Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

The Gibberellin perception system evolved to regulate a pre-existing GAMYB-mediated system during land plant evolution

The Gibberellin perception system evolved to regulate a pre-existing GAMYB-mediated system during... ARTICLE DOI: 10.1038/ncomms1552 Received 5 Aug 2011 | Accepted 18 oct 2011 | Published 22 nov 2011 The Gibberellin perception system evolved to regulate a pre-existing GAm YB-mediated system during land plant evolution 1 2,3 4 4 1 Koichiro Aya , Yuji Hiwatashi , mikiko Kojima , Hitoshi sakakibara , miyako ueguchi-Tanaka , 2,3,5 1 mitsuyasu Hasebe & makoto matsuoka Gibberellin (GA) controls pollen development in flowering plants via the GAm YB transcription factor. Here we show that GAm YB is conserved in Selaginella moellendorffii (lycophyte) and Physcomitrella patens (moss), although the former contains the GA signalling pathway, the latter does not. In the lycophyte, GA treatment promotes the outer wall development on microspores, whereas treatment with GA biosynthesis inhibitors disturbs its development. Contrary, in the moss, GAMYB homologue knockouts also produce abnormal spores that resemble Selaginella microspores treated with GA biosynthesis inhibitors and pollen grains of rice gamyb mutant. moreover, the knockouts fail to develop male organs, instead ectopically forming female organs. Thus, before the establishment of the GA signalling pathway, basal land plants, including mosses, contained a GAm YB-based system for spore and sexual organ development. subsequently, during the evolution from mosses to basal vascular plants including lycophytes, GA signalling might have merged to regulate this pre-existing GAm YB-based system. 1 2 Bioscience and Biotechnology Center, Nagoya University, Nagoya 464-8601, Japan. National Institute for Basic Biology, Okazaki 444-8585, Japan. 3 4 Department of Basic Biology, School of Life Science, The Graduate University for Advanced Studies, Okazaki 444-8585, Japan. RIKEN Plant Science Center, 1-7-22, Suehiro, Tsurumi, Yokohama 230-0045, Japan. ERATO, Japan Science and Technology Agency, Okazaki 444-8585, Japan. Correspondence and requests for materials should be addressed to K.A. (email: [email protected]). nATuRE C ommunICATIons | 2:544 | DoI: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. ARTICLE n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 n flowering plants, gibberellin (GA) regulates the formation of the abnormalities in spore development and showed the transcriptional outer wall on pollen and is therefore essential for pollen develop- reduction of CYP703 homologues, supporting that the regulation of 1–3 Iment . e Th outermost pollen wall, which is known as the exine CYP703 by GAMYB is conserved among land plant lineages. More- or exospore, consists largely of sporopollenin, an organic polymer over, the KOs failed to develop male organs, instead ectopically that is extremely resistant to degradation and is so durable that it forming female organs. On the basis of these results, we propose 4–6 has been found in microfossils that are over 500 million years old . that GA signalling might have merged to regulate this pre-existing In rice, CYP703A3 is an essential enzyme in sporopollenin biosyn- GAMYB-based system in reproductive development. thesis, and its expression is positively regulated by a transcription factor, GAMYB, which is controlled by the GA signalling path- Results 1,7,8 way . Previous anatomical and in silico studies suggest that, similar In silico screening for GAMYB-like genes. To test our hypothesis, we to flowering plants, present-day bryophytes, including mosses, and first searched for homologues of GAMYB in mosses and lycopods. lycophytes (lycopods) contain sporopollenin-containing outer walls A survey of public genomic phytozome database showed that the 9–11 on the surface of their spores and microspores , and genes encod- model moss P. patens has two GAMYB-like genes, PpGAMYB1 and ing CYP703 homologues have been detected in these genomes . PpGAMYB2, whereas the model lycopod S. moellendorffii contains Curiously, however, one of the basal vascular plants, Selaginella, a single GAMYB-like gene, SmGAMYB (Fig. 1a; Supplementary contains a functional GA perception system that is mediated by a Fig. S1a). A comparative analysis of these GAMYB-like genes with GA receptor, GIBBERELLIN-INSENSITIVE DWARF1 (GID1), GAMYBs in flowering plants revealed that these homologues shared and suppressor proteins (DELLAs), whereas one of the basal land three conserved regions (BOX1, BOX2 and BOX3) in addition to a 13–16 plants, Physcomitrella patens, does not contain this system . es Th e R2R3 DNA-binding domain and a putative microRNA 159 target observations led us to speculate that the ancestral land plants devel- site (Fig. 1a; Supplementary Fig. S1b), which are ubiquitous in 18,20 oped sporopollenin-containing outer walls on their spores, which GAMYBs of flowering plants . protected the spores from dryness and ultraviolet radiation on land, A previous study of the CYP703 family reported that three homo- in association with CYP703-like enzymes, and that GA signalling logues of this protein were present in P. patens (PpCYP703B1–3) pathway then evolved to control the development of the spore outer and one was present in S. moellendorffii (SmCYP703C1). wall via the modulation of CYP703-like expressions during the evolution from mosses to lycopods. Mutant and transgenic studies Molecular characterization of SmGAMYB and PpGAMYBs. To in flowering plants indicate that GAMYB preferentially modulates test the function of these putative GAMYB homologues, we fused other aspects of reproductive developments (for example, inflores - the GAMYB coding sequence (CDS) from each species to the cence formation and floral organ development as well as microspore rice GAMYB (OsGAMYB) promoter, and transformed these con- 1,7,8,17,18 development) . Moreover, considering that some GA-like structs into a rice loss-of-function mutant, gamyb-2. As previously substances (known as antheridiogens) are involved in the forma- reported, the observations by the transmission electron microscopy tion of sexual organs in fern gametophytes , it is further likely that (TEM) revealed that the wild type (WT) plants developed micro- this GA signalling pathway that involves GAMYB regulates a broad spores with a normal exine consisting of three layers in the anthers range of reproductive processes, including (micro)spore develop- (tectum, bacula and foot layer ; Fig. 1b and inset of Fig. 1c), whereas ment, in early vascular plants. gamyb-2 plants transformed with an empty vector developed micro- In this study, we identified the functional GAMYB homologues spores with a defective exine consisting of a single layer, resulting in in Selaginella moellendorffii and P. patens. In S. moellendorffii , GA shrunken and whitened anthers that lacked pollen grains relative to treatment promoted the outer wall development on microspores, yellow anthers of WT (Fig. 1b,c). e Th gamyb-2 plants transformed in association with the induction of CYP703 homologue. On the with OsGAMYB developed microspores with a normal exine struc- other hand, in P. patens with no functional GA signalling pathway, ture that consisted of three layers (Fig. 1c). Anther development the knockouts (KOs) of two GAMYB homologues had similar was partially rescued in transgenic plants carrying SmGAMYB R2R3 OsGAMYB Vector pOsGAMYB:: ATG Stop OsGAMYB An R2R3 SmGAMYB ATG Stop R2R3 WT gamyb-2 gamyb-2 PpGAMYB1/2 ATG Stop pOsGAMYB:: pOsGAMYB:: pOsGAMYB:: SmGAMYB PpGAMYB1 PpGAMYB2 WT Vector OsGAMYB Te Ba Ex Fl gamyb-2 gamyb-2 gamyb-2 gamyb-2 gamyb-2 SmGAMYB PpGAMYB1 PpGAMYB2 gamyb-2 gamyb-2 gamyb-2 Figure 1 | Conserved molecular function of SmGAMYB and PpGAMYBs. (a) The exon/intron structure of OsGAMYB and GAMYB homologues. The coding regions and introns are symbolized by boxes and lines, respectively. Black box, R2R3 Dn A-binding domain; grey boxes, Bo X1–3 domains; black line above gene structure, putative miR159 target site. (b,c), Flower (b) and the microspore outer wall (exine) (c) of wild-type (WT) rice plants and of transgenic plants transformed with empty vector, OsGAMYB and GAMYB homologues in the rice gamyb-2 background (n > 11). Inset indicates the magnification of WT exine layer. s cale bars indicate 2 mm in b and 5 µm in c. An, anther; Ba, bacula; Ex, exine; Fl, foot layer; Te, tectum. n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. BOX1 BOX1 miRNA BOX1 miRNA BOX2 miRNA BOX2 BOX2 BOX3 BOX3 BOX3 n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 ARTICLE Young 0.7 25 Premeiosis microspore Premeiosis 0.6 0.5 MMC YM T 15 0.4 0.3 0.2 0.1 as as Sense ND ND ND ND ND Young Old Apices Root Strobili GA GA GA GA GA GA 24 9 4 19 20 1 stem stem MMC YM Microsporangium (kbp) (fold) 0.6 – GA + GA 4 4 0.2 0.5 SmGA3ox 0.5 0.4 as as Sense 0.3 0.2 SmGAMYB 1.7 0.2 0.2 0.1 SmCYP703C1 4.3 YM Young Old Apices Root Strobili 1.1 MMC Sm6PDG stem stem 0.2 as as Sense Figure 2 | The spatioexpression pattern of SmGAMYB and SmCYP703C1 in microspore development of S. moellendorffii . (a,b) Quantitative RT–PCR analysis of SmGAMYB (a) and SmCYP703C1 (b) in different tissues. Data are shown as the mean values ± s.d. (n = 3). (c–e) In situ hybridization of SmGA3ox (c), SmGAMYB (d) and SmCYP703C1 (e) on microsporangium at the premeiosis stage (left and right) and young microspore stage (middle). s cale bars, 50 µm. as, anti-sense; mm C, microspore mother cell; T, tapetal cell; Ym , young microspore. (f) Endogenous GA content of S. moellendorffii microsporangia (grey) and young stems (black). Data are shown as the mean values ± s.d. (n = 3). FW, fresh weight; n D, no expected product was detected. (g) SmGA3ox, SmGAMYB and SmCYP703C1 expression in response to GA treatment in microsporangium. n umber on the right side of each panel indicates the expression ratio in GA treatment versus in m ock control, estimated by Image J software. (Fig. 1b). e Th anthers of these plants were slightly whitened ( Fig. 1b), S3c). Together, these results indicate that SmGAMYB can almost and the microspores exhibited a normal exine consisting of three completely substitute, and PpGAMYB proteins can partially sub- layers (Fig. 1c). In PpGAMYB1 and PpGAMYB2 transgenic lines, stitute, for OsGAMYB in the GA signalling pathway of rice anther anther development was also partially rescued (Fig. 1b), which development. exhibited microspores with a normal exine (Fig. 1c). We also exam- ined whether the CYP703 homologues in P. patens and S. moel- Role of GA in microspore development of S. moellendorffii . Dur- lendorffii had sporopollenin biosynthetic activity. For this analysis, ing the divergence of basal vascular plants including lycopods from we used SmCYP703C1 (the sole homologue detected in S. moellen- the moss lineage, a GA receptor, GID1 and DELLA repressors were dorffii ) and PpCYP703B2 (the dominant gene of three PpCYP703B established as key components for GA perception system in asso- homologues; see below). A defect in exine structure in pollen of the ciation with a functional GA biosynthetic pathway, indicating the 13,14 rice loss-of-function cyp703a3 mutant was fully rescued by intro- occurrence of the entire GA signalling pathway . However, the duction of the CDS for SmCYP703C1 under the control of the rice biological function of GA in basal vascular plants has been unclear, CYP703A3 promoter, and at least partially rescued by introduc- as GA treatment does not cause any developmental defects at tion of PpCYP703B2 with the same rice promoter (Supplementary the vegetative stage, such as increased stem or leaf elongation, in Fig. S2). Taken together, our results show that the biological function Selaginella . er Th efore, we hypothesized that, in such basal vascular of GAMYB and CYP703 homologues for the outer wall develop- plants, GA may be exclusively involved in the outer wall develop- ment is conserved among the land plant lineage. However, the level ment of microspores via GAMYB. Quantitative RT–PCR analysis of activity varied among each family member, at least in rice cells. using various tissues of S. moellendorffii revealed SmGAMYB and To further confirm the function of the GAMYB homologues in SmCYP703C1 were preferentially expressed in the reproductive the GA signalling pathway in rice cells, we compared the expres- organs, strobili (Fig. 2a,b). e Th microsporangium in the strobili of a sion pattern in the anthers of OsGAMYB transgenic plants with heterosporous lycophyte, S. moellendorffii, is an orthologous organ those of SmGAMYB or PpGAMYB2 transgenic plants, using two- of the anther of flowering plants. Microspores within the microspo - color microarrays hybridized with RNA isolated from rice gamyb-2 rangium correspond to pollen grains of flowering plants. Previous anthers as a common reference (Supplementary Methods, Data are observations suggested that GA signalling mediated by GAMYB is 1,7,8 available at Gene Expression Omnibus, accession code: GSE32652.). important for anther and pollen development in flowering plants . Most OsGAMYB-regulated genes that were expressed in response er Th efore, we next performed an in situ hybridization experiment to OsGAMYB expression (x axis) were also, at least partially, regu- to examine the spatiotemporal expression of SmGAMYB and lated at the transcriptional level by the SmGAMYB and PpGAMYB2 SmCYP703C1 during microsporangium development (Fig. 2c–e). expression (y axis; Supplementary Fig. S3a,b and Supplementary During the transition from premeiosis to young microspore stage, Data 1). e Th transcriptional recovery by SmGAMYB expression the expression of these genes was mainly limited to the tapetal cells was more tightly correlated to OsGAMYB-mediated modulation and microspores in a pattern similar to that of a GA biosynthesis than was the transcriptional recovery by PpGAMYB2 expression gene, GA3 oxidase (SmGA3ox; Fig. 2c–e), where the sporopol- 9–11,21–23 (correlation coefficient; 0.84 in Supplementary Fig. 3a and 0.71 in lenin material is produced for the outer wall development . Supplementary Fig. 3b). e Th difference in correlation coefficient Considering that GAMYBs, CYP703As and GA biosynthesis genes agrees with the rescued phenotype of transgenic rice plants (Fig. in flowering plants are also co-expressed in the tapetal cells and 1,8,12,24–26 1b,c). Consistently, the expression of an OsGAMYB target gene, microspores , these observations indicate that SmGAMYB CYP703A3, which was hardly detected in the gamyb-2 background, activity in these tissues of Selaginella is required to mediate GA was complemented by the introduction of SmGAMYB to a similar signalling during the outer wall formation of microspores. or higher degree than it was by OsGAMYB, whereas PpGAMYBs To further verify the importance of GA in spore wall formation, partially restored CYP703A3 expression, with PpGAMYB2 hav- we analysed the endogenous GA level in the microsporangium. We ing a stronger effect than did PpGAMYB1 (Supplementary Fig. detected a 13 non-hydroxylated bioactive GA, GA, in microsporangia, n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Expression levels Expression levels Content (pmol per gFW) ARTICLE n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 Mock GA Uni ** ** ** 5.0 ** ** 4.5 4.0 3.5 3.0 2.5 Uni+GA Pro Pro+GA 4 4 2.0 1.5 1.0 0.5 Mock GA Uni Pro Uni+GA Pro+GA 4 4 4 Figure 3 | Effects of GA or GA biosynthesis inhibitors on the exospore projection of the microspore surface of S. moellendorffii . (a) s Em images of the exospore of microspores developed by the treatment of GA or GA biosynthesis inhibitors in S. moellendorffii . s cale bars, 50 µm. Pro, prohexadione- calcium; u ni, uniconazole. (b) The height of the exospore projection on the microspore following treatment with GA or GA biosynthesis inhibitors. m easurements were made on s Em micrographs using Image J software, and data represent mean values ± s.d. (m ock control: n = 230; GA : n = 219; uniconazole (u ni): n = 321; prohexadione-calcium (Pro): n = 89; GA and uniconazole (u ni + GA ): n = 178; GA and prohexadione-calcium (Pro + GA ): 4 4 4 4 n = 91). **P < 0.01; Tukey’s test. and, to a lesser extent, in young stems (Fig. 2f ). Because bioactive rhizoids from a fraction of the protonema (Fig. 4c and Supplemen- GA negatively regulates the expression of some GA biosynthetic tary Fig. S4a,b). Subsequently, at the shoot apex of the gametophore, genes, including GA3ox, and positively regulates the expression of the archegonium and antheridium (Fig. 4d,e), which contain an egg 1,8,13 GAMYB and CYP703A in flowering plants , we decided to exam- and sperm cells, respectively, develop, and an egg cell is fertilized by a ine the expression levels of these homologues in the presence and sperm cell . Finally, the resulting zygote develops a sporophyte that absence of GA . Application of GA to the microsporangium sup- contains a spore-producing tissue, the sporangium (Fig. 4f ). Quan- 4 4 pressed the expression of SmGA3ox, whereas it induced SmGAMYB titative RT–PCR analysis revealed that transcripts for PpGAMYBs, and SmCYP703C1 (Fig. 2g). As is the case in the pollen grains of PpCYP703B1 and PpCYP703B2 were present in various tissues of flowering plants, these results demonstrate that the GA response is gametophore, but not in the protonema, whereas PpCYP703B3 present in the microspores of S. moellendorffii . transcript was not present in any tested tissues (Fig. 4a,b). In en, Th we examined the effect of GA or a GA biosynthesis sporangium of the sporophyte, PpGAMYB2 and PpCYP703B2 inhibitor, uniconazole, which inhibits a step of the GA biosynthesis were expressed more strongly than the other transcripts analysed pathway that is catalysed by ent-kaurene oxidase , on the exospore (Fig. 4a,b), indicating that they might have a dominant role in spo- development of S. moellendorffii microspores using scanning elec- rangium development. e Th spatiotemporal expression pattern of tron microscopy (SEM). e Th SEM observations revealed that the PpGAMYB1 and PpGAMYB2 was further examined by introduc- outer exospore of microspores in mock-treated plants was formed ing the β-glucuronidase (GUS) reporter gene just upstream of the as numerous sporopolleninous projections (Fig. 3a), the material of stop codon of each gene (Fig. 4c–f ). During the young gametophore 11,23 which originates exclusively from the tapetum and microspores . stage, GUS activity in the PpGAMYB1-GUS and PpGAMYB2-GUS e Th heights of the outer exospore projections of GA -treated micro- plants was mainly detected in the shoot apices and partially in spores were significantly greater than those of mock-treated micro - the rhizoid (Fig. 4c). During the reproductive stage, the GUS sig- spores, whereas the microspores treated with uniconazole produced nal was detected in the developing archegonium and antheridium less and shorter outer exospore projections than did the mock- (Fig. 4d,e). Aer ft the start of sporogenesis, the GUS signal in these treated microspores (Fig. 3a,b). Such defects in outer exospore two lines was detected in the sporangium (Fig. 4f ). In situ hybridiza- projections were not detected in the microspores of plants that were tion revealed that PpGAMYB1 and PpCYP703B1 were co-expressed simultaneously treated with GA and uniconazole, and the number more strongly in spores than in spore sac cells (Supplementary and height of the outer exospore projections of the microspores of Fig. S5a,b; insets). In contrast, the transcripts of PpGAMYB2 and these plants were similar to those of GA -treated plants (Fig. 3a,b). PpCYP703B2, the predominantly expressed genes, co-localized in Moreover, a similar phenotype was induced upon treatment with spore sac cells of the sporangium (Supplementary Fig. S5c,d; insets), another GA biosynthesis inhibitor, prohexadione-calcium, which which correspond to tapetal cells in flowering plants . inhibits another step of the GA biosynthesis pathway catalysed by To further investigate the function of PpGAMYBs, we generated GA3ox , and was rescued by co-treatment with GA (Fig. 3a,b), KO lines of PpGAMYB1 and PpGAMYB2 (PpGAMYB1-KO and strongly suggesting that these phenomena depend on a GA defi - PpGAMYB2-KO) using homologous recombination. er Th e were no ciency. es Th e results show that, similar to flowering plants, basal morphological and growth defects in the protonemata and game- vascular plants require the GA signalling pathway to promote the tophores in each single KO line (Supplementary Fig. S4a–c). On development of the microspore outer exospore wall. the other hand, aer ft 4 or 5 weeks of antheridium and archegon - ium induction, antheridium formation was strongly disturbed in Role of PpGAMYBs in spore development of P. patens. We first PpGAMYB2-KO, whereas archegonium formation was promoted in examined the expression pattern of PpGAMYBs to estimate the PpGAMYB-KOs (Fig. 5a). Most PpGAMYB2-KO plants displayed biological function of PpGAMYBs in P. patens, which has no func- ectopic archegonium formation at the site of antheridium forma- 13,14 tional GA perception system and GA biosynthetic pathway (Fig. tion, and a similar, but milder, trend was seen in PpGAMYB1-KO 4a,b). P. patens develops protonemata colony from spores, and then (Fig. 5b). Moreover, some abnormal antheridia that were light brown differentiates gametophores with haploid leafy shoots and root-like were observed in these KOs (asterisks in Fig. 5c). Consistent with these n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Height of microspore exospore projection (µm) n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 ARTICLE 2.0 3.5 3.0 1.6 2.5 1.2 2.0 1.5 0.8 1.0 0.4 0.5 Protonema Shoot Rhizoid Sporangium Protonema Shoot Rhizoid Sporangium Gametophore Gametophore PpGAMYB1 PpGAMYB2 PpGAMYB1-GUS PpGAMYB2-GUS -GUS -GUS PpGAMYB1-GUS PpGAMYB2-GUS PpGAMYB1-GUS PpGAMYB2-GUS Figure 4 | Expression pattern of PpGAMYBs and PpCYP703Bs in various tissues of P. patens. (a) Quantitative RT–PCR analysis of PpGAMYB1 (grey) and PpGAMYB2 (black) in different tissues of P. patens. Data represent mean values ± s.d. (n = 3). (b) Quantitative RT–PCR analysis of PpCYP703B1 (grey) and PpCYP703B2 (black) in different tissues of P. patens. Data represent mean values ± s.d. (n = 3). (c–f) Gus expression pattern in the leafy gametophore (c), archegonium (d), antheridia (e) and sporangium (f) of PpGAm YB1-Gus (left) and PpGAm YB2-Gus (right) reporter lines. Arrows mark the shoot apex. s cale bars indicate 1 mm in c, 100 µm in d,e, and 200 µm in f. abnormal antheridium phenotypes, the number of sporophytes was Table 1). Considering that the effect of acetolysis is dominantly reduced (Fig. 5d,e), while developed sporophytes appeared to be reflected by the exospore content, the result suggests that dynamic normal in PpGAMYB-KOs (Fig. 5f ). alternation of exospore structure occurs in these KOs. We also Within the sporophyte, however, there were differences in the observed the ultrastructure of spore walls developed in PpGAMYB- surfaces of the spores in each KO (Fig. 6a,b), and the number of KOs by the TEM and revealed structural differences in PpGAMYB- 32,33 spores in each KO was less than in the WT (Table 1). e Th SEM KOs. As described previously , the spore wall in the WT consists observations of mature spores revealed that WT spores produced of four layers, the intine, inner exospore, outer exospore and perine well-developed spiny projections on their surface, whereas develop- (Fig. 6d,e). Among these layers, the perine layer, which contributes to ment of such projections occurred incompletely in PpGAMYB-KOs, the formation of projections of spores, was severely defective in the resulting in a smoother surface (Fig. 6a,b) and a higher frequency of KO spores, especially in PpGAMYB2-KO (Fig. 6d,e). Moreover, the collapsed spores (arrowheads in Fig. 6a). We noticed that the PpGA- continuous outer exospore structure was sporadically interrupted in MYB2-KO had a more severe phenotype than the PpGAMYB1-KO, PpGAMYB1-KO (arrowheads in Fig. 6d), and oen ft in PpGAMYB2- which is consistent with PpGAMYB2 being expressed at a higher KO, whereas there were no obvious defects in the inner exospore level than PpGAMYB1 in WT sporangium (Fig. 4a). e Th collapsed and intine layers of PpGAMYB-KOs (Fig. 6d,e). es Th e observations spores observed in PpGAMYB-KOs suggest that the structural and are in good agreement with previous reports that the sporopollenin physical protection against the vacuum condition of SEM analysis material for perine and exospore is mainly supplied from the spores 11,34 might be reduced in PpGAMYB-KOs spores, compared with WT and spore sac cells , where PpGAMYBs and PpCYP703Bs were spores. We next treated spores with a solution of acetic anhydride expressed (Supplementary Fig. S5). and sulphuric acid, which generally disrupts all spore/pollen content Consistent with the abnormal perine and exospore layers of except for the exospore/exine by hydrolysis and esterification (ace - PpGAMYB-KOs, the expression of PpCYP703B1 and PpCYP703B2 30,31 tolysis) . e Th spores in PpGAMYB-KOs, especially in the PpGA- was lower in the sporangia of the KOs than in those of the WT MYB2-KO, were easily damaged by acetolysis treatment (percent- (Fig. 6f ). e Th suppression level of PpCYP703B1 was significantly age of acetolysis-sensitive spores in PpGAMYB1-KO, 23.32 ± 2.68%; greater in PpGAMYB1-KO than in PpGAMYB2-KO, whereas that PpGAMYB2-KO, 60.33 ± 8.97%; and WT, 14.63 ± 3.21%; Fig. 6c and of PpCYP703B2 was greater in PpGAMYB2-KO than in PpGAMYB1-KO n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Expression levels Rhizoid Shoot Expression levels ARTICLE n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 WT An WT WT WT Ar Sp 60 An PpGAMYB1-KO PpGAMYB1-KO PpGAMYB1-KO PpGAMYB1-KO 0 An Ar An Ar An Ar * * Sp Ar 4 Week 5 Week An ** ** PpGAMYB2-KO PpGAMYB2-KO PpGAMYB2-KO PpGAMYB2-KO An Ar Ar 40 * Ar ** Sp WT PpGAMYB1 PpGAMYB2 -KO -KO Figure 5 | Effect of PpGAMYB knockout on reproductive development in P. patens. (a) Quantification of antheridium and archegonium formation in the WT (blue), PpGAMYB1-Ko (red) and PpGAMYB2-Ko (green) knockout lines. The percentage of gametophores containing antheridia or archegonia relative to the total number of gametophores after 4 or 5 weeks in antheridium/archegonium induction conditions is given (n > 37). An, antheridium; Ar, archegonium. (b,c) m orphology of antheridia and archegonia in WT (upper), PpGAMYB1-Ko (middle) and PpGAMYB2-Ko (lower) P. patens. Ar*, ectopic archegonium. s cale bars indicate 250 µm in b and 100 µm in c. Asterisks mark the abnormal antheridium. (d) Quantification of sporangium formation in the WT and PpGAMYB knockout lines. The percentage of gametophores containing a sporangium relative to the total number of gametophores after 3 months under antheridium/acrhegonium induction conditions is given (n = 4). **P < 0.01; Tukey’s test. (e,f) m orphology of mature gametophores (e) and a sporophyte (f) of the WT (upper), PpGAMYB1-Ko (middle) and PpGAMYB2-Ko (lower). s p, sporophyte. s cale bars indicate 1 mm in e and 0.3 mm in f. (Fig. 6f ). e Th expression level of PpCYP703B2 in PpGAMYB2-KO homologues in non-flowering plants at least partially compensated was decreased to one-third of that in the WT, whereas that in PpGA- for the lack of OsGAMYB in the rice GA signalling pathway, we MYB1-KO was almost half of that in the WT (Fig. 6f ). e Th sup - propose the following evolutional model of the GA signalling path- pression level of PpCYP703B1 was not significantly less than that way (Fig. 7): e Th common ancestor of basal land plants before the of PpCYP703B2 in PpGAMYB1-KO and PpGAMYB2-KO (Fig. 6f ). separation of the mosses had GAMYB-mediated regulation in the However, the level of PpCYP703B1 decreased to a greater degree in proto-tapetal cells (spore sac cells) and spores for spore outer wall PpGAMYB1-KO than in PpGAMYB2-KO, and the opposite was the formation. At this stage, the GAMYB-mediated regulation of spore case for PpCYP703B2 (Fig. 6f ). This corresponds with the co-expres - outer wall formation was independent of the GA signalling path- 13,14 sion pattern of PpGAMYB1 and PpCYP703B1, and PpGAMYB2 and way, as the GA signalling pathway had not yet been established . PpCYP703B2 in sporangium (Supplementary Fig. S5). Consistent with this hypothesis, it is reported that the disruption In the 1-kb promoter region of PpCYP703B2, a paralogue that of multiple DELLA-like genes in P. patens did not ae ff ct the plant’s is predominantly expressed in sporangia (Fig. 4b), there are four life cycle , and that the disruption of PpCPS/KS, which encodes putative GAMYB-binding sequences with a TAAC core sequence ent-kaurene synthase, an enzyme essential for GA synthesis in flow - (filled circles in Supplementary Fig. S6a). Among these DNA frag - ering plants, did not result in abnormal spore development . In ments, the P4 fragment effectively inhibited the interaction between the next step, the common ancestor of basal vascular plants before the truncated PpGAMYB2 containing the DNA-binding domain the separation of the lycopsids acquired GA biosynthesis and GA and the authentic GAMYB-binding sequence of rice α-amylase1A, signalling pathways to regulate the pre-existing GAMYB-mediated RAmy1A (Supplementary Fig. S6b). e Th P4 fragment contains two system, leading to the establishment of the GA signalling pathway putative GAMYB-binding sequences (Supplementary Fig. S6a), and via GAMYB. This conclusion is further supported by the fact that a DNA fragment mutagenized at the lower GAMYB core sequence GA is important for microspore exospore wall development and (mP4-2) failed to inhibit its interaction (Supplementary Fig. S6b,c), SmCYP703C1 expression in Selaginella (Figs 2g and 3). suggesting that PpCYP703B2 is a direct target of PpGAMYB. Taken PpGAMYB2-KO failed to induce antheridia, and instead exhib - together, these data indicate that the direct regulation of CYP703 ited ectopic archegonium formation (Fig. 5a,b). This result sug - by GAMYB has an important role in the development of the spore/ gests that GAMYB is involved in the formation of the gametophytic pollen outer walls in moss and flowering plants. sexual organs in P. patens. GA and its derivatives are known to mod- ulate the formation/development of sexual organs during the repro- Discussion ductive stage in some tree ferns and flowering plants . Although In this study, we demonstrated that, in S. moellendori ffi and P. patens, the downstream factors in this sexual organ developmental process as in flowering plants, GAMYB homologues regulate CYP703 in fern plants have not been identified , these observations sug- expression to promote the development of the outer walls of spores/ gest that, aer ft the establishment of the GA perception system, the microspores, even although the GA signalling and biosynthesis first function of the GA perception system may have been to pro - pathways do not exist in P. patens. Further, since these GAMYB mote sexual organ formation and spore/pollen development at the n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Percentage of gametophores with Percentage of gametophores a antheridium or archegonium (%) with a sporophyte (%) n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 ARTICLE WT WT WT WT Pe OE WT Pe OE IE In IE In OE PpGAMYB1-KO Pe PpGAMYB1-KO IE PpGAMYB-KOs In PpGAMYB1-KO PpGAMYB1-KO Pe OE Pe PpGAMYB2-KO IE In PpGAMYB2-KO IE In ** 1.4 ** PpGAMYB2-KO PpGAMYB2-KO 1.2 1.0 NS 0.8 0.6 0.4 0.2 PpCYP703B1 PpCYP703B2 Figure 6 | Effect of PpGAMYB knockout lines on spore development in P. patens. (a,b) s Em images (a) and their magnification ( b) of spores developed in WT (upper), PpGAMYB1-Ko (middle) and PpGAMYB2-Ko (lower) of P. patens (n = 15). Arrowheads indicate collapsed spores. s cale bars indicate 25 µm in a and 10 µm in b. (c) s pore morphology after the acetolysis incubation. Arrows indicate the debris of spores. s cale bars, 50 µm. (d) The outer wall structure of WT, PpGAMYB1-Ko and PpGAMYB2-Ko spores. Arrowheads indicate the interrupted domains of the outer exospore. s cale bars, 2 µm. IE, inner exospore; In, intine; o E, outer exospore; Pe, perine. (e) Diagrams of spore wall structure in the WT (upper) and PpGAMYB-Ko s (lower). The abnormal wall structures in PpGAMYB-Ko s are marked by dotted lines. (f) PpCYP703Bs expression in WT (blue), PpGAMYB1-Ko (red) and PpGAMYB2-Ko (green) sporangia. Data represent mean values ± s.d. (n = 3). **P < 0.01; *P < 0.05; n -s , not significant; s tudent’s t-test (two-tailed). The ancestor of The ancestor of basal land plants basal vascular plants Table 1 | Characterization of the spores of PpGAMYB knockouts. Spore sac cells Tapetal cells GA perception GA perception Acetolysis experiment (%)† Strain Spore number* system system Resistant‡ Sensitive‡ Wild type 4,995 ± 260 85.37 ± 3.21 14.63 ± 3.21 GAMYB GAMYB PpGAMYB1-Ko 2,821 ± 103** 76.68 ± 2.68** 23.32 ± 2.68** PpGAMYB2-Ko 2,081 ± 316** 39.67 ± 8.97** 60.33 ± 8.97** CYP703 CYP703 Data represent mean values ± s.d. *The total number of spores formed in a sporangium (n > 6). †The spores that collapsed upon incubation in acetolysis solution were judged as being sensi- tive, and the intact ones were considered to be resistant. ‡The percentage of collapsed or intact spores per total number of spores (n > 11). **P < 0.01; Microspore Spore s tudent’s t-test (two-tailed). Figure 7 | Evolutional model of GA-mediated outer wall formation. Before the establishment of GA perception system (left), the common ancestor of basal land plants had GAm YB-mediated regulation for spore outer reproductive stage via GAMYB. In other words, the GA signalling wall formation. In the next evolutional step (right), the common ancestor pathway may have originally arisen to regulate reproduction. of basal vascular plants acquired GA biosynthesis and GA signalling er Th e is a working model that could explain the occurrence of pathways to regulate the pre-existing GAm YB-mediated system, leading GA control of the GAMYB system in the course of plant evolution. to the establishment of the GA signalling pathway via GAm YB. Grey letter Similar to the ancestors of land plants, P. patens, which develops a indicates a loss of indicated molecular system. parasitic sporophyte on the gametophyte, has a gametophyte-domi- nant life cycle, whereas vascular plants, including S. moellendorffii, have a sporophyte-dominant life cycle . e Th evolutional transition tion in a novel manner, the GA perception system and the pre-exist- from a gametophyte- to a sporophyte-dominant life cycle and the ing GAMYB might be used by the early vascular plants, including subsequent adaptation to the land environment have increased the S. moellendorffii . Particularly, in male reproductive development, complexity of the sexual reproduction system: for example, the tran- this GA-dependent control of GAMYB might allow the coordinated sition from spore sac cells to tapetal cells, and the transition from development of various cell types so that it can be synchronized homospory to heterospory. u Th s, to complete the sexual reproduc - with female reproductive development. n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Expression levels ARTICLE nATuRE C ommunICATIons | DoI: 10.1038/ncomms1552 Methods Endogenous GA analysis. Extraction and determination of GA of S. moellendorf- Plant materials and chemical treatment. The rice cultivars Oryza sativa cv i fi microsporangia and young stems were performed using a liquid chromatog- Nipponbare (WT); gamyb-2 (mutant of GAMYB protein) and cyp703a3-1 (mutant raphy–tandem mass chromatography system (AQUITY UPLC System/Quattro 1,7 41 of CYP703A3 protein) were used in this study . The strain of P. patens, WT62, Premier XE, Waters) as described . Data were processed by MassLynx software and the same strain of S. moellendorffii as one of previous report were used as with QuanLynx (version 4.0, Waters). the WT. The plants except P. patens were grown at 28 °C in a greenhouse. P. patens on the sterile peat pellets was cultured at 28 °C under the continuous light for 3 RNA isolation and RT–PCR analysis. Total RNA was isolated from tissue samples using RNeasy plant minikit (Qiagen). The first strand of cDNA was synthesized weeks, and then moved to 15 °C under 8 h light and 16 h dark conditions to induce antheridium/acrhegonium. from 1 µg of total RNA using an Omniscript reverse transcription kit (Qiagen). For the expression analysis using microsporangia of the S. moellendorffii , the Transcripts were quantified by RT–PCR or real-time RT–PCR analyses using one- − 5 twentieth of the resulting cDNA as template. Real-time RT–PCR was performed detached strobili were placed in a solution of 150mM s ucrose containing 10 M GA or ethanol for 12 h at 22 °C, and then microsporangia were collected from the with the LightCycler system (Roche) with the SYBR Green PCR kit (Qiagen). For strobili. For the ultrastructure studies using microspores of S. moellendorffii treated RT–PCR analysis, an adequate volume of each cDNA sample was used for PCR − 4 − 6 − 4 amplification with different numbers of cycles (20–35 cycles) to confirm the linear with mock (1/1,000th EtOH), 10 M GA , 10 M uniconazole, 10 M prohexadi- − 4 − 6 − 4 − 4 one-calcium, 10 M GA and 10 M uniconazole, or 10 M GA and 10 M range of PCR amplification for each gene. The results were confirmed using three 4 4 prohexadione-calcium, samples were sprayed three times weekly with the independent biological replicates. The Sm6PGD, PpUbiquitin or OsActin1 genes respective solution, aer t ft he development of the first strobili. At the end of 1 were used as internal standards for normalizing cDNA concentration variations. e p Th rimer sequences used in this study are listed in Supplementary Table S1. month, a sample of strobili was collected for the SEM observation. Plasmid construction and plant transformation. The sequences of the primers References used in this study are listed in Supplementary Table S1. The PCR fragments were 1. Aya, K. et al. Gibberellin modulates anther development in rice via the sequenced to confirm that no mutations were introduced. transcriptional regulation of GAMYB. Plant Cell 21, 1453–1472 (2009). For the complementation of GAMYB and CYP703 homologues in P. patens and 2. Cheng, H. et al. Gibberellin regulates Arabidopsis floral development via S. moellendorffii , the CDS of each gene was amplified from complementary DNA of suppression of DELLA protein function. Development 131, 1055–1064 (2004). P. patens gametophores or S. moellendorffii strobili using the indicated primer sets. 3. Goto, N. & Pharis, R. P. Role of gibberellin in the development of floral organs e r Th esulting PCR fragments of GAMYB homologues contained a BamHI site at its of the gibberellin-deficient mutant, ga1-1, of Arabidopsis thaliana. Can. J. Bot. 5′ end and a XhoI site at its 3′ end, whereas those of CYP703 homologues contained 77, 944–954 (1999). a NcoI site at its 5′ end and a SmaI or EcoRV site at its 3′ end. These resulting PCR 4. Brooks, J. & Shaw, G. Sporopollenin: a review of its chemistry, palaeochemistry fragments were then cloned into the pCR4 Blunt-TOPO vector (Invitrogen). Aer ft and geochemistry. Grana 17, 91–97 (1978). that, each construct was excised at its enzyme sites, and the resulting fragment was 5. Piffanelli, P., Ross, J. H. & Murphy, D. J. Biogenesis and function of the lipidic ligated into pOsGAMYBnos-pCAMBIA1380 or pOsCYP703A3nos-pCAM- structures of pollen grains. Sex. Plant Reprod. 11, 65–80 (1998). BIA1380 plasmid carrying the hygromycin resistance cassette. These chimeric 6. Ahlers, F., Lambert, J. & Wiermann, R. Acetylation and silylation of piperidine constructs were introduced into Agrobacterium tumefaciens strain EHA105 and solubilized sporopollenin from pollen of Typha angustifolia L. Z. Naturforsch. used to infect rice callus according to previous report . C. 58, 807–811 (2003). To produce the PpGAMYB constructs used to generate P. patens KO lines, the 7. Kaneko, M. et al. Loss-of-function mutations of the rice GAMYB gene impair 5′- and 3′-flanking sequence of each PpGAMYB-coding region was cloned from alpha-amylase expression in aleurone and flower development. Plant Cell 16, genomic DNA of P. patens using the corresponding primers. The resulting 5 ′- and 33–44 (2004). 3′-fragment of each PpGAMYB contained EcoRV and SmaI site at its 5′ and 3′ ends, 8. Murray, F., Kalla, R., Jacobsen, J. & Gubler, F. A role for HvGAMYB in anther and was then introduced into EcoRV and SmaI site of the pTN186 and pTN182 development. Plant J. 33, 481–491 (2003). vectors in the correct orientation, generating PpGAMYB1-pTN186 and PpGA- 9. Hemsley, A. R., Griffiths, P. C., Mathias, R. & Moore, S. E. M. A model for MYB2-pTN182. To obtain GAMYB-GUS fusion transformants, a genomic region the role of surfactants in the assembly of exine sculpture. Grana 42, 38–42 of each PpGAMYB from the middle to the last codon of the CDS, was introduced (2003). in-frame 5′ to the GUS gene of pTN85. The genomic fragment containing the 10. Wellman, C. H., Gensel, P. G. & Taylor, W. A. Spore wall ultrastructure in the 3′-flanking region of each PpGAMYB gene was inserted into the 3′-region of the early lycopsid Leclercqia (Protolepidodendrales) from the Lower Devonian of NPTII expression cassette of this plasmid. To transform these plasmids into North America: evidence for a fundamental division in the lycopsids. Am. J. P. patens, polyethylene glycol-mediated transformation was performed essen- Botany 96, 1849–1860 (2009). tially as described , using 10 µg of linearized plasmid. Stable transformants were 11. Pacini, E., Franchi, G. G. & Hesse, M. The tapetum: its form, function and screened by Southern hybridization using a probe that detected the homologous possible phylogeny in Embryophyta. Plant Syst. Evol. 149, 155–185 (1985). region of the transgene. 12. Morant, M. et al. CYP703 is an ancient cytochromeP450 in land olants catalyzing in-chain hydroxylation of lauric acid to provide building blocks for Microscopy and phenotypic analyses. For GUS staining, tissues were transferred sporopollenin synthesis in pollen. Plant Cell 19, 1473–1487 (2007). to microcentrifuge tubes containing a solution of 50 mM sodium phosphate, pH 13. Hirano, K. et al. The GID1-mediated gibberellin perception mechanism is 7.0, 7% methanol, and 1 mM 5-bromo-4-chloro-3-indolyl-β-d-glucuronide for conserved in the Lycophyte Selaginella moellendorffii but not in the Bryophyte 12 h at 37 °C. Stained tissues were destained with an ethanol series. Physcomitrella patens. Plant Cell 19, 3058–3079 (2007). For in situ hybridization, plant materials were fixed in 4% (w/v) paraformal - 14. Yasumura, Y., Crumpton-Taylor, M., Fuentes, S. & Harberd, N. Step-by-step dehyde and 0.25% (v/v) glutaraldehyde in 0.1M s odium phosphate buffer, pH acquisition of the gibberellin-DELLA growth-regulatory mechanism during 7.4, overnight at 4 °C, dehydrated through a graded ethanol series followed by land-plant evolution. Curr. Biol. 17, 1225–1230 (2007). a t-butanol series, and finally embedded in Paraplast Plus (Sherwood Medical). 15. Ueguchi-Tanaka, M. et al. GIBBERELLIN INSENSITIVE DWARF1 encodes a Microtome sections (8–10 µm thick) were mounted on glass slides treated with soluble receptor for gibberellin. Nature 437, 693–698 (2005). silane. To produce the gene-specific probes, each DNA fragment was generated by 16. Ueguchi-Tanaka, M. et al. Molecular interactions of a soluble gibberellin PCR using complementary DNA and the specific primers. The resulting DNA frag - receptor, GID1, with a rice DELLA protein, SLR1, and gibberellin. Plant Cell ments were introduced between XhoI and EcoRI site of the pBluescript II SK + and 19, 2140–2155 (2007). pBluescript II KS + vectors, and were then transcribed in vitro under T7 promoter 17. Gocal, G. F., Poole, A. T., Gubler, F., Watts, R. J., Blundell, C. & King, R. with T7 RNA polymerase (Invitrogen) using the DIG RNA labeling kit (Roche). W. Long-day up-regulation of a GAMYB gene during Lolium temulentum Hybridization and immunological detection of the hybridized probes were per - inflorescence formation. Plant Physiol. 119, 1271–1278 (1999). formed essentially as described . The primer sequences used in this study are listed 18. Gocal, G. F. et al. GAMYB-like genes, flowering, and gibberellin signaling in in Supplementary Table S1. Arabidopsis. Plant Physiol. 127, 1682–1693 (2001). For observations by TEM, plant samples were vacuum infiltrated three times 19. Tanurdzic, M. & Banks, J. A. Sex-determining mechanisms in land plants. for 10 min with fixative (4% paraformaldehyde and 1% glutaraldehyde in 0.05 M Plant Cell 16, S61–S71 (2004). cacodylate buffer, pH 7.4). Procedures for ultrastructural studies were essentially 40 20. Archard, P., Herr, A., Baulcombe, D. C. & Harberd, N. P. Modulation of the same as described . floral development by a gibberellin-regulated microRNA. Development 131, For observations by SEM, fresh microspores/spores were placed on specimen 3357–3365 (2004). stubs using adhesive carbon tape. The samples were gold coated with Pt and Pd in 21. Toia, R. E. et al. Sporopollenin content of the spore apparatus of Azolla. Am. an ion sputter (E-1030, Hitachi), and imaged using SEM (S-3000N microscope, Fern J. 75, 38–43 (1985). Hitachi). 22. Uehara, K. et al. Ultrastructural study on microspore wall morphogenesis in For the acetolysis experiments, spores were collected in a tube and incubated in Isoetes Japonica (Isoetaceae). Am. J. Botany 78, 1182–1190 (1991). 9:1 (v/v) mixture of acetic anhydride and sulphuric acid at 70°C f or 20 min. Aer ft 23. Rowley, J. R., Morbelli, J. R. & El-Ghazaly, G. Microspore wall structure in the incubation, spores were transferred on a glass slide and observed by the light Selaginella kraussiana (Lycophyta). Taiwania 47, 115–128 (2002). microscopy (Olympus). nATuRE C ommunICATIons | 2:544 | DoI: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. nATuRE C ommunICATIons | DoI: 10.1038/ncomms1552 ARTICLE 24. Millar, A. A. & Gubler, F. The Arabidopsis GAMYB-like genes, MYB33 and 40. Hara-Nishimura, I., Takeuchi, Y. & Nishimura, M. Molecular characterization MYB65, are microRNA-regulated genes that redundantly facilitate anther of a vacuolar processing enzyme related to a putative cysteine proteinase of development. Plant Cell 17, 705–721 (2005). Schistosoma mansoni. Plant Cell 5, 1651–1659 (1993). 25. Kaneko, M. et al. Where do gibberellin biosynthesis and gibberellin signaling 41. Kojima, M. et al. Highly sensitive and high-throughput analysis of plant occur in rice plants? Plant J. 35, 104–115 (2003). hormones using MS-probe modification and liquid chromatography-tandem 26. Hu, J. et al. Potential sites of bioactive gibberellin production during mass spectrometry: an application for hormone profiling in Oryza sativa. Plant reproductive growth in Arabidopsis. Plant Cell 20, 320–336 (2008). Cell Physiol. 50, 1201–1214 (2009). 27. Izumi, K., Kamiya, Y., Sakurai, A., Oshio, H. & Takahashi, N. Studies of sites of action of a new plant growth retardant (E)-1-(4-chlorophenyl)-4, 4- dimethyl- Acknowledgements 2-(1, 2, 4-trizol-1-yl)-penten-3-ol (S-3307) and comparative effects of its We thank Y. Hattori and M. Sato (Nagoya University) for technical assistance. We also stereisomers in a cell free system of Cucurbits maxima. Plant Cell Physiol. 29, thank Y. Nagamura and R. Motoyama (Rice Genome Resource Center of the National 821–827 (1985). Institute of Agribiological Sciences, Japan) for the use of the rice microarray analysis 28. Nakayama, I., Kobayashi, M., Kamiya, Y., Abe, H. & Sakurai, A. Effects of a system and technical support. This study was supported by a Grant-in-Aid from the plant-growth regulat or, prohexadione-calcium (BX-112), on the endogenous Ministry of Education, Culture, Sports, Science and Technology of Japan (18107001 levels of gibberellins in rice. Plant Cell Physiol. 33, 59–62 (1991). and 18075006 to M.M.), a Grant-in-Aid for Scientific Research on Innovative Areas 29. Sakakibara, K., Nishiyama, T., Deguchi, H. & Hasebe, M. Class 1 KNOX genes (23113001 and 23113005 to M.M.) and Global COE Program ‘Advanced Systems- are not involved in shoot development in the moss Physcomitrella patens but do Biology: Designing The Biological Function’, MEXT, Japan. function in sporophyte development. Evol. Dev. 10, 555–566 (2008). 30. Dobritsa, A. A. et al. CYP704B1 is a long-chain fatty acid omega-hydroxylase essential for sporopollenin synthesis in pollen of Arabidopsis. Plant Physiol. Author contributions 151, 574–589 (2009). M.U.-T., M.H. and M.M. conceived the project and designed the experiments together 31. Hesse, M. & Waha, M. A new look at the acetolysis method. Plant Syst. Evol. with K.A. K.A. performed most of the experiments. Y.H. and M.H. generated all 163, 147–152 (1989). transgenic lines of P. patens using homologous recombination. K.M. and H.S. measured 32. Estébanez, B., Yamaguchi, T. & Deguchi, H. Ultrastructure of the spore in four the endogenous GA content of S. moellendorffii (Fig. 2f ). M.M. wrote the paper together Japanese species of Ptychomitrium Furnr. (Musci). Grana 45, 61–70 (2006). with K.A. and M.H. 33. Schuette, S. et al. Novel localization of callose in the spores of Physcomitrella patens and phylogenomics of the callose synthase gene family. Ann. Bot. 103, Additional information 749–756 (2009). Accession Codes: The microarray data have been deposited in the Gene Expression 34. Singer, S. D. & Ashton, N. W. Revelation of ancestral roles of KNOX genes by a Omnibus under accession code GSE32652. functional analysis of Physcomitrella homologues. Plant Cell Rep. 26, 2039–2054 (2007). 35. Gubler, F. et al. Target genes and regulatory domains of the GAMYB Supplementary Information accompanies this paper at http://www.nature.com/ transcriptional activator in cereal aleurone. Plant J. 17, 1–9 (1999). naturecommunications 36. Hayashi, K. et al. Endogenous diterpenes derived from ent-kaurene, a Competing financial interests: The authors declare no competing financial interests. common gibberellin precursor, regulate protonema differentiation of the moss Physcomitrella patens. Plant Physiol. 153, 1085–1097 (2010). Reprints and permission information is available online at http://npg.nature.com/ 37. Banks, J. A. Gametophyte development in ferns. Annu. Rev. Plant Physiol. Plant reprintsandpermissions/ Mol. Biol. 50, 163–186 (1999). How to cite this article: Aya, K. et al. The Gibberellin perception system evolved to 38. Hiei, Y., Ohta, S., Komari, T. & Kumashiro, T. Efficient transformation of rice regulate a pre-existing GAMYB-mediated system during land plant evolution. Nat. (Oryza sativa L.) mediated by Agrobacterium and sequence analysis of the Commun. 2:544 doi: 10.1038/ncomms1552 (2011). boundaries of the T-DNA. Plant J. 6, 271–282 (1994). 39. Kouchi, H. & Hata, S. Isolation and characterization of novel nodulin cDNAs License: This work is licensed under a Creative Commons Attribution-NonCommercial- Share Alike 3.0 Unported License. To view a copy of this license, visit http:// representing genes expressed at early stages of soybean nodule development. Mol. Gen. Genet. 238, 106–119 (1993). creativecommons.org/licenses/by-nc-sa/3.0/ nATuRE C ommunICATIons | 2:544 | DoI: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nature Communications Springer Journals

The Gibberellin perception system evolved to regulate a pre-existing GAMYB-mediated system during land plant evolution

Download PDF
9 pages

Loading next page...
 
/lp/springer-journals/the-gibberellin-perception-system-evolved-to-regulate-a-pre-existing-GhDH94dmn8

References (45)

Publisher
Springer Journals
Copyright
Copyright © 2011 by The Author(s)
Subject
Science, Humanities and Social Sciences, multidisciplinary; Science, Humanities and Social Sciences, multidisciplinary; Science, multidisciplinary
eISSN
2041-1723
DOI
10.1038/ncomms1552
Publisher site
See Article on Publisher Site

Abstract

ARTICLE DOI: 10.1038/ncomms1552 Received 5 Aug 2011 | Accepted 18 oct 2011 | Published 22 nov 2011 The Gibberellin perception system evolved to regulate a pre-existing GAm YB-mediated system during land plant evolution 1 2,3 4 4 1 Koichiro Aya , Yuji Hiwatashi , mikiko Kojima , Hitoshi sakakibara , miyako ueguchi-Tanaka , 2,3,5 1 mitsuyasu Hasebe & makoto matsuoka Gibberellin (GA) controls pollen development in flowering plants via the GAm YB transcription factor. Here we show that GAm YB is conserved in Selaginella moellendorffii (lycophyte) and Physcomitrella patens (moss), although the former contains the GA signalling pathway, the latter does not. In the lycophyte, GA treatment promotes the outer wall development on microspores, whereas treatment with GA biosynthesis inhibitors disturbs its development. Contrary, in the moss, GAMYB homologue knockouts also produce abnormal spores that resemble Selaginella microspores treated with GA biosynthesis inhibitors and pollen grains of rice gamyb mutant. moreover, the knockouts fail to develop male organs, instead ectopically forming female organs. Thus, before the establishment of the GA signalling pathway, basal land plants, including mosses, contained a GAm YB-based system for spore and sexual organ development. subsequently, during the evolution from mosses to basal vascular plants including lycophytes, GA signalling might have merged to regulate this pre-existing GAm YB-based system. 1 2 Bioscience and Biotechnology Center, Nagoya University, Nagoya 464-8601, Japan. National Institute for Basic Biology, Okazaki 444-8585, Japan. 3 4 Department of Basic Biology, School of Life Science, The Graduate University for Advanced Studies, Okazaki 444-8585, Japan. RIKEN Plant Science Center, 1-7-22, Suehiro, Tsurumi, Yokohama 230-0045, Japan. ERATO, Japan Science and Technology Agency, Okazaki 444-8585, Japan. Correspondence and requests for materials should be addressed to K.A. (email: [email protected]). nATuRE C ommunICATIons | 2:544 | DoI: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. ARTICLE n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 n flowering plants, gibberellin (GA) regulates the formation of the abnormalities in spore development and showed the transcriptional outer wall on pollen and is therefore essential for pollen develop- reduction of CYP703 homologues, supporting that the regulation of 1–3 Iment . e Th outermost pollen wall, which is known as the exine CYP703 by GAMYB is conserved among land plant lineages. More- or exospore, consists largely of sporopollenin, an organic polymer over, the KOs failed to develop male organs, instead ectopically that is extremely resistant to degradation and is so durable that it forming female organs. On the basis of these results, we propose 4–6 has been found in microfossils that are over 500 million years old . that GA signalling might have merged to regulate this pre-existing In rice, CYP703A3 is an essential enzyme in sporopollenin biosyn- GAMYB-based system in reproductive development. thesis, and its expression is positively regulated by a transcription factor, GAMYB, which is controlled by the GA signalling path- Results 1,7,8 way . Previous anatomical and in silico studies suggest that, similar In silico screening for GAMYB-like genes. To test our hypothesis, we to flowering plants, present-day bryophytes, including mosses, and first searched for homologues of GAMYB in mosses and lycopods. lycophytes (lycopods) contain sporopollenin-containing outer walls A survey of public genomic phytozome database showed that the 9–11 on the surface of their spores and microspores , and genes encod- model moss P. patens has two GAMYB-like genes, PpGAMYB1 and ing CYP703 homologues have been detected in these genomes . PpGAMYB2, whereas the model lycopod S. moellendorffii contains Curiously, however, one of the basal vascular plants, Selaginella, a single GAMYB-like gene, SmGAMYB (Fig. 1a; Supplementary contains a functional GA perception system that is mediated by a Fig. S1a). A comparative analysis of these GAMYB-like genes with GA receptor, GIBBERELLIN-INSENSITIVE DWARF1 (GID1), GAMYBs in flowering plants revealed that these homologues shared and suppressor proteins (DELLAs), whereas one of the basal land three conserved regions (BOX1, BOX2 and BOX3) in addition to a 13–16 plants, Physcomitrella patens, does not contain this system . es Th e R2R3 DNA-binding domain and a putative microRNA 159 target observations led us to speculate that the ancestral land plants devel- site (Fig. 1a; Supplementary Fig. S1b), which are ubiquitous in 18,20 oped sporopollenin-containing outer walls on their spores, which GAMYBs of flowering plants . protected the spores from dryness and ultraviolet radiation on land, A previous study of the CYP703 family reported that three homo- in association with CYP703-like enzymes, and that GA signalling logues of this protein were present in P. patens (PpCYP703B1–3) pathway then evolved to control the development of the spore outer and one was present in S. moellendorffii (SmCYP703C1). wall via the modulation of CYP703-like expressions during the evolution from mosses to lycopods. Mutant and transgenic studies Molecular characterization of SmGAMYB and PpGAMYBs. To in flowering plants indicate that GAMYB preferentially modulates test the function of these putative GAMYB homologues, we fused other aspects of reproductive developments (for example, inflores - the GAMYB coding sequence (CDS) from each species to the cence formation and floral organ development as well as microspore rice GAMYB (OsGAMYB) promoter, and transformed these con- 1,7,8,17,18 development) . Moreover, considering that some GA-like structs into a rice loss-of-function mutant, gamyb-2. As previously substances (known as antheridiogens) are involved in the forma- reported, the observations by the transmission electron microscopy tion of sexual organs in fern gametophytes , it is further likely that (TEM) revealed that the wild type (WT) plants developed micro- this GA signalling pathway that involves GAMYB regulates a broad spores with a normal exine consisting of three layers in the anthers range of reproductive processes, including (micro)spore develop- (tectum, bacula and foot layer ; Fig. 1b and inset of Fig. 1c), whereas ment, in early vascular plants. gamyb-2 plants transformed with an empty vector developed micro- In this study, we identified the functional GAMYB homologues spores with a defective exine consisting of a single layer, resulting in in Selaginella moellendorffii and P. patens. In S. moellendorffii , GA shrunken and whitened anthers that lacked pollen grains relative to treatment promoted the outer wall development on microspores, yellow anthers of WT (Fig. 1b,c). e Th gamyb-2 plants transformed in association with the induction of CYP703 homologue. On the with OsGAMYB developed microspores with a normal exine struc- other hand, in P. patens with no functional GA signalling pathway, ture that consisted of three layers (Fig. 1c). Anther development the knockouts (KOs) of two GAMYB homologues had similar was partially rescued in transgenic plants carrying SmGAMYB R2R3 OsGAMYB Vector pOsGAMYB:: ATG Stop OsGAMYB An R2R3 SmGAMYB ATG Stop R2R3 WT gamyb-2 gamyb-2 PpGAMYB1/2 ATG Stop pOsGAMYB:: pOsGAMYB:: pOsGAMYB:: SmGAMYB PpGAMYB1 PpGAMYB2 WT Vector OsGAMYB Te Ba Ex Fl gamyb-2 gamyb-2 gamyb-2 gamyb-2 gamyb-2 SmGAMYB PpGAMYB1 PpGAMYB2 gamyb-2 gamyb-2 gamyb-2 Figure 1 | Conserved molecular function of SmGAMYB and PpGAMYBs. (a) The exon/intron structure of OsGAMYB and GAMYB homologues. The coding regions and introns are symbolized by boxes and lines, respectively. Black box, R2R3 Dn A-binding domain; grey boxes, Bo X1–3 domains; black line above gene structure, putative miR159 target site. (b,c), Flower (b) and the microspore outer wall (exine) (c) of wild-type (WT) rice plants and of transgenic plants transformed with empty vector, OsGAMYB and GAMYB homologues in the rice gamyb-2 background (n > 11). Inset indicates the magnification of WT exine layer. s cale bars indicate 2 mm in b and 5 µm in c. An, anther; Ba, bacula; Ex, exine; Fl, foot layer; Te, tectum. n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. BOX1 BOX1 miRNA BOX1 miRNA BOX2 miRNA BOX2 BOX2 BOX3 BOX3 BOX3 n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 ARTICLE Young 0.7 25 Premeiosis microspore Premeiosis 0.6 0.5 MMC YM T 15 0.4 0.3 0.2 0.1 as as Sense ND ND ND ND ND Young Old Apices Root Strobili GA GA GA GA GA GA 24 9 4 19 20 1 stem stem MMC YM Microsporangium (kbp) (fold) 0.6 – GA + GA 4 4 0.2 0.5 SmGA3ox 0.5 0.4 as as Sense 0.3 0.2 SmGAMYB 1.7 0.2 0.2 0.1 SmCYP703C1 4.3 YM Young Old Apices Root Strobili 1.1 MMC Sm6PDG stem stem 0.2 as as Sense Figure 2 | The spatioexpression pattern of SmGAMYB and SmCYP703C1 in microspore development of S. moellendorffii . (a,b) Quantitative RT–PCR analysis of SmGAMYB (a) and SmCYP703C1 (b) in different tissues. Data are shown as the mean values ± s.d. (n = 3). (c–e) In situ hybridization of SmGA3ox (c), SmGAMYB (d) and SmCYP703C1 (e) on microsporangium at the premeiosis stage (left and right) and young microspore stage (middle). s cale bars, 50 µm. as, anti-sense; mm C, microspore mother cell; T, tapetal cell; Ym , young microspore. (f) Endogenous GA content of S. moellendorffii microsporangia (grey) and young stems (black). Data are shown as the mean values ± s.d. (n = 3). FW, fresh weight; n D, no expected product was detected. (g) SmGA3ox, SmGAMYB and SmCYP703C1 expression in response to GA treatment in microsporangium. n umber on the right side of each panel indicates the expression ratio in GA treatment versus in m ock control, estimated by Image J software. (Fig. 1b). e Th anthers of these plants were slightly whitened ( Fig. 1b), S3c). Together, these results indicate that SmGAMYB can almost and the microspores exhibited a normal exine consisting of three completely substitute, and PpGAMYB proteins can partially sub- layers (Fig. 1c). In PpGAMYB1 and PpGAMYB2 transgenic lines, stitute, for OsGAMYB in the GA signalling pathway of rice anther anther development was also partially rescued (Fig. 1b), which development. exhibited microspores with a normal exine (Fig. 1c). We also exam- ined whether the CYP703 homologues in P. patens and S. moel- Role of GA in microspore development of S. moellendorffii . Dur- lendorffii had sporopollenin biosynthetic activity. For this analysis, ing the divergence of basal vascular plants including lycopods from we used SmCYP703C1 (the sole homologue detected in S. moellen- the moss lineage, a GA receptor, GID1 and DELLA repressors were dorffii ) and PpCYP703B2 (the dominant gene of three PpCYP703B established as key components for GA perception system in asso- homologues; see below). A defect in exine structure in pollen of the ciation with a functional GA biosynthetic pathway, indicating the 13,14 rice loss-of-function cyp703a3 mutant was fully rescued by intro- occurrence of the entire GA signalling pathway . However, the duction of the CDS for SmCYP703C1 under the control of the rice biological function of GA in basal vascular plants has been unclear, CYP703A3 promoter, and at least partially rescued by introduc- as GA treatment does not cause any developmental defects at tion of PpCYP703B2 with the same rice promoter (Supplementary the vegetative stage, such as increased stem or leaf elongation, in Fig. S2). Taken together, our results show that the biological function Selaginella . er Th efore, we hypothesized that, in such basal vascular of GAMYB and CYP703 homologues for the outer wall develop- plants, GA may be exclusively involved in the outer wall develop- ment is conserved among the land plant lineage. However, the level ment of microspores via GAMYB. Quantitative RT–PCR analysis of activity varied among each family member, at least in rice cells. using various tissues of S. moellendorffii revealed SmGAMYB and To further confirm the function of the GAMYB homologues in SmCYP703C1 were preferentially expressed in the reproductive the GA signalling pathway in rice cells, we compared the expres- organs, strobili (Fig. 2a,b). e Th microsporangium in the strobili of a sion pattern in the anthers of OsGAMYB transgenic plants with heterosporous lycophyte, S. moellendorffii, is an orthologous organ those of SmGAMYB or PpGAMYB2 transgenic plants, using two- of the anther of flowering plants. Microspores within the microspo - color microarrays hybridized with RNA isolated from rice gamyb-2 rangium correspond to pollen grains of flowering plants. Previous anthers as a common reference (Supplementary Methods, Data are observations suggested that GA signalling mediated by GAMYB is 1,7,8 available at Gene Expression Omnibus, accession code: GSE32652.). important for anther and pollen development in flowering plants . Most OsGAMYB-regulated genes that were expressed in response er Th efore, we next performed an in situ hybridization experiment to OsGAMYB expression (x axis) were also, at least partially, regu- to examine the spatiotemporal expression of SmGAMYB and lated at the transcriptional level by the SmGAMYB and PpGAMYB2 SmCYP703C1 during microsporangium development (Fig. 2c–e). expression (y axis; Supplementary Fig. S3a,b and Supplementary During the transition from premeiosis to young microspore stage, Data 1). e Th transcriptional recovery by SmGAMYB expression the expression of these genes was mainly limited to the tapetal cells was more tightly correlated to OsGAMYB-mediated modulation and microspores in a pattern similar to that of a GA biosynthesis than was the transcriptional recovery by PpGAMYB2 expression gene, GA3 oxidase (SmGA3ox; Fig. 2c–e), where the sporopol- 9–11,21–23 (correlation coefficient; 0.84 in Supplementary Fig. 3a and 0.71 in lenin material is produced for the outer wall development . Supplementary Fig. 3b). e Th difference in correlation coefficient Considering that GAMYBs, CYP703As and GA biosynthesis genes agrees with the rescued phenotype of transgenic rice plants (Fig. in flowering plants are also co-expressed in the tapetal cells and 1,8,12,24–26 1b,c). Consistently, the expression of an OsGAMYB target gene, microspores , these observations indicate that SmGAMYB CYP703A3, which was hardly detected in the gamyb-2 background, activity in these tissues of Selaginella is required to mediate GA was complemented by the introduction of SmGAMYB to a similar signalling during the outer wall formation of microspores. or higher degree than it was by OsGAMYB, whereas PpGAMYBs To further verify the importance of GA in spore wall formation, partially restored CYP703A3 expression, with PpGAMYB2 hav- we analysed the endogenous GA level in the microsporangium. We ing a stronger effect than did PpGAMYB1 (Supplementary Fig. detected a 13 non-hydroxylated bioactive GA, GA, in microsporangia, n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Expression levels Expression levels Content (pmol per gFW) ARTICLE n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 Mock GA Uni ** ** ** 5.0 ** ** 4.5 4.0 3.5 3.0 2.5 Uni+GA Pro Pro+GA 4 4 2.0 1.5 1.0 0.5 Mock GA Uni Pro Uni+GA Pro+GA 4 4 4 Figure 3 | Effects of GA or GA biosynthesis inhibitors on the exospore projection of the microspore surface of S. moellendorffii . (a) s Em images of the exospore of microspores developed by the treatment of GA or GA biosynthesis inhibitors in S. moellendorffii . s cale bars, 50 µm. Pro, prohexadione- calcium; u ni, uniconazole. (b) The height of the exospore projection on the microspore following treatment with GA or GA biosynthesis inhibitors. m easurements were made on s Em micrographs using Image J software, and data represent mean values ± s.d. (m ock control: n = 230; GA : n = 219; uniconazole (u ni): n = 321; prohexadione-calcium (Pro): n = 89; GA and uniconazole (u ni + GA ): n = 178; GA and prohexadione-calcium (Pro + GA ): 4 4 4 4 n = 91). **P < 0.01; Tukey’s test. and, to a lesser extent, in young stems (Fig. 2f ). Because bioactive rhizoids from a fraction of the protonema (Fig. 4c and Supplemen- GA negatively regulates the expression of some GA biosynthetic tary Fig. S4a,b). Subsequently, at the shoot apex of the gametophore, genes, including GA3ox, and positively regulates the expression of the archegonium and antheridium (Fig. 4d,e), which contain an egg 1,8,13 GAMYB and CYP703A in flowering plants , we decided to exam- and sperm cells, respectively, develop, and an egg cell is fertilized by a ine the expression levels of these homologues in the presence and sperm cell . Finally, the resulting zygote develops a sporophyte that absence of GA . Application of GA to the microsporangium sup- contains a spore-producing tissue, the sporangium (Fig. 4f ). Quan- 4 4 pressed the expression of SmGA3ox, whereas it induced SmGAMYB titative RT–PCR analysis revealed that transcripts for PpGAMYBs, and SmCYP703C1 (Fig. 2g). As is the case in the pollen grains of PpCYP703B1 and PpCYP703B2 were present in various tissues of flowering plants, these results demonstrate that the GA response is gametophore, but not in the protonema, whereas PpCYP703B3 present in the microspores of S. moellendorffii . transcript was not present in any tested tissues (Fig. 4a,b). In en, Th we examined the effect of GA or a GA biosynthesis sporangium of the sporophyte, PpGAMYB2 and PpCYP703B2 inhibitor, uniconazole, which inhibits a step of the GA biosynthesis were expressed more strongly than the other transcripts analysed pathway that is catalysed by ent-kaurene oxidase , on the exospore (Fig. 4a,b), indicating that they might have a dominant role in spo- development of S. moellendorffii microspores using scanning elec- rangium development. e Th spatiotemporal expression pattern of tron microscopy (SEM). e Th SEM observations revealed that the PpGAMYB1 and PpGAMYB2 was further examined by introduc- outer exospore of microspores in mock-treated plants was formed ing the β-glucuronidase (GUS) reporter gene just upstream of the as numerous sporopolleninous projections (Fig. 3a), the material of stop codon of each gene (Fig. 4c–f ). During the young gametophore 11,23 which originates exclusively from the tapetum and microspores . stage, GUS activity in the PpGAMYB1-GUS and PpGAMYB2-GUS e Th heights of the outer exospore projections of GA -treated micro- plants was mainly detected in the shoot apices and partially in spores were significantly greater than those of mock-treated micro - the rhizoid (Fig. 4c). During the reproductive stage, the GUS sig- spores, whereas the microspores treated with uniconazole produced nal was detected in the developing archegonium and antheridium less and shorter outer exospore projections than did the mock- (Fig. 4d,e). Aer ft the start of sporogenesis, the GUS signal in these treated microspores (Fig. 3a,b). Such defects in outer exospore two lines was detected in the sporangium (Fig. 4f ). In situ hybridiza- projections were not detected in the microspores of plants that were tion revealed that PpGAMYB1 and PpCYP703B1 were co-expressed simultaneously treated with GA and uniconazole, and the number more strongly in spores than in spore sac cells (Supplementary and height of the outer exospore projections of the microspores of Fig. S5a,b; insets). In contrast, the transcripts of PpGAMYB2 and these plants were similar to those of GA -treated plants (Fig. 3a,b). PpCYP703B2, the predominantly expressed genes, co-localized in Moreover, a similar phenotype was induced upon treatment with spore sac cells of the sporangium (Supplementary Fig. S5c,d; insets), another GA biosynthesis inhibitor, prohexadione-calcium, which which correspond to tapetal cells in flowering plants . inhibits another step of the GA biosynthesis pathway catalysed by To further investigate the function of PpGAMYBs, we generated GA3ox , and was rescued by co-treatment with GA (Fig. 3a,b), KO lines of PpGAMYB1 and PpGAMYB2 (PpGAMYB1-KO and strongly suggesting that these phenomena depend on a GA defi - PpGAMYB2-KO) using homologous recombination. er Th e were no ciency. es Th e results show that, similar to flowering plants, basal morphological and growth defects in the protonemata and game- vascular plants require the GA signalling pathway to promote the tophores in each single KO line (Supplementary Fig. S4a–c). On development of the microspore outer exospore wall. the other hand, aer ft 4 or 5 weeks of antheridium and archegon - ium induction, antheridium formation was strongly disturbed in Role of PpGAMYBs in spore development of P. patens. We first PpGAMYB2-KO, whereas archegonium formation was promoted in examined the expression pattern of PpGAMYBs to estimate the PpGAMYB-KOs (Fig. 5a). Most PpGAMYB2-KO plants displayed biological function of PpGAMYBs in P. patens, which has no func- ectopic archegonium formation at the site of antheridium forma- 13,14 tional GA perception system and GA biosynthetic pathway (Fig. tion, and a similar, but milder, trend was seen in PpGAMYB1-KO 4a,b). P. patens develops protonemata colony from spores, and then (Fig. 5b). Moreover, some abnormal antheridia that were light brown differentiates gametophores with haploid leafy shoots and root-like were observed in these KOs (asterisks in Fig. 5c). Consistent with these n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Height of microspore exospore projection (µm) n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 ARTICLE 2.0 3.5 3.0 1.6 2.5 1.2 2.0 1.5 0.8 1.0 0.4 0.5 Protonema Shoot Rhizoid Sporangium Protonema Shoot Rhizoid Sporangium Gametophore Gametophore PpGAMYB1 PpGAMYB2 PpGAMYB1-GUS PpGAMYB2-GUS -GUS -GUS PpGAMYB1-GUS PpGAMYB2-GUS PpGAMYB1-GUS PpGAMYB2-GUS Figure 4 | Expression pattern of PpGAMYBs and PpCYP703Bs in various tissues of P. patens. (a) Quantitative RT–PCR analysis of PpGAMYB1 (grey) and PpGAMYB2 (black) in different tissues of P. patens. Data represent mean values ± s.d. (n = 3). (b) Quantitative RT–PCR analysis of PpCYP703B1 (grey) and PpCYP703B2 (black) in different tissues of P. patens. Data represent mean values ± s.d. (n = 3). (c–f) Gus expression pattern in the leafy gametophore (c), archegonium (d), antheridia (e) and sporangium (f) of PpGAm YB1-Gus (left) and PpGAm YB2-Gus (right) reporter lines. Arrows mark the shoot apex. s cale bars indicate 1 mm in c, 100 µm in d,e, and 200 µm in f. abnormal antheridium phenotypes, the number of sporophytes was Table 1). Considering that the effect of acetolysis is dominantly reduced (Fig. 5d,e), while developed sporophytes appeared to be reflected by the exospore content, the result suggests that dynamic normal in PpGAMYB-KOs (Fig. 5f ). alternation of exospore structure occurs in these KOs. We also Within the sporophyte, however, there were differences in the observed the ultrastructure of spore walls developed in PpGAMYB- surfaces of the spores in each KO (Fig. 6a,b), and the number of KOs by the TEM and revealed structural differences in PpGAMYB- 32,33 spores in each KO was less than in the WT (Table 1). e Th SEM KOs. As described previously , the spore wall in the WT consists observations of mature spores revealed that WT spores produced of four layers, the intine, inner exospore, outer exospore and perine well-developed spiny projections on their surface, whereas develop- (Fig. 6d,e). Among these layers, the perine layer, which contributes to ment of such projections occurred incompletely in PpGAMYB-KOs, the formation of projections of spores, was severely defective in the resulting in a smoother surface (Fig. 6a,b) and a higher frequency of KO spores, especially in PpGAMYB2-KO (Fig. 6d,e). Moreover, the collapsed spores (arrowheads in Fig. 6a). We noticed that the PpGA- continuous outer exospore structure was sporadically interrupted in MYB2-KO had a more severe phenotype than the PpGAMYB1-KO, PpGAMYB1-KO (arrowheads in Fig. 6d), and oen ft in PpGAMYB2- which is consistent with PpGAMYB2 being expressed at a higher KO, whereas there were no obvious defects in the inner exospore level than PpGAMYB1 in WT sporangium (Fig. 4a). e Th collapsed and intine layers of PpGAMYB-KOs (Fig. 6d,e). es Th e observations spores observed in PpGAMYB-KOs suggest that the structural and are in good agreement with previous reports that the sporopollenin physical protection against the vacuum condition of SEM analysis material for perine and exospore is mainly supplied from the spores 11,34 might be reduced in PpGAMYB-KOs spores, compared with WT and spore sac cells , where PpGAMYBs and PpCYP703Bs were spores. We next treated spores with a solution of acetic anhydride expressed (Supplementary Fig. S5). and sulphuric acid, which generally disrupts all spore/pollen content Consistent with the abnormal perine and exospore layers of except for the exospore/exine by hydrolysis and esterification (ace - PpGAMYB-KOs, the expression of PpCYP703B1 and PpCYP703B2 30,31 tolysis) . e Th spores in PpGAMYB-KOs, especially in the PpGA- was lower in the sporangia of the KOs than in those of the WT MYB2-KO, were easily damaged by acetolysis treatment (percent- (Fig. 6f ). e Th suppression level of PpCYP703B1 was significantly age of acetolysis-sensitive spores in PpGAMYB1-KO, 23.32 ± 2.68%; greater in PpGAMYB1-KO than in PpGAMYB2-KO, whereas that PpGAMYB2-KO, 60.33 ± 8.97%; and WT, 14.63 ± 3.21%; Fig. 6c and of PpCYP703B2 was greater in PpGAMYB2-KO than in PpGAMYB1-KO n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Expression levels Rhizoid Shoot Expression levels ARTICLE n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 WT An WT WT WT Ar Sp 60 An PpGAMYB1-KO PpGAMYB1-KO PpGAMYB1-KO PpGAMYB1-KO 0 An Ar An Ar An Ar * * Sp Ar 4 Week 5 Week An ** ** PpGAMYB2-KO PpGAMYB2-KO PpGAMYB2-KO PpGAMYB2-KO An Ar Ar 40 * Ar ** Sp WT PpGAMYB1 PpGAMYB2 -KO -KO Figure 5 | Effect of PpGAMYB knockout on reproductive development in P. patens. (a) Quantification of antheridium and archegonium formation in the WT (blue), PpGAMYB1-Ko (red) and PpGAMYB2-Ko (green) knockout lines. The percentage of gametophores containing antheridia or archegonia relative to the total number of gametophores after 4 or 5 weeks in antheridium/archegonium induction conditions is given (n > 37). An, antheridium; Ar, archegonium. (b,c) m orphology of antheridia and archegonia in WT (upper), PpGAMYB1-Ko (middle) and PpGAMYB2-Ko (lower) P. patens. Ar*, ectopic archegonium. s cale bars indicate 250 µm in b and 100 µm in c. Asterisks mark the abnormal antheridium. (d) Quantification of sporangium formation in the WT and PpGAMYB knockout lines. The percentage of gametophores containing a sporangium relative to the total number of gametophores after 3 months under antheridium/acrhegonium induction conditions is given (n = 4). **P < 0.01; Tukey’s test. (e,f) m orphology of mature gametophores (e) and a sporophyte (f) of the WT (upper), PpGAMYB1-Ko (middle) and PpGAMYB2-Ko (lower). s p, sporophyte. s cale bars indicate 1 mm in e and 0.3 mm in f. (Fig. 6f ). e Th expression level of PpCYP703B2 in PpGAMYB2-KO homologues in non-flowering plants at least partially compensated was decreased to one-third of that in the WT, whereas that in PpGA- for the lack of OsGAMYB in the rice GA signalling pathway, we MYB1-KO was almost half of that in the WT (Fig. 6f ). e Th sup - propose the following evolutional model of the GA signalling path- pression level of PpCYP703B1 was not significantly less than that way (Fig. 7): e Th common ancestor of basal land plants before the of PpCYP703B2 in PpGAMYB1-KO and PpGAMYB2-KO (Fig. 6f ). separation of the mosses had GAMYB-mediated regulation in the However, the level of PpCYP703B1 decreased to a greater degree in proto-tapetal cells (spore sac cells) and spores for spore outer wall PpGAMYB1-KO than in PpGAMYB2-KO, and the opposite was the formation. At this stage, the GAMYB-mediated regulation of spore case for PpCYP703B2 (Fig. 6f ). This corresponds with the co-expres - outer wall formation was independent of the GA signalling path- 13,14 sion pattern of PpGAMYB1 and PpCYP703B1, and PpGAMYB2 and way, as the GA signalling pathway had not yet been established . PpCYP703B2 in sporangium (Supplementary Fig. S5). Consistent with this hypothesis, it is reported that the disruption In the 1-kb promoter region of PpCYP703B2, a paralogue that of multiple DELLA-like genes in P. patens did not ae ff ct the plant’s is predominantly expressed in sporangia (Fig. 4b), there are four life cycle , and that the disruption of PpCPS/KS, which encodes putative GAMYB-binding sequences with a TAAC core sequence ent-kaurene synthase, an enzyme essential for GA synthesis in flow - (filled circles in Supplementary Fig. S6a). Among these DNA frag - ering plants, did not result in abnormal spore development . In ments, the P4 fragment effectively inhibited the interaction between the next step, the common ancestor of basal vascular plants before the truncated PpGAMYB2 containing the DNA-binding domain the separation of the lycopsids acquired GA biosynthesis and GA and the authentic GAMYB-binding sequence of rice α-amylase1A, signalling pathways to regulate the pre-existing GAMYB-mediated RAmy1A (Supplementary Fig. S6b). e Th P4 fragment contains two system, leading to the establishment of the GA signalling pathway putative GAMYB-binding sequences (Supplementary Fig. S6a), and via GAMYB. This conclusion is further supported by the fact that a DNA fragment mutagenized at the lower GAMYB core sequence GA is important for microspore exospore wall development and (mP4-2) failed to inhibit its interaction (Supplementary Fig. S6b,c), SmCYP703C1 expression in Selaginella (Figs 2g and 3). suggesting that PpCYP703B2 is a direct target of PpGAMYB. Taken PpGAMYB2-KO failed to induce antheridia, and instead exhib - together, these data indicate that the direct regulation of CYP703 ited ectopic archegonium formation (Fig. 5a,b). This result sug - by GAMYB has an important role in the development of the spore/ gests that GAMYB is involved in the formation of the gametophytic pollen outer walls in moss and flowering plants. sexual organs in P. patens. GA and its derivatives are known to mod- ulate the formation/development of sexual organs during the repro- Discussion ductive stage in some tree ferns and flowering plants . Although In this study, we demonstrated that, in S. moellendori ffi and P. patens, the downstream factors in this sexual organ developmental process as in flowering plants, GAMYB homologues regulate CYP703 in fern plants have not been identified , these observations sug- expression to promote the development of the outer walls of spores/ gest that, aer ft the establishment of the GA perception system, the microspores, even although the GA signalling and biosynthesis first function of the GA perception system may have been to pro - pathways do not exist in P. patens. Further, since these GAMYB mote sexual organ formation and spore/pollen development at the n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Percentage of gametophores with Percentage of gametophores a antheridium or archegonium (%) with a sporophyte (%) n ATu RE Commun ICATIons | Do I: 10.1038/ncomms1552 ARTICLE WT WT WT WT Pe OE WT Pe OE IE In IE In OE PpGAMYB1-KO Pe PpGAMYB1-KO IE PpGAMYB-KOs In PpGAMYB1-KO PpGAMYB1-KO Pe OE Pe PpGAMYB2-KO IE In PpGAMYB2-KO IE In ** 1.4 ** PpGAMYB2-KO PpGAMYB2-KO 1.2 1.0 NS 0.8 0.6 0.4 0.2 PpCYP703B1 PpCYP703B2 Figure 6 | Effect of PpGAMYB knockout lines on spore development in P. patens. (a,b) s Em images (a) and their magnification ( b) of spores developed in WT (upper), PpGAMYB1-Ko (middle) and PpGAMYB2-Ko (lower) of P. patens (n = 15). Arrowheads indicate collapsed spores. s cale bars indicate 25 µm in a and 10 µm in b. (c) s pore morphology after the acetolysis incubation. Arrows indicate the debris of spores. s cale bars, 50 µm. (d) The outer wall structure of WT, PpGAMYB1-Ko and PpGAMYB2-Ko spores. Arrowheads indicate the interrupted domains of the outer exospore. s cale bars, 2 µm. IE, inner exospore; In, intine; o E, outer exospore; Pe, perine. (e) Diagrams of spore wall structure in the WT (upper) and PpGAMYB-Ko s (lower). The abnormal wall structures in PpGAMYB-Ko s are marked by dotted lines. (f) PpCYP703Bs expression in WT (blue), PpGAMYB1-Ko (red) and PpGAMYB2-Ko (green) sporangia. Data represent mean values ± s.d. (n = 3). **P < 0.01; *P < 0.05; n -s , not significant; s tudent’s t-test (two-tailed). The ancestor of The ancestor of basal land plants basal vascular plants Table 1 | Characterization of the spores of PpGAMYB knockouts. Spore sac cells Tapetal cells GA perception GA perception Acetolysis experiment (%)† Strain Spore number* system system Resistant‡ Sensitive‡ Wild type 4,995 ± 260 85.37 ± 3.21 14.63 ± 3.21 GAMYB GAMYB PpGAMYB1-Ko 2,821 ± 103** 76.68 ± 2.68** 23.32 ± 2.68** PpGAMYB2-Ko 2,081 ± 316** 39.67 ± 8.97** 60.33 ± 8.97** CYP703 CYP703 Data represent mean values ± s.d. *The total number of spores formed in a sporangium (n > 6). †The spores that collapsed upon incubation in acetolysis solution were judged as being sensi- tive, and the intact ones were considered to be resistant. ‡The percentage of collapsed or intact spores per total number of spores (n > 11). **P < 0.01; Microspore Spore s tudent’s t-test (two-tailed). Figure 7 | Evolutional model of GA-mediated outer wall formation. Before the establishment of GA perception system (left), the common ancestor of basal land plants had GAm YB-mediated regulation for spore outer reproductive stage via GAMYB. In other words, the GA signalling wall formation. In the next evolutional step (right), the common ancestor pathway may have originally arisen to regulate reproduction. of basal vascular plants acquired GA biosynthesis and GA signalling er Th e is a working model that could explain the occurrence of pathways to regulate the pre-existing GAm YB-mediated system, leading GA control of the GAMYB system in the course of plant evolution. to the establishment of the GA signalling pathway via GAm YB. Grey letter Similar to the ancestors of land plants, P. patens, which develops a indicates a loss of indicated molecular system. parasitic sporophyte on the gametophyte, has a gametophyte-domi- nant life cycle, whereas vascular plants, including S. moellendorffii, have a sporophyte-dominant life cycle . e Th evolutional transition tion in a novel manner, the GA perception system and the pre-exist- from a gametophyte- to a sporophyte-dominant life cycle and the ing GAMYB might be used by the early vascular plants, including subsequent adaptation to the land environment have increased the S. moellendorffii . Particularly, in male reproductive development, complexity of the sexual reproduction system: for example, the tran- this GA-dependent control of GAMYB might allow the coordinated sition from spore sac cells to tapetal cells, and the transition from development of various cell types so that it can be synchronized homospory to heterospory. u Th s, to complete the sexual reproduc - with female reproductive development. n ATu RE Commun ICATIons | 2:544 | Do I: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. Expression levels ARTICLE nATuRE C ommunICATIons | DoI: 10.1038/ncomms1552 Methods Endogenous GA analysis. Extraction and determination of GA of S. moellendorf- Plant materials and chemical treatment. The rice cultivars Oryza sativa cv i fi microsporangia and young stems were performed using a liquid chromatog- Nipponbare (WT); gamyb-2 (mutant of GAMYB protein) and cyp703a3-1 (mutant raphy–tandem mass chromatography system (AQUITY UPLC System/Quattro 1,7 41 of CYP703A3 protein) were used in this study . The strain of P. patens, WT62, Premier XE, Waters) as described . Data were processed by MassLynx software and the same strain of S. moellendorffii as one of previous report were used as with QuanLynx (version 4.0, Waters). the WT. The plants except P. patens were grown at 28 °C in a greenhouse. P. patens on the sterile peat pellets was cultured at 28 °C under the continuous light for 3 RNA isolation and RT–PCR analysis. Total RNA was isolated from tissue samples using RNeasy plant minikit (Qiagen). The first strand of cDNA was synthesized weeks, and then moved to 15 °C under 8 h light and 16 h dark conditions to induce antheridium/acrhegonium. from 1 µg of total RNA using an Omniscript reverse transcription kit (Qiagen). For the expression analysis using microsporangia of the S. moellendorffii , the Transcripts were quantified by RT–PCR or real-time RT–PCR analyses using one- − 5 twentieth of the resulting cDNA as template. Real-time RT–PCR was performed detached strobili were placed in a solution of 150mM s ucrose containing 10 M GA or ethanol for 12 h at 22 °C, and then microsporangia were collected from the with the LightCycler system (Roche) with the SYBR Green PCR kit (Qiagen). For strobili. For the ultrastructure studies using microspores of S. moellendorffii treated RT–PCR analysis, an adequate volume of each cDNA sample was used for PCR − 4 − 6 − 4 amplification with different numbers of cycles (20–35 cycles) to confirm the linear with mock (1/1,000th EtOH), 10 M GA , 10 M uniconazole, 10 M prohexadi- − 4 − 6 − 4 − 4 one-calcium, 10 M GA and 10 M uniconazole, or 10 M GA and 10 M range of PCR amplification for each gene. The results were confirmed using three 4 4 prohexadione-calcium, samples were sprayed three times weekly with the independent biological replicates. The Sm6PGD, PpUbiquitin or OsActin1 genes respective solution, aer t ft he development of the first strobili. At the end of 1 were used as internal standards for normalizing cDNA concentration variations. e p Th rimer sequences used in this study are listed in Supplementary Table S1. month, a sample of strobili was collected for the SEM observation. Plasmid construction and plant transformation. The sequences of the primers References used in this study are listed in Supplementary Table S1. The PCR fragments were 1. Aya, K. et al. Gibberellin modulates anther development in rice via the sequenced to confirm that no mutations were introduced. transcriptional regulation of GAMYB. Plant Cell 21, 1453–1472 (2009). For the complementation of GAMYB and CYP703 homologues in P. patens and 2. Cheng, H. et al. Gibberellin regulates Arabidopsis floral development via S. moellendorffii , the CDS of each gene was amplified from complementary DNA of suppression of DELLA protein function. Development 131, 1055–1064 (2004). P. patens gametophores or S. moellendorffii strobili using the indicated primer sets. 3. Goto, N. & Pharis, R. P. Role of gibberellin in the development of floral organs e r Th esulting PCR fragments of GAMYB homologues contained a BamHI site at its of the gibberellin-deficient mutant, ga1-1, of Arabidopsis thaliana. Can. J. Bot. 5′ end and a XhoI site at its 3′ end, whereas those of CYP703 homologues contained 77, 944–954 (1999). a NcoI site at its 5′ end and a SmaI or EcoRV site at its 3′ end. These resulting PCR 4. Brooks, J. & Shaw, G. Sporopollenin: a review of its chemistry, palaeochemistry fragments were then cloned into the pCR4 Blunt-TOPO vector (Invitrogen). Aer ft and geochemistry. Grana 17, 91–97 (1978). that, each construct was excised at its enzyme sites, and the resulting fragment was 5. Piffanelli, P., Ross, J. H. & Murphy, D. J. Biogenesis and function of the lipidic ligated into pOsGAMYBnos-pCAMBIA1380 or pOsCYP703A3nos-pCAM- structures of pollen grains. Sex. Plant Reprod. 11, 65–80 (1998). BIA1380 plasmid carrying the hygromycin resistance cassette. These chimeric 6. Ahlers, F., Lambert, J. & Wiermann, R. Acetylation and silylation of piperidine constructs were introduced into Agrobacterium tumefaciens strain EHA105 and solubilized sporopollenin from pollen of Typha angustifolia L. Z. Naturforsch. used to infect rice callus according to previous report . C. 58, 807–811 (2003). To produce the PpGAMYB constructs used to generate P. patens KO lines, the 7. Kaneko, M. et al. Loss-of-function mutations of the rice GAMYB gene impair 5′- and 3′-flanking sequence of each PpGAMYB-coding region was cloned from alpha-amylase expression in aleurone and flower development. Plant Cell 16, genomic DNA of P. patens using the corresponding primers. The resulting 5 ′- and 33–44 (2004). 3′-fragment of each PpGAMYB contained EcoRV and SmaI site at its 5′ and 3′ ends, 8. Murray, F., Kalla, R., Jacobsen, J. & Gubler, F. A role for HvGAMYB in anther and was then introduced into EcoRV and SmaI site of the pTN186 and pTN182 development. Plant J. 33, 481–491 (2003). vectors in the correct orientation, generating PpGAMYB1-pTN186 and PpGA- 9. Hemsley, A. R., Griffiths, P. C., Mathias, R. & Moore, S. E. M. A model for MYB2-pTN182. To obtain GAMYB-GUS fusion transformants, a genomic region the role of surfactants in the assembly of exine sculpture. Grana 42, 38–42 of each PpGAMYB from the middle to the last codon of the CDS, was introduced (2003). in-frame 5′ to the GUS gene of pTN85. The genomic fragment containing the 10. Wellman, C. H., Gensel, P. G. & Taylor, W. A. Spore wall ultrastructure in the 3′-flanking region of each PpGAMYB gene was inserted into the 3′-region of the early lycopsid Leclercqia (Protolepidodendrales) from the Lower Devonian of NPTII expression cassette of this plasmid. To transform these plasmids into North America: evidence for a fundamental division in the lycopsids. Am. J. P. patens, polyethylene glycol-mediated transformation was performed essen- Botany 96, 1849–1860 (2009). tially as described , using 10 µg of linearized plasmid. Stable transformants were 11. Pacini, E., Franchi, G. G. & Hesse, M. The tapetum: its form, function and screened by Southern hybridization using a probe that detected the homologous possible phylogeny in Embryophyta. Plant Syst. Evol. 149, 155–185 (1985). region of the transgene. 12. Morant, M. et al. CYP703 is an ancient cytochromeP450 in land olants catalyzing in-chain hydroxylation of lauric acid to provide building blocks for Microscopy and phenotypic analyses. For GUS staining, tissues were transferred sporopollenin synthesis in pollen. Plant Cell 19, 1473–1487 (2007). to microcentrifuge tubes containing a solution of 50 mM sodium phosphate, pH 13. Hirano, K. et al. The GID1-mediated gibberellin perception mechanism is 7.0, 7% methanol, and 1 mM 5-bromo-4-chloro-3-indolyl-β-d-glucuronide for conserved in the Lycophyte Selaginella moellendorffii but not in the Bryophyte 12 h at 37 °C. Stained tissues were destained with an ethanol series. Physcomitrella patens. Plant Cell 19, 3058–3079 (2007). For in situ hybridization, plant materials were fixed in 4% (w/v) paraformal - 14. Yasumura, Y., Crumpton-Taylor, M., Fuentes, S. & Harberd, N. Step-by-step dehyde and 0.25% (v/v) glutaraldehyde in 0.1M s odium phosphate buffer, pH acquisition of the gibberellin-DELLA growth-regulatory mechanism during 7.4, overnight at 4 °C, dehydrated through a graded ethanol series followed by land-plant evolution. Curr. Biol. 17, 1225–1230 (2007). a t-butanol series, and finally embedded in Paraplast Plus (Sherwood Medical). 15. Ueguchi-Tanaka, M. et al. GIBBERELLIN INSENSITIVE DWARF1 encodes a Microtome sections (8–10 µm thick) were mounted on glass slides treated with soluble receptor for gibberellin. Nature 437, 693–698 (2005). silane. To produce the gene-specific probes, each DNA fragment was generated by 16. Ueguchi-Tanaka, M. et al. Molecular interactions of a soluble gibberellin PCR using complementary DNA and the specific primers. The resulting DNA frag - receptor, GID1, with a rice DELLA protein, SLR1, and gibberellin. Plant Cell ments were introduced between XhoI and EcoRI site of the pBluescript II SK + and 19, 2140–2155 (2007). pBluescript II KS + vectors, and were then transcribed in vitro under T7 promoter 17. Gocal, G. F., Poole, A. T., Gubler, F., Watts, R. J., Blundell, C. & King, R. with T7 RNA polymerase (Invitrogen) using the DIG RNA labeling kit (Roche). W. Long-day up-regulation of a GAMYB gene during Lolium temulentum Hybridization and immunological detection of the hybridized probes were per - inflorescence formation. Plant Physiol. 119, 1271–1278 (1999). formed essentially as described . The primer sequences used in this study are listed 18. Gocal, G. F. et al. GAMYB-like genes, flowering, and gibberellin signaling in in Supplementary Table S1. Arabidopsis. Plant Physiol. 127, 1682–1693 (2001). For observations by TEM, plant samples were vacuum infiltrated three times 19. Tanurdzic, M. & Banks, J. A. Sex-determining mechanisms in land plants. for 10 min with fixative (4% paraformaldehyde and 1% glutaraldehyde in 0.05 M Plant Cell 16, S61–S71 (2004). cacodylate buffer, pH 7.4). Procedures for ultrastructural studies were essentially 40 20. Archard, P., Herr, A., Baulcombe, D. C. & Harberd, N. P. Modulation of the same as described . floral development by a gibberellin-regulated microRNA. Development 131, For observations by SEM, fresh microspores/spores were placed on specimen 3357–3365 (2004). stubs using adhesive carbon tape. The samples were gold coated with Pt and Pd in 21. Toia, R. E. et al. Sporopollenin content of the spore apparatus of Azolla. Am. an ion sputter (E-1030, Hitachi), and imaged using SEM (S-3000N microscope, Fern J. 75, 38–43 (1985). Hitachi). 22. Uehara, K. et al. Ultrastructural study on microspore wall morphogenesis in For the acetolysis experiments, spores were collected in a tube and incubated in Isoetes Japonica (Isoetaceae). Am. J. Botany 78, 1182–1190 (1991). 9:1 (v/v) mixture of acetic anhydride and sulphuric acid at 70°C f or 20 min. Aer ft 23. Rowley, J. R., Morbelli, J. R. & El-Ghazaly, G. Microspore wall structure in the incubation, spores were transferred on a glass slide and observed by the light Selaginella kraussiana (Lycophyta). Taiwania 47, 115–128 (2002). microscopy (Olympus). nATuRE C ommunICATIons | 2:544 | DoI: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved. nATuRE C ommunICATIons | DoI: 10.1038/ncomms1552 ARTICLE 24. Millar, A. A. & Gubler, F. The Arabidopsis GAMYB-like genes, MYB33 and 40. Hara-Nishimura, I., Takeuchi, Y. & Nishimura, M. Molecular characterization MYB65, are microRNA-regulated genes that redundantly facilitate anther of a vacuolar processing enzyme related to a putative cysteine proteinase of development. Plant Cell 17, 705–721 (2005). Schistosoma mansoni. Plant Cell 5, 1651–1659 (1993). 25. Kaneko, M. et al. Where do gibberellin biosynthesis and gibberellin signaling 41. Kojima, M. et al. Highly sensitive and high-throughput analysis of plant occur in rice plants? Plant J. 35, 104–115 (2003). hormones using MS-probe modification and liquid chromatography-tandem 26. Hu, J. et al. Potential sites of bioactive gibberellin production during mass spectrometry: an application for hormone profiling in Oryza sativa. Plant reproductive growth in Arabidopsis. Plant Cell 20, 320–336 (2008). Cell Physiol. 50, 1201–1214 (2009). 27. Izumi, K., Kamiya, Y., Sakurai, A., Oshio, H. & Takahashi, N. Studies of sites of action of a new plant growth retardant (E)-1-(4-chlorophenyl)-4, 4- dimethyl- Acknowledgements 2-(1, 2, 4-trizol-1-yl)-penten-3-ol (S-3307) and comparative effects of its We thank Y. Hattori and M. Sato (Nagoya University) for technical assistance. We also stereisomers in a cell free system of Cucurbits maxima. Plant Cell Physiol. 29, thank Y. Nagamura and R. Motoyama (Rice Genome Resource Center of the National 821–827 (1985). Institute of Agribiological Sciences, Japan) for the use of the rice microarray analysis 28. Nakayama, I., Kobayashi, M., Kamiya, Y., Abe, H. & Sakurai, A. Effects of a system and technical support. This study was supported by a Grant-in-Aid from the plant-growth regulat or, prohexadione-calcium (BX-112), on the endogenous Ministry of Education, Culture, Sports, Science and Technology of Japan (18107001 levels of gibberellins in rice. Plant Cell Physiol. 33, 59–62 (1991). and 18075006 to M.M.), a Grant-in-Aid for Scientific Research on Innovative Areas 29. Sakakibara, K., Nishiyama, T., Deguchi, H. & Hasebe, M. Class 1 KNOX genes (23113001 and 23113005 to M.M.) and Global COE Program ‘Advanced Systems- are not involved in shoot development in the moss Physcomitrella patens but do Biology: Designing The Biological Function’, MEXT, Japan. function in sporophyte development. Evol. Dev. 10, 555–566 (2008). 30. Dobritsa, A. A. et al. CYP704B1 is a long-chain fatty acid omega-hydroxylase essential for sporopollenin synthesis in pollen of Arabidopsis. Plant Physiol. Author contributions 151, 574–589 (2009). M.U.-T., M.H. and M.M. conceived the project and designed the experiments together 31. Hesse, M. & Waha, M. A new look at the acetolysis method. Plant Syst. Evol. with K.A. K.A. performed most of the experiments. Y.H. and M.H. generated all 163, 147–152 (1989). transgenic lines of P. patens using homologous recombination. K.M. and H.S. measured 32. Estébanez, B., Yamaguchi, T. & Deguchi, H. Ultrastructure of the spore in four the endogenous GA content of S. moellendorffii (Fig. 2f ). M.M. wrote the paper together Japanese species of Ptychomitrium Furnr. (Musci). Grana 45, 61–70 (2006). with K.A. and M.H. 33. Schuette, S. et al. Novel localization of callose in the spores of Physcomitrella patens and phylogenomics of the callose synthase gene family. Ann. Bot. 103, Additional information 749–756 (2009). Accession Codes: The microarray data have been deposited in the Gene Expression 34. Singer, S. D. & Ashton, N. W. Revelation of ancestral roles of KNOX genes by a Omnibus under accession code GSE32652. functional analysis of Physcomitrella homologues. Plant Cell Rep. 26, 2039–2054 (2007). 35. Gubler, F. et al. Target genes and regulatory domains of the GAMYB Supplementary Information accompanies this paper at http://www.nature.com/ transcriptional activator in cereal aleurone. Plant J. 17, 1–9 (1999). naturecommunications 36. Hayashi, K. et al. Endogenous diterpenes derived from ent-kaurene, a Competing financial interests: The authors declare no competing financial interests. common gibberellin precursor, regulate protonema differentiation of the moss Physcomitrella patens. Plant Physiol. 153, 1085–1097 (2010). Reprints and permission information is available online at http://npg.nature.com/ 37. Banks, J. A. Gametophyte development in ferns. Annu. Rev. Plant Physiol. Plant reprintsandpermissions/ Mol. Biol. 50, 163–186 (1999). How to cite this article: Aya, K. et al. The Gibberellin perception system evolved to 38. Hiei, Y., Ohta, S., Komari, T. & Kumashiro, T. Efficient transformation of rice regulate a pre-existing GAMYB-mediated system during land plant evolution. Nat. (Oryza sativa L.) mediated by Agrobacterium and sequence analysis of the Commun. 2:544 doi: 10.1038/ncomms1552 (2011). boundaries of the T-DNA. Plant J. 6, 271–282 (1994). 39. Kouchi, H. & Hata, S. Isolation and characterization of novel nodulin cDNAs License: This work is licensed under a Creative Commons Attribution-NonCommercial- Share Alike 3.0 Unported License. To view a copy of this license, visit http:// representing genes expressed at early stages of soybean nodule development. Mol. Gen. Genet. 238, 106–119 (1993). creativecommons.org/licenses/by-nc-sa/3.0/ nATuRE C ommunICATIons | 2:544 | DoI: 10.1038/ncomms1552 | www.nature.com/naturecommunications © 2011 Macmillan Publishers Limited. All rights reserved.

Journal

Nature Communications – Springer Journals

Published: Nov 22, 2011

There are no references for this article.