample to ensure reproducibility.

Statistical analysis
Data were expressed as mean ± SD. Three to five separate experiments were performed. Data were analyzed using ANOVA with Tukey post hoc tests (SPSS version 18.0). Statistical significance was at p < 0.05.

TNFAIP1 contributes to Aβ 25-35 -induced apoptosis
Previous study suggested that TNFAIP1 is an apoptosisrelated protein [12], implying that TNFAIP1 may also involve in Aβ 25-35 -induced apoptosis. Cleaved caspase-3 is a pivotal executioner and hallmark of apoptosis that is usually activated in the apoptotic cell by both extrinsic and intrinsic pathways. To further determine whether involvement of TNFAIP1 in Aβ 25-35 -induced neuronal apoptosis is associated with changes in the activities of caspases, cleaved caspase-3 was detected by Western blot. As shown in Fig. 3a, b, N2a cells treated with 20 μM of Aβ 25-35 led to a significant increase in the level of cleaved caspase-3, while Aβ 25-35 -induced caspase-3 cleavage was significantly repressed by the knockdown of TNFAIP1. In contrast, 20 μM of Aβ 25-35 treatment or Myc-TNFAIP1 transfection or Aβ 25-35 +Myc-TNFAIP1 co-treatment exhibited a significant increase in the cleavage of caspase-3 (Fig. 3c, d). These results indicate that TNFAIP1 contributes to Aβ 25-35 -induced apoptosis.

TNFAIP1 inhibits Akt/CREB signaling pathway
It has been reported that Akt/CREB signaling plays neuroprotective roles [18] and Aβ could downregulate Akt survival pathway [19]. In agreement with previous studies, our results showed that Aβ 25-35 causes a decrease in Akt phosphorylation (phosphor-Ser473), CREB phosphorylation (phosphor-Ser133-CREB) and Bcl-2 expression in a dose-dependent manner (Fig. 4a). To further explore whether involvement of TNFAIP1 in neuronal apoptosis was associated with changes in Akt/ CREB signaling pathway, we overexpressed TNFAIP1 in N2a cells in dose-dependent manner and then assessed Akt phosphorylation and CREB phosphorylation (Fig. 4a). As expected, overexpression of TNFAIP1 in N2a cells led to significant attenuation of p-Akt and p-CREB in a dose-dependent manner (Fig. 4b-d). Importantly, we also observed that overexpression of TNFAIP1 could significantly reduce the protein levels of Bcl-2 (Fig. 4e), which is an important neuroprotectant and transactivated by CREB in response to Aβ stimulation [20]. Therefore, these results indicate that TNFAIP1 may mediate Aβ 25-35 -induced apoptotic signaling by attenuating p-Akt, p-CREB and Bcl-2 expression.

Discussion
Deposition of Aβ in the brain is a pathological hallmark of AD as it is responsible for progressive neurodegeneration in AD [22]. Accumulating evidences indicated that Aβ could cause neurotoxicity by inducing neuronal apoptosis in vitro and in vivo [23,24]. Elucidating the molecular mechanisms underlying Aβ-induced neuronal apoptosis will help us on developing treatment of AD. However, despite intense research, it remains unclear how Aβ triggers the signaling cascade that results in neuronal dysfunction and neurotoxicity. In the present study, our results demonstrated that TNFAIP1 protein is induced by Aβ [25][26][27][28][29][30][31][32][33][34][35] and involved in Aβ 25-35 -induced neuronal apoptotic pathway because overexpression of TNFAIP1 can accelerate Aβ 25-35 -induced apoptosis while inhibition of TNFAIP1 can significantly reduce The transcript levels of TNFAIP1 was found to be robustly induced in the transgenic C. elegans AD brains and post-mortem AD brain in previous study [12,13], which led us to examine the hypothesis that TNFAIP1 was involved in the pathological development of AD using an in vitro mouse AD model: mouse primary cortical neurons and N2a neuroblastoma cells treated by Aβ [25][26][27][28][29][30][31][32][33][34][35] . Our results suggested that Aβ 25-35 could induce TNFAIP1 protein and mRNA levels in a dosedependent manner. Furthermore, overexpression or knockdown of TNFAIP1 exacerbated or alleviated Aβ 25-35 -induced neurotoxicity. Together, these results indicated that the TNFAIP1 gene expression may be critical for Aβ 25-35 -induced neuronal death and intervention of its expression could be a potential method for inhibiting Aβ 25-35 -induced neurotoxicity. However, the regulatory mechanisms of TNFAIP1 remain largely unclear. Only a recent report described that transcription factor Sp1 could bind to human TNFAIP1 promoter region and transactivate TNFAIP1 promoter [25]. As TNFAIP1 had been originally found to be induced by TNFα [9], which is a potent activator of NFκB, thus it is possible that TNFAIP1 was also regulated by NFκB when exposed to Aβ [25][26][27][28][29][30][31][32][33][34][35] . Further studies will be performed to elucidate whether inhibition of NFκB activation can reduce Aβ 25-35 -induced TNFAIP1 gene expression.
Previous studies described that TNFAIP1 was a proapoptotic protein [12], and apoptosis is a general neuronal death pathway in neurodegenerative diseases which could be triggered by exposure to Aβ. In the present study, we extended our observations to in vitro AD models that TNFAIP1 was also involved in Aβ 25-35 -induced apoptosis. We found that inhibition of TNFAIP1 gene by specific siRNA significantly attenuated Aβ 25-35 induced cleaved-caspase 3. As caspase-3 cleavage is a central event in executing Aβ 25-35 -induced neuronal apoptosis [26,27], these results provide further evidence that TNFAIP1 was involved in Aβ 25-35 -induced apoptosis. However, it is worth noting that Aβ 25-35 +Myc-TNFAIP1 co-treatment did not cause a much higher increase in the cleavage of caspase-3 than Aβ 25-35 treatment or Myc-TNFAIP1 transfection alone. It is possible that 20 μM of It had been demonstrated that CREB is directly phosphorylated and regulated by the protein kinase Akt [28], and Akt/CREB signal has been shown to play a pivotal role in neuroprotection by enhancing cell survival and inhibiting apoptosis [18]. Importantly, Akt and CREB could promote neuronal cell survival by upregulating Bcl-2 protein levels [29]. As expected, TNFAIP1 could reduce p-Akt, p-CREB and Bcl-2 protein levels in a dosedependent manner. Moreover, we observed overexpression of TNFAIP1 could further attenuate Aβ-inhibited p-CREB and Bcl-2 protein expression, whereas inhibition of TNFAIP1 could significantly reverse the effect of Aβ on p-CREB and Bcl-2. These results suggest that TNFAIP1 may mediate the deactivation of CREB induced by Aβ, and then trigger neuronal apoptosis. MAPKs had been reported to be involved in Aβ-induced apoptosis, and MAPKs signaling cascades were activated in brains from AD patients [30]. Moreover, Aβ-induced MAPKs include ERK1/2, JNK and p38 MAPK had been demonstrated to be associated with neuronal cell death [31,32]. Therefore, TNFAIP1 may also modulate MAPKs signaling cascades involved in Aβ-induced neuronal apoptosis. Further study will be conducted to elucidate the role . c Quantitative analysis of p-CREB protein levels (fold of control). Data are expressed as mean ± SD, n = 3, *p < 0.05, **p < 0.01 compared with Myc-TNFAIP1 (0 μg). d Quantitative analysis of p-Akt protein levels (fold of control). Data are expressed as mean ± SD, n = 3, *p < 0.05 compared with Myc-TNFAIP1 (0 μg). e Quantitative analysis of Bcl-2 protein levels (fold of control). Data are expressed as mean ± SD, n = 3, *p < 0.05 compared with Myc-TNFAIP1 (0 μg) of TNFAIP1 in modulating MAPKs signaling pathway involved in the apoptotic effect of Aβ.

Conclusions
In summary, this study clearly demonstrated that TNFAIP1 was significantly upregulated by Aβ [25][26][27][28][29][30][31][32][33][34][35] in mouse primary cortical neurons and N2a cells, and TNFAIP1 may be a key player that mediated Aβ 25-35induced neurotoxicity by inactivating the Akt/CREB signaling pathway, and in turn downregulating anti-apoptotic protein Bcl-2. Our findings implied that TNFAIP1 may be a potential therapeutic target for treatment of AD, but the roles and mechanism of TNFAIP1 in in vivo AD models need to be further elucidated in our future study.