Titanium dioxide and carbon black nanoparticles disrupt neuronal homeostasis via excessive activation of cellular prion protein signaling

Abstract

Background: Epidemiological emerging evidence shows that human exposure to some nanosized materials present in the environment would contribute to the onset and/or progression of Alzheimer's disease (AD). The cellular and molecular mechanisms whereby nanoparticles would exert some adverse effects towards neurons and take part in AD pathology are nevertheless unknown.

Results: Here, we provide the prime evidence that titanium dioxide (TiO₂) and carbon black (CB) nanoparticles (NPs) bind the cellular form of the prion protein (PrP^C), a plasma membrane protein well known for its implication in prion diseases and prion-like diseases, such as AD. The interaction between TiO₂- or CB-NPs and PrP^C at the surface of neuronal cells grown in culture corrupts PrP^C signaling function. This triggers PrP^C-dependent activation of NADPH oxidase and subsequent production of reactive oxygen species (ROS) that alters redox equilibrium. Through PrP^C interaction, NPs also promote the activation of 3-phosphoinositide-dependent kinase 1 (PDK1), which in turn provokes the internalization of the neuroprotective TACE α-secretase. This diverts TACE cleavage activity away from (i) TNFα receptors (TNFR), whose accumulation at the plasma membrane augments the vulnerability of NP-exposed neuronal cells to TNFα -associated inflammation, and (ii) the amyloid precursor protein APP, leading to overproduction of neurotoxic amyloid Aβ40/42 peptides. The silencing of PrP^C or the pharmacological inhibition of PDK1 protects neuronal cells from TiO₂- and CB-NPs effects regarding ROS production, TNFα hypersensitivity, and Aβ rise. Finally, we show that dysregulation of the PrP^C-PDK1-TACE pathway likely occurs in the brain of mice injected with TiO₂-NPs by the intra-cerebro-ventricular route as we monitor a rise of TNFR at the cell surface of several groups of neurons located in distinct brain areas.

Conclusion: Our in vitro and in vivo study thus posits for the first time normal cellular prion protein PrP^C as being a neuronal receptor of TiO₂- and CB-NPs and identifies PrP^C-coupled signaling pathways by which those nanoparticles alter redox equilibrium, augment the intrinsic sensitivity of neurons to neuroinflammation, and provoke a rise of Aβ peptides. By identifying signaling cascades dysregulated by TiO₂- and CB-NPs in neurons, our data shed light on how human exposure to some NPs might be related to AD.

Keywords: Alzheimer’s disease; Aβ peptides; Nanoneurotoxicity; Nanoparticles; Neuroinflammation; PrPC receptor; Signaling; TNFα receptors.

Fig. 1

TiO₂ and CB nanoparticles interact with PrP^C. a Schematic of the experimental procedure to assess the interaction between full-length recombinant mouse PrP^C (recPrP^C, 2 µM) and TiO₂- or CB-NPs (0–80 µg ml⁻¹) exploiting the intrinsic fluorescence of PrP^C or by western blotting (WB). b Fluorescence titration curves showing the binding of TiO₂- or CB-NPs to full-length mouse recPrP^C. The quantity of recPrP^C bound with nanoparticles was deduced by subtracting the fluorescence level of titrated free PrP^C to the fluorescence level of total PrP^C measured in the absence of nanoparticles. Fitting hyperbolic curves were calculated with the help of the Kaleidagraph Software (Abelbeck Software). c Representative Western-blot and quantification histogram showing decrease of free recPrP^C amount in the supernatant of the centrifuged reaction medium between recPrP^C and TiO₂- or CB-NPs. The experiments were performed three times in triplicates. Values are means ± SEM. *denotes p < 0.05, ***p < 0.001, ****p < 0.0001 versus recPrP^C incubated without nanoparticles

Fig. 2

PrP^C facilitates nanoparticle interaction with plasma membrane of 1C11 precursors and 1C11^5−HT neuronal cells. a, b Transmission Electron Microscopy experiments showing large aggregates of TiO₂- and CB-NPs (white arrows) within 1C11 precursor cells and 1C11^5−HT neuronal cells a and small aggregates of TiO₂- and CB-NPs (red arrows) at the plasma membrane of 1C11 and 1C11^5−HT cells b after 24 h exposure to 10 µg cm⁻² nanoparticles. Scale bar = 2 µm in a. Scale bar = 0.5 µm in b. c Light scattering-based FACS analysis of interacting TiO₂-NPs with the surface of 1C11 and PrP^null-1C11 cells exposed up to 10 µg cm⁻² nanoparticles. (Left) Representative bright-field images of 1C11 and PrP^null-1C11 cells exposed to 5 µg cm⁻² TiO₂-NPs merged with the light scattering signal of TiO₂-NPs (pink signal) adsorbed to the plasma membrane after internal masking of cells (see Methods). (Right) Quantification histogram of TiO₂-NPs present at the surface of 1C11 and PrP^null-1C11 cells. Scale bar = 5 µm. The experiments were performed three times in triplicates. Values are means ± SEM. *denotes p < 0.05 versus 1C11 cells exposed to nanoparticles

Fig. 3

TiO₂- and CB-NPs specifically interact with full-length PrP^C in the 1C11 neuronal cell line. a Schematic of the experimental procedure to assess the interaction of TiO₂- and CB-NPs with PrP^C expressed at the plasma membrane of 1C11 precursor cells and their serotonergic 1C11^5−HT neuronal progenies by western blotting. b Quantification histogram deduced from Western-blot experiments showing the amounts of full-length (FL) PrP^C (left) and C1 fragment (right) in the supernatant (S fraction) obtained after lysis of 1C11 cells exposed to increasing concentrations of TiO₂- or CB-NPs (0 to 10 µg cm⁻²) for 15 min and centrifugation of the lysates. c, d Representative Western-blots and quantification histograms showing time-variation in the level of FL PrP^C and C1 fragment in the S and P fractions derived from 1C11 c and 1C11^5−HT d cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²). α-tubulin was used for normalization of PrP^C level in the S fraction. The experiments were performed three times in triplicates. Values are means ± SEM. *denotes p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 versus unexposed cells

Fig. 4

TiO₂ and CB nanoparticle interaction with PrP^C promotes NADPH oxidase-dependent ROS production. a Kinetics of ROS production induced on exposure to TiO₂- and CB-NPs (1 µg cm⁻²) in 1C11 and 1C11^5−HT neuronal cells. b Involvement of NADPH oxidase in NP-induced ROS production (1 µg cm⁻² for 2 h) using the NADPH oxidase inhibitor apocynin (500 µM). c Quantification histogram showing that siRNA-based PrP^C silencing (siPrP) abrogates ROS production induced by TiO₂- or CB-NPs (1 µg cm⁻² for 2 h) in 1C11 and 1C11^5−HT cells. d GSH level keeps constant in 1C11 and 1C11^5−HT neuronal cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) up to 24 h. The experiments were performed three times in triplicates. Values are means ± SEM. *denotes p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 versus unexposed cells

Fig. 5

TiO₂ and CB nanoparticle interaction with PrP^C renders 1C11 and 1C11^5−HT neuronal cells highly sensitive to TNFα insult by promoting TNFR1 accumulation at the plasma membrane. a Kinetics of TNFR1 rise at the plasma membrane of 1C11 cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) and related quantification histogram of immunostained TNFR1. Scale bar = 10 µm. b Viability and c caspase-3 activation in 1C11 cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) in combination or not with TNFα (200 ng ml⁻¹). d TNFR1 expression level as assessed by RT-qPCR and Western-blotting in 1C11 cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) for 1 h. Vinculine was used for normalization. The experiments were performed three times in triplicates. Values are means ± SEM. *denotes p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 versus unexposed cells

Fig. 6

TiO₂ and CB nanoparticle interaction with PrP^C promotes TACE internalization in a PDK1-dependent manner at the root of TNFR1 overexposure at the cell surface. a TACE immunostaining at the plasma membrane of 1C11 cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) for 1 h in the presence or not of a siRNA toward PrP^C (siPrP) or the PDK1 inhibitor, BX912 (1 µM) and related quantification histogram. Cell permeabilization with saponin (0.05%) shows TACE internalization in 1C11 cells exposed to nanoparticles. Scale bar = 10 µm. b, c TACE expression level as assessed by RT-qPCR b and Western-blotting c in 1C11 cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) for 1 h. d PDK1 phosphorylation status at Ser241 (p-PDK1) was assessed by Western-blotting in 1C11 cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) for 1 h. e TNFR1 immunostaining at the plasma membrane of 1C11 cells exposed for 1 h to TiO₂- or CB-NPs (1 µg cm⁻²) in the presence or not of a siRNA toward PrP^C (siPrP) or the PDK1 inhibitor, BX912 (1 µM) and related quantification histogram. Scale bar = 10 µm. The experiments were performed three times in triplicates. Values are means ± SEM. *denotes p < 0.05, **p < 0.01, ***p < 0.001 versus unexposed cells

Fig. 7

Cell exposure to TiO₂ and CB nanoparticles promotes rise of Aβ42 through corruption of the PrP^C-PDK1 pathway. a Kinetics of intracellular Aβ accumulation in 1C11 cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) and related quantification histogram of immunostained Aβ. Scale bar = 10 µm. b ELISA-based quantification of Aβ42 peptides in 1C11 and 1C11^5−HT neuronal cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) up to 24 h. c ELISA-based quantification of Aβ42 peptides in 1C11^5−HT neuronal cells exposed to TiO₂- or CB-NPs (1 µg cm⁻²) for 4 h in the presence or not of a siRNA toward PrP^C (siPrP) or the PDK1 inhibitor, BX912 (1 µM). The experiments were performed three times in triplicates. Values are means ± SEM. *denotes p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 versus unexposed cells

Fig. 8

TiO₂ nanoparticles trigger TNFR1 accumulation in vivo. a TNFR1 immunostaining on two representative brain sections from non-injected, sham-operated, and mice exposed to 10 µg of TiO₂-NPs delivered by an unilateral intra-cerebro-ventricular (ICV) injection 8 weeks before. TNFR1 was detected in the cortex (Cx), septum (Sp) striatum (St), hippocampus (Hip), thalamic (Th) nuclei, and substantia nigra (SN). b Higher magnification of the subicular cortex shows TNFR1 rise at the plasma membrane of neurons that is stronger in TiO₂-NPs-injected mice than in sham-operated mice (arrows). Scale bar = 25 µm