Apoplastic Transport of Abscisic Acid Through Roots of Maize: EEct of the Exodermis

Planta (2000) 210: 222±231

Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis

Elenor Freundl1, Ernst Steudle2, Wolfram Hartung1 1Julius-von-Sachs-Institut fuÈ r Biowissenschaften der UniversitaÈ tWuÈ rzburg, Lehrstuhl Botanik I, Julius-von-Sachs Platz 2, 97082 WuÈ rzburg, Germany 2Lehrstuhl fuÈ r P¯anzenoÈ kologie der UniversitaÈ t Bayreuth, UniversitaÈ tsstraûe 30, 95440 Bayreuth, Germany

Received: 26 January 1999 / Accepted: 26 May 1999

Abstract. The exodermal layers that are formed in Introduction maize roots during aeroponic culture were investigated with respect to the radial transport of cis-abscisic acid In a recent publication, we analysed the radial trans- (ABA). The decrease in root hydraulic conductivity port of water and cis-abscisic acid (ABA) through (Lpr) of aeroponically grown roots was stimulated 1.5- excised roots of young maize and sun¯ower plants, and fold by ABA (500 nM), reaching Lpr values of roots the factors and mechanisms that determine the intensity lacking an exodermis. Similar to water, the radial ¯ow of of the ABA signal in the root xylem (Freundl et al. ABA through roots (JABA) and ABA uptake into root 1998). The results prove that, to some extent, ABA can tissue were reduced by a factor of about three as a result be transported apoplastically across the root cylinder of the existence of an exodermis. Thus, due to the by solvent drag with the transpiration stream into cooperation between water and solute transport the xylem vessels. This conclusion is in line with that of development of the ABA signal in the xylem was not Steudle and Peterson (1998) who have recently sum- aected. This resulted in unchanged re¯ection marised the evidence for an apoplastic transport of coecients for roots grown hydroponically and solutes, such as nutrients. In the endodermis, the aeroponically. Despite the well-accepted barrier proper- apoplastic passage of solutes may be restricted to ties of exodermal layers, it is concluded that the arrays where Casparian bands are not yet mature, such endodermis was the more eective ®lter for ABA. Owing as in the root tip, or where root initials break through to concentration polarisation eects, ABA may accu- the endodermis. In addition, due to low amounts of mulate in front of the endodermal layer, a process suberin the maize endodermis may not be as imperme- which, for both roots possessing and lacking an able as expected (Zeier and Schreiber 1998; Schreiber exodermis, would tend to increase solvent drag and et al. 1999). In the earlier paper by Freundl et al. hence ABA movement into the xylem sap at increased (1998), we argued that the apoplastic transport of ABA water ¯ow (JVr). This may account for the higher ABA could compensate for the dilution resulting from water concentrations found in the xylem at greater pressure uptake, at least to some extent. Hence, apoplastic ABA dierence. transport in the root could aect the root-to-shoot signal of ABA. Key words: Abscisic acid ± Exodermis ± Hydraulic The experiments by Freundl et al. (1998) were per- conductivity ± Root (ABA transport) ± Water transport ± formed with plants that were cultivated hydroponically Zea and were lacking a Casparian band in the hypodermis (Peterson 1988; Zimmermann and Steudle 1998). When grown in soil, most plants, however, develop a Caspar- ian band which may play an important role during the uptake of water and nutrient ions and their retainment within the root (Perumalla et al. 1990; Damus et al. 1997; Enstone and Peterson 1998). The same should be true for solutes produced in the root such as the stress Abbreviations and symbols: ABA = cis-abscisic acid; CABA = o hormone ABA. ABA concentration in the medium; CABA = ABA concentration in x With respect to ABA one may speculate that, the xylem; Lp = hydraulic conductivity of root; r = re¯ec- r ABA depending on the rate of transpiration, this compound tion coecient for ABA; JABA = ABA ¯ow per unit root surface area; JVr = volume ¯ow per unit root surface area could be taken up from the soil solution which contains Correspondence to: W. Hartung; Fax: 49 (931) 888 6158; ABA at concentrations ranging between 1 and 10 nM E-mail: [email protected] (Hartung et al. 1996). On the other hand, release of E. Freundl et al.: Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis 223 endogenous ABA to the soil solution may be retarded by pictures (Intas Colour LC 100C low light camera; GoÈ ttingen, the exodermis, resulting in increased ABA concentra- Germany). Lengths of primary roots were measured with a ruler. tions in the root apoplast. In the present paper, we deal Root surface areas were determined as described in detail by with the role of the exodermis and how this transport Freundl et al. (1998). barrier may aect the uptake of ABA from the root Measurement of xylem sap ¯ow induced by subatmo- medium. We have compared the radial transport of spheric pressures (vacuum). Seedlings were cut at a distance of ABA across maize roots that had developed a complete 20 mm above the root base. Excised roots with the mesocotyl still exodermis with those lacking it. External ABA concen- attached were ®xed to a capillary using a pressure-tight silicone seal trations of 5±500 nM have been applied to the root ®xed by a screw (Freundl et al. 1998). The root medium was medium. This includes the range of naturally occurring aerated. Suction applied to the root system caused xylem sap ¯ow ABA concentrations in the soil. Although measured into a calibrated capillary. The subatmospheric pressure was raised in steps of 0.02 MPa from zero to )0.08 MPa. At each pressure responses are quite variable and sometimes contradic- step, a steady water ¯ow across the root system was waited for. tory, it is known from the literature that ABA aects the Usually, the steady state was attained after 15±30 min. Water ¯ow )1 hydraulic conductivity of roots and hence root water (JVr in m s ) was measured for each step of change in pressure to )1 )1 ¯ow (Markhart et al. 1979; Glinka 1980; Fiscus 1981). calculate hydraulic conductivity Lpr in m s MPa for roots from This, in turn, would also modulate the ABA signal in the aeroponically and hydroponically grown maize plants (Zimmer- xylem. Therefore, changes in root hydraulics have been mann and Steudle 1998). The overall driving force contained an osmotic and a hydraulic measured as well. Rather than applying pressure to the (hydrostatic) component. The former was estimated from the root medium, cut surfaces of maize mesocotyls have dierence in concentration between xylem sap (at steady state at a been subjected to vacuum to induce water ¯ows which given constant hydrostatic pressure gradient) and medium using a simulate conditions in an intact transpiring plant re¯ection coecient of the root of rsr 0:6 for the nutrients in the (Freundl et al. 1998). From the results we conclude that xylem sap and medium since (Steudle and Frensch 1989; the exodermis remarkably alters the transport properties Zimmermann and Steudle 1998): of maize roots. JVr LprPr rsr px po Eq. 1

Here, px means the osmotic pressure of the xylem sap and po the osmotic concentration of the root medium. It held that px po. Materials and methods The correction for the osmotic component would be substantial at low ¯ow rates with a non-linear relationship between water ¯ow Plant material. Seeds of maize (Zea mays L. cv. Garant FAO and driving force (Fiscus 1975). However, this could be omitted, 240; Asgrow, Bruchsal, Germany) were germinated on ®lter paper since the relations were linear and water ¯ows created by osmotic gradients should be small in maize (see Discussion). soaked in 0.5 mM CaSO4 for 4±6 d at 21 °C in the dark. Maize seedlings developed roots of a length of up to 110 mm and primary The root medium was aerated. Root pressure was observed in leaves of a length of up to 30 mm. Some of the seedlings were roots in non-aerated and aerated media using dierent intensities transferred to aerated hydroponic culture vessels as described by of air bubbling. When aeration was switched on, there were no Freundl et al. (1998). Others were grown in mist culture reductions in root pressure, indicating that air bubbling did not (aeroponics) using the same nutrient solution. The latter seedlings cause leakages resulting from cracks and other damage (data not were ®xed by pieces of foam rubber in holes on top of a cubic PVC shown). The Lpr was determined for each individual root system, ®rst in the absence and then, subsequently, in the presence of ABA box of 1 m3 (Zimmermann and Steudle 1998). Roots protruded from the holes and grew into the box. At the bottom of the box, a in the medium. Concentrations used for Lpr measurements were centrally placed timer-regulated air conditioner produced mist for either zero or 500 nM of ABA. Xylem sap was collected with a 10 h a day. (Defensor, Axair; NuÈ rnberg, Germany). Apart from syringe from the capillary attached to the root system. Just that, growing conditions were similar to those used during before raising the vacuum by another step, osmotic concentrations hydroponic culture. Plants employed in experiments were grown of media and xylem sap were measured by freezing-point either in hydroponic or aeroponic culture for 7 d. Roots from depression (Osmomat 030, Gonotec, Berlin, Germany). Through- plants raised hydroponically became 80±400 mm long. Overall out the paper, the atmosphere has been used as a reference (zero shoot length was 100±250 mm. Seedlings grown aeroponically pressure). Therefore, suction forces (subatmospheric pressures) achieved root and shoot lengths of 200±550 mm and 100±250 mm, applied to the root have a negative sign. They were identical to the respectively. Surface areas of the roots of hydroponically and gradients in hydrostatic pressure applied between root medium aeroponically grown plants were between 0.0013 and 0.0110 m2 and and xylem. between 0.0057 and 0.0278 m2, respectively. Flow of ABA across roots of maize grown hydroponically and aeroponically at dierent subatmospheric pressures Root anatomy and root surface areas. Freehand cross sections were prepared from primary roots of plants grown by ()0.02 and )0.06 MPa). Two hours before the experiments both culture methods under conditions identical to those used in were started, maize plants either raised aeroponically or hydropo- the experiments. Samples were taken at distances of 20±400 mm nically were incubated in nutrient solution containing ABA at from the root tip. Sections were either stained for 1 h with 0.1% concentrations ranging between 0 and 500 nM. For 1 h, xylem sap berberine hemisulfate and subsequently for 45 min with 0.5% was collected at intervals of 10 min at )0.02 MPa. Then the toluidine blue O (w/v) (Brundrett et al. 1988), or they were subatmospheric pressure was decreased to )0.06 MPa and sap marked with Sudan III (Merck, Darmstadt, Germany) according collection continued for another hour. Water ¯ow was determined to Gerlach (1984). Samples were immediately examined using an by weighing harvested fractions of xylem sap. Axioplan microscope and ¯uorescence excitation provided by a violet ®lter set (Zeiss, Oberkochen, Germany; excitation ®lter Determination of the apparent re¯ection coecient of 365 nm, dichroitic mirror FT 395, barrier ®lter LP 397). For roots for ABA. The re¯ection coecient of ABA rABAwas sections stained with Sudan III conventional bright-®eld illumi- calculated from solvent drag (Steudle 1994). It has been readily nation was chosen. Microscope pictures were documented using shown that diusional terms (solute permeability) can be a video camera connected to a computer to produce colour neglected (Freundl et al. 1998). Equation 2 was derived from 224 E. Freundl et al.: Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis

ABA steady-state conditions of JABA (JABA =Cx JVr; Freundl et al. distances between 25 and 30 mm from the root tip. As 1998) in the absence of an active transport of ABA into the root described in detail by Freundl et al. (1998), the endo- xylem, i.e. in the absence of an endogenous production of ABA dermis reached its secondary state at 60 mm and its in the root (see Discussion): tertiary state at 180 mm from the root tip. A continuous 2CABA r 1 x Eq. 2 exodermis was never found (Fig. 1A). However, patches ABA ABA ABA Cx Co of cells with suberised walls were occasionally observed (Fig. 1B). Roots grown in aeroponic (mist) culture The term rABA should be denoted as an `apparent' root re¯ection coecient, because the estimation of the concentration of the solute developed a primary endodermis at the same distances (ABA) within the permeation barrier of the root (endodermis and as roots grown hydroponically. Yet, unlike hydroponi- exodermis) as the arithmetic mean of the external (medium) cally grown roots, those from aeroponic culture regu- ABA ABA concentration (Co ) and that of the xylem (Cx ) assumes a larly developed exodermal Casparian bands at distances single homogeneous barrier. However, because of the composite between 30 and 50 mm behind the root tip (Fig. 1C). structure of the root, there may arise problems with an accumu- The exodermis matured 60±90 mm behind the root tip lation of the solute in front of or within the barrier (unstirred layers). As discussed by Freundl et al. (1998), the arithmetic mean into the secondary state. After 250±300 mm, the endo- may represent a lower limit of the actual concentration. In this dermis usually reached its tertiary state with typical case, the rABA calculated from Eq. 2 would be a lower limit as well. U-shaped cell-wall thickenings. In aeroponic culture the Depending on the absolute value of the external ABA concentra- exodermis never reached a tertiary state within a growth tion, accumulation of ABA in the barrier will also tend to increase period of 7 d. the solvent drag component of JABA, the ABA ¯ow per unit root surface area. Morphology. Similar to hydroponically grown roots, Analysis of ABA. Xylem sap samples were taken up with Tris- roots grown aeroponically developed a root hair zone at buered saline buer (Tris-HCl: 50 mM Tris, 150 mM NaCl, a distance of about 30 mm from the tip, but the zone 1 mM MgCl2, pH = 7.8) and analysed immunologically by an was more extended than that of roots grown hydropo- enzyme-linked immunosorbent assay (ELISA) as described by nically (Hofer 1996). Independent of the technique of Weiler (1986). Puri®cation of samples was not necessary. cultivation, lateral roots became visible at distances of more than 100 mm. Their breaking through the endo- Uptake of ABA by maize roots of intact plants and by root dermis should have caused transient bypasses for roots tips under non-transpiring conditions. Seven-day-old maize from both types of culture (Fig. 1D; Steudle et al. 1993; plants from aeroponic and hydroponic cultures were kept for 3 h in aerated nutrient solution, containing 500 nM ABA at 25 °C under Frensch et al. 1996). Compared to hydroponically a mercury-vapour lamp producing 200 lmol m)2 s)1. Controls and grown roots, surface areas of roots of the same age ABA-treated roots were homogenised in liquid nitrogen and grown aeroponically were larger by a factor of 3.5. On extracted with 80% methanol at )25 °C for 2 d after determination average, the root-to-shoot fresh-weight ratio had in- of fresh weight. Extracts were passed through Waters C18-Sep-Pak creased in plants grown aeroponically as compared to cartridges (Millipore Waters, Bed ford, Mass., USA) and eluted those grown hydroponically (data not shown). with 80% methanol. Methanol was removed from eluates by evaporation. Aqueous residues were acidi®ed to pH = 3 with HCl and partitioned three times against an equal volume of ethyl Xylem sap ¯ow and hydraulic conductivity, (Lpr) at low acetate. The combined organic fractions were reduced to dryness, subatmospheric pressures. Figure 2A±C show that, at a taken up in Tris-buered saline buer and analysed by ELISA. given subatmospheric pressure dierence DP, xylem sap The reliability of the ELISA for maize tissues had been con®rmed ¯ow per unit root surface area (JVr t) increased linearly earlier by Hartung et al. (1994). Uptake of ABA under non- with time. Rates of water ¯ow were smaller by a factor of transpiring conditions by root tips was investigated using labelled ()-cis-trans-[214C]ABA (speci®c activity: 833 MBq mmol)1; about 2 in aeroponically grown roots than in roots Amersham Buchler, Braunschweig, Germany). Seven-day-old roots grown hydroponically (Table 1). When water ¯ow from hydroponic and aeroponic cultures were cut 80±110 mm density (JVr) was plotted against DP the resulting curves behind the root tip, immediately before the lateral root zone. Cut were found to be a straight line for most cases. The slope surfaces were sealed with paran (melting point: 42±44 °C; Merck, of the JVr (DP) curves represented the hydraulic conduc- Darmstadt, Germany). Uptake experiments were performed at an tivity of the roots (Lp ). In the presence of 500 nM ABA, external ABA concentration of 10 lM and a pH of 5.5. The r J and Lp increased in both hydroponically and medium contained 1 mM glucose, 2 mM CaCl2, 6 mM KCl, Vr r )1 1.5 mM KNO3, 1 mM MgSO4, 2 mM K2HPO4, 158Umg aeroponically grown roots (Fig. 2B, Table 1) by a factor penicillin and 10 mM Mes-NaOH. Root tips were pretreated with of 3 and 1.5, respectively. The osmotic pressure of xylem 10)5 M tetcyclacis, a norbornanodiacetine derivate that prevents sap decreased for roots grown aeroponically and hydro- oxidative ABA degradation (Daeter and Hartung 1990). Samples ponically with increasing suction force from a mean of were taken at time intervals of 15, 30, 45, 60 min and 2, 3, 4, 5, 6, 60 mosmol kg)1 to 30 mosmol kg)1 (dilution of xylem 8 h. Roots were rinsed twice in cold buer solution for 10 s. Tissues were extracted with 1 ml 80% (v/v) methanol for 24 h. Radioac- sap in the presence of an increased water ¯ow). Dilution tivity was measured in a b-scintillation counter. Data were of xylem sap was similar for roots grown aeroponically expressed on a fresh-weight basis. Experiments were repeated twice. and hydroponically and for ABA-treated and untreated roots. The nutrient solution had an osmotic concentration of 11 mosmol kg)1 (data not shown). Results Flow of ABA across roots of maize grown hydroponically Anatomy. In hydroponic culture, maize roots developed and aeroponically at dierent subatmospheric pressures. a primary endodermis with a Casparian band at When a low suction force of )0.02 MPa was applied to E. Freundl et al.: Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis 225

Fig. 1A±D. Freehand cross-sections of main roots of 7-d-old maize exodermal layer could be observed using berberine hemisulfate. C and plants were stained with Sudan III (A,D) or with berberine D show cross-sections through roots cultivated aeroponically. hemisulfate and counterstained with toluidine blue-O (B,C). A and At 40 mm above the root tip a complete Casparian band was found B show cross-sections through hydroponically grown maize roots. By in the hypodermal layer (C). However, where laterals break through the use of Sudan III (A) no suberin deposits could be detected in the the exodermis, zones with unsuberised hypodermal cells were found, hypodermis at 300 mm above the root tip. At distances of 180 mm resulting in a transient apoplastic bypass (D) above the root tip (B) patches with suberised cell walls in the

the xylem, roots of maize incubated in a nutrient Table 2, ratios of JABA()0.06 MPa)/JABA()0.02 MPa) solution containing 20 nM ABA attained steady ABA are summarised for maize grown hydroponically and in concentrations of the xylem sap of 1 nM to 5 nM mist culture at external ABA concentrations between for both culture conditions (Fig. 3A). Water ¯ow 5 nM and 500 nM. The data show for all measure- (JVr) was reduced by 20% for aeroponically grown ments that the ratio mentioned above was ³1. Unlike ABA roots (Fig. 3B). Flows of ABA (JABA) calculated from other solutes in the xylem, a dilution of Cx in ABA Cx and JVr and referred to unit root surface response to the increased water ¯ow at higher tensions area were, on average, 1.5 ´ 10)14 mol m)2 s)1 and was never observed. On the contrary, the higher )14 )2 )1 0.6 ´ 10 mol m s , respectively (Fig. 3C). After suction force caused JABA to increase to a greater decreasing the pressure from )0.02 MPa to )0.06 MPa, extent than the water ¯ow (JVr). Thus, ABA reacted in ABA Cx increased from 5 nM up to 14 nM (on average) a dierent way from nutrient ions present in the root for maize grown hydroponically and from 2 nM up to medium and xylem solution. 6 nM (on average) for aeroponically grown maize (Fig. 3A); therefore, JABA increased by a factor of 8 Determination of the apparent re¯ection coecient of and 5, respectively. Thus, for both cultivation methods, ABA (rABA) at dierent external ABA concentrations. JABA increased with increasing suction force. In Equation 2 was used to calculate rABA for maize at 226 E. Freundl et al.: Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis

Fig. 2A±C. Xylem sap ¯ow in the presence of hydrostatic pressure shown in A and B, are plotted against the pressure dierence applied gradients (root cultivated aeroponically). The water volume driven to drive water ¯ow across the root. It can be seen that the relation is through the roots increased linearly with time at dierent constant linear, but that the lines do not pass through the origin because of an pressure gradients DP. A Control conditions (®lled symbols). B Eect active nutrient uptake which causes an osmotic exudation. The of 500 nM ABA (open symbols). C Water ¯ows per unit root surface hydrostatic pressure gradient represents the driving force. Hydraulic 3 )2 )1 area (JVr in m m s ) as calculated from the slopes of the graphs conductivity (Lpr) was calculated from the slope in C dierent external ABA concentrations (Table 3). For pressure dierences (0.02 MPa; Freundl et al. 1998). hydroponically grown plants, rABA increased to average The rABA of aeroponically cultivated roots ranged values of between 0.6 and 0.9, when the ABA concen- nearly within the same limits and showed the same tration grew from 10 to 500 nM. The rABA was pressure dependency. The ABA concentrations in the consistently lower at high (0.06 MPa) than at low xylem sap of maize plants from both culture systems showed no dierences at )0.02 MPa (Fig. 3A). Under Table 1. Root hydraulic conductivity (Lp ) from young maize increased subatmospheric pressure ()0.06 MPa) the r amount of ABA in the xylem sap tended to increase roots cultivated hydroponically and aeroponically. The Lpr was measured for both cultivation methods before and after treatment more remarkably for hydroponic than for aeroponic with 500 nM ABA. Root hydraulic conductivity increased after culture, which resulted in a lower rABA. However, this ABA treatment. Thus, aeroponically grown roots which exhibited a was not signi®cant for any measured ABA concentra- lower Lpr than roots from hydroponics reached a hydraulic con- tions (Table 3). ductivity of roots lacking an exodermis The ¯ow JABA was always larger for roots grown Plant )ABA +ABA hydroponically than for those grown aeroponically. No. Lpr Lpr Ratios between ABA ¯ows (hydroponics vs. aeroponics) 107 (m s)1 MPa)1)107 (m s)1 MPa)1) for dierent concentrations ranged between 2 and 3.7 Hydroponics with a mean of 2.9 0.6 (SD, n = 7). The level of H1 1.7 5.3 the endogenous ABA ¯ow into the xylem sap was also H2 1.8 6.7 determined in the absence of ABA in the nutrient H3 2.7 4.4 solution. For roots from hydroponic culture, the H4 1.8 7.0 endogenous ABA concentration in the xylem ranged H5 1.2 5.3 from 1.5 to 10 nM, for aeroponic culture from 1 to Mean 1.8 5.7 9 nM. Thus, there was an endogenous JABA of 1.6 SD 0.5 1.1 and 0.8 ´ 10)14 mol m)2 s)1 at )0.06 MPa, respectively 108 (m s)1 MPa)1)108 (m s)1 MPa)1) (Fig. 4). To increase the ABA ¯ow into the xylem over Aeroponics the endogenous limit by solvent drag, ABA had to be A1 8.2 11.5 added to the medium at least at concentrations of 20 nM A3 7.7 13.7 for hydroponics and 50 nM for aeroponics. For both A4 7.2 11.3 treatments, JABA increased in mean with increasing the A5 8.2 15.2 external ABA concentration. However, the increase was A6 8.9 12.1 A7 10.5 14.4 not signi®cantly dierent after addition of ABA concentrations which are higher than 50 nM for aeroponic- Mean 8.4 13.0 ally grown roots and 100 nM for hydroponically grown SD 1.2 1.6 roots. E. Freundl et al.: Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis 227

Table 2. Pressure perfusion experiments with excised root systems of maize grown hydroponically and aeroponically at dierent external ABA concentrations. The ratios of ABA (JABA) and water (JVr) ¯ows at high (0.06 MPa) and low (0.02 MPa) pressure dif- ferences applied to the cut surface are given. For all cases there was a stronger increase in JABA than in JVr as suction force (water ¯ow) increased. Mean values are given SD (n = number of replicates)

ABA Co Ratio of Ratio of JABA ()0.06 MPa)/ JVr ()0.06 MPa)/ (nM) JABA ()0.02 MPa) JVr ()0.02 MPa) Hydroponics 500 5.9 1.8 (n = 5) 1.9 0.3 (n =5) 100 3.4 1.1 (n = 5) 2.0 0.4 (n =7) 50 3.4 1.4 (n = 8) 1.7 0.3 (n = 12) 20 8.0 3.5 (n = 5) 2.3 0.4 (n =5) 10 2.6 1.1 (n = 6) 1.7 0.3 (n =6) 5 3.5 1.2 (n = 4) 1.7 0.3 (n =5) Aeroponics 500 3.1 1.4 (n = 9) 1.9 0.3 (n =9) 100 2.9 1 (n = 9) 1.7 0.4 (n =9) 50 1.9 0.5 (n = 8) 1.5 0.3 (n =8) 20 4.9 3.1 (n = 9) 1.9 0.5 (n =9) 10 4.0 1.7 (n = 9) 1.8 0.4 (n =9) 5 3.1 1.5 (n = 8) 1.7 0.6 (n =8)

Table 3. Apparent re¯ection coecients of abscisic acid (rABA) for young root systems of maize cultivated hydroponically and aeroponically. The rABA decreased as the pressure dierence (or volume ¯ow) increased. On the other hand, an increase in the ABA concentration of the root medium caused an increase in the root re¯ection coecient for the hormone. Mean values are given (SD, n = number of replicates)

ABA Co Pressure Apparent rABA dierence (MPa) Hydroponics Aeroponics (nM) (n =6) (n = 6±9)

500 0.02 0.96 0.02 0.89 0.05 0.06 0.83 0.05 0.85 0.03 Fig. 3A±C. Xylem ABA concentration (A), water ¯ow (B), and 100 0.02 0.68 0.06 0.65 0.10 radial ¯ow of ABA (C), through roots of maize grown hydroponically 0.06 0.54 0.04 0.52 0.11 (®lled symbols) and aeroponically (open symbols) at a subatmospheric pressure of )0.02 and )0.06 MPa. The ABA concentration of the 50 0.02 0.71 0.04 0.63 0.13 medium is 20 nM. Data points are means of measurements from 5±9 0.06 0.48 0.09 0.61 0.12 roots SD 20 0.02 0.73 0.27 0.75 0.14 0.06 0.12 0.12 0.59 0.19 10 0.02 0.4 0.07 0.68 0.23 Uptake of ABA in maize root tissues of intact plants. On 0.06 0.28 0.05 0.54 0.26 average, hydroponically and aeroponically grown roots )1 5 0.02 0.08 0.15 0.48 0.26 tissues contained 4 to 5 pmol ABA (g FW) , which 0.06 )0.1 0.10 0.35 0.26 would be typical values for unstressed maize plants (Zhang and Davies 1989). After 3 h incubation in 500 nM ABA, roots from hydroponic culture had taken up 3.4 times more ABA than aeroponically cultivated led 10 lM[14C]ABA solution which contained 10)5 M roots (Fig. 5). It should be noted that these ®gures Tetcyclacis. As compared to plants grown hydroponi- represent a lower limit, because degradation was not cally, uptake of ABA by root tips of aeroponically inhibited in these experiments (no Tetcyclacis was cultivated plants was slower by a factor of 2.7 (on added). average). Using the linear part of the uptake kinetics within the ®rst 5 h, a net rate of 1.1 nmol (g FW))1 h)1 Uptake of ABA in maize root tissue under non-transpiring for aeroponically grown and 2.8 nmol (g FW))1 h)1 for conditions (diusional uptake of ABA at high external hydroponically grown plants was calculated, respectively concentrations). Root tips were incubated in radiolabel- (Fig. 6). 228 E. Freundl et al.: Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis

Fig. 4. Solvent drag JABA at a pressure dierence of 0.06 MPa for dierent external ABA concentrations in the medium (0±500 nM). The ABA ¯ows through roots grown hydroponically (®lled columns) are consistently higher than for roots grown aeroponically (open columns). Mean values are givenSD (n = 3±9 roots)

Fig. 6. Time course of [14C]ABA uptake into root tips (80±110 mm long) from hydroponically (d) and aeroponically (s) grown roots. External pH was 5.5, external ABA concentration was 10 lM, Tetcyclacis was present at a concentration of 10)5 M. Mean values are givenrange (n = 2 roots) Fig. 5. Abscisic acid content of roots grown hydroponically and aeroponically in the light before (open columns) and after (®lled columns) incubation in 500 nM ABA for 3 h. The pH of the nutrient solution was 5.5. Mean values are givenSD (n = 4±6 roots) conductivity. Published data about ABA eects on root Lpr are contradictory. Fiscus et al. (1981) and Markhart et al. (1979) found a decrease in root Lpr after ABA Discussion treatment. On the other hand, Glinka (1977, 1980) and Ludewig et al. (1988) showed a stimulation of water In the present paper, the eect of the root exodermis on ¯ow through plant roots after addition of ABA to the apoplastic transport of ABA has been examined by root medium and the apical bud, respectively. In the comparing hydroponically and aeroponically cultivated present experiments, ABA increased the overall hydrau- maize roots. Treatments caused large dierences in the lic conductivity of excised maize root systems. The radial transport of both water and ABA through the hydraulic conductivity of roots from hydroponically roots. However, since both ABA and water ¯ow were grown maize plants was in a range similar to that similarly retarded due to the presence of a Casparian reported earlier (e.g. Steudle and Frensch 1989; Zhu and band in the hypodermis, the ABA concentration in the Steudle 1991). Unlike these studies, the root Lpr given xylem (the root-to-shoot signal) remained virtually here represents an average value for entire root systems. unaected. The results extend earlier ®ndings on hydro- It is known that transport coecients and ¯ows change ponically grown roots of maize and sun¯ower which along developing roots (Melchior and Steudle 1993; indicated a signi®cant apoplastic bypass ¯ow of ABA Frensch et al. 1996). However, as pointed out in more contributing to the hormonal stress signal in the xylem detail by Freundl et al. (1998), these complications may (Freundl et al. 1998). be not so relevant in the context of this paper, since the In discussing ABA transport through roots, one has average hydraulic conductivity of the maize root systems to keep in mind that ABA alters the root hydraulic was similar to that of individual roots. Simultaneous E. Freundl et al.: Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis 229 measurements of Lpr from roots grown aeroponically the root, which would tend to increase the solvent-drag (in mist culture) showed a decreased Lpr by a factor of eect. However, this unstirred-layer eect is not more 2. This is in accordance with Zimmermann and pronounced for aeroponic than for hydroponic culti- Steudle (1998), who found that root Lpr decreased by vation. Thus, we conclude that, even in the presence of a factor of 3.6 in the presence of a Casparian band in the an exodermis, there is a substantial apoplastic trans- hypodermis. The exodermal layer is built up 30 mm port of ABA. The endodermal layer appears to be the behind the root tip, resulting in a barrier to water more eective barrier for apoplastic ABA transport and ions (Peterson 1987, 1989; Zimmermann and than the exodermis. This is in accordance with the Steudle 1998). As maize plants develop an exodermal results of Zimmermann and Steudle (1998) who studied layer under ®eld conditions (Peterson 1992), the the eect of Casparian bands in the hypodermis on the aeroponic culture is closer to the natural growing transport of the ionic apoplastic dye trisodium 3- conditions. hydroxy-5,8,10-pyrenetrisulfonate (PTS). Maize plants are able to compensate for reduced The fact that there is an endogenous ¯ow of ABA Ã hydraulic conductivity caused by a hypodermal Casp- into the root xylem (JABA) complicates the calculation arian band (as happens during water stress in dry or of re¯ection coecients of ABA and their interpretat- saline soil) by an increased ABA production. Addition ion, especially at low external concentrations. The of ABA increased Lpr of aeroponically grown maize results show that external concentrations of 20±50 nM roots to values of roots lacking an exodermal layer. The are required for the solvent drag to override the eect Ã Ã mechanism by which ABA aects the overall hydraulic of JABA. When JABA cannot be neglected, the overall conductivity of roots is not known, but water channels solute ¯ow measured in the steady state will be given may be involved (Tyerman et al. 1999). Accordingly, by: ABA could either trigger existing channels to open or ABA could aect the densitiy of water channels (Steudle and JABA JVrCX Henzler 1995). It should be noted that in the present 1 r 0:5 CABA CABAJ JÃ experiments we have varied the suction force (hydro- ABA x o Vr ABA static pressure gradient). This implied simultaneous Eq. 3 changes in the osmotic pressure in the xylem sap, which is a driving force as well. However, these eects should Solving this for rABA yields: be small, mainly because osmotic gradients do not cause 2CABA JÃ large volume ¯ows. This is expressed by a lower r 1 x 1 ABA Eq. 4 hydraulic conductivity for osmotic pressure gradients ABA ABA ABA J Cx Co ABA than for hydrostatic ones, which has been known for a Ã long time (e.g. Steudle and Frensch 1989; Zimmermann whereby JABA > JABA. Equation 4 diers from Eq. 2 by and Steudle 1998). the last term in brackets on the right-hand side. If it In the previous paper by Freundl et al. (1998), it has were precisely known, this term could be used to been shown that ABA is transported largely by solvent correct the endogenous ¯ow of ABA. Equation 4 shows drag in the apoplast. The conclusion was drawn due to that the correction would tend to increase rABA. On the Ã the observation that, after increasing radial water ¯ow, other hand, when JABA 0 or is negligible as compared ABA with the solvent-drag component, Eq. 4 reduces to Eq. the ABA concentration in the xylem (Cx ) was never ABA 2. In this case, CABA CABA means that the osmotic diluted; Cx even increased with increasing JVr. For x o maize grown hydroponically the increase tended to be barrier in the root is not completely selective for ABA ABA higher than for aeroponically cultivated roots. How- and rABA 0. In the other extreme, Cx 0 means ever, this was not signi®cant for all investigated plants that all ABA molecules are re¯ected at the barrier and ABA at all external ABA concentrations (Co ). According- rABA 1. Equation 4 demonstrates how the endoge- ly, the ``passive selectivity'' as expressed by the re¯ec- nous ¯ow of ABA would lower the overall re¯ection tion coecient (rABA) is not always signi®cantly lower coecient depending on the ABA concentration in the for hydroponically grown roots. Thus, one could medium as found experimentally. The other eect conclude that although the exodermis is substantially which tends to lower rABA is concentration polarisation reducing JABA it is not completely interrupting it, at in front of the osmotic barrier, a factor which is not least in the young roots studied here. For roots grown included in Eq. 4 (Freundl et al. 1998). It should aeroponically, ABA concentrations in the xylem are in become more important at higher ¯ow rates (pressure the same range as for those grown hydroponically. This dierences) which would tend to increase concentration is due to reduced water ¯ow (JVr) through roots with polarisation eects. This tendency has been found in Casparian bands. The ABA ¯ows (JABA) through the experiments. So, for two reasons, the values of aeroponically cultivated roots are reduced. Simulta- rABA given here are apparent ones and may underes- neously, Lpr was reduced by a factor of 2.2 up to 4.4 timate the true value which should be closer to unity. A depending on external ABA concentrations. For maize more appropriate way of determining rABA would be roots grown aeroponically, JABA increased with in- to use ABA as an osmotic solute and to measure creasing JVr, such that JABA went up even more than changes in root pressure in response to changes in the JVr. The eect may be caused by concentration osmotic pressure of the medium at JVr = 0 as exercised polarisation eects in front of the osmotic barrier in with roots for other solutes in the past (nutrient salts, 230 E. Freundl et al.: Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis sugars etc.; e.g. Steudle and Frensch 1989). How- Frensch J, Hsiao TC, Steudle E (1996) Water and solute transport ever, for ABA this method is not applicable because it along developing maize roots. Planta 198: 348±355 would require external concentrations in the millimolar Freundl E, Steudle E, Hartung W (1998) Water uptake by roots of maize and sun¯ower aects the radial transport of abscisic range. acid and the ABA concentration in the xylem. Planta 207: 8± In soil solution, ABA occurs at concentrations 19 ranging between 1 and 10 nM. This external ABA could Gerlach D (1984) Botanische Mikrotechnik. Thieme, Stuttgart be taken up by roots. It may play an important role for Glinka Z (1977) Eects of abscisic acid and of hydrostatic pressure root systems establishing equilibrium of ABA between gradients on water movement through excised sun¯ower roots. endogenous and external ABA (Hartung et al. 1996), Plant Physiol 59: 933±935 Glinka Z (1980) Abscisic acid promotes both volume ¯ow and ion and may initiate developmental processes of the root release to the xylem in sun¯ower roots. Plant Physiol 65: 537± system (MuÈ ller et al. 1989). The precise way by which 540 ABA is transported through the root apoplast is not Hartung W, Zhang J, Davies WJ (1994) Does abscisic acid play a known. The ABA could be translocated by solvent drag stress physiological role in maize plants growing in heavily across the endodermis. The permeability of the lipophil- compacted soil? J Exp Bot 45: 221±226 ic, apoplastic barriers exo- and endodermis is likely to be Hartung W, Sauter A, Turner NC, Fillery I, Heilmeier H (1996) Abscisic acid in soils: What is its function and which factors higher for ABA than for ions and other osmotically and mechanisms in¯uence its concentration? Plant Soil 184: active compounds. This may explain the dierent 105±110 transport behaviour of ABA compared to nutrient ions Hofer R-M (1996) Root hairs. In: Waisel Y, Eshel A, Kafka® U and other hydrophilic solutes. (eds) Plant roots: the hidden half. Dekker, New York, pp 111± In conclusion, the data of this paper show that the 126 exodermis ± besides the more eective endodermal layer Ludewig M, DoÈ ring K, Seifert H (1988) Abscisic acid and water transport in sun¯owers. Planta 175: 325±333 ± represents an additional barrier of radial ABA Markhart AH, Fiscus EL, Naylor AW, Kramer PJ (1979) Eect of transport through maize roots. Thus, both radial ABA abscisic acid on root hydraulic conductivity. Plant Physiol 64: and water ¯ows are similar. This explains why the 611±614 ABA concentration in the xylem driven by water ¯ow is Melchior W, Steudle E (1993) Water transport in onion (Allium not aected by the existence of an exodermis, while cepa L.) roots. Changes of axial and radial hydraulic diusional uptake into root tissue is retarded. Therefore, conductivity during root development. Plant Physiol 101: 1305±1315 extracellular barriers have a signi®cant in¯uence on MuÈ ller M, Deigele C, Ziegler H (1989) Hormonal interactions ABA uptake, transport and distribution, without bio- in the rhizosphere of maize (Zea mays L.) and their eects membranes and membrane-located mechanisms being on plant development. Z P¯anzenernaÈ hr Bodenk 152: 247± involved. 254 Perumalla CJ, Peterson CA, Enstone DE (1990) A survey of angiosperm species to detect hypodermal Casparian bands. I. We are grateful to the Deutsche Forschungsgemeinschaft [SFB 251; Roots with a uniserate hypodermis and epidermis. Bot J Linn Graduiertenkolleg (W.H.) and the Schwerpunktprogramm `Apo- soc 103: 93±112 plast' (E.S.)] for ®nancial support and to Prof. E.W. Weiler Peterson CA (1987) The exodermal Casparian band of onion roots (UniversitaÈ t Bochum) for generous supply with immunochemicals. blocks the apoplastic movement of sulphate ions. J Exp Bot 38: We thank PD Dr. Lukas Schreiber, Lehrstuhl fuÈ r Botanik II, 2068±2081 UniversitaÈ tWuÈ rzburg for giving us the opportunity of using Peterson CA (1988) Exodermal Casparian bands, their signi®cance the ¯uorescence microscope. The expert technical assistance of for ion uptake by roots. Physiol Plant 72: 204±208 B. Dierich, Lehrstuhl Botanik I, UniversitaÈ tWuÈ rzburg, and of Peterson CA (1989) Signi®cance of the exodermis in root function. Burkhard Stumpf, Lehrstuhl fuÈ r P¯anzenoÈ kologie, UniversitaÈ t In: Longhman BC, Gasparikova O, Kolek JS (eds) Develop- Bayreuth, is gratefully acknowledged. ments in plant and soil sciences, structural and functional aspects. Kluwer, Dordrecht, pp 35±40 Peterson RL (1992) Adaptations of root structure in relation to biotic and abiotic factors. Can J Bot 70: 661±675 References Schreiber L, Hartmann K, Skrabs M, Zeier J (1999) Apoplastic barriers in roots: chemical composition of endodermal and Brundrett MC, Enstone DE, Peterson CA (1988) A berberine- hypodermal walls. J Exp Bot 50: 1269±1280 aniline blue ¯uorescent staining procedure for suberin, lignin, Steudle (1994) Water transport across roots. Plant Soil 167: 79± and callose in plant tissue. Protoplasma 146: 133±142 90 Daeter W, Hartung W (1990) Compartmentation and transport of Steudle E, Frensch J (1989) Osmotic responses of maize roots: abscisic acid in mesophyll cells of intact leaves of Valerianella water and solute relations. Planta 177: 281±295 locusta. J Plant Physiol 36: 306±312 Steudle E, Henzler T (1995) Water channels in plants: do basic Damus M, Peterson RL, Enstone DE, Peterson CA (1997) concepts of water transport change? J Exp Bot 46: 1067± Modi®cations of cortical cell walls in roots of seedless vascular 1076 plants. Bot Acta 110: 190±195 Steudle E, Peterson CA (1998) How does water get through roots? Enstone DE, Peterson CA (1998) Eects of exposure to humid air J Exp Bot 49: 775±788 on epidermal viability and suberin deposition in maize (Zea Steudle E, Murrmann M, Peterson CA (1993) Transport of water mays L.) roots. Plant Cell Environ 21: 837±844 and solutes across maize roots modi®ed by puncturing the Fiscus EL (1975) The interaction between osmotic- and pressure- endodermis. Plant Physiol 103: 335±349 induced water¯ow in plants. Plant Physiol 55: 917±922 Tyerman SD, Bohnert HJ, Maurel C, Steudle E, Smith JAC Fiscus EL (1981) Eects of abscisic acid on the hydraulic (1999) Plant aquaporins: their molecular biology, biophysics conductance of and the total ion transport through Phaseolus and signi®cance for plant water relations. J Exp Bot 50: 1055± root systems. Plant Physiol 68: 169±174 1071 E. Freundl et al.: Apoplastic transport of abscisic acid through roots of maize: eect of the exodermis 231

Weiler EW (1986) Plant hormone immunoassay based on mono- Zhang J, Davies WJ (1989) Abscisic acid produced in dehydrating clonal and polyclonal antibodies. In: Linskens HF, Jackson JF roots may enable the plant to measure the water status of the (eds) Modern methods of plant analysis 4. Springer, Berlin, pp soil. Plant Cell Environ 12: 73±81 1±17 Zhu GL, Steudle E (1991) Water transport across maize roots. Zeier J, Schreiber L (1998) Comparative investigation of primary Plant Physiol 95: 305±315 and tertiary endodermal cell walls isolated from the roots of ®ve Zimmermann HM, Steudle E (1998) Apoplastic transport across monocotyledoneous species: chemical composition in relation young maize roots: eect of the exodermis. Planta 206: 7± to ®ne structure. Planta 206: 349±361 19