logistics of water and salt transport through the plant

Plant, Cell and Environment

(2003)

26,

87–101

© 2003 Blackwell Publishing Ltd

87

Blackwell Science, LtdOxford, UKPCEPlant, Cell and Environment0016-8025Blackwell Science Ltd 200226Original Article

A. H. de Boer & V. VolkovStructural and functional features of xylem

Correspondence: Dr A. H. de Boer. Fax:

+

31 20 4447155; e-mail: [email protected]

Logistics of water and salt transport through the plant: structure and functioning of the xylem

A. H. DE BOER

1

& V. VOLKOV

2,3

1

Vrije Universiteit, Faculty of Earth and Life Sciences, Department of Developmental Genetics, Section Mol. Plant Physiol. & Biophysics, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands,

2

K.A. Timiriazev Institute of Plant Physiology, Root Physiology Laboratory, Botanischeskaya 35, 127276 Moscow, Russia and

3

Horticultural Production Chains Group, Wageningen University, Marijkeweg 22, 6709 PG Wageningen, The Netherlands

ABSTRACT

The xylem is a long-distance transport system that is uniqueto higher plants. It evolved into a very sophisticated plumb-ing system ensuring controlled loading/unloading of ionsand water and their effective translocation to the requiredsinks. The focus of this overview will be the intrinsic inter-relations between structural and functional features of thexylem. Taken together the xylem is designed to preventcavitation (entry of air bubbles), induced by negative pres-sures under transpiration and to repair the cavitated ves-sels. Half-bordered pits between xylem parenchyma cellsand xylem vessels are on the one hand the gates to thevessels but on the other hand a serious ‘bottle-neck’ fortransport. Hence it becomes evident that special transportsystems exist at the interface between the cells and vessels,which allow intensive fluxes of ions and water to and out ofthe xylem. The molecular identification and biophysical/biochemical characterization of these transporters has juststarted. Paradigms for the sophisticated mechanism ofcontrolled xylem transport under changing environmentalconditions are SKOR, a Shaker-like channel involved in K

+

-loading and SOS1, a Na

+

/H

+

antiporter with a proposeddual function in Na

+

transport. In view of the importanceof plant water relations it is not surprising to find that waterchannels dominate the gate of access to xylem. Future stud-ies will focus on the mechanism(s) that regulate water chan-nels and ion transporters and on their physiological role in,for example, the repair of embolism. Clearly, progress inthis specific field of research will greatly benefit from anintegration of molecular and biophysical techniques aimedto understand ‘whole-plant’ behaviour under the ever-changing environmental conditions in the daily life of allplants.

Key-words

: cavitation repair; ion channels; ion transport;ion transporters; potassium; sodium; water transport;xylem.

Abbreviations

: CNGC, cyclic nucleotide gated channels;PD, electric potential difference; SKOR, stelar potassiumoutward rectifying channel; TRP,

trans

-root potential.

INTRODUCTION

Like most multicellular organisms plants need structuresthat enable the mass flow of solvents (mainly water) andsolutes such as ions, sugars, amino acids, hormones, etc. Inanalogy to the blood circulation system in animals, plantsuse tube-like structures called the xylem and phloem.Whereas the study of the blood circulation system is tech-nically relatively easy, xylem and phloem transport path-ways are notably difficult to study due to their location deepinside the plant tissue, the presence of reinforced cell wallsand the large positive or negative pressures inside the tubes.Despite, or maybe challenged by these hurdles, a large bodyof research has been devoted to unravelling the mecha-nisms that sustain the functioning of the plant long-distancetransport systems. For recent insights in phloem transportwe refer to reviews by Patrick

et al

. (2001) and Oparka &Cruz (2000).

Here we try to link the structure–function relationship ofthe dead xylem conduits and living adjacent parenchymacells with a focus on the mechanisms that facilitate andcontrol the flow of ions and water into (loading) and out ofthe xylem (unloading). Although it was hypothesized in theearly days of xylem research that, for example, xylem load-ing in the root was a passive process due to the lack ofoxygen in the core of root tissue (Crafts & Broyer 1938), itis clear now that we are dealing with sophisticated regula-tion mechanisms aimed to sustain growth and developmentin highly dynamic environments. Since structure and func-tion are intimately interwoven, the structure of the xylemtransport system will be discussed first. Thereafter, currentknowledge about a number of essential biophysical param-eters namely the electrical, pH and pressure gradientsbetween the vessels and surrounding symplast/apoplast willbe summarized. The molecular mechanisms of potassium,sodium, and water transport will be discussed in more detailat the end.

88

A. H. de Boer & V. Volkov


,


,

26

, 87–101

HOW STRUCTURAL FEATURES OF THE XYLEM MEET THE DEMANDS OF SHORT- AND LONG-DISTANCE TRANSPORT

The main function of the xylem is to transport large quan-tities of water and solutes. This requires a continuous sys-tem of interconnected tubes with a relatively low resistanceto the flow of water. As the driving force for mass flow is apressure gradient (positive with respect to the atmosphericpressure as during root pressure, or negative during tran-spiration) the walls of the vessel or tracheary elements mustbe reinforced. On the one hand leakage of water and sol-utes out of the tubes must be minimized (so the walls mustbe as leak-proof as possible), but on the other hand special-ized structures in the walls must enable rapid exchange ofwater and solutes between the vessel interior and surround-ing cells. Another feature that is vital for survival of thexylem is to prevent and repair the interruption of the con-tinuous water columns by air bubbles (embolism).

Low resistance allows high water flows

The principal conducting cells of the xylem are the trache-ary elements of which there are two types: the tracheidsand the vessel elements. Both tracheary elements developnormally from proliferating cells from the procambium andvascular cambium. When fully differentiated the end wallsof the vessel elements are perforated or removed (what isleft is called the perforation plate) and the cells undergoprogrammed cell death (Roberts & McCann 2000). Thus,files of open interconnected cells are formed. Tracheids alsolack protoplasts at maturity, but in contrast to vessels theend walls are not open but contain pits, which are areaslacking the secondary wall. The so-called pit membrane, amodified thin primary cell wall consisting of a dense net-work of hydrophilic cellulose polymers, is water permeable(Holbrook & Zwieniecki 1999). Water flowing through ves-sels sometimes has to pass pit membranes as well, as vesselsare not continuous over the entire length of the plant andare interconnected through pits or perforation plates(Fig. 1). The principal water-conducting cells inangiosperms are tracheids, whereas vessels are found bothin angiosperms and gymnosperms.

Important determinants of the hydraulic resistance arethe diameter of the vessels and the pit membranes or per-foration plates. In xylem vessels, water is assumed to flowaccording to Poiseuille’s law with a hydraulic conductivityproportional to the fourth power of vessel’s radius. Thus, asmall change in the radius will give a large change in resis-tance. The radius of a vessel may also be related to a func-tion in solute loading. For example, in the first 10–20 cm ofroots of monocots only vessels having a small diameter arefunctional in transport. The reason may be that narrowvessels with a high surface-to-volume ratio are more suit-able for controlling the composition of the xylem sap thanwide vessels. The large late metaxylem vessels in the apical10–20 cm of a maize root do not conduct water (St. Aubin,Canny & McCully 1986), because they mature just in the

upper zones of the root. In these upper zones, the widevessels collect the sap and function as simple conductiveelements, whereas narrow xylem vessels function not onlyas conductive elements but also as xylem sap-controllingelements.

The prevailing view that tracheary elements have eithera constant or a zero (after embolism) hydraulic conduc-tance has recently been challenged (Van Ieperen, Van Mee-teren & van Gelder 2000; Zwieniecki

et al

. 2001) althoughthe observation was already made by Zimmermann (1978)in a nice example of serendipity. Observing a decreasingflow rate of distilled water through a piece of stem (a prob-lem already described by Huber & Merz 1958), Zimmer-man hypothesized that tiny gas bubbles obstructed pitpores. Therefore he replaced the distilled water with tapwater (assuming that tap water contained more bubblesthan distilled water), but contrary to expectation the flowrate instantaneously increased. He discovered that thiseffect was due to the presence of ions in the tap water and

Figure 1.

Bordered pit pairs interconnect xylem vessels and half-bordered pits connect the vessels to xylem parenchyma cells. (a) Cryo-planed transverse section through the secondary xylem of a chrysanthemum stem. (b) Bordered pit pair in high magnification (length of pit membrane is 3·7

µ

m). V, vessel; bp, bordered pit; hp, half-bordered pit; pc, parenchyma cell; lf, libriform fibre; sp, simple pit; pm, pit membrane (Micrograph kindly provided by Dr J. Nijsse, Wageningen University).

pm

a

b

Structural and functional features of xylem

89


,


,

26

, 87–101

concluded that ‘the phenomenon might be based uponswelling or shrinking of the vessel-to-vessel pit mem-branes’. Van Ieperen

et al

. (2000) concluded that the pres-ence of small amounts of cations (even the effect of 10

µ

M

KCl was significant) and not the osmotic strength influ-enced the conductance. However, their conclusion thatswelling and shrinking of vessel-to-vessel pit membranesplayed no role in this phenomenon was contradicted byZwieniecki

et al

. (2001). The latter authors reached thesame conclusion as formulated by Zimmermann (1978)namely that the observed change in hydraulic resistancewhen changing the ionic composition of the xylem sap isdue to the pit membranes connecting one vessel to theother. In these ‘membranes’ water flows through micro-channels of the modified primary cell wall, made up bycellulose microfibrils, hemicelluloses and pectins. It wassuggested that the size of the microchannels increases whenthe pectin matrix (

D

-galacturonic acid being a main com-ponent) shrinks in response to the binding of ions to thenegatively charged matrix. If charge compensation is themechanism that determines the flow resistance, then it issurprising that no significant effect of pH within the naturalpH range of xylem sap (5·8–8·0) was observed (Zwieniecki

et al

. 2001), although pH affected swelling and shrinking ofisolated cell walls (Meychik & Yermakov 2001). Relevantof course is the question whether these ion-mediatedchanges in hydraulic resistance do have a physiological role.The K

+

concentrations having the strongest effect (0–20 m

M

; Zwieniecki

et al

. 2001), are within the range mea-sured in transpiring intact plants (Herdel

et al

. 2001). How-ever, it remains to be seen whether these K

+

variations ina constant background of, for example, divalent cations stillhave an effect on the hydraulic properties.

Reinforcement and waterproofing to withstand negative pressures and retain water

Tracheary elements not only have to withstand the com-pressive forces from surrounding cells, but especially thelarge negative pressures that develop in actively transpiringplants (see below). Therefore, secondary walls composed ofcellulose and xylan are deposited on top of the primary cellwall (Fukuda 1996). Early-formed protoxylem elements ofthe primary xylem have ringlike (annular) or spiral (helical)thickenings that still allow a certain degree of cell elonga-tion. In the late-formed metaxylem elements and elementsof the secondary xylem the secondary cell walls cover theentire primary walls, except at pit membranes shared withadjoining tracheids/vessels or xylem parenchyma cells andcontact cells. Subsequently, the secondary cell walls arelignified: a network is formed that strengthens and water-proofs the wall by the polymerization of aromatic com-pounds called monolignols (McCann 1997). Notsurprisingly, inhibitors of lignification result in a plant phe-notype with collapsed vessels (Smart & Amrhein 1985).Primary cell wall polymers are also important for the sub-sequent lignification and strengthening of a vessel, andmutants in cellulose deposition in the cell wall with col-

lapsed xylem vessels were isolated as well (Turner & Som-erville 1997). In order to prevent ‘leakage’ of water andsolutes from the vessels it seems logical that the lignifiedsecondary walls are relatively impermeable to thesecompounds.

One of the few studies on xylem cell wall permeability isthe one by Wisniewski, Ashwort & Schaffer (1987) whostudied deep supercooling in woody plants and used lan-thanum nitrate as an apoplastic tracer. Lanthanum pene-trates capillaries as small as 2 nm, so a cell wall withoutlanthanum staining is relatively impermeable. They foundthat the primary cell wall of cortical tissues was very per-meable to the lanthanum solution, but primary (exceptparts associated with the pit membranes) and secondarywalls of vessel elements were impermeable. Interestingly,the pit membranes composed of the middle lamella, pri-mary walls of the adjoining cells and the protective layer(in xylem parenchyma cells and contact cells) were highlypermeable. These results are in line with the conclusion thatafter occlusion of pit membranes, vessel elements canassume a closed-cell structure and may have a permeabilityapproaching zero (Siau 1984).

These permeability properties of the xylem walls are rel-evant to the question whether water and solutes can directlyenter the vessels from the stelar apoplast. If not, then theyhave to be taken up by stelar cells into the symplast andloaded into the vessel lumen after passing the membraneof a xylem parenchyma cell underlying a pit membrane. Forexample, the stelar potassium outward rectifying channel(SKOR; activated at depolarizing voltages) is thought tohave a function in loading K

+

into the xylem (Wegner &Raschke 1994; Wegner & de Boer 1997a; Gaymard

et al

.1998). Besides expression in xylem parenchyma cells, clearGUS-staining of root pericycle cells was observed in plantsexpressing the GUS reporter gene under the control of theSKOR promoter region (Gaymard

et al

. 1998). Note alsothat pericycle cells of barley roots stained strongly for theplasma membrane ATPase (Samuels, Fernando & Glass1992). The pericycle cells do not directly border a xylemvessel. Therefore, the question is how K

+

ions released fromthe pericycle cells into the stelar apoplast do reach thexylem lumen if the xylem vessel walls are as impermeableto ions and water as shown by Wisniewski

et al

. (1987).

Pit structures facilitate water-/solute (un)loading and embolism repair

Pits constitute the gates between vessels and from vesselsto adjacent xylem parenchyma cells (Fig. 1). The total areaof pit fields, their shape and pattern of lignification are oflarge variation between species. For example, maize pitfields constitute about 15% of the vessel area, with pitmembranes having a diameter of about 2

µ

m with a hydrau-lic conductivity about 10 times over that of the vessel wall(Peterson & Steudle 1993). Typically, between vessels bor-dered pits are found that are characterized by overarchingwalls that form a bowl-shaped chamber (Fig. 1b). It haslong been known that these pits are vital structural ele-

90



,


,

26

, 87–101

ments in the plant’s defence against cavitation (rupture ofwater columns) and subsequent embolism (filling of vesselsand tracheids with air) (Zimmermann 1983; Tyree & Sperry1989). This is because the modified primary cell wall thatconstitutes the pit membrane is permeable to water, butdoes not allow the passage of even the tiniest air bubbles.For this reason tracheids, wherein all elements are linkedthrough pits are thought to be safer for plants than vesselsalthough the latter elements are more efficient conductorsof water. In some monocots, such as rye, in the root–shootjunction tracheids separate the vessels in the root fromthose in the shoot to protect the vessels of the roots fromembolism originating in the shoot (Aloni & Griffith 1992).

However, pits are not only important structural elementsthat prevent spreading of air bubbles through the water-conducting elements of plants. They are instrumental inrepairing the ‘broken water pipes’ as well (Tyree

et al

. 1999;Holbrook & Zwieniecki 1999; Zwieniecki & Holbrook2000). Repair means removal of the air bubble; how canthis be achieved? As we have seen above, an air bubblecannot be simply ‘flushed’ out of a conduit. The existingparadigm suggests that an air bubble is removed throughgas dissolution in the surrounding water (see Table 1). Animportant structural demand is that in order to contain thispressurization the perimeter of the embolized vessels mustbe sealed. As we have seen above, the secondary wall meetsthis requirement since the thickness and lignification makeit waterproof. However, what about the pits in which thewater permeability of the pit membrane is high (seebelow)? Let us first consider the hypothesis that a plant hasto generate positive (above atmospheric) pressures in thevessels. The best understood mechanism that generates pos-itive pressures is root pressure; namely pressure that isgenerated in the root during the night or in the shoot inearly spring when transpiration is low through osmoticwater flow into the vessels. There is good experimentalevidence that this mechanism operates in a range of

angiosperms and gymnosperms under certain conditions(Tyree

et al

. 1986; Sperry

et al

. 1987; Tyree & Yang 1992;Cochard, Ewers & Tyree 1994). However, not all plantsseem to develop root pressure (20 out of 109 species studiedhad no root pressure, Fisher

et al

. 1997) and it must be keptin mind that root pressure is dissipated by gravity at a rateof 10 kPa per m height. Therefore, tall trees are unlikely todevelop positive pressures in the shoot as a consequence ofpressure built-up in the root. So, how do plants dissolveembolism without root pressure? This question is evenmore relevant in the light of recent studies reporting embo-lism repair under conditions when the threshold xylempressure (

P

x

) (see Table 1) is much less than

−

2

τ

/

r

(where

τ

is the surface tension of water in the vessel and

r

is theradius of curvature of the air–water interface) (Salleo

et al

.1996; Tyree

et al

. 1999). The only possible explanation forthis repair mechanism seems to be that locally (i.e. in theembolized vessel or tracheid element) the xylem pressureis higher than

−

2

τ

/

r

. There are a number of intriguing prob-lems with this model:

1 The data suggest that this mechanism of embolismremoval is concurrent with transpiration (Salleo

et al

.1996; Canny 1997b; Tyree

et al

. 1999). If so, how canxylem pressure increase in the embolized element whilethe adjacent vessels are still under tension?

2 What is the mechanism that generates locally a xylempressure

>

−

2

τ

/

r

?

The first problem was distinguished by Tyree

et al

. (1999)and Holbrook & Zwieniecki (1999) came up with a hypoth-esis as to why in two adjacent vessels (one embolized, theother filled and under tension) connected by pits, a pressuredifference can be maintained until the embolized vesselsare water filled again (Holbrook & Zwieniecki 1999; Zwie-niecki & Holbrook 2000). Without going into much detail,the idea is that the structure of the bordered pits ensuresthat the embolized element remains hydraulically isolatedfrom the neighbouring vessels under tension until themoment that all air has been dissolved. Zwieniecki & Hol-brook (2000) showed that the contact angle of water andthe angle of the bordered pit chamber walls are such thata convex gas–water interface forms in the pit chamber. Thesurface tension of this convex curvature opposes the hydro-static pressure within the filling lumen as long as the pres-sure in the filling element does not exceed the force due tothe surface tension. Clearly, this model deserves furtherexperimental testing.

The second problem seems rather easy to solve. Namely,xylem conduits are generally in contact with numerous liv-ing cells, the xylem parenchyma cells and contact cells(Zimmermann 1983; Lachaud & Maurousset 1996). Thesecells could be triggered by embolism to load osmoticallyactive ions such as K

+

and Cl

–

across the pit membrane intothe vessel lumen. Such a mechanism was proposed for thefirst time by Grace (1993) who studied embolism repair intracheids from

Pinus sylvestris

. Water would then be drivenby osmosis into the vessel lumen. This idea is consistentwith the observation that convex drops protrude through

Table 1.

Threshold xylem pressure

Removal of air from an embolized vessel must occur by dissolution of the gas in the surrounding water. The solubility of a gas in water is proportional to the partial pressure of the gas species adjacent to the water (Henry’s law). Water in plants is saturated with air at atmospheric pressure and therefore the pressure of the gas in the bubble must be above atmospheric for the gas to dissolve in the surrounding water.What determines the pressure of the gas in the bubble (

P

g

)?(1) The pressure inside the xylem element,

P

x

; and(2) the pressure inside the gas bubble that develops due to the

surface tension at the air–water interface, 2

τ

/

r

, where

r

is the radius of the curvature of the air–water interface and

τ

is the surface tension of the water in the vessel.

Thus the pressure of the gas in the bubble equals:

P

g

=

P

x

+

2

τ

/

r

.

Therefore when the xylem pressure is greater than

−

2

τ

/

r

(the threshold xylem pressure), the gas in an embolism will exceed atmospheric pressure and start to dissolve. For a narrow vessel of 10

µ

m diameter this threshold pressure is around

−

30 kPa (

−

0·3 bar). See Tyree

et al

. (1999) for further details.

Structural and functional features of xylem

91


,


,

26

, 87–101

the pits into the lumen of embolized vessels (see Fig. 2)from sunflower petioles (Canny 1997b), roots of maize(McCully, Huang & Ling 1998) and

Laurel nobilis

branches(Tyree

et al

. 1999). However, thus far there is one majorproblem: attempts to measure high concentrations of saltsin the water that enters an embolized vessel have failed(Canny 1997b; McCully

et al

. 1998; McCully 1999; Tyree

et al

. 1999). All of the methods used to estimate the xylemsap ion concentration, rely on the technique of X-raymicroanalysis except one (Tyree

et al

. 1999). In the latterstudy ‘native’ xylem sap from well-watered and cavitatedtwigs was collected by perfusion. Treatments wherein cavi-tation recovery was high also showed a strong increase (upto four-fold) in K

+

concentration. This positive correlationdoes suggest that the increase in ion concentration hassomething to do with the refilling of embolized elements. Ifso, then the increase in concentration of solutes in the refill-ing vessels may be much higher since the sap recoveredfrom the perfused stems will be diluted by water from thefull vessels. So, the discussion on the mechanism of refilling

embolized vessels seems to hinge on the question whetherdroplets entering an embolized element through the pits docontain high enough levels of solutes. If X-ray microanaly-sis gives a reliable estimate of the xylem sap ion concentra-tion, then we need a new paradigm of how positive xylempressures build-up in embolized vessels and how root pres-sure in general is generated (Enns, Canny & McCully2000).

Clearly, there is a need for reliable methods for samplingand analysis of ‘native’ xylem sap. In the literature manymethods have been described, as summarized by Schurr(1998), none of which is without its own specific problems.The methods range from the collection of xylem sap underroot pressure to such an unusual method as using xylemfeeding

Philaenus spumarius

(spittlebugs) to get the sapsample (Watson, Pritchard & Malone 2001). As a generalrule, it is important to define the question to be addressedand then define the requirements a sampling method mustmeet. Ideally, the method of analysis does not disturb theplant in general and the conditions of transpiration in par-ticular. Non-invasive techniques such as nuclear magneticimaging (NMRi), nuclear magnetic resonance (NMR) spec-trometry and magnetic resonance imaging (MRI) do inprinciple meet these requirements (Scheenen

et al

. 2000;Peuke

et al

. 2001; Holbrook

et al

. 2001). Spatial and tempo-ral resolution was initially an important disadvantage ofthese methods, but recent methods have decreased mea-surement time drastically (Rokitta

et al

. 1999; Scheenen

et al

. 2000; Peuke

et al

. 2001).If the refilling of embolized vessels is not driven osmoti-

cally by salts loaded into the vessel lumen by the surround-ing xylem parenchyma cells, then maybe reverse osmosis isa feasible mechanism (McCully 1999; Tyree

et al

. 1999) asoriginally proposed by Canny (1997a, b). Reverse osmosiscould occur if cells surrounding the xylem decrease theirosmotic potential (through ion uptake or conversion ofstarch to sugars), built up a higher turgor pressure, pressur-ize xylem parenchyma cells that then squeeze water (with-out ions) into the embolized vessel element. An auxin-induced loading of solutes in the phloem may be part ofthis mechanism (Salleo

et al

. 1996; Tyree

et al

. 1999). How-ever, Tyree

et al

. (1999) have pointed out the problems thatcome with this mechanism. One important problem tosolve is how to prevent water from flowing in all directionsfrom the ‘squeezed’ parenchyma cells. In other words, howto generate a direction in the flow of water (a problemsolved in the ‘osmosis’ hypothesis by specific loading of ionsinto the embolized element across the pit membrane). Onepossibility is that water channels in the patch of membraneunderlying the pit membrane are activated, thus creating apath of low resistance to waterflow into the vessel element(see section on water channels).

Vessel-associated cells are equipped for a task in (un)loading

Living vessel-associated cells, the xylem parenchyma cellsand contact cells, are the cells that have direct access to the

Figure 2.

Part of an embolized late metaxylem vessel in maize root. Note the pits between the vessel and bordering xylem par-chyma cells. Arrows indicate droplets of liquid exuding through pits into the lumen (

×

1580) [Reproduced from McCully

et al

. (1998), with permission of Blackwell Publishing, Oxford, UK].

92



,


,

26

, 87–101

vessel lumen. Therefore, they play a key role in the loadingand unloading of the xylem. In line with the energy demandof these (un)loading processes, structural features of cellsindicate high metabolic activities that could fuel massivetransmembrane transport. These features are a dense cyto-plasm, well-developed endoplasmic reticulum, numerousribosomes, mitochondria and peroxisomes (Läuchli

et al

.1974). The specialized transport function of these cells isevidenced by the development of cell wall invaginations inthe area underlying the pit membrane bordering the vessels(Kramer

et al

. 1977; Yeo

et al

. 1977; Kuo

et al

. 1980; Mueller& Beckman 1984). This folding pattern increases the mem-brane surface area and is characteristic for transfer cells(Pate & Gunning 1972; Gunning 1977) and the develop-ment is stimulated by salt stress (Kramer

et al

. 1977; Yeo

et al

. 1977). Relatively little is known thus far about thedistribution of ion transporters in the membrane of vessel-associated cells. Sauter, Iten & Zimmermann (1973)observed high ATP hydrolysing activity at the large pits ofcontact cells facing the vessels and suggested a role for thisenzymatic activity in sugar loading of the xylem. This activ-ity was only observed during starch mobilization in the rayscells in spring or when starch mobilization was artificallyinduced by defoliation. Xylem parenchyma cells of barleyand oat roots were strongly labelled by antibodies againstthe plasma membrane ATPase, but the resolution wasinsufficient to discern differences in subcellular distributionof the protein (i.e. patches on the vessel-facing side)(Parets-Soler, Pardo & Serrano 1990; Samuels

et al

. 1992).These results are in line with the localization of ATPaseactivity in barley xylem parenchyma cells as determinedwith a cytochemical lead precipitation method (Winter-Sluiter, Läuchli & Kramer 1977).

The external cytoplasmic layer of xylem parenchymacells has specific features at the point of contact pits (pitsbetween xylem parenchyma cells and xylem vessels). There,the network of actin filaments contains fewer corticalmicrotubules in comparison with other parts of the cyto-plasmic layer where an intensive cortical microtubulecytoskeleton is present (Chaffey, Barlow & Barnett 2000;Chaffey & Barlow 2001). The function of these structuralchanges of the cortical cell layer may be two-fold: to pre-vent the deposition of a thick secondary wall, a processwhich depends on cortical microtubules, and the regulationof ion and water channel activity. Actin involvement in ionchannel regulation was shown for guard cells (Hwang

et al

.1997).

THE PROTON MOTIVE FORCE AT THE XYLEM/SYMPLAST INTERFACE: DRIVING FORCE FOR XYLEM (UN)LOADING

The energy necessary to drive transport of ions, aminoacids, sugars, etc. across the membrane of vessel-associatedcells into or out of the vessel lumen, is derived from theelectrical gradient across this membrane, the pH gradientor both. Therefore, it is important to know the value ofthese two parameters and the dynamic regulation thereof.

Xylem electrical potential

The initial hypothesis postulated by Crafts & Broyer (1938)that primary active transport (e.g. an ATP driven H+-ATPase) at the symplast/vessel boundary was precludeddue to lack of oxygen, proved not to be correct (Okamoto,Katou & Ichino 1979; De Boer et al. 1983; Clarkson, Will-iams & Hanson 1984). In roots of Plantago and hypocotylsof Vigna unguiculata the electric potential difference (PD)between vessel and parenchyma cells contains an electro-genic component, which in both roots and hypocotyls canamount to 70 mV. The primary H+-ATPase is thought to beresponsible for this electrogenic component because it dis-appears in the presence of inhibitors of the oxidative phos-phorylation (anoxia and azide) and increases with auxinand fusicoccin, the activator of the plasma membrane H+-ATPase (De Boer et al. 1985; De Boer 1997). It must bekept in mind that the electrical potential of the cortical andxylem (stelar) apoplast are not necessarily the same. Aninsulating layer formed by the ring of endodermis cells withtheir Casparian strips, can maintain a PD between the twoapoplasts. Because the cortical, endodermal and stelar cellsare connected through plasmodesmata (forming the sym-plast) the apoplast PD is in fact the difference of two mem-brane potentials in series: (PDsymplast/cortical apoplast)minus (PDsymplast/xylem apoplast) (Okamoto et al. 1979;De Boer et al. 1983; De Boer et al. 1985). This PD is alsocalled trans-root potential (TRP), but the same principleholds for stems and other tissues.

All measurements mentioned above were carried out onexcised plants or perfused root and stem segments. Thisapproach can be criticized because in excised roots theturgor and intracellular osmotic pressure gradients collapse(Rygol et al. 1993). Therefore, measurements on intactplants are necessary. Dunlop (1982) succeeded for the firsttime to measure directly with a microelectrode in the ves-sels of intact Trifolium repens plants. The electrical poten-tial of the vessels was 90 mV negative with respect to thecortical apoplast. Because xylem parenchyma cells had amembrane potential of −168 mV it can be calculated thatthe potential difference across the symplast/xylem bound-ary was around 80 mV (xylem positive). These values areclose to the values measured in excised Plantago rootswhere the electrogenic component of the symplast/xylemPD was zero (De Boer et al. 1983). Recently, a new methodwas introduced, using a modified pressure probe, enablingthe simultaneous measurement of the trans-root potentialand the xylem pressure in intact transpiring plants (Wegner& Zimmermann 1998). One of the conclusions of thesemeasurements was that the roots of intact transpiring plantshave electrical properties similar to those of excised roots(Wegner et al. 1999). These measurements also demon-strated that the regulation of the trans-root electrical poten-tial is a dynamic process. Namely, a switch to a higher lightintensity induced strong oscillations in TRP that precededoscillations in the negative xylem pressure. Interestingly,the depolarization of TRP correlated with the most nega-tive xylem pressure. Since the membrane potential of the

Structural and functional features of xylem 93

© 2003 Blackwell Publishing Ltd, Plant, Cell and Environment, 26, 87–101

cortical cells showed no oscillations, the oscillations in TRPmust be the result of a change in ion transport activityacross the symplast/xylem boundary or as suggested byWegner et al. (1999) by a change in streaming potential dueto variations in the transpiration flow. However, thus far itis not clear whether streaming potentials contribute signif-icantly to the potential differences within a xylem vessel orbetween the xylem and cortical apoplast.

Although we now know that xylem transport processesare not simply passive, the early suggestion that the supplyof oxygen to the xylem might be a limiting factor (Crafts &Broyer 1938; De Boer & Prins 1984) deserves renewedattention in the light of recent analysis of the oxygen supplyto sapwood using a novel optical method for oxygen detec-tion (Gansert, Burgdorf & Lösch 2001).

Xylem sap pH

The pH of the xylem sap is relevant for (un)loading pro-cesses for two reasons: (i) as driving force for antiport andsymport processes (see below) and (ii) as regulator of iontransporters or signalling molecules. There are no measure-ments of the xylem sap pH in intact plants as far as we areaware. Measurements on xylem sap collected from rootexudates or by means of centrifugation agree on a slightlyacidic pH in the range 5·5–6·5 (Schurr & Schulze 1995;Wilkinson & Davies 1997; López-Millán et al. 2000). Some-how the xylem has a high capacity to maintain its pH to aset-value (Clarkson et al. 1984; Senden et al. 1992). How-ever, in response to changes in environmental conditionsthe xylem sap pH does fluctuate. Under drought stress thexylem sap pH from Commelina communis increased from6·1 to 6·7 (Wilkinson & Davies 1997) and iron stress condi-tions reduced the pH from sugar beet xylem sap from 6·0to 5·7 (López-Millán et al. 2000). What is the mechanismthat on the one hand buffers and on the other hand changesthe xylem pH? Buffering is certainly a consequence of thecation exchange capacity of the xylem cell walls due to thepresence of negatively charged groups (Ferguson & Bollard1976; Senden et al. 1992) and the presence of organic acidssuch as succinate and malate (Schurr & Schulze 1995). Butthe physiological changes may be due to regulation of theactivity of the H+-ATPase and/or H+-exchange systems inthe plasma membrane of vessel associated cells. If forexample the H+-ATPase activator fusicoccin is perfusedthrough xylem vessels the pH of the emerging sapdecreases within minutes after application, reaching valuesas low as 4·2 (Fig. 3). This effect is explained by the uptakeof fusicoccin by xylem parenchyma cells around the vessels.Fusicoccin then activates the plasma membrane H+-ATPase (De Boer 1997), which results in a strong efflux ofprotons into the xylem sap. Further evidence for theinvolvement of metabolically active cells is the observationthat the pH rapidly increases when roots are subjected toanoxia (De Boer & Prins 1985; Lacan & Durand 1996).Several roles for changes in xylem sap pH have been pro-posed. Thus, the increase in pH initiated by drought stress,is thought to increase the efficiency of abscisic acid in stim-

ulating the closure of stomata (Wilkinson & Davies 1997).Moreover, K+ loading into the root xylem may be a pH-sensitive process through modulation of the activity ofSKOR, the outward rectifying K+ channel involved in K+

release into the xylem (Lacombe et al. 2000 and see below).

MOLECULAR MECHANISM OF SOLUTE AND WATER (UN)LOADING

Potassium

Xylem loading of K+ is regulated separately from K+ uptakefrom the external solution (Cram & Pitman 1972; Engels &Marschner 1992). Studies at the cellular and molecular levelin recent years have shown two principal types of K+ chan-nels in the root (Gaymard et al. 1998; Spalding et al. 1999):AKT1 for K+ uptake from the medium and SKOR for K+

loading into the xylem. The in planta properties of SKOR[formerly KORC in barley (Wegner & De Boer 1997a)]were first characterized through patch-clamp studies onxylem parenchyma protoplasts isolated from barley andmaize roots (Wegner & Raschke 1994; Roberts & Tester

Figure 3. Effect of fusicoccin, activator of the plasma membrane H+-ATPase, on the ion concentration and ion fluxes into/out of the xylem as measured by means of a xylem perfusion technique. (a) Perfusion of taproots from Plantago maritima and (b) of nodal roots from barley. In both roots, fusicoccin (10 µM) strongly acidi-fies the xylem sap, reduces the flow of K+ into the xylem from the surrounding cells and stimulates the release of Na+ into the xylem stream. [(a), De Boer unpubl. res.; (b), Volkov and De Boer unpubl. res.].

94 A. H. de Boer & V. Volkov


1995; Wegner & De Boer 1997a; Roberts 1998). Theseoutward rectifying channels activate in a time-dependentmanner at membrane potentials that shift with and remainslightly positive of the reversal potential of K+ (EK). Theyare selective for K+ but do facilitate the passage of Ca++ aswell. They cannot be responsible for Na+ loading of thexylem as they are virtually impermeable to Na+. Underphysiological conditions a depolarization of the membranepotential to values positive of EK will result in a K+ effluxand an influx of Ca++. Therefore, a non-functional SKORmay result in a reduction of the amount of K+ and anincrease in the amount of Ca++ in the shoot (see below).

In addition to its sensitivity to membrane voltage andapoplastic K+ concentration, the activity of SKOR is regu-lated by cytosolic Ca++, the stress hormone abscisic acid(ABA) and pH. A rise in the cytosolic Ca++ concentrationfrom 150 nM to 5 µM reduced the frequency at which SKORwas recorded in barley xylem parenchyma cells (whole-cellconfiguration) from 90 to 10% (Wegner & De Boer 1997a).As this effect was absent when inside-out patches wereused, the involvement of a cytosolic intermediate was sug-gested. ABA has a dual effect on SKOR activity: a short-term effect measurable within minutes after addition ofABA and a long-term effect (Roberts 1998; Gaymard et al.1998). The short-term effect was observed when ABA wasadded to protoplasts during a patch-clamp experiment:within minutes it reduced the outward current in 55% ofthe cells; the inward current was unaffected. Drought, orovernight pre-treatment of intact plants with ABA alsoresulted in a strong reduction of the outward K+ current. Inline with earlier experiments (Cram & Pitman 1972; Pitman1977) ABA pre-treatment had no effect on in- and out-ward-currents in cortical root cells (Roberts 1998). Itremains to be shown if Ca++ plays a role as second messen-ger in the xylem parenchyma ABA signalling pathway, asit does in guard cells (Allen et al. 2000). An indicationmaybe that both Ca++ and ABA affect SKOR in a similarfashion.

The gene encoding the channel responsible for K+ load-ing in the xylem, called SKOR, belongs to the Shaker super-family (Gaymard et al. 1998). SKOR expressed in Xenopusoocytes, exhibited most if not all of the characteristics ofxylem parenchyma K+ channel characterized in barley andmaize: sigmoidal activation, shift in activation potentialwith EK, sensitivity to blockers, Ca++ permeability andincreased magnitude of the outward current with increasingexternal K+ concentration. Northern blot analysis showedthat SKOR is expressed exclusively in roots and there theexpression is limited to pericycle cells and cells surroundingthe xylem vessels; hence the name stelar K+ outward recti-fier. Treatment of intact plants with the stress hormoneabscisic acid resulted in a rapid decline in SKOR mRNAabundance (Gaymard et al. 1998), thus shedding light onABA inhibition of xylem K+ loading in intact barley plants(Cram & Pitman 1972) and the activity of K+-outward rec-tifying channels in maize stelar protoplasts as measuredwith the patch-clamp technique (Roberts 1998). The con-clusion that xylem loading of K+ is regulated separately

from K+ uptake from the external solution (Cram & Pitman1972; Pitman & Wellfare 1978; Engels & Marschner 1992)is underlined by the fact that the K+ channel (activity andexpression) responsible for K+ uptake into the root sym-plast, AKT1 is not affected by ABA (Roberts 1998; Gay-mard et al. 1998).

Roberts (1998) also observed a short-term effect of ABAupon addition to the bath medium during a patch-clampmeasurement: within 5 min after ABA addition the out-ward K+ current in maize stelar protoplasts declined byabout 50%. This effect might be mediated by changes incytosolic pH since a decrease in cytosolic pH from 7·4 to7·2 induced a strong (80%) voltage-independent decreaseof the macroscopic SKOR current in a heterologous expres-sion system (oocytes) (Lacombe et al. 2000). However, bothin dark-grown coleoptiles and in barley aleurone proto-plasts ABA induced an alkalinization of 0·1–0·2 units(Gehring, Irving & Parish 1990; Van der Veen, Heimo-varaa-Dijkstra & Wang 1992). If the same were true forxylem parenchyma cells, then an ABA-induced alkaliniza-tion of the cytoplasm would increase the outward K+ cur-rent. It remains necessary to determine the pH sensitivityof SKOR in planta as well as the effect of ABA on the pHof xylem parenchyma cells.

The high level of expression of SKOR in the pericyclecells raises some questions. As these cells are not in directcontact with the xylem vessels, K+ ions released throughSKOR will end up in the stelar apoplast. If the ion perme-ability of the secondary walls of the xylem vessels is low(see above) then loading of these ions into the xylem mightbe preceded by uptake into the xylem parenchyma cellsfollowed by release into the vessels. This could explain therelatively large number of inward rectifying K+ channels(i.e. channels that activate at hyperpolarizing voltages) inbarley xylem parenchyma cells, two of which are G-proteinregulated (Wegner & de Boer 1997b). The explanation forsuch a complex mechanism of loading is not immediatelyobvious, but in some plants sugar loading into the phloemalso includes an apoplast unloading step followed byuptake into the companion cells and symplast loading intothe sieve elements (Van Bel et al. 1992). These xylem paren-chyma K+ inward rectifying channels may also have a func-tion in the apoplast resorption of K+ ions released by thephloem in the root stelar apoplast. Retranslocation of K+

from the shoot to the root in the phloem is a well-knownphenomenon and it has been speculated that the amountof K+ returning from the shoot acts as a signal for root K+

uptake from the soil (White 1997; Jeschke & Hartung2000).

Xylem perfusion experiments, using tetraethylammo-nium as blocker of SKOR, showed the in vivo function ofSKOR in loading the xylem with K+ (Wegner & De Boer1997a). The phenotype of the skor-1 knockout mutant(Gaymard et al. 1998) indeed showed a reduced transport(50%) of K+ to the shoot of skor-1. Electrophysiologicalcharacterization of the selectivity properties of SKOR inbarley yielded a PCa++/PK+ ratio of 0·68 (Wegner & De Boer1997a). This outcome suggested a role for SKOR in Ca++



unloading of the xylem in roots, which was confirmed bythe higher Ca++ content of the skor-1 leaves (Gaymard et al.1998).

On the way up through the stem into the leaves K+ ionswill be taken up into the symplast by cells in contact withthe vessels and tracheids. The uptake of K+ in the elonga-tion zone of the hypocotyl of Vigna unguiculata is stimu-lated by the growth-promoting hormone auxin andfusicoccin (De Boer et al. 1985). The actual mechanism (ionchannel or H+/K+-symporter) of uptake is not known. Keu-necke et al. (1997) used a patch-clamp approach to unravelthe molecular mechanism of xylem unloading. In proto-plasts isolated from xylem contact cells in leaves of Viciafaba and maize two different types of K+ channels wereidentified, but a role in K+ unloading could not be assigned.The small veins of leaves are often surrounded by a layerof compactly arranged cells, the bundle sheath. These cellsisolate the veins from the intercellular spaces (the apoplast)of the leaf. Keunecke et al. (2001) showed that uptake ofK+ from the vessels of small veins in maize leaves occurs atthe xylem/bundle sheath cell interface.

An intriguing question is whether the flow of K+ (andother solutes) to the different parts of the shoot is simplydetermined by the transpiration stream or whether there isat certain points a ‘special filter’ diverting more K+ to, forexample, rapidly growing buds/leaves that still have a rela-tively low transpiration rate? Gunning, Pate & Green(1970) showed that xylem transfer cells in stems of manyplants were almost exclusively associated with departingfoliar traces (Fig. 4). They estimated that in these regionsthe membrane surface area of cells surrounding the vesselsmay be amplified by as much as 20-fold. This, suggests apotential for very intensive lateral exchange of solutes. Fur-thermore, xylem–phloem exchange is a very important

transport step in nutrient distribution and the C/N economyof the plant (Van Bel 1990; White 1997; Jeschke & Hartung2000). On the other hand, it seems contra-productive totransport large amounts of, for example, K+ to rapidly tran-spiring mature leaves which are not in need of K+. It willbe interesting to study whether the flow of K+ into thepetiole of these leaves is restricted by active resorptionfrom the transpiration stream entering that leaf.

Sodium

Salt-sensitive (glycophytes) and salt-tolerant (halophytes)plants use different strategies when exposed to high Na+

levels in their root medium. In general, glycophytes try toreduce net Na+ uptake from the soil and translocation ofNa+ to the shoot, whereas halophytes rapidly absorb andtranslocate Na+ and sequester it in the vacuole where it isused as ‘cheap’ osmoticum (see Fig. 5a). Especially in halo-phytes the accumulation of Na+ is at the expense of K+

(Fig. 5b). Therefore it is to be expected that different trans-port mechanisms for Na+ are present at the symplast/xylemboundary in the root where xylem loading takes place. Cer-tain glycophytes (such as bean, maize and reed) have devel-oped a xylem Na+ unloading mechanism in the basal partof the root and shoot where Na+ ions are taken up from thetranspiration stream (Jacoby 1965; Johanson & Cheeseman1983; Lacan & Durand 1996; Ehwald 2001). Although thecapacity of this system is limited, it may play a role inavoiding salt damage in the young growing parts of theshoot at moderate salt stress.

Sodium entry into the root symplast is probably medi-ated by weakly voltage-dependent non-selective cationchannels (Roberts & Tester 1995; Tyerman & Skerrett 1999;Davenport & Tester 2000; Demidchik, Davenport & Tester

Figure 4. Conspicuous wall ingrowths of transfer cells lining the xylem of the cotelydonary traces of a seedling of Jacob’s ladder (Polemonium caeruleum) (× 600). V, xylem vessel; XP, xylem parenchyma cell. [Reproduced from Gunning et al. (1970), with permission of Springer-Verlag Austria].

VV

XP

XP

Figure 5. Different strategies of Plantago media (glycophyte) and P. maritima (halophyte) towards the accumulation of Na+ ions in the root and shoot. The glycophyte has a reduced translocation of Na+ from the root to the shoot, whereas the halophyte prefer-entially accumulates Na+ in the leaves. Plants were grown on sand supplied with nutrient solution with Na+ concentrations as indi-cated [from De Boer (1985), where further details can be obtained].

1 2

Na+

(m

e.kg

–1 d

.m.)

0

100

200

300

400

500

600

700

1 2

0 mM NaCl 1 mM NaCl25 mM NaCl

root shoot

P.media P.maritima P.media P.maritima

(a) (b)



2002; Demidchik & Tester 2002). The properties of thesechannels resemble those of cyclic nucleotide-gated chan-nels (CNGC) (Leng et al. 1999). Evidence for a role ofCNGCs in Na+ uptake is two-fold: (i) the cngc1 mutantshows enhanced Na+ tolerance when grown on moderateexternal Na+ (Talke et al. 2001); and (ii) membrane-perme-able cAMP/cGMP analogues reduced unidirectional Na+

influx and overall Na+ accumulation in NaCl grown Ara-biodopsis (Maathuis & Sanders 2001).

In the plasma membrane of barley xylem parenchymacells a non-selective ion channel, called NORC, has beencharacterized but in contrast to the CNGCs it is non-selec-tive for both cations and anions and it is voltage dependent(Wegner & Raschke 1994; Wegner & De Boer 1997a).Although NORC is equally permeable to K+ and Na+, itremains to be shown whether, under physiological condi-tions, NORC has a function in loading of Na+ into thetranspiration stream. NORC also passes glutamate (Weg-ner & de Boer 1997a). In this respect it is noteworthy thata glutamate receptor gene, AtGluR2, is expressed in cellsadjacent to the conducting vessels in Arabidopsis leaves(Kim et al. 2001). AtGluR2 is homologous to mammalianionotropic glutamate receptors that are non-selective cat-ion channels permeable to Ca++, K+ and Na+, but not Mg++

(Lam et al. 1998). The primary physiological function ofthese channels is thought to be Ca++ influx into cells (Den-nison & Spalding 2000). Over-expression of AtGluR2resulted in Ca++-deficiency symptoms and according to Kimet al. (2001) this might be due to increased Ca++ unloadingfrom the transpiration stream in the root and stem causingCa++ deficiency in the young, growing parts of the plant. Thetransgenic plants also exhibited hypersensitivity to Na+ andK+ ionic stress, which might be due to reduced Ca++ uptakein the presence of excess Na+ or K+ in combination withincreased Ca++ unloading on the way up to the top of theplant.

Na+ loading into the xylem of barley and maize roots isnot mediated by the outward rectifying K+ channel of xylemparenchyma cells since this channel is virtually imperme-able to Na+ (Roberts & Tester 1995; Wegner & De Boer1997a). Heterologous expression of SKOR in Xenopusoocytes corroborates this conclusion (Gaymard et al. 1998).A prominent candidate gene encoding a Na+ transporterinvolved in Na+ transport across the symplast/xylem bound-ary is SOS1 (Ding & Zhu 1997; Shi et al. 2000, 2002). SOS1was identified as a NaCl-hypersensitive Arabidopsis mutantand initially described as a mutant defective in high-affinitypotassium uptake (Wu, Ding & Zhu 1996). The 22Na-uptakeand efflux measurements indicated that the sos1 mutantswere impaired both in high-affinity K+-and low-affinity Na+-uptake (Ding & Zhu 1997). However, in these experimentsthe whole seedling was exposed to the uptake/efflux solu-tion and this makes interpretation of the results in terms ofthe SOS1 function in uptake and translocation in plantahazardous. Subsequent sequencing of the locus showed thatsos1 encodes a 127 kDa protein with 12 transmembranespanning domains and with significant sequence similarityto plasma membrane Na+/H+ antiporters from bacteria and

fungi (Shi et al. 2000; see also Hasegawa, Bressan & Pardo2000). In the same article it was reported that SOS1 geneexpression is concentrated in the cells surrounding thexylem, suggesting that SOS1 may function in loading Na+

into the xylem for long-distance transport. Recently, thesuggestion that SOS1 encodes a Na+/H+ antiporter wasexperimentally substantiated using purified plasma mem-brane vesicles from salt-stressed Arabidopsis thalianaleaves (Qiu et al. 2001). Salt-treated sos1 mutant plantsshowed a 60% lower Na+/H+ exchange activity in compari-son with wild-type. SOS1 expression is up-regulated byNaCl stress and this up-regulation is abolished in plants witha mutation in the SOS3 gene encoding a phosphatase withsignificant sequence similarity with the calcineurin B sub-unit from yeast (Liu & Zhu 1998). In the presence of Ca++

SOS3 interacts with SOS2 (a protein kinase) and activatesthe kinase activity of SOS2 (Halfter, Ishitani & Zhu 2000).Apparently, SOS2 directly modulates SOS1 since in vitroaddition of SOS2 to plasma membrane vesicles isolatedfrom NaCl stressed plants resulted in a two-fold increase inthe Na+/H+ exchange activity (Qiu et al. 2001).

During SOS1-facilitated loading of Na+ into the xylem(Shi et al. 2002), one proton flowing from the xylem to thesymplast (a ‘down-hill’ process energetically speaking)drives the transport of one Na+ ion into the xylem vessels.Na+ loading into the xylem is unlike K+ loading probablynot a channel-mediated process. Much remains to belearned from a comparison of K+ and Na+ exchange acrossthe symplast/xylem boundary in experiments in which thexylem of root or shoot segments is perfused with definedsolutions. As discussed above, the high K+ concentration inthe xylem parenchyma cells (Drew, Webb & Saker 1990)allows K+ loading down the electrochemical K+ gradientfacilitated by the K+-outward channel SKOR (Wegner & deBoer 1997a; Gaymard et al. 1998). In line with the proper-ties of SKOR (depolarization activated and inactivated byexternal acidification; Lacombe et al. 2000), xylem perfu-sion of the H+-ATPase activator fusicoccin stops the K+

loading process in roots of the halophyte Plantago maritima(Fig. 3a). Interestingly, the response of Na+ is very differentfrom that of K+: in the first 45 min after the start of fusic-occin perfusion the Na+ concentration in the xylemdecreases (i.e. Na+ unloading of the xylem sap is stimu-lated). This increase in uptake is unlikely to be caused by apH change of the xylem sap (because this is still small), butrather due to the rapid hyperpolarization of the membranepotential induced by fusicoccin (De Boer & Prins 1985; DeBoer et al. 1985). This uptake may be a channel-mediatedprocess. However, when the xylem sap starts to acidifythere is a turning point resulting in Na+ loading into thexylem when fusicoccin has acidified the xylem sap pH withmore than 1 unit (Fig. 3a). The large pH gradient betweenxylem sap and the cytoplasm of adjacent parenchyma cellsmay now facilitate a Na+/H+ antiport system similar toSOS1 (Shi et al. 2002). Similar fusicoccin-induced responsesof xylem sap pH and K+/Na+ (un)loading were observed inxylem perfusion experiments using nodal roots of barley(Fig. 3b).



Shi et al. (2002) proposed a model where SOS1 plays adual role in Na+ loading and reabsorption from the xylemsap, depending on salinity and pH difference between sym-plast and xylem, in analogy to a model proposed by Lacan& Durand (1996). The latter authors observed an increasein Na+ unloading and an increase in K+ loading of the xylemsap when the initial pH of the perfusion solution was moreacidic. They conclude that Na+ is taken up from the xylemthrough a Na+/H+ antiport mechanism, but now with H+

driven ‘uphill’ by a ‘down-hill’ Na+ gradient. However, amechanistic interpretation of their results is difficult sinceit is not clear how the xylem pH changes in view of thestrong pH-stat properties of the xylem (Clarkson et al.1984). Moreover, it is also not clear how the different initialpH values of the perfusion solution affect the membranepotential. The observed positive correlation between K+

loading and acidity of perfusion solution as reported byLacan & Durand (1996) is not in line with a SKOR-mediated loading mechanism since SKOR activity is verylow at acidic external pH (Lacombe et al. 2000). It must bekept in mind that perfusion of an acidic buffer will result inan exchange of protons with other cations such as K+ at thenegatively charged cell walls of the xylem and that this maycause an apparent increase in K+ loading (Senden et al.1992). The reabsorption of Na+ from the upgoing transpira-tion stream in the basal parts of the root and shoot has beenwell documented both in intact plants and in perfused seg-ments (Jacoby 1965; Johanson & Cheeseman 1983; Lacan& Durand 1996; Ehwald 2001). However, the storagecapacity of the stelar cells is limited and therefore a retrans-location of Na+ by the phloem to the root and extrusioninto the medium has been suggested as a mechanism toincrease the detoxification capacity. Although Van Bel et al.1992) has already pointed to the importance of xylem–phloem exchange processes nothing is yet known about themechanism of Na+ loading into the phloem and unloadingin the root.

Water

On its route to the root xylem, water flow is restricted by anumber of morphological structures such as the epidermis,outer cortex layer (exodermis), several layers of corticalcells, endodermis, pericycle, parenchyma cells of centralcylinder, and cell walls of xylem vessels (Steudle & Peter-son 1998). Under conditions of high transpiration, waterflow is passively dragged across the roots at a high rate,whereas under high humidity and absence of transpirationthe ascending water flow to the shoot is generated by rootpressure with low flow rates. In the permanent fluctuatingriver of water, the aim of a plant is to control ion fluxes tothe xylem and to keep most of them relatively independentof water flow. In fact, transpiration may not be required forthe supply of minerals to the shoot (Tanner & Beevers2001). The latter authors provided evidence that xylem flowbrought about by root pressure and Münch’s phloem coun-terflow, might be sufficient for long-distance mineral supply.

The plant can discriminate well between water and ions

because water molecules have no charge, whereas ions do.Water can move via the apoplast (within cell walls), sym-plast (moving via cells without leaving them) or via a cell-to-cell route (moving from one cell to another, but crossingthe membranes, including the tonoplast, of the cells) (Steu-dle 2000). Generations of physiologists before the 1980shad the firm opinion that water flow occurs mainly in theapoplast apart from the endodermis, where water has tocross the endodermal barrier due to the suberized cell wallsof endodermis, known as the Casparian strip (Schreiber1996). Modelling and hydration kinetic experiments, exper-iments with cell and root pressure probes and NMR studiesshowed that water probably moves mainly via the transcel-lular pathway (= a combination of symplastic and cell-to-cell pathways) and therefore via membranes. The argu-ments in favour of a transcellular route were provided byNewman (1976). In fact, the cross-section of apoplast issmall compared with that of the symplast, and water, havinga fast equilibrium, would enter the cells even under rela-tively high flow rates in the apoplast. Much experimentalsupport for the transcellular pathway came from pressureprobe measurements, in which the membrane resistance ofone cell was compared with the resistance of the wholeroot. The data demonstrated that the results for a rootagreed well with those, when modelling a root as a seriesof membranes (Jones et al. 1988; Steudle & Peterson 1998).The idea that water molecules not just diffuse through thelipid bilayer of a membrane, but that proteins facilitaterapid exchange of water was corroborated 10 years ago bythe cloning of genes encoding water channels or aquaporins(for reviews see Maurel 1997; Kaldenhoff & Eckert 1999;Johansson et al. 2000; Maurel & Chrispeels 2001). Becausethe water permeability of the lignified secondary walls ofthe xylem vessels is low, water very likely enters the vesselsacross the pit membranes separating the xylem parenchymacells and the vessels. At this point water has to pass thexylem parenchyma plasma membrane underlying the pitmembranes. Rough calculations show that water flowacross the xylem parenchyma/xylem interface increasesone-hundred-fold compared with the flow into the epider-mis cells (the vessel radius is about 50 times less than theroot radius, with pit fields occupying about 10–20% of thevessel’s surface). So, entry into the xylem is clearly a ‘bottle-neck’ in the plants plumbing system and it is to be expectedthat the density of water channels in these membranepatches is high in order to create a pathway of low resis-tance to water flow.

As early as 1991 it was demonstrated that a putativeplasma membrane aquaporin gene (TobRB27) is highlyexpressed in the central cylinder of roots (Yamamoto et al.1991). Since the cell-to-cell route of water transportincludes passage through the vacuolar membranes (Steudle2000) it is not surprising that the Arabidopis aquaporin γ-TIP was found to be highly expressed in the vascular bun-dles of roots and leaves (Ludevid et al. 1992). In roots ofthe common ice plant (Mesembryanthemum crystallinum)the plasma membrane aquaporins MIP A and MIP B areexpressed in the youngest portions of the xylem and in



xylem parenchyma cells, respectively (Yamada et al. 1995,1997). Kirch et al. (2000) corroborated by means of immu-nolocalization that the MIP-B protein is present in xylemparenchyma cells of Mesembryanthemum roots. An unex-pected observation was the presence of MIP-B betweenxylem vessels, namely in places not expected to containmembranes. Moreover, in these areas and between vesselsand xylem parenchyma cells MIP-B was found in smallareas in a patchy fashion. Recently, the tobacco plasmamembrane aquaporin NtAQP1 was found to be expressedclose to the xylem vessels in roots, stems and petioles (Otto& Kaldenhoff 2000). Electron microscopy analysis of a rootxylem cross-section showed an uneven distribution of theNtAQP1 protein resembling pitch-like structures (Siefritzet al. 2001). This could indicate that this aquaporin is con-centrated at the contact pits between xylem parenchymacells and xylem vessels ensuring a polarity in the hydraulicresistance of cells adjacent to the vessels.

Such a polar distribution of aquaporins could lend sup-port to the hypothesis put forward by Canny (1997a, b)describing the filling of embolized vessels by means ofreverse osmosis. Pressurization of cells bordering an embo-lized vessel could then direct the flow of water along thepathway of least resistance, namely across the pit mem-brane into the vessel. Although there are still other prob-lems with this model, as indicated by Tyree et al. (1999), itseems worthwhile to study the behaviour of aquaporins incells bordering an embolized vessel.

Fine-tuned control over the fluxes of water into, insideand out of the plant is vital for their struggle to survive inever-changing environmental conditions. Intense researchin the field of aquaporins during the last 10 years has shownthat regulation of membrane water permeability by differ-ential expression of a large number of aquaporin genes incombination with post-translational regulation are instru-mental for this fine-tuned control (Maurel 1997; Clarksonet al. 2000; Maurel & Chrispeels 2001). A very nice exampleof such a dynamic regulation is the recently describedchange in osmotic water permeability of leaf cells inresponse to the magnitude of the transpiration stream(Morillon & Chrispeels 2001).

ACKNOWLEDGMENTS

We acknowledge support from the Commission of theEuropean Union via contract FMRX-CT96-0007 (RYP-LOS) to A.H. de B. V.V acknowledges support by the Foun-dation for Single Cell Research and by an IAC fellowship.

REFERENCES

Allen G.J., Chu S.P., Schumacher K., et al. (2000) Alteration ofstimulus-specific guard cell calcium oscillations and stomatalclosing in Arabidopsis det3 mutant. Science 289, 2338–2342.

Aloni R. & Griffith M. (1992) Functional xylem anatomy in root-shoot junctions of 6 cereal species. Planta 184, 123–129.

Canny M.J. (1997a) Vessel contents of leaves after excision – a test

of Scholander’s assumption. American Journal of Botany 84,1217–1222.

Canny M.J. (1997b) Vessel contents during transpiration- embo-lisms and refilling. American Journal of Botany 84, 1223–1230.

Chaffey N.J. & Barlow P.W. (2001) The cytoskeleton facilitates athree-dimensional symplasmic continuum in the long-lived rayand axial parenchyma cells of angiosperm trees. Planta 213, 811–823.

Chaffey N.J., Barlow P.W. & Barnett J.R. (2000) A cytoskeletalbasis for wood formation in angiosperm trees: the involvementof microfilaments. Planta 210, 890–896.

Clarkson D.T., Carvajal M., Henzler T., Waterhouse R.N., SmythA., Cooke D.T. & Steudle E. (2000) Root hydraulic conduc-tance: diurnal aquaporin expression and the effects of nutrientstress. Journal of Experimental Botany. 51, 61–70.

Clarkson D.T., Williams L. & Hanson J.B. (1984) Perfusion ofonion root xylem vessels: a method and some evidence of con-trol of the pH of the xylem sap. Planta 162, 361–369.

Cochard H., Ewers F.W. & Tyree M.T. (1994) Water relations ofa tropical vine-like bamboo (Rhipidocladum racemiflorum):root pressures, vulnerability to cavitation and seasonal changesin embolism. Journal of Experimental Botany 45, 1085–1089.

Crafts A.S. & Broyer T.C. (1938) Migration of salts and water intoxylem of the roots of higher plants. American Journal of Botany25, 529–535.

Cram W.J. & Pitman M.G. (1972) The action of abscisic acid onion uptake and water flow in plant roots. Australian Journal ofBiological Sciences 25, 1125–1132.

Davenport R.J. & Tester M. (2000) A weakly voltage-dependent,non-selective cation channel mediates toxic sodium influx inwheat. Plant Physiology 122, 823–834.

De Boer A.H. (1985) Xylem/symplast ion exchange: mechanismand function in salt-tolerance and growth. PhD Thesis, Univer-sity of Groningen, The Netherlands.

De Boer A.H. (1997) Fusicoccin- a key to multiple 14–3−3 locks?Trends in Plant Science 2, 60–66.

De Boer A.H. & Prins H.B.A. (1984) Trans-root potential in rootsof Plantago media L. as affected by hydrostatic pressure: theinduction of an O2 deficient root core. Plant and Cell Physiology25, 643–655.

De Boer A.H. & Prins H.B.A. (1985) Xylem perfusion of tap rootsegments of Plantago maritima: the physiological significance ofelectrogenic xylem pumps. Plant, Cell and Environment 8, 587–594.

De Boer A.H., Katou K., Mizuno A., Kojima H. & Okamoto H.(1985) The role of electrogenic xylem pumps in K+ absorptionfrom the xylem of Vigna unguiculata: the effects of auxin andfusicoccin. Plant, Cell and Environment 8, 579–586.

De Boer A.H., Prins H.B.A. & Zanstra P.E. (1983) Biphasic com-position of trans-root potential of Plantago species: involvementof spatially separated electrogenic pumps. Planta 157, 259–266.

Demidchik V. & Tester M. (2002) Sodium fluxes through nonse-lective cation channels in the plasma membrane of protoplastsfrom Arabidopsis roots. Plant Physiology 128, 379–387.

Demidchik V., Davenport R.J. & Tester M. (2002) Nonselectivecation channels in plants. Annual Review of Plant Physiologyand Plant Molecular Biology 53, 67–107.

Dennison K.L. & Spalding E.P. (2000) Glutamate-gated calciumfluxes in Arabidopsis. Plant Physiology 124, 1511–1514.

Ding L. & Zhu J.-K. (1997) Reduced Na+ uptake in the NaCl-hypersensitive sos1 mutant of Arabidopsis thaliana. Plant Phys-iology 113, 795–799.

Drew M.C., Webb J. & Saker L.R. (1990) Regulation of K+ uptakeand transport to the xylem in barley roots: K+ distribution deter-mined by electron probe X-ray microanalysis of frozen-hydratedcells. Journal of Experimental Botany 41, 815–825.



Dunlop J. (1982) Membrane potentials in the xylem in roots ofintact plants. Journal of Experimental Botany 33, 910–918.

Ehwald R. (2001) Salt and water stress of common reed growingin wind-protected flat brackish locations. In International Sym-posium ‘Plants Under Environmental Stress’, Moscow, October,23–28, pp. 66–67. Timiriazev Institute, Moscow, Russia.

Engels C. & Marschner H. (1992) Adaptation of potassium trans-location into the shoot of maize (Zea mays) to shoot demand:evidence for xylem loading as a regulating step. PhysiologiaPlantarum 86, 263–268.

Enns L.C., Canny M.J. & McCully M.E. (2000) An investigationof the role of solutes in the xylem sap and in the xylem paren-chyma as the source of root pressure. Protoplasma 211, 183–197.

Ferguson I.B. & Bollard E.G. (1976) The movement of calcium inwoody stems. Annals of Botany 40, 1057–1065.

Fisher J.B., Angeles G., Ewers F.W. & Lopez-Portillo J. (1997)Survey of root pressure in tropical vines and woody species.International Journal of Plant Science 158, 44–50.

Fukuda H. (1996) Xylogenesis: initiation, progression, and celldeath. Annual Review of Plant Physiology and Plant MolecularBiology 47, 299–325.

Gansert D., Burgdorf M. & Lösch R. (2001) A novel approach tothe in situ measurement of oxygen concentrations in the sap-wood of woody plants. Plant, Cell and Environment 24, 1055–1064.

Gaymard F., Pilot G., Lacombe B., Bouchez D., Bruneau D.,Boucherez J., Michaux-Ferrière N., Thibaud J.-B. & SentenacH. (1998) Identification and disruption of a plant Shaker-likeoutward channel involved in K+ release into the xylem sap. Cell94, 647–655.

Gehring C.A., Irving H.R. & Parish R.W. (1990) Effects of auxinand abscisic acid on cytosolic calcium and pH in plant cells.Proceedings of the National Academy of Sciences USA 87, 9645–9649.

Grace J. (1993) Refilling of embolized xylem. In Water Transportin Plants Under Climatic Stress (eds M. Borghetti, J. Grace, &A. Raschi), pp. 52–62. Cambridge University Press, Cambridge,UK.

Gunning B.E.S. (1977) Transfer cells and their roles in transportof solutes in plants. Scientific Progress, Oxford 64, 539–568.

Gunning B.E.S., Pate J.S. & Green L.W. (1970) Transfer cells inthe vascular system of stems: taxonomy, association with nodes,and structure. Protoplasma 71, 147–171.

Halfter U., Ishitani M. & Zhu J.-K. (2000) The Arabidopsis SOS2protein kinase physically interacts with and is activated by thecalcium-binding protein SOS3. Proceedings of the NationalAcademy of Sciences USA 97, 3735–3740.

Hasegawa M., Bressan R. & Pardo J.M. (2000) The dawn of plantsalt tolerance genetics. Trends in Plant Science 8, 317–319.

Herdel K., Schmidt P., Feil R., Mohr A. & Schurr U. (2001)Dynamic of concentrations and nutrient fluxes in the xylem ofRicinus communis – diurnal course, impact of nutrient availabil-ity and nutrient uptake. Plant, Cell and Environment 24, 41–52.

Holbrook N.M. & Zwieniecki M.A. (1999) Embolism repair andxylem tension: do we need a miracle? Plant Physiology 120, 7–10.

Holbrook N.M., Ahrens E.T., Burns M.J. & Zwieniecki M.A.(2001) In vivo observation of cavitation and embolism repairusing magnetic resonance imaging. Plant Physiology 126, 27–31.

Huber B. & Merz W. (1958) Über die Bedeutung des Hoftüpfelv-erschlusses für die axiale Wasserleitfähigkeit von Nadelhölzern.Planta 51, 645–672.

Hwang J.-U., Suh S., Kim J. & Lee Y. (1997) Actin filamentsmodulate both stomatal opening and inward K+-channelactivities in guard cells of Vicia faba L. Plant Physiology 115,335–342.

Jacoby B. (1965) Sodium retention in excised bean stems. Physio-logia Plantarum 18, 730–739.

Jeschke W.D. & Hartung W. (2000) Root–shoot interactions inmineral nutrition. Plant and Soil 226, 57–69.

Johanson J.G. & Cheeseman J.M. (1983) Uptake and distributionof sodium and potassium by corn seedlings. Role of the meso-cotyl in sodium exclusion. Plant Physiology 73, 153–158.

Johansson I., Karlsson M., Johanson U., Larsson C. & Kjellbom P.(2000) The role of aquaporins in cellular and whole plant waterbalance. Biochimica Biophysica Acta 1465, 324–342.

Jones H., Leigh R.A., Wyn Jones R.G. & Tomos A.D. (1988) Theintegration of whole-root and cellular hydraulic conductivitiesin cereal roots. Planta 174, 1–7.

Kaldenhoff R. & Eckert M. (1999) Features and function of plantaquaporins. Journal of Photochemistry and Photobiology B 52,1–6.

Keunecke M., Lindner B., Seydel U., Schulz A. & Hansen U.P.(2001) Bundle sheath cells of small veins in maize leaves are theloction of uptake from the xylem. Journal of Experimental Bot-any 52, 709–714.

Keunecke M., Sutter J.U., Sattelmacher B. & Hansen U.P. (1997)Isolation and patch clamp measurements of xylem contact cellsfor the study of their role in the exchange between apoplast andsymplast of leaves. Plant and Soil 196, 239–244.

Kim S.A., Kwak J.M., Jae S.-K., Wang M.-H. & Nam H.G. (2001)Overexpression of the AtGluR2 gene encoding an Arabidopsishomolog of mammalian glutamate receptors impairs calciumutilization and sensitivity to ionic stress in transgenic plants.Plant and Cell Physiology 42, 74–84.

Kirch H.-H., Vera-Estrella R., Golldack D., Quigley F., Micha-lowski C.B., Barkla B.J. & Bohnert H.J. (2000) Expression ofwater channel proteins in Mesembryanthemum crystallinum.Plant Physiology 123, 111–124.

Kramer D., Läuchli A., Yeo A.R. & Gullasch J. (1977) Transfercells in roots of Phaseolus coccineus: ultrastructure and possiblefunction in exclusion of sodium from the shoot. Annals of Bot-any 41, 1031–1040.

Kuo J., Pate J.S., Rainbird R.M. & Atkins C.A. (1980) Internodesof grain legumes – new location for xylem parenchyma transfercells. Protoplasma 104, 181–185.

Lacan D. & Durand M. (1996) Na+-K+ exchange at the xylem/symplast boundary. Plant Physiology 110, 705–711.

Lachaud S. & Maurousset L. (1996) Occurrence of plasmodesmatabetween differentiating vessels and other cells in Sorbus tormi-nalis L. Crantz and their fate during xylem maturation. Proto-plasma 191, 220–226.

Lacombe B., Pilot G., Gaymard F., Sentenac H. & Thibaud J.-B.(2000) pH control of the plant outward-rectifying potassiumchannel SKOR. FEBS Letters 466, 351–354.

Lam H.-M., Chiu J., Hieh M.-H., Meisel L., Oliveira I.C., Shin M.& Coruzzi G. (1998) Glutamate-receptor genes in plants. Nature396, 125–126.

Läuchli A., Kramer D., Pitman M.G. & Lüttge U. (1974) Ultra-structure of xylem parenchyma cells of barley roots in relationto ion transport to the xylem. Planta 119, 85–99.

Leng Q., Mercier R.W., Yao W. & Berkowitz G.A. (1999) Cloningand functional characterization of a plant nucleotide-gated cat-ion channel. Plant Physiology 121, 753–761.

Liu J. & Zhu J.-K. (1998) A calcium sensor homolog required forplant salt tolerance. Science 280, 1943–1945.

López-Millán A.F., Morales F., Abadía A. & Abadía J. (2000)Effects of iron deficiency on the composition of the leaf apoplas-tic fluid and xylem sap in sugar beet. Implications for iron andcarbon transport. Plant Physiology 124, 873–884.

Ludevid D., Höfte H., Himmelblau E. & Chrispeels M.J. (1992)The expression pattern of the tonoplast intrinsic γ-TIP in Ara-



bidopsis thaliana is correlated with cell enlargement. Plant Phys-iology 100, 1633–1639.

Maathuis F.J.M. & Sanders D. (2001) Sodium uptake in Arabidop-sis roots is regulated by cyclic nucleotides. Plant Physiology 127,1617–1625.

Maurel C. (1997) Aquaporins and water permeability of plantmembranes. Annual Review of Plant Physiology and PlantMolecular Biology 48, 399–429.

Maurel C. & Chrispeels M.J. (2001) Aquaporins. A molecularentry into plant water relations. Plant Physiology 125, 135–138.

McCann M.C. (1997) Tracheary element formation: building up toa dead end. Trends in Plant Science 2, 333–338.

McCully M.E. (1999) Root xylem embolisms and refilling. Rela-tion to water potentials of soil, roots, and leaves, and osmoticpotentials of root xylem sap. Plant Physiology 119, 1001–1008.

McCully M.E., Huang C.X. & Ling L.E.C. (1998) Daily embolismand refilling of xylem vessels in the roots of field-grown maize.New Phytologists 138, 327–342.

Meychik N.R. & Yermakov I.P. (2001) Swelling of root cell wallsas an indicator of their functional state. Biochemistry (Moscow)66, 178–187.

Morillon R. & Chrispeels M.J. (2001) The role of ABA and thetranspiration stream in the regulation of the osmotic water per-meability of leaf cells. Proc. Natl. Acad. Sci. USA 98, 14138–14143.

Mueller W.C. & Beckman C.H. (1984) Ultrastructure of the cellwall of vessel contact cells in the xylem of tomato stems. Annalsof Botany 53, 107–114.

Newman E.I. (1976) Water movement through root systems.Philosophical Transactions of Royal Society London. Series B.273, 463–478.

Okamoto H., Katou K. & Ichino K. (1979) Distribution of electri-cal potential and ion transport in the hypocotyl of Vigna ses-quipedalis. VI. The dual structure of radial electrogenic activity.Plant and Cell Physiology 20, 103–114.

Oparka K.J. & Cruz S.S. (2000) The great escape: phloem trans-port and unloading of macromolecules. Annual Review of PlantPhysiology and Plant Molecular Biology 51, 323–347.

Otto B. & Kaldenhoff R. (2000) Cell-specific expression of themercury-insensitive plasma-membrane aquaporin NtAQP1from Nicotiana tabacum. Planta 211, 167–172.

Parets-Soler A., Pardo J.M. & Serrano R. (1990) Immunolocaliza-tion of plasma membrane H+-ATPase. Plant Physiology 93,1654–1658.

Pate J.S. & Gunning B.E.S. (1972) Transfer cells. Annual Reviewof Plant Physiology 23, 173–196.

Patrick J.W., Zhang W.H., Tyerman S.D., Offler C.E. & WalkerN.A. (2001) Role of membrane transport in phloem transloca-tion of assimilates and water. Australian Journal of Plant Phys-iology 13, 695–707.

Peterson C.A. & Steudle E. (1993) Lateral hydraulic conductivityof early metaxylem vessels in Zea mays L. roots. Planta 189, 288–297.

Peuke A.D., Rokitta M., Zimmermann U., Schreiber L. & HaaseA. (2001) Simultaneous measurement of water flow velocityand solute transport in xylem and phloem of adult plants ofRicinus communis over a daily time course by nuclear magneticresonance spectrometry. Plant, Cell and Evironment 24, 491–503.

Pitman M.G. (1977) Ion transport into the xylem. Annual Reviewof Plant Physiology 28, 71–88.

Pitman M.G. & Wellfare D. (1978) Inhibition of ion transport inexcised barley roots by abscisic acid: relation to water perme-ability of the roots. Journal of Experimental Botany 29, 1125–1138.

Qiu Q.-S., Guo Y., Dietrich M.A., Schumaker K.S. & Zhu J.-K.(2001) Characterization of the plasma membrane Na+/H+

exchanger in Arabidopsis thaliana. In The 12th InternationalWorkshop Plant Membrane Biology, Madison, p. 235. Universityof Wisconsin, Madison, WI, USA.

Roberts S.K. (1998) Regulation of K+ channels in maize roots bywater stress and abscisic acid. Plant Physiology 116, 145–153.

Roberts K. & McCann M.C. (2000) Xylogenesis: the birth of acorpse. Current Opinion in Plant Biology 3, 517–522.

Roberts S.K. & Tester M. (1995) Inward and outward K+-selectivecurrents in the plasma membrane of protoplasts from maize rootcortex and stele. Plant Journal 8, 811–825.

Rokitta M., Peuke A.D., Zimmermann U. & Haase A. (1999)Dynamic studies of phloem and xylem flow in fully differenti-ated plants by fast nuclear-magnetic-resonance microimaging.Protoplasma 209, 126–131.

Rygol J., Pritchard J., Zhu J.J. & Zimmermann U. (1993) Transpi-ration induces radial turgor pressure gradients in wheat andmaize roots. Plant Physiology 103, 493–500.

Salleo S., Lo Gullo M.A., De Paoli D. & Zippo M. (1996) Xylemrecovery from cavitation-induced embolism in young plants ofLaurus nobilis: a possible mechanism. New Phytologist 132, 47–56.

Samuels A.L., Fernando M. & Glass A.D.M. (1992) Immunofluo-rescent localization of plasma membrane H+-ATPase in barleyroots and effects of K nutrition. Plant Physiology 99, 1509–1514.

Sauter J.J., Iten W. & Zimmermann M.H. (1973) Studies on therelease of sugar into the vessels of sugar maple (Acer saccha-rum). Canadian Journal of Botany 51, 1–8.

Scheenen T.W.J., Van Dusschoten D., De Jager P.A. & Van As H.(2000) Quantification of water transport in plants with NMRimaging. Journal of Experimental Botany 51, 1751–1759.

Schreiber L. (1996) Chemical composition of Casparian strips iso-lated from Clivia miniata Reg. roots: evidence for lignin. Planta199, 596–601.

Schurr U. (1998) Xylem sap sampling – new approaches to an oldtopic. Trends in Plant Science 3, 293–298.

Schurr U. & Schulze E.-D. (1995) The concentration of xylem sapconstituents in root exudate, and in sap from intact, transpiringcastor bean plants (Ricinus communis L.). Plant, Cell and Envi-ronment 18, 409–420.

Senden M.H.M.N., Van Paassen F.J.M., van der Meer A.J.G.M. &Wolterbeek H.Th. (1992) Cadmium – citric acids – xylem cellwall interactions in tomato plants. Plant, Cell and Environment15, 71–79.

Shi H., Ishitani M., Kim C. & Zhu J.-K. (2000) The Arabidopsisthaliana salt tolerance gene SOS1 encodes a putative Na+/H+

antiporter. Proceedings of the National Academy of SciencesUSA 97, 6896–6901.

Shi H., Quintero F.J., Pardo J.M. & Zhu J.-K. (2002) The putativeplasma membrane Na+/H+ antiporter SOS1 controls long-dis-tance Na+ transport in plants. Plant Cell. 14, 465–477.

Siau J.F. (1984) Transport Processes in Wood. Springer, Berlin.Siefritz F., Biela A., Eckert M., Otto B., Uhlein N. & Kaldenhoff

R. (2001) The tobacco plasma membrane aquaporin NtAQP1.Journal of Experimental Botany 52, 1953–1957.

Smart C.C. & Amrhein N. (1985) The influence of lignification onthe development of vascular tissue in Vigna radiata L. Proto-plasma 124, 87–95.

Spalding E.P., Hirsch R.E., Lewis D.R., Qi Z., Sussman M.R. &Lewis B.D. (1999) Potassium uptake supporting plant growth inthe absence of AKT1 channel activity. Inhibition by ammoniumand stimulation by sodium. Journal of General Physiology 113,909–918.

Sperry J.S., Holbrook N.M., Zimmermann M.H. & Tyree M.T.



(1987) Spring filling of xylem vessels in wild grapevine. PlantPhysiology 83, 414–417.

St. Aubin G., Canny M.J. & McCully M.E. (1986) Living vesselelements in the late metaxylem of sheathed maize roots. Annalsof Botany 58, 577–588.

Steudle E. (2000) Water uptake by plant roots: an integration ofviews. Plant and Soil 226, 45–56.

Steudle E. & Peterson C.A. (1998) How does water get throughroots? Journal of Experimental Botany 49, 775–788.

Talke I.N., Bouché N., Bouchez D., Fromm H., Sanders D. &Maathuis F.J.M. (2001) Characterization of CNGC1, a putativecyclic nucleotide-gated channel from Arabidopsis thaliana. InThe 12th International Workshop Plant Membrane Biology,Madison, p. 69. University of Wisconsin, Madison, WI, USA.

Tanner W. & Beevers H. (2001) Transpiration, a prerequisite forlong-distance transport of minerals in plants? Proceedings of theNational Academy of Sciences USA 98, 9443–9447.

Turner S.R. & Somerville C.R. (1997) Collapsed xylem phenotypeof Arabidopsis identifies mutants deficient in cellulose deposi-tion in the secondary cell wall. Plant Cell 9, 689–701.

Tyerman S.D. & Skerrett I.M. (1999) Root ion channels and salin-ity. Scientia Horticulturae 78, 175–235.

Tyree M.T. & Sperry J.S. (1989) Vulnerability of xylem to cavita-tion and embolism. Annual Review of Plant Physiology andPlant Molecular Biology 40, 19–38.

Tyree M.T. & Yang S. (1992) Hydraulic conductivity recoveryversus water pressure in xylem of Acer saccharum. Plant Phys-iology 100, 669–676.

Tyree M.T., Fiscus S.D., Wullschleger M.A. & Dixon M.A. (1986)Detection of xylem cavitation in corn under field conditions.Plant Physiology 82, 597–599.

Tyree M.T., Salleo S., Nardini A., Lo Gullo M.A. & Mosca R.(1999) Refilling of embolized vessels in young stems of Laurel.Do we need a new paradigm? Plant Physiology 120, 11–21.

Van Bel A.J.E. (1990) Xylem-phloem exchange via the rays: theundervalued route of transport. Journal of Experimental Botany41, 631–644.

Van Bel A.J.E., Gamalei Y.V., Ammerlaan A. & Bik L.P.M.(1992) Dissimilar phloem loading in leaves with symplastic orapoplastic minor-vein configurations. Planta 186, 518–525.

Van der Veen R., Heimovaraa-Dijkstra S. & Wang M. (1992) Cyto-solic alkalinization mediated by abscisic acid is necessary, butnot sufficient, for abscisic acid-induced gene expression in barleyaleurone protoplasts. Plant Physiology 100, 699–705.

Van Ieperen W., Van Meeteren U. & van Gelder H. (2000) Fluidcomposition influences hydraulic conductance of xylem con-duits. Journal of Experimental Botany 51, 769–776.

Watson R., Pritchard J. & Malone M. (2001) Direct measurementof sodium and potassium in the transpiration stream of salt-excluding and non-excluding varieties of wheat. Journal ofExperimental Botany 52, 1873–1881.

Wegner L.H. & De Boer A.H. (1997a) Properties of two outward-rectifying channels in root xylem parenchyma cells suggest a rolein K+ homeostasis and long-distance signaling. Plant Physiology115, 1707–1719.

Wegner L.H. & De Boer A.H. (1997b) Two inward K+ channels inthe xylem parenchyma cells of barley roots are regulated by G-protein modulators through a membrane-delimited pathway.Planta 203, 506–516.

Wegner L.H. & Raschke K. (1994) Ion channels in the xylem

parenchyma of barley roots. Plant Physiology 105, 799–813.

Wegner L.H. & Zimmermann U. (1998) Simultaneous recordingof xylem pressure and trans-root potential in roots of intactglycophytes using a novel xylem pressure probe technique.Plant, Cell and Environment 2, 849–865.

Wegner L.H., Sattelmacher B., Läuchli A. & Zimmermann U.(1999) Trans-root potential, xylem pressure, and root corticalmembrane potential of ‘low-salt’ maize plants as influenced bynitrate and ammonium. Plant, Cell and Environment 22, 1549–1558.

White P.J. (1997) The regulation of K+ influx into roots of rye(Secale cereale L.) seedlings by negative feedback via the K+ fluxfrom the shoot to root in the phloem. Journal of ExperimentalBotany 48, 2063–2073.

Wilkinson S. & Davies W.J. (1997) Xylem sap pH increase: adrought signal received at the apoplastic face of the guard cellthat involves the suppression of saturable abscisic acid uptakeby the epidermal symplast. Plant Physiology 113, 559–573.

Winter-Sluiter E., Läuchli A. & Kramer D. (1977) Cytochemicallocalization of K+-stimulated adenosine-triphosphate activity inxylem parenchyma of barley roots. Plant Physiology 60, 923–927.

Wisniewski M., Ashworth E. & Schaffer K. (1987) The use oflanthanum to characterize cell wall permeability in relation todepp supercooling and extracellular freezing in woody plants. I.Intergeneric comparisons between Prunus, Cornus and Salix.Protoplasma 139, 105–116.

Wu S.-J., Ding L. & Zhu J.-K. (1996) SOS1, a genetic locus essen-tial for salt tolerance and potassium acquisition. Plant Cell 8,617–627.

Yamada S., Katsuhara M., Kelly W., Michalowski C. & BohnertH. (1995) A family of transcripts encoding the water channelproteins: tissue specific expression in the common ice plant.Plant Cell 7, 1129–1142.

Yamada S., Nelson D.E., Ley E., Marquez S. & Bohnert H.J.(1997) The expression of an aquaporin promoter from Mesem-bryanthemum crystallinum in tobacco. Plant and Cell Physiology38, 1326–1332.

Yamamoto Y.T., Taylor C.G., Acedo G.N., Cheng C.L. & Con-kling M.A. (1991) Characterization of cis-acting sequences reg-ulating root-specific gene expression in tobacco. Plant Cell 3,371–382.

Yeo A.R., Kramer D., Läuchli. A. & Gullasch J. (1977) Ion distri-bution in salt-stressed mature Zea mays roots in relation toultrastructure and retention of sodium. Journal of ExperimentalBotany 28, 17–29.

Zimmermann M.H. (1978) Hydraulic architecture of some diffuse-porous trees. Canadian Journal of Botany 56, 2286–2295.

Zimmermann M.H. (1983) Xylem Structure and the Ascent of Sap.Berlin: Springer-Verlag.

Zwieniecki M.A. & Holbrook N.M. (2000) Bordered pit structureand vessel wall surface properties. Implications for embolismrepair. Plant Physiology 123, 1015–1020.

Zwieniecki M.A., Melcher P.J. & Holbrook N.M. (2001) Hydrogelcontrol of xylem hydraulic resistance in plants. Science 291,1059–1062.

Received 11 December 2001; received in revised form 15 May 2002;accepted for publication 20 May 2002

logistics of water and salt transport through the plant

Documents