Visualization and quantitative analysis of nanoparticles in the respiratory tract by transmission electron microscopy

Abstract

Nanotechnology in its widest sense seeks to exploit the special biophysical and chemical properties of materials at the nanoscale. While the potential technological, diagnostic or therapeutic applications are promising there is a growing body of evidence that the special technological features of nanoparticulate material are associated with biological effects formerly not attributed to the same materials at a larger particle scale. Therefore, studies that address the potential hazards of nanoparticles on biological systems including human health are required. Due to its large surface area the lung is one of the major sites of interaction with inhaled nanoparticles. One of the great challenges of studying particle-lung interactions is the microscopic visualization of nanoparticles within tissues or single cells both in vivo and in vitro. Once a certain type of nanoparticle can be identified unambiguously using microscopic methods it is desirable to quantify the particle distribution within a cell, an organ or the whole organism. Transmission electron microscopy provides an ideal tool to perform qualitative and quantitative analyses of particle-related structural changes of the respiratory tract, to reveal the localization of nanoparticles within tissues and cells and to investigate the 3D nature of nanoparticle-lung interactions.This article provides information on the applicability, advantages and disadvantages of electron microscopic preparation techniques and several advanced transmission electron microscopic methods including conventional, immuno and energy-filtered electron microscopy as well as electron tomography for the visualization of both model nanoparticles (e.g. polystyrene) and technologically relevant nanoparticles (e.g. titanium dioxide). Furthermore, we highlight possibilities to combine light and electron microscopic techniques in a correlative approach. Finally, we demonstrate a formal quantitative, i.e. stereological approach to analyze the distributions of nanoparticles in tissues and cells.This comprehensive article aims to provide a basis for scientists in nanoparticle research to integrate electron microscopic analyses into their study design and to select the appropriate microscopic strategy.

Figure 1

Survey on several transmission electron microscopy strategies. The figure provides help to chose a specific strategy for TEM preparation based on the scientific purpose of the study. It shows the crucial decisions in specimen preparation from the fixation level to the investigation at the TEM. Abbreviations: GA = glutaraldehyde; PFA = paraformaldehyde; CTEM = conventional TEM; EFTEM = energy-filtered TEM; ET = electron tomography.

Figure 2

Chemical and physical fixation of the lung. Alveolar epithelial type II cells were studied by cTEM either after chemical (A and B) or after physical (C and D) fixation. The overview in A shows a well-preserved type II cell from a newborn rat lung fixed by instillation of 1.5% GA, 1.5% PFA in Hepes buffer and processed according to Table 1. Lamellar bodies (LB), nucleus (Nu) and mitochondria (Mt) are well preserved. At a higher magnification, details of the endoplasmic reticulum (ER) as well as an ER related multivesicular transport vesicle (MvTV) can be visualized. The overview in C shows a well-preserved type II cell from an adult rat lung. A small piece of tissue was cut from the whole lung, put in a syringe with 1-hexadecene and air was extracted from the tissue block by negative pressure. Afterwards, the specimen was high-pressure frozen (Leica EMPact 2.0, Leica, Vienna, Austria), freeze-substituted with acetone containing 1% osmium tetroxide (AFS 2.0, Leica, Vienna, Austria) and embedded in epoxy resin. Most likely due to the lack of uranyl acetate during freeze-substitution the lamellar bodies are not well preserved, with almost complete loss of the surfactant material, only the limiting membrane can be seen. However, the ultrastructure of other organelles like multivesicular bodies (MvB) is highly increased (D) due to the excellent preservation of the membrane structures (Me). Since this is the first description of high-pressure frozen lung tissue, systematic studies are needed to determine the ideal processing both for conventional and immuno TEM. Bars = 1 μm (A, C), 250 nm (B, D).

Figure 3

Conventional TEM of polystyrene nanoparticles. This figure demonstrates the impossibility to distinguish between NP and cellular structures by conventional TEM unambiguously. In A, five polystyrene NP (NP!) with a mean diameter of 78 nm are observed next to an A549 cell. Once taken up by the cells, they may have an appearance as shown in B. It is very likely that the spherical structures in B (NP?) are not NP but vesicular structures like caveolae. CC = Clathrin coated vesicle; PM = Plasma membrane; AJ = Adherens junction. Chemical fixation, Epon embedding, 40–70 nm sections. Bar = 1 μm (A and B are at identical magnification).

Figure 4

Immuno TEM of rat lung labeled for caveolin-1. Caveolae are cholesterol-rich regions of the plasma membrane involved in endocytosis. One of the constituting proteins of caveolae is caveolin-1 which was labeled here using newborn rat lung tissue fixed by instillation of 4% PFA, 0.1% GA in 0.2 M Hepes buffer. After freeze-substitution and embedding in acrylic resin (Table 1), ultrathin sections (40–70 nm) were cut and mounted on formvar-coated Ni mesh grids. Immunogold labeling was performed according to standard protocols [99]. The primary antibody was a rabbit anti-caveolin-1 antibody (BD Biosciences, Pharmingen, Germany) diluted 1:50. The secondary antibody was a goat-anti-rabbit antibody coupled to 10 nm gold particles (British Biocell, Cardiff, United Kingdom). A strong signal is found for caveolae in capillary endothelium and alveolar epithelium. Unspecific background labeling was weak (note the gold particle in the interstitium) but not completely absent. Immunogold labeling requires good knowledge about the biology of the target antigen and the specificity of the antibody. Before going to the TEM level, one is well advised to perform pilot light microscopic experiments. CL = Capillary lumen; EC = Endothelial cell; IC = Interstitial cell; AEI = Alveolar epithelial type I cell; AL = Alveolar lumen. Bar = 1 μm.

Figure 5

Titanium detection by ESI and pEELS. A and B show an energy loss imaging (ESI) series of the L3 orbital of Ti at 440 eV (background) and 464 eV (signal) in an erythrocyte culture. In C, the resulting difference is calculated, revealing the distribution of Ti in the sample. Not shown, but used in the three-window calculation of the Ti distribution is a second background window recorded at 390 eV. The occurrence of Ti is confirmed by parallel electron energy loss spectroscopy (pEELS), shown in the graph. The energy a beam electron looses when interacting with the sample is representative for the atom and orbital it is interacting with. The graph shows a peak intensity around 460 eV, indicative for the interaction with the L3 orbital of Ti (the background follows a negative exponential progression). The zeroloss overview and detail images show a dark particle near a red blood cell. With pEELS confirming the occurrence and ESI revealing the distribution, the particle can be appointed as containing Ti. Bars = 1 μm (D), 250 nm (A to F except D). Titanium dioxide particles were incubated with erythrocytes and fixed and embedded conventionally. The section was placed on a Ni-grid. No supportive film and no staining were used. Zeroloss measuring of the relative thickness of the section at 10 different positions revealed a t/λ of 0.38 (+/- 0.04).

Figure 6

Electron tomography of a 250 nm thick section. A tilt series (three stills are shown in A, B and C) between +/- 60° with an increment of 1° provides the information for a volume reconstruction by weighted backprojection (D) of an in vitro grown alveolar epithelial cell (cell line A549), exposed to polystyrene NP prior to chemical fixation. The 2 nm thin slice (E, at a depth of 58 nm in the section) reveals a crisper and clearer depiction of the polystyrene NP than the 250 nm thick Epon section (F). Arbitrary digital slices can be made (G, position shown by the two arrowheads in D) in order to provide unquestionable recognition of the NP. Bars = 250 nm in A-D, 50 nm in E-G. Software based 3D rendering offers a way for segmentation according to the interpretation of the user and allows full perspective freedom (H and I). Moreover, clipping planes can partially dissect the scene (J), segmented objects can be omitted (the membrane surrounding the NP in K) and specific quantitative information on rendered objects obtained. Blue: nanoparticle, shades of green: rough endoplasmic reticulum (RER), orange: plasma membrane, transparent white: membrane surrounding the NP, red: ribosomes (Ri) on the RER. Me = membrane, NP = nanoparticle, PM = plasma membrane.