ORIGINAL ARTICLE

Ichirou Karahara Æ Atsuo Ikeda Æ Takanori KondoYuzo Uetake

Development of the Casparian strip in primary roots of maizeunder salt stress

Received: 17 June 2003 / Accepted: 17 December 2003 / Published online: 19 February 2004� Springer-Verlag 2004

Abstract The Casparian strip in the endodermis of vas-cular plant roots appears to play an important role inpreventing the influx of salts into the stele through theapoplast under salt stress. The effects of salinity on thedevelopment and morphology of the Casparian strip inprimary roots of maize (Zea mays L.) were studied.Compared to the controls, the strip matured closer tothe root tip with increase in the ambient concentrationof NaCl. During growth in 200 mM NaCl, the numberand the length of the endodermal cells in the regionbetween the root tip and the lowest position of theendodermal strip decreased, as did the apparent rate ofproduction of cells in single files of endodermal cells (therate of cell formation being equal to the rate at whichcells are lost from the meristem). The estimated timerequired for an individual cell to complete the formationof the strip after generation of the cell in the presence of200 mM NaCl was not very different from that requiredin controls. Thus, salinity did not substantially affect theactual process of formation of the strip in individualcells. The radial width of the Casparian strip, a mor-phological parameter that should be related to theeffectiveness of the strip as a barrier, increased in thepresence of 200 mM NaCl. The mean width of the lig-nified region was 0.92 lm in distilled water and 1.33 lmin 200 mM NaCl at the lowest position of the strip. Themean width of the strip relative to that of the radial wallat this position was significantly greater after growth inthe presence of 200 mM NaCl than in the controls,namely, 20.5% in distilled water and 33.9% in 200 mMNaCl. These observations suggest that the function ofthe strip is enhanced under salt stress.

Keywords Casparian strip (band) Æ Salinity (saltstress) Æ Endodermis Æ Lignin Æ Zea

Introduction

Roots are supposed to take up water and necessarysolutes from the soil, simultaneously avoiding the influxby diffusion of unnecessary solutes into the stele. Sinceroots manage to fulfill these contradictory requirements,it seems very probable that the internal tissues that areresponsible for these functions must be precisely regu-lated as the roots respond to the environment. Thesedevelopmental changes are poorly understood. In thecase of salinity stress, the only organ that is directlyexposed to excess salt is the root. Thus, it is important tounderstand the developmental and morphologicalchanges in the barrier structure that protects the rootsagainst an influx of elevated salt. Such an understandingshould help select or generate salt-tolerant variants foragricultural purposes.

The endodermis of virtually all vascular plants(Clarkson and Robards 1975; Haas and Carothers 1975)and the exodermis (Perumalla et al. 1990; Peterson andPerumalla 1990; Lehmann et al. 2000) of the roots ofmany angiosperms develop a Casparian strip that is lo-cated in the transverse and the radial walls of cells ofthese tissues. When a root segment in which the Casp-arian strip has already formed is immersed in hypotonicsolution, the plasmolysis figures of the endodermal cellsappear to be stretched across the cells in cross-sections(Bryant 1934; Enstone and Peterson 1997). This phe-nomenon, known as band plasmolysis, indicates thetight adhesion of the plasma membrane to the wall at theCasparian strip. The cell wall where the strip forms isimpregnated with hydrophobic materials, in particularlignin and suberin (Schreiber 1996; Zeier and Schreiber1997; Zeier et al. 1999a, 1999b; Schreiber et al. 1999).Since the strip is considered to play a pivotal role as thebarrier to apoplastic transport in roots, it seems very

I. Karahara (&) Æ A. Ikeda Æ T. Kondo Æ Y. UetakeDepartment of Biology, Faculty of Science,Toyama University, 930-8555 Toyama, JapanE-mail: karahara@sci.toyama-u.ac.jpTel.: +81-76-4456630Fax: +81-76-4456549

Planta (2004) 219: 41–47DOI 10.1007/s00425-004-1208-7

probable that the development and the morphology ofthe strip should be regulated by environmental factors(Karahara and Shibaoka 1994).

The relationship between the development of theCasparian strip and the growth of roots was studiedinitially in incense cedar (Wilcox 1962). It was subse-quently demonstrated that the distance from the root tipto the lowest position of the exodermal strip decreased inmaize roots under osmotic stress (Perumalla and Peter-son 1986) and in cotton roots under salt stress (Rein-hardt and Rost 1995). However, the mechanismresponsible for Casparian strip development closer tothe root tip remains to be fully characterized.

Reinhardt and Rost (1995) concluded that salinityaccelerates the development of the Casparian strip fromtheir observation that the distance from the root tip tothe lowest position of the strip decreased under saltstress. However, this distance depends on the length andnumber of cells in this region which, in turn, depends onthe rate of cell division in the meristem, the rate of cellelongation, and the time required for formation of thestrip in each individual cell. Therefore, one cannotsimply conclude that the development of the strip hasaccelerated exclusively from the observation that thestrip is formed closer to the root tip. In the presentstudy, whether the development of the strip is actuallyaccelerated in individual cells as a result of elevatedsalinity was ascertained.

Azaizeh and Steudle (1991) reported that salinity re-duces the hydraulic conductivity of roots, and Azaizehet al. (1992) proposed that the symplastic and transcel-lular pathway, or cell-to-cell path, might be responsiblefor the decrease in the hydraulic conductivity. However,hydrostatic experiments with maize have shown that theradial flow of water appears to be predominantly apo-plastic (Frensch and Steudle 1989). We cannot excludethe possibility that the apoplastic path of water is alsoaffected by NaCl (Azaizeh et al. 1992). One factor thatmight regulate hydraulic conductivity would be a changein the development and morphology of the Casparianstrip, which acts as a transport barrier. Differences in thehydraulic conductivity of maize roots caused by variousgrowth conditions were shown to be due to differences inthe development of the exodermal Casparian strip andthe suberin lamellae, and concomitant changes in levelsof suberin and lignin were detected (Zimmermann et al.2000). Monitoring morphological changes in the Casp-arian strip during its development, Barnabas and Pet-erson (1992) found that the width of the endodermalstrip increased as its distance from the root tip increasedin adventitious roots of onion. Moreover, natural radialwidening of the strip occurs during the development ofthe roots of vegetable marrow (Harrison-Murray andClarkson 1973). However, the cited studies did not in-clude any examination of roots under stress and it isclear that the morphology, for example the radial width,of the strip itself should also be related to the effective-ness of the barrier (Lux and Luxova 2001). Majormorphological characteristics of the Casparian strip

include the modification of the cell wall with lignin andsuberin and the tight adhesion of the plasma membraneto the cell wall. In the present study, the effect of saltstress on the morphology of the strip was examined, andthe relationship between anatomy and the function ofthe strip as a barrier to the transport of solutes isdiscussed.

Materials and methods

Plant materials

Kernels of maize (Zea mays L. ssp. sacchorata; Sakata Seed Co.,Yokohama, Japan) were soaked in distilled water for 24 h indarkness at 25�C and then sown on moist vermiculite (Asahi Ko-gyo Co., Okayama, Japan) that had been moistened with distilledwater (dH2O; control), or with a 100 mM or 200 mM solution ofNaCl in dH2O (100 ml liquid per 250 ml of vermiculite). Using thisbrand of vermiculite, it has been confirmed that the root growth ofthe controls (grown in dH2O) was typical of roots with a supply of1 mM calcium sulfate (data not shown). Thus, calcium was notadded to the vermiculite. Seedlings were grown in darkness at 25�Cfor 4–8 days. The ages of seedlings are expressed in terms of thenumber of days from the start of imbibition.

Observations of the Casparian strip and band plasmolysisby fluorescence and light microscopy

The lengths of primary roots that had grown for 8 days (countedfrom the start of imbibition) were measured and a frequency dis-tribution was constructed. When the class width of the frequencydistribution was set at 50 mm, the most frequent length classeswere 150–199 mm in the control, 100–149 mm in 100 mM NaCl,and 50–99 mm in 200 mM NaCl. The primary roots in the mostfrequent length class for each treatment were used in all experi-ments. Hand-cut sections were prepared from the roots at intervalsof 1 mm for observations of the Casparian strip in the endodermisand at intervals of 5 mm for observations of the strip in the exo-dermis, starting at the root tip. The term ‘‘root tip’’ is used todenote the apical end of the root in this study. Thus, the distanceincludes the root cap region. The sections were stained with 0.5%(w/v) berberine hydrochloride (Nacalai tesque, Kyoto, Japan),counterstained with 0.1% (w/v) aniline blue (Chroma Geselschaft,Darmstadt, Germany; Brundrett et al. 1988) and then observedunder a fluorescence microscope as described previously (Karaharaand Shibaoka 1998; Fig. 1). For fluorescence micrographs, 50-lm-thick frozen sections were prepared as described previously (Ka-rahara and Shibaoka 1992).

Band plasmolysis was examined as follows. Segments (2 mmlong) of roots were cut from the region near the lowest position ofthe Casparian strip, and immersed in phosphate-buffered saline(PBS; 10 mM NaH2PO4, 130 mM NaCl, pH 7.3) that contained3.7% (v/v) formalin and 1 M urea for 1 h at room temperature.Then, 50-lm cross-sections were prepared as described above.Light photomicrographs were taken with a digital camera (HC-300Z; Fuji Photo Film Co, Tokyo, Japan) fitted to the microscope.

Quantification of cells in longitudinal sections

Segments (2 mm long) of roots were sampled from the meanlocation of the lowest position of the Casparian strip in controlsand in roots grown in 200 mM NaCl (see Table 1). Segments werefixed for 12 h at 0�C in 50 mM sodium-cacodylate buffer (pH 7.2)that contained 2.5% glutaraldehyde. After washing in dH2O, seg-ments were post-fixed in 1% (w/v) osmium tetroxide in the samebuffer. The fixed segments were washed and dehydrated by passage

42

through a graded ethanol series and embedded in Spurr’s resin(Spurr 1969). Longitudinal semi-thin 4-lm sections were cut suchthat they included the center of the stele; sections were stained withan aqueous solution of 0.1% (w/v) sodium carbonate containing0.5% (w/v) toluidine blue O. The cells in two files of endodermalcells were counted on a longitudinal section. Total numbers of cellsin two files were averaged to give an estimate of cells per file. Sincethe part of the root that contained this region had been cut intosegments, the number of cells did not equal the actual number inone complete cell file but was equal to the total number of cells inthe segment. Since the apical organization of cells in maize root isof the closed type, it was possible to count the number of endo-dermal cells entirely as far as the initial cells.

Apparent rate of production of cells in an endodermal cell file

The time required for a cell to complete the formation of the stripwas estimated by dividing the number of cells in an endodermal cellfile in the region between the root tip and the lowest position of thestrip by the number of cells produced in an hour in an endodermalcell file, namely, the apparent rate of production of endodermalcells in a cell file. Even though the rate of production of cells in agiven file is not equal to the rate of cell division itself, it is pro-portional to the rate of cell division when this rate remains constantand it can be examined relatively easily in each file of cells.Moreover, in this study, it was necessary to know the rate ofproduction of endodermal cells specifically and not the general rateof cell division.

Since the growth rate of roots has been reported to increase forup to 3 days after the start of imbibition (Perumalla and Peterson1986), the apparent rate of production of cells was examined using4-day-old roots. Seedlings were grown in the presence or the ab-sence of 200 mM NaCl for 4 days. For determination of theapparent rate of production of cells, roots of the range within mean± 5 mm in length in the case of controls and of treatment with200 mM NaCl were used (see Results). Roots were picked clean ofvermiculite and marked with indelible black ink (Magic Ink,Tokyo, Japan) under dim green light at 30 mm, in the case ofcontrols, and at 14.5 mm, in the case of treatment with 200 mM

Fig. 1a–c The Casparian strip in maize (Zea mays) primary roots,as observed under a fluorescence microscope. Frozen cross-sectionswere cut from 8-day-old roots that had been grown in the presenceof 200 mM NaCl. a Section at a point 80 mm from the root tip.Exodermal Casparian strips were present but no endodermal stripcould be observed at this point because their staining was maskedby further modification of the wall. ex Exodermis, en endodermis.b Exodermal Casparian strip (arrowheads) at a point 55 mm fromthe root tip. c Endodermal Casparian strip (arrowheads) at a point5 mm from the root tip. Bars = 100 lm (a), 20 lm (b,c)

Table

1Effects

ofsalinityonthedistance

from

thelowestpositionoftheCasparianstripto

themaize(Z

eamays)

roottiponthenumber,length,andapparentrate

ofproductionof

endodermalcellslocatedin

thisregion,andontheestimatedtimerequired

forform

ationoftheendodermalstripin

individualcells.TheMann-W

hitney

U-testorthet-testwasusedto

determinewhether

theeff

ects

ofsalinitywerestatisticallysignificant

Parameter

DW

100mM

NaCl

200mM

NaCl

Statistics/Probability

Mean±

SE

nMean±

SE

nMean±

SE

nStatistics

P

Length

ofprimary

roots

examined

(mm)

173.4±5.2

8127.1±

3.7

878.2±

5.2

10

––

Distance

from

thelowestpositionoftheexodermalCasparianstripto

theroottip(m

m)

84.8±

4.0

871.9±

4.9

834.9±

4.0

10

––

Distance

from

thelowestpositionoftheendodermalCasparianstripto

theroottip(m

m)

12.9±

0.8

86.1±

0.7

83.2±

0.2

10

––

Number

ofcellslocatedbetweenthelowestpositionoftheendodermalCasparianstrip

andtheroottip

169±7

5–

–87±

46

U=

0=

0.0061

Length

ofcellslocatedbetweenthelowestpositionoftheendodermalCasparianstripand

theroottip(lm)

62.7±

2.1

1,694

––

37.7±

1.4

1029

t=8.701

<0.0001

Apparentrate

ofproductionofendodermalcellsin

onecellfile

(cells/h)

8.7±0.8

6–

–4.1±0.7

6U=

0.5

=0.0050

Estim

atedtimerequired

forform

ationoftheendodermalCasparianstripin

each

cell(h)

19.3

––

–21.2

––

–

43

NaCl, from the root tip. The distances were 54% of the mean totallength of the roots of 4-day-old seedlings from the root tip for eachtreatment, namely, at a site at which elongation of cells had alreadyceased. The roots were returned to the vermiculite for 7.5 h. Thenroots were again removed from the vermiculite and marked at thesame distance from the root tip for each treatment. The distancebetween the two marks represented the elongation of roots duringthe 7.5-h period. Segments (2 mm long) were cut from this region,fixed and embedded as described above. The endodermal cells inlongitudinal sections were quantified as described above. Thenumbers of endodermal cells in two files observed in each sectionwere added together, divided by two and divided by 7.5 h. Theresult corresponded to the number of endodermal cells produced in1 h in one cell file, namely, the apparent rate of production of cellsin an endodermal cell file.

Quantitative analysis of the radial width of the Casparian strip

Ultrathin sections were cut from plasmolyzed segments preparedfrom the lowest position of the strip of roots grown either in dH2Oor in the presence of 200 mM NaCl. Segments (2 mm long) weresampled from the region 12–14 mm above the root tip in controlsand 2–4 mm above the root tip in the case of treatment with200 mM NaCl. The mean location of the lowest position of theCasparian strip was located in this region. Each segment was im-mersed in a solution of 0.7 M mannitol in dH2O for 30 min, fixedand embedded as described above. Ultrathin cross-sections werestained with aqueous uranyl acetate, with lead citrate and with 1%KMnO4 for 1 min each. The lignified region of the strip wasidentified as the electron-dense region that was stained specificallywith KMnO4 (Hayat 1986). The region of the cell wall to which theplasma membrane adhered tightly on plasmolysis and the radialwidth of the lignified region were measured for all Casparian stripsin a given section. The radial width of the radial wall was defined asthe width of the region in which the two radial walls of neighboringendodermal cells appeared to be continuous. The experiment wasrepeated four times to confirm the reproducibility of the results.Since it was conceivable that the radial wall of endodermal cellshad also increased in width, the width of the radial wall wasmeasured and the widths of the lignified region and the tightlyadhering region of the plasma membrane relative to that of theradial wall were calculated, as percentages.

Results

Effects of salinity on the distance from the root tipto the lowest position of the Casparian strip

The distance from the root tip to the lowest position ofthe endodermal and the exodermal Casparian strip inroots of 8-day-old seedlings that had been grown in thepresence or the absence of NaCl was examined(Table 1). The distance from the root tip to the lowestposition of the strip decreased with increase in theconcentration of NaCl in both the endodermis and theexodermis.

Quantification of cells in endodermal cell filesbetween the root tip and the lowest positionof the Casparian strip

To ascertain why the distance from the root tip to thelowest position of the Casparian strip decreased under

salt stress, the numbers and the lengths of cells inendodermal cell files in the region between the root tipand the mean lowest position of endodermal Casparianstrips were determined. Both the number and the lengthof endodermal cells in this region decreased significantlyunder salt stress (Table 1).

If it is assumed that the rate of cell division remainedconstant during the time when the cells in this regionwere produced, the numbers of cells produced in 7.5 h inan endodermal cell file should be equal to the number ofcells produced per cell file at the root meristem in 7.5 h.The apparent rate of production of cells was examinedusing 4-day-old roots grown in the presence or the ab-sence of 200 mM NaCl. The roots of 4-day-old seedlingswere 55±3 mm (mean ± SE, n=67) long in the controlsand 26±2 mm (n=73) long in 200 mM NaCl. Theapparent rate of production of endodermal cells in onecell file (Table 1) was reduced significantly under saltstress. If it is assumed that the apparent rate of pro-duction of cells remained constant from 4 days after thestart of imbibition, the estimated time required for anindividual cell to complete the formation of the stripunder salt stress appears not very different from the timerequired in distilled water.

Effects of salinity on the morphology of the Casparianstrip

The width of plasmolyzed protoplasm in the radialdirection was larger in the salt-stressed than in thecontrol samples (Fig. 2). Major morphological charac-teristics of the Casparian strip were examined quanti-tatively to determine whether the width of the lignifiedregion and that of the tightly adhering region of theplasma membrane increased under salt stress. Thewidths of the lignified region and the tightly adheringregion of the plasma membrane were significantly largerunder salt stress than in the controls (Fig. 3, Table 2).The percentages were significantly larger under saltstress than in the controls (Table 2).

Discussion

Effects of salinity on the development and morphologyof the Casparian strip at the cellular level

In the primary roots of maize, the Casparian strip ma-tured closer to the root tip under conditions of saltstress. This result is consistent with results from cottonroots (Reinhardt and Rost 1995). It has been shown thatformation of the strip at positions closer to the root tipunder salt stress was due to a decrease in the number ofcells and in the lengths of cells in the endodermis be-tween the root tip and the lowest position of the strip.The age of a cell at the lowest position of the strip isequivalent to the time required for the cell to completethe formation of the strip since the cell itself had been

44

produced. It has been reported that the apparent rate ofproduction of epidermal cells decreases under salt stress(Kurth et al. 1986; Zidan et al. 1990). In this study, theestimated time that was required for a cell to completeformation of the strip under salt stress was not verydifferent from the time in distilled water or was onlyslightly longer. These observations indicate that salinitydid not have a major effect on the time-dependent for-mation of the strip itself in individual cells.

Previous reports that the endodermal Casparian stripincreases in radial width as it develops suggest that it isnecessary to compare the width of the strip among cellsin which the strip itself is at the same developmentalstage. At present, the developmental stage of the stripcan be specified only at the initial stage of its formation.

Therefore, the width was compared at the lowest posi-tion of the strip.

Functional significance of changes in Casparian stripdevelopment and morphology

The available evidence indicates that the root apicalzone is impermeable to ion movement (Luttge and Weigl1962; Enstone and Peterson 1992). On the other hand, inthe elongation zone of individual maize roots, namely,the region between the meristem and the lowest positionof the endodermal strip, the mRNA that encodes atonoplast aquaporin is expressed at high levels in the

Fig. 2a,b The endodermis in cross-sections prepared from plasmo-lyzed segments of maize roots grown under salt stress, as viewedunder a light microscope. The plasmolysed protoplasm (arrow-heads) appeared to be stretched across the cells. a Section at a point13 mm from the root tip in a control root. b Section at a point7 mm from the tip of a root grown in the presence of 200 mMNaCl. The plasmolyzed protoplasm appears to be wider in theradial direction than that of the control. Bars = 10 lm

Fig. 3a,b Effects of salinity on the ultrastructural morphology ofthe Casparian strip. Ultrathin cross-sections were prepared fromplasmolyzed segments of maize roots. The radial walls between twoneighboring endodermal cells are shown and the electron-denseportion of the wall, which was stained with KMnO4, correspondsto the strip. Brackets show the lignified region (L) and the tightlyadhering region of the plasma membrane (P) of the Casparianstrip, and the radial wall (R). a Section at a point 13 mm from theroot tip in a control root. b Section at a point 7 mm from the roottip in a root grown in the presence of 200 mM NaCl. The width ofthe radial wall was measured along the wall, even though thebrackets have been drawn as straight lines for convenience. Bars =1 lm

Table 2 Effects of salinity on the width of the lignified region of theendodermal Casparian strip and the tightly adhering region ofthe plasma membrane. The widths of the lignified region and thetightly adhering region of the plasma membrane relative to

the width of the radial wall (as percentages) are also shown. TheMann-Whitney U-test was used to determine whether the effects ofsalinity were statistically significant

Parameter Distilled water 200 mM NaCl Statistics/Probability

Mean ± SE n Mean ± SE n U P

Width of the lignified region (lm) 0.92±0.03 26 1.33±0.07 22 77 <0.0001Width of the tightly adhering region of the plasma membrane (lm) 0.85±0.02 50 1.16±0.05 44 386 <0.0001Width of the lignified region relative to the width of the radial wall (%) 20.5±0.7 26 33.9±1.5 22 26 <0.0001Width of the tightly adhering region of the plasmamembrane relative to the width of the radial wall (%)

18.7±0.4 50 29.5±1.0 44 51 <0.0001

45

endodermis and the pericycle (Barrieu et al. 1998). Thissuggests that water transport occurs in this region. Thedecrease in the distance of this region may reduce loss ofwater to the soil under salt stress.

The change in the width of the strip with the changein the environment raises an important possibility withregard to the function of the strip. Initially, the radialwidening of the strip might serve to reinforce the func-tion of the strip as an apoplastic transport barrier.However, for the reinforcement of the hydraulic resis-tance, at least the density of hydrophobic materials, suchas lignin and suberin, must not decrease. This issue re-mains to be examined in future studies. However, if it isassumed that the widening of the strip results in rein-forcement of its hydraulic resistance, some usefulworking hypotheses can be formulated, as follows.

If the Casparian strip itself actually allows limitedpassage of water and solutes, the radial width might berelated to the effectiveness of the strip as a transportbarrier since the length of the resistant pathway wouldbe increased. The available evidence indicates that thestrip is impermeable to ion movement (Baker 1971;Nagahashi et al. 1974; Robards and Robb 1972; Leh-mann et al. 2000). On the other hand, there is no directevidence that the strip is a complete barrier to watertransport. And there are reports that the apoplast is amajor pathway for the radial transport of water undersome conditions (Frensch and Steudle 1989; Petersonet al. 1993), indicating that water can pass through thestrip. The radial width of the lignified region and thetightly adhering region of the plasma membrane bothincreased under salt stress. Those changes would hinderthe passage of water and, thus, they would reinforce thefunction of the strip as the barrier to apoplastic trans-port.

In addition to the morphological parameters of thestrip, the density in the strip and the composition oflignin and suberin, the hydrophobic components of thestrip, should also be related to the effectiveness of thestrip as a barrier. These parameters should be analyzedin future studies by isolating strips (Karahara andShibaoka 1992; Zeier et al. 1999a, 1999b) from roots.

A correlation between the width of the radial walland that of the strip has been observed even in intactcells (Yokoyama and Karahara 2001). This suggests apre-determined relationship between the radial growthof wall and Casparian strip. However, in the presentstudy, any reproducible results indicative of the widen-ing of the radial wall under salt stress were not obtained(data not shown). The width of the strip increased by afactor greater than would be expected from the increasein length of the radial wall during growth (see Yokoy-ama and Karahara 2001); thus, some additional factormust determine growth of the strip. Since salinity ap-pears to be an environmental factor that regulates thewidth of the Casparian strip, we plan to study themechanisms that control how and when the width ofthe strip is determined during the course of developmentof endodermal cells.

Acknowledgements This work was supported by a Grant-in-Aid forScientific Research (no. 11740454) from the Ministry of Education,Science, Sports and Culture of Japan, by a Sasakawa ScientificResearch Grant from the Japan Science Society, and a grant for theimprovement of education from Toyama University to I.K. Theauthors are grateful to Professors M. Sugai, K. Masuda andS. Kamisaka of Toyama University for their helpful suggestionsand to Ms. H. Miyake for her skilled technical assistance.

References

Azaizeh H, Steudle E (1991) Effects of salinity on water transportof excised maize (Zea mays L.) roots. Plant Physiol 97:1136–1145

Azaizeh H, Gunse B, Steudle E (1992) Effects of NaCl and CaCl2on water transport across root cells of maize (Zea mays L.)seedlings. Plant Physiol 99:886–894

Baker DA (1971) Barriers to the radial diffusion of ions in maizeroots. Planta 98:285–293

Barnabas AD, Peterson CA (1992) Development of Casparianbands and suberin lamellae in the endodermis of onion roots.Can J Bot 70:2233–2237

Barrieu F, Chaumont F, Chrispeels M (1998) High expression ofthe tonoplast aquaporin ZmTIP1 in epidermal and conductingtissues of maize. Plant Physiol 117:1153–1163

Brundrett MC, Enstone DE, Peterson CA (1988) A berberine-aniline blue fluorescent staining procedure for suberin, lignin,and callose in plant tissue. Protoplasma 146:133–142

Bryant AE (1934) A demonstration of the connection of the pro-toplasts of the endodermal cells with the Casparian strips in theroots of barley. New Phytol 33:231

Clarkson DT, Robards AW (1975) The endodermis, its structuraldevelopment and physiological role. In: Torrey JG, ClarksonDT (eds) The development and function of roots. AcademicPress, London, pp 415–436

Enstone DE, Peterson CA (1992) The apoplastic permeability ofroot apices. Can J Bot 70:1502–1512

Enstone DE, Peterson CA (1997) Suberin deposition and bandplasmolysis in the maize (Zea mays L.) root exodermis. CanJ Bot 75:1188–1199

Frensch J, Steudle E (1989) Axial and radial hydraulic resistance toroots of maize (Zea mays L.) Plant Physiol 91:719–726

Haas DL, Carothers ZB (1975) Some ultrastructural observationsof endodermal cell development in Zea mays roots. Am J Bot62:336–348

Harrison-Murray RS, Clarkson DT (1973) Relationships betweenstructural development and the absorption of ions by the rootsystem of Cucurbita pepo. Planta 114:1–16

Hayat MA (1986) Basic techniques for transmission electronmicroscopy. Academic press, Orlando, FL

Karahara I, Shibaoka H (1992) Isolation of Casparian strips frompea roots. Plant Cell Physiol 33:555–561

Karahara I, Shibaoka H (1994) The Casparian strip in peaepicotyls: effects of light on its development. Planta 192:269–275

Karahara I, Shibaoka H (1998) Effects of Brefeldin A on thedevelopment of the Casparian strip in pea epicotyls. Protopl-asma 203:58–64

Kurth E, Cramer GR, Laeuchli A, Epstein E (1986) Effects of NaCland CaCl2 on cell enlargement and cell production in cottonroots. Plant Physiol 82:1102–1106

Lehmann H, Stelzer R, Holzamer S, Kunz U, Gierth M (2000)Analytical electron microscopical investigations on the apo-plastic pathways of lanthanum transport in barley roots. Planta211:816–822

Luttge U, Weigl J (1962) Mikroautoradiographische Untersuch-ungen der Aufnahme und des Transportes von 35SO4

2) und45Ca2+ in Keimwurzeln von Zea mays und Pisum sativum L.Planta 58:113–126

Lux A, Luxova M (2001) Secondary dilatation growth in the rootendodermis. In: Gasparikova O, Ciamporova M, Mistrik I,

46

Baluska F (eds) Recent advances of plant root structure andfunction. Kluwer, Dordrecht, pp 31–37

Nagahashi G, Thompson WW, Leonard RT (1974) The Casparianstrip as a barrier to the movement of lanthanum in maize roots.Science 183:670–671

Perumalla CJ, Peterson CA (1986) Deposition of Casparian bandsand suberin lamellae in the exodermis and endodermis of youngmaize and onion roots. Can J Bot 38:1872–1878

Perumalla CJ, Peterson CA, Enstone DE (1990) A survey ofangiosperm species to detect hypodermal Casparian bands. I.Roots with a uniseriate hypodermis and epidermis. Bot J LinnSoc 103:93–112

Peterson CA, Perumalla CJ (1990) A survey of angiosperm speciesto detect hypodermal Casparian bands. II. Roots with a mul-tiseriate hypodermis or epidermis. Bot J Linn Soc 103:113–125

Peterson CA, Murrmann M, Steudle E (1993) Location of themajor barriers to water and ion movement in young roots ofZea mays L. Planta 190:127–136

Reinhardt DH, Rost TL (1995) Salinity accelerates endodermaldevelopment and induces an exodermis in cotton seedling roots.Environ Exp Bot 35:563–571

Robards AW, Robb ME (1972) Uptake and binding of uranyl ionsby barley roots. Science 178:980–982

Schreiber L (1996) The Casparian strip of Clivia miniata Reg. roots:evidence for lignin. Planta 199:596–601

Schreiber L, Hartmann K, Skrabs M, Zeier J (1999) Apoplasticbarriers in roots: composition of endodermal and hypodermalcell walls. J Exp Bot 50:1267–1280

Spurr AR (1969) A low viscosity epoxy resin embedding mediumfor electron microscopy. J Ultrastruct Res 26:31–43

Wilcox H (1962) Growth studies of the root of incense cedar,Libocedrus decurrens I. The origin and development of primarytissues. Am J Bot 49:221–236

Yokoyama M, Karahara I (2001) Radial widening of the Caspar-ian strip follows induced radial expansion of endodermal cells.Planta 213:474–477

Zeier J, Schreiber L (1997) Chemical composition of hypodermaland endodermal cell walls and xylem vessels isolated fromClivia miniata. Plant Physiol 113:1223–1231

Zeier J, Goll A, Yokoyama M, Karahara I, Schreiber L (1999a)Structure and chemical composition of endodermal and rhizo-dermal/hypodermal walls of several species. Plant Cell Environ22:271–279

Zeier J, Ruel K, Ryser U, Schreiber L (1999b) Chemical analysisand immunolocalisation of lignin and suberin in endodermaland hypodermal/rhizodermal cell walls of developing maize(Zea mays L.) primary roots. Planta 209:1–12

Zidan I, Azaizeh H, Neumann PM (1990) Does salinity reducegrowth in maize root epidermal cells by inhibiting their capacityfor cell wall acidification? Plant Physiol 93:7–11

Zimmermann HM, Hartmann K, Schreiber L, Steudle E (2000)Chemical composition of apoplastic transport barriers in rela-tion to radial hydraulic conductivity of maize roots (Zea maysL.). Planta 210:302–311

47