Developmental Physiology of Cluster-Root ... - Plant Physiology

Developmental Physiology of Cluster-Root Carboxylate Synthesis and Exudation in Harsh Hakea. Expression of Phosphoenolpyruvate Carboxylase and the Alternative Oxidase1 Michael W. Shane*, Michael D. Cramer2, Sachiko Funayama-Noguchi3, Gregory R. Cawthray, A. Harvey Millar, David A. Day, and Hans Lambers School of Plant Biology, Faculty of Natural and Agricultural Sciences (M.W.S., M.D.C., S.F.-N., G.R.C., H.L.); and School of Biomedical and Chemical Sciences, Faculty of Life and Physical Sciences (H.M., D.A.D.), The University of Western Australia, Crawley, Western Australia 6009, Australia

Harsh hakea (Hakea prostrata R.Br.) is a member of the Proteaceae family, which is highly represented on the extremely nutrientimpoverished soils in southwest Australia. When phosphorus is limiting, harsh hakea develops proteoid or cluster roots that release carboxylates that mobilize sparingly soluble phosphate in the rhizosphere. To investigate the physiology underlying the synthesis and exudation of carboxylates from cluster roots in Proteaceae, we measured O2 consumption, CO2 release, internal carboxylate concentrations and carboxylate exudation, and the abundance of the enzymes phosphoenolpyruvate carboxylase and alternative oxidase (AOX) over a 3-week time course of cluster-root development. Peak rates of citrate and malate exudation were observed from 12- to 13-d-old cluster roots, preceded by a reduction in cluster-root total protein levels and a reduced rate of O2 consumption. In harsh hakea, phosphoenolpyruvate carboxylase expression was relatively constant in cluster roots, regardless of developmental stage. During cluster-root maturation, however, the expression of AOX protein increased prior to the time when citrate and malate exudation peaked. This increase in AOX protein levels is presumably needed to allow a greater flow of electrons through the mitochondrial electron transport chain in the absence of rapid ATP turnover. Citrate and isocitrate synthesis and accumulation contributed in a major way to the subsequent burst of citrate and malate exudation. Phosphorus accumulated by harsh hakea cluster roots was remobilized during senescence as part of their efficient P cycling strategy for growth on nutrient impoverished soils.

In some plant species, a shortage of phosphorus induces the development of dense clusters of determinate branch roots (rootlets) that arise, en masse, from a localized region of the parent root axis. These short-lived structures have been termed proteoid roots because they were first described for Proteaceae (Purnell, 1960) but have since been found in a wide range of other species and families and are now often referred to as cluster roots (Lambers et al., 2003). Most of our advances in cluster-root biology have been derived from studies of the crop species Lupinus albus (Fabaceae family) (Gardner et al., 1983; Johnson et al., 1994, 1996a, 1996b; Keerthisinghe et al., 1998; Neu1 This work was supported by the Australian Research Council. M.W.S. was the recipient of an International Postgraduate Research Scholarship and a University of Western Australia Postgraduate Award. 2 Permanent address: Department of Botany, University of Cape Town, Private Bag, Rondebosch 7701, South Africa. 3 Present address: Department of Biology, Graduate School of Science, Osaka University, 1–16 Machikaneyama, Toyonaka, Osaka 560–0043, Japan. * Corresponding author; e-mail [email protected]; fax 61–8–6488–1108. Article, publication date, and citation information can be found at www.plantphysiol.org/cgi/doi/10.1104/pp.103.035659.

mann et al., 1999; Watt and Evans, 1999a, 1999b). The importance of developing short-lived cluster roots, in terms of nutrient acquisition from the rhizosphere, is illustrated by the tremendous increase in fine-root surface area following cluster-root initiation (Lamont, 1972; Dell and Kuo, 1980). Furthermore, detailed timecourse studies of cluster-root development in L. albus have demonstrated that large amounts of carboxylates (e.g. citrate) are released from cluster roots during a brief developmental window, lasting only 1 to 2 d (Dinkelaker et al., 1995; Keerthisinghe et al., 1998; Watt and Evans, 1999a; Hagstro¨m et al., 2001). Carboxylates such as citrate play an important role in P mobilization from both inorganic (Hoffland et al., 1989; Jones, 1998) and organic (Gerke, 1992) sources. It is now generally accepted that cluster root development provides an enhanced surface area for the release of large amounts of phosphate-solubilizing compounds in to a relatively small volume of soil (Gardner et al., 1983; Lambers et al., 2003). Given that the rates of carboxylate exudation from cluster roots are among the fastest known for plant roots (Jones, 1998; Roelofs et al., 2001), many investigations have sought to obtain evidence of specialized carbon metabolism to allow this enhanced carboxylate exudation. The evidence to support metabolic specialization comes predominately from studies with L. albus

Plant Physiology, May 2004, Vol. 135, pp. 549–560, www.plantphysiol.org Ó 2004 American Society of Plant Biologists

549

Shane et al.

(Johnson et al., 1994, 1996a, 1996b; Keerthisinghe et al., 1998; Neumann and Ro¨mheld, 1999; Neumann et al., 1999; Watt and Evans, 1999a, 1999b), and is based on three sets of observations: (1) genes involved in citrate synthesis are up-regulated when citrate exudation occurs from cluster roots (e.g. phosphoenolpyruvate carboxylase [PEPC]; Neumann et al., 1999; Kania et al., 2003; Uhde-Stone et al., 2003); (2) genes involved in citrate catabolism are down-regulated when citrate exudation occurs from cluster roots (e.g. aconitase; Neumann et al., 1999); and (3) transport, rather than net synthesis, is the rate-determining step for carboxylate exudation (Watt and Evans, 1999a). Cytosolic PEPC has been implicated as a key enzyme facilitating synthesis and exudation of carboxylates (e.g. citrate, malate; Johnson et al., 1996a, 1996b; Watt and Evans, 1999a; Ryan et al., 2001). The role of PEPC is to replenish TCA intermediates by catalyzing the carboxylation of phosphoenolpyruvate to form oxaloacetate (Johnson et al., 1994). Johnson et al. (1994) found that PEPC-mediated nonphotosynthetic CO2 fixation provided a quarter of the carbon exuded as citrate from cluster roots of L. albus. These authors also showed that enhanced synthesis of carboxylates coincided with elevated in vitro PEPC activity, increased expression of mRNA encoding PEPC, and PEPC protein abundance. By contrast, Keerthisinghe et al. (1998) and Watt and Evans (1999a) found that in vitro PEPC activities in cluster-root extracts were not correlated with the stage of maximum citrate exudation from the roots. Although PEPC is undoubtedly important for the carbon metabolism of citrate-exuding cluster roots, its expression relative to that of other proteins and to changes in internal carboxylate concentrations, carboxylate exudation rates, and respiration rates needs to be critically resolved by conducting all these measurements on the same roots along a time course of cluster-root development. The continued flux through part of the TCA cycle when intermediates like citrate are being siphoned off during the exudative burst, with a concomitantly enhanced production of NADH, requires an active mitochondrial electron transport chain. Rates of root respiration are, however, generally slower in plants grown with a severely limiting P supply (Rychter and Mikulska, 1990), and under conditions of P deficiency, low levels of cellular inorganic phosphorus (Pi) may limit the cytochrome (cyt) electron transport pathway (Theodorou and Plaxton, 1993). Moreover, in vitro activities of cyt oxidase may be severely decreased under Pi deficiency (e.g. in bean roots; Rychter and Mikulska, 1990). However, in plants, the cyanideresistant, alternative oxidase (AOX) pathway branches from the main electron transport pathway and is not linked to ATP production (Millenaar and Lambers, 2003). In L. albus grown under low-P conditions, the amount of AOX protein in cluster roots increased prior to citrate exudation (Kania et al., 2003). In cluster roots that exude a substantial amount of carbon as citrate and other carboxylates, the increased expression of 550

AOX presumably reflects an increased in vivo activity. This may be needed to ensure continued electron flow in the electron transport chain and to oxidize the NADH generated during the synthesis of citrate. Although some progress has been made toward an understanding of cluster-root metabolism in L. albus, information on other species is much sparser. Since many of these other species with this root type are important components of endangered ecosystems, it is important that we extend the studies begun with L. albus to other plants. Since Purnell’s discovery of proteoid roots in Australian Proteaceae (Purnell, 1960), their physiology has hardly been investigated (Grierson and Attiwill, 1989; Lambers et al., 2002a; Shane et al., 2003). Proteaceae in Australia inhabit severely nutrientimpoverished soils and are typically nonmycorrhizal (Dinkelaker et al., 1995; Pate and Bell, 1999). This is remarkable since mycorrhizas are generally considered an adaptation to phosphate-impoverished soils (Smith and Read, 1997). Considering that Proteaceae is one of the most species-rich families in southwest Australia, which is one of the world’s hot spots for plant biodiversity (Myers et al., 2000), and that the Proteaceae probably provide the most spectacular example of cluster roots (McCully, 1975), we examined cluster-root development in harsh hakea (Hakea prostrata R.Br.). Our aim was to follow cluster-root physiology and biochemistry in harsh hakea along a developmental time-course, from initiation to senescence. This pioneer study provides new insights into cluster-root metabolism by linking the relative abundance of AOX and PEPC protein in cluster roots at each stage of development with the physiological data (i.e. CO2 and O2 flux, internal phosphorus concentration, and concentrations and exudation rates of carboxylate) obtained on the same cluster roots.

RESULTS Cluster-Root Development

Cluster roots of harsh hakea completed their entire development, from rootlet emergence to senescence, over the course of approximately 25 d (Fig. 1A). Lamont (1972, 2003) described the developmental changes in morphology and anatomy of harsh hakea, and here we extend this to characterize the associated changes in both physiological and biochemical events. Under our growth conditions, thousands of new rootlets emerged in unison from the swollen parental axis in five to six longitudinal rows (Fig. 1B). Seven to 10 d after emergence, rootlets were fully elongated, and 12 to 13 d after initiation, rootlets were without a root cap. By then, the meristem had grown out (terminology of McCully, 1999), with vascular tissues fully differentiated to the rootlet tip. Root hairs developed at the base of each rootlet before elongation was complete, and by 12 to 13 d after rootlet emergence, root-hair development had extended to the tip. Plant Physiol. Vol. 135, 2004

Carboxylate Synthesis and Exudation in Harsh Hakea

Figure 1. Roots of harsh hakea plants grown hydroponically at low P supply. A, The root at the far left is the noncluster root. The six stages of cluster-root development, labeled left to right, are according to the number of days following rootlet emergence from the swollen axis (day 0) until cluster-root senescence (days 20–21). White bar is 1.75 cm. B, Young cluster root (2–3 d old) as above showing longitudinal rows of emerged rootlets. White bar is 0.4 cm.

The rate of elongation was fast during early clusterroot development and slowed after 7 to 10 d (data not shown). Cluster-root biomass increased rapidly from rootlet emergence to about 13 d after emergence. During this time the cluster-root weight increased approximately 5-fold (Fig. 2A). The relative growth rate (RGR) was very fast during the earliest stages of cluster root development and decreased steadily after 4 to 5 d (Fig. 2B). The concentration of total protein in the developing cluster roots, on a fresh mass basis, was greater than that of the noncluster roots and peaked 3 to 4 d after rootlet emergence (Fig. 2C). The tissue P concentration of the young cluster roots (1–3 d old) was higher than that of the noncluster roots and declined thereafter (Fig. 3). Cluster-Root Respiration: O2 Uptake and CO2 Release

The rate of oxygen consumption was greatest during the first 7 d of cluster-root development, and Plant Physiol. Vol. 135, 2004

the rate was fastest for 2- to 3-d-old cluster roots (Fig. 2D). Thereafter, cluster-root respiration decreased, and the rate of oxygen consumption of 12- to13-d-old cluster roots was similar to that of noncluster roots. The slowest rate of oxygen consumption was measured for 25-d-old cluster roots. The rates of root CO2 release followed a similar pattern of change with cluster-root development to that of O2 consumption (Fig. 4A). Mostly the rate of CO2 release exceeded that of O2 consumption, resulting in a respiratory quotient (RQ) . 1 (Fig. 4B). The RQ increased to a value of 1.4 when the cluster roots were approximately 10 d old and declined thereafter (Fig. 4B). Carboxylate-Exudation Rates and Tissue Concentrations of Carboxylates

The carboxylates detected in noncluster and clusterroot exudates were malate, citrate, cis-aconitate, transaconitate, succinate, and lactate (Fig. 5A). Initially, trans-aconitate and lactate were major components of 551

Shane et al.

Figure 2. Fresh mass accumulation (A) and RGR (B) of developing cluster roots of harsh hakea. In A, each point represents the mean 6 SE (n 5 4). O2-uptake rates (C) and quantity of protein (D) of noncluster roots (Non-CR) and during cluster-root development in harsh hakea. Each point in C and D represents the mean 6 SE (n 5 4). Note that all data points refer to the same plants.

the exuded carboxylates, but in the later stages of cluster-root development, malate and citrate dominated. The rate of total carboxylate exudation was slow during the early phases of cluster-root development, increasing to a peak rate after 12 to 13 d that was considerably faster than that of noncluster roots (Fig. 5A, inset). The peak of carboxylate exudation occurred 8 d after the peak of respiratory O2 consumption. The

Figure 3. The concentration of P ([P]) in the noncluster roots (Non-CR) and during cluster-root development in harsh hakea. Each point is the mean 6 SE (n 5 4). These measurements were from the same roots as in Figure 2. DM, dry mass. 552

amount of carbon used in respiration during the peak of respiratory activity (Fig. 2D; 5.5–8.5 nmol C g21 root fresh weight [FW] s21, using an RQ of 1, as presented in Fig. 4B) was of a similar order of magnitude to that used for exudation during the peak of carboxylate exudation (Fig. 5A, inset; approximately 4 nmol C g21 root FW s21). Except for the 20- to 21-d-old cluster roots, tissue carboxylates comprised malate, cis- and trans-aconitate, and large amounts of iso-citrate and citrate (Fig. 5B). The internal malate concentration was nearly constant during cluster-root development. Although the concentration of trans-aconitate was greater than that of cis-aconitate, both decreased steadily with clusterroot age. The largest changes in internal carboxylate concentrations were for the concentrations of citrate and iso-citrate. The concentrations of these carboxylates increased with cluster root age and peaked in 12- to 13-d-old cluster roots (Fig. 5B). The internal concentration of malate remained approximately constant (5 mmol g21 FW between days 1.5 and 7.5) just before the burst of malate exudation at 13.5 d. The internal (iso)citrate concentration started at a low concentration (12–18 mmol g21 FW) and peaked during the burst of citrate and malate exudation (32–37 mmol g21 FW). After the exudative burst, the concentrations of all carboxylates declined to very low levels. The accumulation of iso-citrate and citrate in cluster-root tissues was in phase with the timing of the peak rates of citrate and malate exudation (Fig. 5A). In Plant Physiol. Vol. 135, 2004


On a fresh mass basis, the amount of porin protein increased to a peak in young (2 to 5-d-old) developing cluster roots. This shows that the increase and decrease in AOX abundances between 1- to 2-d-old and especially in 12- to 13-d-old cluster roots were due to a real increase in AOX abundance in mitochondria, rather than a general increase in mitochondrial protein. PEPC was detected as two protein bands at 110 and 100 kD apparent molecular mass (Fig. 8). These two protein bands could represent two isoforms, but it is also possible that the two bands represent different phosphorylation states or that the 100-kD band is a degradation product. According to Sanchez and Cejudo (2003), Arabidopsis has at least four PEPC isoforms (100.3–116.6 kD), and rice has at least three PEPC isoforms. In L. albus, PEPC occurs as a single 110-kD subunit (Uhde-Stone et al., 2003). The relative intensity of the two bands varied throughout clusterroot development, with the lower band increasing in intensity and the upper band decreasing. However, the combined protein abundance of both bands was largely unaffected by cluster-root development. As was observed with AOX and porin, no PEPC was detected in 20-d-old cluster roots (Fig. 8). Figure 4. O2-consumption rates and CO2 efflux (A) and the RQ (CO2 release:O2 uptake; B) of noncluster roots (Non-CR) and during clusterroot development in harsh hakea. Each point represents the mean 6 SE (n 5 4). These measurements were made on a different batch of plants from those used for Figures 2 and 3.

20- to 21-d-old cluster roots, the concentration of carboxylates fell dramatically to levels much lower than those found in noncluster roots. This, together with the very slow respiration rates of these 20- to 21-d-old cluster roots, indicates that they were metabolically relatively inactive. Concentration of Key Enzymes

Changes in specific enzyme abundances were expressed on the basis of total extracted protein. AOX protein appeared as two immunoreactive bands of approximately 30 kD in apparent molecular mass. The mitochondrial AOX protein abundance increased during early cluster-root development and peaked when cluster roots were 7 to 8 d old; thereafter, the amount of AOX protein decreased dramatically, and no AOX was detected in extracts of 20- to 21-d-old cluster roots (Fig. 6). The concentration of mitochondrial marker protein porin was assessed to determine whether changes in AOX simply reflected changes in root mitochondrial protein content. We confirmed that the protein bands were indeed AOX and porin using isolated harsh hakea mitochondria (data not shown). Porin protein abundance on a protein basis was the same in noncluster and cluster roots throughout most of their development; only in 20- to 21-d-old cluster roots was porin protein abundance decreased (Fig. 7). Plant Physiol. Vol. 135, 2004

DISCUSSION Time Course of Cluster-Root Physiology Respiration, RQ, and the Concentration of Protein and Phosphorus

The fast RGRs of cluster roots at 2 to 3 d after rootlet emergence combined with their increasing protein concentration would require large amounts of ATP, possibly accounting for the fast respiration rates observed. These rates were 2 to 6 times faster than those reported for whole root systems of a range of herbaceous species (table III in Lambers et al., 2002b). Once the rootlets reached their final length, there was a decrease in RGR and protein concentration and a concomitant reduction in the rate of respiration. The RQ values for developing cluster roots were greater than 1, which might be a consequence of nitrate reduction, protein synthesis, and/or lipid synthesis. The RQ values were never below 1, not even when the cluster roots had stopped growing. During the exudative burst, malate and citrate but not iso-citrate were the major exuded carboxylates. However, the amount of internal malate in the tissues at the time of the exudative burst could sustain only approximately 4 h of malate exudation at the observed rate. Therefore, if some of the exuded malate were derived from the decarboxylation of internal isocitrate, then this would increase the RQ values. This is because each iso-citrate decarboxylated to malate generates two CO2, whereas only one O2 is needed to reoxidize the two NADH that are produced during this conversion. 553

Shane et al.

Figure 5. Rates of carboxylate exudation (A), quantity of carbon associated with exudation (A, inset), and internal carboxylate concentrations (B) from noncluster roots (Non-CR) and during clusterroot development in harsh hakea. All measurements from the same roots as shown in Figures 2, 3, and 4. Each point represents the mean 6 SE (n 5 4).

Cluster-root P concentration was greatest 2 to 3 d after rootlet emergence. As rootlet elongation was finished approximately 7 to 10 d after rootlet emergence, the cluster-root [P] was diluted to approximately half the initial concentration. We found that 20- to 21-d-old cluster roots had a root [P] that was further reduced by 6-fold, showing very efficient remobilization of P from senescing cluster roots. The remobilization of P from these senescing cluster roots of harsh hakea is similar to the changes in Pi content described for maturing cluster roots of L. albus (Neumann et al., 1999; Massonneau et al., 2001) and roots of Stylosanthes hamata (Smith et al., 1990). Proteaceous species are well known for their efficient use of P (Handreck, 1997), 554

and efficient remobilization of phosphate from senescing cluster roots appears to be a mechanism to further enhance their ability to survive on severely P-impoverished soils. Carboxylate Exudation and Internal Carboxylate Concentration

The main focus of our physiological investigation of cluster roots has been on the exudation of carboxylates during cluster-root development. The burst of citrate and malate exudation during cluster-root development in harsh hakea is similar to the peak in malate and fumarate exudation measured over a few days in Plant Physiol. Vol. 135, 2004


Figure 6. A, Typical immunoblot of AOX proteins in whole-root tissue extracts from noncluster roots (Non CR) and during cluster-root development in harsh hakea. Each lane was 50 mg of total protein. B, Relative AOX protein abundance (intensity of the bands). The intensity of the upper band of 7- to 8-d-old cluster roots was set to 1. Gray columns, upper bands; white columns, lower bands. All measurements from the same roots as shown in Figures 2, 3, and 5. Each column is the mean 6 SD (n 5 3).

tion/export. The internal carboxylate concentration of cluster roots on day 12.5 was approximately 75 mmol (malate 1 isocitrate 1 citrate) g21 FW, which would give a maximum potential exudation rate of 0.86 nmol g21 FW s21, if all carboxylates were released over a period of 24 h; the measured peak exudation rate was 0.6 nmol (malate 1 citrate) g21 FW s21. Therefore, a gradual buildup of carboxylate stores in cluster roots of harsh hakea may supply a major part of carboxylates during the exudative burst. Our findings with harsh hakea agree with those of Neumann et al. (1999), who also showed that the citrate exudation rates of L. albus cluster roots were positively correlated with the internal citrate concentration, but contrast with the findings of Keerthisinghe et al. (1998), who showed that the internal citrate concentration in developing and mature cluster roots of L. albus did not always correlate with the rates at which citrate was exuded. This discrepancy might be due to species-specific differences or to the fact that Keerthisinghe et al. (1998) only measured citrate, as opposed to all the major carboxylates as in this study. As found here for harsh hakea, there may be conversion of one carboxylate into another, prior to exudation. Is Cellular Metabolism Specialized during Cluster-Root Carboxylate Exudation? The Role of PEPC

cluster roots of H. undulata (Dinkelaker et al., 1997). The finding that the exudative burst of carboxylates in harsh hakea followed the cessation of rootlet growth and vascular development is similar to that described for cluster roots of L. albus (Watt and Evans, 1999a; Hagstro¨m et al., 2001). The shift to fast rates of malate and citrate exudation from cluster roots 10 d after the peak in O2 consumption rate probably reflects a change from metabolism of citrate through the TCA cycle and NADH oxidation via the cyt path driving cluster-root development to metabolism associated with the exudation of citrate and malate (Fig. 5A). The oxidation of one molecule of hexose to two (iso)citrate molecules produces three molecules of NADH, oxidation of which requires 1.5 molecules of O2. Assuming that all carbon enters the mitochondria as malate (Lambers, 1997) and that hexose is the only substrate for respiration, some key components of the respiratory metabolism were calculated (Table I). According to these calculations, the O2 uptake associated with the exudation of citrate during the exudative burst accounted for 24% of the total O2 consumption, assuming that this citrate was produced from hexose. At all other stages of cluster-root development, when the rate of citrate exudation was slow, this value would be 0% to 5%. For noncluster roots, which showed faster exudation rates, this calculated value amounted to 14%. The tissue carboxylate concentration represents the net balance between synthesis/import and consumpPlant Physiol. Vol. 135, 2004

The loss of citrate from the TCA cycle through exudation requires the stoichiometric provision of anaplerotic oxaloacetate derived from PEPC for the

Figure 7. A, Typical immunoblot of porin proteins in whole-root tissue extracts from noncluster roots (Non CR) and during cluster-root development in harsh hakea. Each lane was 50 mg of total protein. B, Relative porin protein abundance (intensity of the bands). The intensity of the upper band of 7- to 8-d-old cluster roots was set to 1. Gray columns, upper bands; white columns, lower bands. All measurements from the same roots as shown in Figures 2, 3, 5, and 6. Each column is the mean 6 SD (n 5 3). 555

Shane et al.

addition to pyruvate) for the mitochondria (Lambers, 1997). The PEPC activity associated with full TCAcycle activity was calculated from the rate of O2 consumption (Table I). Similarly, the PEPC activity associated with carboxylate synthesis was calculated from the rate of carboxylate exudation (Table I). These calculations show that PEPC would be expected to reach a maximum activity of 8.6 nmol g21 FW s21 peak when respiration peaked and then to decline gradually to a value of 2.8 nmol g21 FW s21 during the exudative burst. These values may have been overestimated slightly, since not all carbon enters the mitochondria as malate, as we assumed in calculating the values shown in Table I, but they convincingly demonstrate that there is no necessity for a major change in PEPC during cluster-root development and carboxylate exudation in harsh hakea. An alternative explanation for the lack of change in PEPC protein abundance despite changes in carboxylate synthesis may be related to changes in the activation state of PEPC. PEPC is a highly regulated enzyme that is allosterically activated by Glc-6-P and triose-P and inhibited by L-malate and Asp (Schuller et al., 1990). Furthermore, the sensitivity of the protein to allosteric regulation is modified by its phosphorylation state (Chollet et al., 1996). When phosphorylated, the protein is less sensitive to allosteric regulation. There was a 4- to 5-fold increase in the exudation of citrate and malate from the cluster roots of harsh hakea as compared with that from noncluster roots (Fig. 5A). Under steady-state conditions, the loss of citrate from the TCA cycle through exudation requires the stoichiometric provision of oxaloacetate derived from PEPC. The consequence of increased PEPC activity is a decrease in net CO2 release due to an increase in CO2 consumption relative to respiratory CO2 release, which should result in a decrease in the RQ corresponding to the amount of citrate exuded (Table I). Every malate produced from hexose also requires the anaplerotic

Figure 8. A, Typical immunoblot of PEPCase protein in whole-root tissue extracts from noncluster roots (Non CR) and during cluster-root development in harsh hakea. Each lane was 50 mg of total protein. B, Relative PEPCase protein abundance (intensity of the bands). The intensity of the upper band of 12- to 13-d-old cluster roots was set to 1. Gray columns, upper bands; white columns, lower bands. All measurements from the same roots as shown in Figures 2, 3, 5, 6, and 7. Each column is the mean 6 SD (n 5 3).

maintenance of the TCA cycle. Some evidence supports the involvement of enhanced anaplerotic carbon provision to the TCA cycle for carboxylate biosynthesis in L. albus cluster roots (Johnson et al., 1994, 1996a, 1996b; Neumann et al., 1999). Western blots for PEPC in harsh hakea indicated that the relative abundance of two PEPC isoforms varied only marginally during cluster-root development. Why would there be little change in PEPC abundance? PEPC plays a role in normal respiration, providing malate as a supplementary carbon substrate (in

Table I. Calculated values for key components of the respiratory metabolism in developing cluster roots, based on the assumption that all carbon entering the mitochondria is malate, and Glc is the only substrate for respiration Rootlet Age

O2 Uptake % Associated with Carboxylate Synthesis

Expected PEPC activity Expected RQ

Respiration

Days

1.5 2.5 5.5 8.5 13.5 25 Non-CR

0.3 0.1 0.1 4.7 24.3 0.0 14.4

1.00 1.00 0.99 0.96 0.68 1.00 0.95

6 6 6 6 6 6 6

0.001 0.002 0.003 0.035 0.119 0.000 0.004

5.63 8.61 5.65 4.38 1.72 0.42 2.19

Carboxylate Synthesis

nmol CO2 g21 FW s21

0.02 0.01 0.01 0.34 1.06 0.00 0.54

6 6 6 6 6 6 6

0.01 0.00 0.00 0.11 0.27 0.00 0.07

Total

5.65 8.62 5.67 4.73 2.78 0.42 2.73

The percentage of respiratory O2 consumption associated with carboxylate synthesis was calculated by summing the component metabolic reactions for each carboxylate detected. The expected RQ (6SE) was derived from assuming a baseline RQ of 1.0 in the absence of carboxylate accumulation or exudation. PEPC activity associated with respiration was calculated from the measured values of O2 uptake, corrected for O2 uptake that was expected to be associated with carboxylate synthesis. PEPC activity for carboxylate synthesis (6SE) was calculated from the measured values of carboxylate exudation data. The change in the concentrations of the carboxylate pools within the cluster-root tissue equated to a negligible rate of PEPC activity required for the synthesis of the carboxylates (,0.08 nmol CO2 g21 FW s21) and were therefore not included. Non-CR, noncluster roots. 556

Plant Physiol. Vol. 135, 2004


provision of carbon, with no net reduction of NADH, and thus should equally result in a decrease in the RQ corresponding to the rate of malate production. Such changes in RQ values were not observed during peak carboxylate exudation (Fig. 4B). The RQ never fell below 1.0, and, in fact, increased to 1.3 in 8- to 9-d-old cluster roots. Thus, there must be (1) an additional demand for NADH without O2 consumption (e.g. NO32 reduction), (2) a decarboxylation event, such as the conversion of internal iso-citrate to malate prior to exudation, and/or (3) carboxylates provided from another source as yet unaccounted for (e.g. imported via the phloem). The second explanation is supported by our own results, whereas our data cannot provide information to support or reject the others.

The Role of AOX

The fact that maximum rates of O2 consumption were out of phase with the burst of citrate and malate exudation indicates that cluster-root metabolism does not simultaneously sustain high levels of both carbonrequiring processes. In young cluster roots, carbon is mainly used in respiration (for ATP synthesis) and to provide the carbon skeletons to produce the cluster roots. At a later stage, respiration slows down, and carbon is diverted to enhance carboxylate synthesis. When the cluster-root growth rate slows and protein concentrations decrease, NADH production associated with (iso)citrate synthesis may be in excess of ATP demand. Under these circumstances, diversion of electrons from NADH through the nonphosphorylating alternative oxidase pathway may provide a mechanism for regulating ATP production from NADH. The AOX protein abundance was maximal in western blots of 7- to 8-d-old harsh hakea cluster roots, which corresponded with the stage of cluster root development just prior to the peak carboxylate exudation rate. There may be an additional need for enhanced AOX activity when plants are grown with a limiting P supply. Rates of root respiration were decreased in L. albus grown with a severely limiting P supply (Neumann and Martinoia, 2002), as was the case for Phaseolus vulgaris (Rychter and Mikulska, 1990). In this study AOX abundance increased well before the tissue P concentration reached very low levels. In this study with harsh hakea, AOX abundance initially changed slightly (Fig. 6) when mitochondrial numbers increased (as determined by western blots for porin) in young cluster roots. Thereafter, there was a major peak in AOX abundance immediately prior to the exudative burst of citrate and malate from mature cluster roots, without a further concomitant increase in porin concentrations. Thus, in 7- to 8-d-old cluster roots, the increased expression of AOX presumably reflects an increased in vivo activity that facilitates carboxylate synthesis and/or restricted cyt pathway activity, recycling NADH. However, further studies using the O2-fractionation technique (Guy et al., 1992; Day et al., Plant Physiol. Vol. 135, 2004

1996) are required to determine the AOX activity of intact cluster-roots.

CONCLUSION

The large amount of carboxylates exuded by cluster roots is generally assumed to be produced locally rather than imported via the phloem. In the cluster roots, the carboxylate synthesis would require anaplerotic PEPC activity. Evidence from the present investigations of the entire developmental sequence of cluster roots in harsh hakea indicates that a burst of carboxylate exudation occurs from mature cluster roots. Although PEPC is undoubtedly important for the carbon metabolism of carboxylate-exuding cluster roots, it does not exhibit changes in expression relative to other proteins or to changes in internal carboxylate concentrations, exudation rates, or respiration rates in cluster roots. Instead, it appears that PEPC activity results in a gradual loading of (iso)citrate in clusterroot tissue of harsh hakea during development, followed by a burst of citrate and malate exudation when roots have matured. The burst of exudation is preceded by a reduced metabolic activity as judged from decreased protein levels and respiratory O2 consumption. Thus, it appears that one of the strategies used by harsh hakea cluster roots is directed toward carboxylate synthesis and storage, partly as precursors, during development, and then to release carboxylates in a burst. Our observation that clusterroot tissues first accumulated carboxylates indicates that carboxylate exudation rates from cluster roots of harsh hakea might be controlled by their transport out of root tissues. The production of large quantities of carboxylates for exudation during cluster-root development, which is associated with rapid rates of NADH production, coincided with increased expression of AOX protein, which is presumably needed to provide for greater reoxidation of NADH in the mitochondrial electron transport chain, in the absence of a great need for ATP synthesis.

MATERIALS AND METHODS Plant Growth Harsh hakea (Hakea prostrata R.Br.) seedlings (2 months old) were obtained from Lullfitz nurseries (Wanneroo, Australia). The root systems were washed free of soil with water, and all lateral roots were removed from the short (20-mm) taproot. Seedlings of uniform size were selected and transferred to 7-L black plastic pots, with their shoots supported by a gray foam disc that made a light-tight seal. Each pot contained 6 L of continuously aerated nutrient solution (pH 5.8) of the following composition (in mM): 400 NO32, 200 Ca21, 200 K1, 154 SO422, 54 Mg21, 20 Cl2, 2.0 Fe-EDTA, 0.24 Mn21, 0.10 Zn21, 0.02 Cu21, 2.4 H3BO3, and 0.3 Mo41. The basal nutrient solution was supplemented with P (supplied as KH2PO4) at 1 mmol P L21 week21. To allow recovery after the transfer from soil to nutrient solution, the plants were grown under shade cloth for 4 weeks, after which they received full daylight in the glasshouse at minimum/maximum temperature of 20°C/32°C, while root temperature was maintained between 18°C and 20°C by standing the pots in a root-cooling tank. The nutrient solutions were replaced weekly.

557

Shane et al.

Development of Cluster Roots

Analysis of Carboxylates

Cluster-root development was followed by marking very young cluster roots with a basic fuchsin solution and then sampling cluster roots of different stage of development. Swollen regions of harsh hakea lateral roots were identified as future locations for cluster roots. Cluster roots were considered to be 1 d old when the rootlets were just visible after emerging from the cortex of the lateral axis. These lateral roots were marked just proximal from the tip with a drop of basic fuchsin. The development of cluster roots was divided into six age classes (i.e. 0, 1–2, 4–5, 7–8, 12–13, and 20–21 d).

Carboxylates in cluster-root exudates and tissue extracts were separated on an Alltima C-18 column (250 mm long 3 4.6 mm i.d. with 5-mm diameter packing; Alltech, Deerfield, IL), and identified using Waters HPLC (600E pump, 717 auto injector, and 996 photodiode-array detector; Milford, MA). The mobile phase for tissue carboxylates was 25 mM KH2PO4 (pH 2.50) and for exudates was a mixture of 25 mM KH2PO4 (pH 2.50) and MeOH (i.e. 93%:7%; pH 2.50) at a flow rate of 1 mL min21. Detection was at 210 nm, but data from 195 to 400 nm were collected and used for spectrum matching and peak purity analysis according to Cawthray (2003). The sample injection volume was generally 100 mL but was reduced for higher carboxylate concentrations. The column was completely flushed (gradient elution using 60% [v/v] methanol) prior to the subsequent sample injection to eliminate the transfer of highly nonpolar compounds. Data acquisition and processing was with Millennium software (Waters). Retention times of organic acid standards, including tartaric, formic, malic, iso-citric, malonic, lactic, acetic, maleic, citric, succinic, fumaric, cis-aconitic, and trans-aconitic acids, were used to identify carboxylates in root tissue extracts and root exudates.

Gas Exchange Measurements Over the course of several weeks, we measured the O2-uptake and CO2release rates of noncluster roots and cluster roots of six age classes. Measurements were made in summer, from mid-December 2002 to midJanuary 2003. The plants were transferred to fresh nutrient solution additionally buffered with 1 mM MES (pH 6) and without P 1 d before the measurements. The excised, intact cluster roots of a specific age class were harvested and then placed in nutrient solution containing 1 mM MES (pH 6.0) without P for 1 h prior to transfer of between 0.8 and 2.7 g of root tissue to each of the gas-exchange cuvettes. Four replicate cuvettes were used with at least seven cluster roots per cuvette harvested from more than 10 individual plants. Oxygen consumption was measured polarographically, using Clark-type oxygen electrodes (Lambers et al., 1993) to follow O2 depletion from stirred, gas-free, temperature-controlled cuvettes (approximately 55-mL volume). The O2 electrodes were calibrated between nutrient solution saturated with air and nutrient solution depleted of O2 by addition of sodium dithionite. Before and after O2-depletion measurements, the cuvettes were aerated at a flow rate of approximately 4.5 L h21 with compressed air. O2 consumption was measured immediately after transfer to the cuvettes, the cuvettes were then aerated for 1 h, and the O2 consumption was measured again. The flow rate to each cuvette was measured with a Brooks model 5850E series rotameter (Brooks Instrument, Veenendaal, The Netherlands). The efflux gas was passed through a mini water trap, a column of dry silica gel, and then into an infra-red gas analyzer (ADC 225-Mk3; Analytical Development Company, Hoddesdon, UK) in order to measure the CO2 release. The infrared gas analyzer was used in absolute mode, and the CO2 concentration of the incoming compressed-air stream determined after each measurement. Immediately after the measurements, the cluster roots were blotted dry, weighed, quenched in liquid nitrogen, and stored at 280°C until analysis of protein concentration enzyme levels (porin, PEPC, and AOX), organic acid concentration, and tissue [P]. The modified nutrient solution from each cuvette was collected after the CO2 and O2 flux measurements, immediately filtered through 0.2-mm Supor syringe filters (Gelman, MI) and stored at 220°C until analysis for organic anions (carboxylates). The O2-consumption rate was calculated from linear rates of O2 depletion in the cuvettes, and the RQ was calculated.

Total Root Protein Extraction Frozen cluster roots were homogenized in a chilled mortar and pestle with liquid N2 and suspended in 4 volumes of sample buffer per milligram of tissue. The sample buffer contained 4% (v/v) SDS, 125 mM Tris-HCl (pH 6.8), 0.15% (v/v) glycerol, 20% (v/v) mecaptoethanol, and Complete protease inhibitor cocktail (Roche, Castle Hill, Australia). The samples were incubated for 5 min at approximately 90°C, cooled on ice, and then centrifuged at 20,000g for 10 min. Protein concentrations were estimated by the method of Peterson (1977).

Carboxylate Extraction from Root Tissues Organic anions (carboxylates) were extracted from the frozen cluster roots according to Keerthisinghe et al. (1998) by grinding tissues in a prechilled mortar and pestle with 1.0 mL of ice-cold 0.6 M perchloric acid, then centrifuged at 15,000g for 5 min, and 0.75 mL of the supernatant was neutralized with 50 mL of 5 M K2CO3. The solution was centrifuged again at 15,000g for 5 min, and the supernatant was used to determine the carboxylate concentrations.

558

Measurements of Tissue [P] Frozen ground samples were dried in an oven at 80°C for 1 week. Samples were digested in concentrated HNO3:HClO4 (3:1) at 175°C. Total P concentration in digests was determined using the Malachite green colorimetric method (Motomizu et al., 1983).

Mitochondrial Isolation Mitochondria were isolated from fresh, 4-, 7-, and 17-d-old cluster roots. Cluster roots were excised from root systems and kept on ice until extraction (,1 h). Mitochondria were isolated from cluster roots, using a self-forming Percoll gradient, modified after Day et al. (1985). Briefly, in a cold room (4°C) cluster roots were chopped into small pieces and then homogenized in grinding medium (0.3 M Suc, 25 mM tetrasodium pyrophosphate, 2 mM EDTA, 10 mM KH2PO4, and 1% [w/v] polyvinylpyrrolidone [PVP]-40) with a mortar and pestle. The homogenate was then passed through four layers of wet Miracloth (Calbiochem, La Jolla, CA) and centrifuged at 1,000g for 5 min. The pellets were discarded and the supernatant centrifuged for 20 min at 15,000g. The supernatant was discarded and the pellet resuspended in 1 mL of 1 3 washing solution (0.3 M Suc, 10 mM TES, and 0.1% [w/v] bovine serum albumin). This was pipetted onto Percoll gradients (0%–10% [w/v] PVP gradient) prepared by mixing heavy (40 mL of 2 3 wash medium, 22.4 mL of Percoll, and 17 mL of 20% PVP-40, final pH 7.5) and light (40 mL of 2 3 wash medium, 22.4 mL Percoll, and 17 mL of water, final pH 7.5) using a peristaltic pump. The gradient was spun for 40 min at 40,000g. The mitochondrial layer banded at the bottom of the tube. The mitochondria were aspirated off, diluted with 1 3 wash buffer, and washed by centrifugation at 30,000g for 10 min. The pellets were resuspended in 1 3 wash buffer centrifuged for 1 min at 1,500g; the supernatant contained purified mitochondria and was further concentrated by centrifugation as required.

Western Blotting and Immunodetection Samples (50 mg) of total root protein or mitochondrial protein solubilized in sample buffer were separated by SDS-PAGE (Laemmli, 1970) using 12% (w/v) polyacrylamide gels for AOX and porin and 7.5% gel for PEPC. After electrophoresis, the separated proteins were transferred to nitrocellulose membranes and blocked. Western-blot immunoanalysis was performed to detect PEPC, AOX, and porin protein abundances. Rabbit polyclonal antibodies raised against alfalfa nodule PEPC were obtained from C. Vance (University of Minnesota, St. Paul), and a mouse monoclonal antibody raised against Sauromatum guttatum AOX was obtained from T. Elthon (Elthon et al., 1989; University of Nebraska, Lincoln, NE). Soybean cotyledon mitochondria were used as a positive control for AOX (Kearns et al., 1992), and mouse monoclonal antibody raised against Zea mays porin was obtained from T. Elthon (University of Nebraska). Immunoreactive proteins were detected using chemiluminescence and visualized by a luminescence analyzer (LAS 1000; Fuji Photo Film, Elmsford, NY). Signal intensities were estimated using ImageGauge version 3.0 software (Fuji Photo Film, Tokyo).



ACKNOWLEDGMENTS We thank Martin de Vos, Sytze de Roock, and Christel Norman for their help with the protein extraction protocol for cluster roots of harsh hakea. We thank Margaret McCully (CSIRO, Canberra, Australia) for help with Figure 1B. Received November 2, 2003; returned for revision February 16, 2004; accepted February 22, 2004.

LITERATURE CITED Cawthray GR (2003) An improved reversed-phase liquid chromatographic method for the analysis of low-molecular mass organic acids in plant root exudates. J Chromatogr A 1011: 233–240 Chollet R, Vidal J, O’Leary MH (1996) Phosphoenolpyruvate carboxylase: a ubiquitous, highly regulated enzyme in plants. Annu Rev Plant Physiol Plant Mol Biol 47: 273–298 Day DA, Krab K, Lambers H, Moore AL, Siedow JN, Wagner AM, Wiskich JT (1996) The cyanide-resistant oxidase: to inhibit or not to inhibit, that is the question. Plant Physiol 110: 1–2 Day DA, Neuburger M, Douce R (1985) Biochemical characterisation of chlorophyll-free mitochondria from pea leaves. Aust J Plant Physiol 12: 219–228 Dell B, Kuo J, Thomson CJ (1980) Development of proteoid roots in Hakea obliqua R.Br. (Proteaceae) grown in water culture. Aust J Bot 28: 27–37 Dinkelaker B, Hengeler C, Marschner H (1995) Distribution and function of proteoid root clusters and other root clusters. Bot Acta 108: 183–200 Dinkelaker B, Hengeler C, Neumann G, Eltrop L, Marschner H (1997) Root exudates and mobilization of nutrients. In H Rennenberg, W Eschrich, H Ziegler, eds, Trees: Contributions to Modern Tree Physiology. Backhuys Publishers, Leiden, The Netherlands, pp 441–452 Elthon TE, Nickels RL, McIntosh L (1989) Monoclonal antibodies to the alternative oxidase of higher plant mitochondria. Plant Physiol 89: 1311–1317 Gardner W, Barber D, Parbery D (1983) The acquisition of phosphorus by Lupinus albus L. III. The probable mechanism by which phosphorus movement in the soil/root interface is enhanced. Plant Soil 70: 107–124 Gerke J (1992) Phosphate, aluminium and iron in the soil solution of three different soils in relation to varying concentrations of citric acid. Z Pflanzenernaehr Bodenk 155: 339–342 Grierson PF, Attiwill PM (1989) Chemical characteristics of the proteoid root mat of Banksia integrifolia L. f. Aust J Bot 37: 137–143 Guy RD, Berry JA, Fogel ML, Turpin DH, Weger HG (1992) Fractionation of the stable isotopes of oxygen during respiration by plants—the basis of a new technique to estimate partitioning to the alternative path. In H Lambers, LHW Van der Plas, eds, Plant Respiration. Molecular, Biochemical and Physiological Aspects. SPB Academic Publishing, The Hague, The Netherlands, pp 443–453 Hagstro¨m J, James WM, Skene KR (2001) A comparison of structure, development and function in cluster roots of Lupinus albus L. under phosphate and iron stress. Plant Soil 232: 81–90 Handreck KA (1997) Phosphorus requirements of Australian native plants. Aust J Soil Res 35: 241–289 Hoffland E, Findenegg GR, Nelemans JA (1989) Solubilization of rock phosphate by rape. II. Local root exudation of organic acids as a response to P-starvation. Plant Soil 113: 161–165 Johnson JF, Allan DL, Vance CP (1994) Phosphorus stress-induced proteoid roots show altered metabolism in Lupinus albus. Plant Physiol 104: 657–665 Johnson JF, Allan DL, Vance CP, Weiblen G (1996a) Root carbon dioxide fixation by phosphorus-deficient Lupinus albus. Contribution to organic acid exudation by proteoid roots. Plant Physiol 112: 19–30 Johnson JF, Vance CP, Allan DL (1996b) Phosphorus deficiency in Lupinus albus. Altered lateral root development and enhanced expression of phosphoenolpyruvate carboxylase. Plant Physiol 112: 31–41 Jones DL (1998) Organic acids in the rhizosphere—a critical review. Plant Soil 205: 25–44 Kania A, Langlade N, Martinoia E, Neumann G (2003) Phosphorus deficiency-induced modifications in citrate catabolism and in cytosolic pH as related to citrate exudation in cluster roots of white lupin. Plant Soil 248: 117–127 Kearns A, Whelan J, Young S, Elthon TE, Day DA (1992) Tissue-specific


expression of the alternative oxidase in soybean and siratro. Plant Physiol 99: 712–717 Keerthisinghe G, Hocking P, Ryan PR, Delhaize E (1998) Proteoid roots of lupin (Lupinus albus L.): effect of phosphorus supply on formation and spatial variation in citrate efflux and enzyme activity. Plant Cell Environ 21: 467–478 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680–685 Lambers H (1997) Oxidation of mitochondrial NADH and the synthesis of ATP. In DT Dennis, DH Turpin, DD Lefebrve, DB Layzell, eds, Plant Metabolism. Longman Singapore Publishers, Singapore, pp 200–219 Lambers H, Atkin OA, Millenaar FF (2002b) Respiratory patterns in roots. In Y Waisel, A Eshel, U Kafkafi, eds, Plant Roots the Hidden Half, Ed 3. Marcel Dekker, New York, pp 521–552 Lambers H, Cramer MD, Shane MW, Wouterlood M, Poot P, Veneklaas EJ (2003) Structure and functioning of cluster roots and plant responses to phosphate deficiency. Plant Soil 248: ix–xix Lambers H, Juniper D, Cawthray GR, Veneklaas EJ, Martinez E (2002a) The pattern of carboxylate exudation in Banksia grandis (Proteaceae) is affected by the form of phosphate added to the soil. Plant Soil 238: 111–122 Lambers H, Van der Werf A, Bergkotte M (1993) Assessment of the capacity and activity of the alternative respiratory pathway in intact tissues. In GAF Hendry, JP Grime, eds, Methods in Comparative Plant Ecology—A Laboratory Manual. Chapman and Hall, London, pp 140–144 Lamont B (1972) The morphology and anatomy of proteoid roots in the genus Hakea. Aust J Bot 20: 155–174 Lamont BB (2003) Structure, ecology and physiology of root clustersa review. Plant Soil 248: 1–19 Massonneau A, Langlade N, Le´on S, Smutny J, Vogt E, Neumann G, Martinoia E (2001) Metabolic changes associated with cluster root development in white lupin (Lupinus albus L.): relationship between organic acid excretion, sucrose metabolism and energy status. Planta 231: 534–542 McCully ME (1975) The development of lateral roots. In JG Torrey, DT Clarkson, eds, The Development and Function of Roots, Third Cabot Symposium. Academic Press, New York, pp 105–124 McCully ME (1999) Roots in soil: unearthing the complexities of roots and their rhizospheres. Annu Rev Plant Physiol Plant Mol Biol 50: 695–718 Millenaar FF, Lambers H (2003) The alternative oxidase: in vivo regulation and function. Plant Biol 5: 2–15 Motomizu S, Wakimoto T, Toei K (1983) Spectrophotometric determination of phosphate in river waters with molybdate blue and malachite green. Analyst 108: 361–367 Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity hotspots for conservation priorities. Nature 403: 853–858 Neumann G, Martinoia E (2002) Cluster roots—an underground adaptation for survival in extreme environments. Trends Plant Sci 7: 162–167 Neumann G, Massonneau A, Martinoia E, Ro¨mheld V (1999) Physiological adaptations to phosphorus deficiency during proteoid root development in white lupin. Planta 208: 373–382 Neumann G, Ro¨mheld V (1999) Root excretion of carboxylic acids and protons in phosphorus-deficient plants. Plant Soil 211: 121–130 Pate JS, Bell TL (1999) Application of the ecosystem mimic concept to the species-rich Banksia woodlands of Western Australia. Agrofor Syst 45: 303–341 Peterson GL (1977) A simplification of the protein assay method of Lowry et al. which is more generally applicable. Anal Biochem 83: 346– 356 Purnell HM (1960) Studies of the family Proteaceae. 1. Anatomy and morphology of the roots of some Victorian species. Aust J Bot 8: 38–50 Roelofs RFR, Rengel Z, Cawthray GR, Dixon KW, Lambers H (2001) Exudation of carboxylates in Australian Proteaceae: chemical composition. Plant Cell Environ 24: 891–904 Ryan PR, Delhaize E, Jones DL (2001) Function and mechanism of organic anion exudation from plant roots. Annu Rev Plant Physiol Plant Mol Biol 52: 527–560 Rychter AM, Mikulska M (1990) The relationship between phosphate status and cyanide-resistant respiration in bean roots. Physiol Plant 79: 663–667 Sanchez R, Cejudo FJ (2003) Identification and expression analysis of

559

Shane et al.

a gene encoding a bacterial-type phosphoenolpyruvate carboxylase from Arabidopsis and rice. Plant Physiol 132: 949–957 Schuller KA, Turpin DH, Plaxton WC (1990) Metabolite regulation of partially purified soybean nodule phosphoenolpyruvate carboxylase. Plant Physiol 94: 1429–1435 Shane MW, De Vos M, De Roock S, Cawthray GR, Lambers H (2003) Effect of external phosphorus supply on internal phosphorus concentration and the initiation, growth and exudation of cluster roots in Hakea prostrata R.Br. Plant Soil 248: 209–219 Smith FW, Jackson WA, Vanden-Berg PJ (1990) Internal phosphorus flows during development of phosphorus stress in Stylosanthes hamata. Aust J Plant Physiol 17: 451–464 Smith SE, Read DJ (1997) Mycorrhizal Symbiosis. Academic Press, London

560

Theodorou ME, Plaxton WC (1993) Metabolic adaptations of plant respiration to nutritional phosphate deprivation. Plant Physiol 101: 339–344 Uhde-Stone C, Gilbert G, Johnson JMF, Litjens R, Zinn KE, Temple SJ, Vance CP, Allan DL (2003) Acclimation of white lupin to phosphorus deficiency involves enhanced expression of genes related to organic acid metabolism. Plant Soil 248: 99–116 Watt M, Evans JR (1999a) Linking development and determinacy with organic acid efflux from proteoid roots of white lupin grown with low phosphorus and ambient or elevated atmospheric CO2 concentration. Plant Physiol 120: 705–716 Watt M, Evans JR (1999b) Proteoid roots: physiology and development. Plant Physiol 121: 317–323
