Nano-Infrared Imaging of Primary Neurons

Abstract

Alzheimer's disease (AD) accounts for about 70% of neurodegenerative diseases and is a cause of cognitive decline and death for one-third of seniors. AD is currently underdiagnosed, and it cannot be effectively prevented. Aggregation of amyloid-β (Aβ) proteins has been linked to the development of AD, and it has been established that, under pathological conditions, Aβ proteins undergo structural changes to form β-sheet structures that are considered neurotoxic. Numerous intensive in vitro studies have provided detailed information about amyloid polymorphs; however, little is known on how amyloid β-sheet-enriched aggregates can cause neurotoxicity in relevant settings. We used scattering-type scanning near-field optical microscopy (s-SNOM) to study amyloid structures at the nanoscale, in individual neurons. Specifically, we show that in well-validated systems, s-SNOM can detect amyloid β-sheet structures with nanometer spatial resolution in individual neurons. This is a proof-of-concept study to demonstrate that s-SNOM can be used to detect Aβ-sheet structures on cell surfaces at the nanoscale. Furthermore, this study is intended to raise neurobiologists' awareness of the potential of s-SNOM as a tool for analyzing amyloid β-sheet structures at the nanoscale in neurons without the need for immunolabeling.

Keywords: Alzheimer’s disease; O-PTIR; amyloid-beta; neuron; s-SNOM.

Figure 1

Sample preparation and characterization of intraneuronal Aβ aggregates. (a) Isolation of primary neurons. Cortex homogenate extracted from day 15 embryos was seeded on 100 nm gold-coated Si. After 19 days of growth and maturation in culture, neurons were fixed and air-dried. (b) Optical image of cultured APP/PS1 transgenic neurons on AFM substrate cultured for 19 days. A scanning electron micrograph of cultured neurons is shown in Figure S2. (c) Confocal micrograph of APP/PS1 transgenic neurons, where the blue channel corresponds to the neuronal marker Map2, and the red channel shows the post-synaptic marker Drebrin; Aβ42 labeled with 12F4 antibody is green. (d) Confocal micrograph of APP/PS1 transgenic neurons, where the green channel is used to map Aβ along the neurite labeled with Drebrin. (e,f) STED images of APP-KO neurons treated with Aβ(1-42), indicated by the red channel. (g) STED signal intensity profile of a single Aβ aggregate (shown in the middle panel) with a size of ca. 50 nm. Aβ size distribution is neurons is shown in Figure S3.

Figure 2

Fourier-transform infrared (µ-FTIR) point spectral analysis. (a) Schematic diagram of the working principle, penetration depth, and lateral resolution of µ-FTIR. For µ-FTIR reflectance geometry (right panel), the IR beam is focused on the sample, whose spot size is comparable to the illumination wavelength (diffraction-limited, 3–20 µm). The sampling depth is usually defined by the IR optical path inside the sample (~5–10 µm). (b) A bright-field visible-light image of cultured primary neurons. The white square in the digital insert shows the apertures used during the synchrotron-based µ-FTIR measurements. A white square indicates the aperture set to 8 × 8 μm² and indicates the background position; Crosses show the position for spectra acquisition. (c) Averaged and normalized IR absorption spectra from untreated APP/PS1 transgenic neurons (red), APP/PS1 transgenic neurons (green), and APP-KO (black). Each point spectrum is an average of 256 spectral scans. FTIR spectra were acquired from different neurons (10 to 15 cells per condition), *** corresponds to p < 0.001 (d) Averaged and normalized second derivatives of the absorbance spectra mentioned in (c) show changes in the intensity at the 1630 cm⁻¹ position (arrow). μ-FTIR spectra were acquired at 2 cm⁻¹ spectral resolution with 256 averages in transmission mode. Statistics: a paired t-test for comparing two groups, *** indicates a significant difference at p < 0.001.

Figure 3

Optical photothermal infrared (O-PTIR) imaging of primary neurons. (a) Schematic diagram of the working principle, penetration depth, and lateral resolution for O-PTIR. The O-PTIR instrument employs two co-propagating beams: a 532 nm visible probe beam (shown as green in the illustration) and a tunable IR pump beam (shown as light red in the cartoon). The photothermal response is detected as the partial intensity loss of green light in response to the absorption of a pulsed IR beam. Thus, the spatial resolution is enhanced to ~500 nm. In O-PTIR (right panel), visible light is used as a probe; visible light scattering evolution is connected to the sample’s volume expansion produced by IR illumination (pump). The visible light scattering is then translated to the spectral response, enabling an IR probe, whose lateral resolution, ~500 nm, is defined by the visible probe (in this case, green light). Although the elastic scattering is limited to the surface, the penetration depth of the O-PTIR is defined by the reach of the IR beam (a few microns). (b) Representative optical image of cultured primary neurons grown on the CaF₂ using 10× objective. (c) Zoom-in of the area indicated in (b) using 40× objective. Dots indicate representative O-PTIR spectra locations. (d) Averaged and normalized IR spectra were recorded from wild-type neurons (blue) and APP/PS1 transgenic neurons (red). Spectra were acquired at 2 cm⁻¹ spectral resolution with 50 averages in reflection mode. (e,f) O-PTIR maps were acquired at frequencies 1650 cm⁻¹ and 1630 cm⁻¹ from wild-type and APP/PS1 transgenic neurons, respectively. Ratio maps are used to locate β-sheet structures (1630 cm⁻¹) along the neurons. Mask was rendered using threshold set to 0.03 a.u. of photothermal amplitude using 1650 cm⁻¹ map; white line separates threshold. (Figure S4).

Figure 4

Scattering-type scanning near-field optical microscopy (s-SNOM) nanoimaging and nanospectroscopy. (a) Schematic diagram of the working principle, penetration depth, and lateral resolution for s-SNOM. In s-SNOM, a metallic AFM tip acts as an antenna for the incident IR radiation, whose scattering carries local information (or dielectric response) of the material’s surface (left panel). The tip-confined fields in s-SNOM are strongly attached to the metallic tip (evanescent fields); therefore, both lateral resolution and sampling depth are comparable to the tip radius: in this case, their values are ~25 nm and ~50 nm, respectively. (b–f) s-SNOM measurements of individual APP/PS1 transgenic neurons cultured on the AFM substrate: 10 × 10 µm² AFM topography map of individual neurons with the dots indicating representative locations for spectra acquisition. The topography amps are followed by corresponding reflectivity and optical phase (IR absorption) images acquired at 1656 cm⁻¹, 1630 cm⁻¹, and digital zoom-in of the area indicated by the red square in 1630 cm⁻¹ optical phase image to show amyloid patches on neuronal surface. (g–j) s-SNOM measurements of wild-type neurons cultured on the AFM substrate: 10 × 10 µm² AFM topography map of individual neurons, followed by the corresponding reflectivity image and optical phase (IR absorption) images acquired at 1656 cm⁻¹ and 1630 cm⁻¹. (k) Averaged and normalized s-SNOM nano-FTIR broadband spectra recorded from wild-type (blue) and APP/PS1 transgenic neurons (red). Vertical dashed lines indicate spectral shoulders at around 1656 cm⁻¹, 1630 cm⁻¹, and 1740 cm⁻¹. Each point spectrum is an average of 15–30 spectral scans. Error bars show standard deviation. s-SNOM spectra were acquired from different neurons (3 to 5 cells per genotype, using at least 3 independent preparations of neuronal cultures). Statistics: a paired t-test for comparing two groups; *** correspond to p < 0.001. (l–p) s-SNOM measurements of exosomes isolated from cultured APP/PS1 transgenic primary neurons: 1 × 1 µm² AFM topography map of exosomes deposited on AFM substrate, followed by the corresponding reflectivity image and optical phase (IR absorption) images acquired at 1656 cm⁻¹, 1630 cm⁻¹, and 1740 cm⁻¹. White arrows indicate lipid–protein-rich spots that were identified as exosomes.