BAS1 and SOB7 act redundantly to modulate ... - Wiley Online Library

The Plant Journal (2005) 42, 23–34

doi: 10.1111/j.1365-313X.2005.02358.x

BAS1 and SOB7 act redundantly to modulate Arabidopsis photomorphogenesis via unique brassinosteroid inactivation mechanisms Edward M. Turk1, Shozo Fujioka2,3, Hideharu Seto2,3, Yukihisa Shimada3, Suguru Takatsuto4, Shigeo Yoshida2,3, Huachun Wang5, Quetzal I. Torres1, Jason M. Ward1, Girish Murthy1, Jingyu Zhang1, John C. Walker5 and Michael M. Neff1,* 1 Department of Biology, Washington University, Campus Box 1137, One Brookings Drive, St Louis, MO 63130, USA, 2 Plant Functions Laboratory, RIKEN (The Institute of Physical and Chemical Research), Wako-shi, Saitama 351 0198, Japan, 3 Plant Science Center, RIKEN (The Institute of Physical and Chemical Research), Wako-shi, Suehirocho, Tsurumi, Yokohama, Kanagawa 230 0045, Japan, 4 Department of Chemistry, Joetsu University of Education, Joetsu-shi, Niigata 943 8512, Japan, and 5 Division of Biological Sciences, University of Missouri, Columbia, MO 65211 7400, USA Received 22 October 2004; revised 9 December 2004; accepted 14 December 2004. * For correspondence (fax þ314 935 4432; e-mail [email protected]).

Summary Active brassinosteroids (BRs), such as brassinolide (BL) and castasterone (CS), are growth-promoting plant hormones. An Arabidopsis cytochrome P450 monooxygenase (CYP734A1, formerly CYP72B1), encoded by the BAS1 gene, inactivates BRs and modulates photomorphogenesis. BAS1 was identified as the overexpressed gene responsible for a dominant, BR-deficient mutant, bas1-D. This mutant was isolated in an activationtagged screen designed to identify redundant genes that might not be identified in classic loss-of-function screens. Here we report the isolation of a second activation-tagged mutant with a BR-deficient phenotype. The mutant phenotype is caused by the overexpression of SOB7 (CYP72C1), a homolog of BAS1. We generated single and double null-mutants of BAS1 and SOB7 to test the hypothesis that these two genes act redundantly to modulate photomorphogenesis. BAS1 and SOB7 act redundantly with respect to light promotion of cotyledon expansion, repression of hypocotyl elongation and flowering time in addition to other phenotypes not regulated by light. We also provide biochemical evidence to suggest that BAS1 and SOB7 act redundantly to reduce the level of active BRs, but have unique mechanisms. Overexpression of SOB7 results in a dramatic reduction in endogenous CS levels, and although single null-mutants of BAS1 and SOB7 have the same level of CS as the wild type, the double null-mutant has twice the amount. Application of BL to overexpression lines of BAS1 or SOB7 results in enhanced metabolism of BL, though only BAS1 overexpression lines confer enhanced conversion to 26-OHBL, suggesting that SOB7 and BAS1 convert BL and CS into unique products. Keywords: brassinosteroids, photomorphogenesis, activation-tagging, cytochrome P450.

Introduction Brassinolide (BL) was identified in 1979 as a unique steroid in that it contains a B-ring lactone and is the active growthpromoting component in Brassica napus pollen (Grove et al., 1979). To date, over 50 naturally occurring BL analogs have been identified exclusively in plants, are collectively termed brassinosteroids (BRs), and are recognized as hormones critical to plant development (Clouse and Sasse, 1998; Fujioka and Yokota, 2003). Castasterone (CS) is the immediate precursor to BL and is also considered to be an ª 2005 Blackwell Publishing Ltd

active BR due to the high level of CS and absence of BL in tomato and mung bean (Yokota et al., 1991). Intensive study in the past few years utilizing Arabidopsis has identified several of the BR signaling components, including BRI1, a critical component of the BR receptor (Li and Chory, 1997). The BL metabolic pathway has been largely determined by feeding deuterium-labeled precursors to cultured cells or seedlings followed by GC-MS analysis of the resulting metabolites (Choi et al., 1996; Fujioka et al., 2000). 23

24 Edward M. Turk et al. Subsequent genetic studies in Arabidopsis have led to the identification of some of the corresponding genes, three of which were identified in photomorphogenesis screens (Noguchi et al., 2000). Loss-of-function screens have been employed to identify Arabidopsis mutants conferring light-grown seedling phenotypes, such as short hypocotyls, when grown in the dark. This screen identified the DEETIOLATED2 (DET2) gene, which encodes the steroid 5a-reductase responsible for an early step in the BR biosynthetic pathway (Li et al., 1996). The CONSTITUTIVE PHOTOMORPHOGENIC AND DWARF (CPD) gene, which encodes a cytochrome P450 (CYP90A) responsible for a biosynthetic step in the BR pathway downstream of DET2, was isolated in a similar screen (Szekeres et al., 1996). Activation-tagging screens for photomorphogenesis mutants in Arabidopsis were later used to overcome some of the limitations of loss-of-function screens, such as the inability to identify genes whose loss can be compensated for by a functionally redundant homolog. One such screen identified the PHYB-4 ACTIVATION-TAGGED SUPPRESSOR 1 (BAS1) gene (Neff et al., 1999). BAS1 is a BR catabolic gene and a positive modulator of photomorphogenesis in Arabidopsis (Turk et al., 2003) and encodes a cytochrome P450, originally named CYP72B1. CYP72B1 has been renamed CYP734A1 based on recent phylogenetic analysis of cytochrome P450 genes from Arabidopsis and rice (Nelson et al., 2004). Here we report the identification of the SUPPRESSOR OF PHYB-4 7 (SOB7) gene, which encodes the cytochrome P450 CYP72C1, and is functionally redundant to BAS1 in that they both inactivate BRs and modulate photomorphogenesis. Although both genes encode cytochrome P450s with close homology, they are likely to have unique biochemical mechanisms of BR inactivation. Results Identification of the Arabidopsis SOB7 gene as a BR inactivator An Arabidopsis dominant dwarf mutant, sob7-D, with the characteristic BR-deficient phenotype typified by the det2-1 BR biosynthetic mutant (Figure 1a), was identified in an activation-tagging mutagenesis screen for suppressors of the long-hypocotyl phenotype conferred by the phyB-4 mutation. A phenotypically similar mutant was identified in an activation-tagging screen for suppressors of the adult phenotypes conferred by the er-116 mutation. Sequence analysis of flanking genomic DNA recovered from these mutants by plasmid rescue identified SOB7/CYP72C1 (At1g17060) as the nearest gene. We confirmed that the sob7-D phenotype was caused by the overexpression of SOB7 by transforming wild-type Arabidopsis (Col-0) with the

(a)

(b)

(c)

ATG

Figure 1. Activation-tagging identifies an Arabidopsis BR-deficient mutant. (a) The activation-tagged mutant sob7-D is similar to a BR-deficient mutant (det2-1) and the sob7-DR recapitulation line. SOB7 loss-of-function line (sob71) is similar to the Col-0 wild type (data not shown). Plants were grown for 9 weeks in short days. Bar ¼ 2.54 cm. (b) RT-PCR of SOB7 shows that the activation-tagged and recapitulation lines are both overexpressing SOB7 compared to Col-0. The bas1-2 and sob7-1 lines are both RNA-null alleles of their respective genes. (c) Gene model of SOB7 (3361 bp) compared to BAS1 (3107 bp) with the ATG start codon and the T-DNA insertion sites indicated. Exons are boxed. UTRs and introns are lines.

rescued plasmid, which resulted in several dominant recapitulation lines (sob7-DR) with a range of similar phenotypes from strong to weak BR-deficient (Figure 1a). Overexpression of SOB7 in sob7-D and sob7-DR was confirmed by RTPCR (Figure 1b). Overexpression of the SOB7 cDNA driven by the constitutive CaMV 35S promoter also resulted in several BR-deficient lines (data not shown). ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34

Genetic interactions between BAS1 and SOB7 25 Sequence analysis of several cloned SOB7 cDNAs revealed a 519-amino acid protein that deviated from the 1 genomic prediction (Arabidopsis Genome Initiative, 2000) in that the predicted fourth exon, fourth intron and fifth exon are actually a single exon (Figure 1c). A thymidine is added at position 2202 to restore the reading frame and make a comprehensive sequence. This addition is also supported by ESTs from Brassica napus (CD828642, CD829912) (Bak and Paquette, 2004). The activation-tagged mutant bas1-D has the same BRdeficient phenotype as sob7-D, but is caused by overexpression of the BR catabolic gene BAS1 (Neff et al., 1999). An alignment of their amino acid sequences revealed that BAS1 and SOB7 are 53% similar and 36% identical. Both contain the cytochrome P450 cysteine heme–iron ligand signature and the positions of the three BAS1 introns are identical to SOB7 (Figure 2). A recent phylogenetic analysis of the plant cytochrome P450 family indicates that BAS1 and SOB7 are close homologs in the CYP72 clan (Nelson et al., 2004). BAS1 and SOB7 have overlapping and distinct expression patterns Because BAS1 and SOB7 may both function to inactivate BRs we suspected that they might also affect similar Arabidopsis physiology processes. Indeed, RT-PCR analysis of total RNA from adult Arabidopsis tissues revealed both overlapping as well as distinct expression patterns for BAS1

Figure 2. Sequence alignment of BAS1 and SOB7 from Arabidopsis. The positions of the three BAS1 introns (down arrows) are identical to SOB7 (up arrows), which has an additional fourth intron. Similar residues are shaded gray (53%) and identical residues are shaded black (36%). The position of the T-DNA alleles is marked by X, the endoplasmic reticulum targeting sequences are underlined, and the cytochrome P450 cystein heme-iron ligand signature is marked by *.

ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34

and SOB7 (Figure 3). Both are expressed in all adult organs tested, although to varying degrees. For example, the SOB7 transcript level is greatest in the silique, while the BAS1 transcript level is noticeably reduced in the rosette leaves and stem compared to other tissue (Figure 3a). Both are also expressed in 5-day-old Col-0 seedlings, although there was slightly more BAS1 transcript in the light compared to BAS1 transcript in the dark, and slightly more SOB7 transcript in the dark compared to SOB7 transcript in the light (Figure 3b). T-DNA insertional mutants of BAS1 and SOB7 are less responsive to light A BAS1 null allele, bas1-1, is available in the Ws-2 background (Turk et al., 2003). However, the only sob7-null allele is in the Col-0 background. To investigate the physiological role of BAS1 in conjunction with SOB7, T-DNA insertional mutants in the Col-0 background, bas1-2 and sob7-1, were obtained from the SALK institute (Alonso et al., 2003). The bas1-2 mutant contains head-to-head T-DNA insertions in the 387th codon while the sob7-1 mutant contains head-tohead T-DNA insertions in the 198th codon (Figure 1c). RT-PCR analysis of total RNA isolated from homozygous mutants revealed an absence of BAS1 and SOB7 transcript in their respective mutants, suggesting that both are null alleles (Figure 1b). The homozygous bas1-2 and sob7-1 mutants were crossed and the bas1-2sob7-1 double mutant was

26 Edward M. Turk et al.

(a)

(b)

identified in the F2 generation by PCR and confirmed by RTPCR (Figure 1b). The Arabidopsis hypocotyl provides a simple bioassay for light responsiveness, which is measured as a reduction in length after exposure to light (Koornneef et al., 1980). In a fluence rate response analysis the bas1-2 and sob7-1 hypocotyls were taller/less-responsive than Col-0 to white light (Figure 4a). Hypocotyls of the bas1-2sob7-1 double mutant were even taller/less-responsive, indicating that BAS1 and SOB7 are indeed involved in hypocotyl response to white light. To determine if the role of BAS1 and SOB7 in hypocotyl response to light is specific to a particular photoperception pathway, fluence rate response analysis in monochromatic light was investigated. In all monochromatic light conditions tested the single and double mutants were taller/lessresponsive than Col-0, indicating that the role of BAS1 and SOB7 in hypocotyl response to light is not specific to a single photoperception pathway (Figure 4b–d). Arabidopsis seedlings germinated in the dark produce diminutive cotyledons, which respond to light with a dramatic expansion (Neff and Van Volkenburgh, 1994). We utilized this phenotype to further test the hypothesis that BAS1 and SOB7 are involved in light responses. While bas1-2 and sob7-1 cotyledons had a wild-type response to white light, the double mutant showed enhanced expansion. In red, blue, or far-red light bas1-2 showed enhanced Figure 4. BAS1 and SOB7 are genetically redundant with respect to repression of hypocotyl elongation. Under certain white (a), blue (b), far-red (c), and red (d) fluence rates, BAS1 can compensate for the loss of SOB7, while the double-null mutant has a more dramatic phenotype than either of the singlenull mutants (down arrows). Seedlings pictured are the average Col-0 (left) and bas1-2sob7-1 (right) at the fluence rate pointed out by the arrow.

Hypocotyl length (mm)

Figure 3. BAS1 and SOB7 have overlapping and distinct expression patterns. RT-PCR of total RNA isolated from 5-week-old organs (a) and 5-day-old whole seedlings grown in the dark or continuous white light (b). UBQ10 was used as an internal loading control.

(c)

(d)

Fluence rate (µmol m–2 s–1) ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34

Genetic interactions between BAS1 and SOB7 27

(a)

(b)

Overexpression mutants of BAS1 and SOB7 are more responsive to light The hypothesis that an aberrant phenotype is caused by a null mutant can be strengthened by analysis of a second independent null mutant that has the same phenotype, or by an overexpression mutant that has the opposite phenotype. Although several putative T-DNA insertional mutants in the SOB7 gene were analyzed, only one null mutation was available at the time of this report. Therefore, we assayed the sob7-D overexpression mutant for opposing phenotypes relative to the wild type and the sob7-1 null mutant. We also assayed the BAS1 overexpression mutant, bas1-D, after backcrossing to Col-0 to remove the linked phyB-4 mutation. The homozygous PHYB locus was confirmed by dCAPS analysis (data not shown) (Neff et al., 2002). The bas1-D and sob7-D mutants were more responsive to all light conditions in our hypocotyl assay, strengthening the hypothesis that both genes are involved in light responsiveness (Figure 4a–d). BAS1 and SOB7 are not exclusive to light responses

Figure 5. BAS1 and SOB7 are genetically redundant with respect to cotyledon size (a) and epidermal cell size of the cotyledon in white light (b). *Mean significantly different from the Col-0 mean using an unpaired t-test, two-tailed P < 0.05, n ¼ 60–95.

cotyledon expansion, and the double mutant had even a greater degree of expansion, indicating that both BAS1 and SOB7 are involved in cotyledon growth in response to light (Figure 5a). BRs function in part through the promotion of cell expansion (Bishop and Yokota, 2001). In white light, cotyledons of the double mutant had significantly larger cells than Col-0, suggesting that increased expansion of the cotyledon is in part a result of larger cells (Figure 5b). We also examined flowering time in long-day growth conditions to further test the hypothesis that BAS1 and SOB7 are involved in light responses. The phyB-9 null allele of PHYB was utilized as a control for early flowering, the phyA-211 null allele of PHYA was utilized as a control for late flowering and the cry1-103 null allele of CRY1 was a negative control (Koornneef et al., 1998). While bas1-2 and sob7-1 showed no difference from Col-0, the double mutant flowered earlier, indicating that both genes are involved in this light response (Figure 6). ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34

Adult phenotypes not associated with a direct light response were investigated to determine if BAS1 and SOB7 are involved exclusively in photomorphogenesis. The bas1-2sob7-1 mutant had larger rosettes and longer siliques than Col-0 or either single mutant, suggesting that BAS1 and SOB7 affect many aspects of Arabidopsis growth in a redundant manner (Figure 7a,b). In addition, the primary inflorescence stem height of bas1-2 and bas1-2sob7-1 was taller than sob7-1 and Col-0, indicating that BAS1 and SOB7 do not act redundantly in all respects (Figure 7c). BAS1 and SOB7 metabolize BRs Endogenous BL levels are extremely low, making accurate quantification difficult. By utilizing approximately 100 g of rosette tissue the endogenous levels of CS and several of its precursors can be accurately measured (Noguchi et al., 1999). If SOB7 uses CS as a substrate, then we would predict sob7-D to accumulate less CS than Col-0. In fact sob7-D accumulated less CS and 6-deoxocastasterone than Col-0, presumably due to enhanced metabolism. While there was no difference in the level of BRs in sob7-1, the bas1-2sob7-1 double mutant had increased levels of CS and 6-deoxocastasterone indicating that these BRs are putative substrates of SOB7 (Figure 8). 3-dehydro-6-deoxoteasterone and 6deoxotyphasterol were also significantly reduced in sob7-D, though there was no increase in any of the null mutants, suggesting only tentatively that these precursors may also be SOB7 substrates. Similarly, typhasterol levels appear to be reduced in sob7-D compared to the other genotypes, though there was no increase in any of the null mutants.


Figure 6. BAS1 and SOB7 are genetically redundant with respect to long-day flowering time. The single-null mutants flower the same time as the wild type, but the double-null mutant flowers earlier. Measured as number of leaves at bolting (a), and number of days from sowing to bolting (c). Pictures are shown of representative plants (b). Bar ¼ 2.54 cm. *Mean significantly different from the Col-0 mean using an unpaired t-test, two-tailed P < 0.05.

Because BL is not detectable in the above assay, we utilized a physiological assay to investigate the possibility that BL is a substrate of SOB7. The Arabidopsis hypocotyl provides a simple bioassay for BL responsiveness, which is measured as an increase in hypocotyl length after growth on BL-containing media. BL is a substrate of BAS1, and a null allele of BAS1 is more responsive to exogenous BL, while the overexpressor is less responsive (Turk et al., 2003). We therefore compared the BL responsiveness of BAS1 mutants to SOB7 mutants. Null alleles of both genes conferred hypocotyls that were more responsive to BL than the wild type, while both overexpression lines were less responsive, indicating that BL is a putative substrate of SOB7 (Figure 9). BAS1 and SOB7 have unique biochemistries Because BAS1 is a carbon-26-hydroxylase of BL and CS (Turk et al., 2003) and SOB7 appears to utilize BL and CS as substrates (Figures 8 and 9), we tested the hypothesis that SOB7 is also a carbon-26-hydroxylase. We incubated 2week-old seedlings in 10 lg of BL or CS for 24 h and then measured the remaining levels of unhydroxylated BL and CS

as well as the amount hydroxylated at carbon 26 (26-OHBL and 26-OHCS, respectively) (Figure 10). The sob7-D overexpression mutant showed enhancement of BL catabolism (Figure 10a), but surprisingly there was no change in the resulting amount of 26-OHBL (Figure 10b), suggesting that SOB7 can use BL as a substrate but does not hydroxylate it at carbon 26. Overexpression of BAS1 raised the level of 26OHCS and underexpression reduced 26-OHCS to an undetectable level (Figure 10d). In contrast, the genetic state of SOB7 had no effect on 26-OHCS levels, again suggesting that SOB7 is not a carbon-26-hydroxylase. Discussion In this study, we identified two SOB7 overexpression mutants in two independent activation-tagging mutagenesis screens. The sob7-DphyB-4 mutant was identified in a photomorphogenesis screen for dominant suppressors of a weak allele of PHYB and the sob7-Der-116 mutant was identified in a screen for dominant suppressors of a weak allele of ERECTA. These mutants are similar in that they have the characteristic BR-deficient dwarf phenotype typified by ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34


(a)

(b)

(c)

Figure 7. BAS1 and SOB7 are genetically redundant with respect to rosette diameter (a) and silique length (b) but not plant height (c). *Mean significantly different from the Col-0 mean using an unpaired t-test, two-tailed P < 0.05.

the det2-1 mutant (Figure 1a). Backcrossing an overexpression mutant, recapitulation with a rescued plasmid, and cDNA overexpression driven by the constitutive CaMV 35S promoter each resulted in the same BR-deficient dwarf phenotype, leading to the conclusion that the BR-deficient dwarf phenotype is caused only by SOB7 overexpression and does not depend on phyB-4, er-116 or second site mutations (Figure 1a). We therefore conducted all SOB7 overexpression experiments with sob7-D, an overexpressor in the Col-0 background obtained by backcrossing the sob7Der-116 mutant. Interestingly, the sob7-D phenotype is nearly identical to that of the bas1-D mutant identified in the same photomorphogenesis screen (Neff et al., 1999). To supplement the SOB7 overexpression studies, an SOB7 loss-of-function mutant was desired. An RNAi approach was deemed tenuous due to the propensity for ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34

RNAi to inactivate homologous genes or produce only partial loss-of-function. We therefore concentrated only on the identification of severe T-DNA disruptions of the SOB7 open reading frame. Analysis of all SOB7 T-DNA insertional mutants available at the time of this study resulted in the identification of sob7-1, which contains a head-to-head T-DNA insertion in the third exon (Figure 1c). Such a dramatic disruption of the open reading frame is likely to lead to the complete loss of gene function, especially considering the position of the insertion upstream of the critical heme-binding domain required for cytochrome P450s to function (Poulos et al., 1985). In addition, there was no SOB7 transcript detected in the sob7-1 mutant by RT-PCR (Figure 1b). The sob7-1 phenotypes reported in this study are likely caused by the loss of SOB7 due to the fact that the phenotypes are vectorally opposite from that of sob7-D (Figures 4, 8 and 9) and are similar to the two independent loss-of-function alleles of the homologous BAS1 gene (Turk et al., 2003; this study). BAS1 and SOB7 function redundantly in photomorphogenesis BRs are considered to be growth-promoting hormones that act to enhance cell expansion, and as such their function is counter to that of hypocotyl photomorphogenesis. The hypocotyl elongates to move the seedling meristem through the dark soil, but upon light perception at the surface the hypocotyl stops elongating. Hypocotyl cell expansion via BR action would therefore provide negative regulation of hypocotyl photomorphogenesis. In our previous study we proposed that light signaling upregulates the BAS1 protein, which functions to inactivate BRs and thereby promote hypocotyl photomorphogenesis through the removal of a negative regulator (Turk et al., 2003). Based on the reduced responsiveness of the bas1-2sob7-1 double mutant to light (Figure 4), SOB7 also appears to provide positive modulation of hypocotyl photomorphogenesis by inactivating the negative regulation via BRs. In the cotyledon however, the response to light is to expand, which makes it reasonable to assume that BRs would act as positive modulators of photomorphogenesis in this organ. In fact the bas1-2sob7-1 double mutant is more responsive to light (greater expansion) (Figure 5), presumably due to the reduced ability to inactivate BRs and the concomitant increase in active BR levels. As endogenous BRs appear to generally promote growth, whereas responses to light result in growth inhibition in some cases (e.g., hypocotyl growth inhibition) and promotion in others (e.g., cotyledon expansion), BRs can be thought of as both positive and negative modulators in photomorphogenesis. The transition from vegetative growth to reproductive growth is marked by elongation of the inflorescence stem, making it likely that BRs act as positive modulators of


Figure 8. Castasterone and 6-Deoxocastasterone are the likely substrates of CYP72C1. Endogenous BR levels (ng g)1 fresh weight). *Mean significantly different from Col-0 (unpaired t-test, two-tailed P < 0.05). ND, not detected; NA, not analyzed. Arrows indicate biosynthetic order starting with 24-Methylenecholesterol.

ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34


Figure 9. BAS1 and SOB7 mutants have similar hypocotyl responses to exogenous BL. Hypocotyl length of seedlings grown in white light (85 lmol m)2 sec)1) on media without BL (open/left) and with BL (closed/ right). *Mean significantly different from Col-0 (unpaired t-test, two-tailed P < 0.05).

photomorphogenesis in this case. This assumption is supported by the fact that the bas1-2sob7-1 double mutant flowers early (Figure 6), presumably due to an increase in BR levels during this developmental transition. Germination, shade avoidance, and phototropism are other light responses that require changes in cell expansion. With elevated levels of active BRs, the bas1-2sob7-1 double mutant may be useful for dissection of BR action in these pathways, especially when combined with previously identified mutants specific to each pathway. BAS1 and SOB7 have unique biochemical mechanisms of BR inactivation To date, six unique cytochrome P450s (not counting paralogs) responsible for BR biosynthesis have been identified. DWF4 (CYP90B) (Choe et al., 1998) and CPD (CYP90A) (Szekeres et al., 1996) are close homologs responsible for C-22 and C-23 hydroxylation reactions, respectively, in the Arabidopsis BR biosynthetic pathway. DDWF1 and D2/ CYP90D2 are BR biosynthetic enzymes identified in pea and rice, respectively, and neither has redundant or paralogous cytochrome P450s identified as yet (Hong et al., 2003; Kang et al., 2001). The Arabidopsis CYP85A1 and CYP85A2 proteins are both BR C-6 oxidases and the only known instance of redundant BR biosynthetic enzymes (Shimada et al., 2003). In addition, cytochrome P450 inhibitors block the conversion of CS to BL indicating that the BL biosynthetic enzyme is also likely to be a cytochrome P450 (Kim et al., 2004). In this report we provide evidence to suggest that BAS1 and SOB7 are BR inactivation enzymes with redundant physiological functions but unique enzymatic activities. They both appear to utilize BL and CS as substrates, as well ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34

Figure 10. BAS1 and SOB7 have distinct and shared biochemistries. Metabolism of exogenous BL (a) and detection of the resulting 26-OHBL (b). Metabolism of exogenous CS (c) and detection of the resulting 26-OHCS (d). *Mean significantly different from Col-0 (unpaired t-test, two-tailed P < 0.05).

as one or more of the 6-deoxo precursors. Both enzymes reduce the endogenous levels of CS when overexpressed, while the double-null mutant has increased endogenous levels of CS (Figure 8). Although BL is not detectable in our endogenous BR assay, both overexpression mutants were less responsive to exogenous BL while both null mutants were more responsive (Figure 9). The BAS1 overexpression mutant has enhanced carbon 26-hydroxylase activity in our feeding assay, while the SOB7 overexpression mutant has no change (Figure 10), suggesting that SOB7 is not a carbon

32 Edward M. Turk et al. 26-hydroxylase. This conclusion is supported by the fact that the BAS1 null mutant has reduced carbon 26-hydroxylase activity while the SOB7 null mutant is unchanged (Figure 10). SOB7 may catalyze hydroxylation of carbons other than C-26 on active BRs, possibly creating unstable products. It is also possible that SOB7 hydroxylates and inactivates/degrades precursors to CS and BL. Our previous report demonstrated that BAS1 has CS and BL carbon 26-hydroxylase activity when expressed in yeast cells (Turk et al., 2003), which carry a clone for the Arabidopsis NADPH-P450 reductase (Urban et al., 1997). The enzymatic activity of three different BR C-6 oxidase cytochrome P450s have also been determined by this assay (Bishop et al., 1999; Shimada et al., 2001, 2003). In contrast, the enzymatic activity of D2/CYP90D2 was not detectable by this assay (Hong et al., 2003). Likewise, we could not detect SOB7 activity using this assay (unpublished data). Elucidation of the SOB7 biochemistry may require the development of novel in vitro or in vivo techniques. Once these techniques have been established, we can then address questions related to differential specificity of BAS1 and SOB7 substrate binding. Future studies with transgenic plants expressing translational fusions between each of these genes (under the control of their endogenous promoter) and reporters such as GUS or GFP, as well as studies incorporating antibodies raised against these two proteins, may uncover the degree to which these two enzymes act independently. In addition, these experiments will allow us to address the possibility that the loss of one protein could be compensated for by a shift in the expression pattern of the functional homolog. These studies set the stage for future biochemical analysis addressing the mechanism of redundancy and synergy between these two cytochrome P450s. They also illustrate the importance of BR inactivation with regard to both photomorphogenesis and plant development. The identification of both BAS1 and SOB7 in a gain-of-function mutant screen, coupled with the dramatic genetic interactions conferred between loss-of-function alleles for each of these genes, illustrates the usefulness of activation-tagging mutagenesis for the initial identification of potentially redundant genes involved in developmental processes such as photomorphogenesis.

Identification of sob7-DphyB-4 and molecular cloning of SOB7 The activation-tagging construct, pDSK2-7, was used in a largescale activation-tagging suppressor screen of phyB-4 as previously described in Neff et al. (1999). The pDSK2-7 plasmid was constructed by inserting a 254-base pair SpeI/XbaI fragment from the 5¢ end of a DS transposon (Wirtz et al., 1997) into the unique SpeI site of pSKI074 (Weigel et al., 2000), followed by inserting a 331-base pair SalI/XhoI fragment from the 3¢ end of the same transposon into the unique SalI site of the resulting plasmid (Z. Wang, Carnegie Institution of Washington, Stanford CA, USA, personal communi2 cation). Plants were transformed via the floral dip method using Agrobacterium tumefaciens strain GV3101 as described in Clough and Bent (1998). T1 seeds were grown on plates containing kanamycin (30 mg l)1) in continuous white light (35 lM m)2 sec)1) for 7 days. The sob7-DphyB-4 double mutant was identified by its shorter hypocotyl when compared with phyB-4 mutant plants. T2 seeds from self-fertilized sob7-DphyB-4 were screened on plates with and without kanamycin to confirm that the mutant phenotype is linked to the transgene. T2 seedlings segregated 3:1 kanr:kans, indicating that there was one locus containing a T-DNA. T3 seeds from self-fertilized T2 plants were screened in the same way to confirm the genetic heritability of the mutant phenotype and linkage to the transgene. The presence of the phyB-4 mutation was confirmed in the sob7-DphyB-4 mutant as in Neff et al. (1999). Southern blot analysis identified one T-DNA insertion in the sob7DphyB-4 mutant. Digestion of genomic DNA from the sob7-D mutant with EcoRI and ligation with T4 DNA ligase (New England BioLabs, Beverly, MA, USA) resulted in an 8-kb-rescued plasmid, pSOB7-RI. Genomic DNA flanking the T-DNA was sequenced with primer 5¢-TAA GAT CAC GGA ATT TCT GAC AGG A-3¢ to determine the T-DNA insertion site (3696 bp upstream of the start codon of SOB7/CYP72C1/At1g17060).

Identification of sob7-Der-116 and molecular cloning of SOB7 The activation-tagging construct, pSKI105 (Weigel et al., 2000), was used in a large-scale activation-tagging suppressor screen of er-116. To clone the activation-tagged gene, genomic DNA from 2-week-old sob7-D/er-116 seedlings was isolated utilizing the Nucleon Phytopure DNA Extraction Kit (Amersham Life Science, Buckinghamshire, UK). Five micrograms of genomic DNA was digested with KpnI and purified with the GENECLEAN Turbo DNA Preps (QBIOGENE, Carlsbad, CA, USA). Half of the digested DNA was ligated in a 100 ll volume overnight at 16C. Four microliters of the ligated DNA was used as a template for inverse PCR with primers 5¢-AAT TAA CCC TCA CTA AAG GGA AC-3¢ and 5¢-GAT ATC TAG ATC CGA AAC TAT CAG TG-3¢. The PCR product was sequenced to determine the T-DNA insertion site (546 bp upstream of the start codon of SOB7/ CYP72C1/At1g17060).

Experimental procedures Recapitulation of the sob7-D mutant phenotype Plant material All plants used in this study are Arabidopsis thaliana ecotype Columbia (Col-0). The cry1-103 allele is described in Liscum and Hangarter (1991). The phyB-9 allele is described in Reed et al. (1998). The phyA-211 allele is described in Reed et al. (1994). The bas1-2 allele and the sob7-1 allele were identified from the SALK_006781 and SALK_120416 seed pools, respectively (Alonso et al., 2003).

The sob7-D recapitulation construct was generated by ligation of an SmaI PCR fragment (primers 5¢-TCT CCC GGG GTA ATA CGA CTC ACT ATA GGG CGA-3¢ and 5¢-TCT CCC GGG CTG CAA GCA CCA AAG GAG CAG CAA-3¢) from sob7-Der-116 into pCAMBIA2300 cut with SmaI. This PCR fragment contained the four enhancer elements from the constitutive CaMV 35S promoter, 546 bp noncoding 5¢-UTR, the SOB7 open reading frame, and 1.5 kb of noncoding 3¢ UTR of the SOB7 gene. Col-0 was transformed with this ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34

Genetic interactions between BAS1 and SOB7 33 construct via Agrobacterium-mediated transformation as described in Clough and Bent (1998). Primary transformants were screened on kanamycin (50 mg l)1)-containing plates.

RT-PCR RT-PCR analysis was performed on total RNA isolated, using the RNeasy Plant Mini kit with on-column DNase-treatment (Qiagen, Valencia, CA USA), from 5-day-old seedlings grown in continuous white light. Total cDNA was synthesized using SuperScriptIII FirstStrand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA, USA). PCR was performed on dilute cDNA using Immolase hot start DNA polymerase (BioLine, Randolph, MA, USA). The BAS1 gene was amplified using primers 5¢-CACC ATG GAG GAA GAA AGT AGC AGC TGG TTC-3¢ and 5¢-TCA ATC CTC ATG ATT GGT CAA TCT CCG GAA GG-3¢. The SOB7 gene was amplified using primers 5¢-CACC ATG TTA GAG ATC ATT ACG GTA AGA AAA GTG-3¢ and 5¢-CTA CAG TTT TCG GAT GAT CAA ATG AGC-3¢. The ubiquitin10 (UBQ10) gene was used as an internal control. UBQ10 was amplified using primers 5¢-GGT ATT CCT CCG GAC CAG CAG C-3¢ and 5¢-CGA CTT GTC ATT AGA AAG AAA GAG ATA ACA GGA ACG G-3¢. The linear range of accuracy for the detection of each transcript was established by comparing samples run for different numbers of cycles. All final RT-PCR experiments were performed at least in duplicate.

Sequence analysis The predicted amino acid sequence of BAS1 and SOB7 from Arabidopsis was aligned with the CLUSTALX program (Thompson et al., 1997) with a gap opening penalty of 35.00 and a gap extension penalty of 0.75 (4 gaps). Amino acids were shaded with BOXSHADE 3 3.2 (http://www.ch.embnet.org). SIGFIND (139.91.72.10/sigfind/ sigfind.html) predicted the endoplasmic reticulum targeting sequences and SignalP 3.0 (http://www.cbs.dtu.dk/services/SignalP/) predicted that the BAS1 targeting sequence was a signal anchor (no peptide cleavage site), while the SOB7 targeting sequence was a signal peptide (cleavage site between amino acids 26 and 27). The cytochrome P450 cysteine heme-iron ligand signature was determined by PROSITE scan (http://www.expasy.ch/).

Hypocotyl measurements Seed sterilization, media preparation, growth conditions, and hypocotyl measurements were carried out as detailed in Turk et al. (2003).

Cotyledon area measurements Cotyledons from 5-day-old seedlings were removed from solid media plates as described in Turk et al. (2003) and placed on transparent tape and digitized with a flatbed scanner at a resolution of 600 dpi. Cotyledons were measured using ImageJ 1.29J (NIH, Bethesda, MD, USA; http://rsb.info.nih.gov/ij/java1.3.1). All experiments were performed in triplicate (n ‡ 30) on media without sucrose.

Epidermal imprints and epidermal cell area measurements Five-day-old seedlings grown in continuous white light were used for epidermal imprints. Imprints were taken by coating the top of the cotyledon with clear fingernail polish. After drying the polish was removed with transparent tape. The tape was fastened to a microscope slide, and the imprint was viewed and photographed at 200· ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34

4 using the Nikon Eclipse E600 microscope (Nikon, Tokyo, Japan), QImaging Retiga Ex camera (QImaging, Burnaby BC, Canada), and OpenLab 3.1.4 software (Improvision, Lexington, MA, USA). At least six cotyledons from each of three replicate experiments were digitized. The area from between five and 10 epidermal cells from each cotyledon was measured using ImageJ 1.29J (NIH; http:// rsb.info.nih.gov/ij/java1.3.1).

GC-MS analysis GC-MS analysis was carried out on a mass spectrometer (JMS-AM SUN200; Jeol, Tokyo, Japan) connected to a gas chromatograph (6890A; Agilent Technologies, Wilmington, DE, USA) with a capillary column DB-5 (0.25 mm · 15 m, 0.25 lm film thickness; J&W Scientific, Folsom, CA, USA). Analysis was conducted under the following conditions: ionization, EI (70 eV); carrier gas, helium at a flow rate of 1 ml min)1; injection temperature, 280C; column temperature, 80C for 1 min, elevated to 320C at 30C min)1, then maintained at 320C.

Metabolism of applied BL and CS in Arabidopsis seedlings All metabolic experiments were carried out as described in Turk et al. (2003).

Quantification of BRs in Arabidopsis adult tissue BR purification and quantification from 6-week-old rosette tissue grown under short-day growth-chamber conditions (8 h light:16 h dark) were carried out according to the methods described in Noguchi et al. (1999). Upon request, all novel materials described in this article will be made available for non-commercial research purposes in a timely manner.

Acknowledgements We thank Makoto Kobayashi and Masayo Sekimoto for their technical assistance; Zhiyong Wang, Joanne Chory, Detlef Weigel, and Barbara Baker for the pDKS2-7 clone; as well as Ian H. Street and Leeann Thornton for critical review of the manuscript. This project was directly funded by the National Science Foundation (IBN no. 0114726 to M.M.N. and MCB no. 0112278 to J.C.W.). Q.I.T. and G.M. were supported by the National Science Foundation Research Experience for Undergraduates (IBN no. 0114726 to M.M.N.). J.M.W. was supported in part by the Department of Energy (no. DEFG02-02ER15340 to M.M.N.). J.Z. was supported in part by the Monsanto Corporation (no. 46011J to M.M.N.) and the Department of Energy (no. DE-FG02-02ER15340 to M.M.N.).

References Alonso, J.M., Stepanova, A.N., Leisse, T.J. et al. (2003) Genomewide insertional mutagenesis of Arabidopsis thaliana. Science, 301, 653–657. Arabidopsis Genome Initiative (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature, 408, 796–815. Bak, S. and Paquette, S. (2004) The Arabidopsis P450, Cytochrome b5, P450 Reductase, and Glycosyltransferase Family 1 Site at PlaCe. (http://www.place.kvl.dk/): The Center for Molecular Plant 5 Physiology. http://www.P450.kvl.dk/Arab_cyps/CYP72C1.html.

34 Edward M. Turk et al. Neff, M.M., Turk, E. and Kalishman, M. (2002) Web-based primer design for single nucleotide polymorphism analysis. Trends Genet. 18, 613–615. Nelson, D.R., Schuler, M.A., Paquette, S.M., Werck-Reichhart, D. and Bak, S. (2004) Comparative genomics of rice and Arabidopsis. Analysis of 727 cytochrome P450 genes and pseudogenes from a monocot and a dicot. Plant Physiol. 135, 756–772. Noguchi, T., Fujioka, S., Choe, S., Takatsuto, S., Yoshida, S., Yuan, H., Feldmann, K.A. and Tax, F.E. (1999) Brassinosteroid-insensitive dwarf mutants of Arabidopsis accumulate brassinosteroids. Plant Physiol. 121, 743–752. Noguchi, T., Fujioka, S., Choe, S., Takatsuto, S., Tax, F.E., Yoshida, S. and Feldmann, K.A. (2000) Biosynthetic pathways of brassinolide in Arabidopsis. Plant Physiol. 124, 201–209. Poulos, T.L., Finzel, B.C., Gunsalus, I.C., Wagner, G.C. and Kraut, J. (1985) The 2.6-A crystal structure of Pseudomonas putida cytochrome P-450. J. Biol. Chem. 260, 16122–16130. Reed, J.W., Nagatani, A., Elich, T.D., Fagan, M. and Chory, J. (1994) Phytochrome A and phytochrome B have overlapping but distinct functions in Arabidopsis development. Plant Physiol. 104, 1139– 1149. Reed, J.W., Elumalai, R.P. and Chory, J. (1998) Suppressors of an Arabidopsis thaliana phyB mutation identify genes that control light signaling and hypocotyl elongation. Genetics, 148, 1295– 1310. Shimada, Y., Fujioka, S., Miyauchi, N., Kushiro, M., Takatsuto, S., Nomura, T., Yokota, T., Kamiya, Y., Bishop, G.J. and Yoshida, S. (2001) Brassinosteroid-6-oxidases from Arabidopsis and tomato catalyze multiple C-6 oxidations in brassinosteroid biosynthesis. Plant Physiol. 126, 770–779. Shimada, Y., Goda, H., Nakamura, A., Takatsuto, S., Fujioka, S. and Yoshida, S. (2003) Organ-specific expression of brassinosteroidbiosynthetic genes and distribution of endogenous brassinosteroids in Arabidopsis. Plant Physiol. 131, 287–297. Szekeres, M., Nemeth, K., Koncz-Kalman, Z., Mathur, J., Kauschmann, A., Altmann, T., Redei, G.P., Nagy, F., Schell, J. and Koncz, C. (1996) Brassinosteroids rescue the deficiency of CYP90, a cytochrome P450, controlling cell elongation and de-etiolation in Arabidopsis. Cell, 85, 171–182. Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F. and Higgins, D.G. (1997) The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876–4882. Turk, E.M., Fujioka, S., Seto, H., Shimada, Y., Takatsuto, S., Yoshida, S., Denzel, M.A., Torres, Q.I. and Neff, M.M. (2003) CYP72B1 inactivates brassinosteroid hormones: an intersection between photomorphogenesis and plant steroid signal transduction. Plant Physiol. 133, 1643–1653. Urban, P., Mignotte, C., Kazmaier, M., Delorme, F. and Pompon, D. (1997) Cloning, yeast expression, and characterization of the coupling of two distantly related Arabidopsis thaliana NADPHcytochrome P450 reductases with P450 CYP73A5. J. Biol. Chem. 272, 19176–19186. Weigel, D., Ahn, J.H., Blazquez, M.A. et al. (2000) Activation tagging in Arabidopsis. Plant Physiol. 122, 1003–1013. Wirtz, U., Osborne, B. and Baker, B. (1997) Ds excision from extrachromosomal geminivirus vector DNA is coupled to vector DNA replication in maize. Plant J. 11, 125–135. Yokota, T., Ogino, Y., Suzuki, H., Takahashi, N., Saimoto, H., Fujioka, S. and Sakurai, A. (1991) Metabolism and biosynthesis of brassinosteroids. ACS Symp. Ser. – Am. Chem. Soc. 474, 86– 96.

Bishop, G.J. and Yokota, T. (2001) Plants steroid hormones, brassinosteroids: current highlights of molecular aspects on their synthesis/metabolism, transport, perception and response. Plant Cell Physiol. 42, 114–120. Bishop, G.J., Nomura, T., Yokota, T., Harrison, K., Noguchi, T., Fujioka, S., Takatsuto, S., Jones, J.D. and Kamiya, Y. (1999) The tomato DWARF enzyme catalyses C-6 oxidation in brassinosteroid biosynthesis. Proc. Natl Acad. Sci. USA, 96, 1761–1766. Choe, S., Dilkes, B.P., Fujioka, S., Takatsuto, S., Sakurai, A. and Feldmann, K.A. (1998) The DWF4 gene of Arabidopsis encodes a cytochrome P450 that mediates multiple 22alpha-hydroxylation steps in brassinosteroid biosynthesis. Plant Cell, 10, 231–243. Choi, Y.H., Fujioka, S., Harada, A., Yokota, T., Takatsuto, S. and Sakurai, A. (1996) A brassinolide biosynthetic pathway via 6deoxocastasterone. Phytochemistry, 43, 593–596. Clough, S.J. and Bent, A.F. (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16, 735–743. Clouse, S.D. and Sasse, J.M. (1998) Brassinosteroids: essential regulators of plant growth and development. Ann. Rev. Plant Physiol. Plant Mol. Biol. 49, 427–451. Fujioka, S. and Yokota, T. (2003) Biosynthesis and metabolism of brassinosteroids. Annu. Rev. Plant Biol. 54, 137–164. Fujioka, S., Noguchi, T., Watanabe, T., Takatsuto, S. and Yoshida, S. (2000) Biosynthesis of brassinosteroids in cultured cells of Catharanthus roseus. Phytochemistry, 53, 549–553. Grove, M.D., Spencer, G.F., Rohwedder, W.K., Mandava, N., Worley, J.F., Warthen, J.D., Jr, Steffens, G.L., Flippen-Anderson, J.L. and Cook, J.C., Jr (1979) Brassinolide, a plant growth-promoting steroid isolated from Brassica napus rape pollen. Nature, 281, 216–217. Hong, Z., Ueguchi-Tanaka, M., Umemura, K., Uozu, S., Fujioka, S., Takatsuto, S., Yoshida, S., Ashikari, M., Kitano, H. and Matsuoka, M. (2003) A rice brassinosteroid-deficient mutant, ebisu dwarf (d2), is caused by a loss of function of a new member of cytochrome P450. Plant Cell, 15, 2900–2910. Kang, J.G., Yun, J., Kim, D.H. et al. (2001) Light and brassinosteroid signals are integrated via a dark-induced small G protein in etiolated seedling growth. Cell, 105, 625–636. Kim, T.W., Chang, S.C., Lee, J.S., Hwang, B., Takatsuto, S., Yokota, T. and Kim, S.K. (2004) Cytochrome P450-catalyzed brassinosteroid pathway activation through synthesis of castasterone and brassinolide in Phaseolus vulgaris. Phytochemistry, 65, 679–689. Koornneef, M., Rolff, E. and Spruit, C. (1980) Genetic control of lightinhibited hypocotyl elongation in Arabidopsis thaliana (L.) Heynh. Z. Pflanzenphysiol, 100S, 147–160. Koornneef, M., Alonso-Blanco, C., Peeters, A.J. and Soppe, W. (1998) Genetic control of flowering time in Arabidopsis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 49, 345–370. Li, J. and Chory, J. (1997) A putative leucine-rich repeat receptor kinase involved in brassinosteroid signal transduction. Cell, 90, 929–938. Li, J., Nagpal, P., Vitart, V., McMorris, T.C. and Chory, J. (1996) A role for brassinosteroids in light-dependent development of Arabidopsis. Science, 272, 398–401. Liscum, E. and Hangarter, R.P. (1991) Arabidopsis mutants lacking blue light-dependent inhibition of hypocotyl elongation. Plant Cell, 3, 685–694. Neff, M.M. and Van Volkenburgh, E. (1994) Light-stimulated cotyledon expansion in Arabidopsis seedlings (the role of phytochrome B). Plant Physiol. 104, 1027–1032. Neff, M.M., Nguyen, S.M., Malancharuvil, E.J. et al. (1999) BAS1: A gene regulating brassinosteroid levels and light responsiveness in Arabidopsis. Proc. Natl Acad. Sci. USA, 96, 15316–15323.

Accession numbers: BT010564/At2g26710, AAD50024/At1g17060 ª Blackwell Publishing Ltd, The Plant Journal, (2005), 42, 23–34