Modulation of Root Signals in Relation to Stomatal Sensitivity to Root-sourced Abscisic Acid in Drought-affected Plants

Transcription

1 Journal of Integrative Plant Biology 2007, 49 (10): Invited Review Modulation of Root Signals in Relation to Stomatal Sensitivity to Root-sourced Abscisic Acid in Drought-affected Plants Huibo Ren 1, Kaifa Wei 1, Wensuo Jia 1, William John Davies 2 and Jianhua Zhang 3 ( 1 College of Agronomy and Biotechnology, State Key Laboratory of Plant Physiology and Biochemistry, China Agricultural University, Beijing 10094, China; 2 Department of Biological Sciences, Lancaster Environment Centre, University of Lancaster, Lancaster LA1 4YQ, UK; 3 Department of Biology, Hong Kong Baptist University, Kowloon Tong, Hong Kong, China) Abstract Stomatal sensitivity to root signals induced by soil drying may vary between environments and plant species. This is likely to be a result of the interactions and modulations among root signals. As a stress signal, abscisic acid (ABA) plays a central role in root to shoot signaling. ph and hydraulic signals may interact with ABA signals and thus, jointly regulate stomatal responses to changed soil water status. ph itself can be modified by several factors, among which the chemical compositions in the xylem stream and the live cells surrounding the vessels play crucial roles. In addition to the xylem ph, more attention should be paid to the direct modulation of leaf apoplastic ph, because many chemical compositions might strongly modify the leaf apoplastic ph while having no significant effect on the xylem ph. The direct modulation of the ABA signal intensity may be more important for the regulation of stomatal responses to soil drying than the ABA signal per se. The ABA signal is also regulated by the ABA catabolism and the supply of precursors to the roots if a sustained root to shoot communication of soil drying operates at the whole plant level. More importantly, ABA catabolism could play crucial roles in the determination of the fate of the ABA signal and thereby control the stomatal behavior of the root-sourced ABA signal. Key words: abscisic acid; drought; ph; root signal; stomatal sensitivity. Ren H, Wei K, Jia W, Davies WJ, Zhang J (2007). Modulation of root signals in relation to stomatal sensitivity to root-sourced abscisic acid in drought-affected plants. J. Integr. Plant Biol. 49(10), Available online at Traditionally stomatal conductance is believed to be closely correlated with leaf water potential ( l ), such that a decrease in l, as caused by soil drying would lead to stomatal closure. In Received 15 Mar Accepted 10 May 2007 Supported by for grants from the State Basic Research and Development Plan of China (2003CB114300), the National Natural Science Funds of China ( ), the Research Grant Council of Hong Kong (HKBU 2149/04M, HKBU 2165/05M) and the University Grants Council of Hong Kong (AoE/B- 07/99). Author for correspondence. Tel: ; Fax: ; <Jiaws@cau.edu.cn>. C 2007 Institute of Botany, the Chinese Academy of Sciences doi: /j x the past decades, numerous studies have revealed that in mild drought conditions the stomatal conductance can be regulated by long distance chemical signaling from the roots. Soil pressure chamber and root-split techniques have successfully broken the link between soil drying and reduced water uptake (Passioura 1980, 1987, 1988; Passioura and Tanner 1985; Blackman and Davies 1985; Gowing et al. 1990), and a chemical signal of soil drying from root to shoot has been strongly indicted. It has long been known that abscisic acid (ABA) is able to strongly promote stomatal closure (Jones and Mansfield 1970), and dehydration is able to induce an accumulation of ABA in plant cells (Wright 1977). Because of this, much attention has been paid to the roles of ABA in the stomatal regulation in response to soil drying. It is now well established that ABA is the predominant component controlling stomatal behavior in root to shoot signaling (Davies and Zhang 1991; Wilkinson and Davies 2002).

2 Modulation of Root Signals in Relation to Stomatal Sensitivity 1411 Although the stomatal regulation in relation to the root to shoot signaling has been extensively studied, it is still a controversial issue. One controversy is still about the relative importance of the chemical and hydraulic signaling. While a wealth of data have been accumulated to strongly support the roles of chemical signaling in the stomatal regulation (Davies and Zhang 1991; Davies et al. 1994; Sauter et al. 2001; Hartung et al. 2002; Wilkinson and Davies 2002), it is also difficult to reject the strong evidence that hydraulic signaling is responsible for the stomatal closure (Saliendra et al. 1995; Fuchs and Livingston 1996; Comstock and Mencuccini 1998; Yao et al. 2001; Comstock 2002). Another controversy is the nature of chemical signals. Abscisic acid as a root signal seems to be well established. Many studies have suggested that the increase in xylem ABA concentration can account for most of the antitranspiration activity in the xylem sap of many plant species, such as maize (Zhang and Davies 1991; Tardieu et al. 1992, Tardieu and Davies 1993; Zhu and Zhang 1997). However, there are also studies demonstrating that the stomatal closure in response to soil drying may not be caused by the ABAbased signaling (Munns et al. 1993; Munns and Cramer 1996; Holbrook et al. 2002). Using the graft technique, Holbrook et al. (2002) constructed different kinds of plants grated with the ABAdeficient (sitiens) and wild type (flacca) tomato. No matter the method of drying, differences were observed in the changing pattern in the stomatal conductance among the different kinds of grafting combinations, and with this Holbrook et al. (2002) clearly demonstrated that the stomatal closure in response to drought stress does not require ABA production by the root. The third controversy is the large variation in the stomatal sensitivity to the root signals. While some studies have demonstrated that in response to soil drying the chemical signaling can account for a decrease in stomatal conductance by more than 50% (Zhang and Davies 1989, 1990) or even more than 70% (Khalil and Grace 1993), the decrease in stomatal conductance from less than 40% to less than 20% was commonly reported in different plant species (Gollan et al. 1986; Auge et al. 1995; Yao et al. 2001). So how can we understand these arguments? In recent years it has been increasingly suggested that the drought-induced root to shoot signaling is very complicated, such that the stomatal regulation may actually not be controlled by a sole signal even though some signals such ABA may be able to play a predominant role. It seems that the stomatal regulation may be operated by the interactions or coordinations of different signals or factors produced in the root system, and furthermore, the root signals may be modified on the way from the sites of formation in the roots to sites of action in the leaves (Sauter et al. 2001; Hartung et al. 2002; Wilkinson and Davies 2002). Many questions in relation to the coordination of responses to stresses have been reviewed by Wilkinson and Davies (2002). In this article we will focus the discussion on the modulation of the root signals with emphasis on the mechanisms for the modulation of ph and ABA signals in response to soil drying. Coordination of Root Signals The ph signal As early as the 1980s, Hartung s group reported that the distribution of ABA in mesophyll cells was ph-dependent (Kaiser and Hartung 1981; Hartung et al. 1988; Hartung and Radin 1989). Based on the determination of ABA uptake in response to ph in isolated cells or tissues, Hartung s group generated a mathematic model, in which it was predicted that an increase as detected in vivo in response to drought would be enough to induce stomatal closure (Slovik and Hartung 1992a, 1992b). Based on these fundamental works on the relationship between ABA distribution and ph, it can be hypothesized that ph may be able to act as a long-distance signal if drought can cause a change in xylem ph. A few early works found that drought might indeed be able to trigger a ph change in different plant species. For example, Hartung and Radin (1989) observed that in response to drought stress the ph of xylem sap from Phaseolus coccineus roots increased from 6.3 to 7.2; and Gollan et al. (1992) found that in sunflower plants the ph of xylem sap from shoot increased from a range between 5.8 and 6.6 in wellwatered plants to 7.0 in the drought-affected plants. The data of Wilkinson and Davies (1997) strongly suggested that in response to soil drying, ph was able to act as a root to shoot signal to regulate stomatal movement in the Commelina communis plant. Drought increased the ph of xylem sap from 6.1 to 6.7. Artificial xylem sap buffered to different ph was fed to detached leaves of Commelina communis, and it was found that an increase in ph from 6.0 to 7.0 caused reduction of transpiration rate by about 50% in the presence of low concentrations of ABA. In another research in tomato, Wilkinson et al. (1998) demonstrated that as soil dried, the ph of xylem sap from root increased from 5.0 to 8.0. When artificial xylem sap buffered to difference ph was fed to the detached leaves of either wild type or ABA-deficient mutant flacca, it was found that the wild type but not flacca leaves exhibited reduced transpiration rates with the increase of ph. However, a wellwatered concentration of ABA added in the sap, just like the wild type flacca was able to exhibit a transpirational reduction with ph increased from 6.25 to These data demonstrated that the stomatal regulation by xylem ph absolutely requires the presence of ABA. It is well known that for stomatal regulation, the root-sourced ABA should accumulate in its action sites, which is normally thought as the apoplast of guard cells, and the stomatal sensitivity to root-sourced ABA should thus be modified by any factors affecting the accumulation of ABA in the action sites. As discussed above, the distribution of ABA in plant cells is strongly determined by ph. Because of this, it can be inferred that the stomatal regulation by the increased xylem ph is a function of the promoted accumulation of ABA in the action sites. The co-ordination of ph with ABA signal has

3 1412 Journal of Integrative Plant Biology Vol. 49 No been reviewed by Wilkinson and Davies (2002). It was proposed that the soil drying-induced ph increase in xylem would make apoplast of the leaf more alkaline, which would contribute to a sequestration of more ABA in the apoplast of guard cells, thus promoting the stomatal closure in the presence of ABA (Wilkinson 1999; Wilkinson and Davies 2002). Besides the ph-regulated sequestration of ABA, theoretically the stomatal sensitivity to ABA should also be determined by the binding between ABA and its receptor. The maximum binding between ABA and its receptor was reported to occur normally at a ph of 6.0 to 8.0 depending on different plants or tissues, thus a shift of ph from below 6.0 (normally detected in xylem sap) to neutrality may be able to promote the stomatal sensitivity to the root-sourced ABA (Hornberg and Weiler 1984; Fawzi et al. 2004). The roles of ph in the mediation of root to shoot signaling seem to be well established (Wilkinson 1999). Nevertheless, for each plant species much attention should be paid to the particular effect of ph in the stomatal regulation or other leaf behavior because different plant species usually exhibits rather different ph responses as the soil dries. For example, while soil drying is able to increase the ph of xylem sap in tomato, Commelina communis and barley, it has no effect on the ph of xylem sap in Hydrangea macrophylla cv Bluewave and Cotinus coggyria cv Royal Purple, and in Forsythia intermedia cv Lynwood it can even lead to a decrease in the ph (Wilkinson and Davies 2002). As discussed above, the stomatal sensitivity to root signals can greatly vary among different plant species. It remains to be investigated whether this variation is caused as a function of the big difference in the ph responses to soil drying. Hydraulic signal While the root to shoot chemical signaling has been increasingly demonstrated, there is plenty of evidence supporting a hydraulic signaling in response to soil drying (Petersen et al. 1991; Saliendra et al. 1995; Fuchs and Livingston 1996; Yao et al. 2001; Comstock 2002; Sperry et al. 2002). In studies emphasizing the hydraulic regulation of stomatal behavior, it is even thought that a lack of response to shoot water potential would be potentially fatal to plants (Comstock 2002). According to Ohm s law, Comstock set up a equation: g s = hc( s l)/d, where D is the leaf to air vapor gradient, g s the stomatal conductance, hc the hydraulic conductance, and s and l the soil and leaf water potentials, respectively. Based on this equation, it is predicted that in response to soil drying the stomatal conductance should be directly proportional to the hydraulic conductance when l has approached a constant minimum value. Consistent with this prediction, in Douglas fir Fuchs and Livingston (1996) found that the reduction in leaf conductance as a result of soil drying could be progressively reversed by the pressurization of the root system, and furthermore, once the pressurization was released the leaf conductance would return to its prepressurization levels within minutes. This finding indicated that the hydraulic signal was a predominant regulator of the stomatal behavior. Similar findings were made by Saliendra et al. (1995) in woody plants Betula occidentalis and Yao et al. (2001) in bell pepper (Capsicum annuum L. vau. Maor). Although in some cases the hydraulic signal may be able to play crucial roles in the stomatal regulation, the above equation, g s = hc( s l)/d, may not definitely imply that the leaf conductance should be directly operated by the hydraulic signal, because in terms of the cause-effect relationship between the stomatal and hydraulic conductance, it is difficult to say which is positive and which is passive. Also, it can be argued that the stomatal reponses to root pressure as outlined in the above published reports may be simply caused as a function of the mechanical influence on the guard cells, and whether the stomatal conductance may be able to directly and steadily respond to a sudden change in the hydraulic conductance remains to be further investigated. As discussed above, both the chemical and hydraulic signals seem to be involved in the root to shoot signaling as soil dries. How can we address these arguments? One possible hypothesis is that the two patterns of signaling interact with each other and jointly control the stomatal behavior. Such an idea was supported by the studies of Tardieu et al. (1992) and Tardieu and Davies (1993). Epidermal pieces of Commelina communis were incubated in media with different water potential adjusted by polyethylene glycol. In the media without ABA, the water potentials between 0.3 and 1.5 megapascals had no significant effect on stomatal aperture, however, when ABA was added to the media with a decrease of the water potential from 0.3 to 1.5 megapascals, the stomatal aperture significantly decreased, which strongly suggested that the stomatal sensitivity to ABA could be modified by water potential (Tardieu and Davies 1993). The same observation was obtained by feeding ABA into the field-grown plants over different ranges of leaf water potential (Tardieu and Davies 1993). Besides the effect of leaf water potential on stomatal sensitivity to ABA, hydraulic signals might possibly have an effect on the apparent stomatal sensitivity because it can modify the concentration or flux of the chemical message as a function of the changes of water flux (Tardieu and Davies 1993). The concept that stomatal sensitivity to a chemical message can be modulated by a hydraulic signal seems to have reconciled the conflicting reports on chemical signaling versus hydraulic signaling. Nevertheless, in regards to which signaling is playing a central role in the stomatal regulation, there is clearly plenty of debate between the chemical and hydraulic signaling. An essential component of this debate may actually arise from a somewhat misleading statement. To describe stomatal movement, the phrase stomatal closure was commonly used in published reports. Actually, under a mild drought condition any signal may not be powerful enough to cause a stomatal closure, and the commonly observed reduction in the stomatal conduction can be only between 20 50% in the absence of any detectable change

4 Modulation of Root Signals in Relation to Stomatal Sensitivity 1413 in leaf water potential. A more likely hypothesis is that both chemical and hydraulic signaling could play a central role in the stomatal regulation depending on different stages in the whole course of soil drying. Although the cases may be rather different among plant species, for a specific plant species, it is important to determine at what stage of soil drying does the chemical or hydraulic signaling play a central or integrated role, and what is the magnitude of each signal to control stomatal behavior? Whatever it may be, the chemical signaling should work in the early stages of soil drying, and for a more severe inhibition of the stomatal conductance, the hydraulic signaling should play a predominant role at a later stage of soil drying (Davies and Zhang 1991). Modulation of the Root-sourced ph Signal Modulation by xylem chemical composition In regard to the mechanisms for the modulation of ph signaling, the crucial questions that should be addressed with priority are as follows: (i) How can soil drying cause an increase in the ph of xylem sap? (ii) Supposing the leaf apoplast is just the action site, is the apoplastic ph solely governed by the xylem ph? (iii) Is there any modulation for the ph signal on the way from its production sites to action sites? In this part we mainly focus the discussion on the first two questions, and the third question will be discussed in the next part below. While substantial studies have demonstrated that soil drying can lead to an increase of ph in the xylem sap in many plant species (Wilkinson 1999), the mechanism for the soil drying-induced ph increase in xylem sap is largely un-known. Nevertheless, from some fundamental investigations, we can still understand many aspects of the ph regulation. It is well known that the ph in solution is determined by a balance among ions and anions, such that a change in the composition of ions or anions may possibly result in a change in ph. More importantly, many transport processes of chemical composition across the cell membrane are associated with H + bumps (H + -ATPase), and this must cause a change of ph in different compartments of the cells. Many ions, such as malate together with other organic acid, nitrate and ammonium are known to be powerful regulators of ph. L-Malate is a prominent organic acid in many plant tissues especially the dicotyledonous plants (Vickery 1963; Buttz and Long 1979), and in monocotyledonous plants transaconitate is the major organic acid with L-malate present in smaller quantities (Clark 1969). Clearly, any changes in the uptake, transport and assimilation of these components may be able to cause a ph shift. It has been increasingly suggested that N-containing compounds play crucial roles in the ph regulation. It was reported that under nitrate nutrition in castor oil plants, the organic acids were present in xylem sap only with trace amounts, and nitrate deprivation considerably increased the amounts of organic acids, thus increasing the ph from 5.6 to 7.3 (Kirkby and Armstrong 1980). In pepper plants, Dodd et al. (2003) also found that N-deprivation caused an alkalization of xylem sap. It was proposed that when nitrate is plentiful, nitrate will mainly be reduced in leaf cells, but when soil nitrate availability is low the nitrate reduction will be switched from shoot to root (Lips 1997), and this will produce hydroxyl ions, which are then converted to malate, thus leading to an alkalization of xylem sap (Wilkinson and Davies 2002). These observations seem to suggest that the soil drying-induced ph increase in xylem sap is closely related with possible changes in the rate of nitrate reduction in root. Consistent with this hypothesis, there is plenty of evidence demonstrating that soil drying can strongly affect the activity of the nitrate reductase in root, but in contrast, most of the studies demonstrate that soil drying results in a decrease, not an increase in nitrate reductase (Solomonson and Barber 1990). Whether the soil drying-induced ph increase in the xylem sap is directly governed by nitrate reduction remains to be further investigated. It was demonstrated that the transport of NH 4 + or NO 3 across cell membrane is commonly coupled with H + -ATPase (van Beusichem et al. 1985; Schumaker and Sze 1987), and hence NH 4 + or NO 3 may be able to affect the apoplastic ph. Kosegarten et al. (1999) found that nitrate nutrition could cause a high apoplastic ph in immature sunflower leaves, but they were not able to observe such a phenomenon in mature leaves or in immature leaves with sole NH 4 + or NH 4 + /NO 3 as nutrition. Mühling and Lauchli (2001) also found that nitrate nutrition caused more alkaline leaf apoplastic sap than ammonium nutrition, in both Phaseolus vulgaris and sunflower, but not in Vicia faba or Zea mays. More recently, using a ph ratio imaging technique, Jia and Davies (2007) clearly showed that feeding NO 3 induced a significant alkalization of the leaf apoplast, while the feeding of NH 4 + had a contrary effect (i.e. it gave rise to a significant reduction of the apoplastic ph in Commelina communis) (Figure 1A,B). More interestingly, feeding NH 4 + or NO 3 significantly decreased (for NH 4 + )or increased (for NO 3 ) the stomatal sensitivity to ABA. This evidence demonstrates that the leaf apoplastic ph can strongly be controlled by the chemical transport process coupled with the proton bump. As mentioned above, many studies demonstrate that soil drying can usually reduce the uptake of nitrate. Theoretically, the reduction in nitrate uptake should give rise to a decrease in the leaf apoplstic ph, and this seems unreasonable, because the reduction of apoplastic ph would decrease the stomatal sensitivity to ABA. To explain this question, Wilkinson and Davies (2002) proposed that the N availability may influence the species of the N-containing molecule, and under drought conditions, low nitrate availability would contribute to increases in the content of malate or other organic acid in xylem sap, thus giving rise to a higher apoplastic ph. It should be noted that the

5 1414 Journal of Integrative Plant Biology Vol. 49 No Figure 1. (A, B) Pseudo color image of xylem sap ph in the stem base and the leaf midrib of a sunflower plant determined using 2,7 -Bis- (2-carboxyethyl)-5-(and -6)-carboxyfluoresce (BCECF). (A) stem base; (B) leaf midrib; (C, D) fluorescence images of ph indicator 5-(and -6)-carboxylic acid (SNARF) in a Commelina communis leaf, showing apoplastic ph in relation to nitrate and ammonium ions fed through the transpiration stream. 20 mm nitrate or ammonium containing the ph indicator SNARF was fed to transpiring Commelina communis leaves. C, nitrate;d, ammonium. Modified from Jia and Davies (2007). soil drying-induced xylem ph increase and the changes of xylem chemical composition or content may be two different concepts. This is because NO 3,NH 4 + and some other ions have actually no significant effect on xylem ph, but once transported to the leaf they may strongly modify the appolastic ph, and such a viewpoint has been proved by the recent study of Jia and Davies (2007). In regards to ph signaling in response to soil drying, we should not only pay attention to whether soil drying may directly affect xylem ph, but also pay close attention to whether it may affect the chemical composition or content, and also whether a type of soil may directly affect the chemical composition or content. Modulation by tissues or cells in the vascular systems of plants In root to shoot signaling, a crucial question is the relationship between xylem sap in stem and apoplastic sap in leaf. It is normally thought that the xylem sap characters may be basically the same with the apoplastic sap, or at least predominantly governs the apoplastic characters. Based on this hypothesis, most of the studies on root to shoot signaling commonly try to correlate the leaf behavior with the changes in xylem sap characters, such as ABA, ph and some other chemical composition although the leaf apoplast is well known to be the action site of root signals. In the vascular system, the xylem vessel is closely related to different kinds of parenchyma cells, such that the communications of material and message between xylem vessels and the surrounding cells may commonly occur. Because of this, it is likely that the xylem signals might be modified by surrounding cells during their transport from root to shoot. Several studies reported that the leaf apoplastic ph is actually much different from that in xylem sap (Hoffmann and Kosegarten 1995; Mühling and Lauchli 2000). Interestingly, there are also some reports that some factors that affect transpiration (such as high vapour pressure deficit [VPD], photosynthetic photon flux density [PPFD] and temperature) can give rise to a changed ph in sap expressed from shoot in F. intermedia and H. macrophylla (Wilkinson and Davies 2002). Wilkinson and Davies proposed that the climatically induced changes in sap ph are a result of changes within the leaf apoplast rather than from the incoming xylem sap itself. It is not known how the aerial factor can cause the changes in sap ph. It is hypothesized that high VPD may affect leaf cell H + - ATPase activity by causing slight changes in localised water relation, and high PPFD may increase the removal of CO 2 from apopalst, thus causing an apoplastic alkalization (Hartung and Radin 1989; Wilkinson and Davies 2002). Apparently consistent with the findings described above, in sunflower, tomato and Commelina communis, Jia and Davies (2007) found that a great ph gradient exists between the stem base xylem and the leaf vascular system, and in sunflower this ph gradient may be as high as 1.5 ph units (Figure 1C,D). Also, it was found that a change in transpiration rate would be able to cause a change in the leaf xylem sap or apoplastic ph; but in contrast, increasing the transpiration rate caused a decrease rather than an increase in leaf apoplasic ph. By application of pressure on the root, the sap flow rate from the root to the leaf vascular system was manipulated, and with this, it was demonstrated that the quicker the flow rate, the lower the apoplastic ph becomes. Moreover, it was found that H + - ATPase inhibitor could inhibit the sap flow rate-induced changes in the leaf apoplastic ph, which suggested that H + -ATPase in the leaf vascular system is able to strongly modify the ph in the xylem vessel by effectively removing proton from the xylem vessel. Understandably, the removal rate of proton from the xylem vessel should be affected by sap flow rate, and this may be why transpiration may be able to affect leaf apoplastic ph. As discussed above, in F. intermedia and H. macrophylla higher VPD or PPFD was found to give rise to a higher apoplastic ph. It is not known whether the ph modulation mechanism is different

6 Modulation of Root Signals in Relation to Stomatal Sensitivity 1415 among species. Regardless, these studies have revealed that xylem sap ph can be strongly modified when passing through plant vascular systems, and the leaf vascular system actually plays a crucial role in buffering leaf apoplastic ph by reducing the direct effect of the sap coming in from the root on the leaf apoplast. It seems that the leaf apoplastic ph is strongly buffered as a function of the leaf vascular system and mesophyll cells. It remains to be investigated whether or how the ph modulation by the leaf vascular system may affect the strength of the ph signal. Modulation of Root-sourced ABA Signal Effect of radial transport As a stress signal to regulate stomatal behavior, the ABA signal needs to be transported from its production sites to the action sites. During the long-distance transport process along the xylem vessel, a radial transport between the xylem vessel and its surrounding cells or tissues also occurs (Hartung et al. 2002). It was hypothesized that the root, stele and cortex possess an equal capacity to synthesize ABA, and for a long-distance transport ABA needs to be radially transported from the cortex to the xylem vessel. Understandably, xylem ABA concentration should be determined by the radial transport rate of both ABA and water, thus transpiration-caused changes in the rate of the lateral water flow may affect xylem ABA concentration (Else et al. 1994, 1995; Hartung et al. 2002). One argument is the role of Casparian in the radial transport of ABA in root. It is proposed that Casparian may retard the efflux of ABA from the cortex and thus cause a high apoplastic ABA concentration to built up in cortex (Hartung et al. 2002). However, in Aarabidopsis, it has been recently demonstrated that the gene encoding the key enzyme AAO3, which acts in the last step of the ABA biosynthesis pathway, is solely expressed in stele, and also, Jia s group found that besides AAO3, the gene encoding the most important enzyme (i.e. NCED3) also absolutely expressed in stele (HB Ren, KF Wei and WS Jia, unpubl. data, 2007). This evidence strongly suggests that ABA may predominantly be produced in stele. The accumulation of ABA in stele rather than in cortex seems more reasonable for the ABA physiology. This is because, on one side, the production of ABA in stele can greatly retard the efflux of ABA into soil as a function of the presence of exodermis and endodermis (site of Casparian formation); and on the other side the accumulation of ABA in stele can greatly promote long-distance transport in the xylem vessel due to the lack of difficulties in ABA lateral transport in the cortex. Assuming ABA is synthesized in stele, Casparian will play crucial roles in retarding ABA efflux from root to soil rather than in retarding the uptake of ABA into the xylem vessel. Sauter and Hartung (2002) carried out a bean internode perfusion with ABA and ABA-glucose ester (ABA-GE) in the concentrations that occur in the xylem under natural conditions. By doing this, they demonstrated that ABA can be fed into the xylem from stem parenchyma, and on the other hand xylem ABA can be redistributed to the stem parenchyma (Hartung et al. 2002). In regard to the long-distance transport of ABA from the root to the shoot, it seems that much attention should be paid to the lateral inter-transport of ABA between the xylem and stem parenchyma. It is not known what the significance is for the lateral transport of ABA in stem to affect the xylem ABA concentration. The rate of ABA lateral transport is dependent on the ABA concentration gradient between xylem and stem parenchyma, thus under soil drying conditions, the ABA transport from xylem to stem parenchyma will be greatly promoted because the xylem ABA concentration is elevated, and this may give rise to losses of ABA from the xylem and decrease the ABA signal intensity. However, supposing that with the progress of soil drying the lateral transport of ABA may finally cause equilibrium of the ABA concentration between xylem and stem parenchyma, the lateral transport of ABA will posses little effect on the modulation of the ABA signal. The key point lies in how the lateral gradient of ABA is controlled. It is likely that the lateral ABA gradient is strongly regulated by ABA catabolism, because without ABA catabolism the ABA equilibrium would be formed sooner or later. More detailed discussions on the modulation of the ABA signal in relation to the ABA catabolism are given below. Catabolism As a stress signal, the most important character of the ABA signal is its intensity triggered by soil drying, which is clearly determined by its accumulation level. It is well known that ABA content in plant cells is determined by a dynamic relationship between biosynthesis and catabolism (Zeevaart 1980, 1983; Zeevaart and Creelman 1988). Theoretically, as soil dries, if the rate of ABA catabolism is kept unchanged or becomes lower, the ABA accumulation would be predominantly determined by ABA biosynthesis, because once ABA biosynthesis is promoted, the ABA content would be expected to immediately increase, and unless this promotion is stopped, the increase in ABA content would never stop. Thus, it seems that the final level of ABA accumulated is determined by the sustained time of the promotion of ABA biosynthesis. However, if the rate of the ABA catabolism is high, and more importantly, if soil drying may be able to promote the ABA catabolism, the final level of ABA accumulated would be determined by a new equilibrium between the promoted biosynthesis and catabolism. In such a case, the ABA catabolism would play crucial roles in the determination of the ABA accumulation. It is traditionally thought that the rate of ABA catabolism is relatively slow. Using an 18 O-labelling method (Creelman et al. 1987), a half-life of 15.5 h for ABA in Xanthium leaves was calculated. In response to soil drying, the initiation of ABA

7 1416 Journal of Integrative Plant Biology Vol. 49 No accumulation is very fast, which takes less than half an hour (Gowing et al. 1993; Jia et al. 1996; Jia and Zhang 1997; Zhang and Jia 1997; Zhang et al. 1997;) and within several hours the ABA accumulation may reach its maximum. Compared with this, the half-life of 15.5 h for the ABA catabolism clearly implies that the rate of ABA catabolism is very low. The 18 O-labelling method actually did not take into account the effect of biosynthesis on the disappearance of stress ABA, thus leading to a much longer apparent half-life for ABA catabolism. In contrast to the observation by Creelman et al. (1987), several lines of evidence strongly suggest that the ABA catabolism is very fast (e.g. in several species such as maize, Commelina communis and Vicia faba, the half-lives of ABA catabolism were reported to be only 1 h) (Gowing et al. 1993; Jia et al. 1996; Jia and Zhang 1997; Zhang et al. 1997). More recently, a study by Ren et al. (2007) demonstrated that under both water-stressed and non-stressed conditions, the half-lives of ABA catabolism in maize were basically the same, and moreover, it demonstrated that under both water-stressed and non-stressed conditions the dynamic curve of ABA catabolism exhibited a pattern of exponential decay, which means that the absolute rate of ABA catabolism (i.e. amount of ABA catabolized per time unit) is proportional to the amount of ABA (Figure 2). The recent study by Ren et al. (2007) together with some previous studies (Gowing et al. Figure 2. Catabolic time-course of fed abscisic acid (ABA) in leaf tissues of maize. Maize leaves were fed with 3 H-ABA through the transpiration stream for 15 min, and 3 H-ABA was then extracted and separated by the ABA antibody. The catabolic data were expressed as a percentage of the remaining 3 H-ABA of the total 3 H-ABA fed. The catabolic time-course of the fed ABA is modeled by equation: y = 100 [ exp( ln2 t/0.8119)], r 2 = , where y is the remaining amount of 3 H-ABA and t is the incubation time (h). Points are means ± SE of four samples. Modified from Ren et al. (2007). 1993; Jia et al. 1996; Jia and Zhang 1997; Zhang et al. 1997) all demonstrated that the ABA catabolism possesses two basic characters; it is very fast and proportional to the amount of ABA. Based on the fact that the rate of ABA catabolism is proportional to the amount of ABA as described above, it can be inferred that the increase in the ABA level would, sooner or later, make the rate of ABA catabolism be equal with the rate of ABA biosynthesis, when the ABA accumulation would reach its maximum. Clearly, the ABA catabolism has inhibited an unlimited accumulation of ABA, thus playing a central role in the regulation of the ABA signal intensity (Ren et al. 2007). It has been well established that a 9 cis-epoxycarotenoid dioxygenase (NCED) catalyzed reaction is a rate-limiting step in the ABA biosynthesis pathway due to the presence of a large pool of ABA precursors upstream from the NCED, and that the activation of NCED is responsible for the soil drying-induced ABA accumulation (Schwartz et al. 1997; Qin amd Zeevaart 1999; Seo and Koshiba 2002). To keep a sustained accumulation of ABA, a sustained activation of NCED would be required. Otherwise, the elevated ABA level would rapidly recover to its basic level. Such an inference is consistent with many observations (i.e. once water deficit was relieved, the elevated ABA level would rapidly recover to its basic level). Under soil drying conditions, the sustained activation of NCED would finally give rise to a depletion of the ABA precursor pool, especially in root, where the ABA precursor pool is much small (Li and Walton 1987; Norman et al. 1990; Parry et al. 1990; Parry and Horgan 1991; Ren et al. 2007). After the biosynthesis of the ABA precursor was blocked by fluridone, it was found that the ABA precursor pool could only support ABA accumulation for several hours in maize leaf (Figure 3), and in root, this supporting time was much shorter (Ren et al. 2007). Thus, to keep an elevated level of ABA under soil drying conditions, it is likely that the ABA precursor pool must be imported from the leaf. This hypothesis has been supported by the recent finding that shading leaf (i.e. inhibiting the production of ABA precursor in leaf) greatly inhibited water deficit-induced ABA accumulation in maize root (Ren et al. 2007). It is commonly thought that circulation of ABA from leaf to root plays a role in the root to shoot ABA signaling (Hartung et al. 2002). Based on the discussion above, we propose that the importing of the ABA precursor from leaf to root may play a more important role in the root to shoot ABA signaling. Besides the roles in soil drying-induced ABA accumulation in root, ABA catabolism may also play crucial roles in the modulation of ABA signaling during its long-distance transport and the final fate of the signal in leaf. As discussed above, during long-distance transport, radial transport of ABA may occur, thus possibly resulting in losses of ABA from the xylem vessel. The amount of ABA lost is determined by membrane permeability and the concentration gradient between the xylem vessel and the surrounding cells. Assuming the permeability is relatively small (Hartung et al. 2002) the ABA gradient would be an

8 Modulation of Root Signals in Relation to Stomatal Sensitivity 1417 Figure 3. Effect of fluridone treatment on the time-course of water deficit-induced abscisic acid (ABA) accumulation in attached maize leaves. Plant seedlings were fed with 200 µmol/l fluridone for 1 h, then transplanted and subjected to water deficit for different lengths of time., distilled water treatment;, fluridone treatment. The area filled with oblique lines denotes the amount of ABA contributed by the initial ABA precursor pool. Points are means ± SE of four samples. Modified from Ren et al. (2007). important factor controlling the losses of ABA. Understandably, the ABA gradient should predominantly be determined by the rate of ABA catabolism in the surrounding cells of the xylem vessel, hence it should play important roles in the regulation of the losses of ABA from the xylem vessel. There is plenty of evidence to suggest that herbaceous plant species compared with woody species are less sensitive to the root ABA signal (Saliendra et al. 1995; Fuchs and Livingston 1996; Comstock and Mencuccini 1998; Yao et al. 2001). Schulze (1991) suggested that large woody species would lack a chemical root signal, because the long transport time would make root-signaling ineffective for short-term stomatal regulation. However, Saliendra et al. (1995) argued that referring to a root signal transported at the relatively sluggish velocity as a feed-forward response to soil water status is misleading because a chemical signal will not necessarily arrive at the leaf (Yao et al. 2001). We propose that less stomatal sensitivity to root-sourced ABA may result from ABA catabolism and longer time in transport would make more ABA catabolized, thus greatly decreasing the signal intensity. The fate of root signals in leaf is the most important issue in root to shoot signaling. Any factors that may affect the fate of root-signals should play important roles in the modulation of the stomatal sensitivity to root-sourced ABA. The common fate of the ABA signal in living cells is its removal. Like any signal, the final fate of the ABA signal should also be its removal making unlimited accumulation of the root-sourced ABA in leaf impossible. Clearly, the catabolism of ABA must be the central factor controlling the fate of root-sourced ABA. An attracting issue in root to shoot signaling is the stomatal regulation in relation to the xylem ABA concentration and its flux. Although there are studies demonstrating that stomatal behavior can both respond to ABA concentration and its flux (Gowing et al. 1993; Jackson 1993), more studies seem to suggest that the stomatal behavior may predominantly respond to ABA concentration rather than ABA flux (Trejo et al. 1995; Loewenstein and Pallardy 1998; Jia and Zhang 1999). One may argue that the increase in either ABA concentration or flux can cause an increase in the amount of ABA that enters into leaf, thus theoretically both ABA concentration and flux should be able to control the stomatal behavior to root-sourced ABA. The answer may lie in the fact that the action sites and the removal sites of ABA are different. The action sites are known to be the sites for the ABA perception, and the removal sites are likely where the root-sourced ABA finally accumulate. Assuming ABA catabolism is not fast enough to remove much of the rootsourced ABA in the removal sites, and this would give rise to an ABA accumulation in the action sites. Comparatively, if the rate of ABA catabolism is fast enough to remove the root-sourced ABA even when xylem ABA concentration is much elevated, it is likely that the ABA catabolism would be able to prohibit the ABA accumulation in the action sites, thus whether ABA flux is changed, it would not directly affect the ABA concentration in the action sites. The two basic characters of ABA catabolism described above seem to imply that the ABA catabolism is powerful enough to effectively govern the accumulation of ABA and thus plays a central role in the regulation of stomatal behavior to root-sourced ABA signal. Conclusion The root to shoot stress signaling has been substantially studied in the past decades. Nevertheless, many questions remain unanswered as to why stomatal sensitivity to root signals may vary greatly between environments and plant species. A possible explanation is the coordination and modulation of the signals. Although the chemical signaling is now well established, the big issue still exists between the hydraulic and chemical signaling. To clarify this issue, more investigations are needed to determine at what stage of the soil drying and to what extent the chemical or hydraulic signaling may play an individual role, and how the two patterns of signaling may be integrated to control the stomatal behavior. For the modulation of chemical signaling, while attention has now been paid to the effect of ph on the modulation of the ABA signal, much attention should be paid to the direct modulations of both ph and ABA signals themselves. For soil drying-induced ph signaling, modulations

9 1418 Journal of Integrative Plant Biology Vol. 49 No of the xylem and leaf apoplastic ph are two different concepts. The direct modulation of the ABA signal intensity may be equally or even more important in the stomatal regulation in response to soil drying. ABA catabolism undoubtedly plays a central role in both ABA signal production and the fate of the ABA signal. With the increasing availability of a range of new technology, a full understanding of the mechanisms for the stomatal regulation in response to soil drying will be possible in the future. References Auge RM, Stodola AJW, Ebel RC, Duan X (1995). Leaf elongation and water relations of mycorrhizal sorghum in response to partial soil drying: two Glomus species at varying phosphorus fertilization. J. Exp. Bot. 46, Blackman PG, Davies WJ (1985). Root to shoot communication in maize plants of the effects of soil drying. J. Exp. Bot. 36, Buttz RG, Long RC (1979). L-Malate as an essential component of the xylem fluid of corn seedling roots. Plant Physiol. 64, Clark RB (1969). Organic acids from leaves of several crop plants by gas chromatography. Crop Sci. 9, Comstock JP (2002). Hydraulic and chemical signaling in the control of stomatal conductance and transpiration. J. Exp. Bot. 53, Comstock J, Mencuccini M (1998). Control of stomatal conductance by leaf water potential in Hymenoclea salsola (T. & G.), a desert subshrub. Plant Cell Environ. 21, Creelman RA, Gage DA, Stults JT, Zeevaart JAD (1987). Abscisic acid biosynthesis in leaves and roots of Xanthium strumarium. Plant Physiol. 85, Davies WJ, Zhang J (1991). Root signals and the regulation of growth and development of plants in drying soil. Annu. Rev. Plant Physiol. Plant Mol. Biol. 42, Davies WJ, Tardieu F, Trejo CL (1994). How do chemical Signal work in plants that grow in drying soil? Plant Physiol. 104, Dodd IC, Tan LP, He J (2003). Do increases in xylem sap ph and/or ABA concentration mediate stomatal closure following nitrate deprivation? J. Exp. Bot. 54, Else MA, Davies WJ, Whitford PN, Hall KC, Jackson MB (1994). Concentrations of abscisic acid and other solutes in xylem sap from root systems of tomato and castor-oil plants are affected by the method of sap collection. J. Exp. Bot. 45, Else MA, Hall KC, Arnold GM, Davies WJ, Jackson MB (1995). Export of abscisic acid, 1-aminocyclopropane-1 carboxylic acid, phosphate, and nitrate from roots to shoots of flooded tomato plants. Plant Physiol. 107, Fawzi A, Razem, Luo M, Liu JH, Abrams SR, Hill RD (2004). Purification and characterization of a Barley Aleurone abscisic acidbinding protein. J. Biol. Chem. 279, Fuchs EE, Livingston NJ (1996). Hydraulic control of stomatal conductance in Doughlas fir and alder seedlings. Plant Cell Environ. 19, Gollan T, Passioura JB, Munns R (1986). Soil water status affects the stomatal conductance of fully turgid wheat and sunflower plants. Aust. J. Plant Physiol. 13, Gollan T, Schurr U, Schulze ED (1992). Stomatal response to drying soil in relation to changes in the xylem sap composition of Helianthus annuus. I. The concentration of cations, anions, amino acids in, and ph of, the xylem sap. Plant Cell Environ. 15, Gowing DJ, Davies WJ, Jones HG (1990). A positive root sourced signal as an indicator of soil drying in apple, Malus domestica Borkh. J. Exp. Bot. 41, Gowing DJG, Jones HG, Davies WJ (1993). Xylem-delivered abscisic acid: the relative importance of its mass and its concentration in the control of stomatal aperture. Plant Cell Environ. 16, Hartung W, Radin JW (1989). Abscisic acid in the mesophyll apoplast and in the root xylem sap of water-stressed plants: the significance of ph gradients. Curr. Topics Plant Biochem. Physiol. 8, Hartung W, Radin JW, Hendrix DL (1988). Abscisic acid movement into the apoplastic solution of water-stressed cotton leaves: Role of apoplastic ph. Plant Physiol. 86, Hartung W, Sauter A, Hose E (2002). Abscisic acid in the xylem: where does it come from, where does it go to? J. Exp. Bot. 53, Hoffmann B, Kosegarten H (1995). FITC-dextran for measuring apoplast ph and apoplastic ph gradients between various cell types in sunflower leaves. Physiol. Plant. 95, Hornberg C, Weiler EW (1984). High affinity binding sites for abscisic acid on the plasmalemma of Vicia faba guard cells. Nature 310, Holbrook NM, Shashidhar VR, James RA, Munns R (2002). Stomatal control in tomato with ABA-deficient response of grafted plants to soil drying. J. Exp. Bot. 53, Jackson MB (1993). Are plant hormones involved in root to shoot communication? Adv. Bot. Res. 19, Jia WS, Zhang J (1997). Comparison of exportation and metabolism of xylem-delivered ABA in maize leaves at different water status and xylem sap ph. Plant Growth Regul. 21, Jia WS, Zhang JH (1999). Stomatal closure is induced rather by prevailing xylem abscisic acid than by accumulated amount of xylemderived abscisic acid. Physiol. Plant. 106, Jia WS, Davies WJ (2007). Modification of leaf apoplastic ph in relation to stomatal sensitivity to root-sourced abscisic acid signals. Plant Physiol. 143, Jia WS, Zhang J, Zhang DP (1996). Metabolism of xylem-delivered ABA in relation to ABA flux and concentration in leaves of maize and Commelina communis L. J. Exp. Bot. 47, Jones RJ, Mansfield TA (1970). Suppression of stomatal opening in leaves treated with abscisic acid. J. Exp. Bot. 21, Kaiser WM, Hartung W (1981). Uptake and release of abscisic acid by isolated photoautotrophic mesophyll cells, depending on ph gradients. Plant Physiol. 68, Khalil AAM, Grace J (1993). Does xylem ABA control the stomatal behaviour of water-stressed sycamore (Acer pseudoplatanus L.) seedlings? J. Exp. Bot. 44,