Fluorescence Resonance Energy Transfer-Sensitized Emission of Yellow Cameleon 3.60 Reveals Root Zone-Specific Calcium Signatures in Arabidopsis in Response to Aluminum and Other Trivalent Cations^,,

Abstract

Fluorescence resonance energy transfer-sensitized emission of the yellow cameleon 3.60 was used to study the dynamics of cytoplasmic calcium ([Ca²⁺]_cyt) in different zones of living Arabidopsis (Arabidopsis thaliana) roots. Transient elevations of [Ca²⁺]_cyt were observed in response to glutamic acid (Glu), ATP, and aluminum (Al³⁺). Each chemical induced a [Ca²⁺]_cyt signature that differed among the three treatments in regard to the onset, duration, and shape of the response. Glu and ATP triggered patterns of [Ca²⁺]_cyt increases that were similar among the different root zones, whereas Al³⁺ evoked [Ca²⁺]_cyt transients that had monophasic and biphasic shapes, most notably in the root transition zone. The Al³⁺-induced [Ca²⁺]_cyt increases generally started in the maturation zone and propagated toward the cap, while the earliest [Ca²⁺]_cyt response after Glu or ATP treatment occurred in an area that encompassed the meristem and elongation zone. The biphasic [Ca²⁺]_cyt signature resulting from Al³⁺ treatment originated mostly from cortical cells located at 300 to 500 μ m from the root tip, which could be triggered in part through ligand-gated Glu receptors. Lanthanum and gadolinium, cations commonly used as Ca²⁺ channel blockers, elicited [Ca²⁺]_cyt responses similar to those induced by Al³⁺. The trivalent ion-induced [Ca²⁺]_cyt signatures in roots of an Al³⁺-resistant and an Al³⁺-sensitive mutant were similar to those of wild-type plants, indicating that the early [Ca²⁺]_cyt changes we report here may not be tightly linked to Al³⁺ toxicity but rather to a general response to trivalent cations.

The role of calcium ions (Ca²⁺) as a ubiquitous cellular messenger in animal and plant cells is well established (Berridge et al., 2000; Sanders et al., 2002; Ng and McAinsh, 2003). Cellular signal transduction pathways are elicited as a result of fluctuations of free Ca²⁺ in the cytoplasm ([Ca²⁺]_cyt) in response to external and intracellular signals. These changes in [Ca²⁺]_cyt influence numerous cellular processes, including vesicle trafficking, cell metabolism, cell proliferation and elongation, stomatal opening and closure, seed and pollen grain germination, fertilization, ion transport, and cytoskeletal organization (Hepler, 2005). [Ca²⁺]_cyt fluctuations occur because cells have a Ca²⁺ signaling “toolkit” (Berridge et al., 2000) composed of on/off switches and a multitude of Ca²⁺-binding proteins. The on switches depend on membrane-localized Ca²⁺ channels that control the entry of Ca²⁺ into the cytosol (Piñeros and Tester, 1995, 1997; Thion et al., 1998; Kiegle et al., 2000a; White et al., 2000; Demidchik et al., 2002; Miedema et al., 2008). On the other hand, the off switches consist of a family of Ca²⁺-ATPases and Ca²⁺/H⁺ exchangers in the plasma membrane or endomembrane that remove Ca²⁺ from the cytosol, bringing the [Ca²⁺]_cyt down to the initial resting level (Lee et al., 2007; Li et al., 2008).

The numerous cellular processes regulated by Ca²⁺ have led investigators to ask how specificity in Ca²⁺ signaling is maintained. It has been proposed that specificity in Ca²⁺ signaling is achieved because a particular stimulus elicits a distinct Ca²⁺ signature, which is defined by the timing, magnitude, and frequency of [Ca²⁺]_cyt changes. For instance, tip-growing plant cells such as root hairs and pollen tubes exhibit oscillatory elevations in [Ca²⁺]_cyt that partly mirror the oscillatory nature of growth in these cell types (Cárdenas et al., 2008; Monshausen et al., 2008). Another example is nuclear Ca²⁺ spiking in root hairs of legumes exposed to NOD factors (Oldroyd and Downie, 2006; Peiter et al., 2007). Recently, it was shown that mechanical forces applied to an Arabidopsis (Arabidopsis thaliana) root can trigger a stimulus-specific [Ca²⁺]_cyt response (Monshausen et al., 2009). Translating the Ca²⁺ signature into a defined cellular response is governed by a number of Ca²⁺-binding proteins such as calreticulin that act as [Ca²⁺]_cyt buffers, which shape both the amplitude and duration of the Ca²⁺ signal or Ca²⁺ sensors such as calmodulin that impact other downstream cellular effectors (Berridge et al., 2000; White and Broadley, 2003; Hepler, 2005).

A deeper understanding of Ca²⁺ signaling mechanisms in plants has been driven in large part by our ability to monitor dynamic changes in [Ca²⁺]_cyt in the cell. Such measurements have been conducted using Ca²⁺-sensitive fluorescent indicator dyes (e.g. Indo and Fura), the luminescent protein aequorin (Knight et al., 1991, 1996; Legué et al., 1997; Wymer et al., 1997; Cárdenas et al., 2008), and more recently the yellow cameleon (YC) Ca²⁺ sensor, a chimeric protein that relies on fluorescence resonance energy transfer (FRET) as an indicator of [Ca²⁺]_cyt changes in the cell (Allen et al., 1999; Miwa et al., 2006; Qi et al., 2006; Tang et al., 2007; Haruta et al., 2008). The YC reporter is composed of cyan fluorescent protein (CFP), the C terminus of calmodulin (CaM), a Gly-Gly linker, the CaM-binding domain of myosin light chain kinase (M13), and a yellow fluorescent protein (YFP; Miyawaki et al., 1997, 1999). The increased interaction between M13 and CaM upon binding of Ca²⁺ to CaM triggers a conformational change in the protein that brings the CFP and YFP in close proximity, resulting in enhanced FRET efficiency between the two fluorophores (Miyawaki, 2003). Thus, changes in FRET efficiency between CFP and YFP in the cameleon reporter are correlated with changes in [Ca²⁺]_cyt.

Since it was first introduced, improved versions of the cameleon reporter have been selected to more accurately report [Ca²⁺]_cyt levels in the cell. For instance, the YC3.60 version was selected because of its resistance to cytoplasmic acidification and its higher dynamic range compared with the earlier cameleons. The higher dynamic range of YC3.60 is due to the use of a circularly permutated YFP called Venus (cpVenus) that is capable of absorbing a greater amount of energy from CFP (Nagai et al., 2004). Recently, the utility of YC3.60 for monitoring [Ca²⁺]_cyt was demonstrated in Arabidopsis roots and pollen tubes using ratiometric imaging approaches (Monshausen et al., 2007, 2008, 2009; Haruta et al., 2008; Iwano et al., 2009). Here, we further evaluated YC3.60 as a [Ca²⁺]_cyt sensor in plants using confocal microscopy and FRET-sensitized emission imaging. Unlike the direct ratiometric measurement of cpVenus and CFP reported in previous studies using YC3.60-expressing plants (Monshausen et al., 2008, 2009), the sensitized FRET approach we describe here involves the use of donor-only (CFP) and acceptor-only (YFP) controls, allowing us to correct for bleed-through and background signals from the FRET specimen (van Rheenen et al., 2004; Feige et al., 2005).

For this study, we focused on monitoring [Ca²⁺]_cyt changes in Arabidopsis seedling roots after aluminum (Al³⁺) exposure. Although Ca²⁺ signaling has long been implicated in mediating Al³⁺ responses in plants (Rengel and Zhang, 2003), the [Ca²⁺]_cyt changes evoked by Al³⁺ reported in the literature have been inconsistent, and as such, the significance of these [Ca²⁺]_cyt responses to mechanisms of Al³⁺ toxicity are not very clear. For instance, some studies reported that Al³⁺ caused a decrease in [Ca²⁺]_cyt in plants (Jones et al., 1998b; Kawano et al., 2004), and others demonstrated elevated [Ca²⁺]_cyt upon Al³⁺ treatment (Nichol and Oliveira, 1995; Lindberg and Strid, 1997; Jones et al., 1998a; Zhang and Rengel, 1999; Ma et al., 2002; Bhuja et al., 2004).

Here, we report that Arabidopsis roots expressing the YC3.60 reporter exhibited transient elevations in [Ca²⁺]_cyt within seconds of Al³⁺ exposure. The general pattern of [Ca²⁺]_cyt changes observed after Al³⁺ treatment were distinct from those elicited by ATP or Glu, reinforcing the concept of specificity in [Ca²⁺]_cyt signaling. We also observed root zone-dependent variations in the [Ca²⁺]_cyt signatures evoked by Al³⁺ in regard to the shape, duration, and timing of the [Ca²⁺]_cyt response. Other trivalent ions such as lanthanum (La³⁺) and gadolinium (Ga³⁺), which have been widely used as Ca²⁺ channel blockers (Monshausen et al., 2009), also induced a rapid rise in [Ca²⁺]_cyt in root cells that were similar to those elicited by Al³⁺. Al³⁺, La³⁺, and Gd³⁺ elicited similar [Ca²⁺]_cyt signatures in the Al³⁺-tolerant mutant alr104 (Larsen et al., 1998) and the Al³⁺-sensitive mutant als3-1 (Larsen et al., 2005), indicating that the early [Ca²⁺]_cyt increases we report here may not be tightly linked to mechanisms of Al³⁺ toxicity but rather to a general trivalent cation response. Our study further shows that FRET-sensitized emission imaging of Arabidopsis roots expressing YC3.60 provides a robust method for documenting [Ca²⁺]_cyt signatures in different root developmental zones that should be useful for future studies on Ca²⁺ signaling mechanisms in plants.

RESULTS

FRET-Sensitized Emission Imaging of YC3.60 Faithfully Reports [Ca²⁺]_cyt Dynamics in Arabidopsis Root Cells

We generated Arabidopsis plants expressing YC3.60 under the control of the cauliflower mosaic virus 35S promoter (35S::YC3.60) to investigate Ca²⁺ signaling events in root cells after Al³⁺ treatment. All of our imaging experiments were conducted with one 35S::YC3.60-expressing line that we first established to be capable of displaying Ca²⁺-dependent fluorescence changes in response to stimuli that are known to induce [Ca²⁺]_cyt elevations in plants. The ability of the YC3.60 reporter to monitor changes in [Ca²⁺]_cyt relies on the efficiency of energy transfer between CFP and cpVenus upon Ca²⁺ binding (Nagai et al., 2004). Thus, we utilized a preloaded application wizard from the Leica TCS SP2 AOBS confocal microscope to monitor FRET-sensitized emission in 35S::YC3.60-expressing roots. To efficiently measure FRET-sensitized emission with the Leica application wizard, we needed to account for background and bleed-through artifacts that could compromise accurate FRET efficiency readings (van Rheenen et al., 2004; Feige et al., 2005). For these corrections, it was necessary to acquire donor-only (CFP) and acceptor-only (YFP) fluorescence. Therefore, we generated Arabidopsis plants expressing 35S::CFP or 35S::YFP to serve as reference plants for the background subtraction and bleed-through correction algorithms needed for measuring FRET-sensitized emission from the 35S::YC3.60-expressing lines (see “Materials and Methods”).

We first confirmed that growing root hairs from our independently generated 35S::YC3.60 line displayed the tip-focused [Ca²⁺]_cyt gradients and oscillations that were reported recently using ratiometric cpVenus::CFP imaging (Monshausen et al., 2008). When FRET efficiency at the tip of growing root hairs was plotted over time, we observed Ca²⁺-dependent FRET efficiency oscillations at the root hair tip with a frequency of about two to four peaks per min (Supplemental Fig. S1; Supplemental Movie S1). These root hair oscillations were similar to those reported by Monshausen et al. (2008).

To set the stage for the Al³⁺ experiments, we next asked whether we could elicit FRET efficiency changes in YC3.60-expressing roots by applying compounds previously shown to induce elevations in [Ca²⁺]_cyt. We first tested whether we could detect global Ca²⁺-dependent FRET efficiency changes in a region of the root located 200 to 400 μ m from the root cap junction (RCJ; Fig. 1A, white rectangular box). This region of the root has been referred to as the distal elongation zone or transition zone and is one of the proposed four distinct zones in the Arabidopsis primary root that has characteristic cellular activities (Verbelen et al., 2006). We chose this region for our initial Ca²⁺-dependent FRET efficiency measurements because it has been shown to be the most responsive to various environmental stimuli (Baluška et al., 2001; Verbelen et al., 2006). Glu has been previously shown to cause transient elevations in [Ca²⁺]_cyt in plants (Dennison and Spalding, 2000; Qi et al., 2006). We therefore examined Ca²⁺-dependent FRET efficiency changes in the root transition zone of 3- to 4-d-old seedlings in response to exogenous Glu. During the conduct of our experiments, we found that the consistency of eliciting Ca²⁺-dependent FRET efficiency changes in roots, particularly when it involved exogenous application of chemicals, was strongly influenced by the manner in which seedlings were grown. To efficiently apply exogenous solutions to living Arabidopsis roots, it was necessary to germinate seeds directly on coverslips coated with a thin layer of low-melting-point agarose supplemented with half-strength Murashige and Skoog (MS) nutrients (see “Materials and Methods”). This allowed us to securely anchor the root so that it did not drift away from the microscope field of view upon application of the solution. Using this system, we observed that 1 mm Glu added directly to the root surface caused a rise in [Ca²⁺]_cyt in the root transition zone starting at 19.63 ± 1.47 s after application and reached a maximum after 32.18 ± 2.80 s (values are means ± se of 11 roots; Fig. 1, B and C; Supplemental Movie S2). Application of the solvent control solution did not induce a rise in [Ca²⁺]_cyt (data not shown).

Figure 1.

[Ca²⁺]_cyt-dependent FRET efficiency changes in Arabidopsis primary roots after Glu and ATP application. A, An image of a root from a 4-d-old Arabidopsis seedling indicating different regions of the root. The white rectangle marks part of the transition zone of the root where FRET efficiency traces were obtained. B, Application of 1 mm Glu triggers a rapid increase in [Ca²⁺]_cyt-dependent FRET efficiency. C, Pseudocolored time-lapse sequence of cells in the transition zone after treatment with Glu showing the peak of the [Ca²⁺]_cyt response at 30 s and its return to basal levels within 60 s. Supplemental Movie S2 shows the complete movie sequence. D, ATP-induced elevation of [Ca²⁺]_cyt-dependent FRET efficiency. Note the broader shape of the [Ca²⁺]_cyt response of ATP-treated roots compared with Glu-treated roots. Arrows in B and D indicate the time of chemical addition. Bars = 100 μ m in A and 20 μ m in C.

Since ATP has been shown to induce [Ca²⁺]_cyt increases in plants using methods such as aequorin (Demidchik et al., 2003; Jeter et al., 2004), we also tested the effect of extracellular ATP on [Ca²⁺]_cyt in the transition zone. When 1 mm ATP was applied to YC3.60-expressing roots, we observed a rapid increase in Ca²⁺-dependent FRET efficiency in cells of the transition zone. However, compared with Glu, the onset of the [Ca²⁺]_cyt increase took longer and began at 37.5 ± 3.87 s after ATP application, with a maximum [Ca²⁺]_cyt response occurring at 63 ± 5.84 s (values are means ± se of 11 roots). Furthermore, unlike the Glu-induced [Ca²⁺]_cyt spike, which declined rapidly after reaching maximum values (Fig. 1B), the ATP-induced increase in [Ca²⁺]_cyt was generally followed by a more gradual decline that spanned a period of approximately 3 to 5 min (Fig. 1D). The shapes of the ATP-induced [Ca²⁺]_cyt signature that we observed were similar to those reported in mammalian cells expressing YC3.60 (Nagai et al., 2004).

Al³⁺ Induces a Range of [Ca²⁺]_cyt Signatures in the Transition Zone That Are Different from Those Elicited by Glu and ATP

On the basis of the results described above with Glu and ATP, we were confident that FRET-sensitized emission measurements with YC3.60-expressing plants presented a simple and highly reproducible approach to begin a more detailed evaluation of [Ca²⁺]_cyt signaling in Arabidopsis root cells upon Al³⁺ treatment. Like Glu and ATP, we first evaluated the effect of Al³⁺ on global changes in [Ca²⁺]_cyt in the transition zone, as this region of the root is a primary target of Al³⁺ (Sivaguru and Horst, 1998). Since the toxic species of Al³⁺ forms at low pH (Kinraide and Parker, 1987), we prepared a working solution of AlCl₃ in half-strength MS medium at pH 4.5. Application of the low-pH solution alone had no impact on [Ca²⁺]_cyt-dependent FRET efficiency in cells of the transition zone, but when supplemented with Al³⁺, we observed transient [Ca²⁺]_cyt-dependent FRET efficiency increases (Fig. 2). In contrast to the largely uniform shapes of the [Ca²⁺]_cyt signatures induced by Glu or ATP (Fig. 1), the root transition zone showed variations in the patterns of the [Ca²⁺]_cyt response after Al³⁺ treatment. We found that 35% of the roots imaged exhibited a [Ca²⁺]_cyt signature that had a single broad peak (Fig. 2A), whereas 59% of the roots displayed a [Ca²⁺]_cyt signature that was biphasic (i.e. it had two peaks; Fig. 2, B and C). The magnitude of the two peaks varied from root to root such that peaks were either unequal in height (Fig. 2B), which occurred in most of the roots with a biphasic response (31%), or roughly similar in height (28%; Fig. 2C). A representative time-lapse sequence of a heat map of the root transition zone after Al³⁺ application shows the onset of the first and second [Ca²⁺]_cyt peaks (Fig. 2E; Supplemental Movie S3). In a few roots (6%), we observed three distinct [Ca²⁺]_cyt peaks during the entire period of imaging (Fig. 2D). The onset of the [Ca²⁺]_cyt rise in the transition zone in response to Al³⁺ generally took longer compared with either Glu or ATP. The [Ca²⁺]_cyt increase started at 66.0 ± 6.9 s after Al³⁺ application, and the first peak reached a maximum at 186.5 ± 6.9 s (values are means ± se from 55 roots).

Figure 2.

[Ca²⁺]_cyt-dependent FRET efficiency changes in the transition zone of Arabidopsis primary roots after Al³⁺ application. A to D, The [Ca²⁺]_cyt signature had one broad peak (A) and two to three peaks (B–D). The dashed line in A is a representative trace after addition of the pH-4.5 control solution. B shows two representative tracings with peaks that were unequal in height, and C shows a trace with almost two identical peaks. D shows a [Ca²⁺]_cyt signature with three distinct peaks. Numbers in each panel indicate the percentage of roots imaged that displayed a particular [Ca²⁺]_cyt signature. Arrows indicate the time of Al³⁺ application. E, Pseudocolored time-lapse sequence of cells in the transition zone after treatment with Al³⁺. The middle panels show the timing of two peaks that correspond to maximum [Ca²⁺]_cyt response. Supplemental Movie S3 shows the complete movie sequence. Bar = 20 μ m.

Imaging of FRET-Sensitized Emission Allows Monitoring of [Ca²⁺]_cyt Changes in Different Root Zones

We extended our sensitized FRET efficiency imaging to the meristem (0–100 μ m from the RCJ), root cap, and maturation zone to determine whether cells in these different zones also exhibited a [Ca²⁺]_cyt increase in response to the compounds tested (Fig. 1A; Verbelen et al., 2006). We found that cells in the meristem and cap displayed a Ca²⁺-dependent increase in FRET efficiency in response to Glu and ATP that had similar shapes as those observed in the transition zone (Fig. 3, A and B). The meristem and root cap exhibited an almost simultaneous elevation in [Ca²⁺]_cyt upon Glu treatment, but the magnitude of FRET efficiency increases was lower in the cap (Fig. 3A). The rise in [Ca²⁺]_cyt in the cap started at 16.33 ± 2.29 s, whereas in the meristem this increase occurred at 15.0 ± 1.67 s (values are means ± se; n = 10). On the other hand, the [Ca²⁺]_cyt increase in the cap after ATP application was slightly delayed compared with Glu application, and the magnitude of the response was also lower compared with that of the meristem (Fig. 3B). The rise in the [Ca²⁺]_cyt in the cap after ATP treatment began at 54.0 ± 4.29 s, whereas in the meristem the [Ca²⁺]_cyt increase occurred at 40.28 ± 4.48 s (values are means ± se; n = 7). A time-lapse movie sequence of a heat map representing FRET efficiency values shows the almost simultaneous increase in [Ca²⁺]_cyt in the cap and meristem after Glu treatment and a slight delay in the onset of the [Ca²⁺]_cyt rise in the cap compared with the meristem after ATP application. The movies also show that upon ATP treatment, [Ca²⁺]_cyt remained elevated for a longer duration compared with that of Glu treatment (Supplemental Movies S4 and S5).

Figure 3.

[Ca²⁺]_cyt-dependent FRET efficiency changes in the cap, meristematic zone, and maturation zone of Arabidopsis primary roots after Glu, ATP, and Al³⁺ application. A to C, The shape of the [Ca²⁺]_cyt response after Glu (A), ATP (B), and Al³⁺ (C) application was roughly similar between the cap and the meristem, but the magnitude of the response was generally lower in the cap. Arrows indicate the time of treatment application. Insets show pseudocolored images before and after chemical application. Supplemental Movies S4 to S6 show the complete movie sequences for the insets. Yellow rectangles in A indicate areas corresponding to the cap and meristem where [Ca²⁺]_cyt traces were obtained. D, Glu application triggered a [Ca²⁺]_cyt increase in the maturation zone that had similar shapes as those elicited in the cap and meristem. E, ATP application induced a [Ca²⁺]_cyt signature in the maturation zone that oscillated slightly while gradually declining to basal levels. F, In the maturation zone, Al³⁺ treatment induced a [Ca²⁺]_cyt response with a sharp peak and a second broader peak. Bars in insets = 20 μ m.

Unlike in the transition zone, [Ca²⁺]_cyt signatures in the cap and meristem induced by Al³⁺ treatment were more uniform, such that only one [Ca²⁺]_cyt peak was observed in all roots imaged (Fig. 3C). Like ATP, the onset of the [Ca²⁺]_cyt increase in the cap in response to Al³⁺ was delayed compared with that of the meristem and the magnitude of the response was lower. The [Ca²⁺]_cyt increase in the cap after Al³⁺ treatment started at 99.4 ± 8.5 s, whereas in the meristem this increase occurred at 85.2 ± 6.7 s (values are means ± se; n = 14–15). A time-lapse movie sequence of a heat map for FRET efficiency values shows the [Ca²⁺]_cyt rise in the meristem starting ahead of the [Ca²⁺]_cyt increase in the cap (Supplemental Movie S6).

The maturation zone of the root also displayed a [Ca²⁺]_cyt increase after application of Glu, ATP, or Al³⁺. The maturation zone of the root can clearly be identified based on the presence of initiating root hairs (Fig. 1A). Although all three chemicals triggered a rise in [Ca²⁺]_cyt in the maturation zone, the magnitude of the [Ca²⁺]_cyt increase was noticeably less compared with that in the meristem and transition zone. The Glu-induced [Ca²⁺]_cyt spike was similar to that observed in all other regions of the root (Fig. 3D). On the other hand, with ATP, the majority of the tracings from the maturation zone showed a rise in [Ca²⁺]_cyt followed by oscillations that declined gradually over time (Fig. 3E). Like in the transition zone, Al³⁺-induced [Ca²⁺]_cyt signatures in the maturation zone were biphasic, but the [Ca²⁺]_cyt peaks were not as well separated. The first [Ca²⁺]_cyt peak was narrow, followed by a second peak that was broader compared with the second peak observed in the transition zone (compare Figs. 3F and 2C). The [Ca²⁺]_cyt increase in the maturation zone evoked by Al³⁺ was also longer in duration compared with that triggered by Glu or ATP (Fig. 3, D–F). The onset of the [Ca²⁺]_cyt rise in the maturation zone occurred at 47.6 ± 6.8 s, 26.25 ± 2.17 s, and 57.0 ± 4.70 s for Al³⁺, Glu, and ATP, respectively (values are means ± se; n = 8). Likewise, the duration of the [Ca²⁺]_cyt elevation triggered by Al³⁺ was longer in the maturation zone than in the transition zone, meristem, and root cap. Table I summarizes the timing of the Al³⁺-triggered [Ca²⁺]_cyt signatures in the different zones of the root.

Table I. Time analysis of Al³⁺-induced [Ca²⁺]_cyt elevations in cells from different developmental regions of Arabidopsis roots expressing YC3.60²⁺.

Values are means ± se from at least three independent experiments. Number of replicates is in parentheses. N/A, Not applicable.

Parameter	Root Cap	Meristem	Transition Zone	Maturation Zone
Response onset	99.4 ± 8.5 (n = 14)	85.2 ± 6.7 (n = 15)	66.0 ± 6.9 (n = 55)	47.6 ± 6.8 (n = 8)
Time to reach maximum [Ca2+]cyt elevation (first peak)	211.8 ± 10.4 (n = 14)	197.0 ± 5.3 (n = 15)	186.5 ± 6.9 (n = 55)	115.9 ± 8.7 (n = 8)
Time to reach maximum [Ca2+]cyt elevation from onset (first peak)	67.4 ± 12.0 (n = 14)	66.8 ± 5.3 (n = 15)	66.4 ± 5.2 (n = 55)	23.3 ± 2.6 (n = 8)
Time to reach maximum [Ca2+]cyt elevation (second peak when observed)	N/A	N/A	260.5 ± 12.5 (n = 33)	11.0 ± 37.2 (n = 6)
Time to reach maximum [Ca2+]cyt elevation from onset of second peak	N/A	N/A	31.5 ± 2.9 (n = 33)	28.5 ± 4.9 (n = 6)
Total duration of [Ca2+]cyt elevation	282.6 ± 31.3 (n = 14)	197.7 ± 23.6 (n = 15)	280.8 ± 15.7 (n = 55)	416.6 ± 41.4 (n = 8)

Based on the onset of [Ca²⁺]_cyt increase in the different root zones, we were able to infer the likely directions in which the [Ca²⁺]_cyt wave propagates in response to Glu, ATP, and Al³⁺. In Al³⁺-treated roots, the fastest [Ca²⁺]_cyt response appeared to occur in the maturation zone, with subsequent [Ca²⁺]_cyt increases propagating toward the root cap (Table I). Upon Glu treatment, the [Ca²⁺]_cyt increase was fastest in the meristem and root cap and propagated rapidly toward the transition zone, elongation zone, and maturation zone. Treatment with ATP induced the fastest [Ca²⁺]_cyt response in the meristem and transition zone, and subsequent propagation of the [Ca²⁺]_cyt signal appeared to occur toward the root cap and the maturation zone. A summary of the timing and potential directions of [Ca²⁺]_cyt increases in response to Al³⁺, Glu, and ATP in the primary root is shown in Figure 4.

Figure 4.

Summary of the onset of the [Ca²⁺]_cyt increase in the root after Glu, ATP, and Al³⁺ application. Numbers indicate the average time for the onset of the [Ca²⁺]_cyt rise ± se. Arrows indicate the proposed direction of the [Ca²⁺]_cyt wave based on the timing of the [Ca²⁺]_cyt rise for each root developmental zone. Bar = 100 μ m.

The Biphasic [Ca²⁺]_cyt Signature Triggered by Al³⁺ Occurs Primarily in Cells Located at 300 to 500 μ m from the Root Cap Junction

As shown in Figure 2, the [Ca²⁺]_cyt signatures in the region designated as the transition zone triggered by Al³⁺ were either monophasic or biphasic. This variability in the pattern of [Ca²⁺]_cyt increases prompted us to probe further into the reasons behind this variability. Our initial experiments were conducted with a 63 × objective with a confocal zoom setting of 2.0. With such settings, we were only able to acquire data from a 100- μ m longitudinal area of the root. To improve our spatial resolution and ask whether a particular [Ca²⁺]_cyt signature was associated with a more defined area of the transition zone, we used a 20 × objective with a confocal zoom of 1.0. At this magnification, we were able to acquire data that covered the apical 500 μ m of the primary root tip, which encompassed the cap, meristem, transition zone, and a zone before the initiating root hairs, which is likely the zone that Verbelen et al. (2006) refer to as the fast elongation zone. At this low-magnification setting, we found that the biphasic [Ca²⁺]_cyt rise after Al³⁺ treatment was primarily associated with the region of the root that was about 300 to 500 μ m from the RCJ (Fig. 5A).

Figure 5.

[Ca²⁺]_cyt-dependent FRET efficiency changes in the apical 500 μ m of Arabidopsis primary roots after Al³⁺ (A), La³⁺ (B), and Gd³⁺ (C) application. FRET efficiency imaging and calibration were conducted using a 20 × objective. Representative FRET efficiency tracings were acquired from the regions indicated by the black rectangles. Cations were applied 45 s after the start of imaging. A clear biphasic [Ca²⁺]_cyt signature was obtained primarily from the region 300 to 500 μ m from the root cap junction after exposure to all three trivalent cations. FRET efficiency measurements are representative of at least eight independent experiments. Bar = 100 μ m.

Gd³⁺ and La³⁺ Trigger Increases in [Ca²⁺]_cyt That Are Similar to Those Induced by Al³⁺

To evaluate whether the increases in [Ca²⁺]_cyt were specific to Al³⁺ and not due to a general effect of trivalent ions, we analyzed the effect of Gd³⁺ and La³⁺ on Ca²⁺-dependent FRET efficiency changes in the root transition zone. Both Gd³⁺ and La³⁺ induced Ca²⁺-dependent FRET efficiency transients in root cells of the transition zone that were similar to those triggered by Al³⁺. Like Al³⁺, La³⁺ and Gd³⁺ treatments induced both biphasic and monophasic [Ca²⁺]_cyt signatures (Supplemental Fig. S2). Because the variability of La³⁺- and Gd³⁺-induced [Ca²⁺]_cyt signatures was reminiscent of those induced by Al³⁺, we resorted to low-magnification imaging to determine whether a particular [Ca²⁺]_cyt signature was associated with a specific part of the transition zone. Like Al³⁺ treatment, the origin of the biphasic [Ca²⁺]_cyt response after La³⁺ and Gd³⁺ application occurred primarily in cells located at 300 to 500 μ m from the root tip (Fig. 5, B and C).

The Biphasic [Ca²⁺]_cyt Signature Induced by Trivalent Cations Originated from Epidermal and Cortical Cells

In order to further resolve the origin of the biphasic [Ca²⁺]_cyt response in the transition zone triggered by Al³⁺, Gd³⁺, and La³⁺, we acquired FRET efficiency data in a region of the root that was approximately 400 to 500 μ m from the tip, where we consistently observed the biphasic [Ca²⁺]_cyt signature (Fig. 5). For these experiments, we used a 63 × objective and a confocal zoom setting of 2.0 to obtain better cellular resolution. We imaged the root at a depth that allowed us to clearly distinguish epidermal cells from the underlying cortical cells. This was only possible if the root was growing flat across the surface of the coverslip. This ensured that the outlying epidermal cells, which were mostly vacuolated, were in the same focal plane as the cytoplasm-rich cortical cells. We collected FRET efficiency traces from a region of the root that corresponded to epidermal or cortical cells (Fig. 6A, white rectangles). In a majority of the roots imaged (61%), the biphasic [Ca²⁺]_cyt response after Al³⁺ treatment was associated with cortical cells, whereas the [Ca²⁺]_cyt signature from the epidermal cells had only one peak (Fig. 6A). We found in 39% of the roots that both epidermal and cortical cells exhibited the biphasic [Ca²⁺]_cyt signature (Fig. 6B). Similar observations were made when the roots were treated with Gd³⁺ (Fig. 6C) or La³⁺ (data not shown).

Figure 6.

Epidermal and cortical cell measurements of [Ca²⁺]_cyt-dependent FRET efficiency changes in the root transition zone. Measurements were obtained from the areas marked by the white rectangles to delineate epidermal and cortical cells. Measurements were obtained at 400 to 500 μ m from the tip, where the biphasic [Ca²⁺]_cyt signature is consistently observed. Cations were applied 45 s after the start of imaging. In traces with a biphasic [Ca²⁺]_cyt response, two general patterns were observed. A, Cortical cells exhibited biphasic [Ca²⁺]_cyt signatures, whereas epidermal cells displayed a monophasic [Ca²⁺]_cyt signature. B, Both epidermal and cortical cells displayed biphasic [Ca²⁺]_cyt responses. Numbers in A and B indicate the total number of roots displaying a particular cell type response over the total number of roots imaged. C, Representative [Ca²⁺]_cyt signature traces after Gd³⁺ treatment showing biphasic and monophasic signatures in the cortical and epidermal cells, respectively. Bars = 20 μ m.

EGTA, BAPTA, Verapamil, La³⁺, and Gd³⁺ Pretreatment Disrupt the Shape, Amplitude, and Timing of the Trivalent Al³⁺-Triggered Biphasic [Ca²⁺]_cyt Signature

To further characterize the nature of the biphasic [Ca²⁺]_cyt response to Al³⁺, we pretreated the roots with reagents that block Ca²⁺ channels or chelate Ca²⁺ from the external medium. We first utilized verapamil, which is known to block plasma membrane voltage-dependent Ca²⁺ channels (Demidchik et al., 2002). Roots pretreated with the solvent control solution for 20 min and then with Al³⁺ displayed the typical Al³⁺-induced biphasic [Ca²⁺]_cyt response (Fig. 7A). In roots pretreated with 1 mm verapamil, Al³⁺ evoked only a single [Ca²⁺]_cyt peak (Fig. 7B). In roots pretreated with compounds that chelate free Ca²⁺ ions, such as EGTA or 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′ -tetraacetic acid (BAPTA), the Al³⁺-induced biphasic [Ca²⁺]_cyt was generally dampened (Fig. 7, C and D). Similar observations were made on Gd³⁺- and La³⁺-treated roots after pretreatment with verapamil, EGTA, or BAPTA (data not shown).

Figure 7.

[Ca²⁺]_cyt-dependent FRET efficiency changes in the transition zone of Arabidopsis primary roots pretreated with verapamil, EGTA, or BAPTA for 20 min prior to Al³⁺ application. Imaging was conducted at 400 to 500 μ m from the root cap junction, which consistently shows a biphasic [Ca²⁺]_cyt signature in response to Al³⁺. A, Pretreatment with deionized water (solvent control) does not affect the Al³⁺-induced biphasic [Ca²⁺]_cyt signature. B, Only a monophasic [Ca²⁺]_cyt signature is observed in verapamil-pretreated roots after Al³⁺ application. Note that the peak of the [Ca²⁺]_cyt signature in verapamil-pretreated roots coincides with the second peak of solvent control-pretreated roots. EGTA (C) and BAPTA (D) dampened the amplitude of the Al³⁺-elicited [Ca²⁺]_cyt signature. Arrows indicate the time of treatment application. FRET efficiency measurements are representative of at least eight independent experiments.

Since La³⁺ and Gd³⁺ are commonly used as Ca²⁺ channel blockers (Sivaguru et al., 2003; Monshausen et al., 2009), we investigated whether pretreatment of roots with these two trivalent cations had any impact on the Al³⁺-induced [Ca²⁺]_cyt transients. We found that pretreatment with La³⁺ or Gd³⁺ for 20 min produced obvious effects on the pattern of the Al³⁺-elicited [Ca²⁺]_cyt signatures in the root transition zone, but these effects were quite variable (Supplemental Fig. S3). In some experiments, pretreatment with either cation appeared to completely inhibit any [Ca²⁺]_cyt increase after Al³⁺ application (Supplemental Fig. S3B), whereas in other experiments, either the first or second [Ca²⁺]_cyt peak was dampened or inhibited (Supplemental Fig. S3, B and C). In the majority of the tracings, pretreatment with La³⁺ or Gd³⁺ followed by Al³⁺ application triggered a [Ca²⁺]_cyt rise with a broad peak (Supplemental Fig. S3B). An Al³⁺-induced biphasic signature was rarely observed when the root was pretreated with La³⁺ or Gd³⁺. In one case (6% of tracings), we observed a biphasic [Ca²⁺]_cyt signature in Gd³⁺-pretreated roots, but the peaks were widely separated and smaller in area (Supplemental Fig. S3C). The basis for this variability is unclear, but it is evident that these two cations altered the Al³⁺-induced [Ca²⁺]_cyt signatures. Further experiments are needed to uncover the reasons for these observations.

In another set of experiments, roots were treated with elevated external Ca²⁺ by applying 10 mm CaCl₂ and [Ca²⁺]_cyt-dependent FRET changes were monitored shortly thereafter. Application of 10 mm CaCl₂ alone did not trigger a rise in [Ca²⁺]_cyt in the root transition zone (Supplemental Fig. S4A). When Al³⁺ was added to CaCl₂-pretreated roots, we observed a rapid increase in [Ca²⁺]_cyt that remained elevated throughout the 10-min sampling period (Supplemental Fig. S4B). These results indicate that roots have a mechanism to prevent Ca²⁺ influx into the cytosol despite high Ca²⁺ in the external medium, which is altered by Al³⁺ application.

The Al³⁺-Induced Biphasic [Ca²⁺]_cyt Signature Is Modified by an Antagonist of Neuronal Glu Receptors and an Anion Channel Blocker

It was previously reported that the impact of Al³⁺ on Arabidopsis root growth and the cytoskeleton could be mediated in part by interactions with Glu signaling pathways (Sivaguru et al., 2003). We asked whether imaging of [Ca²⁺]_cyt using sensitized FRET of YC3.60 could reveal additional insights into the nature of this interaction. We first asked how Glu pretreatment affected the Al³⁺-induced biphasic [Ca²⁺]_cyt signature in the root transition zone. The transition zone of control roots (i.e. roots pretreated with pH-4.5 solvent control solution) displayed the characteristic biphasic [Ca²⁺]_cyt signature after Al³⁺ application (Fig. 8A). When roots were pretreated with Glu for 20 min prior to application of Al³⁺, only a single [Ca²⁺]_cyt peak was observed that roughly coincided with the second peak of control roots (Fig. 8B). When Al³⁺ and Glu were applied simultaneously (Fig. 8C), roots exhibited [Ca²⁺]_cyt signatures that mirrored those of roots treated with Glu alone (Fig. 1B).

Figure 8.

[Ca²⁺]_cyt-dependent FRET efficiency changes in the transition zone of Arabidopsis primary roots pretreated with Glu, AP-5, and NPPB for 20 min prior to Al³⁺ application. Imaging was conducted at 400 to 500 μ m from the root cap junction, which consistently shows a biphasic [Ca²⁺]_cyt signature in response to Al³⁺. A, Pretreatment with the solvent control solution does not affect the Al³⁺-induced biphasic [Ca²⁺]_cyt signature. B, Only a monophasic [Ca²⁺]_cyt signature is observed in Glu-pretreated roots after Al³⁺ application. C, Simultaneous application of Glu and Al³⁺ triggers a [Ca²⁺]_cyt response that resembles that induced by Glu alone. D, AP-5 treatment evokes a [Ca²⁺]_cyt transient similar to that induced by Glu alone. E and F, Preatreatment with AP-5 appears to block or dampen the first Al³⁺-induced [Ca²⁺]_cyt peak (E) while inhibiting the Glu-induced [Ca²⁺]_cyt increase (F). G, The anion channel inhibitor NPPB does not cause a [Ca²⁺]_cyt response when applied on its own. H and I, NPPB pretreatment modifies both the Al³⁺-induced (H) and Glu-induced (I) [Ca²⁺]_cyt response. Arrows indicate the time of treatment application. FRET efficiency measurements are representative of at least eight independent experiments.

We next asked whether the Al³⁺-induced [Ca²⁺]_cyt transients could be facilitated through ionotropic Glu receptors, which are known to conduct cations across the plasma membrane upon binding to Glu (Mayer, 2005). We pretreated roots with 2-amino-5-phosphonopentanoate (AP-5), an antagonist of neuronal Glu receptors (Davies et al., 1981), which was shown by Sivaguru et al. (2003) to block Al³⁺-induced microtubule disruption and membrane depolarization. Interestingly, 1 mm AP-5 on its own was able to induce rapid [Ca²⁺]_cyt transients resembling that of roots treated with Glu alone (Fig. 8D). When Al³⁺ was applied to roots pretreated with AP-5, only a single broad [Ca²⁺]_cyt peak was observed in the transition zone, which was comparable to the Al³⁺-induced [Ca²⁺]_cyt peak of Glu-pretreated roots (Fig. 8, compare E and B). Consistent with its function as an antagonist of ionotropic Glu receptors, AP-5 pretreatment completely blocked the Glu-induced [Ca²⁺]_cyt increases in the root transition zone (Fig. 8F).

We also tested whether the Al³⁺-induced biphasic [Ca²⁺]_cyt signature in the root transition zone was affected by the anion channel blocker 5-nitro-2-(3′-phenylpropyl-amino) benzoate (NPPB), since this compound was also shown to prevent Al³⁺-triggered cellular changes in Arabidopsis roots (Sivaguru et al., 2003). NPPB alone did not evoke any [Ca²⁺]_cyt changes in the root transition zone (Fig. 8G), but pretreatment with this compound altered the shape and amplitude of the Al³⁺-induced [Ca²⁺]_cyt signature. Treatment with this anion channel inhibitor either abolished the first Al³⁺-induced [Ca²⁺]_cyt peak while dampening the second peak or caused a late rise in [Ca²⁺]_cyt that was sustained throughout the entire 10-min sampling period (Fig. 8H). Like AP-5, NPPB pretreatment inhibited the Glu-induced [Ca²⁺]_cyt spike (Fig. 8I).

Trivalent Cations Elicit Similar Early [Ca²⁺]_cyt Increases in an Al³⁺-Resistant and an Al³⁺-Sensitive Mutant

The observation that La³⁺ and Gd³⁺ elicited [Ca²⁺]_cyt signatures similar to those evoked by Al³⁺ indicates that these early [Ca²⁺]_cyt changes represent a general response to trivalent cations rather than to mechanisms that are specific to Al³⁺ toxicity or signaling. To address this issue further, we transformed one Al³⁺-resistant mutant (alr104; Larsen et al., 1998) and one Al³⁺-sensitive mutant (als3-1; Larsen et al., 2005) with the YC3.60 reporter and evaluated the [Ca²⁺]_cyt response after treatment with Al³⁺, La³⁺, or Gd³⁺. We again focused our imaging on the region of the root that exhibited the biphasic [Ca²⁺]_cyt response, since this pattern of [Ca²⁺]_cyt change appears to be the most distinguishing and common feature of the trivalent cation response. We observed similar trivalent-induced biphasic [Ca²⁺]_cyt signatures in the root transition zone in the wild type and the two Al³⁺ response mutants (Fig. 9, A–F). Like wild-type roots, roots of alr104 and als3-1 pretreated with 1 mm verapamil for 20 min, followed by Al³⁺ addition, displayed only a single [Ca²⁺]_cyt peak (Fig. 9, G and H).

Figure 9.

[Ca²⁺]_cyt-dependent FRET efficiency measurements in alr104 and als3-1 mutants after treatment with 1 mm Al³⁺, La³⁺, and Gd³⁺. Imaging was conducted at 400 to 500 μ m from the root cap junction, which consistently shows a biphasic [Ca²⁺]_cyt signature in response to the trivalent cations. Both alr104 (A–C) and als3-1 (D–F) show similar [Ca²⁺]_cyt signatures as the wild type. Like wild-type roots, only a monophasic [Ca²⁺]_cyt signature is observed in verapamil-pretreated roots of alr104 (G) and als3-1 (H) after Al³⁺ application. Cations were applied 45 s after the start of imaging. FRET efficiency measurements are representative of at least five independent experiments for each genotype.

Growth Analysis of Arabidopsis Roots Expressing 35S::YC3.60

Since alr104 and als3-1 mutants expressing the YC3.60 reporter did not show any obvious differences in their [Ca²⁺]_cyt response as compared with YC3.60-expressing wild-type roots after exposure to Al³⁺, we conducted short- and long-term growth assays on all three genotypes to determine whether YC3.60 transformation altered the tolerance and sensitivity of these lines to Al³⁺. We followed the growth kinetics of wild-type Arabidopsis primary roots expressing the 35S::YC3.60 reporter maintained under the identical conditions used for [Ca²⁺]_cyt measurements (i.e. the same 1 mm Al³⁺ concentration and volume used to elicit the [Ca²⁺]_cyt changes reported here). Images of growing wild-type roots were obtained at 1-min intervals for 60 min. Arabidopsis primary roots growing unperturbed on the coverslip system for 10 min elongated at a rate of 2.50 ± 0.17 μ m min⁻¹ (n = 11). When treated with 1 mm Al³⁺, the growth rate dropped to an average of 1.46 ± 0.20 μ m min⁻¹ (n = 11) for the next 50 min of imaging (Supplemental Fig. S5A). We also monitored the growth kinetics of alr104 and als3-1 lines expressing YC3.60. The growth rates of untreated alr104 and als3-1 were 3.08 ± 0.40 (n = 6) and 2.30 ± 0.37 μ m min⁻¹ (n = 6), respectively. Upon application of 1 mm Al³⁺, the growth rates of alr104 and als3-1 dropped to an average of 1.62 ± 0.1 (n = 6) and 1.09 ± 0.07 μ m min⁻¹ (n = 6), respectively, during the next 50 min of imaging (Supplemental Fig. S5A). Based on ANOVA, these growth rates were not statistically different among the three genotypes. However, two-sample t tests show that there were statistical differences between the growth rates before adding Al³⁺ (0–10 min) and after (10–60 min) in all three genotypes.

For long-term Al³⁺ exposure growth assays, we followed the protocol of Sivaguru et al. (2003), as this proved to be the most effective method for long-term application of Al³⁺ on Arabidopsis seedlings growing on agar plates. Six-day-old seedlings expressing YC3.60 of the wild type, alr104, and als3-1 were transplanted onto agar plates perfused with a range of Al³⁺ concentrations. The position of the root tip was marked on the back of the petri dish at 0 and 24 h after transplanting, and the increment between marks was measured (Sivaguru et al., 2003). The data were analyzed using ANOVA and Duncan's multiple comparison test. In the absence of Al³⁺, there was no statistical difference in root growth among the three genotypes. As expected, als3-1 was more sensitive to the growth inhibitory effects of Al³⁺ compared with the wild type at 40 and 80 μm Al³⁺, whereas alr104 was more resistant at 40 to 160 μm Al³⁺ (Supplemental Fig. S5B). The differences in growth were statistically significant at α = 0.05 as determined by Duncan's multiple comparison test. Thus, these data indicate that differences in growth among the three genotypes are most strongly manifested after long-term exposure to Al³⁺ and that expressing the YC3.60 in the mutant backgrounds did not alter the tolerance or sensitive nature of these plants to Al³⁺.

DISCUSSION

In this investigation, we employed a preloaded application wizard from the Leica TCS SP2 AOBS confocal microscope to monitor Ca²⁺-dependent FRET efficiency in Arabidopsis primary roots expressing the YC3.60 sensor. The Leica application wizard takes into account spectral bleed-through (i.e. the amount of light detected in the FRET channel that is not due to energy transfer) and normalizes for differences in donor and acceptor levels (van Rheenen et al., 2004; Feige et al., 2005) for collecting FRET-sensitized emission images. Using this approach, we were able to demonstrate [Ca²⁺]_cyt-dependent FRET efficiency oscillations in growing root hair tips (Supplemental Fig. S1; Supplemental Movie S1). We were also able to elicit [Ca²⁺]_cyt-dependent FRET efficiency changes in root cells after treatment with Glu and ATP, two compounds that are known to cause rapid spikes in [Ca²⁺]_cyt in plant cells (Fig. 1). In contrast to previous studies with Glu and ATP, which relied mostly on luminescence-based [Ca²⁺]_cyt measurements with aequorin-expressing plant cells (Dennison and Spalding, 2000; Demidchik et al., 2003; Jeter et al., 2004; Qi et al., 2006), the improved spatial resolution of FRET-based confocal imaging of YC3.60 allowed us to define regions of the root from which a particular [Ca²⁺]_cyt response originated. In doing so, we showed that all developmental regions of the root exhibited a [Ca²⁺]_cyt response to Glu and ATP. Furthermore, based on the timing of the [Ca²⁺]_cyt rise, we were able to infer the direction in which the [Ca²⁺]_cyt signal propagated in the root (Figs. 1, 3, and 4; Supplemental Movies S4 and S5).

Having established that imaging of FRET efficiency can be used to monitor [Ca²⁺]_cyt transients in roots expressing YC3.60, we asked whether this [Ca²⁺]_cyt detection method would allow us to gain additional insight into the early responses of plants to Al³⁺ and the underlying mechanisms behind the Al³⁺-induced [Ca²⁺]_cyt changes that are prevalent in the literature. Increased [Ca²⁺]_cyt in response to Al³⁺ has been reported in intact rye (Secale cereale; Ma et al., 2002), barley (Hordeum vulgare; Nichol and Oliveira, 1995), and wheat (Triticum aestivum; Zhang and Rengel, 1999; Bhuja et al., 2004) roots. In these studies, the Al³⁺-induced [Ca²⁺]_cyt increase was observed 10 to 20 min (Zhang and Rengel, 1999; Ma et al., 2002) or hours (Nichol and Oliveira, 1995; Bhuja et al., 2004) after Al³⁺ treatment. Here, we show an Al³⁺-induced [Ca²⁺]_cyt elevation after 48 to 99 s of application (Fig. 2; Table I). Lindberg and Strid (1997) observed increases in [Ca²⁺]_cyt within approximately the same time frame as those reported here using wheat root protoplasts loaded with the Ca²⁺-sensitive dye Fura 2. However, with constitutive expression of the YC3.60 reporter, we were able to expand on these earlier observations by demonstrating that different regions of an intact root have distinct responses to Al³⁺ application with regard to the onset, shape, and duration of the [Ca²⁺]_cyt response (; Table I). Furthermore, we showed that the [Ca²⁺]_cyt rise after Al³⁺ treatment generally begins in the maturation zone and propagates toward the root cap (Fig. 4). The manner of propagation of the [Ca²⁺]_cyt response is distinct from that elicited by ATP or Glu (Fig. 4). The reason for differences in [Ca²⁺]_cyt responses among the different root developmental zones upon Al³⁺ application remains unclear. However, one possibility is that cells from each root developmental region differ in their efficiency in removing Ca²⁺ because of differences in the expression of Ca²⁺-ATPases in the plasma membrane and/or endoplasmic reticulum and Ca²⁺/H⁺ exchangers in the plasma membrane and tonoplast (Sze et al., 2000; Sanders et al., 2002). Another possibility is that cells in the different root regions express certain receptor proteins that perceive the applied chemical signal, which in turn activate a specific type of Ca²⁺ channel (Kiegle et al., 2000a). Nonetheless, our ability to distinguish unique patterns of [Ca²⁺]_cyt increases in different root developmental zones in response to various chemical treatments paves the way for unraveling spatial [Ca²⁺]_cyt signaling mechanisms within a growing, intact root.

The majority of our measurements of Al³⁺-induced [Ca²⁺]_cyt elevations in the transition zone of the root were biphasic (Fig. 2). The spatial resolution provided by FRET efficiency imaging allowed us to narrow down the region of the root from where the biphasic [Ca²⁺]_cyt signature originated to cells located approximately 300 to 500 μ m from the RCJ (Fig. 5). However, the significance of the Al³⁺-induced biphasic [Ca²⁺]_cyt response is unclear. From our results, it seems likely that the first Al³⁺-induced [Ca²⁺]_cyt peak is linked to Ca²⁺ influx across the plasma membrane through Ca²⁺-permeable channels (Demidchik et al., 2002; Miedema et al., 2008) and that the second peak is triggered by the release of Ca²⁺ from internal stores (e.g. vacuole, endoplasmic reticulum, mitochondria, or plastids; Trewavas and Malhó, 1998; Sanders et al., 2002; Nomura et al., 2008). In this context, the initial rise in [Ca²⁺]_cyt may activate Ca²⁺ channels located in the tonoplast, endoplasmic reticulum, and other internal Ca²⁺ stores, resulting in the appearance of the second peak. Indeed, such a phenomenon of Ca²⁺-induced Ca²⁺ release from internal stores has been described in plant cells (Ng and McAinsh, 2003). It is also possible that the observed biphasic elevation in [Ca²⁺]_cyt represents [Ca²⁺]_cyt transients originating from two cell types, which have [Ca²⁺]_cyt signatures that differ in their amplitude and timing. To address this hypothesis, we collected [Ca²⁺]_cyt-dependent FRET efficiency data from epidermal and cortical cells in the region of the root that consistently exhibited the biphasic [Ca²⁺]_cyt response upon Al³⁺ treatment. Our measurements of FRET efficiency showed that the biphasic [Ca²⁺]_cyt signature could be traced to either epidermal or cortical cells, with the majority of the biphasic [Ca²⁺]_cyt elevations originating from the cortex (Fig. 6). However, we cannot rule out the possibility that other cell types, such as vascular, endodermal, or pericycle cells, contribute to the biphasic [Ca²⁺]_cyt signature. Expressing the YC3.60 reporter in various root cell types (Kiegle et al., 2000b) combined with imaging of FRET efficiency in the intact root should help uncover the precise nature of the Al³⁺-evoked biphasic [Ca²⁺]_cyt signature.

Despite the difference in the direction of propagation of the [Ca²⁺]_cyt wave between Al³⁺- and Glu-treated roots (Fig. 4), there is reason to believe that some aspects of the [Ca²⁺]_cyt response evoked by each chemical occurs through a common signaling pathway. Sivaguru et al. (2003) found that Al³⁺-induced microtubule reorganization and membrane depolarization could involve Glu receptors, since both cellular responses were blocked by pretreatment with AP-5, a specific neuronal Glu receptor antagonist. Furthermore, they showed that Al³⁺ and Glu applied simultaneously did not have an additive effect on microtubule reorganization and membrane depolarization. In agreement with this study, we found that Glu and Al³⁺ applied together evoked [Ca²⁺]_cyt transients that were similar to those induced by Glu alone, whereas Glu and AP-5 pretreatment inhibited the onset of the first Al³⁺-induced [Ca²⁺]_cyt peak (Fig. 8). These results reinforce the notion that an ionotropic Glu receptor that conducts cation transport across the membrane partly contributes to the Al³⁺-induced Ca²⁺ influx in the root transition zone and broadly support the findings of Sivaguru et al. (2003).

The inhibitor studies depicted here reveal additional insights into the nature of the Al³⁺-evoked biphasic [Ca²⁺]_cyt signature in the root transition zone, since pretreatment with either AP-5 or Glu inhibited only the first Al³⁺-induced [Ca²⁺]_cyt peak (Fig. 8). This suggests that in addition to the yet to be identified plant Glu receptor(s), the biphasic Al³⁺-evoked [Ca²⁺]_cyt signature in the transition zone could be a reflection of Al³⁺ acting on a cellular target that triggers the release of Ca²⁺ from internal stores rather than Ca²⁺-induced Ca²⁺ release. This notion is supported by the fact that most of the inhibitors we used in this study, including the plasma membrane Ca²⁺ channel blocker verapamil and the Ca²⁺ chelators EGTA and BAPTA, dampened or abolished the first Al³⁺-induced [Ca²⁺]_cyt peak, with minimal or no effect on the second peak. Although our experiments with AP-5, verapamil, and Ca²⁺ chelators support the idea that the Al³⁺-induced [Ca²⁺]_cyt elevations are linked to Ca²⁺ entry across the plasma membrane, it is possible that other channels could also contribute to Ca²⁺ influx, including hyperpolarization-activated Ca²⁺ channels and depolarization-activated Ca²⁺ channels (Kiegle et al., 2000a; Miedema et al., 2008). Al³⁺ has been shown to hyperpolarize (Kinraide, 1993, and refs. therein) or depolarize the plasma membrane of root cells (Sivaguru et al., 2003; Illéš et al., 2006). This could trigger both hyperpolarization-activated Ca²⁺ channels and depolarization-activated Ca²⁺ channels to open, resulting in Ca²⁺ influx into the cytoplasm and contributing to the [Ca²⁺]_cyt transients in the root. Pretreatment with NPPB, an anion channel inhibitor, also prevented the initial Al³⁺-induced [Ca²⁺]_cyt peak in the transition zone, similar to its effect on membrane depolarization and microtubule organization (Sivaguru et al., 2003). Since Glu-induced membrane depolarization was not blocked by NPPB, Sivaguru et al. (2003) hypothesized that Al³⁺ might facilitate the opening of a channel through which Glu permeates, triggering a [Ca²⁺]_cyt rise upon binding to its receptor. Since we show that NPPB also prevented the Glu-induced [Ca²⁺]_cyt increase (Fig. 8), a more likely scenario would be Al³⁺ acting on a plasma membrane receptor, which then triggers a Ca²⁺ influx. Electrophysiological studies in conjunction with [Ca²⁺]_cyt imaging and pharmacological approaches would be a fruitful area of research to better understand the precise nature of the early Al³⁺-induced [Ca²⁺]_cyt increases in the root.

Gd³⁺ and La³⁺ also triggered increases in [Ca²⁺]_cyt that were similar to those induced by Al³⁺, particularly with regard to the biphasic [Ca²⁺]_cyt response in the root transition zone (Fig. 5). These trivalent cations have been shown to inhibit specific types of Ca²⁺ currents in plants (Kiegle et al., 2000a; Demidchik et al., 2002; Qu et al., 2007) and as such have been commonly used as Ca²⁺ channel blockers. Thus, it was surprising to observe elevations in [Ca²⁺]_cyt in response to all three trivalent cations. This indicates that, at least in Arabidopsis, the biphasic [Ca²⁺]_cyt changes we observed are part of a more general signal transduction pathway in response to trivalent cations and are not directly linked to mechanisms of Al³⁺ toxicity. Despite the ability of La³⁺ and Gd³⁺ to trigger early [Ca²⁺]_cyt increases in the root, we found that a 20-min pretreatment with these trivalent cations had obvious but inconsistent effects on the Al³⁺-triggered biphasic [Ca²⁺]_cyt signature (Supplemental Fig. S3). Although such observations somewhat validate the use of La³⁺ and Gd³⁺ as Ca²⁺ signaling antagonists, additional studies are needed to clarify the mode of action of these cations on [Ca²⁺]_cyt responses. Pretreatment with La³⁺ and Gd³⁺, particularly at the concentrations used here, could on their own have a negative impact on the growth of the root. Thus, any subsequent chemical application on pretreated roots will display an altered [Ca²⁺]_cyt signature that could be more reflective of the growth status of the root. For example, the broad Al³⁺-induced [Ca²⁺]_cyt shapes that we occasionally observed in the transition zone of La³⁺- or Gd³⁺-pretreated roots (Supplemental Fig. S3B) might be indicative of responses of nongrowing cells like those observed in the mature zone or differential effects on specific cell types (Figs. 3 and 6).

An additional observation indicating that the early [Ca²⁺]_cyt signatures shown here are not specifically related to Al³⁺ toxicity is the fact that [Ca²⁺]_cyt changes were first initiated in the maturation and elongation zones, which are less sensitive to Al³⁺ compared with the transition zone (Sivaguru and Horst, 1998; Illéš et al., 2006). This notion is further supported by the fact that alr104 and als3-1 mutants, which were previously shown to be resistant and sensitive to Al³⁺_, respectively (Larsen et al., 1998, 2005), displayed the same [Ca²⁺]_cyt responses to all three trivalent cations as wild-type seedlings (Fig. 9). The als3-1 mutant, which is disrupted in a gene that encodes a phloem-localized ATP-binding cassette-like transporter, was hypersensitive to Al³⁺ but not to La³⁺ (Larsen et al., 1998), indicating that the trivalent ion-induced [Ca²⁺]_cyt rise in Arabidopsis root cells is unrelated to the Al³⁺ response mechanism involving the ALS3 gene. To the best of our knowledge, the ALR104 gene has not yet been cloned, but given the similar nature of the Al³⁺-induced [Ca²⁺]_cyt response between the wild type and alr104, these early [Ca²⁺]_cyt responses might be independent of ALR104. However, since significant differences in sensitivity to Al³⁺ among all three genotypes arose hours after Al³⁺ application (Supplemental Fig. S5B), we cannot rule out the possibility that differences in [Ca²⁺]_cyt responses will be more strongly manifested at later time points and other root zones. Careful documentation of [Ca²⁺]_cyt responses or oscillations in all root developmental regions during long periods of Al³⁺ exposure will allow us to more precisely describe the relationship between the Al³⁺-sensitive and -resistant natures of the mutants to Ca²⁺ signaling.

In conclusion, we demonstrate an alternative method for imaging dynamic changes in [Ca²⁺]_cyt in living roots of Arabidopsis seedlings expressing the YC3.60 reporter. This method is based on a standardized FRET-sensitized emission approach involving the use of donor-only and acceptor-only control lines to correct for optical cross talk and spectral bleed-through. Using this approach, we were able to monitor elevations in [Ca²⁺]_cyt in different growth zones of intact Arabidopsis roots treated with Al³⁺. The results we report here provide evidence that the rapid [Ca²⁺]_cyt transients induced by Al³⁺ are not tightly linked to mechanisms of Al³⁺ toxicity but rather are a feature of a common signaling pathway in response to trivalent cations, part of which involves modulation by a Glu receptor.

MATERIALS AND METHODS

Preparation of Plant Material

Surface sterilization and planting of transgenic Arabidopsis (Arabidopsis thaliana) seeds (Columbia ecotype) were conducted as described by Wang et al. (2004). The seeds were planted on sterile 48- × 64-mm coverslips layered with 2 mL of 0.5% NuSieve agarose (FMC BioProducts) containing half-strength MS salts (Caisson Laboratories) and vitamins. The coverslips were placed inside square petri dishes and placed in a growth chamber at 24°C and 40% humidity with a 16-h-light (124 μ E m⁻² s⁻¹)/8-h-dark cycle for 3 to 4 d. The petri dishes were vertically positioned to promote root growth along and against the coverslip (Wang et al., 2004).

Subcloning of the YC3.60 and Arabidopsis Transformation

The plasmid pcDNA3 containing the gene YC3.60 (GenBank accession no. AB178712) was kindly provided by Dr. Atsushi Miyawaki (RIKEN Brain Science Institute). The pcDNA3 plasmid was cut with EcoRI (New England Biolabs), and the recessed 3 ′ ends were filled in using the large Klenow fragment (Promega) of the DNA PolI. The linearized plasmid was purified using a PCR purification kit (Qiagen) and then digested with NcoI (New England Biolabs). The products were separated by 1% agarose gel electrophoresis, and the YC3.60 construct was recovered from the gel and purified using a Qiagen gel purification kit. The YC3.60 construct was subcloned into the NcoI/PmlI cloning sites of the plant binary vector 35S/pCAMBIA1390 using standard protocols. 35S::YC3.60/pCAMBIA1390 plasmid was isolated from overnight cell cultures and introduced into competent Agrobacterium tumefaciens cells (C58C1 strain). Flowering Arabidopsis plants were transformed by dipping them in medium containing A. tumefaciens cells harboring 35S::YC3.60/pCAMBIA1390 as described by Clough and Bent (1998). Similarly, A. tumefaciens harboring 35S::eCFP/pCAMBIA1390 or 35S::eYFP/pCAMBIA1390 was used to obtain transgenic Arabidopsis expressing eCFP or eYFP only. Plants were then transferred to a growth chamber and grown at 24°C/22°C day/night temperature, 50% relative humidity, under a cycle of 16 h of light (87 μ E m⁻² s⁻¹) and 8 h of dark.

Confocal Laser Microscopy and Measurement of FRET-Sensitized Emission

Imaging was done with an inverted Leica TCS SP2 AOBS confocal laser scanning microscope (Leica Microsystems) with an argon ion laser using a 20 × dry lens (numerical aperture 70) or a 63 × water-immersion lens (numerical aperture 1.2). Excitation wavelengths were 458 and 514 nm for CFP and YFP, respectively. Images were taken every 3 s for 10 min without any line averaging. Images were captured at a scanning speed of 200 MHz and a pixel resolution of 256 × 256. Single optical sections were taken at a depth of about 20 μ m from the root epidermal cell layer. At this focal plane, both epidermal and cortical cells in longitudinal view were visible. Exogenous application of chemicals did not affect the focus of the roots during imaging, since roots were securely anchored on the agarose that coated the coverslips. In cases where the root drifted slightly, focus was restored by quickly rotating the fine-adjustment knob of the microscope. Roots that drifted significantly during imaging were not included in the analysis.

Because the spectra of CFP and YFP overlap, the sensitized emission must be corrected for bleed-through of the emission of the donor (CFP) into the acceptor (YFP or cpVenus) channel and for direct excitation of YFP during the excitation of CFP (van Rheenen et al., 2004; Feige et al., 2005). Although bleed-through ratios are theoretically constant, they can vary with fluorophore intensities. Thus, corrections to account for bleed-through and for the differences in intensity between the donor and the acceptor must be performed.

The Leica application wizard for FRET-sensitized emission imaging was used to make the necessary corrections. For this, Arabidopsis roots expressing YC3.60 and either CFP or YFP alone were used each time to obtain calibration images. To maintain consistency in imaging, calibration was done for each objective used (20 × versus 63 ×) and root developmental region. Also, new calibrations were always performed on the day of a new experiment (i.e. old calibration images from the previous day were not used for an experiment conducted on another day). In this procedure, an image of the root zone of interest expressing YC3.60 was acquired first. The image was taken after optimizing settings (pinhole, laser intensity, and photomultiplier gain) for the CFP, FRET, and YFP channels. Then, images of reference roots (CFP and YFP alone) were obtained using the same settings and used to generate calibration coefficients. FRET efficiency was calculated by the Leica software using the formula described by van Rheenen et al. (2004):

where E_A is the apparent FRET efficiency; A, B, and C are the intensities of the three channels (CFP, FRET, and YFP); and a, b, and c are the calibration coefficients. The CFP-only reference generates the correction factor b, which is CFP cross talk in the FRET image, b = B/A. The YFP-only reference generates correction factors a and c. The ratio A/C is equal to a, which corrects for YFP cross-excitation in the CFP image; c = B/C and corrects for YFP cross-excitation in the FRET channel.

Chemical Treatments

Three- to 4-d-old wild-type and mutant seedlings expressing YC3.60 were kept in the agarose growth medium on coverslips and carefully placed on the stage of the confocal microscope. Stock solutions of 1 m AlCl₃, LaCl₃, GdCl₃, and Glu were made in deionized water, and a 10 mm stock solution was made for ATP. Half-strength MS medium (pH 5.7 or 4.5 adjusted with 1.5 m Tris) without vitamins was used to prepare all treatment solutions. A 1 mm working solution was used for all treatments. Using a Hamilton syringe, 25 to 50 μ L of treatment solution was delivered gently on top of the agarose layer where the root was embedded. All experiments were replicated and repeated at least five times.

Stock solutions of 50 mm verapamil (Sigma-Aldrich), 10 mm EGTA, 10 mm BAPTA, 10 mm AP-5, and 10 mm NPPB were prepared in deionized water and further diluted in half-strength MS medium (pH 4.5) to a final concentration of 1 mm. NPPB was used at a final concentration of 100 μm. Pretreatment with half-strength MS medium (pH 4.5) was used as a control for the inhibitor work.

Root Growth Assessment and Statistical Analysis

For short-term growth measurements, 4-d-old seedlings of the wild type, als3-1, and alr104 expressing YC3.60 were grown on agarose-coated coverslips as described above. The time course of root growth was monitored using a Nikon Eclipse TE300 microscope equipped with a Hamamatsu image processor CCD camera (Hamamatsu Photonics). Images were taken every 60 s for 60 min. Twenty-five microliters of 1 mm Al³⁺ solution was applied at 10 min after starting the time course. Change in root elongation was measured from individual frames using Metamorph 6.3r5 software (Molecular Devices). For long-term growth analysis, 6-d-old seedlings of the wild type, als3-1, and alr104 expressing YC3.60 were transplanted onto agar plates with or without Al³⁺ according to the methods of Sivaguru et al. (2003). ANOVA, Duncan's multiple comparison tests, and two-sample t tests were performed using the software package Number Cruncher Statistical System 97 (JL Hintze).

Supplemental Data

The following materials are available in the online version of this article.

• Supplemental Figure S1. Time course of the Ca²⁺-dependent FRET efficiency oscillations at the tip of a growing Arabidopsis root hair from a root expressing YC3.60.

• Supplemental Figure S2. Ca²⁺-dependent FRET efficiency changes in Arabidopsis primary roots after La³⁺ and Gd³⁺ application.

• Supplemental Figure S3. Ca²⁺-dependent FRET efficiency changes in the root transition zone induced by Al³⁺ after La³⁺ or Gd³⁺ pretreatment.

• Supplemental Figure S4. Al³⁺-induced Ca²⁺-dependent FRET efficiency changes in the root transition zone pretreated with 10 mm CaCl₂.

• Supplemental Figure S5. Short- and long-term growth responses of wild-type, als3-1, and alr104 roots to Al³⁺.

• Supplemental Movie S1. Ca²⁺-dependent FRET efficiency oscillations at the tip of a growing root hair.

• Supplemental Movie S2. Ca²⁺-dependent FRET efficiency increase in the root transition zone induced by Glu.

• Supplemental Movie S3. Ca²⁺-dependent FRET efficiency increase in the root transition zone induced by Al³⁺.

• Supplemental Movie S4. Ca²⁺-dependent FRET efficiency increase in the root cap and meristematic zone induced by Glu.

• Supplemental Movie S5. Ca²⁺-dependent FRET efficiency increase in the root cap and meristematic zone induced by ATP.

• Supplemental Movie S6. Ca²⁺-dependent FRET efficiency increase in the root cap and meristematic zone induced by Al³⁺.