Plant Physiol. (1996) 11 1: 671-678

A Brassinosteroid-lnsensitive Mutant in Arabidopsis thaliana Exhibits Multiple Defects in Growth and Development'

Steven D. Clouse*, Mark Langford, and Trevor C. McMorris

Department of Horticultural Science, North Carolina State University, Raleigh, North Carolina 27695 (S.D.C.); Pharmingen Co., 10975 Torreyana Road, San Diego, California 921 21 (M.L.); and Department of Chemistry,

University of California, 9500 Gilman Drive, San Diego, California 92093 (T.C.M.)

Brassinosteroids are widely distributed plant compounds that modulate cell elongation and division, but little is known about the mechanism of action of these plant growth regulators. To investi- gate brassinosteroids as signals influencing plant growth and devel- opment, we identified a brassinosteroid-insensitive mutant in Ara- bidopsis thaliana (L.) Henyh. ecotype Columbia. The mutant, termed bril , did not respond to brassinosteroids in hypocotyl elon- gation and primary root inhibition assays, but it did retain sensitivity to auxins, cytokinins, ethylene, abscisic acid, and gibberellins. The bril mutant showed multiple deficiencies in developmental path- ways that could not be rescued by brassinosteroid treatment, in- cluding a severely dwarfed stature; dark green, thickened leaves; male sterility; reduced apical dominance; and de-etiolation of dark- grown seedlings. Cenetic analysis suggests that the Bril phenotype is caused by a recessive mutation in a single gene with pleiotropic effects that maps 1.6 centimorgans from the cleaved, amplified, polymorphic sequence marker DHSl on the bottom of chromosome IV. The multiple and dramatic effects of mutation of the BRIl locus on development suggests that the BRIl gene may play a critical role in brassinosteroid perception or signal transduction.

BRs are plant-growth-promoting natural products with structural similarities to animal steroid hormones (Man- dava, 1988). When BRs are applied exogenously at nano- molar to micromolar levels in a number of test systems, they have a marked effect on cell elongation and/or pro- liferation that is distinct from that of auxins, cytokinins, and GAs, although they interact with these plant hormones and also with environmental signals, such as light and temperature, in complex ways (Adam and Marquardt, 1986). BRs are found throughout the plant kingdom and have been shown by microchemical techniques to occur endogenously at levels sufficient to promote, in planta, the physiological effects observed in bioassay systems (Sasse et al., 1992). Taken together, these facts suggest that BRs are endogenous plant-growth regulators and may serve as one of the signals controlling plant growth and development (Sasse, 1991).

Despite the large body of physiological data gathered since the first publication of a BR structure (Grove et al.,

' This work was supported in part by grant no. 92-03404 from the U.S. Department of Agriculture, National Research Initiative Competitive Grants Program (S.D.C.).

* Corresponding author; e-mail steve-clouse@ncsu.edu; fax 1-919 -515-2505.

671

1979), unequivocal proof that BRs are essential for plant growth and development has so far been unavailable. Fur- thermore, the molecular mechanism of BR action is cur- rently unclear. One might argue from structural consider- ations that they act by a mechanism similar to that of animal steroid hormones. In general, such hormones act via a soluble receptor/ligand complex that binds to nuclear sites to regulate the expression of specific genes (Evans, 1988; Mangelsdorf et al., 1995). We and others have shown that BR regulates gene expression in Glycine max (Clouse and Zurek, 1991; Clouse et al., 1992; Zurek and Clouse, 1994; Zurek et al., 1994) and Arabidopsis thaliana (Clouse et al., 1993; Xu et al., 1995) and that this modulation of specific gene expression is correlated with cell elongation, but in- formation on a BR receptor is lacking.

The use of mutants of A. thaliana that are deficient in or insensitive to plant hormones has been invaluable in dis- secting the molecular mechanisms of ABA, auxin, ethylene, and GA action (Estelle and Klee, 1994; Finkelstein and Zeevart, 1994; Ecker, 1995). Hormone-deficient mutants usually result from lesions in genes encoding hormone biosynthetic enzymes and are rescued to wild-type pheno- type by treatment with the hormone. For example, the gal, ga2, and ga3 mutants of A. thaliana are dwarfs that are restored to wild-type height by spraying with GA (Finkel- stein and Zeevart, 1994). The GA1 locus has been cloned by genomic subtraction (Sun et al., 1992) and shown to encode ent-kaurene synthetase A, a critica1 enzyme in GA biosyn- thesis (Sun and Kamiya, 1994). Hormone-insensitive mu- tants may show the same phenotype as hormone-deficient mutants but are not rescued by hormone treatment. Hormone-insensitive mutants usually result from lesions in genes encoding the hormone receptor or elements of the signal transduction pathway. The ETRl locus of A. tkaliana was identified by a screen for ethylene resistance (Bleecker et al., 1988), and positional cloning of the ETRl gene with subsequent characterization showed that the ETRl locus encoded an ethylene receptor (Chang et al., 1993).

Our goal is to identify the BR receptor and other components of the BR signal transduction pathway using

Abbreviations: BR, brassinosteroid; CAI'S, cleaved, amplified, polymorphic sequences; Col-O, Columbia ecotype; EBR, 24- epibrassinolide; EMS, ethylmethane sulfonate; Ler, Landsberg erecta ecotype; SSLP, simple sequence length polymorphism; XET, xyloglucan endotransglycosylase.

672 Clouse et al. Plant Physiol. Vol. 11 1 , 1996

a mutational analysis in A. thaliana. BRs have a range of physiological effects, but promotion of stem elongation is perhaps the best characterized (Clouse et al., 1992). Employing the inability to elongate stems in the presence of BR as a screen for BR insensitivity was found to be impractical for technical reasons (S.D. Clouse, M. Lang- ford, T.C. McMorris, unpublished data). However, like auxin, BR often inhibits primary root elongation even though shoot growth is stimulated by hormone applica- tion (Roddick and Juan, 1991). The axrl (Lincoln et al., 1990), axr2 (Wilson et al., 1990), and auxl (Pickett et al., 1990) auxin-resistant mutants and the ckrllein2 (Su and Howell, 1992; Ecker, 1995) cytokinin-resistant mutant were selected because of the ability of the mutant plants to elongate roots in the presence of hormone concentra- tions that inhibited root growth in wild-type A. thaliana. We have previously shown that BRs exhibit a similar inhibitory effect on Arabidopsis root growth and that EMS-mutagenized plants can be selected that are insen- sitive to this response (Clouse et al., 1993). We recently presented a preliminary report on the first BR-insensi- tive mutant in A. thaliana, bril, that has a severely dwarfed phenotype and reduced apical dominance and is male-sterile (Clouse and Langford, 1995). In this paper we provide additional genetic and physiological analysis of bril.

MATERIALS AND METHODS

Screening of Mutagenized Plants

M, seeds of A. thaliana (Col-O background) mutagenized with EMS were obtained from Lehle Seeds (San Antonio, TX). The M, seeds were supplied in bulked parenta1 groups of 8,000 seeds representing 1,000 M, parents. M, seeds (approximately 71,000 total) were sterilized by treatment with absolute ethanol(l0 min) followed by a 30% commer- cial bleach solution containing 0.1% Triton X-100 (20 min). Seeds were then rinsed five times in a large volume of sterile water and suspended in 0.2% agarose, and between 100 and 200 seeds were pipetted along straight lines in 150 x 100 mm Petri plates containing 1% agar, 2% SUC, and standard mineral salts (Estelle and Somerville, 1987). EBR was synthesized as previously described (McMorris and Patil, 1993) and added to the medium after autoclaving at either 10-6 or 10p7 M final concentration. The plates were placed vertically in a growth chamber at 23°C with a 16-h light/8-h dark cycle at 50 PE m-' s-l intensity. After 6 d of incubation, M, plants, whose roots elongated substantially (>9 mm), were picked and grown to maturity to produce M, seeds. A large number of M, progeny from individual M, picks were then compared to wild type for their ability to elongate roots in the presence of 10-7 M EBR. M, plants that were male-sterile, such as bril, were backcrossed to wild-type Col-O to generate the M,, which was selfed to create a segregating M, population. EBR insensitivity of the mutant phenotype was confirmed in M, individuals fol- lowed by a second round of backcrossing and selfing to generate M, mutant individuals for additional analysis.

Hormone Sensitivity Assays

Because bril is male-sterile, pure stocks of bril seed were not available. Therefore, segregating F, progeny from a backcross of bril X Col-O were used for a11 experiments. Excess F, seeds were sterilized and plated on agar medium as described above with the exception that plates were incubated for 3 d at 4°C in the dark to increase uniform germination before transfer to the growth chamber. A11 hormones were added after autoclaving and cooling of the media to <55"C. Root length of a11 seedlings was measured to the nearest 0.1 mm after 7 or 8 d using a dissecting scope with a stage micrometer, and plants were tagged individ- ually and grown in soil for 2 months to determine pheno- type. Data points represent 20 measurements 2 SE for those plants determined to be bril, compared with 20 measure- ments 5 SE for Col-O seedlings treated in the same manner.

Ethylene Experiments

Seeds were sterilized as described above and plated on 1% agar containing mineral salts (Estelle and Somerville, 1987) and 2% SUC, pH 5.8, poured into the bottom of 1-quart Mason jars whose lids were fitted with rubber septums. Jars were wrapped in aluminum foi1 and placed at 4°C for 3 d. Ethylene was injected to a final concentration of 1 p L / L and the jars were placed at 23°C for 4 d. Ethyl- ene, CO,, and O, were monitored daily by GC. Final mea- surements of root and hypocotyl length were made using a dissecting microscope with a stage micrometer.

Plant Growth

Plants for crossing and phenotypic analysis were grown in growth chambers or shaking liquid culture as previously described (Clouse et al., 1993).

Chromosome Mapping

A single plant showing the Bril phenotype (Col-O back- ground) was pollinated with Ler pollen yielding a fertile F, hybrid, which was selfed to generate a segregating F, population of 70 mutant and 212 wild-type individuals. DNA was isolated from individual mutant plants by plac- ing the entire plant in a microfuge tube with a small amount of liquid nitrogen, followed by grinding with a plastic pestle. After sublimation of the nitrogen, 500 pL of buffer (100 mM Tris, 100 mM Na,EDTA, 250 mM NaC1, pH 8.0) were added and grinding was continued. Fifty micro- liters of 10% sarkosyl were then added, followed by 5 pL of proteinase K (10 mg/mL), and the tubes were incubated at 55°C for approximately 1 h. After centrifugation (14,000 rpm, 10 min), the supernatant was extracted twice with pheno1:chloroform:isoamyl alcohol (24:24:1, v / v), and 0.35 volumes of 5 M NaCl and 2 volumes of 100% ethanol were added to the final aqueous phase to precipitate DNA free from polysaccharides (Mettler, 1987). The pelle'ts were washed twice in 75% ethanol and resuspended in 100 pL of sterile water, and 1.5 PL were used per 2 0 - ~ L PCR reaction. DNA from individual plants showing the wild-type phe- notype was isolated in a similar manner except that a small

A Brassinosteroid-lnsensitive Mutant in Arabidopsis thaliana 673

mortar and pestle was used for grinding whole plants andthe buffer was increased to 5 mL/g tissue. Genotypes of theF2 wild-type plants were determined by F3 progeny anal-ysis. Linkage to known molecular markers was determinedby SSLP (Bell and Ecker, 1994) and CAPS (Konieczny andAusubel, 1993) analysis using primers from Research Ge-netics (Huntsville, AL).

RESULTS

Phenotype and Genetics of the br'(\ Mutant

From a screen of approximately 70,000 EMS-mutagenizedM2 seedlings of A. thalicma, an individual was selected thatshowed root elongation in the presence of 10~6 or 10~7 MEBR, a concentration that inhibited the elongation of wild-type Col-0 roots by up to 75%. This mutant consistentlyexhibited long roots and very short hypocotyls when placedon various EBR concentrations and showed a striking pheno-type when grown to maturity, which suggested that a funda-mental pathway controlling growth and development hadbeen compromised by the mutation. Figure la shows that thebril mutant is an extreme dwarf that reaches less than 10% ofthe height of wild-type Col-0 plants grown under the sameconditions. Two-month-old bril plants did not exceed 1.5 cm,whereas Col-0 plants of the same age were greater than 15 cmin height. Figure 1, b and c, shows that the leaf morphology ofthe bril mutant was distinctly different from that of the wildtype. Rosette leaves were shorter in the mutant than in the

wild type, had a thickened, curled appearance, and weredarker green, and their petioles failed to elongate. Prelimi-nary microscopic evaluation of the bril mutant suggests thatreduced cell size rather than cell number is responsible for thedwarf phenotype (data not shown). Figure Id shows that4-month-old plants had a bushy appearance, indicating re-duced apical dominance when compared to mature wild-typeplants.

bril plants failed to yield viable seeds even when self-pollinated by hand and grown at either 23 or 15 to 18°C.However, when pollen from Col-0 or Ler wild-type plantswere placed on bril stigmas, siliques containing fertile F,seeds developed, indicating that the bril mutant is mostlikely male-sterile but female-fertile. We have not yet ex-amined the mechanism of male sterility. The F, resembledwild type in appearance, and selfed F, plants gave an F2population that segregated in a 3:1 wild-type-to-mutantphenotypic ratio, suggesting that the Bril phenotype iscaused by a single recessive Mendelian allele. Table I sum-marizes the F2 phenotypic segregation of a number of brilX Col-0 and bril X Ler crosses. In all cases examined, thoseplants showing the dwarf phenotype also were BR insen-sitive, as determined by the root-elongation assay. More-over, after two generations of backcrossing interspersedwith two generations of selfing, there was no segregationof the dwarf phenotype and BR insensitivity.

To determine the map position of the bril locus, DNAwas isolated from individual F2 progeny of a bril X Lercross and analyzed for linkage to SSLP (Bell and Ecker,

Figure 1. Phenotype ot the far/1 mutant, a, Two-month-old mutant (left) and wild-type (right) plants grown in 50-mLcentrifuge tubes in a 23°C growth chamber (16-h light/8-h dark), b and c, Side and top views, respectively, of a 2-month-oldfar/I mutant plant, d, The same plant after 4 months. All bars = 1 cm.

674 Clouse et ai. Plant Physiol. Vol. 11 1, 1996

Table 1. Phenotypic ratios in segregating Fz populations of br i l x COLO or br i l X Ler crosses

Cross No. Wild Type Mutant Ratio

1 (bril X COLO, F2) 2 (bril X Col-O, F,) 3 (bril x COLO, F,) 4 (bril x Col-O, F,) 5 (bril X COLO, F,) 6 (bril X COI-O, F,) 7 (bril X Col-O, F,) 8 (bril x Col-O, F,) 9 (bril x COLO, F,) Total 10 (bril X Ler, F,)

480 21 2 234 289

1585 427 661 192 220

601 4080 1

164 2.93:l 70 3.03:l 87 2.69:l 83 3.48:l

495 3.20:l 127 3.36:l 214 3.09:l 70 2.74:l

103 2.1 4:l 1310 3.1 1 :1 21 3 3.03:l

1994) and CAPS (Konieczny and Ausubel, 1993) markers. No linkage between the Bril phenotype and the following SSLP markers was observed: nga63, nga248, nga280, and ATPase on chromosome I; nga168 on chromosome 11; nga126, nga162, and nga6 on chromosome 111; and ctrl, nga225, ngal51, and ngal06 on chromosome V. The Bril phenotype, however, was found to be tightly linked to the CAPS marker DHSl on the bottom of chromosome IV (1 recombinant out of 63 mutant F, plants examined), with a map distance of 1.61 centimorgans estimated by the method of Kosambi and 1.59 centimorgans by the method of Haldane (Koornneef and Stam, 1992). Comparison of the map position and phenotype of bril with those of other dwarf and hormone-insensitive Arabidopsis mutants indi- cates that bril is not allelic to other mutants in the pub- lished literature. A collection of unpublished T-DNA dwarfs is also nonallelic to bril based on map position or physiological responses to BR (K. Feldmann, personal communication).

Hormone Sensitivity of the bril Mutant

The bril mutant was identified by the presence of long roots when grown on a concentration of EBR that inhib- ited wild-type root elongation. Figure 2 shows that 0.5 PM EBR caused, on average, 82% inhibition of root elon- gation in wild-type Col-0 when compared to a buffer control. In a segregating F, population from a cross of the bril mutant with wild-type Col-O, those individuals showing Col-O phenotype gave a similar response to the wild-type parent with respect to inhibition of root elon- gation by 0.5 p~ EBR. However, those individuals with the Bril phenotype were completely insensitive to the effect of 0.5 ~ L M EBR on root elongation. This insensitiv- ity was confirmed over a wide range of EBR concentra- tions, as shown in Figure 3. Concentrations of EBR greater than 5.0 IIM resulted in a rapid decrease in root length of wild-type Col-O seedlings, reaching a maxi- mum inhibition at 0.5 WM. In contrast, bril mutants main- tained the same root length on EBR concentrations rang- ing from 10.0 nM to 0.5 p~ as they did when they were grown on control medium. An EBR concentration of 1.0 p~ did result in a slight inhibition of bril root elonga- tion, but the mutant roots were still over five times longer than roots of Col-O seedlings grown on 1.0 p~

EBR. EBR concentrations greater than 1.0 p~ were not attempted so as to preserve the limited amounts of the compound available.

The inhibition of root elongation in wild-type Arabidop- sis seedlings seems to be a general effect of most plant hormones tested (Bleecker et al., 1988; Lincoln et al., 1990; Pickett et al., 1990; Wilson et al., 1990; Su and Howell, 1992; Clouse et al., 1993). To determine if the bril mutation confers insensitivity specifically to BRs or results in insen- sitivity to multiple hormones, as has been observed for axrl, axr2, auxl, and ckrl, we compared root elongation of Col-O and bril seedlings on a number of plant hormones. Figure 4 shows that concentrations of 2,4-D from 5 to 100 nM resulted in even greater inhibition of bril root elonga- tion than in wild-type, and that at 0.5 and 1.0 p~ the inhibition was equally severe in both genotypes. Therefore, the bril mutant retains complete sensitivity to 2,4-D. Figure 5 shows that bril and Col-O are equally sensitive to 0.5 FM IAA, 6-benzylaminopurine, and kinetin, and that bril is hypersensitive to 0.5 p~ ABA. bril appears to be slightly less sensitive to 0.5 WM GA, than Col-O, but nowhere does it approach the insensitivity observed with 0.5 p~ EBR (cf. Fig. 2).

To determine the sensitivity of bril to ethylene, dark- grown seedlings were used in an experiment similar to that described by Bleecker et al. (1988) for the analysis of the etrl ethylene-insensitive mutant. It is interesting that the bril mutant grown in the dark in air had short, thickened hypocotyls and fully opened cotyledons characteristic of the det2 and cop class of photomorphogenic mutants (Chory and Susek, 1994). Figure 6 shows that even though

30

CONTROL 0.5, M 24-EPI BR

WILD-TYPE bril X COL. F2 bril X COL. Fz COLUMBIA W.T. PHENOTYPE MUTANT PHENOTYPE

Figure 2. Effect of EBR on root length in F, progeny of a bril X Col-O backcross. Because bril is male-sterile, pure stocks of bril seed were not available. Therefore, segregating F, progeny from a backcross of bril X COLO were plated on agar medium with or without 0.5 p~ EBR as described in "Materials and Methods." Root length was measured under a dissecting microscope after 7 d. The graph repre- sents the mean of 20 measurements t SE. All seedlings with long roots on EBR showed the bril dwarf phenotype when mature.

A Brassinosteroid-lnsensitive Mutant in Arabidopsis thaliana 675

----t Columbia (w.t . ) * br i l (mutant)

O 05 1 5 1 5 10 5 0 100 500 1000

24-EPIBRASSINOLIDE CONCENTRATION (nM)

Figure 3. Dose-response curve of root elongation on EBR. Segregat- ing F, progeny from a backcross of bril X COLO were plated on agar medium with the indicated concentration of EBR. Root length of all seedlings was measured after 7 d and plants were tagged individually and grown in soil for 2 months to determine phenotype. Data points represent 20 measurements 2 SE for those plants determined to be bril, compared with 20 measurements -t SE for COLO seedlings treated in the same manner.

roots of bril plants grown in the dark in air had shorter roots than the wild type, both Col-O and bril root elonga- tion were equally inhibited by approximately 60% upon treatment with 1.0 pL/L ethylene, indicating that bril re- tains sensitivity to ethylene. Therefore, of a11 compounds tested, bril shows marked insensitivity only to BRs.

Finally, to determine if the Bril phenotype could be rescued by hormone treatment, plants were sprayed twice weekly with 1.0 PM GA, or 1.0 PM EBR, or were grown for 21 d in shaking liquid culture containing 1.0 ~ L M GA, or 1.0 PM brassinolide. None of the treatments had any observ- able effect on the Bril phenotype. Mutant plants retained the same long roots, short hypocotyls, and dwarf stature with or without hormone treatment, whereas wild-type plants showed substantial hypocotyl elongation and inhi- bition of root growth with GA, or BR treatment (data not shown).

D I SC U SSI ON

The inhibitory effect of BRs on primary root elongation in wild-type A. thaliana proved to be a facile and reproduc- ible screen for BR insensitivity (Clouse et al., 1993). While the root assay allowed the initial identification of bril, mutant plants exhibited a phenotype that extended far beyond the simple root response to BR, which suggested that the entire developmental program had been altered. Analysis of severa1 thousand F, progeny showed no seg- regation between root elongation in the presence of BR and the altered developmental morphology. Therefore, it is most likely that the short stature, altered leaf structure,

reduced apical dominance, and male sterility are caused by a mutation in a single gene with pleiotropic effects that also confers insensitivity to BRs. However, after two back- crosses and two rounds of selfing, it cannot be completely ruled out that the dwarf phenotype and the BR insensitiv- ity are caused by independent mutations in genes that are very tightly linked.

Another Arabidopsis dwarf that has a very similar phe- notype to bril is det2, a de-etiolated mutant that shows characteristics of a light-grown plant even when it is grown in the dark (Chory and Susek, 1994). Very recently, the DET2 locus has been cloned by chromosome walking and has been found to have significant homology to a mamma- lian 5a steroid reductase (Li et al., 1996), an enzyme in- volved in steroid metabolism. Moreover, the Det2 mutant phenotype is rescued by BR treatment, indicating that det2 is a BR biosynthetic mutant (Li et al., 1996). The fact that det2 (a single locus that is nonallelic to bril and that can be rescued to wild type by BR treatment) has a closely related phenotype to bril strongly suggests that the single locus conferring insensitivity to BR is the same locus controlling the developmental aspects of the Bril phenotype. In the cases of both det2 and br i l , it is likely that the dwarf phenotype is caused by a lack of BR action, either through reduced synthesis or reduced response, respectively.

It is not surprising that both a BR-insensitive and a BR-deficient mutant show a dwarfed growth habit, since BRs have long been known to promote stem elongation. BR-promoted elongation of young vegetative tissue has been observed in mung bean, azuki bean, and pea epicotyls (Yopp et al., 1981; Gregory and Mandava, 1982; Mandava,

h

E E Y

I I-

Lu J

ò O a!

40 - Columbia (w.t.) - bril (mutant)

30

20

10

O . . . . O .O5 .I .5 1 5 10 50 100 5 0 0 1000

2,4-D CONCENTRATION (nM)

Figure 4. Dose-response curve of root elongation on 2,4-D. As in Figure 3 , segregating F, progeny from a backcross of bril X Col-O were plated on agar medium but with the indicated concentration of 2,4-D replacing EBR. Root length of all seedlings was measured after 7 d and plants were tagged individually and grown in soil for 2 months to determine phenotype. Data points represent 20 measure- ments t SE for those plants determined to be bril, compared with 20 measurements -t- SE for Col-O seedlings treated i n the same manner.

Clouse et al. Plant Physiol. Vol. 11 1 , 1996 676

100

T O E 80

8 Y O 5 60 W O o! W

I 40 I- o 2 W A

5

ô 20 s O

COLUMEIA (W.T.) br i l (MUTANT)

T

FpM'0.5 FM ' 0.5pM ' 0.5 pM ' 0.5 ,M ' 100 p M

=A3 =A, IAA BAP KlNETlN ABA

Figure 5. Effect of other hormones on root elongation. Segregating F progeny from a backcross of bril X Col-O were plated on agar medium as described in Figure 3 with the indicated concentration of hormone. Root length of all seedlings was measured after 8 d and plants were tagged individually and grown in soil for 2 months to determine phenotype. Data points represent 20 measurements t SE

for those plants determined to be bril, compared with 20 measure- ments 2 SE for Col-O seedlings treated in the same manner.

1988); bean, sunflower, and cucumber hypocotyls (Man- dava et al., 1981; Katsumi, 1985); Arabidopsis peduncles (Clouse et al., 1993); and maize mesocotyls and wheat coleoptiles (Yopp et al., 1981; Sasse, 1985). We have previ- ously shown that BR is a potent enhancer of epicotyl elon- gation in soybeans and we have found that BRs modulated cell-wall mechanical properties and gene expression in this system (Clouse et al., 1992; Zurek et al., 1994). It is well known that auxin also promotes stem elongation (Taiz, 1984), and comparisons of auxin- versus BR-stimulated growth showed that the kinetics of elongation and effects on gene expression were quite different for BR and auxin (Clouse et al., 1992; Zurek and Clouse, 1994).

Also of interest is whether BR-induced growth, like auxin-induced cell elongation, is ultimately mediated by cell-wall loosening. The biochemical basis of wall relax- ation was predicted nearly 20 years ago to involve en- zymes that cleave and rejoin cell-wall polymers, partic- ularly xyloglucans (Albersheim, 1976). XET is an enzyme that specifically cleaves a xyloglucan chain and transfers a fragment of that chain to an acceptor xyloglucan (Fry, 1995). Severa1 groups (Smith and Fry, 1991; Nishitani and Tominaga, 1992; Fry et al., 1992; Fanutti et al., 1993) have proposed a wall-loosening function for this enzyme during cell elongation, or it may have other important roles in elongation such as xyloglucan biosynthesis or inte- gration of new xyloglucan polymers into the wall (McQueen-Mason et al., 1993).

We have cloned a BR-regulated gene from soybean called BRUl that shows high sequence homology to XETs from other species (Zurek and Clouse, 1994) and whose

transcript levels are correlated with the extent of BR- promoted stem elongation and increases in plastic extensi- bility of the cell wall (Zurek et al., 1994). We recently found that the recombinant BRUl protein has XET activity and that extractable XET activity in soybean epicotyls is in- creased by BR treatment (W. Romanow, R. Smith, S.D. Clouse, unpublished data). Therefore, XETs may play a role in BR-modulated stem elongation in soybean.

A similar story is unfolding in A. thaliana. TCH4, which is 70% identical to BRUl at the amino acid level, shows increased transcript levels in BR-treated Arabidopsis tissue and has also been found to encode a XET (Xu et al., 1995). Moreover, transgenic A. thaliana plants harboring TCH4 promoter:GUS fusions show highest reporter gene expres- sion in elongating tissue (Xu et al., 1995). Therefore, a likely explanation for the dwarf phenotype of bril is that BR insensitivity results in reduced XET activity (along with reduced expression of other unidentified molecular com- ponents of the cell-elongation machinery), which results in shortened cells. We are currently examining TCH4 expres- sion and XET activity in the bril mutant.

The specificity of the hormone insensitivity of bril lends additional support to the importance of BRs in growth and development. Based on the root-elongation assay, bril is as sensitive to auxins, GAs, cytokinins, ABA, and ethylene as wild-type Col-O. Therefore, if insensitivity to BR in the root assay and the developmental phenotype of bril are caused by a single gene, it is compelling evidence that BR activity is essential for growth and development. Furthermore, be- sides cell elongation, BRs must play a role in male fertility and apical dominance. Based on early physiological stud- ies, it was suggested that the effects of BR are mediated through auxin or that BR increases tissue sensitivity to endogenous auxins (Mandava, 1988). We previously showed that BR can act independently of auxin in elongat- ing soybean epicotyls (Clouse et al., 1992) and that auxin-

h

E E - I !- (3 2 W -1

ò

f in4 T

2.0 _i O 0 1 .o

AIR W E N E

T

.~

Columbia w.t. b r i l mutant

Figure 6. Effect of ethylene on root elongation. Segregating F, prog- eny from a backcross of bril X Col-O were plated on agar medium and placed at 4°C for 3 d. Ethylene or air was then injected to a final concentration of 1 pL/L and the jars were placed at 23°C for 4 d in the dark. Final measurements of root length were made using a dissecting microscope with a stage micrometer. Data points repre- sent the mean of 20 measurements 5 SE.

A Brassinosteroid-lnsensitive Mutant in Arabidopsis thaliana 677

insensitive mutants such as axrl (Lincoln et al., 1990) and dgt (Daniels et al., 1989) retained sensitivity to BR (Clouse et al., 1993; Zurek et al., 1994). The complete sensitivity of bril to auxin demonstrated here provides additional evi- dente of the independence of BR and auxin action.

Our primary motivation in studying BR-insensitive mutants is to identify the BR receptor and clarify the BR signal transduction pathway. Hormone insensitivity generally results from one of three mechanisms: (a) a mutation in the receptor prevents the hormone from binding, or the receptor fails to become activated even when bound to the ligand; (b) the receptor i s functional but a mutation in a critical step in the signal transduction pathway has occurred; or (c) the applied hormone is not actually the active endogenous compound and an en- zyme that converts the applied hormone to the active form has been mutated. In view of the proven biological activity of numerous forms of BRs, including brassinol- ide, EBR, 28-homobrassinolide, and castesterone (Mc- Morris et al., 1994), and the fact that we have shown insensitivity of the bril mutant to more than one BR (S.D. Clouse, M. Langford, T.C. McMorris, unpublished data), option c is unlikely and bril probably represents a mu- tation in class a or b. A collaborative effort between the Chory and Sakurai groups to determine endogenous BRs in Arabidopsis will help to clarify this issue (J. Chory, personal communication).

Original views of the mechanism of animal steroid hor- mone action assumed a single protein receptor 1 signal transduction model in which the steroid ligand diffused through the cell membrane, bound to the steroid receptor in the cytosol or nucleus, and the receptor 1 ligand complex interacted directly with hormone-responsive elements in the promoters of steroid-regulated genes (Evans, 1988). It is now known that steroid receptors interact with heat-shock proteins in the cytosol, are regulated by phosphorylation, and can complex with nonsteroidal transcription factors in the nucleus (Beato et al., 1995). Therefore, although the steroid signal transduction pathway in animals is more complex than previously thought, it is still quite short. If BRs have a similar mechanism of action in plants, a small set of BR-insensitive mutants will define the entire signal transduction pathway. Even if the BR pathway differs from that of animals and is more complex, isolation of hormone- insensitive mutants in Arabidopsis is still a very fruitful approach, as has been amply demonstrated in ethylene (Ecker, 1995) and ABA (Finkelstein and Zeevart, 1994) sig- na1 transduction studies.

We are engaged currently in the positional cloning of the BRIl locus, which is greatly facilitated by the recent pub- lication of a complete physical map of chromosome IV (Schmidt et al., 1995). The dramatic and pleiotropic effects of the bril mutation suggests that the B X l l gene lies very early in the signal perceptionl transduction pathway. Therefore, cloning and characterization of the BRIl locus should lead to a major advance in our understanding of BR action. The BR-insensitive (bril) and BR-deficient (def2) mutants will be valuable resources for further understand- ing the molecular mode of BR action. Furthermore, the

phenotype of these mutants provides strong genetic evi- dente that BRs are essential for normal plant growth and development.

ACKNOWLEDCMENTS

Thanks to Gregory Scott for excellent technical assistance and to Sylvia Blankenship for performing the GC. We also thank Joanne Chory, Ken Feldmann, Punita Nagpal, and Jason Reed for helpful discussions.

Received February 7, 1996; accepted April 4, 1996. Copyright Clearance Center: 0032-0889/96/111 /O671 /OS.

LITERATURE CITED

Adam G, Marquardt V (1986) Brassinosteroids. Phytochemistry

Albersheim P (1976) The primary cell wall. In J Bonner, JE Varner, eds, Plant Biochemistry, Ed 3. Academic Press, New York, pp

Beato M, Herrlich P, Schutz G (1995) Steroid hormone receptors: many actors in search of a plot. Cell 83: 851-857

Bell CJ, Ecker JR (1994) Assignment of 30 microsatellite loci to the linkage map of Arabidopsis. Genomics 19: 137-144

Bleecker AB, Estelle MA, Somerville C (1988) Insensitivity to ethylene conferred by a dominant mutation in Arabidopsis thali- unu. Science 241: 1086-1089

Chang C, Kwok SF, Bleecker AB, Meyerowitz EM (1993) Arabi- dopsis ethylene-response gene ETRI: similarity of product to two-component regulators. Science 262: 539-544

Chory J, Susek RE (1994) Light signal transduction and the control of seedling development. In E Meyerowitz, C Somerville, eds, Arabidopsis. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp 579-614

Clouse SD, Langford M (1995) A brassinosteroid-insensitive mu- tant in Arabidopsis thaliana (abstract no. 375). Plant Physiol 108: s-81

Clouse SD, Langford M, Hall AF, McMorris TC, Baker ME (1993) Physiological and molecular effects of brassinosteroids on Ara- bidopsis thuliuna. J Plant Growth Regul 12: 61-66

Clouse SD, Zurek D (1991) Molecular analysis of brassinolide action in plant growth and development. In HG Cutler, T Yokota, G Adam, eds, Brassinosteroids: Chemistry, Bioactivity and Applications. ACS Symposium Series 474. American Chem- ical Society, Washington, DC, pp 122-140

Clouse SD, Zurek DM, McMorris TC, Baker ME (1992) Effect of brassinolide on gene expression in elongating soybean epicotyls. Plant Physiol 100: 1377-1383

Daniels SG, Rayle DL, Cleland RE (1989) Auxin physiology of the tomato mutant diageotropica. Plant Physiol 91: 804-807

Ecker JR (1995) The ethylene signal transduction pathway in plants. Science 268: 667-675

Estelle M, Klee H (1994) Auxin and cytokinin. In E Meyerowitz, C Somerville, eds, Arabidopsis. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp 555-578

Estelle MA, Somerville CR (1987) Auxin-resistant mutants of Arabidopsis witli an altered morphology. Mo1 Gen Genet 206:

Evans RM (1988) The steroid and thyroid hormone receptor su- perfamily. Science 240: 889-895

Fanutti C, Gidley MJ, Reid JS (1993) Action of a pure xyloglucan endo-transglycosylase from the cotyledons of germinated nas- turtium seeds. Plant J 3: 691-700

Finkelstein RR, Zeevart JAD (1994) Gibberellin and abscisic acid biosynthesis and response. In E Meyerowitz, C Somerville, eds, Arabidopsis. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp 523-554

Fry SC (1995) Polysaccharide-modifying enzymes in the plant cell wall. Annu Rev Plant Physiol Plant Mo1 Biol 46: 497-520

25: 1787-1799

225-274

200-206

678 Clouse et al. Plant Physiol. Vol. 11 1, 1996

Fry SC, Smith RC, Renwick KF, Martin DJ, Hodge SK, Matthews KJ (1992) Xyloglucan endotransglycosylase, a new wall-loosening enzyme activity from plants. Biochem J 282: 821-828

Gregory LE, Mandava NB (1982) The activity and interaction of brassinolide and gibberellic acid in mung bean epicotyls. Physiol Plant 54: 239-243

Grove MD, Spencer GF, Rohwedder WK, Mandava N, Worley JF, Warhen JD (1979) A unique plant growth promoting steroid from Brassica napus pollen. Nature 281: 216-217

Katsumi M (1985) Interaction of a brassinosteroid with IAA and GA, in the elongation of cucumber hypocotyl sections. Plant Cell Physiol 26: 615-625

Konieczny A, Ausubel FM (1993) A procedure for mapping Ara- bidopsis mutations using co-dominant ecotype-specific PCR- based markers. Plant J 4: 403410

Koornneef M, Stam P (1992) Genetic analysis. In C Koncz, N Chua, J Schell, eds, Methods in Arabidopsis Research. World Scientific, Singapore, pp 83-99

Li J, Nagpal P, Vitart V, McMorris TC, Chory J (1996) A role for brassinosteroids in light-dependent development of Arabidopsis. Science 272: 398-401

Lincoln C, Britton JH, Estelle M (1990) Growth and development of the a x r l mutants of Arabidopsis. Plant Cell 2: 1071-1080

Mandava NB (1988) Plant growth-promoting brassinosteroids. Annu Rev Plant Physiol Plant Mo1 Biol 39: 23-52

Mandava NB, Sasse JM, Yopp JH (1981) Brassinolide, a growth- promoting steroidal lactone. 11. Activity in selected gibberellin and cytokinin bioassays. Physiol Plant 53: 453-461

Mangelsdorf DJ, Thummel C, Beato M, Herrlich P, Schutz G, Umesono K, Blumber B, Kastner P, Mark M, Chambon P, Evans RM (1995) The nuclear receptor superfamily: the second decade. Cell83: 835-839

McMorris TC, Patil PA (1993) Improved synthesis of 24- epibrassinolide from ergosterol. J Org Chem 58: 2338

McMorris TC, Patil PA, Chavez RG, Baker ME, Clouse SD (1994) Synthesis and biological activity of 28-homobrassinolide and analogs. Phytochemistry 3 6 585-589

McQueen-Mason SJ, Fry SC, Durachko DM, Cosgrove DJ (1993) The relationship between xyloglucan endotransglycosylase and in-vitro cell wall extension in cucumber hypocotyls. Planta 190:

Mettler IJ (1987) A simple and rapid method for minipreparation of DNA from tissue cultured plant cells. Plant Mo1 Biol Rep 5:

Nishitani K, Tominaga R (1992) Endo-xyloglucan transferase, a nove1 class of glycosyltransferase that catalyzes transfer of a segment of xyloglucan molecule to another xyloglucan mole- cule. J Biol Chem 267: 21058-21064

Pickett FB, Wilson AK, Estelle M (1990) The auxl mutation of Arabidopsis confers both auxin and ethylene resistance. Plant Physiol94: 1462-1466

327-331

346-349

Roddick JG, Juan M (1991) Brassinosteroids and root develop- ment. In HG Cutler, T Yokota, G Adam, eds, Brassinosteroids: Chemistry, Bioactivity and Applications. ACS Symposium Series 474. American Chemical Society, Washington, DC, pp

Sasse JM (1985) The place of brassinolide in the sequential re- sponse to plant growth regulators in elongating tissue. Physiol Plant 63: 303-308

Sasse JM (1991) The case for brassinosteroids as endogenous plant hormones. In HG Cutler, T Yokota, G Adam, eds, Brassino- steroids: Chemistry, Bioactivity and Applications. ACS Sympo- sium Series 474. American Chemical Society, Washington, DC,

Sasse JM, Yokota T, Taylor PE, Griffiths PG, Porter QN, Cam- eron DW (1992) Brassinolide-induced elongation. In C Karssen, L van Loon, D Vreugdenhil, eds, Progress in Plant Growth Regulation. Kluwer Academic, Dordrecht, The Netherlands, pp

Schmidt R, West J, Love K, Lenehan Z, Lister C, Thompson H, Bouchez, Dean C (1995) Physical map and organization of Ara- bidopsis thaliana chromosome 4. Science 270: 480-483

Smith RC, Fry SC (1991) Endotransglycosylation of xyloglucans in plant cell suspension cultures. Biochem J 279: 529-535

Su W, Howell SH (1992) A single genetic mutant, c k r l , defines Arabidopsis mutants in which root growth is resistant to low concentrations of cytokinin. Plant Physiol99: 1569-1574

Sun T, Goodman HM, Ausubel FM (1992) Cloning the Arabidop- sis G A 1 locus by genomic subtraction. Plant Cell 4: 119-128

Sun T, Kamiya Y (1994) The Arabidopsis GA1 locus encodes the cyclase ent-kaurene synthetase A of gibberellin biosynthesis. Plant Cell 6: 1509-1518

Taiz L (1984) Plant cell expansion: regulation of cell wall mechan- ical properties. Annu Rev Plant Physiol 35: 585-657

Wilson AK, Pickett FB, Turner JC, Estelle M (1990) A dominant mutation in Arabidopsis confers resistance to auxin, ethylene and abscisic acid. Mo1 Gen Genet 222: 377-383

Xu W, Purugganan MM, Polisensky DH, Antosiewicz DM, Fry SC, Braam J (1995) Arabidopsis TCH4, regulated by hormones and the environment, encodes a xyloglucan endotransglycosy- lase. Plant Cell 7: 1555-1567

Yopp JH, Mandava NB, Sasse JM (1981) Brassinolide, a growth- promoting steroidal lactone. I. Activity in selected auxin bioas- says. Physiol Plant 53: 445-452

Zurek DM, Clouse SD (1994) Molecular cloning and characteriza- tion of a brassinosteroid-regulated gene from elongating soy- bean epicotyls. Plant Physiol 104: 161-170

Zurek DM, Rayle DL, McMorris TM, Clouse SD (1994) Investi- gation of gene expression, growth kinetics and wall extensibility during brassinosteroid-regulated stem elongation. Plant Physiol

231-245

pp 158-166

319-325

104: 505-513