The microtubule cytoskeleton does not integrate auxin transport and gravitropism in maize roots.

K. H. Hasenstein^a*, E.B. Blancaflor^a,b and J.S Lee^c

^aBiology Department, University of Southwestern Louisiana, P.O. Box 42451, Lafayette, LA 70504-2451

^bcurrent address: Biology Department, 208 Mueller Laboratory, The Pennsylvania State University, University Park, PA 16802, USA

^cBiology Department, Ehwa Womans University, Seoul, Korea

^*Corresponding author: Biology Department, University of Southwestern Louisiana, Lafayette, LA 70504-2451, USA (e-mail: hasenstein@usl.edu)

The Cholodny-Went hypothesis of gravitropism suggests that the graviresponse is controlled by the distribution of auxin. However, the mechanism of auxin transport during the graviresponse of roots is still unresolved. To determine whether the microtubule (MT) cytoskeleton is participating in auxin transport, the cytoskeleton was examined and the movement of ³H-IAA measured in intact and excised taxol, oryzalin, and naphthylphthalamic acid (NPA) treated roots of Zea mays cv. Merit. Taxol and oryzalin did not inhibit the graviresponse of roots but the auxin transport inhibitor NPA greatly inhibited both auxin transport and graviresponse. NPA had no effect on MT organization in vertical roots but caused MT reorientation in horizontally placed roots. Regardless of treatment, the organization of MTs in intact roots differed from that in root segments. The MT inhibitors taxol and oryzalin, had opposite effects on the MTs, namely depolymerization (oryzalin) and stabilization and thickening (taxol) but both treatments caused swelling of the roots. The data indicate that the MT cytoskeleton does not directly interfere with auxin transport or auxin mediated growth responses in maize roots.

Abbreviations - NPA: naphthylphthalamic acid, MT: microtubules, TIBA: triiodobenzoic acid

Introduction

The Cholodony-Went hypothesis for gravitropism suggests that differential elongation between opposite flanks of graviresponding organs is linked to auxin transport, concentration, and/or sensitivity of the responding tissue. When a root is positioned horizontally, auxin redistributes to the lower side of the root. This redistribution promotes growth on the upper side and inhibits growth on the lower side of roots, resulting in downward curvature (Hasenstein and Evans 1988, Ishikawa et al. 1991, Evans 1991). Auxin transport is clearly involved in root gravitropism since auxin transport inhibitors such as naphthylphthalamic acid (NPA) and triiodobenzoic acid (TIBA) inhibit root curvature (Lee et al. 1984, Evans et al. 1992, Muday and Haworth 1994) and the differential induction of an auxin-responsive promoter (Li et al. 1991). Despite the strong correlation between auxin transport and gravitropism, the regulation of auxin transport during graviresponse remains unclear.

In plant cells, the cortical MTs beneath the plasma membrane have been proposed to control the deposition of cellulose microfibrils. The orientation of cellulose microfibrils in turn determines the extent and direction of cell expansion (Giddings and Staehelin 1991). Therefore, the orientation of MTs may be a useful indicator for the growth status of cells within a tissue. Studies on plant tropisms have shown that MT organization and cell growth are strongly correlated (Nick et al. 1990, Blancaflor and Hasenstein 1993). For example, in vertically-growing roots, MTs in the elongation zone are transverse to the longitudinal axis of the root. Positioning the root horizontally (gravistimulation) results in downward curvature, and MTs of the outer cortical cells on the slower growing, concave side reorient longitudinally (Blancaflor and Hasenstein 1993). Both graviresponse and MTs reorientation depend upon the auxin content of roots (Blancaflor and Hasenstein 1995). The pattern of MT reorientation during the graviresponse of both shoots (Nick et al. 1990) and roots (Blancaflor and Hasentein 1995) suggests an interaction between the MT cytoskeleton and the auxin transport system.

The association of the actin cytoskeleton to the NPA binding protein (Cox and Muday 1994) implicates the actin microfilaments in the regulation of auxin transport. However, the failure of actin disrupting compounds to inhibit root curvature (Blancaflor and Hasenstein 1997, Staves et al. 1997), suggests that the regulation of auxin transport during gravitropism is independent of the actin cytoskeleton. Although disruption of microfilaments with cytochalasins inhibited auxin transport (Butler et al. 1998), the involvement of MTs in auxin transport has not yet been studied.

The objective of this study was to determine whether disrupting the MT cytoskeleton affects the transport of exogenously applied auxin in maize roots. We also report on the effect of MT disrupting compounds on the graviresponse of roots. The data show that disruption of MTs does not inhibit the polar transport of exogenously applied auxin and confirm that an intact and dynamic MT cytoskeleton is not required for gravitropic curvature in maize roots.

Materials and methods

Plant material

Caryopses of corn (Zea mays) cv. Merit were soaked in deionized water for 10 h then planted between wet paper towels in opaque plastic trays. The trays were positioned vertically in a growth chamber at 24C under 18 h light (Sylvania Cool White 50 W m^-2) and 6 h dark cycle. After 3 days, seedlings with 25-30 mm long roots were selected and used for the experiments.

Drug treatments

Stock solutions (20 mM) of taxol (Calbiochem, La Jolla, CA), oryzalin and naphthylphthalamic acid (NPA) (Chem Services, West Chester, PA) where prepared in dimethyl sulfoxide (DMSO). Working solutions of 20 µM taxol, 5 µM oryzalin and 5 µM NPA were prepared by adding the appropriate volume of the stock solution to 5 mM MES/Tris buffer, pH 6.5. Seedlings with straight roots were transferred to 1-ml microfuge tubes containing working solutions (1 root per tube) and the terminal 10 mm of the roots were immersed for 3 h. Roots immersed in buffer with the corresponding concentration of DMSO were used as controls.

Growth measurement and gravistimulation

After pretreatment with the drugs, seedlings were transferred to 9-cm plastic petri dishes lined with 3 layers of wet filter paper. Seedlings were gravistimulated by rotating the petri dishes by 90 degrees. Curvature was measured using a video digitizer system (Hasenstein 1991) which continuously acquired and recorded in 2 min intervals the angular orientation of the root tip. Curvature data were obtained from at least 12 roots per treatment.

Auxin transport experiments

Auxin transport was measured in both root segments and intact roots. The terminal 10 mm of < 30 mm long maize roots were immersed in vials containing 20 µM taxol, 5 µM oryzalin or 5 µM NPA. Alternatively, 5 mm long apical segments were excised, transferred to solutions containing the drugs (same concentration) and agitated on an orbital shaker (30 rpm). After 3 h the segments were mounted vertically by inserting either their tips or their basal cut surfaces into a slab of agar (1.5% [w/v], 20×5×2 mm) which served as the receiver block for transported auxin. Acropetal transport was measured downward and basipetal transport upward, corresponding to the normal, vertical orientation of a root. A donor block ([1.5 mm]³) containing ³H-IAA (3×10^-7 M, 999 TBq mol^-1, Amersham, Arlington Heights, IL) was applied either to the basal cut surface or to the tip of each segment for acropetal and basipetal transport, respectively. The time course of radioactivity transport into the receiver at the opposite end was measured by transferring the donors and tissue segments to new blocks in hourly intervals.

Auxin transport in intact roots was measured by applying donor blocks (same dimensions and IAA concentration as described above) to groups of five seedlings either at the root cap or 10 mm from the root tip for basipetal and acropetal transport, respectively. Receiver blocks were applied either at the elongation zone or the root tip. After 1 to 4 h transport periods the respective agar blocks were combined and the roots dissected into three segments, namely 2 mm tissue adjacent to the donor block, 5 mm transport tissue and >5 mm receiver tissue. The corresponding parts of sets of 5 roots were combined, suspended in scintillation cocktail and stored over night.

Radioactivity in root segments, receiver and donor blocks was measured by liquid scintillation counting (Packard 1900-CA). All transport experiments were repeated five times with three replicates each. Transport data were normalized based upon uptake since the area of donor block-tissue interfaces was different for the root cap and the root proper.

Immunofluorescence

Roots treated with taxol, oryzalin, and NPA were fixed in their experimental position and processed for microscopy as previously described (Blancaflor and Hasenstein 1993). Briefly, the terminal 6 mm of the roots were excised and fixed in 4% [w/v] formaldehyde in PHEMD buffer (see Blancaflor and Hasenstein 1993). Longitudinal Vibratome sections were treated with an enzyme solution (1% [w/v] cellulase, 0.5% [w/v] pectolyase) for 15 min and incubated in 0.1% [w/v] Tritron X-100 for 30 min. The primary antibody (monoclonal rat anti-yeast alpha-tubulin (clone YOL1/34; Accurate Chemical and Scientific Corp., Westbury, NY, 1:100) was visualized with DTAF-labeled goat anti-rat IgG (Accurate Chemicals, 1:200). Sections were mounted in 20% [w/v] Mowiol 4-88 (Calbiochem) in PBS, pH 8.5 containing 0.1% [w/v] phenylenediamine (Sigma) and examined with a scanning confocal microscope (MRC-600; BioRad, Richmond, CA).

Results

Microtubule organization in vertically growing intact roots treated with taxol and oryzalin

Confocal microscopy verified the action of oryzalin and taxol on the MT cytoskeleton of maize root sections. MTs in the elongation zone of vertically-grown control roots were oriented perpendicular to the longitudinal axis of the cell (Fig. 1A). Incubation for 3 h in taxol had no effect on the orientation but resulted in an increase in the density of MTs. Taxol-treated roots showed brighter fluorescence and reduced space between MT strands (Fig. 1B). In contrast, cells in the elongation zone of roots treated for 3 h with oryzalin were devoid of MTs (Fig. 1C). In the meristematic region of control roots, spindle MTs and phragmoplasts were numerous and indicative of active cell division in this region (Fig. 1D). Taxol treatment resulted in the appearance of abnormal spindle structures. Some spindles had more than 2 poles, while others had a brightly staining central region with highly flourescent strands radiating from the center (Fig. 1E). No mitotic spindles were present in roots treated with oryzalin and weak fluorescence, probably due to diffusely distributed tubulin dimers, characterized the cytoplasm (Fig. 1F).

Fig 1. Arrangement of cortical microtubules in vertically-grown maize roots after incubation for 3 h in 20 µM taxol and 5 µM oryzalin. The root tip is toward the bottom of the page for all images. Within the elongation zone (3-4 mm from tip), microtubules in outer cortical cells of control roots are oriented perpendicular to the longitudinal axis of the root (A). After treatment with taxol, the orientation of microtubules remained perpendicular to the longitudinal axis of the root but microtubule density increased (B). Treatment with oryzalin resulted in depolymerization of microtubules (C). In the meristematic region of control roots, spindle microtubules (arrows) and phragmoplast (arrowhead) were common (D). Taxol treatment resulted in the formation of multipolar spindles (arrow, E) and spindles with a radial array of microtubules originating from a brightly staining region (arrowhead, E). Oryzalin also caused depolymerization of microtubules in the meristematic zone as shown by the absence of mitotic spindles and diffuse cytoplasmic fluorescence (F). Bar: 25 µm.

Roots that were treated with taxol or oryzalin for 3 h and were then allowed to grow vertically developed swelling within 48 h. Taxol-treated roots showed maximum swelling along the elongation zone while expansion due to oryzalin was closer to the root tip but excluded the root cap (Fig. 2). After taxol treatment the zone of oblique MTs moved from the maturation zone toward the tip of the root. Immediately after removal from taxol, MTs of cortical cells located 6 mm from the root tip were all oblique or longitudinal in orientation and similar to control roots but with greater fluorescence. After 12 h some cortical cells exhibited oblique MTs at 4 mm from the tip, and after 24 h oblique and dense MTs appeared less than 3 mm from the root tip (data not shown).

Fig. 2. Morphology of vertically-grown roots treated for 3 h with 20 µM taxol (A) and 5 µM oryzalin (B) 48 h after treatment and controls (C). Swelling in taxol-treated roots occurred along the elongation zone while oryzalin caused more extensive expansion in the apical region. Bar: 1 mm.

The swelling of the elongation zone 48 h after taxol treatment was due to some unusual cellular morphology. Some cells showed highly aligned MTs transverse to the longitudinal axis of the root (data not shown) while some cells exhibited anomalous shapes with slightly oblique but highly bundled MTs (Fig. 3A). Unlike cortical cells, vascular parenchyma cells in the stele region retained their regular shape. However, these cells also showed dense MTs of typically oblique but irregular orientations (Fig. 3B). Similar to the swelling induced by taxol, the enlarging of oryzalin-treated roots was due to changes in the shape of cortical cells. Cortical cells closer to the root tip were devoid of MTs and assumed a spherical shape (Fig. 3C). Epidermal cells showed fragmented MTs but retained the columnar shape (Fig. 3D).

Fig. 3. Orientation of microtubules in the elongation zone of vertically grown roots 48 h after application of taxol and oryzalin. By 48 h, most cells in the elongation zone (> 2 mm from the tip) were irregularly shaped and dense arrays of obliquely oriented and bundled microtubules were observed (A). Vascular parenchyma cells also contained dense microtubules showing different orientations but rectangular shape (B). Oryzalin-treated roots exhibited cortical cells devoid of microtubules with a spherical shape (C); while epidermal cells close to the root tip remained columnar in shape and showed fragmented MTs (D). Bar: 25 µm.

Microtubule organization in root segments

Since auxin transport was measured in root segments, and cutting may have affected the cellular organization, possibly due to a wound response (Hush et al. 1991) that could interfere with the interpretation of our results, we studied the MT organization in drug-treated root segments. The organization of cortical MTs in excised root segments was different from intact roots. MTs in the basal elongation zone (>3 mm from the root tip) of control roots were fragmented and predominantly oblique in orientation (Fig. 4A). Segments exposed to NPA also showed some oblique and fragmented but mostly longitudinal MTs along the basal elongation zone (Fig. 4B). MTs in the distal elongation zone remained intact and transversely oriented (Fig. 4C). In taxol-treated maize root segments, cortical MTs in elongating cells located 2 mm from the tip were transverse to the longitudinal cell axis (Fig. 4D). Like controls and NPA-treated root segments, MTs in the basal elongation zone (> 3 mm from the root tip) of taxol-treated root segments were obliquely oriented. However, MTs were dense and unlike control or NPA-treated root segments, MTs were not fragmented (Fig. 4E). Similar to its effect on intact roots, oryzalin caused the depolymerization of MTs in root segments (Fig. 4F).

Fig. 4. Organization of microtubules in excised segments of maize roots. The root tip is toward the bottom in all images. Microtubules in the elongation zone (>3mm from the root tip) of control roots were highly fragmented and predominantly oblique in orientation (A), similar to NPA treated tissue (B). Microtubules in the distal elongation zone (ca 2 mm from the root tip) of control segments remained intact and transversely oriented (C). In taxol-treated root segments, cortical microtubules in elongating cells located 2 mm from the tip were transverse to the longitudinal cell axis (D). Taxol-treated segments 3 mm from the root tip showed obliquely oriented (arrow, E) or longitudinal (arrow heads, E) microtubules but in contrast to controls and NPA-treated roots microtubules were not fragmented (E). Root segments incubated in oryzalin showed background fluorescence indicative of depolymerized MTs (F). Bars: 25 µm.

Microtubule orientation and graviresponse in roots pretreated with taxol, oryzalin or NPA

In control roots, MTs in cells along the outer cortex of the lower flank reoriented longitudinally within 2 h after gravistimulation (Fig. 5A) and the orientation of MTs on the upper side remained perpendicular, as in vertical roots (data not shown). In contrast, MTs in taxol-treated roots remained completely transverse and very dense in both the inner and outer cortex on either side of the root (Fig. 5B). Interestingly, when 2 h-gravistimulated roots (which showed reorientation of MTs) were treated with 20 µM taxol for 1 h without changing their position, the MTs reverted back to the original, transverse orientation (Fig. 5C). Roots pretreated with oryzalin remained devoid of MTs after 2 h of gravistimulation (data not shown).

In the meristematic region of NPA treated roots (Fig. 5D), the appearance of mitotic spindles, pre-prophase bands and cortical MTs was similar to controls (see Fig. 1D). Cortical MTs in cells of the elongation zone (2 to 4 mm from the root tip) also remained transverse to the longitudinal axis of the root (Fig. 5E) and similar to controls (see Fig. 1A). However, reorienting NPA-pretreated roots horizontally resulted in the reorganization of cortical MTs on the upper (Fig. 5F) and lower (Fig. 5G) flank of the elongation zone after 2 h. MTs in the stele region, 2.5-3 mm from the tip, also became randomly arranged (data not shown).

Fig. 5. Microtubules along the lower side of the elongation zone in 2 h gravistimulated roots. The root tip in A, B, C, F and G is to the left and in D and E to the bottom. The gravity vector in all pictures is toward the bottom of the page. (A). Microtubules in control roots became parallel to the long cell axis. (B) Microtubules in roots pretreated for 3 h with 20 µM taxol before gravistimulation remained transverse and very dense throughout the cortex. (C) Taxol treatment of roots after they had been positioned horizontally for 2 h caused the microtubules to revert back to the original transverse orientation. Organization of microtubules in vertical (D, E) and 2 h gravistimulated roots (F, G) after pretreatment for 3 h in 5 µM NPA. Microtubules in the meristematic region show preprophase bands (arrow) mitotic spindles (arrowhead) and cortical microtubules similar to controls(D). Cortical cells in the elongation zone also exhibit transverse microtubules (E). Two h after gravistimulation, microtubules in the outer cortical cells of the elongation zone become randomly oriented on the upper (F) and lower flank (G). Bars: 25 µm (A-C, E-G); 10 µm (D).

Despite their effects on the MTs, neither taxol nor oryzalin inhibited the graviresponse of roots. However, in taxol-treated roots, the rate of curvature in the first 60 min after gravistimulation was faster than in controls, but the final angle of curvature was similar. During the initial phase of gravicurvature, oryzalin-treated roots showed no difference compared with controls but curvature stopped about 70 - 80 min after reorientation, resulting in reduced curvature compared with control and taxol-treated roots. NPA pretreated roots did not curve during the initial 2 h of gravistimulation (Fig. 6A). Taxol, oryzalin and NPA treatment resulted in a reduction of root growth rate. In contrast to taxol and NPA treated roots, the growth rate of oryzalin-treated roots dropped sharply 50 min after gravistimulation and after 2 h, the growth rate of oryzalin pretreated roots was reduced to about 0.3 mm h^-1 (Fig. 6B). NPA-treatment blocked the root graviresponse in a concentration dependent manner (Fig. 6C).

Fig. 6. Curvature and growth rate of maize roots after 3-h pretreatment with taxol, oryzalin and NPA. The final angle of curvature of taxol pretreated roots () and controls () were identical. However, the rate of curvature was initially faster in taxol-treated roots. Oryzalin-treated roots () curved but the final angle of curvature was significantly less than in control and taxol-treated roots. However, propizamide (+) did not inhibit the graviresponse. Curvature in NPA-treated roots () was completely inhibited (A). The growth rate of vertical roots pretreated with the above substances (same symbols) showed considerable variability. The growth rate of taxol and NPA pretreated roots was constant within the time period tested, but the growth rate of oryzalin-treated roots declined significantly after 80 min (B). The inhibition of graviresponse by NPA (concentrations shown as neg. log) is time and concentration dependent (C). Data are means ± SE, n = 8 to 10.

Effects of taxol, oryzalin and NPA on IAA transport

The effect of anti-MT drugs on longitudinal auxin transport was weak (Fig. 7.) In all treatments, the transport of ³H-IAA was preferentially in the basipetal direction (Fig. 7A). Basipetal transport intensity was about 1.5-fold greater than in the acropetal direction (Table 1). NPA-treatment strongly inhibited while oryzalin treatment slightly promoted ³H-IAA transport in basipetal direction. Acropetal transport was not affected by any of the treatments. The basal transport velocity for IAA was estimated to be > 3 mm hr^-1 and did not differ between treatments. Acropetal transport velocities could not be reliably determined, because of delayed uptake into the stele (the presumptive acropetal transport pathway).

Table 1: Transport characteristics of ³H-IAA through intact roots of Zea mays, cv Merit. The measured transport equations (= regression lines of data in fig. 7A, with Q= quantity of transported radioactivity, t= time). The transport intensities (coefficients of t) were not statistically different from one another, except for NPA and basipetal transport.

The uptake of radioactivity was similar for all roots (and segments, data not shown), increasing from 20 to 40% after one hour to over 70% after 4 hours for acropetal transport. Uptake for basipetal transport started at a higher initial rate but leveled off during the last two hours (Fig. 7B). The similarity of the rate of uptake for both acropetal and basipetal transport and the decline of uptake after 2 hours for basipetal transport despite continuing higher rates of transport indicate that the differences in transported IAA are not due to differences in uptake.

Fig. 7. Time course of basipetal (dashed) and acropetal (solid lines) transport of ³H-IAA through intact root tips pretreated for 3 h with oryzalin (), taxol (), NPA () or untreated controls () (A). Uptake of radioactivity was similar regardless of transport direction (site of uptake) or for the different treatments (B), same symbols as in (A).

Discussion

The role of the MT cytoskeleton in root graviresponse and auxin transport was investigated using drugs with opposite effects on MTs. Biochemical studies have shown that taxol, a plant alkaloid which stabilizes MTs lowers the critical concentration of monomeric tubulin (Bokros et al. 1993, Falconer and Seagull 1985) while oryzalin causes MT depolymerization by binding to tubulin dimers and blocking their incorporation into polymers (Morejohn et al. 1991). Three hours incubation in 5 µM oryzalin depolymerized MTs in epidermal and cortical cells of the meristem, elongation and maturation zone and caused extensive fragmentation of MTs in the stele and rhizodermal tissue. The changing status of MTs translated into conspicuous swelling along the apex of the root after 24 h and was similar to the effects observed in roots of Arabidopsis (Baskin et al. 1994) and several other plant species (Cleary and Hardham 1988). These data indicate that an intact and modifiable MT cytoskeleton is necessary for normal, anisotropic growth in root cells.

Despite the depolymerizing effect of oryzalin on MTs in all examined areas of the root, graviresponse was still possible. However, curvature in oryzalin-treated roots stopped about 80 min after horizontal orientation leading to a lower angle of curvature as compared with controls. The cessation of curvature in oryzalin-pretreated roots coincided with strong reduction in the growth rate (compare Fig. 6A and B). However, the reduced curvature may be an oryzalin-specific response since experiments with propizamide, another MT-depolymerizing herbicide, had similar kinetics and final angle of curvature as controls despite complete depolymerization of MTs (Fig. 6).

Although the data show that an intact MT cytoskeleton appears not to be required for gravitropic curvature in maize roots (see also Baluška et al. 1996a), maize coleoptiles treated with ethyl-N-phenylcarbamate and propizamide showed MT depolymerization and inhibited gravitropic bending (Nick et al. 1991) and auxin transport in rice shoots (Nick et al. 1997). The contrasting effect of MT depolymerization on root and shoot gravitropism and auxin transport is unclear but suggests either that roots and shoots have different cellular mechanisms for gravity response or they differ in their sensitivity to MT disruption.

The reorientation of cortical MTs along the outer cortical cells of the lower side accompanies the gravitropic curvature of maize roots (Blancaflor and Hasenstein 1993). The application of taxol prevented MT reorientation in the elongation zone (see Fig. 5). However, gravitropic curvature was initially faster than in controls. Therefore, MT reorientation or instability is not essential for the graviresponse of maize roots.

Not only did taxol prevent MTs on the lower side from reorienting, it also restored transverse MTs that had been reoriented horizontally for 2 h (Fig. 5C). However, taxol did not interfere with the oblique and longitudinal orientations of MTs in the maturation zone (data not shown) a characteristic feature of intact roots, that typically occurs 5 to 6 mm from the root tip (Blancaflor and Hasentein 1993) indicating that the MT reorientation that accompanies normal cell differentiation and maturation is different from the reorientation resulting from graviresponse. The observation that taxol can also prevent auxin-induced MT reorientation (Baluška et al. 1996b; Blancaflor and Hasenstein, unpublished data) suggests that the MT reorienting mechanism due to auxin is similar to that of the graviresponse and can be used as indirect evidence for the involvement of IAA in the gravitropic response of roots. It appears that the factors responsible for maintaining MTs in transverse orientation are not modified in response to either auxin or gravistimulation. The MT orienting factors that are not affected by auxin or graviresponse could include components of the cell wall such as the cellulose microfibrils themselves (Williamson 1991, Fisher and Cyr 1998) or transmembrane links to the cell wall (Shibaoka 1994), possibly in the form of integrins (Katembe et al. 1997). Aside from elevated auxin concentration (Blancaflor and Hasenstein 1995), the reorientation of MTs during the graviresponse could be a consequence of changes in internal tissue stresses as a result of bending (Hush and Overall 1991, Zandomeni and Schopfer 1994; Baluška et al. 1996a).

Another interesting observation with taxol-treated roots is the induction of radial expansion in maize roots particularly along the elongation zone. Taxol induced the appearance of abnormal cell morphology which was very prominent after 48 h. The aberrant cell shapes were observed even in cells with an overall transverse alignment of MTs. A similar phenomenon was observed in the Arabidopsis root mutant cobra-1, which shows expansion in the radial direction without changing microtubule orientation (Hauser et al. 1995). Taxol also induced radial expansion of cells in the elongation zone of Arabidopsis roots (Baskin et al. 1994) and cultured cells of Vicia (Weederburg and Seagull 1988) without changing the alignment of microtubules. Therefore, the role of cortical microtubules in shifting the growth direction of cells maybe more complex than previously proposed (Hush et al. 1990).

Despite regular MT orientation, there was extensive bundling/thickening of MTs in taxol-treated roots which could affect the putative interaction between MTs and cellulose forming complexes and hence lead to abnormal cell expansion. Thus, dynamic MTs may be necessary to maintain normal growth polarity (Weederburg and Seagull 1988).

The difference in MT organization between intact and excised root segments was not surprising in view of observations that wounding can induce MT reorientation in roots (Hush and Overall 1992). Excision of the root could also induce the production of ethylene, a MT-reorienting agent (Roberts et al. 1985, Baluška et al. 1993). Since excised NPA-treated and control roots had similar MT organization but only NPA inhibited IAA transport, differences in auxin transport are not due to the organization of MTs.

The Cholodny-Went theory proposes that gravitropism occurs as a result of changes in the distribution of auxin. When roots are reoriented, auxin transport is modified such that more auxin accumulates on the lower side of the root inhibiting growth and causing the root to curve downward (Hasenstein and Evans 1988, Lomax et al. 1995). The auxin transport inhibitor NPA has been shown to abolish asymmetric auxin redistribution and gravitropic curvature in maize roots (Lee et al. 1984, Evans et al. 1992). The transport inhibitor NPA had no effect on the MT organization in vertical roots but caused the reorientation of MTs in all cells of the elongation zone on the top and bottom side of horizontally placed roots (see Fig. 5 F, G). These data indicate that in addition to inhibiting (longitudinal) auxin transport, NPA may cause omnidirectional release of auxin, presumably from the stele (see Hasenstein and Evans 1988) during graviresponse. The lateral accumulation of auxin would then lead to two effects, inhibition of growth acceleration on the upper side of the root (Ishikawa et al. 1991) which would prevent curvature, and reorientation and depolymerization of MTs on all sides of the root. The duration of such omnidirectional auxin release and thus inhibition of graviresponse is temporary and likely to be dependent upon the NPA concentration (see Fig. 6C).

Oryzalin slightly enhanced basipetal transport of auxin. Thus, in addition to its proposed mode of action, namely binding to tubulin monomers (Morejohn 1991), oryzalin may cause auxin accumulation in the tip region which then induces root swelling, presumably due to auxin-enhanced ethylene production (Yoshii and Imaseki 1981). This notion is supported by an increase in free IAA levels in roots after application of trifluralin, a herbicide that also causes MT depolymerization (Locher and Pilet 1994).

The data presented here emphasize an intrinsic connection between auxin and the MT cytoskeleton but confirm that, at least in maize roots, the cortical MTs do not participate in auxin transport.

Acknowledgements - This work was supported by NASA grants NAGW-3565 and NAG10-0190

References