Integrated into a Stress Response Pathway

The response of Arabzdqsis thaliana etiolated seedlings to the plant hormone ethylene is a conspicuous phenotype known as the triple response. We have identified genes that are required for ethylene perception and response by isolating mutants that fail to display a triple response in the presence of exogenous ethylene. Five new complementation groups have been identified. Four of these loci, designated Sn4, Sn5, Sn6 and Sn7, are insensitive to ethylene. The fifth complementation group, eirl, is defined by a novel class of mutants that have agravitropic and ethylene-insensitive roots. Double-mutant phenotypes have allowed the positioning of these loci in a genetic pathway for ethylene signal transduction. The ethylene-response pathway is defined by the following loci: ETRI, EIN4, CTRI, EIN2, EIN3, ELV5, EIN6, EIN7, EIRI, AUXl and HLSI. ctrl-I is epistatic to etrl-3 and ein4, indicating that CTRI acts after both ETRI and EIN4 in the ethylene-response pathway. Mutations at the EIN2, EIN3, EIN5, EIN6 and E m 7 loci are all epistatic to the ctrl seedling phenotype. The EIRI and AUXl loci define a root-specific ethylene response that does not require EIN3 or EIN5 gene activity. HLSl appears to be required for differential cell growth in the apical hook. The EIRI, AUXl and HLSl genes may function in the interactions between ethylene and other plant hormones that occur late in the signaling pathway of this simple gas. E THYLENE is a simple gaseous molecule that regulates many complex processes in plant growth and physiology. The effect of ethylene on pea seedling development, studied by DIMITRY NELJUBOV in 1901, was the first demonstration that a gas could act as a signaling molecule in a biological system (described in ABELES et al. 1992) . Ethylene since has been shown to play a fundamental role in fruit ripening, germination, sex determination, leaf abscission, flower senescence, responses to mechanical stress and several different pathogenic responses (MATTOO and SUTTLE 1991; ABELES et al. 1992) . These effects are regulated by factors that control either the biosynthesis or the perception of this gas. The biosynthetic pathway for ethylene has been established and several of the rate-limiting enzymes have been cloned (YANC and HOFFMAN 1984; SATO and THEOLOGIS 1989; SPANU et al. 1991). In contrast, the mechanisms by which plants perceive and respond to this hormone are just beginning to be revealed ( K ~ E B E R and ECKER 1993) . In the presence of ethylene, Arabidopsis thaliana seedlings undergo dramatic morphological changes referred to as the triple response (KNIGHT et al. 1910). This seedling phenotype consists of an exaggerated curCorresponding aulhur: Joseph R. Ecker, Plant Science Institute, Department of Biology, University of Pennsylvania, Philadelphia, PA 191046018. E-mail: [email protected] ' Present address: University of Illinois, Laboratory for Molecular Biology, Chicago, IL 60607-7020. Genetics 139: 1393-1409 (March, 1995) vature of the apical hook, radial swelling of the hypocotyl and an inhibition of root and hypocotyl elongation ( BLEECKER et al. 1988; GUZMAN and ECJER 1990) . It long has been suggested that these changes, which result from stress-induced ethylene, are necessary for the seedling to penetrate the soil without damaging the apical meristem (DARWIN and DARWIN 1881 ) . The formation of the triple response relies on the plant's ability to respond to ethylene. The silver ion ( A g f ) , a very effective noncompetitive inhibitor of ethylene action (BEER 1976), and transcyclooctene, a strong competitive inhibitor of ethylene responses ( SISLER 1990), inhibit the triple-response phenotype in A. thaliana ( GUZMAN and ECKER 1990) . One mechanism by which ethylene may regulate hypocotyl and root length is by its effects on the structure of the extracellular matrix ( MELBAUM and BURG 1971; LANG et al. 1982). Ethylene treatment of pea epicotyls causes the orientation of cellulose microfibrils in the secondary cell wall to shift from a transverse to a longitudinal direction (MELBAUM and BURG 1971; LANC et al. 1982; YUAN et al. 1994). This change may be responsible for the increased lateral growth of these cells. Ethylene also affects the polar transport of the auxin indole acetic acid (IAA) , thus controlling the levels of this hormone in various target tissues (BURG and BURG 1967; BEYER 1973). Modified levels of auxin may result in reduced cell elongation and, therefore, reduce the length of the seedling. Ethylene also has been shown 1394 G. Roman et al. to inhibit the rate of cell division in pea seedlings (&EL BAUM and BURG 1972). The effect of ethylene on the apical hook appears complex and may involve intricate interactions with auxin to coordinate the rate of elongation between cells throughout he hook structure ( SCHWARK and SCHIERLE 1993; A. LEHMAN and J. R. ECKER, unpublished results). Some of these and other effects of ethylene on plant growth and development are likely to be mediated by changes in gene expression. Several genes whose transcription is induced by ethylene have been identified ( ECKER and DAVIS 1987; LAWTON et al. 1994; MELLER et al. 1993) . Complex patterns of ethylene induced gene expression have been found in tomato fruit ripening (LINCOLN and FISHER 1988), and there is some evidence that ethylene also acts posttranscriptionally ( THEOLOGIS 1992 ) . Mutations that affect the triple response in Arabidopsis have defined a large number of genes involved in the synthesis and perception of ethylene ( BLEECKER et al. 1988; GUZMAN and ECKER 1990; HARPHAM et al. 1991; VAN DER STRAETEN et al. 1993). Four mutant loci with an ethylene-insensitive phenotype have been identified and characterized genetically: etrl ( BLEECKER et al. 1988) , ein2 ( GUZMAN and ECKER 1990) , ein3 (KIEBER et al. 1993; M. ROTHENBERG, . ROMAN and J. R. ECKER, unpublished data) and a in l (VAN DER STRAETEN et al. 1993) . Additional screens identified mutants that overproduce ethylene ( e t o l , et02 and eto3) and have a constitutive triple response phenotype ( KIEBER et al. 1993) , as well as mutants that failed to form an exaggerated apical hook (hookless, h k l ; GUZMAN and ECKER 1990). Mutations at the C T R l locus result in severe constitutive triple-response phenotypes that are not reverted by inhibitors of ethylene biosynthesis or action ( KIEBER et al. 1993) . Several pleiotropic mutants have been identified with auxin defects that are also insensitive to ethylene in the seedling root (LINCOLN et al. 1990; PICKETT et al. 1990). The C T R l gene was cloned and found to show similarity to the Raf family of protein kinases, implicating a kinase cascade in this ethylene response ( KIEBER and ECKER 1993; KIEBER et al. 1993). The dominant etrl mutant gene also has been cloned; it shows significant similarity to the SLNl gene of Saccharomyces cerevisiae and the bacterial two-component histidine kinases ( CHANG et al. 1993; OTA and VARSHAVSKY 1993). In this study, we describe mutations in four new loci, all of which produce an ethylene-insensitive phenotype. We also describe a novel class of ethylene-insensitive and agravitropic root mutants. A genetic framework has been established for the action of these genes on the basis of epistasis interactions. Most of the mutations that we and others have described act in linear pathway. Genes that control branch points leading to tissue-specific phenotypes occur late in this pathway; these branches may provide critical insights into the nature of hormonal interactions in plants. MATERIALS AND METHODS Plant material: Arabidopsis plants were grown on MetroMix 200 ( Grace-Sierra Horticultural Products, Milpitas, C A ) under the following conditions: 23", 100-200 pE/m'/sec constant light. Plants were fertilized once between 2 and 4 wk with either a modified Hoaglands solution (FELDMANN and MARKS 1987) or Pete's Lite 14-15-16 ( Grace-Sierra Horticultural Products, Milpitas, C A ) . Plants were out-crossed as previously described ( GUZMAN and ECKER 1990). All mutants (with the exception of ctrl-5, tin32 and ein6 ) were isolated from the A. thalzana strain Columbia. Wassilewskija is the parental strain of both c tr l -5 and ein3-2 ( FELDMANN and MARKS 1987; KIEBER et al. 1993) . The ein6mutant was isolated from mutagenized Landsberg seed. X-ray and diepoxybutane mutagenesis were performed as described (WEBER et al. 1993). Fast neutron treated seed were a gift of LAURA CONWAY (University of Pennsylvania) . Surface-sterilized seedlings were grown on medium pH 5.7 containing MS salts (GIBCO, Gaithersburg, MD) supplemented with 1 mg/ml thiamine, 0.5 mg/ml pyridoxine, 0.5 mg/ml nicotinic acid, 100 mg/ liter inositol, 10 g / liter sucrose and 0.8% bacto-agar (DIFCO, Detroit, MI). Surface sterilization of seeds was performed as previously described ( GUZMAN and ECKER 1990). Mutagen treated seeds were plated at a density of -5000 seeds per 132 mm diameter petri dish. Assays for ethylene insensitivity or constitutive triple response were as previously described ( GUZMAN and ECKER 1990) . Genetic mapping: The ein2-1 mutant was mapped in a cross to the Landsberg marker line DP28. DNA was prepared from F, Einfamilies and scored with RFLPs ( CH&% et al. 1988; NAM et al. 1990) and the ATCTRl Simple Sequence Length Polymorphism (SSLP; BELL and ECKER 1994). The ein4, ein5-l and ein7 mutants were crossed onto the M10 marker line ( KOORNNEEF and STAM 1992). The genotypes of the F2 progenies from the ein4 cross were determined by testing the F3 progenies for segregation of mutant phenotypes; the F3 seedlings were then used for SSLP and Cleaved Amplified Polymorphic Sequences (CAI's; KONIECZNY and AUSUBEL 1993) mapping. An up1 ein5-l double mutant was isolated from the F2 progenies of the cross to M10; this plant was then crossed with CS2 ( d i s l , a n ) . F3 families were tested for segregation of the flanking apl, and disl markers and were mapped with nga280 (BELL and ECKER 1994). F3 families were also progeny tested in the ein7 cross for the segregation of up1 and nga280. The ein6 mutant was mapped in a cross to the wild-type Columbia strain. The eirl locus was mapped in a cross to the WlOO marker strain ( KOORNNEEF et al. 1987). DNAs for the SSLP and CAPS mapping experiments were prepared from a single rosette leaf from each Fz progeny or a minimum of 20 F3 seedlings (EDWARDS et al. 1991 ) . The ATCTRl and nga280 SSLPs were amplified using a capillary thermocycler (Idaho Technologies, Idaho Falls, ID) with the following conditions: 94", 1 sec; 56" 1 sec; 72", 15 sec; 3 mM magnesium, 50 cycles. The LFYCAPS products were amplified on a capillary thermocycler (Idaho Technologies, Idaho Falls, ID) with the following conditions: 94", 2 sec; 55", 2 sec; 72", 60 sec; 3 mM magnesium, 50 cycles. All other amplifications were performed as previously described ( KONIECZNY and AUSUBEI. 1993; BELL and ECKER 1994). Map distances were determined using the Map Manager v2.5 program (MANLY 1993) . Determination of epistasis and isolation of double mutants: A list of mutant strains used in this analysis is provided Ethylene Signal Transduction 1395 in Table 1. Most strains were backcrossed at least twice to remove potential background effects on the ethylene phenotype before double mutant construction. Chi-square analysis was performed on the F4 segregation ratios to examine possible epistasis relationships. Double mutan& then were isolated to demonstrate the genetic interaction between the mutant phenotypes. The etrl-3 ctrl-1, an?-2 ctrl-1, ein5-I ctrl-1, ein7 ctrl1, ein2-1 eirl-1, ein3-1 eirl-1, ein5-1 eirl-2, ctrl-1 eirl-1, ctrl-1 auxl-21, ein2-1 e t o l I , an2-1 hlsl-1 and eirl-1 auxl-21 double mutants were obtained by progeny testing F2 plants of each parental mutant phenotype to find a plant of the genotype ml /ml, + /m2; the double mutants were identified as seedlings with a new phenotype in the Fs generation. The genotype of each double mutant was verified by failure to complement the hypostatic mutation. The c tr l -1 hlsl 1, et01 1 hlsl 1 and eirl 1 hlsl I double mutants were identified in F2 progenies as seedlings expressing both mutant phenotypes. The ctrl-5 allele was created by the insertion of a kanamycin-resistant marked T-DNA into the CTRl gene ( KIEBER et al. 1993) . This selectable marker was used to isolate recombinants between ein2-1 and ctrl-5. F2 seeds from a ctrl-5 by ein2-1 cross were examined for ethylene insensitive seedlings in the presence of 50 pg/ml kanamycin. Kanamycin-resistant (kan') F2 Einplants were progeny tested for the presence of both the kan' marker and Einphenotype. The double mutant was then verified by the failure to complement ctrl-1. Quantifying the seedling phenotypes: Seedling measurements were obtained using a WILD dissecting scope with an ocular micrometer. Hook angles were quantified using a protractor reticle (Edmunds Scientific, Edison, NJ) . The severity of the triple-response phenotype was measured as elongation of root or hypocotyl in the presence either of 10 pl/liter ethylene or in hydrocarbon-free air. Statistical analysis was carried out with the Statview 512+ software (Brainpower Inc., Ventura, CA) . The Student's t-test was used for statistical comparison of the means. Sensitivity to 2,4dichlorophenoxyacetic acid (2,4D; Sigma) was assayed as inhibition of root elongation. Seedlings were surface sterilized and plated on MS agar media containing 2,4D. The seedlings were vernalized at 4" for 4 days before germination and root elongation was measured after 5 days at 23" in the dark. Gravity response was measured using a stage protractor on an Olympus dissecting scope. Surface-sterilized seeds were plated along a straight line marked on the petri dish. Seedlings were vernalized for 4 days at 4" and then placed perpendicular to the horizontal line. After 3 days in the dark at 23", the angle of root growth was measured as the deviation from vertical; a perfect positive gravitropic response (completely vertical root) was measured as 0".