In the present work, intracellular penetration of carbon-coated iron nanoparticles applied in planta was tracked using correlative microscopy, including a first screening of the samples with optical microscopy (bright field, dark field and phase contrast) followed by analysis through transmission electron microscopy. This strategy has allowed us to confirm both the progressive penetration of particles through the plant tissues and their presence in the form of intracellular aggregates, when injected into tissues near the application site, in as short a time as 48 h. Furthermore, 48 h after injection, isolated nanoparticles in the cytoplasm of individual cells close to a vascular core were observed in the proximity of a magnetic device located far from the application point. When applied by spraying, isolated nanoparticles were also found in the cytoplasm of epidermic cells, in regions near the application point.
Cell and tissue response to the presence of nanoparticles
Nanoparticles move away from the application point, on the interior of the stem, where they were injected, to the outer epidermis through the tissues. The fact that a dense cytoplasm with starch-containing organelles was observed concomitantly with nanoparticle aggregates in the cytosol, suggests that plant cells could respond to the presence of a high density of nanoparticles by changing their subcellular organization. Results also showed that the nanoparticles had already appeared in the outer surface of the plant, both inside and outside of the trichomes, 24 h after application, indicating that at least part of the bioferrofluid can be expelled in a short time. Several examples of cytotoxicity of carbon nanoparticles have been described in various animal systems [18–23] as well as different effects derived from the application of magnetic ferrofluid [24, 25]. Cytotoxicity has also been associated with the dose of nanoparticles . This correlation between the number of nanoparticles and cytotoxicity agrees with our results, which showed that no subcellular rearrangements were detected in those plant cells in which only a few isolated nanoparticles were detected (Figs. 8 and 9), compared with the response detected in cells carrying aggregates of nanoparticles. But, to date, there has been no description of changes in cell architecture and organization after nanoparticle internalization in vivo, so it is not possible to assess if the observed reaction is specific to our material or if it is a general trait of cell reaction to high concentrations of nanoparticles within the cell. Moreover, the solution used for nanoparticle suspension, namely gelafundine, contains calcium, an important second messenger in plants. The possibility that some calcium remains attached to the carbon coat of the nanoparticle and could have an effect on cell response, does not seem to be very plausible due to the chemical properties of the nanoparticle surface (hydrophobic, not negatively charged...), even so, it cannot be completely excluded.
Internalization of nanoparticles near the application point
The observation of particle aggregates in adjacent parenchymatic cells suggests a movement from cell to cell. Another question to be addressed is the presence of nanoparticles as intracellular aggregates, as it seems quite unlikely that nanoparticles have the capability of entering cells other than as individual particles. The mechanism involved in this aggregation is unclear. As indicated in MM, the carbon-iron nanoparticles employed show a tendency to aggregate in aqueous solutions because of the chemical characteristics of the carbon coat. Aggregation of nanoparticles is a common phenomenon , so it is possible that this reclustering takes place spontaneously after individual internalization into the cytoplasm, although it could not be discarded that the nanoparticles were redirected by the cell to a specific subcompartment or region in the cytoplasm.
The presence of nanoparticles in epidermal cells after application by spraying is of special interest. As stated before, one of the main drawbacks of other methods is that they cannot be employed for agronomic purposes. The method used in this work resembles the procedures which would be used by breeders and coordinators of phytosanitary control, employing both large scale and hand-on spraying to leaf surfaces. The fact that nanoparticles passed through the epidermal cell wall opens up the possible application of these nanotechnology tools for agronomical purposes. Given the special characteristics of the epidermic outer cell wall, specifically its considerable thickness, and the presence of protective waxes, a possible particle penetration point could be through the stomata and the subestomatic chambers. In fact, this aperture is a route used by pathogens of different species, such as the white pine blinter rust . Interestingly, water-suspended 43 nm hydrophilic particles have been described as occasionally penetrating Vicia faba leaves through stomatal pores . Recently, the uptake of magnetite nanoparticles through the root system of Cucurbita maxima plants was reported, though a significant uptake of particles was only found in plants growing in liquid media . So, in order to make the system suitable for agronomical purposes, methodological improvements would need to be made.
Long range transport and putative movement through the vascular system
The main reason for developing the carbon-iron nanoparticles and their application in this work was to accumulate them at a specific site in the plant by means of magnetic fields applied specifically to a certain area. As shown in Fig. 9, isolated nanoparticles were detected in the cytoplasm of cells close to the vascular core, far from the application point and near the magnet. The position of these cells suggests that the route taken by the particles involved the use of the plant vascular system. Direct observation of freshly cut material revealed the presence of bioferrofluid, specifically in the interior of the xylem vessels , along with nanoparticle aggregates, 48 h after application (Fig. 7) suggesting that the particles can use this system for long range transport. It has to be noted that those particles found far from their application point (Fig. 9) were quite homogeneous in size, around 46 nm in diameter on average, when compared with the variable sizes found in the original mixture detected in the aggregates in the pith cavity and cells close to the application point. (Figs. 2, 3, 4, 5, 6, 7). This may suggest that a certain critical size is required for the appropriate movement of particles through the plant by long range transport mechanisms.