The application of confocal laser scanning microscopy (CLSM) to the study of xenobiotic uptake into plant foliage is explored in this paper. Three fluorescent dyes of low molecular weight and contrasting polarities (hydrophilic, moderately lipophilic and lipophilic) were selected to represent foliage-applied pesticides. These model compounds were applied as droplets to the surfaces of various leaves and/or fruits according to the particular experiment. The transcuticular diffusion behaviour, the compartmentation into epidermal cells and the influence of a surfactant on the uptake of these fluorescent compounds were visualized by CLSM. Distinct differences in diffusion speed across the cuticle and distribution in cell compartments were found between different fluorescent compounds. The presence of a surfactant significantly accelerated the uptake of the moderately lipophilic dye into both thin- and thick-cuticled leaves. The results are discussed in relation to the current knowledge on pesticide uptake and translocation. The advantages and limitations of this technique are highlighted.
The efficacy of foliage-applied herbicides, plant growth regulators and systemic fungicides is closely related to the quantity of active ingredients penetrating into the target species. A radiolabelling method is generally used to measure the uptake of pesticides into plant foliage. Uptake is generally defined as the proportion of the radioactive dose not recovered from washing the treated leaves (Gauvrit, 1994). Although this method allows a fast screening of pesticide formulations and tank-mix additives based on the gross uptake of the active ingredients, it cannot provide information on the location of absorbed chemicals inside leaf tissues and cell compartments. Of the pesticide molecules absorbed by plant leaves, only those that reach the target cells can exert their designed function. Unfortunately, there has previously been no easy method by which to estimate the relative proportion of absorbed molecules that can move from the leaf surface layer (comprising wax and cuticle) to the internal tissues and whether a particular type pesticide is prone to vacuolar sequestration by epidermal cells. The few studies conducted using fluorescence microscopy and micro-autoradiography techniques only provided limited information on the location of a few compounds at the leaf surface (Dybing & Currier, 1959, 1961; Kirkwood et al., 1982).
Confocal laser scanning microscopy (CLSM) has been widely applied in plant science over the last decade (for review see Hepler & Gunning, 1998). This technique has also been used to validate the phloem mobility model of xenobiotics with the help of a wide range of fluorescent compounds (Wright et al., 1996), and to visualize stomatal infiltration of surfactant solutions into leaves (Gaskin et al., 1998). However, to our knowledge, there has been no report concerning the application of CLSM to the study of the foliar uptake of xenobiotics. The fact that CLSM is able to reveal at the subcellular level the location of fluorescent molecules in thick samples in a non-destructive manner makes it an attractive tool for studying in vivo uptake of chemicals into plant leaves.
Most pesticides are non-fluorescent and cannot be easily visualized by CLSM. Therefore, several fluorescent compounds of low molecular weight and differing lipophilicities were selected to represent a range of pesticides. Their Lipinski parameters (molecular mass, log P, H-bond donors and acceptors, and rotatable bonds) were all within the ranges of typical foliage-applied herbicides (Tice, 2001). Such model xenobiotics are therefore expected to diffuse into plant leaves in the same way as pesticides.
Considerable effort has been made during the last few years in our laboratory to test the utility of CLSM in visualizing xenobiotic diffusion after foliar application. Some preliminary results have been reported elsewhere (Liu & Zabkiewicz, 2001). This short communication describes the in vivo visualization of the transcuticular diffusion, the compartmentation into epidermal cells and the effect of a surfactant on the uptake of three representative fluorescent compounds by using CLSM.
Materials and methods
Mature green pepper (Capsicum annuum) and apple (Malus domestica cv. Granny Smith) fruits were purchased locally.
Three other plant species were used: bean (Vicia faba cv. Evergreen), wheat (Triticum aestivum cv. Advantage) and cabbage (Brassica oleracea cv. Grow fresh). Bean and wheat seeds were sown in 7-cm square pots filled with a potting mix (‘Bloom’, Yates NZ Ltd). Cabbage seedlings were purchased locally. All plants were then raised in a controlled environment (20 °C/15 °C, day/night temperatures, 70% relative humidity, 12 h photoperiod with light intensity c. 500 µmol m−2 s−1). Bean and wheat plants were used 4 weeks after sowing, cabbage plants at the 5–6-leaf stage.
Three fluorescent dyes of low molecular weight (< 500 Da) and contrasting polarities were selected: Oregon Green 488 (2,7-difluorofluorescein, hereafter Oregon Green), Rhodamine B (N-[9-2-carboxyphenyl-6-(diethylamino)-3H-xanthen-3-ylidene]-N-ethylethanaminium chloride) and Nile Red [(5H-benzo[α]phenoxazin-5-one, 9-(diethylamino)-]. Oregon Green was used as its sodium salt. The physicochemical properties of these compounds are summarized in Table 1. These three fluorescent compounds were selected to simulate hydrophilic, moderately lipophilic and lipophilic pesticides, respectively. All three dyes were applied at a concentration of 0.5 g L−1.
Table 1. Properties of the fluorescent dyes used.
Logo/w was determined using the shake flask method (Anon., 1996).
The surfactant, Lutensol AO5 (C13/C15 linear alcohol ethoxylate, mean ethylene oxide content = 5, BASF, Germany) was used where appropriate at a concentration of 2 g L−1.
Visualization of xenobiotic diffusion into plant tissues
Droplets (0.5 µL) of the fluorescent dye solutions (prepared in 50% acetone) were applied with a microsyringe (Hamilton) to the intact surface of pepper or apple fruit, or to the adaxial surface of bean, wheat or cabbage leaves. Apple and pepper fruits and bean leaves were examined 24 h after application of the dye solutions, wheat leaves at 2 h, and cabbage leaves at both 2 h and 16 h after treatment. All treated materials were washed thoroughly with 50% acetone to remove the residual (unabsorbed) chemicals and then examined directly, without any preparation, on a confocal microscope (Leica TCS NT) using a 63× oil-immersion objective.
Two wavelengths (488 and 568 nm), generated by an Ar/Kr laser, were selected for exciting the fluorescent dyes. The localization of Oregon Green, Rhodamine B and Nile Red in plant tissues was detected using 530-nm (Oregon Green) or 600-nm (Rhodamine B and Nile Red) filters. Laser intensities and detector pinholes were adjusted such that a minimum noise level and the optimum resolution and brightness were achieved for each image. Over-saturation of signal (> 256 grey level) in certain parts of the images was tolerated in order to obtain maximal information contained in one optical section.
Cross-section (z) and/or horizontal (xy) section images of the treated leaves or fruits were acquired immediately (within 5 min after excision) to avoid sample dehydration. Each image was the average of four scans conducted at the maximum resolution (1024 × 1024 pixels) from one single optical section.
Transcuticular diffusion of xenobiotics
Owing to the relatively low resolution of CLSM (Pawley, 1995), it was difficult to resolve the cuticle layer when its thickness is below 1 µm, which is the case for most herbaceous plant species. Pepper and apple fruits were therefore used in this experiment.
Cross-sectional images revealed the vertical distribution of the two fluorescent compounds across the cuticle layer. The hydrophilic compound, Oregon Green, diffused readily through the thick cuticle of apple and moved into the cytoplasm of the epidermal cells, with little dye being retained by the cuticle layer (Fig. 1A). By contrast, the lipophilic compound, Nile Red, was strongly retained by the cuticle and no fluorescence was detected below the cuticle layer (Fig. 1B).
Pepper cuticle showed lower permeability than apple cuticle despite being thinner. Although a substantial quantity of Oregon Green reached the epidermal cells, a significant proportion of the dye was still found in the cuticle 24 h after treatment (Fig. 1C). In the case of Nile Red, fluorescence was again mainly observed in the cuticle layer (Fig. 1D).
It is widely known that the foliar uptake of lipophilic pesticides is much faster than that of hydrophilic pesticides (de Ruiter et al., 1993). This experiment reveals that as a result of their high affinity with the lipoidal cuticle, the desorption of lipophilic compounds from the cuticle to the internal tissues is much slower. By contrast, hydrophilic compounds, which have low affinity with the cuticle, could diffuse readily through it and reach the internal tissues. So higher uptake in the case of lipophilic pesticides may not lead to greater efficacy because of the slow dose transfer towards the target cells. Indeed, the most successful foliage-applied herbicide, glyphosate, is a water-soluble compound.
Subcellular compartmentation of xenobiotics
Cross-sectional images of bean leaves showed that after traversing the cuticle, Oregon Green diffused into the cytoplasm of both the epidermal and the palisade cells (Fig. 2A), whereas Rhodamine B was only present in the vacuole of the epidermal cells (Fig. 2B). The horizontal section images, acquired from the same samples about 10 µm below the leaf surface, confirmed the cytoplasmic accumulation of Oregon Green and vacuolar sequestration of Rhodamine B in the epidermal cells (Fig. 2C,D). Such contrasting compartmentation of the xenobiotics in the epidermal cells may have significant implication in phloem translocation and efficacy in the case of herbicides. Oregon Green, which was found in the cytoplasm of the epidermal and mesophyll cells, displayed clear phloem translocation, as evidenced by the presence of strong fluorescence in the petiole of the treated leaves (results not shown). By contrast, the vacuolar sequestration of Rhodamine B by the epidermal cells may prevent further transport of this compound to the mesophyll tissues. Indeed, no fluorescence could be detected in any of the mesophyll cells at 24 h after treatment and this dye did not show any phloem mobility (results not shown).
Most post-emergence herbicides are weakly acidic compounds and phloem mobility of such chemicals results from an ‘ion trapping’ mechanism caused by the difference in pH between the sieve tubes and the apoplast (Devine et al., 1993). As Oregon Green is also weakly acidic, the diffusion pattern observed above is expected to be applicable to other weakly acidic chemicals.
When wheat leaves were treated with Rhodamine B alone, fluorescence was only detected in the wax-cuticle layer (Fig. 3A). The isolated fluorescent spots present at the palisade cell level represent chlorophyll autofluorescence, also detected by the optical filter used for Rhodamine B detection. Addition of the surfactant substantially increased the total fluorescence intensity inside the vacuoles of treated leaves (Fig. 3B).
On cabbage leaves, only very weak fluorescence was observed in the wax-cuticle layer when Rhodamine B was applied alone (Fig. 3C). In the presence of the surfactant, there was a marked increase in fluorescence intensity in the wax-cuticle layer, but the dye was not detected in any other parts of the epidermal cells 2 h after treatment (Fig. 3D). Although extending the treatment time to 16 h only slightly increased the amount of the dye in the cuticle-wax layer in the absence of surfactants (Fig. 3E), the presence of the surfactant not only increased the fluorescence intensity in the cuticle, but also enhanced the diffusion of the dye through the cuticle. Consequently, fluorescence was clearly detectable in the epidermal cells (Fig. 3F). The weak fluorescence observed in the palisade cells again originated from the chloroplasts.
Results presented in this report have demonstrated the value of CLSM for understanding the possible modes of uptake and compartmentation of xenobiotics into plant foliage. The ability of CLSM in optically sectioning thick samples makes it possible to acquire both cross-sectional (z) and horizontal section (xy) images at required depth directly from intact leaves. Owing to the negligible depth of field of optical sections, high-resolution images were captured at both the epidermal and the mesophyll cell levels. A time-consuming and delicate sample preparation procedure would be required with a micro-autoradiography technique. Although fluorescence microscopy also revealed the distribution of fluorescent compounds on the leaf surface, the resolution is dramatically reduced with increasing depth. The advantages of CLSM have been confirmed in this application.
CLSM permitted the visualization of the desorption of hydrophilic and lipophilic compounds from the cuticle to the internal tissues and the compartmentation of different xenobiotics in epidermal cells. The information obtained provides novel insight into understanding the possible uptake pathways of pesticides after foliar application and the possible difference in phloem mobility between different pesticides. In addition, it has been shown, for the first time, how a surfactant influences the penetration of a moderately lipophilic compound into both thin- and thick-cuticled leaves.
It should be born in mind, however, that the confocal micrographs essentially provide qualitative information on the distribution of fluorescent compounds inside plant tissues and cells. Using CLSM to perform quantitative measurement of such compounds remains a challenge. It is still necessary to combine the surface wash-off method and CLSM to obtain both qualitative and quantitative information on the uptake of xenobiotics into plant foliage. As the resolution of CLSM is still relatively low compared with TEM, the cuticle layer and the outer epidermal cell wall cannot be clearly distinguished in the case of thin-cuticled leaves. Thick-cuticled materials are thus required to study the transcuticular diffusion characteristics of chemicals.
In conclusion, CLSM has proved to be an attractive tool in studying various aspects of xenobiotic diffusion into plant tissues. The results obtained in this study are believed to be applicable to the foliar uptake of pesticides. Further work is needed to refine the technique, especially in real-time imaging of the uptake process and phloem loading of model xenobiotics.
I would like to thank the anonymous referees for valuable suggestions. This work was supported by the New Zealand Foundation for Research, Science and Technology.