Translocation in Legumes: Assimilates, Nutrients, and Signaling Molecules

Translocation or long distance transport in plants is achieved by a vascular network that connects and is an integral part of all organs. The vasculature comprises two distinctly different and separate cellular translocation pathways: xylem and phloem. The principal xylem pathway is the transpiration stream that moves nutrients and water taken up by roots to the shoot. This stream also bears products of root metabolism and solutes that reflect features of the internal and external root environment. Phloem provides the means for redistributing xylem-delivered solutes to weakly transpiring organs, but most significantly phloem distributes the carbon assimilated by photosynthesis (principally as Suc) to heterotrophic organs like roots, vegetative and reproductive apices, flowers, fruits, and developing seeds. Together these two translocation streams provide all the nutrients and assimilates, in appropriate forms and proportions, to enable growth and development in an ordered and regulated fashion. Because translocation connects distant components of the plant body, xylem and phloem have long been considered to fulfill a role in communicating between organs, through the movement of plant hormones and other signaling molecules. Such signals are envisaged to move with assimilates by mass flow. However, phloem also transmits pressure/concentration (turgor) information at rates greatly in excess of mass flow of solutes (Thompson and Holbrook, 2004) and long distance electrical signaling is also thought to be directionally propagated via vascular bundles (Brenner et al., 2006). These action potential or osmotic signals may prove to have a significant regulatory role in terms of phloem function but are outside the scope of this article. Most recently our understanding of the functional significance of phloem has been extended with the realization that it also provides a conduit for trafficking macromolecules (nucleic acids and proteins), some of which may regulate gene expression as a consequence of their translocation (Banerjee et al., 2006; Lough and Lucas, 2006; Jones-Rhoades et al., 2006). Similarly root-derived signals that are postulated to regulate shoot processes are believed to move in xylem (Beveridge, 2006; Kinkema et al., 2006) together with a suite of secreted proteins (Buhtz et al., 2004).

While the supporting evidence for these diverse roles of translocation has been gathered from many species, this article will highlight information that is specific to legumes where it is available, drawing particularly on data from the author's laboratory for members of the genus Lupinus.

SAPS/EXUDATES AND THEIR COLLECTION

Much of our knowledge of what is found in and translocated by phloem comes from analyses of sieve tube (ST) exudate and in xylem by analyses of sap, displaced from the vasculature by applying either increased or decreased pressure. A range of techniques has been developed and exploited to sample transport fluids and it is appropriate to consider the likely limitations that collection methods impose on interpretation of the compositional data they have generated.

The gold standard for phloem has been considered to be analyses of exudate collected from the detached stylets of sap sucking insects such as aphids. Aphid stylectomy has been applied mostly to woody and herbaceous dicotyledon species (Peel, 1975) but also to wheat (Triticum aestivum; Fisher et al., 1992) and using brown leaf hoppers to rice (Oryza sativa; Aoki et al., 2005). While stylet exudate may be regarded as ideal and least subject to artifact, stylets can be extremely variable in both rate and extent of exudation, and, even though the insect body is removed, salivary secretions are likely to persist as contaminants (Miles, 1999). The most common method has been to collect exudate from incisions in the bark of woody plants with compositional data for more than 500 species from 100 dicotyledonous families collated by Zimmermann and Ziegler (1975). Among herbaceous species there are very few that exude freely from severed vasculature, most occlude rapidly after damage preventing loss of phloem contents. The rapid blocking of ST has been interpreted as a phloem defense mechanism triggered by the release of Ca²⁺ causing constriction of sieve pores with extracellular callose (a 1,3-β-glucan polymer) and their plugging with coagulated structural phloem proteins (Will and van Bel, 2006). Interestingly, ST in legumes contains a unique protein crystalline body, the forisome, which disperses in response to damage or disturbance of phloem turgor to physically block the sieve plates, again in response to an increase in intracellular Ca²⁺ levels (Knoblauch et al., 2001). Known exceptions to rapid wound response are the hemophiliacs of the plant world. They include a number of cucurbits (cucumber [Cucumis sativus], pumpkin [Cucurbita pepo]), castor bean (Ricinus communis), Yucca flaccida, the axes of some palms, Brassica napus, and members of the genus Lupinus. Reasons for the sluggish response to vascular damage in these species are not known. Vascular contents have also been collected from nonspontaneous phloem bleeders by bathing wounds in a solution containing chelators (e.g. EDTA) to preclude rapid Ca²⁺-induced occlusion (King and Zeevart, 1974; Terce-Laforgue et al., 2004) or following rapid freezing and thawing that apparently also retards normal wound response (Pate et al., 1984). Pate (1976) outlined reasons why exudate collected from an incision may not accurately represent the solute composition moving in intact phloem and these considerations remain important in interpreting compositional data in relation to translocation.

Even though an aphid or leaf hopper stylet may enter and draw initially on a single ST, each microliter of exudate is equivalent to the lumenal volume of about 2,500 STs (Dixon, 1975). On this basis in lupin (Lupinus albus), where 50 μL of sap containing approximately 10% Suc is readily collected within minutes from an incision in the bundle of vascular elements at the stylar tip or along the sutures of a fruit, the exudate could derive from more than 10⁵ ST, extending some considerable distance beyond the wound. Initial piercing of a ST would cause a rapid fall in turgor pressure, decreasing local water potential and resulting in both longitudinal and lateral fluxes of water, diluting the exudate (Peel, 1975). It is reasonable to expect that the impact of a wound could potentially increase solute fluxes from closely associated companion cells (CCs) and phloem parenchyma. Pressure release on piercing may also result in exudate containing structural components of the STs that are not normally translocated. Because the exudate from a wound is typically high in sugar and other solutes, the contents of damaged cells from both vascular and adjacent tissues are likely to be contaminants, along with water and solutes from the surrounding apoplast. For this reason the first drop of exudate from a vascular incision is typically discarded, ostensibly to minimize contamination. While this precaution may reduce the impact from damaged cells, analyses of solutes from sequentially collected samples of exudate indicate that there is no progressive decline in these despite prolonged bleeding (Pate et al., 1974; Eschrich and Heyser, 1975).

The issue of purity of phloem exudate has been addressed by a number of groups. Giavalisco et al. (2006) compared the two-dimensional protein profile of B. napus phloem exudate with that of adjacent stem tissue and found that there was very little overlap in spot pattern between the two samples. Other groups have tested for proteins or mRNAs that would be expected in stem tissue but not phloem. Most commonly used is Rubisco or the transcript for its small subunit (sieve elements do not contain chloroplasts but specialized plastids; van Bel, 2003) and these assays suggest that there is limited contamination from outside cells (Ruiz-Medrano et al., 1999; Yoo et al., 2004; Haywood et al., 2005; Giavalisco et al., 2006).

Xylem contents are collected as tracheal sap by centrifugal pressure or vacuum displacement from stems and from root systems under natural root pressure, or by applying external pressure in a closed vessel, following detachment of the shoot. Xylem vessels or tracheids connect directly to the apoplast of surrounding tissues and it is reasonable to expect that water and solutes in the apoplast will be included in the collected sap. While generally positive or negative pressures applied are not great, nevertheless, materials from the apoplast or in the xylem but not normally translocated are likely to be mobilized. The contents of nonvascular cells damaged at the severed stem sections including phloem might be expected to contaminate xylem sap.

While these considerations will not have a significant impact on the nature and form of the major solutes in phloem and xylem they are likely to be important for minor constituents and especially on the macromolecules that have been found in exudates or saps and for which long distance signaling roles have been postulated, based on their assumed translocation. Confirming the constituents identified in phloem exudate move across a graft union becomes an important criterion to prove translocation.

ASSIMILATES AND NUTRIENTS

The principal assimilates translocated from sites of synthesis (sources) to sites of their utilization in growth and development (sinks) are those of carbon and of nitrogen. In legumes Suc is the predominant sugar (Zimmermann and Ziegler, 1975), and among nitrogenous solutes, amino acids, principally the amides Gln and Asn predominate, both in xylem and phloem (Atkins, 1991). In some species the ureides, allantoin and allantoic acid or citrulline may predominate, particularly in xylem, as translocated products of nitrogen assimilation in nodulated root systems (Atkins, 1991). In addition nodulated legumes also translocate unique solutes formed as result of the symbiosis with rhizobium and that can influence plant development. These include very high levels of cytokinin (CK; Upadhyaya et al., 1991), a bioactive product of riboflavin hydrolysis, lumichrome (Matiru and Dakora, 2005), and no doubt others yet to be discovered.

There is a massive body of experimental data that relates to assimilate translocation ranging from the pioneering studies of ST exudates collected from woody species through the ¹⁴C-labeling studies that began in the late 1940s and led to a continuing area of research seeking to describe mechanisms of translocation (for review, see Canny, 1973). More recently, expression analysis and localization of specific transporters for major assimilates have provided the basis from which molecular transport mechanisms involved in phloem loading and unloading and regulation of translocation have been developed (Lalonde et al., 2003; van Bel, 2003). Legumes have provided an ideal model in this area particularly for studies of translocation of photoassimilates and nitrogen compounds to seeds (Peoples et al., 1985; Patrick, 1997; Zhou et al., 2007).

In addition to the major solutes involved in translocation, analysis of exudates from lupin species has revealed an extraordinarily wide range of low M_r minor constituents (Atkins, 1999). These include inorganic nutrients, all protein amino acids as well as nonprotein amino compounds, organic acids, sugars and sugar alcohols, plant growth regulators, alkaloids, and no doubt many other solutes yet to be detected. As noted above, the site of synthesis for each of these and whether all are normally translocated or are in exudates as a consequence of collection is not clear.

From quantitative accounting of changes in carbon and nitrogen in component organs together with knowledge of carbon gains and losses through photosynthesis and respiration coupled with carbon and nitrogen translocation as solutes in xylem and phloem it proved possible to develop empirical models that quantified translocation and source/sink relations in the legume white lupin. This approach has been extended in lupin to the carbon and nitrogen nutrition of individual organs (Layzell et al., 1981; Fig. 1), to translocation of inorganic nutrients (Jeschke et al., 1985), specific amino acids (Pate et al., 1981), plant growth regulators (Emery et al., 2000), and quinolizidine alkaloids (Lee et al., 2007). These models have been summarized (Pate et al., 1998) and provide a basis for experimental perturbation of carbon and nitrogen partitioning, identification of specific solute/assimilate exchanges between xylem and phloem, and for development of an integrated view of translocation in relation to plant growth (Fig. 1). The quantitative description of source/sink relations in a grain legume also provides a framework upon which to view the emerging roles for translocation in regulating organ development.

Figure 1.

The models are based on the data for nodulated white lupin plants over a 1 week period (51–58 d after sowing) in midvegetative growth from Layzell et al. (1981). The net inputs of carbon and nitrogen were 1,061 and 34.8 mg, respectively, and are depicted in the models of partitioning as 1,000 units in each case. The flows of carbon and nitrogen were established from analysis of phloem exudates collected from petioles in each of four strata of leaves (A, B, C, and D; each stratum comprising four leaves), from the stem in each stratum, at the base of the stem just above the first lateral root, from the main stem apex, and from the axillary shoots forming in the top stratum of leaves. Xylem exudate due to root pressure was collected from the root system following removal of the shoot and from each stratum of the stem as tracheal sap by vacuum displacement. Measurements of respiration, net photosynthesis, and the incremental increases (or decreases) in carbon and nitrogen in dry matter were used to establish the fluxes of these commodities into sinks and out of sources. The models identified transfers of carbon or nitrogen from xylem to xylem, from xylem to phloem, and from phloem to xylem as processes essential for the establishment of source/sink relations over the period of study. These processes provide information as to the sites where molecular mechanisms of loading and unloading regulate the extent of nutrient partitioning and where similar processes are likely to participate in the distribution of low M_r signal compounds and perhaps to some extent the movement and partitioning of small macromolecules.

PROTEINS

Ever since early studies that measured protein content and recorded the myriad of enzyme activities assayed in ST exudate (Eschrich and Heyser, 1975) the origin and functional significance of these phloem proteins have provided subjects for considerable discourse and debate. Levels of protein may be as high as 35 to 60 μg/mL in exudate from cucurbit species and two-dimensional gel electrophoresis has resolved 350 to 400 polypeptides (Walz et al., 2004). Protein levels in exudate collected by stylectomy in rice are much lower (<1 μg/mL) but electrophoresis revealed more than 100 polypeptides (Hayashi et al., 2000). Giavalisco et al. (2006) recently identified 140 proteins in B. napus phloem. Mass spectrometric analysis of lupin phloem exudate also revealed a large number of proteins and peptides (Marentes and Grusak, 1998). The most recent estimates suggest more than 1,500 species in the phloem proteome of angiosperms (Lough and Lucas, 2006). Although newly differentiating ST are nucleate and engage in protein synthesis, the proteins in enucleate mature sieve elements are most likely products of gene expression in CC, and possibly phloem parenchyma, transferred through plasmodesmata. As noted above, the normal transfer of proteins through this route may be enhanced as a consequence of wounding and the wide range of enzyme activities assayed in exudates from a number of species (Eschrich and Heyser, 1975), including Robinia pseudoacacia, may be in part an artifact. Despite these considerations sequential sampling during prolonged exudation indicates that protein content and polypeptide composition remains relatively stable, suggesting that the initial impact of wounding is slight (Hayashi et al., 2000).

The localization of tagged antibodies at either light or electron microscopy levels of resolution have led to positive associations of proteins with structure and this has also been the case for phloem CCs (Terce-Laforgue et al., 2004). Molecular studies that have exploited expression of fluorescent proteins as reporters for the cloned promoter regions of genes in transiently or stably transformed plants (Stadtler et al., 2005) have also revealed close associations with the vasculature, inferring localization of the encoded protein. In many cases localization has been linked to phloem and the further inference that such proteins might be translocated.

Stadtler et al. (2005) generated a series of transgenic Arabidopsis (Arabidopsis thaliana) lines expressing a number of GFP-fusion proteins (36–67 kD) under a CC-specific AtSUC2 promoter. Results indicated that the size exclusion limit (SEL) for plasmodesmata between the CC and the ST was >67 kD and that free GFP and fusion proteins in the range 36 to 67 kD were loaded onto and translocated in phloem. However, unloading and postphloem mobility of these fusion proteins in root tips were restricted to a narrow cell zone adjacent to mature protophloem and only free GFP was effectively transported beyond this zone. The data indicated a SEL of approximately 27 to 36 kD and consequently restricted access for larger proteins into the root tip. How widespread these SEL values apply is not known but if they represent a reasonable estimate then many of the proteins identified in the phloem proteome, assessed on the basis of analysis of exudates, are unlikely to be transferred beyond the vasculature in sink tissues, even if they are translocated.

There is now little doubt that proteins are mobile in phloem and further that there may be destination-selective translocation. The elegant experiments of Aoki et al. (2005) in which two isolated pumpkin phloem proteins were labeled and injected into the vasculature of intact rice plants through severed leaf hopper stylets provide evidence for specificity in protein translocation. Interaction of one of the proteins, CmPP16-1, with other phloem proteins appeared to positively regulate translocation of this protein to the roots (Aoki et al., 2005). Nevertheless, considerable uncertainty exists about which proteins detected in exudates engage in long distance transport and whether their loading is a nonselective process resulting from their proximity to ST. The likely limitations and questions raised by Oparka and Santa Cruz (2000) in their review of macromolecules in phloem remain to be satisfactorily addressed.

The question about the functional significance of protein translocation is largely just speculation based on likely annotation of sequence data derived from mass spectrometry (MS) proteomic analysis. From cucurbit exudate a complete and apparently functional antioxidant defense system has been recovered (Walz et al., 2002) and a more comprehensive analysis based on partial sequencing of resolved polypeptides by tandem MS has identified 45 proteins with the majority linked to stress responses and defense mechanisms. It seems likely that these proteins provide the vasculature with its own protection systems perhaps to mitigate the impacts of wounding and disease, but it is also possible that responses to a range of stress conditions are transmitted as a consequence of the enzymology that is set in train in ST. For exudates collected from insect stylets it should be noted that the proteins secreted to ST in the watery saliva of aphids include similar defense enzymes that may also be translocated long distances (Will and van Bel, 2006). A number of proteins have been identified in phloem exudates that bind RNA and there is growing evidence for translocation of ribonucleoprotein complexes (Gomez et al., 2005). With the likely roles for translocated RNA in regulating gene expression (see below) this group of binding proteins may provide the means to stabilize and protect signals during long distance transport. Proteins such as CmHsc70 have been identified that may act as chaperones to direct other proteins or RNA into the ST and/or to refold proteins unfolded as they pass through plasmodesmata into the ST (Aoki et al., 2002).

Buhtz et al. (2004) reported that the protein composition found in xylem sap was conserved across two species of Brassica and two cucurbit species. The saps contained 50 to 100 μg protein/mL and following electrophoresis yielded sequence data that identified peroxidases, chitinases, and a number of proteases. While phloem contamination was found to be negligible the likely site of synthesis of the proteins or secretion to xylem is unknown. Significantly, N-terminal secretion signal sequences could be found for all the database entries that were detected, consistent with passage of these proteins to the root apoplast. A similar suite of proteins, but also including prominent Gly-rich proteins, were found in xylem sap from cucumber (Satoh, 2006), and in maize (Zea mays) a much wider range, including putative defense-related proteins (Alvarez et al., 2006b). Proteomic analysis of xylem exudate from nodulated soybean (Glycine max) has identified 25 proteins, many of which were similar to those found in cucurbit and brassica xylem, with 20 having N-terminal signal sequences (M.A. Djordjevic, personal communication). While all these studies speculate that xylem proteins are involved in pathogen or stress resistance their long distance movement in the transpiration stream and their functions, as a consequence of translocation, are yet to be determined.

LONG DISTANCE SIGNALING IN LEGUMES

Legumes provide an opportunity to study two phenomena involving long distance signaling: autoregulation of nodule development and cluster root formation. When nodule development is initiated there is an exchange of signals between the roots and shoot to regulate the number of nodules that develop. This signaling also appears to be related to one of the mechanisms (but not all) by which legumes regulate the symbiosis in response to soil nitrogen conditions. Split root experiments, where one part of a root system was inoculated with rhizobia and subsequently inhibited nodulation on the other part, suggested that the signal was systemic. A number of mutants, super- or hypernodulators (HAR1 in Lotus japonicus, GmNARK in soybean, SYM29 in pea [Pisum sativum], and SUNN in Medicago truncatula) have been identified where this control is lost (Carroll et al., 1985; for review, see Oka-Kira and Kawaguchi, 2006). Grafting experiments with some of these mutants showed that the shoot genotype controls nodule development. A signal derived in the roots is translocated and perceived in the shoot generating a second signal that inhibits further development of nodule primordia. Although the genes above have now been cloned and shown to encode a Leu-rich repeat receptor-like kinase similar to CLAVATA1 (CLV1) from Arabidopsis (Krusell et al., 2002; Nishimura et al., 2002; Searle et al., 2003; Schnabel et al., 2005; see below), the nature of the signal molecules involved have not been defined (for review, see Kinkema et al., 2006; Beveridge et al., 2007).

Proteoid or cluster roots are densely clustered secondary roots with determinant growth that are a feature of many species in the Proteaceae as a response to low soil phosphorus levels (Lamont, 2003). Cluster roots enhance the plants' ability to access phosphorus by increasing the root area for uptake, synthesis, and exudation of acid phosphatase and organic acids, which solubilize phosphorus in the soil that is bound to aluminum, iron, or calcium. White lupin shares this root response to low phosphorus and the species has been used as a model for studying plant adaptations to phosphorus deficiency (Vance, 2001) and the molecular events involved (Uhde-Stone et al., 2003a, 2003b, 2005; Liu et al., 2005). Development of cluster roots in lupin is controlled by a systemic signal from the shoot (Liu et al., 2005). Liu et al. (2005) have recently shown sugar and phloem transport to be essential components in regulating phosphorus deficiency-induced gene expression in white lupin roots, indicating that the ability to sample downward-moving phloem streams could be exploited in identifying the translocated signal(s).

SIGNALING MOLECULES

The vasculature of the plant provides an ideal pathway for rapidly transmitting information from sources to sinks together with assimilates and nutrients. Despite some uncertainty in relating solute composition of exudates to actual translocation there seems little doubt that the channels of xylem and phloem do indeed carry molecular signals. It is likely that some assimilates engage in regulatory functions but similar roles for the myriad of other small molecules, the plant growth regulators, nucleic acids, proteins, and peptides in phloem are probable. The transpiration stream also provides a means for the rapid long distance movement of signals that originate in roots and in legumes also in nodules, and which are postulated to regulate events in the shoot.

Assimilates

Plant physiologists refer to translocation of major solutes in phloem as representing the source/sink relations of a plant, a feature that determines harvest index (ratio of grain yield to plant biomass at final harvest) and has been the single most important trait exploited in domestication of a species and its improvement by plant breeders. Despite the obvious significance of this relationship, the means by which it is regulated remain elusive, but intuitively some sort of signaling mechanism has been invoked. Trewavas (2006) points out that in whole plants “both sugars and N-solutes are obvious candidates to provide information loops enabling some balance between root and shoot development to be maintained”; but adds, in the context of the need for feedback loops, a complication to transmission of information between tissues is “uncertainty in the homeostatic control of the major conducting tissues” (p. 2426). As well as its role in transport, Suc acts as an important signaling molecule in plants (Rolland et al., 2006). Evidence from sugar beet (Beta vulgaris) suggests that phloem loading may be regulated by Suc that stimulates transcription of Suc symporters in CC (Chiou and Bush, 1998; Vaughn et al., 2002), allowing the plant to regulate assimilate partitioning at this level. The model developed suggests the existence of a Suc sensor but this has yet to be identified (Ransom-Hodgkins et al., 2003).

Sugars unloaded from the phloem in seeds may act as regulators of seed development with the high ratio of hexose to Suc produced by cell wall-bound invertases acting to enhance cell division in cotyledons early in development in Vicia faba (for review, see Weber et al., 2005). Later the predominance of Suc in the seed coat apoplast results in development of transfer cells on the cotyledonary epidermis (Wardini et al., 2007) and changes the developmental program in the cotyledon to one of cell expansion after stimulation of transcription of Suc symporters in the transfer cells (Weber et al., 2005; Wardini et al., 2007).

γ-Aminobutyric acid is a powerful neurotransmission signal in mammalian systems and a signaling role has been suggested in plants (Bouché et al., 2004); specifically, as a shoot-to-root signal regulating nitrate uptake and the plant's carbon/nitrogen balance (Beuve et al., 2004). A similar signaling function has been suggested for translocated Gln (Nazoa et al., 2003). Both amino acids occur in phloem and xylem exduates of lupin (Atkins et al., 1983), cowpea (Vigna unguiculata; Pate et al., 1984), and other species. While there is evidence for Glu-gated Ca²⁺-channel receptors in plants that could also bind γ-aminobutyric acid (Bouché et al., 2004) clear evidence linking translocation of these compounds with functional regulatory mechanisms is yet to be gathered.

Plant Growth Regulators

Despite the many analyses of exudates that document the presence of all known plant growth regulators (except the brassinosteroids; Symons and Reid, 2004), their precursors, or conjugates in xylem and/or phloem, a clear functional role(s) in relation to or as a consequence of their translocation has in most cases yet to be established. The data are thus largely descriptive and, while valuable in establishing that changes in development or in environmental factors may be reflected in relative concentrations of these hormones in translocation streams, the connection of such changes to developmental processes is far from clear. Identification of proteins in phloem exudates has also highlighted a likely role for the vasculature in synthesizing some phytohormones in situ (Walz et al., 2004), including jasmonic acid (JA) and ethylene. There is also recent evidence for localization of expression of one of the isopentenyl pyrophosphate transferase alleles (IPT3) in phloem consistent with localized CK synthesis (Miyawaki et al., 2004).

Perhaps the most studied example is acropetal movement of abscisic acid (ABA) through xylem to its site of action in the shoot, linking soil water deficit to stomatal function and the water relations of leaves (Jackson, 1993). However, empirical modeling of translocatory flows of ABA in white lupin (Wolf et al., 1990), using the same sap analysis and accounting methodology that was used to model carbon/nitrogen and water economies in the species (see Fig. 1), found that much of the ABA exiting roots in xylem was in fact recirculating from the shoot following phloem-to-xylem transfer in the root with the majority synthesized in the shoot. When the plant was stressed by applying a 40-mm saline solution to roots the balance of synthesis shifted from the shoot to the root with net biosynthesis increasing more than 20-fold. A recent study (Foo et al., 2007) using increased branching (rms) mutants of pea and equivalent mutants in Arabidopsis has suggested that transfer of phloem CK to xylem in roots and elsewhere between xylem and phloem should be examined to gain a more complete understanding of CK translocation.

The significance of long distance transport of CK in plant development is far from clear (Dodd and Beveridge, 2006; Foo et al., 2007). Both xylem and phloem exudates contain CK but unambiguous information on the forms and their mobility is at best fragmentary. Despite a general belief that the prominent translocated forms are the nucleosides (Sakakibara, 2006), detailed analysis of exudates from white lupin indicates that the composition of CK forms in translocation channels, especially phloem, is far from constant (Emery et al., 2000). Significant shifts from cis- to trans-isomers in both xylem and phloem appear to be related to changes in reproductive development and during a period of 77 d, over which exudates were collected, 12 individual CK species were detected in varying amounts. Cis-CK are also prominent in xylem sap from chickpea (Cicer arietinum) and in developing chickpea seeds are the major accumulating form (Emery et al., 1998). The biological activity of these isomers in legume species is yet to be determined but it seems clear from Arabidopsis that the balance between trans- and cis-isomers reflects the activity of two separate pathways, one plastid localized and the other cytosolic (Sakakibara, 2006).

White lupin has also provided evidence for the presence of gibberellins (GAs) in phloem and xylem and their likely translocation (Hoad, 1995). Gas chromatography/MS analysis confirmed the presence of 16 GAs representing four major pathways and there was some evidence that phloem collected from different sites on the plant (stems versus fruits) exhibited a different spectrum of GAs (Thompson et al., 1988; Hoad et al., 1993). However, there has been conflicting data as to which GA species are translocated, and in pea GA₁ is immobile while its precursor, GA₂₀, is mobile (Reid et al., 1983). Specific roles for each of the GA species are yet to be resolved so that any functional significance for those that are translocated either in xylem or phloem cannot be assigned.

Bioactive Peptides

The systemic and cell-to-cell signaling role of low M_r peptides is a significant feature of metabolic and posttranscriptional gene regulation in animals. A similar role for bioactive peptides is also a feature of plant development and while a number have been implicated in regulation of important processes through intercellular communication (e.g. the 14 amino acid CLV peptides; Fiers et al., 2004), there has been no convincing evidence for their long distance translocation in phloem and action as a truly systemic signal. However, Stacey et al. (2002) has drawn attention to the multitude of putative peptide transporters in the Arabidopsis genome and bioinformatics tools have identified in silico novel genes encoding small secretory peptides (Matsubayashi and Sakagami, 2006) with no doubt many more to be discovered (Lease and Walker, 2006). The plasmalemma-localized high affinity peptide transporter AtPTR1 is expressed predominantly in vascular tissues and in leaves in phloem or phloem parenchyma (Dietrich et al., 2004), placing the protein ideally to load peptides into the ST for long distance transport. A study of proteins in lupin phloem exudates by Marentes and Grusak (1998) using matrix-assisted laser-desorption ionization time-of-flight MS was extended by Hoffmann-Benning et al. (2002) who were able to find more than 100 small proteins and peptides. These authors used this methodology to compare exudate collected into a chelator solution from flowering and nonflowering Perilla ocymoides leaves in the hope of finding the elusive flowering hormone florigen. They obtained sequence data for 16 peptides in the mass range 1 to 9 kD, four of which were specific to or increased in phloem from plants induced to flower. Recent data (Huang et al., 2005) has implicated mRNA of the FLOWERING LOCUS T (FT) gene in Arabidopsis in providing the translocated florigenic signal to the shoot apical meristem. However, there is still speculation that the FT protein may be involved (Corbesier and Coupland, 2006) and this is supported by the presence of FT in phloem exudate of B. napus (Giavalisco et al., 2006). Whether the peptides found by Hoffmann-Benning et al. (2002) are also involved remains to be shown; one possibility is that they associate with the FT mRNA to protect the transcript as a ribonucleoprotein complex in transit.

Systemin (18 amino acid peptide) was thought to be mobile in phloem of tomato (Solanum lycopersicum) causing an induced response to herbivory in leaves and other tissues distant from the initial site of wounding. Phloem mobility was supported by ¹⁴C-labeling studies and more or less specific expression of its precursor protein, prosystemin, in phloem parenchyma (Jacinto et al., 1997). Immunochemical analysis has also located prosystemin in the cytosol of vascular parenchyma cells (Narvaez-Vasquez and Ryan, 2004). However, recent studies indicate that systemin is released to the vasculature where it binds to a specific membrane receptor inducing a number of events, including release of JA precursors and activating localized JA synthesis. Thus, JA appears to be the long distance signal that induces systemic wound response initially attributed to the peptide (Stratmann, 2003; Matsubayashi and Sakagami, 2006).

The gene responsible for the supernodulation phenotype in a number of legumes has been identified as a Leu-rich repeat receptor-like kinase similar to CLV1 (Krusell et al., 2002; Nishimura et al., 2002; Searle et al., 2003; Schnabel et al., 2005). CLV1 regulates shoot and floral meristem proliferation in Arabidopsis as part of a protein complex that is activated by a small peptide CLV3 (for review, see Beveridge et al., 2007). The similarity of the legume genes to CLV1 suggests that the long distance signaling molecule translocated from root to shoot may be a peptide similar to CLV3 that is carried in xylem.

Nucleic Acids

There is little doubt that the recent discovery of a number of RNA species in phloem exudates and the possibility that some of these may function as specific regulators of gene expression has opened up a new and exciting role for long distance translocation. The many studies that over the years have sought to explain developmental responses in plants as a consequence of some translocated cue or signal may now have the possibility for rational explanation. More specifically, the idea of controls that modify gene expression at a level outside local molecular networks (Sachs, 2005) being long distance signals may have credibility. This sort of information will provide a means for the complex mathematical modeling essential for development of a new systems biology approach to whole plant growth and development.

mRNAs

There are a number of examples of mRNAs that are translocated in phloem (Ruiz-Medrano et al., 1999; Xoconostle-Cazares et al., 1999) and some where translocation has been demonstrated to result in changes in plant development (Kim et al., 2001; Haywood et al., 2005; Banerjee et al., 2006). Long distance movement of the ST BEL5 transcript was recently implicated in control of tuber induction in potato (Solanum tuberosum) in response to short day length. BEL5 is part of a conserved transcription factor family in plants (TALE consisting of two main groups BEL1 and KNOX). RNA for six of the seven BEL1 genes in potato were detected in phloem, suggesting that translocation of RNA for other members of this family may regulate other developmental processes (Banerjee et al., 2006). Phloem of legumes has yet to be analyzed to determine which mRNAs are present.

Small RNAs

Small noncoding RNAs play an important role in gene regulation in eukaryotes. Two classes of small RNA, small interfering RNAs (siRNAs) and microRNAs (miRNAs) have been detected in phloem exudate (Yoo et al., 2004; P.M.C. Smith and C.A. Atkins, unpublished data). These small RNAs interact with a multicomponent RNA-induced silencing complex and direct the cleavage of their target RNAs. siRNAs, which were initially identified through their role in cosuppression of transgenes, form a part of the plants defense mechanism against foreign viruses. They act to silence viral RNAs and so restrict spread of the virus. Once production is initiated a systemic signal is generated to allow a response in distant tissues. Silencing of transgenes in RNAi is also mediated by siRNAs (Bonnet et al., 2006). In addition to these siRNAs that target foreign nucleic acids, there are two recently identified types of siRNA that target endogenous genes. Trans-acting siRNAs (ta-siRNAs) are generated after initial cleavage of a noncoding, single stranded precursor transcript by a miRNA. After cleavage, RDR6 initiates double stranded RNA formation and phased siRNAs are produced by DCL4 with the phase determined by the site of miRNA cleavage (Peragine et al., 2004; Vazquez et al., 2004; Allen et al., 2005). Natural antisense transcript derived siRNAs (nat-siRNAs) are produced as a result of cleavage of two convergent and partially overlapping transcripts. The first nat-siRNAs discovered mediate responses to salt stress (Borsani et al., 2005) and pathogen attack (Katiyar-Agarwal et al., 2006).

Much of the evidence for translocation of siRNAs to induce systemic gene silencing is based on grafting experiments showing systemic spread of gene silencing from silenced stocks to wild-type scions and the presence of siRNAs directed against the transgene in tissue from the scion (Palauqui et al., 1997; Voinnet et al., 1998) or transient expression in localized areas of the shoot to generate a systemic silencing signal (Voinnet and Baulcombe, 1997; Klahre et al., 2002). Direct analysis of phloem exudate is not possible in the plants (particularly Nicotiana benthamiama) that provide useful models in this area. However, Yoo et al. (2004) showed movement of 23 nucleotide (nt) siRNAs between a spontaneously silencing squash (Cucurbita pepo) stock and a wild-type cucumber scion, supporting the idea that siRNAs may form the mobile signal for systemic gene silencing. However, sense and antisense transcripts were detected in the phloem of silenced plants and their translocation through the graft was not ruled out. These results and others in which viral suppressors were used to eliminate siRNAs suggest that other RNAs (single or double stranded) could also act as systemic signaling molecules in long distance transport of gene silencing (for review, see Voinnet, 2005). Isolation of a small RNA binding protein (PSRP1) in pumpkin phloem that bound to 25 nt single (and double) stranded RNA further supports the idea of translocation of small RNAs in phloem. PSRP1 was able to mediate cell-to-cell movement of 25 nt single stranded (but not double stranded) RNA molecules but PSRP1-mediated movement of these RNAs in phloem was not demonstrated (Yoo et al., 2004). An interesting observation from this experiment is that although they are small these RNAs could not move between cells when microinjected without PSRP1. A small RNA binding protein of size similar to PSRP1 was also detected in phloem from white lupin, along with a population of small RNAs (Yoo et al., 2004). If siRNAs are able to move systemically to mediate viral defense and silence transgenes then it is possible that trans-acting siRNAs and nat-siRNAs have similar properties, but this remains to be tested.

miRNAs are important regulators of plant development and responses to environmental signals. The majority of their target genes are transcription factors and they play an important role in clearing regulatory transcripts from daughter cell lineages to allow a change in developmental state (Rhoades et al., 2002). Examples are the control of leaf polarity where miR166 acts to cleave transcripts of PHV and PHB to allow abaxialization of the leaf (Juarez et al., 2004; Kidner and Martienssen, 2004) and miR172 that acts during floral development to inhibit expression of APETALA2 (AP2) in whorls three and four of the flower to allow development of stamens and carpels (miRNA inhibits translation of AP2 as well as cleaving the transcript [Aukerman and Sakai, 2003; Chen, 2004]). Other miRNAs are expressed in response to environmental conditions and cleave targets to allow the plant to adapt. An example is miR399 that is expressed in conditions of phosphorus starvation and cleaves the transcript of a ubiquitin-conjugating enzyme to regulate phosphorus homeostasis (Fujii et al., 2005; Aung et al., 2006; Bari et al., 2006; Chiou et al., 2006).

Yoo et al. (2004) identified three miRNAs in cucurbit phloem and showed that a population of small RNA molecules existed in phloem of white lupin, castor bean, and yucca (Yucca spp.). We have identified 11 different miRNAs in white lupin phloem (M. Jordan, C.A. Atkins, and P.M.C. Smith, unpublished data). Analysis of expression of miR166 in maize suggested that the miRNA is expressed below the leaf primordium and is transported to its site of action in cells on the abaxial side of the leaf where it blocks expression of PHV and PHB to allow differentiation (Juarez et al., 2004; Kidner and Martienssen, 2004). In situ hybridization localized the miRNA in phloem. Expression of viral movement proteins that disrupt RNA trafficking cause the formation of adaxialized leaves, lending support for a functional role in normal leaf development (Juarez et al., 2004). We have since cloned miR166 from phloem exudate of white lupin (M. Jordan, C.A. Atkins, and P.M.C. Smith, unpublished data). Bari et al. (2006) also noted a systemic signal that results in expression of miR399 in response to low phosphorus conditions and we also cloned this miRNA from white lupin phloem exudate (M. Jordan, C.A. Atkins, and P.M.C. Smith, unpublished data). However, there is no direct evidence for translocation of these miRNAs and a number of studies suggest that their expression is cell autonomous or that they have only limited cell-to-cell movement. In grafting experiments where a synthetic miRNA was expressed from a 35S promoter in tomato and tobacco (Nicotiana tabacum) there was no translocation of the miRNA across the graft union to a wild-type scion (Alvarez et al., 2006a) and experiments where miRNA action was detected with sensor GFP constructs containing a miRNA binding site and compared with expression from the native MIR gene promoter suggest they act in the cells in which they are expressed (Parizotto et al., 2004; Voinnet, 2005). So while it is tempting to specilate that miRNAs, given their similarity in structure to siRNAs, could move systemically to regulate gene expression in response to environmental signals, there is no clear evidence for this role.

Lupins as a Model to Study Translocated Molecules?

Despite the fact that the genome sequence for lupin is not available and is not likely to become available in the near future the genus offers a valuable model for studies that link long distance translocation with plant growth and development. It could be particularly useful in identifying long distance signal molecules that relate to symbiosis and cluster root development. In addition, a number of lupin species are crops that produce economic yields of high protein grain and in this respect the information on translocation in this genus could serve as models for other pulse (grain legume) crops. Their unique feature that permits ready access to the transport fluids of xylem and phloem at a number of sites on both vegetative plants and plants during reproductive development has potential to provide new information about the molecular events that are regulated as consequence of translocation of solutes and signal molecules, both large and small. Most importantly, phloem exudates can be readily collected at both sources and sinks (Pate et al., 1979), providing unique tools to study molecular events that signal regulation of assimilate partitioning. Members of the genus can be grafted (Pigeaire et al., 1997; Lee et al., 2007) and a number of species can be routinely and stably transformed (Molvig et al., 1997; Pigeaire et al., 1997; Clements et al., 2005). Stable transformation frequencies for white lupin, which bleeds phloem most consistently, are low but the root system can be transformed using Agrobacterium rhizogenes (Uhde-Stone et al., 2005). To further develop this species as a model for translocation more molecular information would be useful, particularly EST libraries, which apart from providing information about the transcriptome, would facilitate further proteomic studies, and development of TILLING systems to identify mutants in key genes (McCallum et al., 2000). The growing genomic data that is available through sequencing projects with other legume species, including soybean, will also provide a timely resource to interpret the information revealed by studies of translocation in lupin.