After inhalation, airborne particles are deposited on the surface structures of the respiratory tract. In the alveoli, surfactant displaces particles to the aqueous hypophase bringing them into close contact with the epithelial cells [12, 13]. Understanding the interactions between inhaled particles and epithelial cells is of crucial importance because epidemiological and experimental studies have proved that inhalation of airborne particles is associated with adverse effects [5–10]. In recent years, it has been emphasized that particle size differences determine the extent of the cellular reactions to particle exposure and the mechanism by which particles are taken up by epithelial cells . Since the APM of epithelial cells is the first cellular structure the particles encounter, it is particularly necessary to understand whether particles induce changes in the plasma membrane and how they are taken up by the cells. In order to address this issue, we utilized an in vitro approach to investigate the effects of particle exposure on the surface area of the APM of A549 epithelial cells. Non-toxic polystyrene particles were used to exclude that cytotoxic or inflammatory effects of the particles influence the observed results. The low toxicity of the particles was confirmed by measuring LDH release and Il-8 secretion (Figure 5). Emphasis was placed on a correlation of dose and effect of different sized particles by analyzing which particle characteristic (number, surface area or mass/volume) showed the strongest correlation between dose and effect. Furthermore, the relation of these changes with the quantitative uptake of particles by epithelial cells was analyzed.
We used A549 cells because it is a widely used and well characterized cell culture line which shares characteristics with alveolar epithelial cells [32, 42]. A549 cells are capable of forming various types of endocytotic mechanisms including caveolin- and clathrin-mediated endocytosis [43, 44] as well as phagocytosis/macropinocytosis . As the uptake mechanisms of fine and ultrafine particles are very likely to be different from each other , the capacity of the cells to perform various endocytic mechanisms is very important. Exposure to a particle suspension was used as it allows an easy and exact dosimetry of particles in terms of particle number. A problem with submersed exposure is that differently sized particles have different diffusion and sedimentation characteristics. Limbach et al.  hypothesized that diffusion and sedimentation processes are responsible for a less pronounced uptake of ultrafine particles compared with fine particles. In order to avoid effects of sedimentation, the volume of the particle suspension was kept at a minimal liquid column of 0.8 mm, thus facilitating access of the particles to the cells. To exclude a significant influence of agglomeration of ultrafine particles on our results, we analyzed the size distribution of ultrafine particles in the submersion medium (Figure 1C). Particles were mainly present as single objects or in small aggregates of two to three particles excluding a significant agglomeration.
The analysis of APM surface area changes required the high resolution of the transmission electron microscope in combination with design-based stereology. Stereology allows the unbiased quantification of morphological cellular characteristics in absolute terms, in this case the total apical surface area of the APM per A549 cell. The uptake of particles was suspected to alter total cell volume which was only the case for fine particles at a number concentration of 6 × 108 particles per well. Since the total APM surface area per cell was calculated from the surface density and the cell volume, this parameter does not depend on changes in the cell volume. The quantification of intracellular particles by confocal LSM and subsequent application of a deconvolution algorithm was shown to be a suitable tool to quantify large numbers of fine and ultrafine fluorescent particles in an efficient way . Nevertheless, it was tested whether only agglomerates of ultrafine particles are recognized or individual ultrafine particles by experiments using 0.05 μm particles emitting fluorescence at different wavelengths. Counting the numbers of particles in the different fluorescence channels separately, and again in the merged channel showed no significant difference between the counts (data not shown), thus verifying the quantification of ultrafine particles.
The main results of our study can be summarized as follows: 1) The surface area of the APM was increased after exposure to 6 × 108 fine particles and to 4.5 × 1011 ultrafine particles per cell culture well. These results provide evidence for an altered cellular lipid metabolism favoring lipid trafficking to the APM after exposure to high particle concentrations. The increase in APM surface area was dependent on the total particle surface area administered to the cells. The unchanged mRNA expression of key genes required for lipid synthesis and uptake suggests that the additional APM originates from intracellular membrane stores, rather than from new synthesis. 2) At similar number concentrations, the uptake of fine particles was greater than that of ultrafine particles, a relationship becoming more pronounced with increasing particle concentrations. However, upon exposure to equal surface area concentrations, the number of ultrafine intracellular particles exceeded that of fine particles.
The quantification of APM surface area was based on the rationale that the interaction between particles and the membrane might interfere with endocytic and exocytic events leading to an increase or decrease in APM surface area. Interestingly, we observed a significant increase for both particle sizes at equal exposure surface area concentrations, indicating a particle-induced lipid trafficking to the APM. This finding is in accordance with studies on lipid trafficking to the APM, due to deformation stress performed in A549 cells . This study provided evidence that the APM enlargement protects the epithelial cells against injury and helps to reseal plasma membrane injuries. It is reasonable to hypothesize that different mechanic stress stimuli lead to lipid trafficking to the APM to compensate for increases in plasma membrane surface tension , loss of membrane after bulk phagocytosis  or plasma membrane injury [24, 28]. Interestingly, exposure of human embryonic kidney cells and immortalized mouse macrophages to 0.77 μm latex particles caused increased lipid synthesis and induced the expression of genes involved in lipid synthesis and uptake . Lipid synthesis was saturated at a particle concentration of about 30 μg/well of a 96 well plate corresponding to a particle number of 1.29 × 108 particle per well. Since the wells of the 96 well plate have an area which is approximately 21.4 times smaller than the wells used in the present study, this effect would be expected to occur at a number concentration of 2.5 × 109 particles or at a surface area concentration of 5 × 109 μm2 particle surface area. However, in our study the mRNA induction of enzymes involved in lipid synthesis and uptake was not observed at comparably high particle numbers or surface area concentrations. This indicates that the additional membrane in our study stems from pre-existing membrane pools, such as vesicles or the ER , rather than from newly synthesized membranes. The lack of the mRNA induction in a non-phagocytic cell line may also explain why the cells die at 1 μm particle number concentrations that are one magnitude higher than those at which the APM increase is observed. The results of this study may not be specific for the respiratory tract epithelium but the comparison with the data of Castoreno et al.  shows that the effects of particle endocytosis on lipid metabolism and membrane turnover may depend on the cell type and specialization. It remains to be determined if the lipid synthesis induced by Castoreno et al.  and the APM increase analyzed in this study have the same functional origin.
The APM of alveolar epithelial cells serves many functions, including the secretion and re-uptake of surfactant components by type II cells, as well as fluid regulation by type I cells. Any changes occurring in the quantitative composition of the APM of pulmonary epithelial cells may therefore have an effect on the normal metabolism of these cells. We admit that the use of cell lines limits the significance of the observed results for in vivo particle exposure in the lung. Not all particles in the alveoli come into contact with the alveolar epithelial cells because alveolar macrophages may take up a major portion of the particles. This makes it difficult to estimate how realistic particle concentrations in a mono cell culture are. However, it has been shown that particularly ultrafine particles interact with the alveolar epithelium , partially because they are not taken up by alveolar macrophages as effectively as larger particles . The doses investigated in the present study are higher than usual normal environmental pulmonary exposure, however, tobacco smoking or occupational exposure may increase the number of inhaled particles manifold.
The dependence of APM surface area increase on particle surface area dose corresponds to studies by Stoeger et al. [49, 50]. These authors exposed mice to six different particle types and measured the inflammatory response from bronchoalveolar lavage fluid and correlated the effects to the number, surface or mass of the particle dose. They found that particle surface area shows the closest correlation with the inflammatory response . Our present results underline the importance of total particle surface area as the most appropriate dose metric of particles for both structural and functional changes of cells induced by particles.
However, quantification of particles within the epithelial cells shows that the uptake characteristics are different between fine and ultrafine particles. Indeed, extrapolation of the trend lines shown in Figure 9 provides approximately the same numbers of intracellular particles after exposure to 6 × 108 fine and 4.5 × 1011 ultrafine particles per well, i.e. at the same concentrations that induced the increase in APM surface area. These results offer two reasonable interpretations: (1) Firstly, the exposed surface area determines the number of particles taken up by the cells and the increase in APM surface area independently. (2) Secondly, the main factor influencing increase in APM surface area is not the total particle surface area the cells are exposed to but the number of particles taken up by the cells. These relationships, however, require further analysis before clear conclusions can be drawn.