• The usefulness of the zinc (Zn)-fluorophore, Zinpyr-1, to examine the localization of Zn in the roots of Arabidopsis has been investigated.
• In wild-type roots Zinpyr-1 fluorescence was predominantly in the xylem. The fluorescence signal was abolished by the application of the Zn-chelator, N,N,N’,N-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN), and was increased by increasing exogenous Zn in the medium, indicating that fluorescence reflected relative Zn concentrations.
• In the hma2, hma4 double mutant, which is deficient in root to shoot Zn translocation, Zinpyr-1 fluorescence was low in the xylem and high in the adjacent pericycle cells in which HMA2 and HMA4 are specifically expressed in a wild type. Zinpyr-1 fluorescence was also increased in the endodermis.
• These results show that Zinpyr-1 can be used to examine the effects of mutations in Zn transporters on the localization of Zn in Arabidopsis roots and should be a useful addition to the tools available for studying Zn homeostasis in plants.
Metal ions such as iron (Fe), zinc (Zn), manganese (Mn) and copper (Cu) are essential for plant growth and are required for structural and catalytic roles in numerous proteins. However, an excess of free metal ions is toxic to cells and plants have mechanisms for metal homeostasis that restrict the potential toxicity of free metal ions while allowing the uptake and translocation of essential metals to tissues and cells (Clemens, 2001; Colangelo & Guerinot, 2006). Molecular genetic analysis of the mechanisms by which plants maintain homeostasis of essential metals has advanced through the identification of mutants, particularly in Arabidopsis but also in other species (Alonso et al., 2003; Grotz & Guerinot, 2006). In addition, significant advances have been made through the analysis of metal hyperaccumulator species and comparisons with their nonhyperaccumulator counterparts (Becher et al., 2004; Hammond et al., 2006).
The measurement and localization of metals in tissues and cells is an important part of these analyses. Metals can be extracted from tissues and measured using techniques such as atomic absorption spectroscopy (AAS), inductively coupled plasma (ICP) atomic emission spectroscopy (AES) and ICP mass spectrometry (ICP-MS). The uptake and translocation of metals can also be measured using radioactive isotopes (Page et al., 2006). Examples of techniques to measure metals in situ at the cellular or subcellular level in animal or plant tissues or cells include energy dispersive X-ray fluorescence (EDXRF) (Zaichick et al., 1999), laser ablation ICP-MS (Becker et al., 2005) and micro-proton-induced X-ray emission (micro-PIXE) (Vogel-Mikus et al., 2006). Recently, fluorescent sensors have been used to examine metals in vivo in cells and subcellular compartments. Much of this work has focused on the use of Zn-fluorophores (Thompson, 2005).
Some of the most commonly used Zn fluorescent probes are (6-methoxy-(2-p-toluenesulfonamido)quinoline (TSQ), Zinquin, Zinpyr and the ZnAF family (Thompson, 2005). Different members of these families may be either membrane-impermeable or membrane-permeable. Particular chemical modifications enable a membrane-permeable probe to be converted to an impermeable derivative through enzymatic modification in the cell. An example of this is the hydrolysis of an ethyl ester group on the Zinquin probe by endogenous cellular esterases (Kimber et al., 2000). Zinc fluorophores have been used for imaging Zn in vivo in animal cells, particularly in neural cell biology where, for example, Zn appears to modulate neurotransmission and the disruption of neuronal zinc homeostasis is associated with Alzheimer's disease (Woodroofe et al., 2004; Kay et al., 2006).
In plants, there are few examples of the use of fluorophores to study metal homeostasis or detoxification. The chemical morin has been used to visualize aluminium (Al) within maize cell roots. While cytoplasmic Al could be visualized, morin was unable to identify Al bound to the cell wall (Eticha et al., 2005). Copper transport across the thylakoid membrane has also been studied using the fluorophore Phen Green SK (PGSK) in thylakoid membranes isolated from P. sativum leaves. This study used the quenching of PGSK fluorescence by Cu2+ to demonstrate the transport of Cu2+ into the lumen (Shingles et al., 2004). The Zn fluorophore, Zinquin, has been used to show the accumulation of Zn at the base of trichomes in the Nicotiana tabacum leaf (Sarret et al., 2006). The capacity to visualize Zn in plant cells would be a valuable addition to the analysis of Zn homeostasis.
In Arabidopsis the Zn-transporting P-type ATPases, HMA2 and HMA4, are expressed predominantly in the vasculature and are believed to be important in the translocation of Zn within the plant. A number of studies have indicated that these transporters are involved in Zn homeostasis (Williams & Mills, 2005). An hma2, hma4 double mutant accumulates Zn in root tissue and exhibits a Zn-deficient growth phenotype because of insufficient Zn translocation from root to shoot (Hussain et al., 2004). This deficiency can be rescued by increasing the level of Zn in the growth medium or soil. Because of the alteration in Zn homeostasis we have taken advantage of the hma2, hma4 mutant to investigate the feasibility of using the Zn fluorophore, Zinpyr-1, as a tool to localize Zn in root tissue. Zinpyr-1 is useful for intracellular work because it is membrane-permeable, it uses a fluorescein-reporting group and it has a high quantum yield and excitation and emission wavelengths exceeding 490 nm. Zinpyr-1 is highly selective for Zn over other biological metals, and metal-dependent fluorescence occurs only upon binding to Zn or Cd (Burdette et al., 2001). No Cd was present in these experiments. In this study we show that Zinpyr-1-dependent fluorescence in vivo in seedling roots is influenced by exogenous Zn levels, by exposure to a Zn-chelator (N,N,N’,N-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN)) and in the hma4 single and hma2, hma4 double mutants indicating that it may be useful in the analysis of a range of mutants affected in Zn homeostasis.
Materials and Methods
Plant materials and growth conditions
Arabidopsis thaliana L. (Heynh.) ecotype Columbia wild type, hma2–4, hma4–2, and hma2–4, hma4–2 lines (Hussain et al., 2004) were used throughout this study. Seedlings were germinated and grown on agar nutrient medium as previously described (Howden et al., 1995). The nutrient medium routinely contained 1 µm added Zn unless otherwise indicated.
Generation of HMA2 promoter–HMA2-GFP transgenic lines
HMA2 contains a NcoI restriction site in the first exon downstream of the ATG codon. A 6.0 kb fragment of genomic DNA extending upstream from this NcoI site was ligated to a 2.7 kb HMA2 cDNA fragment extending from the NcoI site to downstream of the stop codon (Hussain et al., 2004). This HMA2 promoter-cDNA construct was subsequently used to replace the CaMV35S promoter-HMA2 cDNA in the previously described 35S-HMA2 cDNA-GFP in the 35S-GFP-JFH1 vector (Hussain et al., 2004). The junction sites and the entire cDNA were sequenced to verify the fusion construct. hma2–4, hma4–2 mutant plants grown in the presence of additional exogenous Zn to rescue the Zn-deficient phenotype were transformed with the HMA2 promoter–HMA2-GFP construct using the floral-dip method (Clough & Bent, 1998). Ten ammonium 4-(hydroxy(methyl)phosphinoyl)-DL-homoalaninate (BASTA)-resistant transformants were selected using dl-phosphinothricin (Sigma, St Louis, MO, USA). None exhibited the Zn-deficient phenotype demonstrating the construct encoded a functional HMA2–GFP fusion.
Zinpyr-1 treatment and imaging conditions
Working solutions of Zinpyr-1 in 0.9% saline were diluted from 1 mm stock made up in dimethyl sulphoxide (DMSO) and stored at −20°C. Seven- to 8-d-old seedlings were washed alternately three times in deionized water and in 10 mm ethylenediaminetetraacetic acid (EDTA) before being immersed in a working solution of 5–20 µm Zinpyr-1 and incubated at room temperature in darkness for 3 h. Samples were rinsed in deionized water, immersed in 75 µm propidium iodide to stain cell walls red and rinsed again. Samples were mounted in 0.9% saline and images were taken on an MRC-1000/1024 (Bio-Rad, Marnes-la-Coquette, France) Confocal laser-scanning microscope (CLSM), using excitation at 488 nm with a 100 mW Ar ion laser and a ×60 Plan Apo water immersion lens with fluorescein isothiocyanate (FITC) and Texas Red filters.
Quantification of Zinpyr-1 fluorescence
For each comparative analysis images of seedling roots were taken using fixed settings on the confocal microscope. A method of quantification was developed using imagej software (Rasband, 1997–2006) that involved assigning a value for green fluorescence intensity to every pixel in a defined area. A section in each image containing the xylem, pericycle and endodermis cells was boxed (as illustrated in Fig. 2) and the average pixel intensity in the boxed region calculated. In addition, the average maximum fluorescence intensity in three transects across each image was calculated to relate the region of maximum intensity to cell type.
Results and Discussion
Zinpyr-1 fluorescence is consistent with the expression patterns of HMA2 and HMA4
Previous studies using HMA2-GUS and HMA4-GUS reporter gene fusions have shown that, in the root, both genes are expressed in the vasculature (Fig. 1a; Hussain et al., 2004). To more clearly visualize the expression of HMA2 we expressed an HMA2 promoter–HMA2-GFP fusion construct in transgenic hma2–4, hma4–2 mutant plants. The HMA2 promoter region was the same as that used previously to express the GUS reporter. All the transgenic lines examined no longer exhibited the Zn-deficient phenotype of the hma2, hma4 double mutant, demonstrating that the HMA2-GFP fusion was functional and able to complement the hma2 mutation. Confocal laser scanning microscopy showed that the expression of the HMA2 promoter–HMA2-GFP fusion coincided with the location of the HMA2-GUS reporter construct previously described (Hussain et al., 2004; Fig. 1(b)). An optical X–Z cross-section derived from stacked confocal images taken longitudinally through the root showed that expression of HMA2-GFP was confined to the pericycle cells of seedling roots (Fig. 1c). The GFP appeared to be localized to the plasma membrane consistent with previous observations of the localization of both HMA2 and an HMA2-GFP fusion (Hussain et al., 2004). The localization of expression of the HMA2- and HMA4-GUS constructs in roots appears to be indistinguishable (Hussain et al., 2004). However, the precise cellular expression pattern of HMA4 has not been determined using a parallel HMA4 promoter–HMA4-GFP construct.
To visualize the accumulation of Zn in the roots of the hma2, hma4 mutant, 7- to 8-d old seedlings were immersed in 20 µm Zinpyr-1 and examined using CLSM. In the wild type roots the fluorescence signal was highest in the xylem while the pericycle cells showed relatively little fluorescence (Fig. 2a,b). By contrast, the fluorescence signal was significantly higher in the hma2, hma4 mutant roots and confocal X-Z sections show that the Zinpyr-1-dependent fluorescence in the mutant was predominant in the pericycle cells where HMA2 and HMA4 are normally expressed and in the adjacent endodermis cells (Fig. 2c,d). These observations suggested that Zinpyr-1 might be an effective Zn fluorophore in Arabidopsis roots.
Relative Zinpyr-1 fluorescence is reproducible in Arabidopsis roots
To quantify the relative fluorescence signal we used fixed CLSM settings to take longitudinal images of roots of 7- to 8-d-old seedlings stained with Zinpyr-1 for 3 h. We used 10–15 seedlings for each treatment and/or genotype and observed the region of the root above the elongation zone (approximately 1 cm from the root tip) that coincides with the expression of the HMA2-GUS and HMA4-GUS reporter genes in transgenic lines. In both wild-type and mutant roots the fluorescence was largely confined to the xylem and the pericycle and endodermis cell layers, encompassed in the boxed areas in Fig. 2, although the fluorescence intensity in different cells within this region differed between the wild type and mutant. The average pixel intensity in each boxed area gave an overall value of fluorescence (Table 1). In addition, we determined the average maximum pixel intensity from three separate transects across this region in each root to indicate the region in which the fluorescence was most intense and to obtain a further measure of relative intensity (Table 1). Within a treatment or genotype for both of these measures the standard error of the mean was not greater than 20% indicating a relatively reproducible intensity of fluorescence across replicate individuals.
Table 1. Zinpyr-1 fluorescence of roots of wild-type and mutant Arabidopsis (7- to 8-d-old seedlings were exposed to 20 µm Zinpyr-1 for 3 h)
The concentration of Zinpyr-1 was based on that used in previous studies of animal cells. We have tested concentrations of 5, 10 and 20 µm and the same level of fluorescence was observed at all concentrations (Fig. 3a). In addition, we used a time of exposure to Zinpyr-1 ranging from 15 min to 3 h and observed no significant differences in fluorescence intensity (Fig. 3b).
Zinpyr-1 fluorescence is influenced by exogenous Zn levels and a Zn chelator
We have used several approaches to confirm that differences in fluorescence intensity reflect relative levels of intracellular Zn. First, when roots were exposed to the Zn-chelator, TPEN, the fluorescence signal was effectively abolished (Fig. 4), as has been observed in animal cells (Malavolta et al., 2006). Second, when wild-type seedlings were grown on media containing increasing concentrations of Zn, the fluorescence signal increased progressively (Fig. 5). We also attempted to localize Zn in wild-type and mutant roots by using EDXRF. However, the Zn concentration was not sufficiently high to give a reliable measure using this technique. Although we are unable to use the intensity of Zinpyr-1 fluorescence to calculate an absolute level of Zn in cells, these approaches show that the fluorescence intensity can be manipulated through increasing exogenous Zn levels and through using a Zn chelator and indicate that fluorescence intensity reflects relative levels of available intracellular Zn. In addition, the increase in the fluorescence signal in the hma2, hma4 mutant in the pericycle cells that normally express HMA2 is consistent with the previous observation that the roots of the hma2, hma4 mutant accumulate higher levels of Zn than the wild type (Hussain et al., 2004), again indicating that Zinpyr-1 fluorescence intensity reflects intracellular Zn levels.
Zinpyr-1 fluorescence is also increased in the hma4 mutant
We have compared the intensity of Zinpyr-1 fluorescence in the wild type with both single hma2 and hma4 mutants and the hma2, hma4 double mutant (Table 1). In the hma2 mutant fluorescence was predominant in the xylem (not shown) although the intensity was slightly less than in the wild type. This may reflect decreased Zn transport into the xylem although not to the extent that accumulation of Zn in the pericycle was apparent. The hma4 single mutant, however, showed a fluorescence intensity that was similar to the hma2, hma4 double mutant, and significantly higher than the wild type (Table 1; Fig. 2e,f). Although both hma2 and hma4 must be mutated for the Zn-deficient phenotype to be observed, previous studies of the single mutants have shown that the Zn accumulation in shoots of the hma2 mutant is similar to the wild type but is decreased in the hma4 mutant (Hussain et al., 2004; Verret et al., 2004). This suggests that HMA4 has a more significant role in root to shoot translocation consistent with the data in Table 1. It is also striking that in both the hma4 and hma2, hma4 mutants, Zinpyr-1 fluorescence is high in both the pericycle and endodermis cells. This may suggest that, unlike HMA2, HMA4 is expressed in both these cell layers, although the expression of their respective promoter–GUS reporter constructs appears indistinguishable. Alternatively, it may indicate that Zn accumulation in the pericycle as a result of the hma4 mutation under these conditions results in decreased transport of Zn out of the adjacent endodermis through some feedback mechanism.
In summary, the data described here illustrate that Zinpyr-1 can be used as an effective sensor of relative Zn levels in the root cells of Arabidopsis seedlings. The clear correlation of HMA2 expression in pericycle cells with Zn accumulation in the same cells in the hma2, hma4 mutant demonstrates that Zinpyr-1 can be a tool to further analyse the cellular distribution of Zn in the roots. This approach should be useful in the analysis of other mutants of Arabidopsis affected in Zn homeostasis.
This research was supported by a grant from the Australian Research Council. We are grateful to Dr Stephen Cody at the Ludwig Institute for Cancer Research, Melbourne, for invaluable help with Confocal Microscopy.