As part of an effort to isolate new Arabidopsis mutants specifically defective in responsiveness to red light, we identified srl1 (short hypocotyl in red light) by screening an EMS-mutagenized M2 population derived from a phytochrome B (phyB)-overexpressor line (ABO). The srl1 mutant shows enhanced responsiveness to continuous red but not far-red light, in both wild-type and ABO backgrounds, consistent with involvement in the phyB-signaling pathway but not that of phyA. The hypersensitive phenotype of srl1 is not due to overexpression of endogenous phyA or phyB, and the locus maps to the center of chromosome 2, distinct from any other known photomorphogenic mutants. srl1 seedlings display enhancement of several phyB-mediated responses, including shorter hypocotyls, more expanded cotyledons, shorter petioles and modestly higher levels of CAB gene expression under red light than the wild type. Double mutant analyses show that the hypersensitive phenotype of srl1 is completely phyB-dependent. The data suggest, therefore, that SRL1 may encode a negatively acting component specific to the phyB-signaling pathway.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Plants use light not only as an energy source for photosynthesis, but also as an informational signal for growth and development. There are at least three families of photoreceptors in plants: phytochromes, blue/UV-A, and UV-B receptors that monitor light quality, quantity, direction, and duration ( Kendrick & Kronenberg, 1994). Of these, the most extensively characterized are the phytochromes which mediate both inductive and adaptive responses to red (R) and far-red (FR) light throughout a plant's life cycle. Responses include seed germination, seedling de-etiolation, shoot and leaf development, shade avoidance, and flowering ( Kendrick & Kronenberg, 1994). The biological activity of these photoreceptors resides in their capacity for reversible interconversion between the R-absorbing Pr form and the FR-absorbing Pfr form upon sequential photon absorption ( Quail, 1991).
The phytochromes are encoded by a small multigene family with five members (PHYA–E) in Arabidopsis ( Clack et al. 1994 ; Sharrock & Quail 1989) , and related sequences have been found in a diversity of other plants ( Schneider-Poetsch et al. 1998 ). The phytochromes of higher plants are soluble, dimeric, pigmented proteins in which each subunit consists of an approximately 124–130 kDa polypeptide that is covalently linked to a tetrapyrrole chromophore ( Quail, 1997). The polypeptide folds into two major structural domains: a globular NH2-terminal domain which binds the chromophore, and a C-terminal domain which carries the dimerization sites ( Quail, 1997). The NH2-terminal domain provides the photosensory function, defined as perception and interpretation of the nature of the light signals received; the C-terminal domain is necessary for the regulatory function, defined as transmission of the perceived signal to downstream components ( Quail et al. 1995 ).
Although considerable progress has been made in defining the photosensory roles, structural features, and functionally important domains and residues of the phytochromes ( Chory et al. 1996 ; Quail et al. 1995 ; Quail, 1997), the molecular nature of the signal transduction pathway is still poorly understood. The recent evidence for light-dependent kinase activity ( Fankhauser et al. 1999 ; Yeh & Lagarias, 1998) of recombinant phytochrome preparations from yeast supports the previous proposition that phytochromes may be non-conventional Ser/Thr kinases. Cellular signaling molecules such as trimeric G proteins, cGMP, and calcium/calmodulin have also been implicated in the phytochrome signal transduction pathway by using microinjection and pharmacological techniques ( Bowler et al. 1994 ). But the link between this kinase activity, signaling molecules, and gene expression is still missing.
In other studies, both genetic and molecular approaches have been employed to identify putative signaling components. Ni et al. (1998) have identified a basic helix–loop–helix protein, PIF3, by using a yeast two-hybrid approach, and have shown that PIF3 binds to the C-terminal domains of both PHYA & PHYB. Ni et al. (1999) have also shown that PIF3 binds to full-length phyB in the Pfr form, and that the two proteins dissociate when phyB is converted back to Pr form by exposure to FR light. Recently Martinez-Garcia et al. (2000) have shown that phytochrome B binds reversibly to G-box-bound PIF3 specifically upon light-triggered conversion of the photoreceptor to its biologically active conformer, suggesting that the phytochromes may directly regulate the transcriptional machinery of specific genes by physically complexing with promoter-bound transcriptional regulators. Using a similar approach, Fankhauser et al. (1999) have isolated a protein designated phytochrome kinase substrate 1 (PKS1) which binds to both PHYA and PHYB C-terminal domains. Moreover, PKS1 is phosphorylated by recombinant oat phyA preparations from yeast in a light-dependent manner. Choi et al. (1999) have also isolated nucleoside diphosphate kinase 2 as a phyA-interactor which is involved in phytochrome signaling. These studies have thus indentified potential immediate signaling partners for phytochromes A and B.
Genetic approaches have identified two broad classes of mutants defective in phytochrome signaling: one that shows partially etiolated morphology, i.e. long hypocotyl when grown in the light (hy), and another that shows de-etiolated morphology even when grown in complete darkness (det, de-etiolated; or cop, constitutively photomorphogenic). The first class includes the photoreceptor mutants, hy4 (blue-light receptor), hy8 (phyA, Dehesh et al. 1993 ), and hy3 (phyB, Reed et al. 1993 ); mutants defective in phytochrome chromophore biosynthesis (hy1, hy2, hy6;Koornneef et al. 1980 ); as well as putative phytochrome signal transduction mutants such as hy5, fhy1, fhy3 ( Whitelam et al. 1993 ), pef1, pef2, pef3 ( Ahmad & Cashmore, 1996), red1 ( Wagner et al. 1997 ), fin2 ( Soh et al. 1998 ), and far1 ( Hudson et al. 1999 ), which are thought to act as positive regulatory components. In addition, some mutants show enhanced responsiveness to either FR or both R and FR in a light-dependent manner. These include spa1 ( Hoecker et al. 1998 ), psi2 ( Genoud et al. 1998 ), and poc1 ( Halliday et al. 1999 ), of which spa1 and psi2 are thought to act as negative regulatory components. The second class of mutants, containing multiple members designated as cop, det or fus, exhibits seedling development in total darkness (short hypocotyls, expanded cotyledons, etc.) resembling that found in wild-type seedlings grown in white light. The recessive nature and pleiotropic phenotypic effects of these mutations indicate that the wild-type gene products repress photomorphogenic development in darkness. Presumably in the normal plant this repression is relieved in the appropriate tissues by signals from the photoreceptors.
Recently, considerable progress has been made in cloning and characterization of the genetic loci identified in both positive and negative regulatory mutants. HY5 has been shown to be a bZIP class transcription factor ( Oyama et al. 1997 ) which binds to the G-box elements of both CHS and RBCS promoters ( Ang et al. 1998 ; Chattopadhyay et al. 1998 ). FAR1 is a nuclear localized novel protein, apparently a member of a multigene family ( Hudson et al. 1999 ). Both COP1 and DET1 have been shown to be nuclear localized ( von Arnim & Deng, 1994; Pepper et al. 1994 ), and COP1 has been shown to regulate HY5 negatively, thereby exerting its negative effect ( Ang et al. 1998 ). The COP9 signalosome which consists of eight subunits is a conserved nuclear protein complex found in plants and animals, and is closely related to the lid subcomplex of the 26S proteosome ( Wei & Deng, 1999). DET2 has been shown to be a steroid 5α reductase in the brassinolide biosynthetic pathway ( Li et al. 1996 ). SPA1 has been shown to be a novel WD-40 repeat-containing protein which is also localized to the nucleus ( Hoecker et al. 1999 ). Cloning and characterization of all the mutants in this class will shed light on the nature and complexity of this signal transduction pathway.
Apart from cop/det/fus-like mutants, there are reports of hypersensitive mutants which exhibit negative regulation of either the phyA-signaling pathway specifically, such as spa1 ( Hoecker et al. 1998 ; Hoecker et al. 1999 ), or both phyA-and phyB-signaling pathways, such as psi2 ( Genoud et al. 1998 ). Although poc1 is a continuous red light (Rc)-specific hypersensitive mutant, it has been shown that the poc1 phenotype is due to aberrant overexpression of PIF3 which is a positively acting, bHLH-class transcription factor ( Halliday et al. 1999 ; Ni et al. 1998 ). There are no reports of a negative regulatory component, specifically in the phyB-signaling pathway. Here we report the isolation of srl1 which shows hypersensitivity in many phyB-mediated responses.
Isolation of the srl1 mutant
Overexpression of Arabidopsis phyB under the control of the CaMV35S promoter in transgenic Arabidopsis (designated as the ABO line) induces a strong hypersensitive phenotype under Rc ( Wagner et al. 1991 ). Mutagenesis of this ABO line and screening under normal Rc was effective in isolating both intragenic mutants ( Wagner & Quail, 1995), and extragenic signaling mutants ( Wagner et al. 1997 ). The mutagenized ABO seeds were also screened in the present study for reduced sensitivity to end-of-day FR. In this screen, the seeds were grown in white light under a day–night cycle for 4 days, and a saturating far-red light pulse (FRp) was given at the end of each day for 10 min to convert the Pfr form of the phytochromes to Pr before the beginning of the dark period. Although this screen was not initially designed to isolate hypersensitive mutants, we have obtained several such mutants in this way, including poc1 ( Halliday et al. 1999 ). These mutants fall into two groups: one where the mutant phenotype co-segregated with the introduced PHYB transgene and were found to represent intragenic hyperactive phyB mutants (data not shown); and the other where the phenotype segregated independently of the transgene and is therefore extragenic. Analysis of one such extragenic mutant, designated srl1, is described here. This mutant initially showed a hypersensitive phenotype in the ABO background under Rc ( Fig. 1). The mutant in the ABO background was back-crossed to wild type (No-O) and isolated in the No-O background from an F2 population. The mutant was back-crossed again to reduce the background mutations and an F2 population of this cross-segregated 3 : 1 wild type to mutant under Rc, indicating that it represents a completely recessive, monogenic and extratransgenic trait.
Hypocotyl length of srl1 is selectively hypersensitive to continuous red light
To determine whether srl1 is hypersensitive to FRc as well as Rc, we performed fluence-rate response analysis for each wavelength. As shown in Fig. 2(a), srl1 showed shorter hypocotyl lengths compared to the wild type over the range of Rc fluence rates used. In contrast, the hypocotyl lengths of srl1 were similar to wild type under FRc ( Fig. 2b). The hypocotyl lengths, apical hooks and cotyledons are indistinguishable from wild type in the dark, establishing that the hypersensitive phenotype of srl1 is light-dependent ( Fig. 2a–c). Therefore we conclude that srl1 is an Rc-specific, hypersensitive mutant.
Srl1 has more expanded cotyledons and shorter petioles than the wild type
Expansion of cotyledons and shortening of petioles are controlled by Rc through the phyB-signaling pathway. To test whether srl1 has any effect on these responses, we measured cotyledon size and petiole lengths of the mutant and wild type. We also included ABO and phyB-mutant lines as positive and negative controls, respectively, in these experiments. As shown in Fig. 2(d), the cotyledon area of srl1 is larger than No-O and similar to ABO. The phyB mutant has very small cotyledons under these conditions. Similarly, petiole lengths are also very short for the srl1 mutant as compared to No-O ( Fig. 2e). The phyB mutant showed very long petioles, whereas ABO had barely any petioles under these conditions ( Fig. 2e). Thus the srl1 phenotype approaches that of ABO as regards these morphological parameters of seedling de-etiolation.
Apart from the morphogenic effects indicated above, no striking influence of the srl1 mutation was observed on any aspect of the adult phenotype in preliminary experiments (data not shown).
Srl1 has higher levels of CAB gene expression
To test whether srl1 has any effect on light-regulated gene expression, we assayed CAB and RBCS mRNA levels, both of which are induced by red light through the phyB-signaling pathway. We first grew seedlings under Rc for 18 h using 20 μmol m−2 s−1, and did not see any difference in the expression of these genes between wild type and srl1 (data not shown). Then we tested short pulses (30 sec and 5 min at 22 μmol m−2 s−1) of red light, and found that CAB mRNA is induced about twofold more at the higher light fluence in srl1 compared to the wild type ( Fig. 3). RBCS expression in srl1 was similar to the wild type using this amount of light ( Fig. 3). These results indicate that the srl1 mutation increases the sensitivity of R-induced CAB gene expression.
Hypersensitive phenotype of srl1 is phyB-dependent
To further characterize the hypersensitive phenotype of srl1 specifically to Rc, we constructed srl1phyA101 and srl1phyB-1 double mutants. Although phyA can perceive R, we predicted that srl1 would show the hypersensitive phenotype even in the absence of phyA, based on the fact that srl1 showed no phenotype under FRc. On the other hand, because srl1 did not have any phenotype in the dark, and showed hypersensitivity only to Rc, we predicted that the srl1phyB-1 double mutant phenotype might be similar to the phyB-1 single mutant. Indeed, the srl1phyA double mutant showed shorter hypocotyl lengths than the phyA single mutant over the range of Rc fluence rates used ( Fig. 4), whereas the srl1phyB double mutant was hyposensitive to all the Rc fluence rates used, in a manner similar to the phyB single mutant ( Fig. 4). This result indicates that the Rc-specific hypersensitive phenotype of srl1 is strictly dependent on red light perception by phyB.
Hypersensitivity of srl1 is not due to overexpression of phyA and phyB
The Rc-specific hypersensitive phenotypes displayed by srl1 could be caused in several different ways. A mutation that causes overexpression or increased stability of phyB and/or phyA would show similar phenotypes, although such a mutant for phyA would also be hypersensitive to FRc. Alternatively, a signaling mutant which functions as a negative regulator of the phyB-signaling pathway would be hypersensitive specifically to Rc. To distinguish between these possibilities, we performed Western blot analyses for phyA and phyB. As shown in Fig. 5, both phyA and phyB levels in srl1 are comparable to wild type under similar conditions. In addition, the degradation kinetics of the Pfr form of phyA under Rc is similar in both srl1 and wild type. phyB is more stable than phyA under Rc, and a similar level of reduction of phyB was observed in srl1 and wild type after 3 days' Rc ( Fig. 5). This result indicates that the hypersensitive phenotype is not due to overexpression or increased stability of either phyA or phyB. Thus the data are consistent with the possibility that srl1 is a signaling mutant specific to the phyB-signaling pathway.
Map position of srl1
We have mapped srl1 using simple sequence length polymorphism (SSLP) ( Bell & Ecker, 1994) and cleaved amplified polymorphic sequence (CAPS) ( Konieczny & Ausubel, 1993) markers using two mapping populations. Fifty mutant seedlings (100 chromatids) from the F2 population of each mapping population were used. srl1 is located in the middle of chromosome 2 about 3.6 c m away from the PHYB gene. This result is consistent with an approximately 4 c m distance between srl1 and phyB obtained using visible markers (see Experimental procedures for constructing srl1phyB double mutant). No other photomorphogenic mutants have been reported to map to this region. This distinct map position thus provides evidence that srl1 is a new photomorphogenic mutant.
Evidence from both genetic and molecular studies indicates that phytochrome signaling might be initiated as separate pathways specific to each individual member of the photoreceptor family, but which converge to a common point later in the signal transduction cascade ( Deng & Quail, 1999; Neff et al. 2000 ). Isolation and characterization of the Rc-specific hypersensitive mutant, srl1, reinforces that proposition. Moreover, the data presented here show that, similar to phyA-signaling ( Hoecker et al. 1998 ; Hoecker et al. 1999 ), phyB-signaling may also be naturally attenuated in Arabidopsis.
srl1 is the first mutant reported to be hypersensitive specifically to red light as a result of apparent negative regulation of the signaling pathway. One hypersensitive mutant, poc1, which is specific to red light, has been isolated and described previously ( Halliday et al. 1999 ), but is thought to be a positive regulator based on independent molecular and reverse-genetic experiments ( Ni et al. 1998 ). Although PKS1 ( Fankhauser et al. 1999 ) and ATHB2 ( Carabelli et al. 1996 ; Steindler et al. 1999 ) have been proposed to function as negative regulators in the phyB-signaling pathway, no mutant for these genes has yet been reported. The phenotypic responses presented here for srl1, such as shorter hypocotyls, expanded cotyledons, shorter petioles and increased CAB gene expression, compared to the wild type ( Figs 2 3and 3), are all characteristics of enhanced phyB signaling. By contrast, the hypocotyl length of srl1 was very similar to the wild type in the dark ( Fig. 2), indicating that the srl1 phenotype is light-dependent. Moreover, the lack of an FRc effect on hypocotyl elongation in the srl1 mutant provides evidence for Rc specificity of this mutant. The Rc-enhanced cotyledon expansion is also consistent with enhancement of the normal seedling de-etiolation process, and is evidence against a light-induced, global inhibition of growth. Consistent with these results, the srl1phyA double mutant is also hypersensitive to Rc as compared to phyA siblings ( Fig. 4), indicating that phyA signaling is unaffected in srl1. In contrast, srl1phyB double mutants were almost indistinguishable from phyB siblings under Rc ( Fig. 4). These results suggest that phyB is essential for the srl1 phenotype, and the other four phytochromes cannot substitute for phyB in inducing the srl1 phenotype.
Because superficially the srl1 phenotype could have been obtained simply by overexpressing phyB, as observed for ABO ( Wagner et al. 1991 ), we performed both genetic and biochemical tests to determine if this was the case. The data show no perturbation of phyB or phyA levels in the srl1 mutant ( Fig. 5). Thus the evidence for srl1 being a new photomorphogenic mutant are: (i) the srl1 phenotype is not due to overexpression or increased stability of phyA and/or phyB ( Fig. 5); and (ii) the srl1 locus maps to a region on chromosome 2 where no other reported photomorphogenic mutants are located ( Fig. 6). The distinct map position (based on both the CAPS and visible markers) also rules out the possibility that the srl1 phenotype might be due to a mutation in the PHYB gene, leading to expression of a hyperactive phyB molecule. Therefore the srl1 phenotype is potentially due to an amplification or reduction in attenuation of the signal perceived by phyB. Such an enhancement of phytochrome signaling has been found in high pigment (hp) mutants of tomato ( Kendrick et al. 1997 ; Peters et al. 1992 ), and in spa1 which specifically amplifies phyA signaling in Arabidopsis ( Hoecker et al. 1998 ; Hoecker et al. 1999 ). On the other hand, the minimal effect of the srl1 mutation on CAB gene expression ( Fig. 3a), and the absence of any striking influence on adult phenotypic responses observed in preliminary experiments (data not shown), may indicate that SRL1 function is primarily confined to a branch of the phyB-signaling pathway controlling seedling growth responses to light.
Two simple models are proposed for the possible action of SRL1. The recessive nature of the srl1 mutation indicates that SRL1 might act as a negative regulator of phyB signaling. In that case, SRL1 might directly regulate phyB itself ( Fig. 7a), or positive regulators of phyB signaling, such as RED1, PEF2 or PEF3 ( Fig. 7b), or both. In the case of phyB, SRL1 might directly interact, and either physically block signal transfer to downstream components, or post-translationally modify phyB, thereby reducing its activity. Although post-translational modification of phyA has been proposed to attenuate phyA signaling ( Emmler et al. 1995 ; Jordan et al. 1995 ; Stockhaus et al. 1992 ), no such regulation of phyB activity has been reported. However, both phyA and phyB have been shown to translocate into the nucleus upon light exposure ( Nagy & Schäfer 2000) . SRL1 might be involved in nuclear translocation or cytosolic retention of phyB. Recently, Martinez-Garcia et al. (2000) have shown that phyB interacts with PIF3, which is bound to a G-box element present in light-regulated promoters, thereby potentially controlling transcription of these light-regulated genes. Any factor that impairs this interaction could work as a negative regulator of phyB signaling. SRL1 might represent such a factor, which inhibits this positive interaction. Alternatively, SRL1 might regulate the expression or activity of positively acting components of phyB signaling ( Fig. 7b). Again, SRL1 might block the ability of these components to receive signals from phyB by physical interaction.
Why do plants need a negative regulator in addition to positive regulators? One explanation might be that evolution of negative regulators, such as SRL1, in addition to positive regulators, allowed plants to fine-tune their responses to complex environments. Alternatively, SRL1 might function as a mediator for cross-talk with other signaling cascades which either are regulated by, or regulate, phyB signaling.
In summary, we provide genetic evidence for isolation of a new locus, srl1, which appears to function as a negative regulator in phyB signaling. Cloning and characterization of SRL1 will provide more insight into the mechanism of SRL1 action.
Growth conditions for seedlings and measurements
Seeds were surface-sterilized with 20% bleach (1.05% sodium hypochlorite) and 0.03% Triton X-100 for 10 min, then washed five times with sterile water before plating on growth medium (GM) ( Valvekens et al. 1988 ) without sucrose. The plates were kept in the dark at 4°C for stratification. Germination was induced with 3 h white light treatment followed by 21 h incubation in the dark at 21°C. The plates were then placed under appropriate light conditions for 3 days before measuring hypocotyl length. The light sources used were described by Wagner et al. (1991) . Light fluence rates were measured using a spectroradiometer (model LI-1800, LiCor, Lincoln, NE, USA). Hypocotyl lengths were measured according to Wagner et al. (1996) . For cotyledon measurements, the seedlings were grown under Rc (13 μmol m−2 s−1) for 4 days. The cotyledons were pressed gently on the surface of agar medium before taking photographs. The perimeter of each cotyledon was traced and the area determined by NIH image software (Bethesda, MD, USA). For petiole lengths, the seedlings were grown on GM agar plates for 5 weeks in low-level continuous white light (WLc) (2.25 μmol m−2 s−1). The two longest leaves from each plant (n > 8) were photographed, and petiole lengths were measured in the same way as hypocotyl lengths.
Isolation of RNA and Northern blotting
Total RNA was isolated from 4-day-old dark-grown seedlings treated with a single red-light pulse (Rp) of 660 or 6600 μmol m−2. Tissue was harvested 18 h after the light treatment and used for RNA extraction using Qiagen RNeasy Plant mini kit (Qiagen, Valencia, CA, USA). Total RNA (5 μg) was separated on a 1.2% 3-[N-morpholino]propanesulfonic acid (MOPS)-formaldehyde agarose gel and blotted onto a nitrocellulose membrane. The membrane was prehybridized and hybridized using Church's buffer ( Church & Gilbert, 1984) and washed finally with 0.2 × SSC containing 0.1% SDS for 10 min twice at 65°C. Transcript levels for CAB and RBCS were determined using a phosphorimager (Storm 860, Molecular Dynamics, Sunnyvale, CA, USA) quantification and were normalized to 18S rRNA levels.
Construction of srl1phyA and srl1phyB double mutants
For construction of the srl1phyA double mutant, homozygous srl1 was crossed to phyA-101 (hy8-1, RLD ecotype; Dehesh et al. 1993 ). F1 seedlings were selfed and F2 progeny were plated and grown under Rc (13 μmol m−2 s−1). srl1-like short-hypocotyl seedlings were selected along with tall seedlings showing the wild-type phenotype (for selecting homozygous phyA siblings), and grown in the greenhouse. F3 seeds were plated and grown under FRc (7 μmol m−2 s−1) to select for homozygous srl1phyA and homozygous phyA siblings. The presence of the mutant phyA locus was confirmed by CAPS marker.
For construction of the srl1phyB double mutant, homozygous srl1 was crossed to a phyB-1 mutant ( Reed et al. 1993 ) hy3-Bo64 allele that was introgressed into the No-O background. F2 seeds were plated and grown under Rc (13 μmol m−2 s−1). phyB mutant-like seedlings with long hypocotyls and small cotyledons were selected and grown in the greenhouse. At this stage, genomic DNA from these selected seedlings was isolated and the genotype of the PHYB locus was checked using PHYB CAPS marker. The homozygous phyB plants were test-crossed to homozygous srl1 plants to determine the genotype at the srl1 locus. If srl1 is phyB-dependent, the F1 progenies from the test crosses should fall into three classes when grown under Rc: 100% of the seedlings should have short hypocotyls compared to the PHYB/phyB heterozygote if the srl1 locus is homozygous; 50% should have short hypocotyls and the remaining 50% should be similar to the PHYB/phyB heterozygote if the srl1 locus is heterozygous; and 100% should be similar to the PHYB/phyB heterozygote if the srl1 locus is wild type. From test crosses of 25 independent phyB seedlings, we found only two seedlings that segregated 50% for a short-hypocotyl phenotype compared to the PHYB/phyB heterozygote under Rc. This result indicated that the srl1 locus might be approximately 4 c m apart from the PHYB locus, which is consistent with our mapping data using SSLP and CAPS markers. These two seedlings, which were homozygous for the phyB allele but heterozygous for the srl1 allele, were selfed. Twenty different plants from each of these plants were test-crossed again to determine the genotype at the SRL1 locus. When grown under Rc (13 μmol m−2 s−1), the F1 seeds from these test-crosses segregated into three classes as described above. Only seedlings that showed 100% short hypocotyl compared to PHYB/phyB heterozygotes were selected as srl1phyB double mutants. Homozygous phyB siblings were also selected from the same population showing 100% long hypocotyls similar to the PHYB/phyB heterozygous seedlings.
Western blot analyses of phyA and phyB
Crude proteins were extracted from No-O and srl1 seedlings according to Wagner et al. (1991) . Crude protein (5 or 10 μg) was separated on 8% SDS–PAGE for detecting phyA and phyB, respectively. The gels were blotted onto polyvinylidene difluoride membrane and the membranes were blocked in 1× TBST containing 2% fat-free milk powder. The primary and secondary antibody incubations and subsequent washings were performed with 1× TBST containing 0.5% milk powder. For detecting phyA and phyB, 073D and B1–B7 monoclonal antibodies, respectively, were used ( Hirschfeld et al. 1998 ). An anti-mouse antibody (Promega, Madison, WI, USA) conjugated with horseradish peroxidase was used as the secondary antibody and was detected with a chemiluminescent system (SuperSignal, Pierce, Rockfield, IL, USA).
Mapping of srl1
Homozygous srl1 plants were crossed to the Col-O and La-er ecotypes. The F1 plants were selfed and the resulting F2 seeds were plated on GM medium without sucrose and grown under Rc (13 μmol m−2 s−1) for 3 days. The short-hypocotyl seedlings similar to srl1 were selected from this segregating population and were transplanted onto soil to grow in the greenhouse. Leaf tissue was harvested from these selected plants and genomic DNA was isolated according to Edwards et al. (1991) . srl1 was mapped using these DNA preparations with SSLP ( Bell & Ecker, 1994) and CAPS ( Konieczny & Ausubel, 1993) markers. The F3 seeds of the selected plants were harvested and tested for short-hypocotyl phenotype. The F3 seedlings showing 100% short-hypocotyl phenotype under Rc were used for calculating map position.
We thank U. Hoecker for helpful discussions and M. Hudson for critical reading of the manuscript. This work was supported by grants from DOE Basic Energy Sciences number DE-FG03-87ER13742 and US Department of Agriculture Current Research Information Service number 5335-21000-010-00D.