Amyloid Beta-Peptide 25–35 (Aβ_25–35) Induces Cytotoxicity via Multiple Mechanisms: Roles of the Inhibition of Glucosylceramide Synthase by Aβ_25–35 and Its Protection by D609

Abstract

Sphingolipids (SLs), such as ceramide, glucosylceramide (GlcCer), and sphingomyelin, play important roles in the normal development/functions of the brain and peripheral tissues. Disruption of SL homeostasis in cells/organelles, specifically up-regulation of ceramide, is involved in multiple diseases including Alzheimer’s disease (AD). One of the pathological features of AD is aggregates of amyloid beta (Aβ) peptides, and SLs regulate both the formation/aggregation of Aβ and Aβ-induced cellular responses. Up-regulation of ceramide levels via de novo and salvage synthesis pathways is reported in Aβ-treated cells and brains with AD; however, the effects of Aβ on ceramide decomposition pathways have not been elucidated. Thus, we investigated the effects of the 25–35-amino acid Aβ peptide (Aβ_25–35), the fundamental cytotoxic domain of Aβ, on SL metabolism in cells treated with the fluorescent nitrobenzo-2-oxa-1,3-diazole-labeled C6-ceramide (NBD-ceramide). Aβ_25–35 treatment reduced the formation of NBD-GlcCer mediated by GlcCer synthase (GCS) without affecting the formation of NBD-sphingomyelin or NBD-ceramide-1-phosphate, and reduced cell viability. Aβ_25–35-induced responses decreased in cells treated with D609, a putative inhibitor of sphingomyelin synthases. Aβ_25–35-induced cytotoxicity significantly increased in GCS-knockout cells and pharmacological inhibition of GCS alone demonstrated cytotoxicity. Our study revealed that Aβ_25–35-induced cytotoxicity is at least partially mediated by the inhibition of GCS activity.

INTRODUCTION

Alzheimer’s disease (AD), the most common type of dementia, affects many people, specifically those aged over 65–75 years, and causes neurodegeneration and abnormality of the brain. In addition to neurofibrillary tangles containing hyper-phosphorylated tau protein, deposition of amyloid beta (Aβ) plaques is a pathological hallmark of AD.^1,2) The Aβ fibrils and/or plaques are mainly composed of the 1–40- and 1–42-amino acid Aβ peptides derived from a large Aβ precursor protein (AβPP). Aβ_25–35 is regarded as the fundamental cytotoxic and/or biologically active domain of Aβ, and the peptide existing in humans in vivo is observed/detected in neurons of the cortex of AD patients.^1–3)

Sphingolipids (SLs), such as ceramide, glucosylceramide (GlcCer), sphingomyelin (SM), ceramide-1-phosphate (C1P), sphingosine and resulting sphingosine-1-phosphate (S1P), play important roles in multiple functions of the brain.^4–6) In the brains of AD patients, levels of ceramide, a cytotoxic molecule, are high accompanied by increased expression/activity of several ceramide synthases and sphingomyelinases (SMases).^5,7,8) In contrast, levels of S1P, which directly and via S1P receptors plays protective roles against cytotoxicity, were reduced by the up- and down-regulation of S1P phosphatases and sphingosine kinases, respectively.^5,7,8) These studies demonstrated that abnormality in SL metabolism leads to AD. Structural roles of SLs are important for their functions in AβPP/Aβ metabolism, including the processing of AβPP to Aβ peptides and their aggregation.^2,6) In addition, treatment with Aβ peptides, including Aβ_25–35, which increased ceramide levels through the activation of SMases, induced cytotoxicity in cortical neurons,⁹⁾ oligodendrocytes,¹⁰⁾ and astrocytes.⁷⁾ Aβ-induced responses were also altered by S1P⁹⁾ and ganglioside GM1.¹¹⁾ This suggested that Aβ-induced cellular responses are controlled by SL metabolism. However, the effects of inhibitors of ceramide-converting enzymes (CCEs) that metabolize ceramide, such as GC synthase (GCS), SM synthases (SMSs), ceramidases, and ceramide kinase (CerK), on cell viability with and without Aβ_25–35 have not been examined simultaneously to our knowledge.

Symptomatic expression of AD is restricted to the brain, whereas AD pathophysiology is ubiquitous and notable in the peripheral tissues/cells such as blood, skin, and eye lens.¹²⁾ Aβ secretion increases in skin fibroblasts of patients with familial AD, and Aβ treatment caused multiple types of pathophysiological and/or cellular responses in fibroblasts, endothelial cells, and islet β-cells.^13–17) In patients with chronic obstructive pulmonary disease, which is associated with cognitive decline, significantly increased levels of serum Aβ levels were observed and this increase was correlated with poorer pulmonary function.¹⁸⁾ In the present study, we examined the effects of Aβ_25–35 and several inhibitors of CCEs on cell viability using lung and skin fibroblasts as model cells. In some cases, the Aβ_25–35-induced changes in CCE activity, including GCS and SMSs, were examined using fluorescent nitrobenzo-2-oxa-1,3-diazole-labeled C6-ceramide (NBD-ceramide). The roles of GCS in cell viability after Aβ treatment were also examined in two HeLa cell lines, GCS knockout (GCS-KO) cells and GCS-expressing (GCS-wild-type (WT)) cells.

MATERIALS AND METHODS

Cell Culture

We used two types of fibroblasts, HFL-1 cells, a cell line from human fetal lung (American Type Culture Collection, Manassas, VA, U.S.A., diploid, passage 14), and GM00038, a cell line from the skin of healthy humans (Coliell Institute, NJ, U.S.A., passage 10). The cells were cultured in medium containing 10% serum according to the manufacturer’s protocol. In order to examine the effects of genetic inhibition of GCS on cytotoxicity, we used two types of human cervical carcinoma HeLa cells, native HeLa cells (HeLa-mCat cells, GCS-WT cells) and mutant cells genetically lacking GCS function (GCS-KO cells). GCS-knockout HeLa cells were established using transcription activator-like effector nucleases (TALEN) method and cell growth of this cell line is similar to parental HeLa cells. These HeLa cells were kind gifts from Dr. Hanada (National Institute of Infectious Disease, Tokyo, Japan).¹⁹⁾ The cells that achieved 70–80% confluence (sub-confluent stage) at 2–3 d after seeding were used for experiments.

Materials

Reagents used in this study and their sources were as follows. Aβ_25–35 (human, Peptide Institute, Kumamoto, Japan); NB-DNJ (alias miglustat, Toronto Res. Chem., Toronto, ON, Canada); NVP-231, D-threo-1-phenyl-2-hexadecanoylamino-3-morpholino-1-propanol (PPMP), D609, ceranib-2, C2- and C6-ceramide (Cayman, Ann Arbor, MI, U.S.A.); eliglustat (alias GENZ-112638, Selleck Chem., Huston, TX, U.S.A.). Official names and/or structural formulae and pharmacological characters of reagents/drugs, such as NB-DNJ, PPMP, and D609, are described in Supplementary Materials. A stock solution of Aβ_25–35 was prepared with dimethyl sulfoxide as described in the manufacturer’s protocol and kept at −20 °C until use. The stock solutions of other reagents were prepared with distilled water or dimethyl sulfoxide. The reagents including Aβ_25–35 were diluted with medium or buffer before experiments, and final concentration of dimethyl sulfoxide was 1% or less. NBD-ceramide, which has an NBD-bound C6-N-acyl chain and C18-sphingosine, was purchased from Molecular Probes (Eugene, OR, U.S.A.) and was prepared by Prof. Nishida (Chiba University, Japan).

Assay for Cell Viability

Cells cultured with serum were treated with Aβ_25–35 and/or the indicated reagents for 24 h. In some cases, the cells were treated with Aβ_25–35 in the absence of serum. Cell viability was measured by the WST-8, a tetrazolium salt, colorimetric method using the Cell Count Reagent SF kit (#07553, Nacalai, Kyoto, Japan). The final concentrations of dimethyl sulfoxide in assay mixtures were less than 1%, which by itself did not affect cell viability.

Measurement of Changes in NBD-Ceramide Metabolism in Cells Treated with Aβ_25–35

Cells on 12-well plates after treatment with the vehicle, Aβ_25–35, and/or the reagents were incubated with 10 µM NBD-ceramide for 30 min with Hanks’ Balanced Salt Solution containing 0.1% fatty acid-free albumin. Ceramide and its metabolites, including NBD-GlcCer, NBD-SM, and NBD-C1P, were separated on TLC silica gel-60 plates (Merk, Darmstadt, Germany), as described previously.^20,21) NBD fluorescence was detected using LAS4000-Plus (FUJIFILM, Tokyo, Japan; 470 nm excitation and 515 emission). For quantitative analyses of NBD-ceramide metabolites, different amounts (0–20 pmol) of standard NBD-ceramide were spotted in the upper area of the plate after separation by TLC. Absolute values (pmol/well) of each ceramide metabolite varied depending on experiments but had the same order of magnitude; values were 80–200 for NBD-GlcCer, 40–90 for NBD-SM, and 2–8 for NBD-C1P in the control cells. NBD-caproic acid, which was produced by ceramidases, was not analyzed because NBD-caproic acid and NBD-ceramide exhibited similar mobility (Rf values, approximately 0.9) by our TLC method.²⁰⁾ Thus, the effects of Aβ_25–35 on the formation of sphingosine/S1P were unable to be evaluated in this study.

Data Presentation

No sample size calculations were performed and no test of outliers was conducted. Data are the means ± standard deviation (S.D.) of 3–4 independent experiments performed in duplicate or triplicate in many cases. In some cases of the measurement of cell viability, absolute absorbance values were plotted as the means ± S.D. ANOVA with Dunnett’s test and the Student’s two-tailed t-test were used for multiple and pair-wise comparisons, respectively. * p < 0.05 and ** p < 0.01 indicate significant differences compared with the control. Typical images of NBD-ceramide metabolites in the TLC plate were from at least 3 independent and representative experiments.

RESULTS

Reduced Cell Viability by Aβ_25–35 Treatment and Its Alteration by Inhibitors of CCEs in Human Lung Fibroblasts

Cell viability in HFL-1 cells, a human lung fibroblast cell line, was examined with Aβ_25–35 and inhibitors of CCEs. A model of ceramide metabolism in cells and possible target CCEs of inhibitors tested in the present study is shown in Fig. 1A. The pharmacological characteristics of inhibitors tested including their structural formulae are shown in Supplementary Materials. HFL-1 cells were cultured in medium with 10% serum at 70–80% confluency, and then further treated with Aβ_25–35 for 24 h. The absolute values of cell viability in a typical experiment performed in triplicate cell samples that were treated with Aβ are shown in Fig. 1B. Cell viability was significantly reduced by 10 and 20 µM Aβ_25–35 treatment (Fig. 1C). The pretreatment of Aβ_25–35 alone overnight with medium at 37 °C before application to cells did not change the Aβ response including its variations. Next, we investigated the concentration-dependent effects of respective inhibitors for CCEs on cell viability with and without Aβ_25–35 (Table 1, Supplementary Fig. 1). Treatment with 1 µM D609, a putative inhibitor of SMSs,²²⁾ slightly increased cell viability alone and significantly reduced Aβ_25–35-induced cytotoxicity (Fig. 1C). Of note, D609 at less than 1 µM was effective, and more than 1 µM D609 was not shown additional effect (Supplementary Fig. 1A). PPMP and eliglustat, those are ceramide analogs and act as GCS inhibitors, alone exhibited cytotoxicity from 10 µM and from 40 µM, respectively. The cytotoxic effects of these two inhibitors were additive to those of Aβ_25–35. NB-DNJ, an iminosugar class of GCS inhibitor, at any concentration up to 300 µM did not affect cell viability alone or in the presence of Aβ_25–35. We previously reported that treatment of cells with 10 µM PPMP and 200 µM NB-DNJ markedly, over 80%, inhibited the formation of NBD-GlcCer by GCS,^20,23) and confirmed that PPMP and eliglustat at the µM order and NB-DNJ at the 100-µM order almost completely inhibited GCS activity in vitro (data not shown). Thus, the lack of effects of NB-DNJ in intact cells may not have been because the reagent was broken down. NVP-231, a CerK inhibitor, alone exhibited cytotoxicity, as previously reported.²⁴⁾ Ceramib-2, a ceramidase inhibitor, at 20 µM exhibited cytotoxicity with and without Aβ_25–35.

Fig. 1. Effects of Aβ_25–35, D609, and Ceramide on Cell Viability

Ceramide metabolism and its inhibitors used in this study are shown in A. In B–E, HFL-1 cells cultivated in medium with serum were treated with vehicle or 10 and 20 µM Aβ_25–35 for 24 h, and then cell viability was examined. In some cases, D609 or ceramide at the indicated concentrations was supplemented during the period of Aβ treatment. In B, absolute values of WST-8 absorbance, reflecting cell viability, in a typical experiment performed in triplicate are shown. In C, quantitative data of cell viability in Aβ_25–35-treated cells with D609 are shown. Cell viability in Aβ-treated cells with C2- and C6-ceramide is shown in D and E, respectively. Data are the means ± S.D. of 3 independent experiments performed in duplicate or triplicate. Effects of tested inhibitors of CCEs and their concentration-dependency on cell viability are shown in Table 1 and Supplementary Fig. 1.

Table 1. Effects of Inhibitors of CCEs on Cell Viability with and without Aβ_25–35

	Vehicle	20 µM Aβ25–35
	Cell viability (% of control)
Exp. I.
Vehicle (12)	100	50.3 ± 7.6a)
10 µM PPMP (3)	43.6 ± 19.2a)	23.4 ± 14.8
40 µM Eliglustat (3)	74.3 ± 14.4a)	36.7 ± 4.3
300 µM NB-DNJ (3)	102 ± 9	49.4 ± 6.8
10 µM NVP-231 (3)	63.4 ± 8.6a)	26.2 ± 13.7
Exp. II.
Vehicle (3)	100	45.6 ± 11.3a)
20 µM Ceranib-2 (3)	69.5 ± 13.7a)	25.3 ± 8.7
Exp. III.
Vehicle (3)	100	48.6 ± 3.5a)
20 µM D609 (3)	114 ± 11	68.2 ± 3.3b)

Cells were treated with vehicle, 20 µM Aβ_25–35, or respective inhibitors at the indicated concentrations for 24 h, and then cell viability was examined by the WST-8 method. a) p < 0.05, significantly different from the control without Aβ_25–35. b) p < 0.05, significantly different from the value without D609 in Aβ-treated cells. The concentration-dependent effects of each inhibitor are shown in Supplementary Fig. 1.

All inhibitors tested are theoretically considered to up-regulate cellular ceramide levels. Many reports, including ours, demonstrated that extracellular application of ceramides having shorter N-acyl chains, such as C2- and C6-ceramide, was cytotoxic to several cell types.^4,6,25) Indeed, extracellular application of C2- and C6-ceramide exhibited cytotoxicity with and without Aβ_25–35 (Figs. 1D, E). Although D609 exerted protective effects against Aβ_25–35-induced cytotoxicity, the decrease in cell viability by 20 µM C6-ceramide was not affected by co-treatment with 1 µM D609; the values were 45 ± 7 and 49 ± 6% (n = 3) in vehicle- and D609-treated cells, respectively.

Aβ_25–35-Induced Decrease in NBD-GlcCer Formation in Cells

Next, changes in ceramide metabolism by Aβ_25–35 treatment were examined using fluorescently labeled C6-ceramide, NBD-ceramide. HFL-1 cells were treated with Aβ_25–35 for 3 h and then treated with 10 µM NBD-ceramide for 30 min. NBD-ceramide metabolites in the lipids extracted were separated by the TLC method (Fig. 2A). Treatment with 20 µM Aβ_25–35 slightly inhibited the formation of NBD-GlcCer. Next, to analyze the effects of cytokines include in serum, we examined the Aβ response in the absence of serum (Fig. 2B). Treatment with Aβ_25–35 from 5 µM up to 20 µM significantly inhibited the formation of NBD-GlcCer (Fig. 2C). In the presence and absence of serum, Aβ_25–35 did not affect the formation of NBD-SM or NBD-C1P (Supplementary Figs. 2A, B). Treatment with 20 µM Aβ_25–35 for 6 h also reduced the formation of NBD-GlcCer (Supplementary Figs. 2C, D), but not NBD-SM/C1P (data not shown), in cells cultured with serum. This suggested that Aβ_25–35 inhibited the formation of GlcCer in cells.

Fig. 2. Reduced Formation of NBD-GlcCer in Aβ_25–35-Treated Cells

Cells were treated with Aβ_25–35 for 3 h in the presence and absence of serum. Then, cells were treated with 10 µM NBD-ceramide for 30 min, and the lipids including NBD-GlcCer were extracted and separated by the TLC method. In A and B, typical images of NBD-ceramide metabolites in cells with and without serum are shown, respectively. In C, quantitative data for NBD-GlcCer levels in Aβ-treated cells with (○) and without (△) serum are shown. Data are the means ± S.D. of 3 independent experiments performed in duplicate. The levels of other metabolites are shown in Supplementary Fig. 2. ANOVA with Dunnett’s test were used for multiple comparisons. * p < 0.05, significantly different from the control.

Reversal of Aβ_25–35-Induced Cytotoxicity and Inhibition of GCS by D609 Treatment

Aβ_25–35 reduced both cell viability and GCS activity, and the formation of NBD-GlcCer, whereas D609, a putative inhibitor of SMS, reduced the cytotoxicity induced by Aβ_25–35. Thus, we next investigated the effects of D609 on NBD-ceramide metabolism with and without Aβ_25–35 (Fig. 3). Treatment with 5 µM D609 for 3 h alone did not affect NBD-ceramide metabolism, including the formation of NBD-SM in the absence (Fig. 3A) and presence (data not shown) of serum. Of note, D609 treatment canceled the decrease in NBD-GlcCer formation induced by 14 µM Aβ_25–35 (Fig. 3B). D609 also canceled the Aβ_25–35-induced decrease in cells cultured with serum (Fig. 3C and Supplementary Fig. 3A). In the presence and absence of serum, the levels of NBD-SM and NBD-C1P were not affected by Aβ_25–35 or D609 (Fig. 3A, Supplementary Fig. 3). Treatment with 1 µM D609 also restored the formation of NBD-GlcCer in Aβ_25–35-treated cells; 98 ± 12 (% of control) and 81 ± 3% with and without D609, respectively (n = 3).

Fig. 3. Restoration of the Aβ_25–35-Induced Decrease of NBD-GlcCer Formation by D609 Treatment

Cells were treated with 14 µM Aβ_25–35 for 3 h in the absence (A, B) or presence (C) of serum. Five micro mol D609 was further supplemented. Then, NBD-ceramide metabolism in cells was examined. In A, typical images of NBD-ceramide metabolites in cells without serum are shown. Quantitative data for NBD-GlcCer formation in cells without and with serum are shown in B and C, respectively. Data are the means ± S.D. of 3 independent experiments. Typical images of NBD-ceramide metabolites in cells with serum and the levels of other metabolites are shown in Supplementary Fig. 3. ANOVA with Dunnett’s test were used for multiple comparisons. * p < 0.05, significantly different from control.

Increased Aβ_25–35-Induced Cytotoxicity in HeLa Cells Genetically Lacking GCS Function

Our study thus far suggested that pharmacological inhibition of GCS alone causes cytotoxicity, and Aβ_25–35 reduces GCS activity and induces cytotoxicity in fibroblasts. In order to further clarify the roles of GCS in cell viability, we examined the effects of Aβ_25–35 on cytotoxicity in another cell line, human cervical carcinoma HeLa cells. Two types of HeLa cells, native cells (CGS-WT cells) and its mutant genetically lacking GCS function (GCS-KO cells), were used. Under our conditions, cell growth of GCS-KO cells was slightly slower than that of GCS-WT cells in the presence of serum, but marked changes, including in the ratio of non-adhesive cells and cell morphology, were not observed. Cell viability examined for 24 h was slightly, but not significantly, lower in GCS-KO cells than in GCS-WT cells (Fig. 4A, left two columns). Substances/components in serum and of cells are considered to supply and compensate GlcCer and its metabolites directly and/or indirectly in GCS-KO cells. In the absence of serum, cell viability of GCS-KO cells was significantly lower than that of GCS-WT cells (Fig. 4A, right two columns). Treatment of GCS-WT and GCS-KO cells with Aβ_25–35 reduced cell viability after 24 h (Fig. 4B). The cytotoxicity induced by Aβ_25–35 was higher in GCS-KO cells and the response to 30 µM Aβ_25–35 was significantly potentiated in GCS-KO cells compared with that in GCS-WT cells. In the presence of serum, Aβ_25–35-induced cytotoxicity was lower, and the values of cell viability with 20 and 30 µM Aβ_25–35 were approximately 90% and 80%, respectively, in both cell lines. This suggested that Aβ_25–35 exhibits cytotoxicity in the absence of GCS and that GCS activity has protective effects against Aβ-induced cytotoxicity.

Fig. 4. Effects of GCS Knockout on Aβ_25–35-Induced Cytotoxicity in HeLa Cells HeLa-GCS-WT Cells (white Column and ○) and HeLa-CGS-KO Cells (Black Column and ●) Were Cultured in the Presence of 10% Serum to Sub-confluency, from 70 to 80% Confluence

In A, the cells were cultured for an additional 24 h in medium with (left two columns) and without (right two columns) serum, and cell viability was measured. The absolute values of WST-8 absorbance are shown. In B, the cells were treated with the indicated concentrations of Aβ_25–35 for 24 h in the absence of serum. Data are expressed as percentages of viability of each cell line without Aβ_25–35. Data are the means ± S.D. of 3 independent experiments performed in duplicate. Student’s two-tailed t-test were used for pair-wise comparisons. * p < 0.05, significantly different from WT.

DISCUSSION

In the present study, we investigated the effects of Aβ_25–35 on the activities of 3 enzymes to metabolize a common substrate ceramide, GCS, SMSs, and CerK. Treatment of fibroblasts with Aβ_25–35 for 3 and 6 h significantly inhibited the activity of GCS, and induced cytotoxicity after 24 h. The Aβ_25–35-induced responses were ameliorated by D609 treatment. GCS activity is considered to act protectively against Aβ_25–35-induced cytotoxicity.

A Proposed Model of Aβ_23–35-Induced Cytotoxicity via the Inhibition of GCS

Based on our study and previously reported results, we propose a model of the protective action of GCS against Aβ_25–35-induced cytotoxicity (Fig. 5). Up-regulation of ceramide levels induces cytotoxicity, as demonstrated in the present and previous studies.^5,8,25) Aβ treatment up-regulates cellular ceramide levels via the activation of SMases and de novo synthesis of ceramide.^7–10) In the present study, we revealed the involvement of GCS inhibition in Aβ_25–35-induced cytotoxicity. D609, whose mechanisms are unclear, canceled the inhibition of GCS activity and decrease in cell viability induced by Aβ_25–35 treatment. The cytotoxicity induced by Aβ_25–35 treatment was increased by GCS knockout. Treatment with pharmacological inhibitors of GCS, PPMP and eliglustat, alone caused cytotoxicity. In conclusion, GCS activity is considered to play a protective role against Aβ_25–35-induced cytotoxicity by converting ceramide to GlcCer. Marks et al.²⁶⁾ reported that a significant decrease in GCS in the frontal cortex of patients with AD and treatment of primary neurons with an inhibitor of GCS induced an increase in long-chain ceramides parallel to the loss of viability. Our study and their findings suggest the role of GCS dysfunction in damage in cells/neurons in AD.

Fig. 5. A Proposed Model of the Protective Role of GCS Activity against Aβ_23–35-Induced Cytotoxicity

Protective Effects of D609 Treatment against Aβ_25–35-Induced Cytotoxicity

D609 is reported to inhibit SMS activity in vitro (both 1 and 2 types, IC₅₀ = 375 µM),²²⁾ thus we examined possible effects of D609 as a modulator of SL metabolism on cell viability with and without Aβ_25–35. Although the inhibition of SMSs is expected to up-regulate cellular levels of ceramide, a cytotoxic molecule, treatment with D609 alone did not affect cell viability. Interestingly, treatment with D609 at concentrations less than 10 µM significantly attenuated the Aβ_25–35-induced cytotoxicity. Although AβPP/Aβ metabolism, including activities of β/γ-secretases, is regulated by the levels of lipids, including SM, in rafts of the plasma membrane (PM)²⁾ and by D609 treatment in vivo,²⁷⁾ the Aβ_25–35-induced responses were reversed by D609 treatment, suggesting no involvement of Aβ metabolism. The Aβ peptides, including Aβ_25–35, directly bind with SM/cholesterol in the PM, and form pores and/or oligometric channels.^1–3) Thus, changes in SM levels may affect the Aβ_25–35 response, whereas D609 at the used concentrations did not affect the formation of NBD-SM under our conditions. Thus, D609 appeared to modulate Aβ_25–35-induced responses in SL metabolism-independent manner. D609 acts as an antioxidant in vitro and in vivo,^22,27) and the reagent at concentrations higher than 25 µM exhibited antioxidant activity against Aβ-induced responses.²⁸⁾ Considering the used concentrations of D609, SM- and antioxidation-related events of D609 were not involved in the D609-induced protection against Aβ_25–35-induced responses. Many effects of D609 were previously reported.²²⁾ The mechanisms underlying the D609-induced cancelation of Aβ_25–35-induced GCS inhibition should be elucidated in the future.

Possible Protective Mechanisms by GCS Activity against Aβ-Induced Cytotoxicity

How does GCS activity maintain cell viability in Aβ-treated cells? GCS activity may maintain cell viability via two mechanisms: the downregulation of toxic ceramide and up-regulation of putative protective molecules, GlcCer and its metabolites, including gangliosides. The activity of GCS is positively correlated with multidrug resistance protein 1 gene expression, which encodes P-glycoprotein-1 (P-gp, alias ABCB1, one of ATP-binding cassette (ABC) efflux transporters) that increases the efflux of intracellular ceramide into extracellular spaces.²⁹⁾ Thus, the inhibition of GCS may synergistically up-regulate ceramide levels in cells. In addition, Aβ_1–42 treatment reduced the expression of many ABC transporters, including ABCB1, in primary brain endothelial cells.¹⁶⁾ Thus, Aβ_25–35 treatment may increase ceramide levels via multiple pathways. Glyco-SLs, including GlcCer and gangliosides, have protective roles against cytotoxicity via multiple pathways for cell survival.^11,26,30) Knockdown and inhibition of GCS induced autophagy via significant attenuation of the AKT-mammalian target of rapamycin (mTOR) signaling pathway, a key negative regulator of autophagy, in neurons.³¹⁾ Although the decrease in cell viability by Aβ_25–35 (and serum withdrawal) was greater in HeLa-GCS-KO cells than in HeLa-GCS-WT cells, the changes by GCS knockout were not marked in the present study. Thus, the increase in ceramide plays a major role in the cytotoxicity induced by GCS inhibition, and the probable decreases in GlcCer and glyco-SLs levels may play supportive roles in Aβ_25–35-induced cytotoxicity. Moreover, in this study, we analyzed Aβ_25–35-induced cytotoxicity using fibroblast and HeLa cells. Aβ treatment caused multiple types of cellular responses in fibroblasts and HeLa cells as well as neuronal cells.¹³⁾ However, components of ceramide and pattern of ceramide species are different between cell lines.³²⁾ Further studies will help us to understand the relationship of Aβ_25–35-induced cytotoxicity and component of ceramide in various cells and tissues.

Ineffectiveness of NB-DNJ, an Inhibitor of GCS, against Aβ_25–35-Induced Cytotoxicity

NB-DNJ (alias miglustat, Zavesca™, Actelion Pharm. Ltd.), an iminosugar-based inhibitor of GCS, is approved as a therapeutic for Gaucher’s disease and Niemann–Pick type C disease including in the EU and Japan. In the present study, however, NB-DNJ, unlike PPMP and eliglustat, did not exhibit cytotoxicity. NB-DNJ was reported to inhibit the activities of β-glucosidases, specifically type 2 (GBA2) that degrades GlcCer, and other sugar-related enzymes.^33,34) Treatment of Niemann–Pick type C disease mice with NB-DNJ prolonged survival but paradoxically increased brain GlcCer levels.³⁵⁾ Thus, treatment of cells with NB-DNJ, unlike PPMP/eliglustat, may not alter cellular ceramide levels from the normal range.

Possible Mechanisms for GCS Inhibition-Independent Cytotoxic Effects of Aβ and Reagents Tested

In addition to GCS inhibition, Aβ treatment may stimulate un-identified and/or other processes, leading to reduced cell viability. First, Aβ_25–35 treatment induced cytotoxicity in GCS-KO cells. Second, the peptide increased cytotoxicity in the PPMP/eliglustat-treated cells. As described in Results, PPMP and eliglustat at the used concentrations almost completely inhibited GCS activity or at least the formation of NBD-GlcCer. Third, the 20 µM Aβ_25–35-induced decrease in cell viability was greater than 50%, but the inhibition of GCS activity was small, at less than 25%. Aβ treatment was reported to activate multiple signaling pathways and responses, including mitochondrial toxicity/dysfunction, that are involved in cell fate.^{10,15,36–38)} It remains unclear which pathways/responses are directly altered by changes in GCS activity.

Treatments with PPMP and eliglustat alone exhibited cytotoxicity under our conditions. PPMP/PDMP inhibits several enzymes, including a variety of glycosidases,³⁴⁾ and these reagents alter the organization of late endosomes/lysosomes and the endoplasmic reticulum in a GCS inhibition-independent manner.^39,40) Eliglustat has been reported to inhibit GCS activity in vitro, with an IC₅₀ value of 20 nM,³⁴⁾ however, cell viability was reduced 25% by eliglustat at 40 µM, but not at 20 µM. Thus, PPMP and eliglustat may cause cytotoxicity through other un-identified processes, in addition to GCS inhibition. Ceranib-2 alone exhibited cytotoxicity. As described in the Introduction, S1P directly and/or via the S1P receptor functions as a survival factor or anti-apoptotic molecule. Although we were unable to confirm the inhibition of ceramidase(s) by ceranib-2 because of the limitation of the TLC method, a possible decrease in sphingosine levels by ceramib-2 may cause cytotoxicity via the reduced formation of S1P. NVP-231 alone exhibited cytotoxicity. We reported that NVP-231 at less than 1 µM inhibited the formation of NBD-C1P in cells.²¹⁾ C1P has multiple cellular functions regulating cell fate,⁴¹⁾ thus reduced formation of C1P by NVP-231 may induce cytotoxicity. The intracellular levels of sphingosine and C1P, which are produced by ceramidase(s) and CerK, respectively, were reported to be 1/100 approx. 1/10,000 those of ceramide.^4,21,41) Thus, cytotoxic effects of ceranib-2 and NVP-231 were considered to be mediated by the sphingosine/S1P and C1P pathways, respectively, not by ceramide levels.

Acknowledgments

We thank Dr. K. Hanada (National Institute of Infectious Disease, Japan) and Prof. A. Nishida (Chiba University, Japan) for providing the two HeLa cell lines (GCS-WT and GCS-KO cells) and NBD-ceramide, respectively.

Author Contributions

Study Design, TH and TM; Conducted experiments and performed data analysis, ZT, KM and TH; Wrote or contributed to the writing of the manuscript, TH, HN and TM.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

The online version of this article contains supplementary materials.

REFERENCES

• 1) Di Scala C, Chahinian H, Yahi N, Garmy N, Fantini J. Interaction of Alzheimer’s β-amyloid peptides with cholesterol: mechanistic insights into amyloid pore formation. Biochemistry, 53, 4489–4502 (2014).
• 2) Chen GF, Xu TH, Yan Y, Zhou YR, Jiang Y, Melcher K, Xu HE. Amyloid beta: structure, biology and structure-based therapeutic development. Acta Pharmacol. Sin., 38, 1205–1235 (2017).
• 3) Smith AK, Khayat E, Lockhart C, Klimov DK. Do cholesterol and sphingomyelin change the mechanism of Aβ_25–35 peptide binding to zwitterionic bilayer? J. Chem. Inf. Model., 59, 5207–5217 (2019).
• 4) Hannun YA, Obeid LM. Principles of bioactive lipid signaling: lessons from sphingolipids. Nat. Rev. Mol. Cell Biol., 9, 139–150 (2008).
• 5) Hannun YA, Obeid LM. Sphingolipids and their metabolism in physiology and disease. Nat. Rev. Mol. Cell Biol., 19, 175–191 (2018).
• 6) Alessenko AV, Albi E. Exploring sphingolipid implications in neurodegeneration. Front. Neurol., 11, 437 (2020).
• 7) Ayasolla K, Khan M, Singh AK, Singh I. Inflammatory mediator and β-amyloid (25–35)-induced ceramide generation and iNOS expression are inhibited by vitamin E. Free Radic. Biol. Med., 37, 325–338 (2004).
• 8) Czubowicz K, Jęśko H, Wencel P, Lukiw WJ, Strosznajder RP. The role of ceramide and sphingosine-1-phosphate in Alzheimer’s disease and other neurodegenerative disorders. Mol. Neurobiol., 56, 5436–5455 (2019).
• 9) Malaplate-Armand C, Florent-Béchard S, Youssef I, Koziel V, Sponne I, Kriem B, Leininger-Muller B, Olivier JL, Oster T, Pillot T. Soluble oligomers of amyloid-β peptide induced neuronal apoptosis by activating a cPLA₂-dependent sphingomyelinase–ceramide pathway. Neurobiol. Dis., 23, 178–189 (2006).
• 10) Zeng C, Lee JT, Chen H, Chen S, Hsu CY, Xu J. Amyloid-β peptide enhances tumor necrosis factor-α-induced iNOS through neutral sphingomyelinase/ceramide pathway in oligodendrocytes. J. Neurochem., 94, 703–712 (2005).
• 11) Sokolova T, Zakharova IO, Furaev VV, Rychkova MP, Avrova NF. Neuroprotective effect of ganglioside GM1 on the cytotoxic action of hydrogen peroxide and amyloid β-peptide in PC12 cells. Neurochem. Res., 32, 1302–1313 (2007).
• 12) Khan TK, Alkon DL. Peripheral biomarkers of Alzheimer’s disease. J. Alzheimers Dis., 44, 729–744 (2015).
• 13) Palotás A, Kálmán J, Palotás M, Kemény L, Janka Z, Penke B. Long-term exposition of cells to β-amyloid results in decreased intracellular calcium concentration. Neurochem. Int., 42, 543–547 (2003).
• 14) Bai JZ. β-Amyloid peptides ^1–40βA and ^25–35βA suppress human amylin-mediated death of RINm5F islet β-cells with distinct actions on fibril formation. Cell Biol. Int., 32, 447–455 (2008).
• 15) Fonseca AC, Ferreiro E, Oliveira CR, Cardoso SM, Pereira CF. Activation of the endoplasmic reticulum stress response by the amyloid-beta 1–40 peptide in brain endothelial cells. Biochim. Biophys. Acta, 1832, 2191–2203 (2013).
• 16) Shubbar MH, Penny JI. Effect of amyloid beta on ATP-binding cassette transporter expression and activity in porcine brain microvascular endothelial cells. Biochim. Biophys. Acta, Gen. Subj., 1862, 2314–2322 (2018).
• 17) Pavliukeviciene B, Zentelyte A, Jankunec M, Valiuliene G, Talaikis M, Navakauskiene R, Niaura G, Valincius G. Amyloid β oligomers inhibit growth of human cancer cells. PLOS ONE, 14, e0221563 (2019).
• 18) Bu XL, Cao GQ, Shen LL, Xiang Y, Jiao SS, Liu YH, Zhu C, Zeng F, Wang QH, Wang YR, He Y, Zhou HD, Wang YJ. Serum amyloid-β levels are increased in patients with chronic obstructive pulmonary disease. Neurotox. Res., 28, 346–351 (2015).
• 19) Yamaji T, Hanada K. Establishment of HeLa cell mutants deficient in sphingolipid-related genes using TALENs. PLOS ONE, 9, e88124 (2014).
• 20) Tada E, Toyomura K, Nakamura H, Sasaki H, Saito T, Kaneko M, Okuma Y, Murayama T. Activation of caramidase and ceramide kinase by vanadate via a tyrosine kinase-mediated pathway. J. Pharmacol. Sci., 114, 420–432 (2010).
• 21) Hori M, Gokita M, Yasue M, Honda T, Kohama T, Mashimo M, Nakamura H, Murayama T. Down-regulation of ceramide kinase via proteasome and lysosome pathways in PC12 cells by serum withdrawal: its protection by nerve growth factor and role in exocytosis. Biochim. Biophys. Acta. Mol. Cell Res., 1867, 118714 (2020).
• 22) Adibhatla RM, Hatcher JF, Gusain A. Tricyclodecan-9-yl-xanthogenate (D609) mechanism of actions: a mini-review of literature. Neurochem. Res., 37, 671–679 (2012).
• 23) Hashimoto N, Matsumoto I, Takahashi H, Ashikawa H, Nakamura H, Murayama T. Cholesterol-dependent increases in glucosylceramide synthase activity in Niemann–Pick disease type C model cells: abnormal trafficking of endogenously formed ceramide metabolites by inhibition of the enzyme. Neuropharmacology, 110 (Pt. A), 458–469 (2016).
• 24) Pastukhov O, Schwalm S, Zangemeister-Wittke U, Fabbro D, Bornancin F, Japtok L, Kleuser B, Pfeilschifter J, Huwiler A. The ceramide kinase inhibitor NVP-231 inhibits breast and lung cancer cell proliferation by inducing M phase arrest and subsequent cell death. Br. J. Pharmacol., 171, 5829–5844 (2014).
• 25) Honda T, Motoyoshi K, Kasahara J, Yamagata K, Takahashi H, Nakamura H, Murayama T. Tyrosine-phosphorylation and activation of glucosylceramide synthase by v-Src: its role in survival of HeLa cells against ceramide. Biochim. Biophys. Acta. Mol. Cell Biol. Lipids, 1866, 158817 (2021).
• 26) Marks N, Berg MJ, Saito M, Saito M. Glucosylceramide synthase decrease in frontal cortex of Alzheimer brain correlates with abnormal increase in endogenous ceramides: consequences to morphology and viability on enzyme suppression in cultured primary neurons. Brain Res., 1191, 136–147 (2008).
• 27) Yang H, Xie Z, Wei L, Ding M, Wang P, Bi J. Glutathione-mimetic D609 alleviates memory deficits and reduces amyloid-β deposition in an AβPP/PS₁ transgenic mouse model. Neuroreport, 29, 833–838 (2018).
• 28) Sultana R, Newman SF, Abdul HM, Cai J, Pierce WM, Klein JB, Merchant M, Butterfield DA. Protective effect of D609 against amylioid-beta_1–42-induced oxidative modification of neuronal proteins: redox proteomics study. J. Neurosci. Res., 84, 409–417 (2006).
• 29) Lee WK, Torchalski B, Kohistani N, Thévenod F. ABCB1 protects kidney proximal tubule cells against cadmium-induced apoptosis: roles of cadmium and ceramide transport. Toxicol. Sci., 121, 343–356 (2011).
• 30) Wegner MS, Schömel N, Gruber L, Örtel SB, Kjellberg MA, Mattjus P, Kurz J, Trautmann S, Peng B, Wegner M, Kaulich M, Ahrends R, Geisslinger G, Grösch S. UDP-glucose ceramide glucosyltransferase activates AKT, promoted proliferation, and doxorubicin resistance in breast cancer cells. Cell. Mol. Life Sci., 75, 3393–3410 (2018).
• 31) Shen W, Henry AG, Paumier KL, Li L, Mou K, Dunlop J, Berger Z, Hirst WD. Inhibition of glucosylceramide synthase stimulates autophagy flux in neurons. J. Neurochem., 129, 884–894 (2014).
• 32) Valsecchi M, Mauri L, Casellato R, Prioni S, Loberto N, Prinetti A, Chigorno V, Sonnino S. Ceramide and sphingomyelin species of fibroblasts and neurons in culture. J. Lipid Res., 48, 417–424 (2007).
• 33) Ridley CM, Thur KE, Shanahan J, Thillaiappan NB, Shen A, Uhl K, Walden CM, Rahim AA, Waddington SN, Platt FM, van der Spoel AC. β-Glucosidase 2 (GBA2) activity and imino sugar pharmacology. J. Biol. Chem., 288, 26052–26066 (2013).
• 34) Shayman JA, Larsen SD. The development and use of small molecule inhibitors of glycosphingolipid metabolism for lysosomal storage diseases. J. Lipid Res., 55, 1215–1225 (2014).
• 35) Nietupski JB, Pacheco JJ, Chuang WL, Maratea K, Li L, Foley J, Ashe KM, Cooper CG, Aerts JM, Copeland DP, Scheule RK, Cheng SH, Marshall J. Iminosugar-based inhibitors of glucosylceramide synthase prolong survival but paradoxically increase brain glucosylceramide levels in Niemann–Pick C mice. Mol. Genet. Metab., 105, 621–628 (2012).
• 36) Wang Y, Liu L, Hu W, Li G. Mechanism of soluble beta-amyloid 25–35 neurotoxicity in primary cultured rat cortical neurons. Neurosci. Lett., 618, 72–76 (2016).
• 37) Cheng W, Chen W, Wang P, Chu J. Asiatic acid protects differentiated PC12 cells from Aβ_25–35-induced apoptosis and tau hyperphosphorylation via regulating PI3K/Akt/GSK-3β signaling. Life Sci., 208, 96–101 (2018).
• 38) Sharoar MG, Islam MI, Shahnawaz M, Shin SY, Park IS. Amyloid β binds procaspase-9 to inhibit assembly of Apaf-1 apoptosome and intrinsic apoptosis pathway. Biochim. Biophys. Acta. Mol. Cell Res., 1843, 685–693 (2014).
• 39) Nakamura M, Kuroiwa N, Kono Y, Takatsuki A. Glucosylceramide synthesis inhibitors block pharmacologically induced dispersal of the Golgi and anterograde membrane flow from the endoplasmic reticulum: implication of sphingolipid metabolism in maintenance of the Golgi architecture and anterograde membrane flow. Biosci. Biotechnol. Biochem., 65, 1369–1378 (2001).
• 40) Sprocati T, Ronchi P, Raimondi A, Francolini M, Borgese N. Dynamic and reversible restructuring of the ER induced by PDMP in cultured cells. J. Cell Sci., 119, 3249–3260 (2006).
• 41) Presa N, Gomez-Larrauri A, Domingues-Herrera A, Trueba M, Gomez-Muñoz A. Novel signaling aspects of ceramide 1-phosphate. Biochim. Biophys. Acta. Mol. Cell Biol. Lipids, 1865, 158630 (2020).