SEARCH

SEARCH BY CITATION

Keywords:

  • biosensor;
  • calcium;
  • fluorescence resonance energy transfer;
  • pH;
  • reactive oxygen species;
  • redox

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

Many plant response systems are linked to complex dynamics in signaling molecules such as Ca2+ and reactive oxygen species (ROS) and to pH. Regulatory changes in these molecules can occur in the timeframe of seconds and are often limited to specific subcellular locales. Thus, to understand how Ca2+, ROS and pH form part of plants’ regulatory networks, it is essential to capture their rapid dynamics with resolutions that span the whole plant to subcellular dimensions. Defining the spatio-temporal signaling ‘signatures’ of these regulators at high resolution has now been greatly facilitated by the generation of plants expressing a range of GFP-based bioprobes. For Ca2+ and pH, probes such as the yellow cameleon Ca2+ sensors (principally YC2.1 and 3.6) or the pHluorin and H148D pH sensors provide a robust suite of tools to image changes in these ions. For ROS, the tools are much more limited, with the GFP-based H2O2 sensor Hyper representing a significant advance for the field. However, with this probe, its marked pH sensitivity provides a key challenge to interpretation without using appropriate controls to test for potentially coupled pH-dependent changes. Most of these Ca2+-, ROS- and pH-imaging biosensors are compatible with the standard configurations of confocal microscopes available to many researchers. These probes therefore represent a readily accessible toolkit to monitor cellular signaling. Their use does require appreciation of a minimal set of controls but these are largely related to ensuring that neither the probe itself nor the imaging conditions used perturb the biology of the plant under study.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

Plants respond to endogenous and environmental stimuli through a complex, integrated signaling network that then regulates and coordinates processes ranging from the modulation of suites of genes to the control of all aspects of growth and development. Cytosolic Ca2+, reactive oxygen species (ROS), pH and even cellular redox status have emerged as key players in carrying information about stimuli and in the regulation of responses. The question that has often been raised is: if these regulators are so ubiquitously used in so many signaling pathways, how can they lead to specific plant responses? Indeed, it has been argued that in many cases they may actually represent more general regulators, with specificity of response being encoded by other, much more specific, signaling systems (Plieth, 2005; Scrase-Field and Knight, 2003; interested readers are directed to the collection of reviews on cellular regulation described in Gilroy and Davies, 2011). However, it is now generally accepted that simply measuring the changes in production/concentration of Ca2+ and ROS and pH and redox status at the whole plant level does not capture many of the dynamics that are key to understanding the action of these regulators in imposing specificity on signaling. The temporal and spatial kinetics of these regulators are thought to carry information about the stimulus that led to their change and also to impose specific regulatory patterns on responses. Similarly, their organ-wide, cellular and even subcellular dynamics can be highly variable. Changes at the subcellular level are also thought to be responsible for unique stimulus–response coupling even though diverse stimuli are being processed through apparently common signaling components (Pei and Gilroy, 2008; Dodd et al., 2010; Swanson and Gilroy, 2010). Relevant signaling-related changes have been characterized to occur in a timeframe of seconds and are often highly localized within the cell, down to responses limited to specific organelles. Therefore it has become imperative to be able to visualize the spatial and temporal dynamics of Ca2+, redox status, ROS and pH at high resolution to understand how they are involved in signaling and/or regulation.

A wide range of monitoring techniques have been used to make these kinds of measurements, ranging from the application of ion-sensitive microelectrodes to the loading of cells with fluorescent probes selective for each regulator (Swanson et al., 2011). Although such approaches continue to provide extremely revealing insights into signaling dynamics, in general they are technically demanding to both apply and interpret. Fortunately, recent advances in GFP-based sensors for Ca2+, redox status, ROS, and pH have greatly simplified the task of making these kinds of measurements, especially when the sensors are expressed in stably transformed lines. In this review, we will cover the use of some of these GFP-based probes that have been successfully used in planta (Table 1). Although this currently means that the measurements have most likely been made in Arabidopsis, the utility of these probes in other plants such as Medicago and tobacco is now being realized. In addition, we will highlight some of the artifacts and necessary controls that the experimenter should be aware of when applying this technology.

Table 1.   Examples of the application of cytosolic and subcellularly targeted GFP-based biosensors for Ca2+, pH and reactive oxygen species (ROS) in plants
BiosensorSubcellular localeTargeting methodCell typesSpeciesReferences
  1. NLS, nuclear localization signal.

Ca2+
 YC2.1CytosolRoot hair, root epidermis, guard cells, pollen tubesArabidopsis, Medicago, lily, tobaccoMiwa et al. (2006), Kosuta et al. (2008), Allen et al. (1999, 2001, 2002)), Klusener et al. (2002), Hugouvieux et al. (2001), Capoen et al. (2011), Watahiki et al. (2004)
NucleusNucleoplasmin-fusionRoot hairs, pollen tubesMedicago, tobaccoSieberer et al. (2009), Watahiki et al. (2004)
 YC3.6CytosolStomatal guard cell, roots, root hairs, cotyledons, pollen tubesArabidopsis, lotus, tobaccoWeinl et al. (2008), Krebs et al. (2011), Iwano et al. (2009), Monshausen et al. (2008, 2009, 2011)
Plasma membraneN-terminal fusion of YC3.6 with the LT16b proteinRoots, cotyledonsArabidopsisKrebs et al. (2011)
NucleusNLS from SV40 large-T proteinRoots, cotyledonsArabidopsisKrebs et al. (2011)
 YC3.1CytosolPollen, stigmatic papillaeArabidopsis, tobaccoIwano et al. (2004), Certal et al. (2008), Michard et al. (2008, 2011), Watahiki et al. (2004)
 YC4.6ERPumpkin 2S albumin signal peptide/HDEL ER retention signalPollenArabidopsisIwano et al. (2009)
 D3cpvPeroxisomeC-terminal KVK–SKL peptideRoots, cotyledonsArabidopsisCosta et al. (2010a)
TonoplastN-terminus of CBL2Roots, cotyledonsArabidopsisKrebs et al. (2011)
pH
 H148DCytosolRoots, root hairsArabidopsisFasano et al. (2001), Monshausen et al. (2007, 2009, 2011)
 pHluorinCytosolRoots, pollen tubesArabidopsis, tobaccoGao et al. (2004), Moseyko and Feldman (2001), Certal et al. (2008), Michard et al. (2008)
ApoplastChitinase signal peptideRootsArabidopsisGao et al. (2004)
 Pt-GFPCytosolRoots, leavesArabidopsisSchulte et al. (2006)
H2O2
 HyperCytosolLeaf epidermis, stomatal guard cells, suspension cell cultureArabidopsisCosta et al. (2010)
PeroxisomeC-terminal KSRM peptideLeaf epidermis, stomatal guard cellsArabidopsis, TobaccoCosta et al. (2010)
Redox
 RoGFP1/2CytosolRoots, leavesArabidopsis, tobaccoJiang et al. (2006), Meyer et al. (2007), Rosenwasser et al. (2010), Schwarzlander et al. (2009)
MitochondrionFirst 87 amino acids of the tobacco β-ATPaseRoots, leavesArabidopsis, tobaccoJiang et al. (2006), Rosenwasser et al. (2010), Schwarzlander et al. (2008, 2009)
ERChitinase targeting peptide/HDEL retention signalRoots, tobacco leaf cellsArabidopsisMeyer et al. (2007), Schwarzlander et al. (2008)
PeroxisomeC-terminal SKL peptideLeavesArabidopsis, tobaccoRosenwasser et al. (2010), Schwarzlander et al. (2008)
PlastidTransketolase target peptideLeavesArabidopsisRosenwasser et al. (2010), Schwarzlander et al. (2008)

Monitoring cytosolic Ca2+ using GFP-based probes

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

Cytosolic Ca2+ is maintained at approximately 100 nm against extracellular and organellar levels that can range into the mm level (Dodd et al., 2010). Signaling-related elevations in Ca2+ occur up to the μm range, thus the goal has been development of measurement techniques to selectively monitor Ca2+ in the 10−7–10−6 m range and which distinguish these cytosolic signals from the much higher background within organelles or the extracellular space. Microelectrodes, fluorescent dyes and now GFP-based sensors all provide the required selectivity and sensitivity (Gilroy, 1997; Swanson et al., 2011).

In 1997, the first genetically-encoded Ca2+ sensors based on GFP (such as yellow cameleon 2.1, YC2.1) were released (Miyawaki et al., 1997; Persechini et al., 1997) and these heralded a major change in the ease with which Ca2+ imaging in plants could be approached. The most widely used, the yellow cameleon sensors are comprised of cyan fluorescent protein (CFP) linked to yellow fluorescent protein (YFP) by the Ca2+-binding protein calmodulin and the M13 calmodulin-binding peptide taken from mammalian myosin light chain kinase. This yields a single protein with CFP linked to YFP through a Ca2+-dependent ‘hinge’ region. Upon binding of Ca2+, the calmodulin domain binds to the M13 peptide causing a conformational change in the hinge region that results in a change in proximity/orientation between the CFP and YFP. This alteration in the relationship between these two fluorescent proteins causes the fluorescence resonance energy transfer (FRET) between them to increase. Thus, some of the energy formerly emitted from the CFP as fluorescence is now being transferred to the YFP due to FRET, leading to a simultaneous increase in YFP and decrease in CFP fluorescence emission intensities. By monitoring the rise in FRET, a change in Ca2+ can therefore be inferred, with the increase in signal being quantitatively linked to the magnitude of the Ca2+ increase. In practice, the CFP is excited and the ratio of FRET (YFP emission):CFP emission is monitored as this corrects for many optical artifacts that could be present in the FRET signal alone. For example, an increase in FRET in one region of a cell could reflect a Ca2+ increase but could also simply be due to more of the sensor accumulating in that area. The FRET:CFP ratio corrects for these kinds of possible problems (e.g. doubling the intensity of both FRET and the CFP signal leaves the FRET:CFP ratio unchanged) and so provides a much more robust measurement than the FRET signal alone (Gilroy, 1997; Swanson et al., 2011). The original cameleons showed significant pH sensitivity; however, by introducing V68L and Q69K mutations into the YFP, pH sensitivity was greatly reduced leading to the development of sensors such as YC2.1 (Miyawaki et al., 1999) that represent the first generation of these probes to be widely applied to plants.

YC2.1 has been used to provide important insights into, for example, Ca2+ dynamics in the nucleus related to the NOD factor signaling found during the N2 fixing Rhizobium–legume symbiosis (e.g. Miwa et al., 2006; Kosuta et al., 2008). YC2.1 has also been instrumental in allowing characterization of the role for oscillatory Ca2+ signals in the ‘Ca2+ programmed’ but not the ‘Ca2+ reactive’ pathway of stomatal closure (e.g. Allen et al., 1999, 2001, 2002; Young et al., 2006), leading to current models of Ca2+ sensitivity priming in these cells (Hubbard et al., 2011). However, this probe suffers from a relatively small dynamic range in its FRET signal with Ca2+ change, exhibiting a maximal change in FRET signal of 40% on going from 0 to saturating (μm) Ca2+ (Miyawaki et al., 1997). This leads to a minimum FRET/CFP ratio (Rmin) of approximately 1.9 and Rmax of approximately 3.8, i.e. only a two-fold change in signal across its entire Ca2+ response range. Fortunately, a range of FRET sensors with improved characteristics have subsequently been developed and have also been used in plants (Swanson et al., 2011). Thus, for the newer Ca2+ indicator YC3.6, the YFP of YC2.1 was replaced with a circularly permutated version of this fluorescent protein (Nagai et al., 2004). This change increased the dynamic range of the FRET signal from 40 to 84%, resulting in an Rmin of approximately 1.45 and Rmax of approximately 9.0. This means that YC3.6 has a six-fold change in ratio on going from 0 to saturating Ca2+, greatly facilitating measurements. YC3.6 has been used extensively in plants to monitor Ca2+ dynamics, including the Ca2+ signaling linked to tip growth in root hairs and pollen tubes, touch sensing, ROS homeostasis, stomatal responses to extracellular Ca2+, responses to extracellular ATP, and Ca2+-dependent signaling during auxin response (Table 1; Iwano et al., 2004; Monshausen et al., 2009, 2008, 2011; Costa et al., 2010; Tanaka et al., 2010; Weinl et al., 2008).

These Ca2+-sensitive probes are under constant development, and recently new Ca2+ single-wavelength probes (i.e. non-ratiometric) with more color variations and significant increases in dynamic range were described by Zhao et al. (2011). These probes have been validated in animal cells but as yet there are no reports of their use in plants.

Subcellular targeting of Ca2+ probes

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

YC3.6 and other related Ca2+ sensors such as YC4.6 and D3cpv have also been fused to signal sequences to allow subcellular targeting to the cytosolic face of the tonoplast or plasma membrane, endoplasmic reticulum (ER), nucleus (with both nuclear localization and exclusion signals) and peroxisomes (Table 1; Iwano et al., 2004; Costa et al., 2010; Krebs et al., 2011). As a note of caution, in the tonoplast-targeted D3cpv line, cytosolic signal was also observed, limiting the conclusions that can be drawn from this line (Krebs et al., 2011; see also the similar caveat in interpreting the apoplastically targeted pH sensor pHluorin discussed below). There also appears to be a significantly reduced sensitivity of these sensors when localized to the plasma membrane (Heim and Griesbeck, 2004; Krebs et al., 2011), possibly related to interference with endogenous calmodulin (Heim and Griesbeck, 2004; Palmer et al., 2006). This interference arises because yellow cameleons contain the high-affinity M13 calmodulin-binding peptide that could easily bind to endogenous calmodulin. An elegant way to circumvent this potential problem was reported by Palmer et al. (2006): residues in the calmodulin domain and the corresponding M13 peptide interface were mutated to promote selective binding only between these two partners and not to endogenous calmodulins. Although it seems that expression of the cameleons to high levels will most likely simply increase the Ca2+ buffering capacity of the cell rather than interfere with calmodulin action (Miyawaki et al., 1999), controls for potential effects of the sensor on growth and development remain critical for each transgenic line developed. For example, monitoring of the growth rate, maintenance of normal cellular architecture and unaltered dynamics such as cytoplasmic streaming are all easily made measurements that can help spot potential problems with a particular line. This is especially important as most lines drive bioprobe expression using a strong promoter such as the cauliflower mosaic virus 35S (CaMV35S) promoter. Although silencing of sensor expression has been a reported problem with using CaMV35S and the ubiquitin promoter suggested as an alternative (Krebs et al., 2011), many of the measurements outlined in Table 1 have been made with CaMV35S lines. CaMV35S can drive recombinant proteins to several per cent of total plant protein (Guenoune et al., 2002) and so there is a good possibility for some high-expression lines to significantly affect plant processes. In particular, careful thought needs to be given to controls for disruption of cellular function if these sensors are used in transient expression studies where gauging expression levels is more complex and side-effects potentially less noticeable than in a stably transformed line.

There are now a wide variety of Ca2+-reporting GFP-based probes based not only on this M13–calmodulin partnership but also on other Ca2+-binding motifs such as from the Ca2+-binding protein troponin-C (Heim and Griesbeck, 2004). The widespread availability of these tools is making Ca2+ imaging readily accessible to plant researchers, especially those working on Arabidopsis, where many transgenic lines have already been developed. Quantitative FRET measurements are also relatively straightforward to make using a standard confocal microscope where, for example, the 458-nm emission line generated by the argon ion laser excites the CFP:YFP FRET partnership sufficiently to make measurements.

Many of the appropriate controls for using these sensors relate to those generally needed for any fluorescence microscopy experiment. Does imaging the sample on the microscope stage, an inherently stressful locale for most biology, affect the plant independent of any experimental treatment? It has been shown that the illumination required to excite fluorescent probes may lead to rapid accumulation of damaging ROS (Dixit et al., 2006). For example, laser illumination for confocal imaging can arrest cell division in BY2 cells (Dixit and Cyr, 2003). These potential problems are relatively easy to spot by carefully characterizing the dynamics of the sample under study and once understood are easily addressed by changes in the imaging protocol such as lowering laser power or reducing the frequency of sampling (Dixit et al., 2006).

Monitoring ROS and redox status using fluorescent dyes and GFP-based probes

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

Reactive oxygen species such as singlet oxygen (1O2), the superoxide radical (inline image), hydrogen peroxide (H2O2) and the hydroxyl radical (HO) were originally thought to be toxic byproducts of cellular metabolism. However, they are now recognized as key regulators of cellular functions ranging from defense to cell growth in both animals and plants (Apel and Hirt, 2004; D’Autreaux and Toledano, 2007; Giorgio et al., 2007; Moller et al., 2007; Swanson and Gilroy, 2010). To understand how ROS act in so many processes it has become critical to define their temporal and spatial dynamics. However, until recently, limitations on the available imaging tools have made making such measurements with confidence highly challenging.

Fluorescent dyes as ROS sensors

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

Measurements of inline image, 1O2 and H2O2 using small organic-molecule-based probes such as DanePy, singlet oxygen sensor green (SOSG), dihydroethidium (DHE), dihydrorhodamine 123 (DHR), Amplex Red, OxyBurst green and dihydrodichlorofluorescein (H2DCF) have been widely used to monitor ROS in plants (Swanson et al., 2011). These probes have revealed ROS production in processes as varied as light stress responses, tip growth and defense signaling (e.g. Orozco-Cardenas and Ryan, 1999; Foreman et al., 2003; Flors et al., 2006; Agati et al., 2007; Cardenas et al., 2008; Coelho et al., 2008; Driever et al., 2009; Liu et al., 2009). However, all these ROS-sensitive fluorescent probes have a significant limitation in that they respond with a change in fluorescence character that is effectively irreversible. Thus, it is impossible to directly monitor a decrease in ROS production. For example, both DHR 123 and OxyBurst green are oxidized by inline image, resulting in the accumulation of green fluorescent products (Henderson and Chappell, 1993; Probes, 2010). These dyes have been applied to detect the spatial kinetics of inline image accumulation in mitochondria in response to various stimuli such as heavy metal toxicity, heat stress and plant–pathogen interactions (Yamamoto et al., 2002; Baek et al., 2008; Bulgakov et al., 2008), and in the extracellular space during growth and development (Monshausen et al., 2007, 2009; Zhu et al., 2007). However, they impose very significant limitations on measuring the temporal dynamics of inline image production because of their chemically irreversible reaction properties; therefore, only the onset of accumulation of ROS could really be inferred from their use.

These dye-based probes also have other potential problems that make their application technically challenging. For example, extra care has to be taken to handle OxyBurst green variants since this probe can be oxidized by exposure to oxygen in the air (Probes, 2010). Dihydrodichlorofluorescein variants are probably the most widely used ROS sensors in both animals and plants (Dooley et al., 2004; Rhee et al., 2010), but again these probes have significant limitations. The DCF-based ROS sensors are relatively unselective to other ROS species and intracellular oxidants. In addition, DCF-based H2O2 probes are highly susceptible to photooxidation and photobleaching from the illumination system of the microscope that can lead to erroneous views of the dynamics of ROS production (Swanson et al., 2011).

Alternative probes have been developed based on GFP-based biosensors; critically, these probes show reversible changes in fluorescence to alterations in redox/ROS levels allowing both the production and reduction in these levels to be monitored. As described below, of these GFP-based sensors, roGFP1 and roGFP2, are sensitive to redox levels in the cell, whereas the sensor ‘Hyper’ reports H2O2 levels.

Monitoring redox with roGFPs

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

The roGFP1 biosensor was generated by mutating wild-type GFP with three point mutations (C48S, S147C and Q204C). An additional mutation (S65T) then produced roGFP2 (Dooley et al., 2004; Hanson et al., 2004). The roGFPs undergo significant conformational changes in an oxidizing environment by forming disulfide bonds between the two introduced Cys147 and Cys204 residues. These cysteines are located near the chromophore of GFP and so the conformational change results in the shifting of the GFP excitation peak from 400 to 490 nm. Thus, RoGFP1 shows a decreasing emission signal with 400 nm excitation and increasing emission signal with 490 nm excitation (Figure 1). The roGFPs have therefore been used to report spatio-temporal intracellular oxidation status by ratiometric analysis (490excitation/400excitation; Table 1). These biosensors have also been modified by the addition of signal sequences for the purpose of targeting to different subcellular organelles (Table 1). For example, per-roGFP1and px-roGFP2 were targeted to peroxisomes by fusing with the peroxisomal targeting peptide sequence SKL (Jiang et al., 2006; Meyer et al., 2007; Schwarzlander et al., 2008; Rosenwasser et al., 2010). mt-roGFP1 and mt-roGFP2 were localized in mitochondria by fusing with an 87 amino acid (aa) long mitochondrial localization signal peptide from the tobacco β-ATPase to their C-termini (Jiang et al., 2006; Meyer et al., 2007; Schwarzlander et al., 2008, 2009; Rosenwasser et al., 2010). cp-GFP2 and er-roGFP2 were also targeted to the plastid and ER, respectively, by fusing with the plastid targeting signal peptide TKTP and ER retention signal peptide K/HDEL to their C-termini (Jiang et al., 2006; Meyer et al., 2007; Schwarzlander et al., 2008, 2009; Rosenwasser et al., 2010). Such approaches have provided key insights into redox response and signaling such as revealing a role for the mitochondrion in sensing and signaling the cellular redox challenge in response to abiotic stress (Schwarzlander et al., 2009), or the likely central role of glutathione in mediating redox signaling in the plant (Meyer et al., 2007). Changes in the poise of cellular redox is likely to have widespread effects on multiple cellular processes and so, along with pH (discussed below), may represent a coordinating factor for diverse cellular processes. The roGFPs offer a window into the dynamics of this sometimes ignored but nonetheless extremely important cellular regulator.

image

Figure 1.  Emission and excitation spectral response of GFP-based probes. Emission and excitation spectral response of GFP-based probes for Ca2+ (YC3.6; Nagai et al., 2004), redox status (RoGFP1 and RoGFP2; Dooley et al., 2004), H2O2 (Hyper; Belousov et al., 2006) and pH (H148D and ratiometric pHluorin; Elsliger et al., 1999; Miesenbock et al., 1998) commonly used in plants. Arrows show the wavelengths typically used for ratiometric analysis of the spectral shifts in these sensors. Em, typical emission wavelength monitored; Ex, typical excitation wavelength used.

Download figure to PowerPoint

Monitoring H2O2 with Hyper

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

In contrast to the abilities of roGFP to report the overall cellular oxidation state through oxidation/reduction of introduced cysteines, oxidation of cysteines within Hyper allows it to selectively report H2O2 levels. This specificity arises as Hyper consists of the regulatory domain of an Escherichia coli transcription factor OxyR (OxyR-RD) that is naturally used by the bacterium to monitor H2O2 levels (Choi et al., 2001). The OxyR-RD is inserted into a circularly permutated YFP (cpYFP; Nagai et al., 2001). Upon exposure to H2O2, the OxyR-RD in Hyper selectively binds to H2O2 and undergoes significant conformational change because of the formation of an intramolecular disulfide bond between two cysteine residues (Cys199 and Cys208; Choi et al., 2001; Zheng et al., 1998). This conformational change alters the attached cpYFP sufficiently to shift its excitation maximum from 420 to 500 nm (emission maximum at 516 nm; Figure 1; Belousov et al., 2006). This shift in the excitation spectrum allows Hyper to be used as a ratioable H2O2 probe. These ROS-related changes in Hyper are fully reversible, since the disulfide between Cys199 and Cys208 becomes reduced when H2O2 is scavenged.

Recent work by Costa et al. (2010) presented the first successful expression of Hyper in the cytoplasm and peroxisomes of Arabidopsis by fusing the peroxisome localization peptide signal KSRM to the C-terminus of the sensor (Hyper-KSRM) (Table 1; Costa et al., 2010). They demonstrated that alterations in H2O2 could be detected in response to treatments with as little as 50 μm exogenous H2O2. These challenges were thought to raise Ca2+ levels, stimulating catalase activity in peroxisomes and resulting in enhanced peroxisomal H2O2 scavenging efficiency. When we have monitored Hyper in Arabidopsis leaves, an increase in excitation ratio [i.e. an increase in (H2O2)] is seen in response to a high concentration of exogenous H2O2 applied to guard cells, and this increase lasted for ∼30 sec (Figure 2). However, a very low signal increase was observed in response to lower exogenous H2O2 concentrations (e.g. 100 μm). Taken together with the report from Costa et al., these observations indicate that Hyper appears to be a useful tool to measure the real-time in vivo spatio-temporal dynamics of H2O2 production in plants. These findings also highlight that the effects on intracellular ROS of a challenge with exogenous H2O2 are likely to be extremely cell-type specific and transient, with guard cells, for example, showing a remarkably efficient system to maintain ROS homeostasis.

image

Figure 2.  Responsiveness of Hyper in guard cells to exogenous H2O2. (a) Arabidopsis stably expressing Hyper driven by the 35S promoter (Costa et al., 2010) was challenged with 100 mm H2O2 and imaged using a Zeiss LSM 510 confocal microscope (excitation 458/488 nm; emission 530 nm). Images have been pseudocolor coded according to the inset scale. (b) Quantitative analysis of the 488/458 nm ratio from Hyper in a guard cell measured using iVision image analysis software (Biovision Technologies Inc., Exton, PA, USA). Note transient nature of the Hyper response (red time points in a and b) even though exogenous H2O2 is constantly present.

Download figure to PowerPoint

Although Hyper is a powerful tool for imaging the dynamics of H2O2, it is not without its own limitations. In particular, Hyper fluorescence is affected by the pH of the cellular milieu. Thus, deprotonation of Hyper as pH rises results in an increasing Hyper signal that could be mistaken for an increase in ROS production. A change in pH from 7.0 to 7.5 gives an approximately three-fold change in Hyper ratio, or an apparent increase from 0 to ∼100 nm H2O2; Hyper is fully oxidized at 250 nm H2O2, so this pH artifact mimics a significant part of its ROS-sensing range (Belousov et al., 2006). Therefore, without careful controls to monitor any possible pH changes, interpreting results from this probe as reflecting H2O2 levels must remain tentative. Fortunately, GFP-based probes for pH are well developed for plants and offer one avenue to test whether pH changes may be interfering with Hyper’s ROS reporting abilities in a particular experiment.

Indeed, potential artifacts from the pH sensitivity of these GFP-based sensors are not uncommon. For example, in the first genetically encoded Cl sensor (a FRET construct called Clomeleon; Kuner and Augustine, 2000) pH strongly influenced the affinity of YFP for Cl (Wachter et al., 2000). Such pH sensitivity is even seen in subsequently improved versions of Clomeleon (Markova et al., 2007). However, the recently released Cl sensor ClopHensor (Arosio et al., 2010) measures Cl and pH simultaneously, providing a very powerful way to address this potential artifact by allowing calculation of Cl levels corrected for any alterations in pH. Although Clomeleon has been used in Arabidopsis (Lorenzen et al., 2004), the potential of ClopHensor for plant research has yet to be fully explored.

Monitoring cytosolic and cell wall pH using GFP-based probes

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

The cytoplasmic pH of a plant cell is generally thought to be maintained at around 7.2 by the action of proton-translocating ATPases located on the plasma and internal membranes, which remove protons from the cytoplasm. This activity is necessary to counter the influx of protons into the cytoplasm from the acidic wall or from acidic organellar spaces, which may differ from cytoplasmic pH by several orders of magnitude. In addition these elements of the pH-stat of the cell remove protons generated in the cytoplasm by metabolic processes. However, similar to cytoplasmic calcium concentrations, the cytoplasmic pH of a local region in an organ or in a subcelluar region can change rapidly in response to a stimulus. In many cases, blocking a stimulus-induced pH change can prevent the proper response; therefore pH can play a crucial role in plant growth and development (e.g. Scott and Allen, 1999; Gao et al., 2004; Monshausen et al., 2009, 2011). Shifts in pH can cause dramatic changes in protein conformation and activity, along with altering the protonation state of a number of plant hormones. Thus, precise measurement of cytoplasmic pH can provide insight into a key plant cell regulator.

Fluorescent probes able to measure proton concentration in conjunction with fluorescence microscopy have enabled high-resolution measurements of this critical ion with huge improvement in spatial and temporal detail when compared with classical methods of measuring pH such as the use of microelectrodes, 31P-NMR, or assessing the partitioning of a probe and then, based on its pKa, inferring the pH (for a review see Kurkdjian and Guern, 1989). Fluorescent chemical pH probes for measuring cytoplasmic pH, such as carboxySNAFL-1 or 2′7′-bis-(2-carboxyethyl)-5-carboxyfluorescein (BCECF) have proven extremely important in revealing the subtleties of dynamic pH shifts in cells and organs (reviewed by Roos, 2002). However, the need to load these chemical pH indicators into the cytoplasm of plant cells can pose a significant challenge; indeed some loading techniques (e.g. acid loading) can result in stress to the cell or other undesirable effects (Swanson et al., 2011). In addition, limiting the accumulation of a chemical pH indicator to only the cytoplasm is often impossible due to sequestration of the probe into the vacuole or other subcellular compartments. This can lead to fluorescence from multiple compartments potentially confounding measurements (Swanson et al., 1998).

A genetically encoded pH indicator obviates many of the drawbacks of these chemical probes, at least in plant species amenable to transient or stable transformation. The chromophore of GFP is inherently pH sensitive (Miesenbock et al., 1998; Elsliger et al., 1999), but the fluorescence of the native protein is relatively pH insensitive as its beta-barrel structure protects the chromophore from interacting with protons in the surrounding environment. Modifications to this structure allow protons to move in and out of the chromophore region and so have allowed the design of fluorescent protein-based pH indicators. Development efforts have led to a number of pH sensors, including pH GFP H148D, pHluorins and Pt-GFP (Miesenbock et al., 1998; Elsliger et al., 1999; Schulte et al., 2006). Thus, mutation at position 65 (S65T) created a GFP with a pKa of 6.0 and a second mutation at position 148 (H148D) shifted the pKa up to 7.8 (creating GFP H148D; Elsliger et al., 1999). pH GFP H148D is a dual-excitation sensor: the fluorescence emission intensity at 488 nm excitation is pH dependent and emission with 458 nm excitation is relatively pH-independent (Figure 1). Thus, it is possible use this indicator with ratio analysis.

H148D has been used to monitor cytoplasmic pH dynamics in the Arabidopsis root where it has shown, for example, oscillatory cytoplasmic pH changes associated with root hair growth (Monshausen et al., 2007). In this work, correlation to cell elongation rate showed a decrease in pH following a burst of growth, leading to a model where cell expansion triggers proton influx from the wall and concomitant cytoplasmic acidification related to limiting runaway growth (Monshausen et al., 2007). H148D has also revealed cytoplasmic acidification related to mechanical signaling in the Arabidopsis root (Monshausen et al., 2009) and the likely role of ROP GTPases in the sensitivity of root hair pH oscillations and an overall cytoplasmic pH to NO3 (Bloch et al., 2011a,b).

pHluorin is a second ratiometric pH sensor developed by structure-directed combinatorial mutation (Miesenbock et al., 1998). It has a usable range between pH 5.5 and 7.5; pHluorin data are collected by capture of the 535 nm emission with 380–410 nm excitation (increasing signal as pH increases) and 470 nm excitation (decreasing signal as pH increases) with subsequent ratio analysis (Figure 1). pHluorin has been used in Arabidopsis to show pH gradients between different developmental regions with, for example, root cap cell pH reportedly differing from the root meristem (Moseyko and Feldman, 2001). However, this difference was seen in only 50% of vertically growing roots, and so a relationship to the function of these cells has yet to be defined. Tobacco pollen tubes transiently expressing pHluorin have also been reported to show a tip-focused pH gradient, with a relatively acidic tip (Certal et al., 2008; Michard et al., 2008). pHluorin has also been used to infer that cells maintain a remarkably tight pH homeostasis even under toxic levels of salt stress (Gao et al., 2004), a finding confirmed by equivalent measurements with a third kind of pH-sensing biosensor, Pt-GFP (Schulte et al., 2006).

The ‘standard’ GFP that is the basis for most of the sensors described above was isolated from the jellyfish Aequorea victoria. The pH sensor Pt-GFP is based on a different fluorescent protein isolated from the orange seapen Ptilosarcus gurneyi (Schulte et al., 2006). Compared to pHluorins, Pt-GFP is less sensitive to pH; however, it does have a broader pH-responsiveness (usable between pH 4.7 and 7.6). Pt-GFP is a ratioable pH sensor with the excitation ratio of 475/390 nm (emission collected at 540 nm) leading to an increasing ratio with increasing pH. Pt-GFP has been stably expressed in Arabidopsis and used to show root cell cytoplasmic acidification in response to anoxia (Schulte et al., 2006).

These pH sensors have not yet been targeted to specific subcelluar locales, with the exception of pHluorin targeting to the cell wall (Gao et al., 2004). Observation of the proton concentration in the cell wall and areas immediately outside the plant cell can reveal how signaling outside the plasma membrane is coordinated with ionic changes inside the cell during signal transduction and growth. Similar to cytoplasmic pH, numerous chemical pH probes have been used to report apoplastic pH (for a review see Swanson et al., 2011). For example, fluorescein linked to a dextran, to keep it outside the cell, can be used to report extracellular pH concomitantly with a cytosolic pH-sensitive-GFP to indicate intracellular pH (Monshausen et al., 2007, 2009). Wall-targeted and cytosolic pHluorins have been used to monitor both apoplastic pH and cytoplasmic pH in Arabidopsis roots, allowing their coordinated dynamics to be analyzed. For example, measurements with wall-targeted and cytosolic pHluorin have revealed that salt stress but not mannitol treatment leads to alterations in the pH of both compartments (Gao et al., 2004). Although Gao et al. found wall pH was generally close to pH 6.0, the fluorescence from pHluorins, has been shown to be irreversibly quenched at pH ≤ 4.0 due to pH-induced conformational changes in the protein (Hanson et al., 2002), making their use in more highly acidic compartments problematical. Pt-GFP is much more stable at low pHs, even as low as 2.5 (Schulte et al., 2006), offering the possibility of an improved pH sensor for acidic compartments such as the vacuole where GFP probes have traditionally been hard to use due to low-pH-induced quenching (Fluckiger et al., 2003). It is also important to note that as the use of pHluorin as an apoplastic sensor requires it to be secreted to the wall, there is the possibility of a contaminating signal from sensor in transit in the secretory apparatus that must be carefully assessed.

Conclusions and perspectives

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References

The quantitative image analysis needed to use the GFP-based probes described above is not complex and is readily approached with many commercial and public domain image analysis packages such as ImageJ (Rasband, 1997). With the application of appropriate controls and an appreciation for the limitations of these sensors, they offer increasingly more detailed views into the dynamics of signaling within the plant.

Coordinate changes in Ca2+, ROS and pH can often be seen to arise in what may be a conserved response cassette (Monshausen et al., 2007, 2009, 2011). Thus, it is critical to be aware of the potential for changes in each of these factors to appear as artifactual changes in the response of probes to the other signaling molecules. As outlined above, this is especially true when using Hyper to probe for changes in H2O2.

It is also important to note that we have limited our discussion of ROS species to inline image, 1O2 and H2O2, but nitric oxide (NO) also represents a key regulator of plant function with, for example, strong links to Ca2+ signaling in both stomata (Kwak et al., 2006) and pollen tubes (Prado et al., 2004, 2008). A genetically encoded sensor for NO remains an important gap in the arsenal of signaling sensors.

The measurements of Ca2+ dynamics revealed by these probes show that this ion does indeed exhibit the complex spatial and temporal kinetics consistent with the idea of signaling signatures that carry information within the cell. Taken together, the studies using pH-sensitive GFPs have led to the somewhat unexpected finding that rather than a tightly buffered, unchanging cytosolic pH, in response to wide range of endogenous and external stimuli cells can exhibit changes in pH that are highly dynamic. Such shifts in pH should be having widespread effects on cellular biochemistry, suggesting that along with redox, pH could be acting as a global coordinator of cell activities, shifting the poise of the cell between signaling/response states.

Although we have concentrated on describing probes for Ca2+, redox status, ROS and pH, it is important to note that this is just the ‘tip of the iceberg’ of the GFP-based bioprobes that are available or under development. Similar sensors exist for imaging the dynamics of features ranging from cellular metabolites (Bermejo et al., 2011; Hou et al., 2011) to enzyme activities (Ting et al., 2001; Gallegos et al., 2006). As with the Ca2+, ROS and pH probes, these other biosensors are already pushing back the boundaries of our understanding of the in vivo cellular dynamics that underlie key regulatory networks at scales from the subcellular to the whole plant.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Monitoring cytosolic Ca2+ using GFP-based probes
  5. Subcellular targeting of Ca2+ probes
  6. Monitoring ROS and redox status using fluorescent dyes and GFP-based probes
  7. Fluorescent dyes as ROS sensors
  8. Monitoring redox with roGFPs
  9. Monitoring H2O2 with Hyper
  10. Monitoring cytosolic and cell wall pH using GFP-based probes
  11. Conclusions and perspectives
  12. Acknowledgements
  13. References