Arabidopsis thaliana sku Mutant Seedlings Show Exaggerated ... · ing on tilted agar plates (Okada and Shimura, 1990), we surface-sterilized and plated seeds from the collection de-

Plant Physiol. (1 996) 11 1 : 987-998

Arabidopsis thaliana sku Mutant Seedlings Show Exaggerated Surface-Dependent Alteration in Root Growth Vector'

Robert Rutherford and Patrick H. Masson*

Laboratory of Genetics, University of Wisconsin, 445 Henry Mall, Madison, Wisconsin 53706

Roots of wild-type Arabidopsis thaliana seedlings in the Was- silewskija (WS) and landsberg erecta (ler) ecotypes often grow aslant on vertical agar surfaces. Slanted root growth always occurs to the right of the gravity vector when the root i s viewed through the agar surface, and is not observed in the Columbia ecotype. Right-slanted root growth is surface-dependent and does not result directly from directional environmental stimuli or gradients in the plane of skewing. We have isolated two partially dominant mutations in WS (skul and sku.7) that show an exaggerated right-slanting root-growth phenotype on agar surfaces. The right-slanting root- growth phenotype of wild-type and mutant roots is not the result of diagravitropism or of an alteration in root gravitropism. It is accompanied by a left-handed rotation of the root about its axis within the elongation zone, the rate of which positively correlates with the degree of right-slanted curvature. Our data suggest that the right-slanting root growth phenotype results from an endogenous structural asymmetry that expresses itself by a directional root-tip rotation.

Roots use environmental information to direct their growth. For instance, they reorient when they are tilted within the gravity field (gravitropism), when they are exposed to lateral light (phototropism), to gradients in temperature (thermotropism), humidity (hydrotropism), or ions and chemicals (chemotropism), or when they encoun- ter obstacles (thigmotropism) (Hasenstein and Evans, 1988; Okada and Shimura, 1990, 1992; Fortin and Poff, 1991; Takahashi and Scott, 1991; Ishikawa and Evans, 1992; Mas- son, 1995).

More elaborate growth patterns appear when roots are subjected at the same time to severa1 directional stimuli. The "wavy root" phenotype described by Okada and Shimura (1990) illustrates that point. Arabidopsis thaliana roots cannot penetrate a medium containing high concentrations of agar. Therefore, a tilted agar surface constitutes an obstacle to regular growth, because roots tend to grow vertically downward, parallel to the gravity vector. Conse- quently, the roots initiate a wavy pattern of growth on that

This work was supported in part by grants from the National Institutes of Health (NIH) (no. Rol-GM48053) and the National Aeronautics and Space Administration (no. NAGW4053) and by a Packard fellowship in Science and Engineering to P.H.M. R.R. was supported by a Predoctoral Fellowship in Genetics (NIH Training Grant no. 5-T32-6M07133). This is paper no. 3445 of the Labora- tory of Genetics.

* Corresponding author; e-mail pmassonQmacc.wisc.edu; fax 1- 608 -262-2976.

987

surface. Root waving is accompanied by a succession of right- and left-handed rotations occurring within the elongation zone (Okada and Shimura, 1990). Okada and Shimura (1990, 1992) attributed that phenotype to a combination of gravitropism and obstacle-avoidance responses, although the possible involvement of circumnu- tations was not excluded (Okada and Shimura, 1992; Simmons et al., 1995b).

Although tropisms direct much of the growth of plant organs, they exert their effects within the limits dictated by endogenous constraints of organ structure. For example, gravitropic, thigmotropic, or phototropic stimuli generate curvatures that subsequently decline over time (autotropism; see Pickard, 1985; Nick and Schafer, 1988; Hart, 1990). Experiments in microgravity environments have shown that autotropism is not gravity dependent and that it derives from endogenous properties of the responding organ (Chapman et al., 1994).

Other examples of growth processes conditioned by endogenous properties of the growing organ include nuta- tional and nastic movements. Nutations are native, circular patterns of organ growth that are not triggered by a directional environmental gradient and, therefore, reflect the structural and/ or physiological properties of the growing organ. Similarly, nastic movements result from unique or repeated patterns of differential growth that lead to a change in orientation or shape of a plant organ, and are not the direct consequence of a directional environmental stimulus (reviewed by Hart, 1990; Barlow et al., 1994).

Severa1 lines of evidence suggest that native directional growth bias exists in A. thaliana roots. For instance, Mirza (1987) identified a clockwise "horizontal rotation" of Ara- bidopsis roots embedded between two layers of medium containing 1% agar. Under these conditions, agravitropic auxl roots grow in clockwise circles when viewed from the direction of incident light provided perpendicularly to the agar surface, from above or from below. Furthermore, gravitropism contributes to this curvature in wild-type seedlings, acting either additively (seedlings illuminated from above) or antagonistically (seedlings illuminated from below). Even though curvature is dependent on the direction of light and gravity, the directional bias occurs in a plane perpendicular to these stimuli. The existence of clockwise coiling is consistent with a native growth bias (Mirza, 1987).

Abbreviation: CFR, cell file rotation.

www.plantphysiol.orgon October 28, 2020 - Published by Downloaded from Copyright © 1996 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org

988 Rutherford and Masson Plant Physiol. Vol. 11 1, 1996

When A. thaliana seedlings from the Wassilewskija (WS), Landsberg erecta (Ler), and C24 ecotypes grow on the surface of a horizontal agar medium, they tend to deviate away from the vertical, always slanting to the right on that surface (Simmons et al., 199513). We desig- nate this phenotype "right-slanting" or "skewing." In this paper we characterize that phenotype and report on the identification and characterization of two mutants, named skul and sku2, that develop an exaggerated, right- slanting, root-growth phenotype on agar surfaces.

MATERIALS A N D METHODS

Plant Stocks and Manipulation

Wild-type Arabidopsis thaliana seeds (ecotype WS) were kindly provided by Tim Caspar (DuPont), and wild-type Columbus (Col) and Ler seeds were provided by the Ara- bidopsis Biological Resource Center (The Ohio State Uni- versity, Columbus). The collection of T-DNA-mutagenized A. thaliana lines (49 pools of 100 independent transgenic lines each, corresponding to a total of about 60,000 seeds) was also obtained from the Arabidopsis Biological Re- source Center (Feldmann et al., 1994). All techniques aimed at surface sterilizing A. thaliana seeds: germinating and growing them in Petri dishes on medium containing one- half the concentration of salts recommended by Murashige and Skoog (1962) with 1.5% Suc and various concentrations of agar, as indicated in the text; transferring the corresponding seedlings to soil; and growing, self-fertilizing, or cross-pollinating them, were as described by Okada and Shimura (1990), by Fedoroff and Smith (1993), and by Simmons et al. (1995a, 199513).

Screening for A. thaliana Mutants Affected in Root Waving

To identify A. thaliana mutants defective in root waving on tilted agar plates (Okada and Shimura, 1990), we surface-sterilized and plated seeds from the collection described above on the surface of a medium containing 1.5% agar (type E, Sigma) in square Petri dishes. Plates were wrapped in aluminum foi1 and incubated at 4°C for 2 d to promote uniform germination. They were unwrapped and grown vertically at 22°C in a Conviron (Asheville, NC) TC16 growth chamber and subjected to 16-h day/8-h night cycles for 3 d. Light intensity was maintained at around 70 pE m-' spl using cool-white fluorescent bulbs (F20T12/ CW, Philips, Eindhoven, The Netherlands). Then, plates were tilted backward 30" and grown for another 4 d. At the end of that period, the pattern of root waving was compared with that of control wild-type plants of the same ecotype (WS) subjected to the same experimental protocol (Okada and Shimura, 1990). Seedlings that showed altered root waves were transferred to soil and allowed to self- fertilize for four generations. These progenies were tested the same way.

Quantification of Right-Slanting

Surface-sterilized wild-type and mutant seeds were sown on the surface of 0.8 or 1.5% agar medium (type E,

Sigma). After 2 d in the cold (4°C) and darkness, plates were transferred to the Conviron TC16 growth chamber and seedlings were grown under the conditions described above, except that the plates were positioned at an angle of 95 to 100" from the horizontal (tilted slightly forward) to avoid root waving (Okada and Shimura, 1990; Simmons et al., 1995b). In some experiments (see "Results" and "Dis- cussion"), plates were positioned at various angles from the vertical to test the effect of the extent of plate tilting on the response. Pictures were taken every 6 h with a Nikon N8008S camera mounted with a 60-mm Micro Nikkor (To- kyo, Japan) 1:2.8 lens, and Kodak Ektachrome (Rochester, NY) 160T film. Slides were digitized and analyzed with the National Institutes of Health (NIH, Bethesda, MD) Image Analysis software (version 1.54) on a Macintosh Centris 650 computer to determine the root-tip angle from the vertical. The data were transferred to a Microsoft Excel spreadsheet, and the average angle of root growth at each time point and its SD were calculated and plotted using the same software.

The effect of white light, red light, and far-red light on the left-slanted, root-growth phenotype was determined by plating seedlings on agar surfaces and analyzing them as described above, except that the specific light treatment was provided in a Conviron TC16 growth chamber. Red- and far-red-light treatments were performed by filtering white light (generated by Philips F20T12/ CW cool-white fluorescent bulbs) through two layers of monochromatic red and far-red light filters (wavelengths of transmitted light were 620-680 nm with a maximum peak at 659 nm for the red-light filter, and 670-870 nm with a maximum peak at 750 nm for the far-red-light treatment). In both cases, light intensity in the chamber dropped to approximately 6.6 pE m-'s-'.

Quantification of Root Cravitropism and Root Coiling on a Clinostat

Similar strategies were used to quantify root gravitropism, except that vertical plates were rotated by 90" after 2 d of growth (Bullen et al., 1990; Masson et al., 1993; Simmons et al., 1995a). Pictures were taken before and every 6 h after reorientation and analyzed as described above.

Similarly, the rate of root coiling on the clinostat was measured on seedlings that were grown either on the surface of 0.8% agar medium or submerged within the medium (as described in the "Results") in Petri dishes that rotated at 1.5 rpm around an axis either perpendicular or parallel to the agar surface. Petri dishes that rotated around an axis perpendicular to the agar surface were positioned lid-to-lid on the clinostat so that one-half were rotating clockwise and one-half were rotating counterclockwise when viewed through their lids. Plates that rotated around an axis parallel to the agar surface were also placed lid-to- lid, but were scored together because little plate-to-plate variation was observed in that experiment. The roots of approximately 20% of the tested seedlings grew into the agar medium or outward into the air. These seedlings were omitted from our analysis. In each clinostat experiment,



Surface-Dependent Alteration of Root Growth Vector 989

seedlings were allowed to grow for a period of 96 h, afterwhich the plates were analyzed as described above. Theroot-tip angle from a line extending from the hypocotyldownward was measured on each seedling according tothe convention shown in Figure 5, and the mean angle andSD were determined. Between 8 and 31 seedlings of eachgenotype (Col, WS, skul, and sku2) were subjected to eachassay condition.

Quantification of Root-Tip Rotation

The rate of root tip rotation about its axis was deter-mined by microscopic analysis of epidermal cell files usinga dissecting microscope. Pictures were taken of each root,digitized, and analyzed with the NIH software. The root tip(7 mm) was virtually segmented into 0.160-mm units(which corresponds on average to the width of an A. thali-ana root, 0.160 ± 0.026 mm, n = 80). For each segment, thenumber of epidermal cell files crossing the root midlinewas determined and considered positive if associated withleft-handed root-tip rotation, or negative if associated withright-handed root-tip rotation. The average number ofCFRs per segment was determined for each root and de-fined the average CFR for that root.

Time-course studies of CFR were conducted by transfer-ring 6-d-old skul, sku2, and WS seedlings onto plates con-taining freshly poured medium containing half the concen-tration of salts recommended by Murashige and Skoog(1962) with 1.5% agar, realigning them to the vertical, andlabeling the lowest 7 to 10 mm of the root tip with graphitegrains (Okada and Shimura, 1990). Graphite grains werealso placed on adjacent sites on the agar surface to serve asfixed reference points for measurements. Plates were trans-ferred to the growth chamber and incubated under theconditions described above, tilted 5 to 10° forward to avoidroot waving. Pictures were taken every hour for a period of20 h through a dissecting microscope (M3Z, Wild, Heer-brugg, Switzerland) (magnification 200X), digitized, andanalyzed with the NIH Image software to determine theposition of each graphite grain on the surface of the rootrelative to the agar surface. Dynamic data regarding root-tip rotation were obtained by defining the position of each

grain relative to the others, as well as relative to the X-Zaxis defined by the agar surface and a virtual cross-sectionof the root, as shown in Figure 8B and described by Okadaand Shimura (1990).

RESULTS

Direction of Root Growth in Wild-Type A. thalianaSeedlings on Vertical Plates

The direction of root growth in wild-type plants belong-ing to various A. thaliana ecotypes was determined bymeasuring the angle from the vertical root tips on seedlingsgrowing on an agar surface. The roots of wild-type Colseedlings grew almost straight and vertically on verticalagar surfaces (Fig. 1; Table I). However, wild-type WS andLer roots grew progressively away from the vertical, al-ways slanting to the right (to the left when roots were seenthrough the lid of the plate) on such surfaces. On the otherhand, roots from seedlings of all tested ecotypes grewvertically when embedded within the medium (Table I).Therefore, wild-type roots of different A. thaliana ecotypesshowed different growth properties on the surface of anagar medium.

Identification of sku Mutants

To better define the mechanisms involved in slanted rootgrowth on the surface of an agar medium, we identified A.thaliana mutants affected in that phenotype, using thewavy-root screening assay developed by Okada andShimura (1990). From the 4900 T-DNA-mutagenized WSlines tested, three showed an exaggerated right-slantingroot-growth phenotype when grown on the surface of a1.5% agar medium tilted 30° from the vertical, or on thesurface of a 0.8% agar medium tilted 5 to 10° forward toavoid root waving (Fig. 2, A and B). All three mutantsmaintained the right-slanting root-growth phenotype for atleast four generations. The mutations defined by thesethree lines were designated skul, sku2, and sku3; they werenot tagged by a T-DNA insert (data not shown). Two ofthese (skul and sku2) are further characterized here.

C»tFigure 1. Right-slanting root-growth pheno-type of wild-type Col, Ler, and WS seedlings.Photograph of 7-d-old Col, Ler, and WS seed-lings grown on the surface of a 0.8% agarmedium in a plate tilted slightly forward toreduce root waving. The photograph wastaken from the back of the plate through theagar medium. Note that the WS and Ler rootsslant to the right, whereas the Col roots growmore or less vertically.



990 Rutherford and Masson Plant Physiol. Vol. 111, 1996

Table I. Average root-tip angle from the vertical (mean ± so) ofwild type Col, Ler, and WS seedlings (n) grown for 10 d on thesurface or embedded in agar medium

A positive root-tip angle represents a deviation to the right of thevertical, and a negative angle represents a deviation to the left.

Seedling

ColLerWS

Surface

Angle5.98 ± 4.6

17.27 ± 6.719. 76 ±6.1

n

554463

Embedded

Angle-4.41 ± 7.25

-0.6 ± 10.4-0.52 ± 5.76

n

293858

The kinetics of right-slanting in roots of wild-type WSand of skul and sku2 mutant seedlings were defined bytime-lapse studies. A deviation from the vertical in root-growth direction appeared very early after germinationand increased over time (Fig. 2). In the assay shown inFigure 2B, skul and sku2 mutant roots reoriented to thehorizontal after 10 d of growth, and some roots reorientedeven farther, growing upward. Furthermore, WS rootsskewed minimally, illustrating the phenotypic variabilityof the right-slanting root-growth phenotype (see below). Inexperiments in which right slanting of skul and sku2 mu-tant roots did not develop as severely, the roots reorientedto reach an asymptotic growth vector at 30 to 90° to theright from the vertical, after 8 to 10 d of growth (Fig. 2, Aand B). In all cases the exaggerated right-slanting root-growth phenotype of skul and sku2 mutant seedlings wasassociated with a slight reduction in the rate of root growth(77 and 70% of the wild type, respectively).

The degree of right-slanting of wild-type WS and mutantskul and sku2 roots varied from experiment to experiment,reaching an average angle from the vertical that variedfrom 30 to 90° for skul and sku2 mutant plants after 10 d ofgrowth, and it was sometimes absent for WS seedlings (seebelow). However, plate-to-plate variation within each ex-periment was minimal and was controlled by includingplants from all tested genotypes on the same plate. Fur-thermore, skul and sku2 mutant roots were always found toslant rightward more severely than those of wild-type WS

plants, whereas the wild-type Col roots showed little or noslanting.

The right-slanting root-growth phenotype is surface-dependent. Indeed, vertical growth resumed almost imme-diately after right-slanting roots penetrated the agar sur-face. Furthermore, when the roots reached the bottom ofthe plate, right-slanted root growth resumed. Roots alsoslanted to the right when growing across a surface of moistnitrocellulose filter (data not shown).

The agar concentration of the medium used to test right-slanted root growth influenced the intensity of the pheno-type. Roots grown on the surface of 0.4% agar medium didnot skew significantly, regardless of their genotype (TableII). These roots showed a tremendous proliferation of roothairs all around, even though they clearly grew on thesurface. Contact with harder agar media (>0.8% agar) hasbeen described to inhibit root-hair elongation in Arabidop-sis (Okada, 1993). On the other hand, mutant roots grownon 0.8 or 1.5% agar medium skewed more severely, andmutant roots grown on a 3.0% agar surface skewed lessthan roots grown on a 0.8 or 1.5% agar surface. Finally,roots grown between layers of 0.8% agar medium or be-tween layers of 0.8 and 3.0% agar medium did not skew.

Genetic Analysis of sku Mutants

The genetic basis of the exaggerated right-slanted root-growth phenotype in these mutant lines was determinedby backcrossing each mutant with a wild-type WS plantand analyzing the right-slanting root-growth phenotypeof the corresponding progeny. The skul and sku2 muta-tions are at least partially dominant: the distribution ofroot angles from the vertical in Fj plants was moresimilar to the distribution of root angles in homozygousmutant plants than to that of homozygous wild-typeplants (Fig. 3, A and B).

When the backcrossed F, plants were selfed and theright-slanting root-growth genotype of the correspondingF2 plants was determined by progeny typing, the segrega-tion ratios (homozygous wild type:heterozygous:homozy-gous mutant F2 seedlings) were consistent with the 1:2:1

Figure 2. Right-slanting root-growth phenotypeof wild-type Col and WS and of mutant skul andsku2 seedlings. A, Photograph of 7-d-old WS(left of the black lines) and sku2 (right of theblack lines) seedlings grown on the surface of a0.8% agar medium in a plate tilted slightly for-ward to reduce root waving. Note that sku2roots slant more to the right than WS roots. B,Quantification of the right-slanting root-growthphenotype over time for Col, WS, skul, andsku2 seedlings. The x axis represents the time ofgrowth (d) and the yaxis represents the averageroot-tip angle from the vertical (n = 24), usingthe convention mentioned in the legend to Fig-ure 1. At each time point, the so of the measure-ments is represented by vertical bars.

1 \\J '100-90-80

l e o "•5 50£40| 30

2010

-10

* WS~'m"'sluil

* Col— -x— -sku2

\

s*,'

\-'''

( r

;S \

t' /''

s

r'' '' ' •^'

i i.H-5 7 9

Days grown11



Surface-Dependent Alteration of Root Crowth Vector 991

Table II. Average root tip angle from the vertical (mean 2 SD) of seedlings grown for 10 d on the surface or submerged between two layers (top/bottom) of agar medium (same sign convention as in Table 1)

Agar Concentration (70)

Seedl i ng Surface Submerged

0.4 0.8 1.5 3.0 0.810.8 0.8/3.0

0.3 5 0.8 Col -3.2 2 4.7 0.8 2 1.5 6.0 2 2.4 0.5 -+ 1.5 0.6 2 0.9 ws 4.7 2 1.2 -2.0 2 1.8 2.8 2 1.3 -0.4 2 1.1 0.1 2 1.1 0.7 +- 1.6

1.1 2 0.7 0.0 f 0.8 sku 1 8.8 2 1.7 46.0 2 1.6 57.7 2 1.4 sku2 -6.7 2 1.5 41.6 2 1.9 58.5 2 1.7 25.4 2 1.5 1.5 t 0.7 4.0 5 0.9

20.8 2 1.5

segregation ratio expected for mutations in a single locus. For instance, the F, progeny of a cross between skul and wild-type WS plants contained 4 homozygous SKUl 1 SKUl plants, 11 heterozygous SKUl lskul plants, and 5 homozygous skul l sku l mutant plants (2 = 0.3; P = 0.86).

To determine if the skul and sku2 mutations affected the same or different loci, we crossed homozygous skul and sku2 plants and analyzed the right-slanting root-growth phenotype of the corresponding F, and selfed F, families. If the skul and sku2 mutations affect two different loci, one expects to identify wild-type seedlings in the F, population as long as these two loci are not closely linked. Addition- ally, because these two mutations are at least partially dominant, one also expects that some F, plants (those heterozygous for one or both loci) will develop an intermediate right-slanting root-growth phenotype between the wild-type and mutant parents. In the experiment shown in Figure 3C, the distributions of root-tip angles for skul and sku2 seedlings (varying between 10 and 45") did not over- lap with that of wild-type WS seedlings (varying between O and 10"). Yet, some segregating F, seedlings showed a degree of root slanting typical of wild-type seedlings, whereas others slanted to a degree intermediate between that of wild-type and mutant seedlings, and yet others showed a mutant phenotype. These data are consistent with the skul and sku2 mutations affecting two different loci.

Root Cravitropism 1s Not Affected in skul and sku2 Mutants

Surface-grown mutant roots reoriented upon gravi- stimulation as fast as wild-type roots, independent of whether they were reoriented leftward (Fig. 4A) or rightward (Fig. 48). However, the final direction of root growth after reorientation remained rightward of the gravity vector. In contrast, seedlings grown within agar (0.8%) re- sponded like the wild type to the same 90" reorientation stimulus and resumed vertical growth after reorientation (Fig. 4C).

Because skul and sku2 mutant roots were often found to grow horizontally rather than vertically, we tested the possibility that the right-slanted root-growth phenotype derives from diagravitropism (controlled horizontal growth). Vertical plates carrying seedlings with severely right-slanting (horizontal) roots were chosen and tilted horizontally so that roots were still horizontal but lying on the top of a horizontal agar surface, perpendicular to the

gravity vector. If right slanting is simply due to diagravitropism (horizontal growth), one would expect these roots to continue growing straight (horizontally) on the surface of the medium. If not, roots should reorient, penetrate the agar surface, and resume vertical growth. In a11 cases, roots started to coil clockwise on the agar surface before pene- trating it and growing vertically to the bottom of the plate (data not shown). This behavior indicates orthogravitro- pism rather than diagravitropism.

To investigate the possibility that gravitropism could modify the intensity of the right-slanting root-growth phenotype, we quantified the root-tip angle from the vertical of seedlings growing on 1.5% agar surfaces tilted to 70, 90, or 110" from the horizontal. skul and sku2 mutant roots developed a more severe right-slanting root-growth phenotype on plates tilted backward (70") than on vertical plates (90") and on vertical plates than on plates tilted forward (Table 111). Wild-type WS and Col roots did not slant significantly in this experiment (Table 111). However, wild- type WS and Ler, as well as mutant skul and sku2 roots, coiled when the plates were positioned horizontally (Sim- mons et al., 199513; R. Rutherford and P.H. Masson, unpublished data). Therefore, the right-slanting root-growth phenotype is not the consequence of plagiogravitropism, and its intensity correlates with the intensity of the pressure between the root and the agar surface.

Slanted Root Growth 1s Not a Direct Result of Cravitropism or Phototropism

To determine if right-slanting root growth is a result of gravitropism or phototropism, wild-type Col and WS and mutant skul and sku2 seedlings were germinated and grown on the surface of a 0.8% agar medium in clinorotated Petri dishes. The clinorotation negates the gravity and light vectors. The skul, sku2, and wild-type WS roots coiled little or not at a11 in either direction when embedded within the agar medium, regardless of the direction of clinorotation (Fig. 5). However, wild-type Col roots coiled more clockwise or counterclockwise when clinorotated clockwise or counterclockwise, respectively, around an axis perpendicular to the agar surface (Fig. 5). Embedded roots that clinorotated around an axis parallel to the agar surface tended to coil in a11 three dimensions (data not shown).

A very different picture emerged when seedlings that were growing on the surface of an agar medium were clinorotated (Fig. 5). When they were clinorotated clock-



992 Rutherford and Masson Plant Physiol. Vol. 111, 1996

O.45

0.4

0.35

0.3CO 0.25

« O.2u.

0.1 S

O.1

O.O5

D ws• Vt'Sakul HI• skul

Angle Skewed0.45

O.4

0.35

0.3C

•5 O.2S%£ 0-2

0.15

O.1

O.O5

O

BO WS• WSwto.'KI

Lllk!S S

Angle Skewed

O.4

O.35

0.2

0.15

0.1

O.O5 1

a ws

Angle Skewed

Figure 3. Right-slanting root-growth phenotype of homozygouswild-type WS, homozygous skul/skul mutants, and heterozygousSKUl/skul seedlings (A), homozygous wild-type WS, homozy-gous mutant sku2/sku2, and heterozygous SKU2/sku2 seedlings(B), and homozygous wild-type, homozygous skul/skul and sku2/sku2 mutant seedlings, and selfed F2 progeny from a cross be-tween sku1/sku1 and sku2/sku2 mutant plants (C). In all cases, thehomozygous seedlings were derived by self-pollination from theparents used to generate, by cross-pollination, the heterozygotesshown in the graph. Each histogram shows the distribution of rootangles from the vertical after 8 d of growth under the conditionsdescribed in "Materials and Methods." The y axis of each graphrepresents the proportion of seedlings with roots slanting to theright at the angle represented on the x axis. In all cases, 40 to 70seedlings were analyzed per genotype, except for the wild-typeWS seedlings shown in C, of which 9 seedlings were tested.

wise around an axis perpendicular to the agar surface, skuland sku2 roots coiled clockwise, whereas wild-type WS andCol roots coiled counterclockwise. Surface-grown skul andsku2 roots also coiled clockwise more than wild-type WSroots when clinorotated around an axis parallel to the agarsurface. Finally, when the plates were clinorotated coun-terclockwise around an axis perpendicular to the agar sur-face, Col roots coiled more counterclockwise than sku2 or

WS roots, skul roots, on the other hand, maintained aclockwise coiling under these conditions (Fig. 5).

To test the existence of possible links between right-slanted root growth and the direction and intensity ofincident light, vertical plates carrying wild-type WS andCol and mutant skul and sku2 seedlings were exposed tovarious light regimens, including darkness, or light appliedfrom either top or bottom of the plate for a period of 10 d.Mutant skul and sku2 roots slanted equally well whetherthe plates were illuminated from above or below (TableIV). skul and sku2 roots also slanted significantly to theright when seedlings grew in darkness, although to a lesserextent than when they were illuminated. However, dark-grown roots were much shorter than light-grown roots.Wild-type Col and WS roots did not slant significantly inthis experiment (Table IV). Similar experiments showedthat red or far-red light do not affect the extent of rightslanting (data not shown). In summary, light quality, in-tensity, and direction do not modify the right-slantingroot-growth phenotype.

Right Slanting Is Not a Tropic Response to SpecificEnvironmental Gradients in the Plane of Skewing

To determine if right-slanted root growth derived from atropic response to some unidentified environmental vectoror gradient in the plane of skewing, we germinated wild-type Col and WS and mutant skul and sku2 seeds on thesurface of a 0.8% agar medium poured into both the lid andthe base of Petri dishes. The dishes were closed and posi-tioned vertically in the growth chamber. Pictures of theplates were taken after 10 d of growth. sku2 mutant rootsgrew slanted to the right on their own surface, rather thanwith respect to some external gradient (Fig. 6, A and B).Similar results were obtained with skul mutant and wild-type WS roots, whereas Col roots showed little or no slant-ing on either surfaces (Fig. 1, and data not shown).

Root-Tip Rotation Accompanies Right-Slanted Growth

A microscopic analysis of right-slanting roots indicatedthat right slanting is always accompanied by a character-istic rotation of the root tip, resulting in the appearance ofCFR (Simmons et al., 1995b) around the root axis (Fig. 7A).Quantification of the amount of CFR on surface-grownroots indicated that Col roots developing an insignificantamount of right slanting also showed a small amount ofleft-handed CFR (0.14 CFRs per root width; Fig. 7A). Right-slanting wild-type WS and mutant skul and sku2 rootsdeveloped a significant amount of left-handed CFR (0.39,0.70, and 0.69 CFRs per root width, respectively). By con-trast, roots embedded within the agar medium grew ver-tically and showed little or no CFR (data not shown).

Root-tip rotation during right-slanted growth was quan-tified by labeling the surface of the root tip of seven WS,skul, and skul seedlings with graphite grains and followingthe movement of each grain with respect to the agar surfaceand to each other during right-slanted growth (see "Mate-rials and Methods"). The root tip of a typical skul seedlinggrew to the right of the gravity vector, eventually reaching www.plantphysiol.orgon October 28, 2020 - Published by Downloaded from

Copyright © 1996 American Society of Plant Biologists. All rights reserved.



C,

-500 6 12 18 24

Hours after reorientation

-50 -506 12 18 24

Hours after reorientation0 6 12 18 24

Hours after reorientation

Figure 4. Gravitropic response of wild-type WS and mutant skul and sku2 roots grown on the surface of a vertical agarmedium after a 90° counterclockwise (A) or clockwise (B) reorientation, or grown in agar after a 90° counterclockwisereorientation (C). Twenty-four seedlings of each genotype were tested in each case (see "Materials and Methods").Shown here are the averages of root-tip angles from the vertical after reorientation (y axis) over time (x axis). Error barsrepresent 1 so.

an angle of 45° from the vertical after 8 h (Fig. 8A). Thebasal limit of the elongation zone extended to about 0.60mm away from the root tip, even though some elongationstill occurred up to 1 mm away from the tip (Fig. 8, B andD). The apical limit of the elongation zone was locatedabout 0.25 mm from the root tip. Finally, the distancebetween most other markers and the root tip and adjacentmarkers continued to increase over time, indicating thatthey were included within the elongation zone over thatperiod of time (Fig. 8, B and D). The length of the elonga-tion zone on this A. thaliana root was consistent with thedata reported by Okada and Shimura (1990) using similarstrategies.

Root-tip rotation was defined by following the positionof each grain around the root axis over time (Fig. 8B). Allgrains within the elongation zone underwent left-handedrotation about the root axis (Fig. 8C). That rotation re-mained visible until the grains reached the agar surface(represented by 90°; Fig. 8C). The rate of rotation correlatedwith root reorientation (between 3 and 7 h after realign-ment). Grain e ceased its left-handed rotation after 5 h,when it left the elongation zone. Similarly, grain f, whichwas in the elongation zone only for the 1st h after reorien-

Table III. Average root-tip angle from the vertical (mean ± so) ofseedlings after 10 d of growth on the surface of 1.5% agar mediumin Petri dishes tilted to 70, 90 (vertical), or 110° from the horizon-tal (same sign convention as in Table I)

Angle

ColWSskulsku2

70°

1.5 ± 0.9-1.5 ± 1.047.1 ± 1.546.8 ± 1.8

90°

3.6 ± 1.53.3 ± 1.4

32.2 ± 3.133.7 ± 1.1

110°

5.2 ± 1.42.7 ± 1.4

22.6 ± 1.029.3 ± 1.2

tation, did not rotate about its axis, but remained fixed at+ 10° (data not shown). Finally, the rate of grain displace-ment was higher for grains located at the distal end of theelongation zone (between 0.25 and 0.35 mm from the roottip, as determined by the position of grains c and d be-tween 3 and 5 h) than for grains located at the root tip. Thisindicates that the rotation was initiated in the elongationzone (grain a in Fig. 8C).

DISCUSSION

Roots appear radially symmetrical, but they often re-spond asymmetrically to specific environmental stimuli. Inthis report, we show that seedling roots from several A.thaliana ecotypes (Ler and WS) but not others (Col) grewasymmetrically on the surface of an agar medium within anotherwise homogenous environment. Under these condi-tions, roots grew away from the vertical, slanting progres-sively to their right (to the left when viewed through the lidof the plate), tending to reach a specific angle from thevertical when balanced by gravitropism. Two mutants(skul and sku2) were isolated with exaggerated surface-dependent, right-slanted root-growth phenotypes.

The skewing wild-type WS and mutant skul and sku2roots developed similar gravitropic responses to 90° gravi-stimulations when embedded within the agar medium.They also developed similar kinetics of gravitropic reori-entation when growing on the surface of an agar medium.However, in this case they tended to realign rightward ofthe vertical (Fig. 4). These data suggest that skewing rootspossess full gravitropic competence but that they experi-ence a surface-dependent deviation of growth from thevertical that gravitropism cannot completely correct.

The clinostat experiments demonstrate that right-slanting isnot a direct result of gravitropism or phototropism. Wild-




Clockwise Clinostating

1 :;I Y6S sk;; sku22 1 Mean -8.58 35.1 100.7 153.2

Stdev mean 23.77 24.87 18.14 24.99 n

Mean I f 4 ;tf 1 n Stdev mean 25.22 26.98 12.06

“Perpendicular” Cliinstatting

n

Counterclockwise Clinostating

Root Coiling Conventions

Figure 5 . Clinostat effects on directional root growth in and on agar. Wild-type Col and WS and mutant skul and sku2 seedlings were grown for 96 h within or on a 0.8% agar medium in Petri dishes subjected to clockwise (left) or counterclockwise (right) clinorotation around an axis perpendicular to the agar surface, or around an axis parallel to the agar surface (center). Shown are the mean angle of root coiling measured according to the conventions shown in the figure, the SD (Stdev mean), and the number of plants analyzed (n).

type WS and mutant skul and sku2 roots growing on agar surfaces coiled dramatically clockwise (when viewed from the top of the plate, as defined by Mirza, 1987) when rotated clockwise around an axis perpendicular to the agar surface, or around an axis parallel to that surface, whereas wild-type Col roots did not coil more in one direction than in the other under these conditions. Furthermore, skul and sku2 roots coiled more on average than wild-type WS roots. Finally, none of the tested genotypes showed root coiling when embedded in the agar medium (Fig. 5). These data are consistent with the observation that agravitropic mutants in the WS and Ler backgrounds show a clockwise- root-coiling phenotype when growing on the surface of a vertical agar medium (Maher and Martindale, 1980; Sim- mons et al., 199513; R. Rutherford, P. Hilson, J. Sedbrook, and P.H. Masson, unpublished data).

Taken together, these data suggest that clockwise coiling results from the development of an unchecked, left-handed curving in the absence of gravitropism (Simmons et al., 199513; R. Rutherford and P.H. Masson, unpublished data). However, that conclusion should be interpreted with cau- tion. Indeed, Col, WS, and sku2 roots coiled counterclockwise on plates clinorotated counterclockwise (Fig. 5). This coiling behavior is consistent with the data reported by

Table IV. Average root tip angle from the vertical (mean ? SD) of Col, WS, skul, and sku2 seedlings after 10 d of growth on the surface of 1.5% agar medium, illuminated from above or below (70 pE m-’ s-’or grown in darkness

The sign convention is the same as in Table I. Light

Seedl ing Above Below Darkness

Col -3.1 2 9.5 -2.6 ? 10.3 1.1 rf- 13.4 ws -1.8 * 10.2 -2.5 2 11.0 -3.4 ? 11.7 sku 1 22.0 ? 10.8 30.3 2 11.2 12.2? 11.5 sku2 20.8 2 13.0 24.1 ? 9.6 14.4 ? 9.9

Mirza (1988) and may reflect an unusual root-growth response to clinorotation rather than an endogenous pattern of root growth (Mirza, 1988).

When clinorotated, roots remained in contact with the agar surface. Similarly, roots growing on plates tilted to 110” from the horizontal remained in contact with the medium, even though gravitropism tended to pull them away from that surface (Table 111). These data imply that the root is attracted to the agar by specific characteristics of the surface. Such characteristics could include humidity and/or nutrients, for which tropic responses have been documented in severa1 plant species (reviewed by Masson, 1995), or surface tension. On the other hand, the right- slanting root growth phenotype did not appear when roots grew within the agar medium on tilted plates, but it resumed when roots hit the bottom of the plate. In this case, the direction of slanted growth on the plastic surface was similar to that of roots growing on the agar surface, even though the humidity or chemical gradient vectors were opposite, being directed toward the surface when roots grew on agar medium and away from the surface when they grew on the bottom of the dish. These data suggest that the right-slanting root-growth phenotype is not simply derived from hydrotropism or chemotropism at the surface, and that the interaction between the root and the agar surface itself, deriving from a combination of gravitropism, hydrotropism, chemotropism, phototropism, and / or surface tension, is responsible for the clockwise coiling and right-slanting root-growth phenotypes described here.

Therefore, even though gravitropism, phototropism, hydrotropism, or chemotropism are probably not individu- ally required for the development of a right-slanting root- growth phenotype, they seem to contribute to it by enhancing the intensity of interaction between root and agar surface. Indeed, one can alter the degree of right slanting by modifying the character of that interaction. For instance, the degree of left-slanted root growth was larger




B

Figure 6. The right-slanting root-growth phenotype is not the direct consequence of a directional environmental stimulus.Seedlings were grown on agar surfaces poured into both the base and the lid of a square Petri dish, as defined in "Materialsand Methods." Shown is a photograph of the plate after 8 d of growth (A) from the perspective represented in B. The bottomrow of seedlings is nearest the camera, on the inner side of the nearest agar surface, whereas the row at the top of thephotograph sits on the inner side of the farthest surface.

Col WS skul sku2 BRoot#

123456

AveStd

Degreesskewed

211-151893

MeanLHCFR

0.18-0.040.190.150.290.07

5115

0.140.12

Degreesskewed

21015303523

MeanLHCFR

0.320.880.230.540.44

-0.03

19123

0.390.30

Degreesskewed

162213281616

MeanLHCFR

0.340.710.620.651.090.77

185.4

0.700.24

Degreesskewed

5540495342

MeanLHCFR

0.550.48

1.00.600.84

47.86.6

0.690.21

H,,: WSp=0.087p=0.076

Figure 7. The right-slanting root-growth phenotype is accompanied by a left-handed CFR. A, Shown are the average root-tipangle from the vertical (Degrees skewed) and the average number of left-handed CFR (LH CFR) per unit of root width (0.16mm) for the apical 8 mm of five to six roots (Root #) from Col, WS, skul, and sku2 seedlings, calculated as shown in B.Summary statistics for each tested genotype are totaled in the last two rows of the table (Ave and Std, average root-tip angleand CFR for all roots of each genotype, and its associated so, respectively). Also shown under the table are the results of aX2 test aimed at testing the probability that the observed number of CFRs for a specific ecotype does not differ from that ofanother ecotype, as defined on the left (H0) (/iCFR, mean number of CFRs for the designated genotype). B, Microphotographof a root tip showing epidermal cell files. The sectioned bar to the left of the root represents an area of the root tip subjectedto the quantification described in A. The numbers to the left of each bar section represent the number of CFRs for that section,as defined in "Materials and Methods." The numbers in A under the heading "Mean LH CFR" are averages of the CFRsobtained for each 0.16-mm section of a root up to 7 mm from the tip. www.plantphysiol.orgon October 28, 2020 - Published by Downloaded from



996 Rutherford and Masson Plant Physiol. Vol. 11 1 , 1996

Figure 8. Kinetic quantification of right slanting, rotation, and growth of a representative sku2 root. A, Root-tip angle from the vertical over time. 6, Position of the graphite grains (grains a-O on the root surface at time O. C, Position of each graphite grain along a section of the root at each time point (angle from an axis perpendicular to the root and to the agar surface, as shown in 6 and as described by Okada and Shimura [1990]). D, Distance of each graphite grain from the root tip at each time point. Grain f has not been represented in these graphs because it left the elongation zone very early in the analysis and, consequently, did not rotate (see ”Materials and Methods”).

A Grain angle - 1 0 40

C Root tip angle - 6 0 5 0 40 30 2 0 1 0 O

D

’ -pc

2

3

4

5

6

7

8

9

1.0

1 .6

1.4 3 3 1 .2 7

20.8

g 0 . 6

s 1 c

0 4

o .2

O 1 2

when skul and sku2 seedlings were grown on an agar surface tilted backward than when they were grown on the surface of a vertical agar medium or on the surface of an agar medium tilted forward (Table 111). Therefore, the degree of root slanting in skul and sku2 seedlings was highest when gravitropism tended to direct the roots toward the agar medium and lowest when gravitropism tended to direct them away from the surface.

On the other hand, skul and sku2 mutant roots growing vertically on a 0.4% agar surface did not skew significantly, in contrast to the progressively higher levels of skewing seen on 0.8 and 1.5% agar medium. It is interesting that roots grown on a 0.4% agar medium were covered with root hairs, whereas the contact with harder agar surfaces inhibited root-hair formation (Okada, 1993). This suggests that the hardness of the agar medium on which the roots grow is an important determinant of the right-slanting root-growth phenotype. The extremely fluid surface of a 0.4% agar medium may lack the rigidity necessary to cause the root to skew, whereas that of a 0.8 or 1.5% agar medium is sufficient. By contrast, roots grown on the surface of an extremely rigid 3.0% agar medium skewed less than roots

i’ I

e

5 4 5 6 7 0 9 hours

grown on a 0.8 or a 1.5% agar medium (Table 11). Roots specifically inscribe troughs with microscopic structural details (cell files) when grown on the surface of a 3.0% agar medium but not on a 0.8 or 1.5% agar medium, suggesting that such a high concentration of agar (3%) offers an ex- cessive resistance to root-tip movement when compared to the more yielding 0.8 or 1.5% agar surfaces. These data suggest that skewing depends on the mechanical interaction between root and agar surface.

Roots that were grown on a hard 3.0% agar surface but submerged in a soft 0.8% agar medium did not slant, whereas roots that were grown within the same 0.8% agar medium on the bottom of the dish did, suggesting that a minimal differential resistance between the two surfaces is required for the development of that phenotype. Finally, it should be remembered that although the surface is clearly important for the development of a right-slanting root- growth phenotype, that phenotype is fundamentally the same when roots grow across a moist nitrocellulose filter, which has a very different texture.

Right-slanted growth was accompanied by a left-handed rotation of the root about its axis within the elongation




zone. On average, right-slanting skul and sku2 roots ro-tated more about their axes than wild-type WS roots did,and wild-type WS roots rotated more than wild-type Colroots did (Fig. 7). This tendency parallels the right-slanted-growth phenotype of these plants. However, this correla-tion does not always hold when one compares the intensityof CFR and the level of right slanting for individual roots ofthe same genotype, suggesting that the link between root-tip rotation and right slanting might not be directly caus-ative. Furthermore, kinetic studies have shown that rootsrotate faster when they curve and much less so when theydo not curve (Fig. 8, and data not shown). Finally, Okadaand Shimura (1990) have argued that root-tip rotationwithin the elongation zone is directly responsible for thecurvature of roots waving on the surface of tilted agarmedium. Therefore, skewing roots seem to rotate withinthe elongation zone when they curve, rather than rotatingcontinuously (regardless of curvature). Hence, the curvatureof a skewing root may result from a surface-dependent, left-handed rotation of the root tip about its axis.

The mechanism(s) by which root-tip rotation translatesitself into skewing may derive from the resistance to rota-tion and movement encountered by the root tip on the agarsurface. We have observed that the rate of rotation is largerat the distal end of the elongation zone than at the very tipof the root, often resulting in that subterminal region of theroot bowing out of the agar surface (Fig. 8C, and data notshown) (Simmons et al., 1995b). This bowing of the rootresults in a curvature at the elongation zone, which is likelyresponsible for the clockwise coiling of clinorotated roots.Clockwise coiling will lead to a right-slanted root-growthphenotype when corrected by gravitropism.

From these analyses we conclude that the right-slantedroot-growth phenotype on the surface of vertical agarplates results from a surface-dependent endogenous devel-opmental program that tends to force the root to rotatearound its axis at the elongation zone, thereby promoting aclockwise coiling. This directional, circular growth patternis best described as a nutation, as previously discussed byMaher and Martindale (1980).

Simmons et al. (1995b) developed a similar model toexplain the wavy pattern of root growth on tilted agarsurfaces. They proposed that a right-handed circumnu-tation process generates the wavy pattern of root growthfound on tilted agar plates when corrected by gravitro-pism. Their model is illustrated in Figure 9A (Simmonset al., 1995b). However, a more careful, three-dimensional microscopical analysis of roots waving onthe surface of a tilted agar medium reveals that the wavypattern of root growth is more complex (Fig. 9, B and C).The root tends to protrude from the agar surface at eachintersection with the central axis of that wave, ratherthan at every other intersection, as anticipated from thesingle circumnutation model (Simmons et al., 1995b)(Fig. 9, B-D). These data suggest that root waving iscreated by a succession of left-handed and right-handedcircumnutations (Fig. 9D) or by other related growthprocesses accompanied by left-handed and right-handedCFRs, which are brought about by a combination of

Figure 9. Circumnutation models to explain the wavy-root pheno-type of seedlings growing on the surface of a tilted 1.5% agarmedium. A, Circumnutation model proposing that a wavy root un-dergoes a right-handed circumnutation around the hypothetical axisdrawn at the middle (Simmons et al., 1995b). B and C, Microphoto-graphs of a root (r) waving on the surface of a 1.5% agar medium (a),viewed from the top (B) or from the right side on the agar surface (C).The micrographs shown in B and C were aligned parallel to eachother, using as reference points the root hairs represented by num-bers 1 and 2. D, Model of root waving proposing that each wave isformed by a succession of left-handed and right-handed circumnu-tations around the hypothetical axes drawn on that figure. The mi-crograph shown in B was digitized, and a skeletized version of it,obtained using the Adobe Systems (Mountain View, CA) Photoshop2.5.1. software, was used to draw the models represented in A and D.

gravitropism, obstacle avoidance, and other responses tosurface-derived environmental stimuli.

The precise nature of the structural asymmetry respon-sible for the right-slanting root-growth phenotype and forthe root-tip rotation is unknown. However, it is interestingto note that the meristem initial cells giving rise to thelateral root cap cells and to the root epidermal cells followa spiral mode of cell division responsible for a spiral ar-rangement of the lateral root cap cells in A. thaliana (Baumand Rost, 1995). Therefore, the right-slanted root-growthphenotype could be a consequence of that pattern of celldivision in the root apical meristem, possibly resulting in asimilar spiral pattern of cell elongation at the elongationzone (Barlow et al., 1994). Experiments are in progress totest this model.

The analysis of mutants showing alterations in surface-dependent slanted-root growth should provide clues to themechanisms underlying that interesting root-growth phe- www.plantphysiol.orgon October 28, 2020 - Published by Downloaded from




notype. We have shown that seedlings from different A. thaliana ecotypes develop different tendencies to slant t o the right when grown on the surface of a n agar medium (Fig. 1). Preliminary data indicate that the right-slanting phenotypes of WS and Ler segregate genetically as partially dominant traits (R. Rutherford, P. Hilson, and P.H. Mas- son, unpublished data), a n d work is i n progress to determine the genetic basis of the differences between ecotypes.

The exaggerated right-slanting root-growth phenotypes of skul a n d sku2 m u t a n t plants also segregate genetically as partially dominant traits. The skul and sku2 mutat ions m a y fall within loci tha t directly determine root symmetry or, alternatively, tha t may indirectly enhance t h e right-slanted root-growth program dictated by the WS genotype. It will be interesting t o determine whether these mutat ions affect the same locus (loci) as the one(s) responsible for the differences in right- s lant ing root-growth phenotype between wild-type WS and Col seedlings, or different loci.

Finally, we have also identified additional mutants showing an exaggerated right-slanting root-growth phenotype i n the WS background, as well as mutants developing either a right-slanting or a left-slanting root-growth phenotype in the Col background (R. Rutherford, J. Sedbrook, and P.H. Masson, unpublished data). The genetic, physiological, morphological, cytological, and molecular analysis of these m u t a n t s should al low us t o bet ter under - s tand the processes regulat ing root g r o w t h and nuta t ion i n A. thaliana.

ACKNOWLEDCMENTS

We thank Tim Caspar (DuPont) and The Ohio State University Arabidopsis Biological Resource Center for provid- ing wild-type and mutant A . thaliana seeds; John Schiefelbein and Mary K. Barton for assistance with microscopy; John Kiss for directing us to the nitrocellulose membrane as a potential alternative growth surface in our experiments; and Pierre Hilson, John Sedbrook, and Kathleen Carroll for constant sup- port and discussions.

Received December 11, 1995; accepted April 24, 1996. Copyright Clearance Center: 0032-0889/96/ 111 / 0987/ 12.

LITERATURE ClTED

Barlow PW, Parker JS, Brain P (1994) Oscillations of axial plant organs. Adv Space Res 14 149-158

Baum SF, Rost TL (1995) The development of the root cap and epidermis in Arubidopsis thaliana roots: a spiral motif (abstract no. 140). In 6th International Conference on Arabidopsis Re- search, Madison, WI, p 140

Bullen BL, Best TR, Gregg MM, Barsel SE, Poff KL (1990) A direct screening procedure for gravitropism mutants of Arubi- dopsis thaliana (L.) Heynh. Plant Physiol 93: 525-531

Chapman DK, Johnsson A, Karlsson C, Brown A, Heathcote D (1994) Gravitropically-stimulated seedlings show autotropism in weightlessness. Physiol Plant 90: 157-162

Fedoroff NV, Smith DL (1993) A versatile system for detecting transposition in Arabidopsis. Plant J 3: 273-289

Feldmann KA, Malmberg RL, Dean C (1994) Mutagenesis in Arabidopsis. In EM Meyerowitz, CR Somerville, eds, Arabidop- sis. Cold Spring Harbor Laboratory Press, Cold Spring Harbor,

Fortin MA, Poff KL (1991) Characterization of thermotropism in primary roots of maize: dependence on temperature and temperature gradient, and interaction with gravitropism. Planta 184: 410414

Hart JW (1990) Tropisms and other forms of plant movement. In JW Hart, ed, Plant Tropisms and Other Growth Movements. Chapman and Hall, London, pp 1-22

Hasenstein KH, Evans ML (1988) Effects of cations on hormone transport in primary roots of Zea muys. Plant Physiol86: 890-894

Ishikawa H, Evans ML (1992) Induction of curvature in maize roots by calcium or by thigmostimulation: role of the postmitotic isodiametric growth zone. Plant Physiol 100: 762-768

Maher EP, Martindale SJB (1980) Mutants of Arabidopsis thaliana with altered responses to auxins and gravity. Biochem Genet 18:

NY, pp 137-172

1041-1053 Masson P (1995) Root gravitropism. BioEssays 1 7 119-127 Masson P, Sedbrook J, Rutherford R, Hilson P, Carroll K, Caspar

T, Simmons C, Lindsey K, Gallois P (1993) Molecular genetics of root gravitropism and waving in Arabidopsis thaliana. ASGSB Bulletin 7 26-27

Murashige T, Skoog F (1962) A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol Plant 15: 473497

Mirza JL (1987) The effects of light and gravity on the horizontal curvature of roots of gravitropic and agravitropic Arabidopsis thaliuna L. Plant Physiol 83: 118-120

Mirza JL (1988) The clinostat effect on the direction of root growth of wild type and aux-1 seedlings of Arabidopsis thaliana. Arabi- dopsis Information Service 26: 35-38

Nick P, Schafer E (1988) Interaction of gravi- and phototropic stimulation in the response of maize (Zea mays) coleoptiles. Planta 173: 213-220

Okada K (1993) Stimulus response interactions in roots (abstract no. 11). In 5th International Conference on Arabidopsis Research, Columbus, OH, p 11

Okada K, Shimura Y (1990) Reversible root tip rotation in Arabi- dopsis seedlings induced by obstacle-touching stimulus. Science 250: 274-276

Okada K, Shimura Y (1992) Mutational analysis of root gravitropism and phototropism of Arabidopsis thaliana seedlings. Aust J Plant Physiol 19 439448

Pickard B (1985) Roles of hormones in phototropism. In RP Pharis, DM Reid, eds, Hormonal Regulation of Development. 111. Role of Environmental Factors. Encyclopedia of Plant Physiology, Vol 11. Springer, Berlin, pp 364417

Simmons C, Migliaccio F, Masson P, Caspar T, Sol1 D (1995a) A nove1 root gravitropism mutant of Arabidopsis thaliana exhibiting altered auxin physiology. Physiol Plant 93: 790-798

Simmons C, Sol1 D, Migliaccio F (1995b) Circumnutation and gravitropism cause root waving in Arabidopsis thaliana. J Exp Bot

Takahashi H, Scott TK (1991) Hydrotropism and its interaction with gravitropism in maize roots. Plant Physiol 96: 558-564

4 6 143-150



Documents

Arabidopsis thaliana sku Mutant Seedlings Show Exaggerated ... · ing on tilted agar plates (Okada and Shimura, 1990), we surface-sterilized and plated seeds from the collection de-