Upregulation of Thioredoxin-Interacting Protein in Brain of Amyloid-β Protein Precursor/Presenilin 1 Transgenic Mice and Amyloid-β Treated Neuronal Cells

Abstract

Oxidative stress has been hypothesized to play a role in the pathophysiology of Alzheimer’s disease (AD). Previously, we found that total nitrosylated protein levels were increased in the brain of amyloid-β protein precursor (AβPP) and presenilin 1 (PS1) double transgenic mice, an animal model for AD, suggesting that cysteine oxidative protein modification may contribute to this disease. Thioredoxin (Trx) is a major oxidoreductase that can reverse cysteine oxidative modifications such as sulfenylation and nitrosylation, and inhibit oxidative stress. Thioredoxin-interacting protein (Txnip) is an endogenous Trx inhibitor. To understand the involvement of Trx and Txnip in AD development, we investigated Trx and Txnip in the brain of AβPP/PS1 mice. Using immunoblotting analysis, we found that although Trx protein levels were not changed, Txnip protein levels were significantly increased in hippocampus and frontal cortex of 9- and 12-month-old AβPP/PS1 mice when compared to wild-type mice. Txnip protein levels were also increased by amyloid-β treatment in primary cultured mouse cerebral cortical neurons and HT22 mouse hippocampal cells. Using biotin switch and dimedone conjugation methods, we found that amyloid-β treatment increased protein nitrosylation and sulfenylation in HT22 cells. We also found that downregulation of Txnip, using CRISPR/Cas9 method in HT22 cells, attenuated amyloid-β-induced protein nitrosylation and sulfenylation. Our findings suggest that amyloid-β may increase Txnip levels, subsequently inhibiting Trx reducing capability and enhancing protein cysteine oxidative modification. Our findings also indicate that Txnip may be a potential target for the treatment of AD.

Keywords

Alzheimer’s disease amyloid-β nitrosylation oxidative stress sulfenylation thioredoxin thioredoxin-interacting protein

INTRODUCTION

Over the last century, Alzheimer’s disease (AD) has become known as the most common cause of dementia. AD is now recognized as a neurodegenerative disease with symptoms of memory loss, reduced cognitive function, and behavioral changes such as apathy, agitation, and aggression. Although the precise mechanisms are not fully understood, the amyloid plaque is a common feature found in the brain of patients with AD [1]. Because amyloid-β (Aβ) is the principal component of amyloid plaques, the hypothesis of Aβ cascades is considered to be an important mechanism for AD pathophysiology [2, 3]. Aβ is composed of 39–43 amino acids that are proteolytically derived from sequential enzymatic cleavage of the widely distributed transmembrane amyloid-β protein precursor (AβPP) by β-secretase and γ-secretase. Existing literature indicates that mutations in AβPP and presenilins lead to higher amounts of Aβ aggregation [3]. Recently, the toxic Aβ oligomer enriched with β-sheet was found to be a primary cause of cytotoxicity, synaptic impairment, and memory deficiency in rodents [4]. Various animal models with elevated Aβ levels have also showed impaired memory [5]. Although much evidence indicates that Aβ contributes significantly to neuronal impairment and memory deficiency in animal models for AD, anti-amyloid therapies for AD treatment have failed in many clinical trials [6, 7]. This discrepancy indicates that the mechanism of Aβ cascades involvement in AD remains unclear.

Oxidative stress results from reactive oxygen species/reactive nitrogen species (ROS/RNS) overproduction that overrides the cellular antioxidative capacity to remove excess ROS/RNS. ROS/RNS are capable of damaging or modifying several types of molecules within the cell including nucleic acids, lipids, and proteins. The brain is vulnerable to ROS/RNS attack because it has a high demand for oxygen and contains easily peroxidizable unsaturated fatty acids, and also has a relative lack of antioxidant systems [8]. Studies have shown oxidative stress increased in brain of AD animal models and Aβ-treated neuronal cells. For example, AβPP/presenilin 1 (PS1) double transgenic mice, a commonly used animal model for AD, have exhibited elevated Aβ levels and impaired memory [9, 10]. It has also been reported that ROS and lipid peroxidation were increased, but glutathione levels decreased, in brain of AβPP/PS1 mice when compared to wild-type mice [11, 12]. Our laboratory has also found that total nitrosylated protein levels were increased in the brain of 9- and 12-month-old AβPP/PS1 transgenic mice [10]. Treatment with Aβ has also been found to increase the levels of H₂O₂ and lipid peroxides in rodent cortical and hippocampal neurons, and rat PC12 cell line [12 –14]. These findings suggest that oxidative stress contributes to Aβ-induced neurotoxicity. Previous studies also showed that oxidative stress is involved in AD. Higher levels of lipid peroxidation products have been reported in various brain regions of patients with AD and mild cognitive impairment [15 –17]. Studies have also reported high levels of protein carbonyls in frontal cortex of patients with mild cognitive impairment, mild AD, and AD itself when compared to healthy matched controls [18]. However, a recent meta-analysis showed that levels of lipid peroxidation product malondialdehyde were only marginally increased in temporal and occipital lobes, while levels of protein carbonyls were marginally increased in occipital lobe, but not in other brain regions of AD patients [19], which raise a question about the role of oxidative stress in AD pathology. Because only data from bulk tissue were evaluated in this meta-analysis, oxidative stress analysis in specific cellular organelles and cell types, and subpopulations and stages of AD may provide a better picture of the role of oxidative stress in this neurodegenerative disease.

Redox-sensitive signaling may play a role in neurodegenerative diseases [20]. ROS can oxidatively modify thiol groups of redox-sensitive proteins that trigger signaling involved gene expression, apoptosis, and other cellular functions [21]. It was previously found that levels of thiol at reduced state were decreased, but levels of thiol at oxidized state were increased in AD patients when compared to healthy controls [22]. Nitrosylated thiol levels were also shown to be changed in AD patients [23]. Thioredoxin (Trx) is a highly conserved and ubiquitous oxidoreductase protein with cysteine 32 and cysteine 35 as the active sites. Through disulfide exchange, Trx can reverse protein cysteine oxidative modifications including sulfenylation and nitrosylation [21]. Trx can also reduce oxidized peroxiredoxin, facilitating peroxiredoxin-induced scavenging of H₂O₂ and other peroxides [24]. Trx is important in maintaining thiol redox homeostasis. Thioredoxin-interacting protein (Txnip) as a Trx inhibitor interacts with the active site of Trx, thereby inhibiting Trx reducing activity [25]. In the current study, we analyzed Trx and Txnip in brain of 3-, 6-, 9-, and 12-month-old AβPP/PS1 transgenic mice, an animal model for AD, and in Aβ-treated primary cultured mouse cerebral cortical neurons and mouse hippocampal HT22 cells.

MATERIALS AND METHODS

AβPP/PS1 double transgenic mice and genotyping

AβPP/PS1 mice were generated from mating single transgenic mice expressing human mutant AβPP_K670N/M671L and mice expressing human mutant PS1_M146L as described in our previous publication [10]. Mice were grown for 3, 6, 9, or 12 months. All procedures of these animal studies were in accordance with the guidelines from the Canadian Council on Animal Care.

Aβ peptide preparation

Preparation of Aβ₄₂ oligomers was performed as previously described [26]. Aβ₄₂ peptide (Abcam, US) was dissolved in 1,1,1,3,3,3-Hexafluoro-2-propanol (HFIP) (Sigma-Aldrich, Canada) with final concentration at 0.5 mM. Aβ/HFIP solution was incubated for 1 h at room temperature. The solution was dried under a fume hood overnight and the peptide film sealed the next day and stored at –20°C. The peptide film was resuspended in dimethyl sulfoxide (DMSO) with final concentration at 5 mM. Then DMSO solution of Aβ₄₂ was further diluted in phosphate buffered saline (PBS) to a concentration of 100 μM and incubated at 4°C for 24 h. The final DMSO concentration in cultured medium was 0.1% in vehicle and Aβ-treated cells. DMSO at 0.1% did not produce any effect on Trx and Txnip protein levels, or on nitrosylated and sulfenylated protein levels (data not shown).

Cell culture

Primary culturing of mouse cerebral cortical neurons was performed as previously described [27]. Embryonic fetuses at day 17-18 were isolated from the uterus and the cerebral cortex was dissected. Meninges and blood vessels were removed and then the tissue was digested by 0.25% trypsin for 15 min at 37°C, followed by DNase (0.7 mg/ml) digestion for another 15 min at 37°C. Then digestion procedure was terminated with 10% fetal bovine serum. After precipitation, supernatant was discarded. Then tissue was dissociated mechanically by pipetting and resuspended in 1 ml fresh neurobasal medium (Life Technologies Inc) with 1X GlutaMax (Life Technologies Inc), 1X GS21 supplement (Sigma-Aldrich Canada), and 1% of penicillin/streptomycin. Cells were seeded onto poly-D-lysine-coated (50 μg/ml) cell culture plates at a density of 3×10⁵ cells/ml. Medium was half changed twice per week. Cells were maintained in culture at 37°C under 5% CO₂ for 10 days before Aβ treatment.

HT22 mouse hippocampal cells were generously provided by the Salk Institute (La Jolla, CA, USA). HT22 cells were maintained in Dulbecco’s Modified Eagle Media (DMEM) (Life Technologies Inc, Burlington, ON, Canada) supplemented with 1% penicillin/streptomycin, 10% fetal bovine serum, and cultured at 37°C under 5% CO₂.

Tissue processing and protein isolation

Mouse frontal cortex and hippocampus were isolated and homogenized at 10:1 (ml/g) ice-cold lysis buffer. This buffer includes 20 mM HEPES (pH 7.5), 250 mM NaCl, 20% glycerol, 30 mM MgCl₂, 0.5 mM EDTA, 0.1 mM ethylene glycol tetraacetic acid (EGTA), 1% Nonidet P40 and 1× protease inhibitor cocktail (Thermo Scientific, Marietta, OH, USA). The homogenized tissues were kept on ice for 1 h and then centrifuged at 10,000× g for 15 min at 4°C. The supernatants were then collected as protein extract.

Cells were washed twice, scraped with ice-cold PBS and then collected by centrifuge at 1000× g for 5 min at 4°C. Cell lysis procedure was the same as the procedure used for mouse brain tissue as described above. The cell lysates were kept on ice for 1 h and then centrifuged at 10,000× g for 15 min at 4°C. The protein in supernatant was collected. Protein concentrations were determined by the Bradford protein assay [28].

Immunoblotting analysis

Protein samples were mixed with a loading buffer containing 100 mM Tris-HCl (pH 6.8), 200 mM dithiothreitol (DTT), 4% sodium dodecyl sulfate (SDS), 0.2% bromophenol blue, and 20% glycerol. Protein samples were loaded to electrophoresis in 12% SDS polyacrylamide gels for 1 h at 120 V. Then proteins were transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA) for 2 h at 220 mA on ice. Membrane blots were first blocked with 5% milk in Tris-buffered saline (TBS) containing 10 mM Tris-HCl (pH 7.4) and 0.1% Tween-20 at room temperature for 1 h. The membrane was incubated overnight at 4°C with primary antibody of rabbit monoclonal Txnip (1:2000 dilution, Abcam Inc., Toronto, ON, Canada) or rabbit monoclonal Trx1 (1:7000 dilution, Cell Signaling Technology, Danvers, MA, USA). Then the membrane was further incubated with secondary antibody of goat anti-rabbit conjugated to horseradish peroxidase (1:5000, Abcam, Eugene, Oregon, USA) for 1 h at room temperature. The membranes were developed using enhanced chemiluminescence reagents (PerkinElmer, Waltham, MA, USA), and then images were captured by the ChemiDoc MP System (Bio-Rad, Dreieich, Germany). Band signal intensity was quantified by measuring the density of the band using Image Lab (Bio-Rad).

Dimedone conjugation assay for detection of cysteine sulfenylated proteins

Sulfenylated proteins were measured by dimedone conjugation assay [27]. Sulfenylated thiols were first reacted with dimedone to generate irreversible dimedone-derivatized proteins that were then subjected to SDS-PAGE gel, transferred to membranes and analyzed by western blot with anti-dimedone antibody. Cells were washed with PBS and digested with 0.25% trypsin. Then the reaction was terminated with Dulbecco’s modified Eagle medium containing 5% dimedone and 10% fetal bovine serum. After cells were centrifuged at 1000× g, cell pellet was collected and lysed on ice for 30 min with a lysate buffer containing dimedone and catalase (2 mM dimedone, 3 mM citric acid, 0.1% Triton-X, 12 mM sodium phosphate dibasic, 10 mM iodoacetamide, 10 mM N-ethylmaleimide, 200 units/ml catalase, 10 mM EDTA, and 0.01% SDS). The lysates were centrifuged at 10,000 g for 10 min and the supernatants were collected. The protein concentration was measured by Bradford protein assay [28]. Fifteen μg of protein was used for analysis. Sulfenylated proteins were measured using immunoblotting analysis with anti-cysteine sulfenic acid antibody at 1:3000 dilution (Millipore Canada Ltd, Etobicoke, ON, Canada).

Biotin-switch assay for detection of cysteine nitrosylated proteins

Cysteine nitrosylated proteins were detected by biotin-switch assay [27]. First, unmodified thiols in cysteine were blocked with a thiol-specific methylthiolating agent methyl methanethiosulfonate (MMTS) (MMTS cannot block oxidatively modified thiols). Then nitrosylated thiols in cysteine were reduced back to free thiols by ascorbate. The newly formed free thiols were labeled with N-[6-(biotinamido) hexyl]-3’- (2’-pyridyldithio) propionamide (biotin-HPDP). Biotinylated protein (nitrosylated protein) was resolved by SDS-PAGE, transferred to membrane and detected by anti-biotin antibody. One hundred μg of protein in 50 μl was incubated with 200 μl blocking solution (40 mM MMTS, 2.5% SDS, 250 mM HEPES pH 7.7, 1 mM EDTA, and 0.1 mM neocuproine) at 50°C for 40 min, and then added with 1 ml cold 99% acetone and further incubated at –20°C for 40 min for precipitation. The reaction mixture was centrifuged at 13,000× g, at 4°C for 15 min. After supernatant was removed, the pellet was resuspended with 15 μl HENS buffer (250 mM HEPES pH 7.7, 0.1 mM neocuproine and 1 mM EDTA, 1% SDS), and added with 250 μl of 50 mM ascorbate and 4 mM biotin-HPDP, and incubated at room temperature for 1 h. Fifteen μl from each sample was used to analyze nitrosylated proteins using immunoblotting analysis with polyclonal anti-biotin antibody at 1:2000 dilution (Sigma, St. Louis, MO, USA).

Knocking out Txnip in HT22 cells

Txnip knocking out was performed using CRISPR-Cas9 method as described in our previous publication [27]. Scrambled sgRNAs (used as control) or Txnip single guide RNAs (sgRNAs), and CRISPR/Cas9 All-in-One lentivector pLenti-U6-sgRNA-SFFV-Cas9-2A-Puro were bought from ABM Inc (Richmond, BC, Canada). HEK293T cells were transfected with sgRNA-Cas9- constructs and packaging plasmids. Then the media containing lentiviral particles was collected after 48-h HEK293T cell transfection. The lentiviral particles were used to further transfect into HT22 cells. Single clones from scramble and Txnip knockout cells were selected and propagated in media containing 4 μg/ml Puromycin.

Statistical analysis

IBM SPSS 24.0 software (IBM, Armonk, New York, USA) was used to perform statistical analysis. All results were expressed as mean±standard error of the mean (SEM). Significant differences among means of more than two groups were analyzed by one-way analysis of variance (ANOVA) with Tukey post hoc comparisons. Differences in sulfenylation and nitrosylation in Txnip knockout study were analyzed using two-way ANOVA with Aβ treatment and Txnip knockout as between subject factors. Student’s t-tests were used for statistical analysis of two groups. A p value of less than 0.05 was regarded as statistically significant.

RESULTS

First, Trx and Txnip protein levels in frontal cortex and hippocampus of 3-, 6-, 9-, and 12-month-old AβPP/PS1 double transgenic mice were analyzed. We found that Trx protein levels were not changed in frontal cortex and hippocampus of 3-, 6-, 9-, and 12-month-old AβPP/PS1 mice when compared to controls (Fig. 1). However, we found that although Txnip protein levels were not significantly changed in frontal cortex and hippocampus of 3- and 6-month-old AβPP/PS1 mice, Txnip protein levels were significantly increased in frontal cortex and hippocampus of 9-month-old (p = 0.011 for frontal cortex, p = 0.010 for hippocampus) and 12-month-old (p = 0.045 for frontal cortex, p = 0.047 for hippocampus) AβPP/PS1 mice when compared to wild type littermates (Fig. 2).

Fig.1

Thioredoxin (Trx) protein levels in frontal cortex and hippocampus of 3-, 6-, 9-, and 12-month-old wild type (WT) mice and AβPP/PS1 transgenic mice (Tg). Proteins were isolated from frontal cortex and hippocampus of mice followed by immunoblotting analysis. β-actin was used as a normalization standard. Data are displayed as mean±SEM, 3 month: N = 9 for wild type, N = 10 for AβPP/PS1; 6 month: N = 10 for wild type, N = 6 for AβPP/PS1; 9 month: N = 9 for wild type, N = 9 for AβPP/PS1; 12 month: N = 9 for wild type, N = 5 for AβPP/PS1. *indicates p < 0.05 by student’s t test.

Fig.2

Thioredoxin-interacting protein (Txnip) protein levels in frontal cortex and hippocampus of 3-, 6-, 9-, and 12-month-old wild type (WT) mice and AβPP/PS1 transgenic mice (Tg). Proteins were isolated from frontal cortex and hippocampus of mice followed by immunoblotting analysis. β-actin was used as a normalization standard. Data are displayed as mean±SEM. *indicates p < 0.05 by student’s t test.

Second, because Aβ production is significantly increased in brain of AβPP/PS1 transgenic mice [23, 24], we further measured the direct effect of Aβ₄₂ on protein levels of Trx and Txnip in primary cultured mouse cerebral cortical neurons and in HT22 mouse hippocampal cells. Cultured cells were treated with Aβ₄₂ at 1 μM and 3 μM for one day. We found that although treatment with Aβ₄₂ at 1 μM or 3 μM had no effect on Trx protein levels, these treatments significantly increased Txnip protein levels in primary cultured mouse cerebral cortical neurons (p = 0.044 for 1 μM group and p = 0.000 for 3 μM group) (Fig. 3). We also found that treatment with Aβ₄₂ at 1 μM increased Txnip protein levels on trend (p = 0.056) and at 3 μM significantly increased Txnip protein levels (p = 0.019) in HT22 cells (Fig. 4).

Fig.3

The effect of Aβ₄₂ on Trx and Txnip protein levels in primary cultured mouse cerebral cortical neurons. Cultured neurons were treated with vehicle (CTL) or Aβ₄₂ at 1 and 3 μM for 1 day. Protein levels of Trx and Txnip were measured with each antibody using immunoblotting analysis. β-actin was used as a normalization standard. Results are shown as mean±SEM (Trx: N = 6 for each group; Txnip: N = 6 for each group). *indicates p < 0.05 when compared to controls determined by one-way ANOVA followed by Tukey’s post-hoc analysis.

Fig.4

The effect of Aβ₄₂ on Trx and Txnip protein levels in HT22 cells cerebral cortical neurons. HT22 cells were treated with vehicle (CTL) or Aβ₄₂ at 1 and 3 μM for 1 day. Protein levels of Trx and Txnip were measured with each antibody using immunoblotting analysis. β-actin was used as a normalization standard. Results are shown as mean±SEM (Trx: N = 4 for each group; Txnip: N = 6 for each group). *indicates p < 0.05 when compared to controls determined by one-way ANOVA followed by Tukey’s post-hoc analysis.

Third, because Aβ-increased Txnip may promote cysteine oxidative modification, we tested whether Aβ treatment increases protein cysteine sulfenylation and nitrosylation. H₂O₂ and NO radical can attack cysteine thiol groups, causing protein cysteine sulfenylation and nitrosylation respectively, therefore H₂O₂ was used as positive control for sulfenylation and NO donor S-nitrosoglutathione (GSNO) was used as positive control for nitrosylation. HT22 cells were treated with Aβ₄₂ at 3 μM for one day, H₂O₂ at 300 μM for 30 min and GSNO at 200 μM for 30 min. As shown in Fig. 5, treatment with Aβ or H₂O₂ significantly increased total sulfenylated protein levels (p = 0.022 for Aβ group and p = 0.012 for H₂O₂ group). Similarly, treatment with Aβ or GSNO also increased total nitrosylated protein levels (p = 0.016 for Aβ group and p = 0.015 for GSNO group) (Fig. 5).

Fig.5

The effect of Aβ₄₂ on protein sulfenylation and nitrosylation in HT22 cells. Cells were treated with vehicle (CTL) and Aβ₄₂ at 3 μM for 1 day. Cells were also treated with H₂O₂ at 300 μM for 30 min and nitric oxide donor S-nitrosoglutathione (GSNO) at 200 μM for 30 min as positive controls. Sulfenylated protein levels were measured by dimedone conjugation assay followed by immunoblotting analysis. Nitrosylated protein levels were measured by biotin-switch method followed by immunoblotting analysis. Band signal intensity in whole lane was quantitated by densitometry. The membrane was striped and stained with Coomassie blue used as loading control. Data are displayed as mean±SEM (protein sulfenylation: N = 5 for each group; protein nitrosylation: N = 5 for each group). *indicates p < 0.05 when compared to controls determined by one-way ANOVA followed by Tukey’s post hoc test.

To examine the potential role of Txnip in Aβ-induced protein sulfenylation and nitrosylation, we measured the effect of Aβ on sulfenylated and nitrosylated protein levels in Txnip knockout HT22 cells. As shown in Fig. 6, Txnip protein levels were significantly inhibited in cells transfected with Txnip sgRNA/CRISPR Cas9, but not in cells transfected with scrambled sgRNA/CRISPR Cas9 (p = 0.011). Further, we found that two-way ANOVA showed significant Aβ treatment and Txnip gene knockout interactions for sulfenylation (p = 0.001) and nitrosylation (p = 0.001). Treatment with Aβ at 3 μM for 1 day significantly increased sulfenylated protein levels (p = 0.036) and nitrosylated protein levels (p = 0.002) in cells transfected with scrambled sequence, but not in cells transfected with Txnip sgRNAs (Fig. 7).

Fig.6

Knocking out Txnip gene in HT22 cells. Cells were transfected with lentivectors containing Txnip sgRNAs/CRISPR/Cas9 (KO). Lentivectors with scrambled sgRNAs/CRISPR/Cas9 (SCM) were used as control. Txnip protein levels were measured by immunoblotting analysis. β-actin was used as loading control. Data are displayed as mean±SEM (N = 4 for each group). *indicates p < 0.05 when compared to SCM group determined by t test.

Fig.7

Effect of Txnip sgRNAs on Aβ-increased protein sulfenylation and nitrosylation. HT22 cells transfected with Txnip sgRNAs or scrambled sgRNAs were treated with vehicle (CTL) or Aβ₄₂ at 3 μM for 1 day. Sulfenylated protein was measured by dimedone conjugation assay. Nitrosylated protein was measured by biotin switch method. Band signal intensity in whole lane was quantitated by densitometry. The membrane was striped and stained with Coomassie blue used as loading control. Data are displayed as mean±SEM (protein sulfenylation: N = 5 for each group; protein nitrosylation: N = 5 for each group). *indicates p < 0.05 when compared to controls determined by two-way ANOVA followed by Tukey’s post hoc test.

DISCUSSION

In the present study, we found that although Trx protein levels were not changed in frontal cortex and hippocampus of 3-, 6-, 9-, and 12-month-old AβPP/PS1 double transgenic mice, Txnip protein levels were significantly increased in frontal cortex and hippocampus of 9- and 12-month-old AβPP/PS1 mice when compared to controls. 5XFAD mouse is another commonly used animal model for AD. It was found that Txnip protein levels were also increased in the hippocampus of 4-month-old 5XFAD mice when compared to controls [29]. Since AβPP/PS1 and 5XFAD mice exhibit increased Aβ levels in brain, these results suggest that Txnip can be upregulated by Aβ. Indeed, we further found that treatment with Aβ increased Txnip protein levels in primary cultured cerebral cortical neurons and HT22 mouse hippocampal cells.

Txnip cysteine 247 residue can interact with Trx active center cysteine-32 residue, subsequently inhibiting Trx reducing capability [25]. Txnip is considered as an endogenous inhibitor for Trx. Because Trx can reverse cysteine nitrosylation and sulfenylation, upregulation of Txnip may lead to inhibition of Trx reducing capacity, resulting in enhanced protein cysteine sulfenylation and nitrosylation, which affects neuronal functions. Previously we reported that nitrosylated protein levels were significantly increased in hippocampus and frontal cortex of 9- and 12-month-old AβPP/PS1 transgenic mice [10]. In the present study, we found that treatment with Aβ increased sulfenylated and nitrosylated protein levels in HT22 cells. Our findings suggest that Aβ-upregulated Txnip may facilitate protein cysteine sulfenylation and nitrosylation processes. To further determine if Txnip contributes to Aβ-induced protein cysteine oxidative modification, we measured the effect of knocking out Txnip on Aβ-induced protein sulfenylation and nitrosylation. We found that knocking out Txnip totally reversed Aβ-increased sulfenylated protein levels and partially reduced Aβ-increased nitrosylated protein levels. Our findings indicate that Aβ may upregulate Txnip levels, inhibiting Trx reducing activity and promoting protein cysteine oxidation process.

Cysteine is highly vulnerable to attack by ROS/RNS, resulting in cysteine oxidative modification. Protein cysteine residues are involved in many protein functions including enzymatic catalysis, homeostasis, and metal binding [30]. Since cysteine nitrosylation and sulfenylation are reversible, these cysteine modifications may mediate redox-regulated signal transduction, resulting in interruption of many cysteine-mediated cellular processes, which may further contribute to development of AD. Cholinergic deficits and glutamate-induced excitotoxicity play important roles in AD development. Previously our laboratory found that vesicular acetylcholine transporter and vesicular glutamate transporter can be nitrosylated and nitrosylation of these transporters can inhibit their neurotransmitter uptake functioning [31]. Further nitrosylation of vesicular acetylcholine transporter and vesicular glutamate transporter was increased in hippocampus and frontal cortex of AβPP/PS1 transgenic mice [10]. These results suggest that the nitrosylation process may contribute to impairment of cholinergic and glutamatergic neurotransmission in AD. Pin1 is a peptidyl-prolyl cis/trans isomerase and isomerizes only phospho-serine/ threonine/proline motifs. Pin1-induced conformational changes affect various protein functions such as phosphorylation status, protein interaction, and protein stability [32]. It was found that the sulfenylation of Cys113 on Pin1 was significantly elevated in the postmortem AD human brain and in AβPP transgenic mice. The oxidation of Cys113 induced Pin1 inactivation and mislocalization [33], indicating that oxidative modification of Pin1 disrupts protein phosphorylation, protein interaction and protein stability, which may be important in AD pathology. Since Txnip can promote protein cysteine oxidative modification, Aβ-increased Txnip may further enhance cysteine oxidative modification of vesicular transporters and various enzymes, resulting in impairment of neuroplasticity. These finding may also suggest that ROS/RNS-regulated specific redox signaling through cysteine oxidative modification, rather than nonspecific oxidative damage, contributes to AD development. Currently, redox proteomics analysis using mass spectrometry is being developed to identify a wider array of oxidatively modified protein products [23, 34]. Analysis of oxidatively modified protein in whole proteome may lead to the discovery of novel mechanisms of protein functions regulated by redox, which can help to clarify complexity of cysteine oxidative signaling in AD research. It is of interest that reversible cysteine oxidative modification is not solely detrimental, but also sometimes beneficial. For example, it has been found that NMDA receptor subunit NR2A Cys-399 can be nitrosylated in cultured rat cerebral cortical neurons. Nitrosylation of NR2A can reduce Ca^2 + influx and protect against excitotoxicity [35]. Since Txnip can bind to Trx, Txnip/Trx interaction may act as a redox-mediated signaling sensor, and play an important role in regulating cysteine oxidative modification under both physiological and pathological conditions.

Recently, Txnip has been found to directly bind to NLR family pyrin domain containing 3 protein (NLRP3), resulting in activation of NLRP3 inflammasome. NLRP3 inflammasome activation can cleave procaspase-1 into caspase-1 that cleaves interleukin (IL)-1β precursor and IL-18 precursor to mature IL-1β and IL-18, activating proinflammatory responses [25]. Indeed, many studies have indicated that NLRP3 and its downstream effectors caspase-1 and IL-1β are upregulated in AD patients, AD animal models and Aβ-treated cells. For example, NLRP3, caspase-1 and IL-1β levels were found to be upregulated in Aβ-treated monocytes from severe and mild AD patients [36]. Caspase-1 levels were also found to be increased in postmortem hippocampus and frontal cortex of AD patients and 16-month-old AβPP/PS1 transgenic mice [37]. It was also noted that IL-1β levels were increased in postmortem temporal cortex of AD patients and that IL-1β levels were also increased in the cerebrospinal fluid of patients with AD [38, 39]. Reactive IL-1β-immunoreactive astrocytes were found in close proximity to both fibrillary and diffuse Aβ deposits in neocortex and hippocampus of 13-, 16-, and 19-month-old AβPP transgenic mice [40]. Aβ has also been found to promote the processing of pro-IL-1β into mature IL-1β [41]. These studies indicate that NLRP3 inflammasome and proinflammatory processes are activated and lead to the pathology of AD. Aβ-increased Txnip may accelerate the NLRP3 inflammasome-triggered inflammatory process. However, a clear role for Txnip in Aβ-induced inflammation needs to be further investigated.

Our findings should be interpreted with some caution due to limitations in the samples and methods used in the present studies. Firstly, only frontal cortex and hippocampus were sampled, while other brain regions may also be involved in amyloid-caused toxicity. Also, because bulk brain tissues contain heterogeneous cell populations, Txnip could be further explored in multiple brain regions and homogenous cells from central nervous system. Secondly, nitrosylation and sulfenylation are reversible and labile oxidative modifications. The detection of these modifications is technically more challenging than other post-translational protein modifications. Both biotin-switch assay and dimedone conjugation assay used in the current study are indirect methods for analysis of these modifications, and cannot identify oxidatively modified cysteine residues. In addition, using these methods, we only detected total nitrosylated and sulfenylated protein levels. Therefore, identifying nitrosylated and sulfenylated proteins in whole proteome using direct mass spectrometry-based methods will help us to determine the physiology of cysteine oxidation and further reveal its molecular mechanisms.

In summary, we found that Txnip protein levels were increased in frontal cortex and hippocampus of 9- and 12-month-old AβPP/PS1 transgenic mice. Txnip protein levels were also increased by Aβ treatment in cultured neuronal cells. Further, Aβ increased protein cysteine nitrosylation and sulfenylation in cultured neuronal cells, and knocking out Txnip inhibited Aβ-increased nitrosylation and sulfenylation. Our results suggest that Aβ may upregulate Txnip, inhibiting Trx reducing capability and promoting protein cysteine oxidative modifications. Our findings also indicate that Txnip may be a potential target for the treatment of AD. It is noteworthy that clinical trials on existing antioxidant therapies for AD have not produced any satisfactory outcome [42, 43]. Clearly identifying further novel targets for AD drug development is needed.

Footnotes

ACKNOWLEDGMENTS

This work is supported by a Research Grant from the Alzheimer Society Research Program of Canada (JFW) and by a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada (JFW).

Authors’ disclosures available online ().

References

Gouras

, Olsson

, Hansson

(2015) β-Amyloid peptides and amyloid plaques in Alzheimer’s disease. Neurotherapeutics 12, 3–11.

Hardy

, Higgins

(1992) Alzheimer’s disease: The amyloid cascade hypothesis. Science 256, 184–185.

Vetrivel

, Thinakaran

(2006) Amyloidogenic processing of β-amyloid precursor protein in intracellular compartments. Neurology 66(1 Suppl 1), S69–S73.

Cerasoli

, Ryadnov

, Austen

(2015) The elusive nature and diagnostics of misfolded Aβ oligomers. Front Chem 3, 17.

Kim

, Jang

, Jeon

, Han

(2019) Assessment of cognitive phenotyping in inbred, genetically modified mice, and transgenic mouse models of Alzheimer’s disease. Exp Neurobiol 28, 146–157.

Karran

, de Strooper

(2016) The amyloid cascade hypothesis: Are we poised for success or failure? J Neurochem 139(Suppl 2), 237–252.

Kozin

, Barykin

, Mitkevich

, Makarov

(2018) Anti-amyloid therapy of Alzheimer’s disease: Current state and prospects. Biochemistry (Mosc) 83, 1057–1067.

Maher

, Schubert

(2000) Signaling by reactive oxygen species in the nervous system. Cell Mol Life Sci 57, 1287–1305.

Zhang

, Hao

, Liu

, Zhang

, Lei

, Su

, Miao

, Li

(2011) Soluble Aβ levels correlate with cognitive deficits in the 12-month-old APPswe/PS1dE9 mouse model of Alzheimer’s disease. Behav Brain Res 222, 342–350.

10.

Wang

, Zhou

, Tan

, Zhu

, Wang

, Sun

, Wang

(2017) Nitrosylation of vesicular transporters in brain of amyloid precursor protein/presenilin 1 double transgenic mice. J Alzheimers Dis 55, 1683–1692.

11.

Ahmad

, Singh

, Das

, Gowaikar

, Shaw

, Ramachandran

, Rupanagudi

, Kommaddi

, Bennett

, Ravindranath

(2017) Reactive oxygen species-mediated loss of synaptic Akt1 signaling leads to deficient activity-dependent protein translation early in Alzheimer’s disease. Antioxid Redox Signal 27, 1269–1280.

12.

Ali

, Kim

, Rehman

, Khan

, Amin

, Khan

, Ikram

, Kim

(2018) Natural dietary supplementation of anthocyanins via PI3K/Akt/Nrf2/HO-1 pathways mitigate oxidative stress, neurodegeneration, and memory impairment in a mouse model of Alzheimer’s disease. Mol Neurobiol 55, 6076–6093.

13.

Behl

, Davis

, Lesley

, Schubert

(1994) Hydrogen peroxide mediates amyloid β protein toxicity. Cell 77, 817–827.

14.

De Felice

, Velasco

, Lambert

, Viola

, Fernandez

, Ferreira

, Klein

(2007) Aβ oligomers induce neuronal oxidative stress through an N-methyl-D-aspartate receptor-dependent mechanism that is blocked by the Alzheimer drug memantine. J Biol Chem 282, 11590–11601.

15.

Bradley

, Xiong-Fister

, Markesbery

, Lovell

(2012) Elevated 4-hydroxyhexenal in Alzheimer’s disease (AD) progression. Neurobiol Aging 33, 1034–1044.

16.

Greilberger

, Koidl

, Greilberger

, Lamprecht

, Schroecksnadel

, Leblhuber

, Oettl

(2008) Malondialdehyde, carbonyl proteins and albumin-disulphide as useful oxidative markers in mild cognitive impairment and Alzheimer’s disease. Free Radic Res 42, 633–638.

17.

Markesbery

, Kryscio

, Lovell

, Morrow

(2005) Lipid peroxidation is an early event in the brain in amnestic mild cognitive impairment. Ann Neurol 58, 730–735.

18.

Ansari

, Scheff

(2010) Oxidative stress in the progression of Alzheimer disease in the frontal cortex. J Neuropathol Exp Neurol 69, 155–167.

19.

Zabel

, Nackenoff

, Kirsch

, Harrison

, Perry

, Schrag

(2018) Markers of oxidative damage to lipids, nucleic acids and proteins and antioxidant enzymes activities in Alzheimer’s disease brain: A meta-analysis in humanathological specimens. Free Radic Biol Med 115, 351–360.

20.

McBean

, Aslan

, Griffiths

, Torrão

(2015) Thiol redox homeostasis in neurodegenerative disease. Redox Biol 5, 186–194.

21.

Matsuzawa

(2017) Thioredoxin and redox signaling: Roles of the thioredoxin system in control of cell fate. Arch Biochem Biophys 617, 101–105.

22.

Gumusyayla

, Vural

, Bektas

, Deniz

, Neselioglu

, Erel

(2016) A novel oxidative stress marker in patients with Alzheimer’s disease: Dynamic thiol-disulphide homeostasis. Acta Neuropsychiatr 28, 315–320.

23.

Zahid

, Khan

, Oellerich

, Ahmed

, Asif

(2014) Differential S-nitrosylation of proteins in Alzheimer’s disease. Neuroscience 256, 126–136.

24.

Silva-Adaya

, Gonsebatt

, Guevara

(2014) Thioredoxin system regulation in the central nervous system: Experimental models and clinical evidence. Oxid Med Cell Longev 2014, 590808.

25.

Chong

, Chan

, Nguyen

, Liu

, Procter

, Ngo

, Sverdlov

, Chirkov

, Horowitz

(2014) Thioredoxin-interacting protein: Pathophysiology and emerging pharmacotherapeutics in cardiovascular disease and diabetes. Cardiovasc Drugs Ther 28, 347–360.

26.

Batarseh

, Mohamed

, Al Rihani

, Mousa

, Siddique

, El Sayed

, Kaddoumi

(2017) Oleocanthal ameliorates amyloid-β oligomers’ toxicity on astrocytes and neuronal cells: In vitro studies. Neuroscience 352, 204–215.

27.

Bharti

, Tan

, Chow

, Wang

, Nagakannan

, Eftekharpour

, Wang

(2018) Glucocorticoid upregulates thioredoxin-interacting protein in cultured neuronal cells. Neuroscience 384, 375–383.

28.

Bradford

(1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248–252.

29.

Gouget

, Djelloul

, Boucraut

, Weinhard

, Baranger

, Rivera

, Perrone

(2011) TXNIP, the major player in insulin resistance, is early over-expressed in the brain of the 5XFAD Alzheimer’s mice model and is induced by Aβ in vitro: Emerging role of TXNIP and inflammation in Alzheimer’s disease progression. Alzheimers Dement 7, S684.

30.

Giles

, Watts

, Giles

, Fry

, Littlechild

, Jacob

(2003) Metal and redox modulation of cysteine protein function. Chem Biol 10, 677–693.

31.

Wang

, Zhou

, Leylek

, Tan

, Sun

, Parkinson

, Wang

(2015) Protein cysteine S-nitrosylation inhibits vesicular uptake of neurotransmitters. Neuroscience 311, 374–381.

32.

Liou

, Zhou

, Lu

(2011) Prolyl isomerase Pin1 as a molecular switch to determine the fate of phosphoproteins. Trends Biochem Sci 36, 501–514.

33.

Chen

, Li

, Sultana

, You

, Kondo

, Shahpasand

, Kim

, Luo

, Nechama

, Lin

, Yao

, Lee

, Zhou

, Swomley

, Butterfield

, Zhang

, Lu

(2015) Pin1 cysteine-113 oxidation inhibits its catalytic activity and cellular function in Alzheimer’s disease. Neurobiol Dis 76, 13–23.

34.

, Liu

, Ferreira

, Carroll

, Yang

(2019) Proteome-wide analysis of cysteine s-sulfenylation using a benzothiazine-based probe. Curr Protoc Protein Sci 95, e76.

35.

Kim

, Choi

, Rayudu

, Das

, Asaad

, Arnelle

, Stamler

, Lipton

(1999) Attenuation of NMDA receptor activity and neurotoxicity by nitroxyl anion, NO. Neuron 24, 461–469.

36.

Saresella

, La Rosa

, Piancone

, Zoppis

, Marventano

, Calabrese

, Clerici

(2016) The NLRP3 and NLRP1 inflammasomes are activated in Alzheimer’s disease. Mol Neurodegener 11, 23.

37.

Heneka M.

, Kummer M.

, Stutz

, Delekate

, Schwartz

, Vieira-Saecker

, Gelpi

(2013) NLRP3 is activated in Alzheimer’s disease and contributes to pathology in App/Ps1 mice. Nature 493, 674.

38.

Griffin

, Stanley

, Ling

, White

, MacLeod

, Perrot

, Araoz

(1989) Brain interleukin 1 and S-100 immunoreactivity are elevated in Down syndrome and Alzheimer disease. Proc Natl Acad Sci U S A 86, 7611–7615.

39.

Apelt

, Schliebs

(2001) β-amyloid-induced glial expression of both pro-and anti-inflammatory cytokines in cerebral cortex of aged transgenic Tg2576 mice with Alzheimer plaque pathology. Brain Res 894, 21–30.

40.

Blum-Degena

, Müller

, Kuhn

, Gerlach

, Przuntek

, Riederer

(1995) Interleukin-1β and interleukin-6 are elevated in the cerebrospinal fluid of Alzheimer’s and de novo Parkinson’s disease patients. Neurosci Lett 202, 17–20.

41.

Parajuli

, Sonobe

, Horiuchi

, Takeuchi

, Mizuno

, Suzumura

(2013) Oligomeric amyloid β induces IL-1β processing via production of ROS: Implication in Alzheimer’s disease. Cell Death Dis 4, e975.

42.

Ferreira

, de Vasconcelos

, da Costa Vilhena

, da Silva

, Barbosa

, Gomes

, Dolabela

, Percário

(2015) Oxidative stress in Alzheimer’s disease: Should we keep trying antioxidant therapies? Cell Mol Neurobiol 35, 595–614.

43.

Wojsiat

, Zoltowska

, Laskowska-Kaszub

, Wojda

(2018) Oxidant/antioxidant imbalance in Alzheimer’s disease: Therapeutic and diagnostic prospects. Oxid Med Cell Longev 2018, 6435861.