Rozprawy doktorskie na temat „OsMADS4”

Le riz (Oryza sativa L) est la source principale d'alimentation pour plus de la moitié de la population mondiale (Khush, 2005). La production de riz devrait augmenter de plus de 40% en 2030 pour satisfaire la demande de croissance de la population. Chaque année, environ 25% de la production est perdue à cause des insectes ravageurs, des maladies et des mauvaises herbes (Khus, 2005). Des pertes semblables sont dues aux stress abiotiques comme la sécheresse. L'objectif de mon travail de thèse a consisté à étudier la fonction de deux facteurs de transcription (FT) à boîte MADS OsMADS25 et OsMADS26 dans la réponse aux stress ou dans le développement. Pour cela, j'ai généré des lignées de riz surexprimant les ADNc codant ces FT et aussi des lignées interférées pour le gène OsMADS26 en utilisant deux GST différentes pour induire de l'ARN interférence destiné à détruire les ARNm OsMADS26. Dans le cas du gène OsMADS25 qui appartient à un groupe de cinq gènes phylogénétiquement proches, j'ai généré des plantes exprimant la protéine OsMADS25 fusionnée avec le motif répresseur dominant de la transcription EAR. Les lignées T2 exprimant le FT OsMADS25 fusionné au motif EAR présentent un phénotype semblable à celui d'une lignée d'insertion de TDNA dans ce gène. Ces plantes sont caractérisées par une forte réduction du nombre de leur talle et par une hauteur plus importante de la talle principale. Les plantes qui surexpriment OsMADS25 natif ne présentent pas de phénotype particulier. Ceci suggère que le gène OsMADS25 pourrait être impliqué dans la régulation du nombre de talles chez le riz bien qu'il soit exprimé au niveau de la racine. Le mode d'action du gène OsMADS25 sur le contrôle du développement des méristèmes axillaire du riz reste à préciser. Les lignées interférées OsMADS26 présentent une meilleure résistance à Magnaporthe oryzae (Mo) et à Xanthomonas oryzae pv. Oryzae (Xoo), deux principaux pathogènes du riz, et aussi une meilleure capacité de restauration après l'application d'un stress hydrique par rapport aux lignées témoin tandis que les lignées surexprimant OsMADS26 sont plus sensibles à ces stress. Les analyses de QPCR et du transcriptome que nous avons effectuées ont mis en évidence l'expression constitutive plus élevée dans les lignées interférées de plusieurs gènes de réponse aux stress biotique et abiotique. Ces résultats suggèrent que OsMADS26 pourrait être un inhibiteur général des mécanismes de défense de la plante et que les plantes interférée OsMADS26 sont dans un statut physiologique de type primed-like qui leur permettent d'être plus résistantes aux stress. Les lignées interférées pour OsMADS26 sont très peu affectées dans leur développement. Le gène OsMADS26 est donc un gène très intéressant pour les programmes d'amélioration du riz<br>MADS-box transcription factors (TF) have been mostly characterized for their involvement of plant development such as floral organogenesis and flowering time. Some of them are involved in stress related developmental processes such as abscission, fruit ripening and senescence. Overexpression of the rice OsMADS26 TF suggested a function in stress response. Here we report that OsMADS26 interfered lines presented a better resistance against two major pathogens of rice, Xanthomonas oryzae (Xoo) and Magnaportae oryzae (Mo) and a better recovery capacity after a water stress period. Transcriptome analysis revealed that several biotic and abiotic stresses related genes were up regulated in OsMADS26 interfered lines. In addition QPCR analysis showed that the expression of a set of biotic and abiotic genes was induced when OsMADS26 interfered lines were infected by Xoo or submitted to a water stress. This indicated that OsMADS26 is a negative regulator of biotic and abiotic stress response in rice. Taking in account the data previously published that showed that inducible overexpression of OsMADS26 resulted in the activation of expression of genes involved in jasmonic acid or reactive oxygen species biosynthesis, we postulate that OsMADS26 may be a hub regulator of stress response in rice and that it may be posttranscriptional regulated to modulate negatively or positively rice response to various stresses.In addition we have shown in this thesis that an insertion mutant line disrupting the OsMADS25 gene is characterized by a reduced number of tiller. This phenotype was also obtained in transgenic lines expressing the OsMADS25 transcription factor fused with a dominant motif inhibitor of transcription. Thissuggested that OsMADS25 is involved in the control of tiller development in rice.Key words: Rice, stress, blast, tillering, MADS-box, transcription factor, OsMADS26, OsMADS25, transcriptome

Organs in modern dicot flowers are positioned in concentric rings (whorls). The outermost whorl has green protective sepals, internal to which are showy petals, and the reproductive stamen and carpel whorls. Florets of rice, cereals and grasses evolved certain morphological and functionally distinct features in their non-reproductive organs. Striking among them are the highly modified petal analogs; called lodicules, and the large bract-like outermost organs named palea and lemma. The analogy of these modified rice floret organs to sepals and petals is debated. The two lodicules of rice florets are small (limited growth in Proximal-Distal axis), thick (extensive growth in Dorsal-Ventral axis) and are asymmetrically positioned to occupy only one half of the second whorl. They perform an important mechanical role in the partial opening of the flower for stamen emergence and subsequent closing. Their asymmetric position, small fleshy structure with many parenchymatous cell layers and their regulated physiology for swelling and collapse are critical for these functions. Understanding the developmental mechanisms of these organs that underlie their function is of direct interest to evo-devo plant biologists and breeding programs aimed at crop yield improvement. Given these implications, for a deeper understanding of plant development and potential future uses in crop breeding, we define the objectives of this study and report our key findings Objective 1: Uncovering the gene targets of rice class B PISTILLATA-like (PI-like) factors that regulate lodicule and stamen development Objective 2: Characterizing an OsMADS2 target gene; AP2/EREBP86 encoding AINTEGUMENTA-like/PLETHORA (AIL/PLT) family transcription factor. Overall, this study expands our knowledge on traits and molecular mechanisms controlled by rice class B PI-like factors: OsMADS2 and OsMADS4, and provides new insights on their functional divergence that greatly extend our understanding of lodicule and stamen development. This study sheds light on some molecular mechanisms triggered by AP2/EREBP86 that can initiate and maintain shoot meristem fate which have the potential to improve somatic embryogenesis.

碩士<br>國立中興大學<br>分子生物學研究所<br>105<br>M52048, a T-DNA insertion mutant, shows dwarf, early flowering, node bending and impaired in panicle exsertion and it has been demonstrated to have three flanking genes OsMADS14, OsMADS34 and OsCP7 activated. Transgenic plants, Ubi:OsMADS14, ectopically overexpressing OsMADS14 revealed dwarf and early flowering phenotype, and up-regulated expression of florigen genes Hd3a and RFT1 were observed in the previous study. The first part of this study is to investigate the possible mechanisms that how the expression of Hd3a and RFT1 can be up-regulated by OsMADS14. At first, the expression profile of rice florigen regulators such as OsPRR37, OsCO3, DTH2, OsDof12, PPS and RFL were analyzed. Among them, the expression of RFL was promoted and OsCO3 was repressed suggesting that the activation of Hd3a and RFT1 might result from the up-regulation of RFL and/or down-regulation of OsCO3. Secondly, the phase transition miRNAs, miR156 and miR172 were investigated and results showed the expression of miR156 was slightly activated while miR172 was repressed suggesting that the delayed phase transition might occurred in Ubi:OsMADS14. Thirdly, the expression of potential floral repressors such as putative OsTOE1 (TARGET OF EAT1), putative OsTEM1 (OsTEMPRANILLO1), putative OsTEM2 and RCN1 that may regulate by OsMADS14 were analyzed. Results showed that OsTEM1 and RCN1 were activated, suggesting the possible involvement of partial counteraction of the effect of early flowering by OsMADS14. Finally, physiological functions of these putative floral repressors were studied by investigating their correspondent T-DNA insertion activation mutants such as M59289 (for OsTOE1), M78020 (for RCN1) and M89461 (for OsTEM1). Among these mutants, only the putative OsTEM1 gene in mutant M89461 was activated and a slightly late flowering phenotype was observed in mutant M89461 as well suggesting that the putative OsTEM1 might be a floral repressor in rice. Transgenic plants, Ubi:OsMADS34, ectopically overexpressing OsMADS34 revealed impaired panicle exsertion phenotype and that resulted by the reduced length of the first internode was previously demonstrated. The second part of this study is to investigate the possible mechanisms that how the length of the first internode was significant reduced by overexpressing OsMADS34. Results of the microarray assays comparing Ubi:OsMADS34 to TNG67 revealed differential expressed of many stress-related genes. Among them, 31 genes designated as M1 to M31 were selected for further study and their expressions were to be confirmed by RT-PCR. Results showed that the expression profile of M1, M2, M4, M8, M9, M10, M16, M17, M18, M21, M22, M23, M24, M26, M27, M30 and M31 were in accordance with the results of microarray assays suggesting that the expression of these genes were regulated by overexpression of OsMADS34.

碩士<br>國立中興大學<br>分子生物學研究所<br>99<br>A rice T-DNA insertion mutant M0052048 showing extreme early flowering, bending tillers and impairment in panicle exertion was isolated from the Taiwan Rice Insertion Mutant (TRIM) library. This mutant contained a copy of the T-DNA tag inserted in the chromosome number 3 at the 64,867 bp position of OSJBa0032E21 BAC clone where the expression of 048-3 (OsMADS34), 048-4 (OsMADS14) and 048-7 (putative cysteine proteinase, OsCP7) genes were activated by the 35S enhancer located in the T-DNA. The OsMADS14 and OsMADS34 are MADS box-containing transcription factors that have important roles in plant growth and development and furthermore, OsCP7 is a putative cysteine proteinase that may be involved in the program cell death, pollen and xylem maturation, embryogenesis and flowering time in plants. In order to understand the functions of these activated genes and their contribution to the mutant phenotype, transgenic rice over-expressing each of these genes were created and expression profiles of these genes in various rice plant tissues were analyzed. RT-PCR analysis revealed that OsMADS14 and OsMADS34 expressed mainly in panicles but not necessarily in vegetative tissues, with the exception of a relatively high expression level of OsMADS14 in the 90-day-old leaf; For OsCP7, no clear expression signals were detected in all tested tissues. Transgenic rice Ubi:MADS14 showed constitutive expression of OsMADS14 in leaf tissue and revealed extreme early (63.6 days vs 121.3 days in TNG67) flowering and tiller bending phenotypes. The transgenic rice Ubi:MADS34 showed constitutive expression of OsMADS34 and revealed impairment in panicle exertion and slightly early (110 days vs 121.3 days) flowering compared to TNG67. For the overexpression study of OsCP7, no transgenic plant with ubiqutin promoter construct (Ubi:CP7) was obtained, thus a 1.6 kb promoter region from native OsCP7 gene was used to replace the ubiqutin promoter and several ectopically-expressed OsCP7 transgenic rice plants were obtained. These OsCP7 transgenic rice plants showed increased levels of OsCP7 mRNA and protein and higher cysteine protease activity compare to that of TNG67. In addition, these plants had reduced height, approximately 88% of TNG67, and revealed brown lesions on the surfaces of most spikelets that were neither observed in the TNG67 nor in M0052048. In summary, the present study suggests that the activation of OsMADS14 and OsMADS34 genes contributes to early flowering, bending tillers and impairment in panicle exertion phenotypes in the mutant M0052048. However, the effect of OsCP7 activation in mutant M0052048 and the function of OsCP7 in rice plants are still not clear and further investigations will be needed to answer these questions.

Plant reproductive development begins when vegetative shoot apical meristems change their fate to inflorescence meristems which develop floral meristems on the flanks. This process of meristem fate change and organ development involves regulated activation and/or repression of many cell fate determining factors that execute down-stream gene expression cascades. Flowers are formed when floral organs are specified on the floral meristem in four concentric whorls. In the model dicot plant Arabidopsis, the identity and pattern of floral organs is determined by combined actions of MADS-domain containing transcription factors of the classes A, B, C, D and E. Rice florets are produced on a compact higher order branch of the inflorescence and have morphologically distinct non-reproductive organs that are positioned peripheral to the male and female reproductive organs. These unique outer organs are the lemma and palea that create a closed floret internal to which are a pair of lodicules that are asymmetrically positioned fleshy and reduced petal-like organs. The unique morphology of these rice floret organs pose intriguing questions on how evolutionary conserved floral meristem specifying and organ fate determining factors bring about their distinct developmental functions in rice. We have studied the functions for two rice MADS-box proteins, OsMADS2 and OsMADS1, to understand their role as master regulators of gene expression during rice floret meristem specification and organ development. OsMADS2; a transcriptional regulator of genes expression required for lodicule development Arabidopsis B-function genes AP3 and PI are stably expressed in the whorl 2 and 3 organ primordia and they together with other MADS-factors (Class A+E or C+E) regulate the differentiation of petals and stamens (Jack et al, 1992; Goto and Meyerowitz, 1994). Rice has a single AP3 ortholog, SPW1 (OsMADS16) but has duplicated PI-like genes, OsMADS2 and OsMADS4. Prior studies in our lab on one of these rice PI-like genes OsMADS2 showed that it is needed for lodicule development but is dispensable for stamen specification (Kang et al., 1998; Prasad and Vijayraghavan, 2003). Functional divergence between OsMADS2 and OsMADS4 may arise from protein divergence or from differences in their expression patterns within lodicule and stamen whorls. In this study, we have examined the dynamic expression pattern of both rice PI-like genes and have examined the likelihood of their functional redundancy for lodicule development. We show OsMADS2 transcripts occur at high levels in developing lodicules and transcripts are at reduced levels in stamens. In fully differentiated lodicules, OsMADS2 transcripts are more abundant in the distal and peripheral regions of lodicules, which are the tissues that are severely affected in OsMADS2 knock-down florets (Prasad and Vijayraghavan, 2003). The onset of OsMADS4 expression is in very young floret meristems before organ primordia emergence and this is expressed before OsMADS2. In florets undergoing organogenesis, high level OsMADS4 expression occurs in stamens and carpels and transcripts are at low level in lodicules (Yadav, Prasad and Vijayraghvan, 2007). Thus, we show that these paralogous genes differ in the onset of their activation and their stable transcript distribution within lodicules and stamens that are the conserved expression domains for PI-like genes. Since the expression of OsMADS4 in OsMADS2 knock-down florets is normal, our results show OsMADS2 has unique functions in lodicule development. Thus our data show subfunctionalization of these paralogous rice PI-like genes. To identify target genes regulated by OsMADS2 that could contribute to lodicule differentiation, we have adopted whole genome transcript analysis of wild-type and dsRNAiOsMADS2 panicles with developing florets. This analysis has identified potential down-stream targets of OsMADS2 many of which encode transcription factors, components of cell division cycle and signalling factors whose activities likely control lodicule differentiation. The expression levels of few candidate targets of OsMADS2 were examined in various floret organs. Further, the spatial expression pattern for four of these down-stream targets of OsMADS2 was analysed and we find overlap with OsMADS2 expression domains (Yadav, Prasad and Vijayraghvan, 2007). The predicted functions of these OsMADS2 target genes can explain the regulation of growth and unique vascular differentiation of this short fleshy modified petal analog. OsMADS1, a rice E-class gene, is a master regulator of other transcription factors and auxin and cytokinin signalling pathways In Arabidopsis four redundant SEPALLATA factors (E-class) are co-activators of other floral organ fate determining MADS-domain factors (classes ABCD) and thus contribute to floral meristem and floral organ development (Krizek and Fletcher, 2005). Among the grass-specific sub-clade of SEP-like genes, rice OsMADS1 is the best characterized. Prior studies in our lab showed that OsMADS1 is expressed early throughout the floret meristem before organ primordia emergence and later is restricted to the developing lemma and palea primordia with weak expression in carpel (Prasad et al, 2001). Stable expression continues in these floret organs. OsMADS1 plays critical non-redundant functions to specify a determinate floret meristem and also regulates floret organ identities (Jeon et al., 2000; Prasad et al, 2001; 2005; Agarwal et al., 2005; Chen et al., 2006). In the present study, we have adopted two different functional genomic approaches to identify genes down-stream of OsMADS1 in order to understand its mechanism of action during floret development. We have studied global transcript profiles in WT and dsRNAiOsMADS1 panicles and find OsMADS1 is a master regulator of a significant fraction of the genome’s transcription factors and also a number of genes involved in hormone-dependent cell signalling. We have validated few representative genes for transcription factors as targets regulated by OsMADS1. In a complementary approach, we have determined the consequences of induced-ectopic over-expression of a OsMADS1:ΔGR fusion protein in shoot apical meristems of transgenic plants. Transcript levels for candidate target genes were assessed in induced tissues and compared to mock-treated meristems and also with meristems induced for OsMADS1:ΔGR but blocked for new protein synthesis. These analyses show that OsMADS55 expression is directly regulated by OsMADS1. Importantly, OsMADS55 is related to SVP that plays an important role in floral transition and floral meristem identity in Arabidopsis. OsHB3 and OsHB4, homeodomain transcription factors, with a probable role in meristem function, are also directly regulated by OsMADS1. The regulation of such genes by OsMADS1 can explain its role in floret meristem specification. In addition to regulating other transcription factors, OsMADS1 knock-down affects expression of genes encoding proteins in various steps of auxin and cytokinin signalling pathways. Our differential expression profiling showed OsMADS1 positively regulates the auxin signalling pathway and negatively regulates cytokinin mediated signalling events. Through our induced ectopic expression studies of OsMADS1:ΔGR, we show OsMADS1 directly regulates the expression of OsETTIN2, an auxin response transcription factor, during floret development. Overall, we demonstrate that OsMADS1 modulates hormonal pathways to execute its functions during floret development on the spikelet meristems. Functional studies of OsMGH3; an auxin-responsive indirect target of OsMADS1 To better understand the contribution of auxin signalling during floret development, we have functionally characterized OsMGH3, a down-stream indirect target of OsMADS1, which is a member of the auxin-responsive GH3 family. The members of this family are direct targets of auxin response factors (ARF) class of transcription factors. GH3-proteins inactivate cellular auxin by conjugating them with amino acids and thus regulate auxin homeostasis in Arabidopsis (Staswick et al., 2005). OsMGH3 expression in rice florets overlaps with that of OsMADS1 (Prasad et al, 2005). In this study, we have demonstrated the consequences of OsMGH3 over-expression and knock-down. The over-expression of OsMGH3 during vegetative development causes auxin-deficient phenotypes such as dwarfism and loss of apical dominance. Its over-expression in developing panicles that was obtained by driving its expression from tissue-specific promoters created short panicles with reduced branching. The latter is a phenotype similar to that observed upon over-expression of OsMADS1. In contrast, the down-regulation of endogenous OsMGH3 through RNA-interference produced auxin over-production phenotypes such as ectopic rooting from aerial nodes. Knock-down of OsMGH3 expression in florets affected carpel development and pollen viability both of which affect floret fertility. Taken together, this study provides evidence for the importance of auxin homeostasis and its transcriptional regulation during rice panicle branching and floret organ development. Our analysis of various conserved transcription factors during rice floret development suggest that factors like OsMADS2, OsMADS4 and OsMADS1 are master regulators of gene expression during floret meristem specification and organ development. The target genes regulated by these factors contribute to development of morphologically distinct rice florets.

Plant reproductive development begins when vegetative shoot apical meristems change their fate to inflorescence meristems which develop floral meristems on the flanks. This process of meristem fate change and organ development involves regulated activation and/or repression of many cell fate determining factors that execute down-stream gene expression cascades. Flowers are formed when floral organs are specified on the floral meristem in four concentric whorls. In the model dicot plant Arabidopsis, the identity and pattern of floral organs is determined by combined actions of MADS-domain containing transcription factors of the classes A, B, C, D and E. Rice florets are produced on a compact higher order branch of the inflorescence and have morphologically distinct non-reproductive organs that are positioned peripheral to the male and female reproductive organs. These unique outer organs are the lemma and palea that create a closed floret internal to which are a pair of lodicules that are asymmetrically positioned fleshy and reduced petal-like organs. The unique morphology of these rice floret organs pose intriguing questions on how evolutionary conserved floral meristem specifying and organ fate determining factors bring about their distinct developmental functions in rice. We have studied the functions for two rice MADS-box proteins, OsMADS2 and OsMADS1, to understand their role as master regulators of gene expression during rice floret meristem specification and organ development. OsMADS2; a transcriptional regulator of genes expression required for lodicule development Arabidopsis B-function genes AP3 and PI are stably expressed in the whorl 2 and 3 organ primordia and they together with other MADS-factors (Class A+E or C+E) regulate the differentiation of petals and stamens (Jack et al, 1992; Goto and Meyerowitz, 1994). Rice has a single AP3 ortholog, SPW1 (OsMADS16) but has duplicated PI-like genes, OsMADS2 and OsMADS4. Prior studies in our lab on one of these rice PI-like genes OsMADS2 showed that it is needed for lodicule development but is dispensable for stamen specification (Kang et al., 1998; Prasad and Vijayraghavan, 2003). Functional divergence between OsMADS2 and OsMADS4 may arise from protein divergence or from differences in their expression patterns within lodicule and stamen whorls. In this study, we have examined the dynamic expression pattern of both rice PI-like genes and have examined the likelihood of their functional redundancy for lodicule development. We show OsMADS2 transcripts occur at high levels in developing lodicules and transcripts are at reduced levels in stamens. In fully differentiated lodicules, OsMADS2 transcripts are more abundant in the distal and peripheral regions of lodicules, which are the tissues that are severely affected in OsMADS2 knock-down florets (Prasad and Vijayraghavan, 2003). The onset of OsMADS4 expression is in very young floret meristems before organ primordia emergence and this is expressed before OsMADS2. In florets undergoing organogenesis, high level OsMADS4 expression occurs in stamens and carpels and transcripts are at low level in lodicules (Yadav, Prasad and Vijayraghvan, 2007). Thus, we show that these paralogous genes differ in the onset of their activation and their stable transcript distribution within lodicules and stamens that are the conserved expression domains for PI-like genes. Since the expression of OsMADS4 in OsMADS2 knock-down florets is normal, our results show OsMADS2 has unique functions in lodicule development. Thus our data show subfunctionalization of these paralogous rice PI-like genes. To identify target genes regulated by OsMADS2 that could contribute to lodicule differentiation, we have adopted whole genome transcript analysis of wild-type and dsRNAiOsMADS2 panicles with developing florets. This analysis has identified potential down-stream targets of OsMADS2 many of which encode transcription factors, components of cell division cycle and signalling factors whose activities likely control lodicule differentiation. The expression levels of few candidate targets of OsMADS2 were examined in various floret organs. Further, the spatial expression pattern for four of these down-stream targets of OsMADS2 was analysed and we find overlap with OsMADS2 expression domains (Yadav, Prasad and Vijayraghvan, 2007). The predicted functions of these OsMADS2 target genes can explain the regulation of growth and unique vascular differentiation of this short fleshy modified petal analog. OsMADS1, a rice E-class gene, is a master regulator of other transcription factors and auxin and cytokinin signalling pathways In Arabidopsis four redundant SEPALLATA factors (E-class) are co-activators of other floral organ fate determining MADS-domain factors (classes ABCD) and thus contribute to floral meristem and floral organ development (Krizek and Fletcher, 2005). Among the grass-specific sub-clade of SEP-like genes, rice OsMADS1 is the best characterized. Prior studies in our lab showed that OsMADS1 is expressed early throughout the floret meristem before organ primordia emergence and later is restricted to the developing lemma and palea primordia with weak expression in carpel (Prasad et al, 2001). Stable expression continues in these floret organs. OsMADS1 plays critical non-redundant functions to specify a determinate floret meristem and also regulates floret organ identities (Jeon et al., 2000; Prasad et al, 2001; 2005; Agarwal et al., 2005; Chen et al., 2006). In the present study, we have adopted two different functional genomic approaches to identify genes down-stream of OsMADS1 in order to understand its mechanism of action during floret development. We have studied global transcript profiles in WT and dsRNAiOsMADS1 panicles and find OsMADS1 is a master regulator of a significant fraction of the genome’s transcription factors and also a number of genes involved in hormone-dependent cell signalling. We have validated few representative genes for transcription factors as targets regulated by OsMADS1. In a complementary approach, we have determined the consequences of induced-ectopic over-expression of a OsMADS1:ΔGR fusion protein in shoot apical meristems of transgenic plants. Transcript levels for candidate target genes were assessed in induced tissues and compared to mock-treated meristems and also with meristems induced for OsMADS1:ΔGR but blocked for new protein synthesis. These analyses show that OsMADS55 expression is directly regulated by OsMADS1. Importantly, OsMADS55 is related to SVP that plays an important role in floral transition and floral meristem identity in Arabidopsis. OsHB3 and OsHB4, homeodomain transcription factors, with a probable role in meristem function, are also directly regulated by OsMADS1. The regulation of such genes by OsMADS1 can explain its role in floret meristem specification. In addition to regulating other transcription factors, OsMADS1 knock-down affects expression of genes encoding proteins in various steps of auxin and cytokinin signalling pathways. Our differential expression profiling showed OsMADS1 positively regulates the auxin signalling pathway and negatively regulates cytokinin mediated signalling events. Through our induced ectopic expression studies of OsMADS1:ΔGR, we show OsMADS1 directly regulates the expression of OsETTIN2, an auxin response transcription factor, during floret development. Overall, we demonstrate that OsMADS1 modulates hormonal pathways to execute its functions during floret development on the spikelet meristems. Functional studies of OsMGH3; an auxin-responsive indirect target of OsMADS1 To better understand the contribution of auxin signalling during floret development, we have functionally characterized OsMGH3, a down-stream indirect target of OsMADS1, which is a member of the auxin-responsive GH3 family. The members of this family are direct targets of auxin response factors (ARF) class of transcription factors. GH3-proteins inactivate cellular auxin by conjugating them with amino acids and thus regulate auxin homeostasis in Arabidopsis (Staswick et al., 2005). OsMGH3 expression in rice florets overlaps with that of OsMADS1 (Prasad et al, 2005). In this study, we have demonstrated the consequences of OsMGH3 over-expression and knock-down. The over-expression of OsMGH3 during vegetative development causes auxin-deficient phenotypes such as dwarfism and loss of apical dominance. Its over-expression in developing panicles that was obtained by driving its expression from tissue-specific promoters created short panicles with reduced branching. The latter is a phenotype similar to that observed upon over-expression of OsMADS1. In contrast, the down-regulation of endogenous OsMGH3 through RNA-interference produced auxin over-production phenotypes such as ectopic rooting from aerial nodes. Knock-down of OsMGH3 expression in florets affected carpel development and pollen viability both of which affect floret fertility. Taken together, this study provides evidence for the importance of auxin homeostasis and its transcriptional regulation during rice panicle branching and floret organ development. Our analysis of various conserved transcription factors during rice floret development suggest that factors like OsMADS2, OsMADS4 and OsMADS1 are master regulators of gene expression during floret meristem specification and organ development. The target genes regulated by these factors contribute to development of morphologically distinct rice florets.

碩士<br>國立中興大學<br>分子生物學研究所<br>103<br>The T-DNA mutant M52048 identified from Taiwan Rice Insertional Mutant (TRIM) library showed dwarf, early flowering, node bending and impaired in panicle exertion. Three flanking genes, OsMADS34, OsMADS14 and OsCP7 (putative cysteine protease 7) were activated in this mutant. Both OsMADS34 and OsMADS14 belong to MADS-box gene family that may participate in regulation of flowering time and the identity of floral organ. OsCP7 is a member of C1A cysteine proteases. In this study, the function of OsMADS14 and OsCP7 were further investigated. Previous study has demonstrated that over-expression of OsMADS14 could cause early flowering. To understand whether or not any other flowering regulatory genes were affected by the expression of OsMADS14. The flowering regulatory genes including OsGI, OsMADS50, Ehd2, Hd1, Ehd1, Hd3a, RFT1, OsMADS14 and OsMADS18 were investigated in mutant M52048 and Ubi:OsMADS14, Ubi:OsMADS34 and OsCP7:OsCP7 transgenic rice respectively. Results showed two florigen genes, Hd3a and RFT1, expressed much earlier in M52048 and Ubi:OsMADS14 but not in Ubi:OsMADS34 and OsCP7:OsCP7, suggesting that Hd3a and RFT1 were regulated by the expression of OsMADS14. The mechanism how the expression of OsMADS14 could regulate florigen genes requires further investigation. Expression of OsCP7 driven by the maize ubiquitin promoter or the CaMV 35S promoter in transgenic rice cannot be obtained successfully. However transgenic rice, OsCP7:OsCP7 using 1.6 kb of OsCP7 promoter could be easily obtained and OsCP7:OsCP7 transgenic rice plants revealed slightly shorter in plant height, delayed flowering, lower fertility and lesion-like spots on spikelet. In contrast to the wild-type where no OsCP7 was detected in panicles, the RNA and protein expressions of OsCP7 in OsCP7:OsCP7 transgenic rice were detected in leaves at all development stages and panicles, and their expressions in transgenic rice correlated to the observed phenotypes. In addition, the phenotypes of segregated homozygous plants showed more significant than those of heterozygous plants within the same transgenic line, suggesting the dosage effect of transgene. However the expression levels of RNA and protein cannot be differentiated between homo- and hetero-zygous lines. To unravel the causes that lead to lower fertility of OsCP7:OsCP7, the floral organ and pollen viability were investigated. The floral organ showed no obvious differences between wild-type and OsCP7:OsCP7, but the pollen viability of OsCP7:OsCP7 was lower than that of wild-type, indicating that continuing expression of OsCP7 influence pollen development. Further investigation also indicated that the lesion-like spots on spikelet was correlated with the expression levels of OsCP7 and the lesions could possibly due to the programmed cell death caused by the activity of increased mature OsCP7 present in spikelet.

碩士<br>輔英科技大學<br>生物科技系碩士班<br>100<br>Oncidium is one of the most important orchids used for cut flowers and potted plants in Taiwan. The creation of new cultivars with novel traits is important for Oncidium to enhance the competitiveness of orchid industry. However, traditional breeding processes are limited by the long life cycle and self-incompatibility. The objectives of this study intend to analyze rice AP1-like gene OsMADS14 in Arabidopsis and Oncidium through genetic transformation. In this study, the rice gene OsMADS14 driven by maize ubiqutin promoter was introduced into Arabidopsis and Oncidium Gower Ramsey mediated by Agrobacterium tumefaciens. Twenty one independent ubiquitin::OsMADS14 transgenic Arabidopsis plants were produced, confirmed by PCR and RT-PCR analysis. The ectopic expression of OsMADS14 in transgenic Arabidopsis plants showed dwarf, early flowering and terminal flowers. In addition, the OsMADS14 gene were introduced into Oncidium Gower Ramsey using Agrobacterium tumefaciens-mediated transformation, protocorm-like bodies ( PLBs ) of Oncidium were used as explants materials for genetic transformation and selected on medium containing 5ppm hygromycin. The resistant transgenic Oncidium were analyzed by PCR, RT-PCR and histochemical GUS assay, indicating the transgene integrated into the genome. The resistant transgenic Oncidium showed leave bending phenotype. The results of this study suggest that the rice OsMADS14 gene could be applied early-flowering and dwarf traits to ornamental flowers.

碩士<br>國立中興大學<br>分子生物學研究所<br>103<br>The T-DNA mutant M52048 identified from Taiwan Rice Insertional Mutant (TRIM) library showed dwarf, early flowering, node bending and impaired in panicle exertion. Three flanking genes, OsMADS34, OsMADS14 and OsCP7 (putative cysteine protease 7) were activated in this mutant. Both OsMADS34 and OsMADS14 belong to MADS-box gene family that may participate in regulation of flowering time and the identity of floral organ. OsCP7 encode a putative cysteine protease, belongs to C1A cysteine protease family, its function remains unknown. In this study the function of OsMADS34 and OsCP7 were further investigated. Previous study showed that over-expression of OsMADS34, Ubi:OsMADS34 transgenic rice, could cause slightly early flowering and impaired in panicle exertion. Morphological dissection indicated that the impaired in panicle exertion was mainly caused by shortening the first internode (peduncle). Inhibition of the peduncle elongation caused by drought stress and ABA accumulation has been reported. In the present study, some drought-related genes, such as DREB1A, DREB1E and EATB, were regulated in Ubi:OsMADS34 transgenic rice and therefore hypothesized that the phenotype of transgenic rice may regulated by drought stress or plant hormones. However, treated transgenic rice with ABA inhibitor and/or GA could not improve the peduncle elongation and panicle exertion, suggested that the shortened peduncle and impaired in panicle exertion in Ubi:OsMADS34 transgenic rice might not cause by ABA accumulation. Analysis of GA biosynthesis related genes and the cell elongation promoting genes, at the internodes, revealed high expression levels of EATB, EUI1, GA13ox1, GA20ox2, OsPK1 and lower expression of XTH28, suggested that the shortened internode might due to the imbalance expression of these genes. However the mechanism how these genes involved in internode elongation remain to be elucidated. In addition to the shortened peduncle, the anther development was also affected in Ubi:OsMADS34, suggesting that OsMADS34 function as an E class MADS-box gene may interact with other MADS-box genes to regulate the floral organ development. Further study by searching OsMADS34 interaction proteins will help us to unravel the possible function of OsMADS34. For the study of OsCP7 gene, we were unable to obtained stable transgenic rice lines with constitutive promoter constructs, suggesting that constitutively ectopic expression of OsCP7 might cause lethal. Instead, transgenic lines with a 1.6 kb of OsCP7 promoter construct, OsCP7:OsCP7, were successfully obtained. OsCP7:OsCP7 revealed slightly dwarf, delayed flowering, lesion-like spots on panicles and lower fertility, and these phenotypes are correlated to the expression of OsCP7 gene. The possible mechanisms that cause these aberrant panicle developments were under investigated.

碩士<br>國立中興大學<br>分子生物學研究所<br>107<br>Flowering is an important developmental process for seed producing in plants. Previous studies had showed that respective over-expressing OsMADS14, OsMADS15 and OsMADS18 in rice would cause early flowering phenotype, implied that the three genes might play vital roles in regulating flowering process of rice. Moreover, the accumulation of OsMADS14 in both of T-DNA activation tagging mutant M52048 and Ubi:OsMADS14 transgenic rice result in dwarf and early flowering phenotype, which accompanied with activation of Hd3a, RFT1 and OsMADS15, raising the possibility that these genes might cooperate with OsMADS14 in regulating flowering. Nevertheless, the ability of the three genes to promote flowering have not investigated and compared in the same genetic background yet, to elucidate among the three genes, which gene is mainly involved in regulating flowering process, the transgenic approach and CRISPR/Cas9 knockout approach for OsMADS14, OsMADS15 and OsMADS18 were conducted in cultivar TNG67. In the present study, most of the Ubi:OsMADS15 transgenic rice displayed dwarf and early flowering phenotypes, however, instead of promoting flowering, the accumulation of OsMADS18 in Ubi:OsMADS18 result in delayed flowering and had no obvious effect on plant height growth, indicated that OsMADS18 might not have ability to promote flowering in rice. Besides, according to the expression comparison of Hd3a, RFT1 and OsMADS14, OsMADS15 and OsMADS18 among the Ubi:OsMADS15 transgenic rice, we found that early flowering phenotype was not observed from the transgenic rice which had no activated Hd3a, RFT1 and OsMADS14, suggested that overexpression of OsMADS15 along is not enough to promote flowering. On top of that, the flowering time of Ubi:OsMADS15 might promote by coexistence of Hd3a, RFT1, OsMADS14 and OsMADS15, and how of both Hd3a and RFT1 were activated in Ubi:OsMADS14 and Ubi:OsMADS15 transgenic rice still needed to be investigated in future. In addition, the CRIPSR/Cas9 genome editing system was applied for OsMADS14, OsMADS15 and OsMADS18 gene knockout in this study. In the flowering time comparison among these osmads14, osmads15 and osmads18 gene knockout mutants, our preliminary data showed that both of osmads14 and osmads15 had no obvious delay on flowering time. Surprisingly, the flowering time was promoted earlier in T1 progenies of osmads18, which corresponded to the result observed from Ubi:OsMADS18, implied that OsMADS18 might function as a negative regulator rather than positive regulator in regulating rice flowering. Since all of these data were observed and collected from either T0 or T1 progenies of osmads14, osmads15 and osmads18 knockout mutants, the functional inference of OsMADS14, OsMADS15 and OsMADS18 in regulating rice flowering still needed to be further confirmed in genetic stable T2 progenies. In addition, according to the flowering pathway in Arabidopsis, which had showed that TEM1 is a flowering repressor, therefore, we speculated that pOsTEM1 might also play as flowering repressor in rice. To investigate the function of TEM1 homologous gene pOsTEM1, the pOsTEM1 activation mutant M89461 was isolated from TRIM data base, besides, the pOsTEM1 overexpression transgenic rice was also created in current study. Surprisingly, our preliminary data showed that the accumulation of pOsTEM1 in both of T-DNA mutant M89461 and Ubi:pOsTEM1 transgenic rice result in late flowering phenotype, suggested that delayed flowering phenotype might result from accumulation of pOsTEM1. However, whether the pOsTEM1 is a flowering repressor in rice still needed to be further confirmed by further experiment in future.

博士<br>國立中興大學<br>生命科學系所<br>101<br>The rice gene, OsMADS45, which belongs to the MADS-box E class gene, participates in the regulation of floral development. Previous studies have revealed that ectopic expression of OsMADS45 induces early flowering and reduces plant height under short-day (SD) conditions. However, the regulation mechanism of OsMADS45 overexpression remains unknown. We introduce an OsMADS45 overexpression construct Ubi:OsMADS45 into TNG67 plants (an Hd1 (Heading date 1) and Ehd1 (Early heading date 1) defective rice cultivar grown in Taiwan), and we analyzed the expression patterns of various floral regulators to understand the regulation pathways affected by OsMADS45 expression. The transgenic rice exhibit a heading date approximately 40 days earlier than that observed in TNG67 plants, and transgenic rice display small plant size and low grain yield. OsMADS45 overexpression did not alter the oscillating rhythm of the examined floral regulatory genes, Hd3a (Heading Date 3a) and RFT1 (RICE FLOWERING LOCUS T1), but advanced (by approximately 20 days) the up-regulation of the two florigens and suppressed the expression of Hd1 at the juvenile stage. The expression levels of OsMADS14 and OsMADS18, which are two well-known reproductive phase transition markers, were also increased at early developmental stages and are believed to be the major regulators responsible for early flowering in OsMADS45-overexpressing transgenic rice. OsMADS45 overexpression did not influence other floral regulator genes upstream of Hd1 and Ehd1, such as OsGI (OsGIGANTEA), Ehd2/Osld1/RID1 and OsMADS50. These results indicate that in transgenic rice, OsMADS45 overexpressing ectopically activates the upstream genes Hd3a and RFT1 at early developmental stage and up-regulates the expression of OsMADS14 and OsMADS18, which promote early flowering.

In angiosperms, specialized reproductive structures are borne in flowers to ensure their reproductive success. After the vegetative growth, plants undergo reproductive phase change to produce flowers. Floral meristems (FMs) are generated on the flanks of inflorescence and groups of specialized stem cells in the FM differentiate into four whorls of organs of a flower. In dicots, floral meristem successively gives rise to sepals, petals, stamens and carpels; after which it terminates. The fate of organs formed on FM is under the control of genetic regulators, key among which are members of MADS box transcription factor family. Their individual and combined act confers distinct identities to floral organs. Grass flowers are highly modified in structure. Rice flower, a model for grasses, is borne on a short branch called spikelet and they together from the basic structural units of the rice infloresences known as panicle. The outer whorl organs of a grass floret are bract-like structures known as lemma and palea to dicot sepals is highly dibated (see Chapter 1). In grass florets, petal homologs are a pair of highly reduced, fleshy bracts known as lodicules, while stamen and carpel homologs occupy the same position and share the same functions as their dicot counterparts. Aside from these distinct outer whorl organs, the florets are subtended by two pairs of bracts known as empty glumes and rudimentary glumes. The genetic regulators that control their unique identities and those that perform conserved functions are very intriguing and central questions in plant developmental biology. Using various contemporary and complementary technologies, we have analysed the molecular functions and downstream pathways of a MADS box transcription factor, OsMADSI during the rice floret meristem specification and organ development. Further by reverse genetics and overexpression studies, we have also functionally characterized two target genes of OsMADSI, OsETTINI and OsETTINI2 to understand their roles downstream to OsMADSI during the rice floret development.

In angiosperms, specialized reproductive structures are borne in flowers to ensure their reproductive success. After the vegetative growth, plants undergo reproductive phase change to produce flowers. Floral meristems (FMs) are generated on the flanks of inflorescence and groups of specialized stem cells in the FM differentiate into four whorls of organs of a flower. In dicots, floral meristem successively gives rise to sepals, petals, stamens and carpels; after which it terminates. The fate of organs formed on FM is under the control of genetic regulators, key among which are members of MADS box transcription factor family. Their individual and combined act confers distinct identities to floral organs. Grass flowers are highly modified in structure. Rice flower, a model for grasses, is borne on a short branch called spikelet and they together from the basic structural units of the rice infloresences known as panicle. The outer whorl organs of a grass floret are bract-like structures known as lemma and palea to dicot sepals is highly dibated (see Chapter 1). In grass florets, petal homologs are a pair of highly reduced, fleshy bracts known as lodicules, while stamen and carpel homologs occupy the same position and share the same functions as their dicot counterparts. Aside from these distinct outer whorl organs, the florets are subtended by two pairs of bracts known as empty glumes and rudimentary glumes. The genetic regulators that control their unique identities and those that perform conserved functions are very intriguing and central questions in plant developmental biology. Using various contemporary and complementary technologies, we have analysed the molecular functions and downstream pathways of a MADS box transcription factor, OsMADSI during the rice floret meristem specification and organ development. Further by reverse genetics and overexpression studies, we have also functionally characterized two target genes of OsMADSI, OsETTINI and OsETTINI2 to understand their roles downstream to OsMADSI during the rice floret development.

碩士<br>國立中興大學<br>分子生物學研究所<br>101<br>M52048 is a rice mutant with T-DNA insertion which activated flanking genes OsMADS14, OsMADS34 and OsCP7, resulting in darwf, early flowering, node bending and impaired panicle exertion. OsMADS14 and OsMADS34 belong to MADS-box gene family, can regulate the flowering time and the floral meristem development. OsMADS14 regulate the flowering time and has been well studies. OsMADS34 has been known to regulate the floral meristem development and panicle branching, OsCP7 encodes a putative cysteine protease, but the function by overexpressing OsMADS34 and OsCP7 genes remained unknown. The OsMADS34 overexpressing transgenic rice, Ubi:OsMADS34, revealed early flowering and impaired in panicle exertion. Morphological dissection indicated that the impaired panicle exertion was caused by shortening the first and second internodes. Previous study showed drought stress inhibited panicle exertion by reducing the peduncle elongation during flowering and induced the accumulation of ABA. The present study showed some dorught-related genes and miRNAs were regulated in Ubi:OsMADS34, and therefore, hypothesized that the phenotype of Ubi:OsMADS34 may regulated by drought stress or plant hormones. However, ABA inhibitor and GA treatments in this study could not improve the first internode elongation and panicle exertion. Gene expressinon analysis indicated the GA-regulated gene, EATB, was up-regulated and cell elongation promoting genes, XTH19 and XTH28, were reduced in the first and second internodes of Ubi:OsMADS34, suggesting that the limited internode elongation may be due to the differential expression of these internode regulating genes. However, how these genes were regulated remains further elucidation. OsCP7, like a typical cysteine protease, consist of a signal peptide, a prodomain and a protease domain, expressed in vegetative tissues but not in reproductive tissues. No OsCP7 overexpressing transgenic rice driven by ubiquitin or 35S promoters was stably obtained, though few of them could survive in the medium. Instead of these consititutive promoters, a 1.6 kb promoter fragment of OsCP7 could drive the expression of OsCP7 in rice successfully. This 1.6 kb OsCP7 promoter-driven transgenic rice, OsCP7:OsCP7, showed slightly dawrf, delayed flowering, low pollen viabillity, lower fertility seeds, and lesion-like brown spots appeared on spikelet. The lesion-like spots were first occuerd on the outer surface of spikelet when exposed to sunlight after heading. The spots could spread quickly in days and led to extensive cell death with ROS accumulation. However, using RNAi approach to knockdown the expression of OsCP7 did not show any specific phenotype. Further investigation to understand the molecular mechanism of the formation of the lesion-like spot is underway.

Rice have highly derived florets borne on a short branch called ‘spikelet’ comprised of a pair of rudimentary glumes and sterile lemma (empty glumes) that subtends a single fertile floret. The floral organs consist of a pair of lodicules, six stamens and a central carpel that are enclosed by a pair of bract-like organs, called lemma and palea. A progressive reprogramming of meristem identity during the floral development of flowers, on branches on the inflorescence, is correlated with changes in transcriptional status of regulatory genes that execute cascades of distinct developmental events. On the other hand phytohormones such as auxin and cytokinin that are critical in predetermining the sites of new organ primordia emergence and in maintaining the size or populations of meristems. Molecular genetic analyses of mutants have expanded the repository of genes regulating floral organ specification and identity, yet the finer mechanistic details on process downstream to these regulatory genes and co-ordination with phytohormone signalling pathways needs further investigation. One aim of the study presented in this thesis is to develop a tool that would display of spatial description of dynamic auxin or cytokinin accumulation in developing rice inflorescence and floral meristems and to evaluate auxin distribution defects of OsMADS1-RNAi florets using this tool. Additionally, we aim to understand the regulatory effects on OsMADS1 on candidate floral organ and meristem fate determining genes during two temporal phases of flower development to decipher other regulatory cascades controlled by OsMADS1. Spatial distribution profile of phytohormones in young and developing meristems of rice Cytokinin promotes meristem activity (Su et al., 2011) while auxin accumulation, directed by auxin efflux transport PIN proteins predicts sites of new organ initiation (Reinhardt et al., 2003; van Mourik et al., 2012). Previous studies in the lab deciphered that OsMADS1 exerts positive regulatory effects on genes in auxin pathways and repressive effects on cytokinin signaling and biosynthetic genes (Khanday et al., 2013). Thus, the need for a reliable system to understand auxin and cytokinin activity in live inflorescence and floral meristems of rice motivated us to raise promoter: reporter tools to map the spatial and temporal phytohormone distribution. Confocal live imaging conditions in primary roots of IR4DR-GFP and DR5-CyPet lines was performed and responsiveness of the DR5 elements to auxin was authenticated. Auxin maxima were distinctly seen in the epidermal and sub-epidermal cells of inflorescence branch primordia anlagen and apices of newly emerged branch primordia. As floral organs were being initiated, on the floret meristem, we discerned the sequential appearance of auxin accumulation at sites of organ primordia while apices of early floral meristems (FM) showed low auxin content. We clearly detect canalization of auxin streams marking regions of vascular inception. Using this live imaging system we probed auxin patterns and levels in malformed and indeterminate OsMADS1-RNAi florets and we observed a significant reduction in the levels of auxin. Two oppositely positioned peaks of auxin were noted in the persistent FM of OsMADS1-RNAi florets, a pattern similar to auxin dynamics at sites of rudimentary glume primordia on the wild-type (WT) spikelet meristem. These studies were followed up with immunohistochemistry (IHC) on fixed tissues for “PIN” transport proteins that suggest PIN convergence towards organ initiation sites, regions where auxin accumulation was clearly visualized by the IR4DR5-GFP and DR5-CyPet reporters. IHC experiments that detected GFP, in fixed tissues of TCSn-mGFP ER (WT) and TCSn-mGFP ER;OsMADS1-RNAi (OsMADS1-RNAi) inflorescence and florets showed an ectopic increase in the domain of cells with cytokinin response in OsMADS1-RNAi florets, compared to that of WT. Intriguingly, cytokinin responsive cells persisted in the central FM of OsMADS1-RNAi florets that might partially account for some of the FM indeterminacy defects seen in these florets. A correlative observation of these different imaging data hint at some exclusive patterns of the IR4DR5/DR5 and TCSn reporters that in turn lead us to speculate that a cross talk between auxin and cytokinin distribution may contribute to the precise phyllotaxy of lateral organs in rice inflorescence. Studies on novel targets of OsMADS1 in floral organ identity and meristem determinacy Loss of OsMADS1 function results in rice florets with miss specified floral organs and an indeterminate carpel produces new abnormal florets. Despite having several mutants in OsMADS1, mechanisms of how OsMADS1 regulates meristem maintenance and termination is not well understood. Global expression profile in OsMADS1-RNAi vs. WT tissues encompassing a wide range of developing florets (0.2 to 2cm panicles), gave an overview of OsMADS1 functions in many aspects of floret development. Here, a gene-targeted knockout of OsMADS1 named - osmads1ko (generated in a collaborative study) was characterized and found to display extreme defects in floral organs and an indeterminate FM. Strikingly, in addition to loss of determinacy, FM reverts to a prior developmental fate of inflorescence on whose new rachis are leaf-like malformed florets. We suggest these phenotypes reflect the null phenotype of OsMADS1 and its role in meristem fate maintenance. We tested gene expression levels for some proven targets of OsMADS1 (Khanday et al., 2013) and utilized panicles in two developmental phases- young early FMs (panicles of 0.2 to 0.5 cm) and older florets with organ differentiation (panicles of 0.5 to 1cm). We observed temporally different effects on the regulation of OsMADS34 that together with histology of young osmads1ko inflorescences suggest that the mutant is impeded for spikelet to floral meristem transition. In addition, OsMADS1 had a positive regulatory effect on genes implicated for lemma and palea organ identity such as OsIDS1, OsDH1, OsYABBY1, OsMADS15, OsMADS32, OsDP1 and OsSPL16 in both young and old panicles while OsIG1 was negatively regulated in both phases of development. MADS-box genes important for carpel and ovule development - OsMADS13 and OsMADS58 were had significantly reduced expression in florets undergoing organ differentiation. OsMADS1 positively regulated several other non MADS-box developmental genes - OsSPT, OsHEC2 and OsULT1, whose Arabidopsis homologs control carpel development and FM determinacy. These genes are de-regulated in later stages of osmads1ko floret development and are unaffected in younger panicles. Finally, OsMADS1 continually activated meristem maintenance genes - OsBAM2-like and OsMADS6 while the activation of OSH1 in early floral meristems was later altered to a repressive effect in developing florets. Perhaps such dynamic temporal effects on meristem genes are instrumental in the timely termination of the floral meristem after floral organ differentiation. More importantly, we show that regulation of many of these genes is directly affected by OsMADS1, through our studies on expression levels before and after chemical induction of OsMADS1-GR protein in amiRNAOsMADS1 florets. Further, some key downstream targets were re-affirmed by studying expression status in transgenic lines, with the OsMADS1-EAR repressive protein variant. These results provide new insights into the developmentally phased roles of OsMADS1 on floral meristem regulators and determinants of organ identity to form a determinate rice floret. Gene networks regulated by OsMADS1 during early flower development To identify global targets in early floret meristems, we determined the differential RNA transcriptome in osmads1ko tissues as compared to wild-type tissues. These data revealed regulators of inflorescence architecture, floral organ identity including MADS-box floral homeotic factors, factors for meristem maintenance, auxin response, transport and biosynthesis as some of the important functional classes amongst the 2725 differentially expressed genes (DEGs). Integrating DEGs with OsMADS1 ChIP-seq data (prior studies from our lab) we deciphered direct vs. indirect and positive vs. negatively regulated targets of OsMADS1. These datasets reveal an enrichment for functional categories such as metabolic processes, signaling, RNA transcription and processing, hormone metabolism and protein modification. Using Bio-Tapestry plot as a tool we present a visualization of a floral stage-specific regulatory network for genes with likely functional roles in meristem specification and in organ development. Further, to examine if indirect targets regulated by OsMADS1 could be mediated through transcription factors (that are themselves direct targets), we constructed a small network with the transcription factors OSH1, OSH15 and OsYABBY1 as key nodal genes and we predicted their downstream effects. Taken together, these analyses provide examples of the complex networks that OsMADS1 controls during the process of rice floret development. In summary, we surmise that defect in phytohormone distribution in OsMADS1 knockdown florets results in irregular patterns of lateral organ primordia emergence. In addition, the derangements in the developmentally stage specific expression of floral meristems identity and organ identity genes culminates in miss-specified and irregularly patterned abnormal organs in Osmads1 florets. Thus, our study highlights the versatility of OsMADS1 in regulating components of hormone signaling and response, and its effects on various floral development regulators results in the formation of a single determinate floret on the spikelet. References: Khanday I, Yadav S.R, and Vijayraghavan U. (2013). Plant Physiol 161, 1970–1983. van Mourik S , Kaufmann K, van Dijk AD, Angenent G.C, Merks R.M.H, Molenaar J. (2012). PLOS One 1, e28762 Reinhardt D, Pesce E, Stieger P, Mandel T, Baltensperger K, Bennett M, Traas J, Friml J and Kuhlemeier C. (2003). Nature 426, 255-260 Su Y, Liu Y and Zhang X. (2011) Mol Plant 4, 616–625

Rice have highly derived florets borne on a short branch called ‘spikelet’ comprised of a pair of rudimentary glumes and sterile lemma (empty glumes) that subtends a single fertile floret. The floral organs consist of a pair of lodicules, six stamens and a central carpel that are enclosed by a pair of bract-like organs, called lemma and palea. A progressive reprogramming of meristem identity during the floral development of flowers, on branches on the inflorescence, is correlated with changes in transcriptional status of regulatory genes that execute cascades of distinct developmental events. On the other hand phytohormones such as auxin and cytokinin that are critical in predetermining the sites of new organ primordia emergence and in maintaining the size or populations of meristems. Molecular genetic analyses of mutants have expanded the repository of genes regulating floral organ specification and identity, yet the finer mechanistic details on process downstream to these regulatory genes and co-ordination with phytohormone signalling pathways needs further investigation. One aim of the study presented in this thesis is to develop a tool that would display of spatial description of dynamic auxin or cytokinin accumulation in developing rice inflorescence and floral meristems and to evaluate auxin distribution defects of OsMADS1-RNAi florets using this tool. Additionally, we aim to understand the regulatory effects on OsMADS1 on candidate floral organ and meristem fate determining genes during two temporal phases of flower development to decipher other regulatory cascades controlled by OsMADS1. Spatial distribution profile of phytohormones in young and developing meristems of rice Cytokinin promotes meristem activity (Su et al., 2011) while auxin accumulation, directed by auxin efflux transport PIN proteins predicts sites of new organ initiation (Reinhardt et al., 2003; van Mourik et al., 2012). Previous studies in the lab deciphered that OsMADS1 exerts positive regulatory effects on genes in auxin pathways and repressive effects on cytokinin signaling and biosynthetic genes (Khanday et al., 2013). Thus, the need for a reliable system to understand auxin and cytokinin activity in live inflorescence and floral meristems of rice motivated us to raise promoter: reporter tools to map the spatial and temporal phytohormone distribution. Confocal live imaging conditions in primary roots of IR4DR-GFP and DR5-CyPet lines was performed and responsiveness of the DR5 elements to auxin was authenticated. Auxin maxima were distinctly seen in the epidermal and sub-epidermal cells of inflorescence branch primordia anlagen and apices of newly emerged branch primordia. As floral organs were being initiated, on the floret meristem, we discerned the sequential appearance of auxin accumulation at sites of organ primordia while apices of early floral meristems (FM) showed low auxin content. We clearly detect canalization of auxin streams marking regions of vascular inception. Using this live imaging system we probed auxin patterns and levels in malformed and indeterminate OsMADS1-RNAi florets and we observed a significant reduction in the levels of auxin. Two oppositely positioned peaks of auxin were noted in the persistent FM of OsMADS1-RNAi florets, a pattern similar to auxin dynamics at sites of rudimentary glume primordia on the wild-type (WT) spikelet meristem. These studies were followed up with immunohistochemistry (IHC) on fixed tissues for “PIN” transport proteins that suggest PIN convergence towards organ initiation sites, regions where auxin accumulation was clearly visualized by the IR4DR5-GFP and DR5-CyPet reporters. IHC experiments that detected GFP, in fixed tissues of TCSn-mGFP ER (WT) and TCSn-mGFP ER;OsMADS1-RNAi (OsMADS1-RNAi) inflorescence and florets showed an ectopic increase in the domain of cells with cytokinin response in OsMADS1-RNAi florets, compared to that of WT. Intriguingly, cytokinin responsive cells persisted in the central FM of OsMADS1-RNAi florets that might partially account for some of the FM indeterminacy defects seen in these florets. A correlative observation of these different imaging data hint at some exclusive patterns of the IR4DR5/DR5 and TCSn reporters that in turn lead us to speculate that a cross talk between auxin and cytokinin distribution may contribute to the precise phyllotaxy of lateral organs in rice inflorescence. Studies on novel targets of OsMADS1 in floral organ identity and meristem determinacy Loss of OsMADS1 function results in rice florets with miss specified floral organs and an indeterminate carpel produces new abnormal florets. Despite having several mutants in OsMADS1, mechanisms of how OsMADS1 regulates meristem maintenance and termination is not well understood. Global expression profile in OsMADS1-RNAi vs. WT tissues encompassing a wide range of developing florets (0.2 to 2cm panicles), gave an overview of OsMADS1 functions in many aspects of floret development. Here, a gene-targeted knockout of OsMADS1 named - osmads1ko (generated in a collaborative study) was characterized and found to display extreme defects in floral organs and an indeterminate FM. Strikingly, in addition to loss of determinacy, FM reverts to a prior developmental fate of inflorescence on whose new rachis are leaf-like malformed florets. We suggest these phenotypes reflect the null phenotype of OsMADS1 and its role in meristem fate maintenance. We tested gene expression levels for some proven targets of OsMADS1 (Khanday et al., 2013) and utilized panicles in two developmental phases- young early FMs (panicles of 0.2 to 0.5 cm) and older florets with organ differentiation (panicles of 0.5 to 1cm). We observed temporally different effects on the regulation of OsMADS34 that together with histology of young osmads1ko inflorescences suggest that the mutant is impeded for spikelet to floral meristem transition. In addition, OsMADS1 had a positive regulatory effect on genes implicated for lemma and palea organ identity such as OsIDS1, OsDH1, OsYABBY1, OsMADS15, OsMADS32, OsDP1 and OsSPL16 in both young and old panicles while OsIG1 was negatively regulated in both phases of development. MADS-box genes important for carpel and ovule development - OsMADS13 and OsMADS58 were had significantly reduced expression in florets undergoing organ differentiation. OsMADS1 positively regulated several other non MADS-box developmental genes - OsSPT, OsHEC2 and OsULT1, whose Arabidopsis homologs control carpel development and FM determinacy. These genes are de-regulated in later stages of osmads1ko floret development and are unaffected in younger panicles. Finally, OsMADS1 continually activated meristem maintenance genes - OsBAM2-like and OsMADS6 while the activation of OSH1 in early floral meristems was later altered to a repressive effect in developing florets. Perhaps such dynamic temporal effects on meristem genes are instrumental in the timely termination of the floral meristem after floral organ differentiation. More importantly, we show that regulation of many of these genes is directly affected by OsMADS1, through our studies on expression levels before and after chemical induction of OsMADS1-GR protein in amiRNAOsMADS1 florets. Further, some key downstream targets were re-affirmed by studying expression status in transgenic lines, with the OsMADS1-EAR repressive protein variant. These results provide new insights into the developmentally phased roles of OsMADS1 on floral meristem regulators and determinants of organ identity to form a determinate rice floret. Gene networks regulated by OsMADS1 during early flower development To identify global targets in early floret meristems, we determined the differential RNA transcriptome in osmads1ko tissues as compared to wild-type tissues. These data revealed regulators of inflorescence architecture, floral organ identity including MADS-box floral homeotic factors, factors for meristem maintenance, auxin response, transport and biosynthesis as some of the important functional classes amongst the 2725 differentially expressed genes (DEGs). Integrating DEGs with OsMADS1 ChIP-seq data (prior studies from our lab) we deciphered direct vs. indirect and positive vs. negatively regulated targets of OsMADS1. These datasets reveal an enrichment for functional categories such as metabolic processes, signaling, RNA transcription and processing, hormone metabolism and protein modification. Using Bio-Tapestry plot as a tool we present a visualization of a floral stage-specific regulatory network for genes with likely functional roles in meristem specification and in organ development. Further, to examine if indirect targets regulated by OsMADS1 could be mediated through transcription factors (that are themselves direct targets), we constructed a small network with the transcription factors OSH1, OSH15 and OsYABBY1 as key nodal genes and we predicted their downstream effects. Taken together, these analyses provide examples of the complex networks that OsMADS1 controls during the process of rice floret development. In summary, we surmise that defect in phytohormone distribution in OsMADS1 knockdown florets results in irregular patterns of lateral organ primordia emergence. In addition, the derangements in the developmentally stage specific expression of floral meristems identity and organ identity genes culminates in miss-specified and irregularly patterned abnormal organs in Osmads1 florets. Thus, our study highlights the versatility of OsMADS1 in regulating components of hormone signaling and response, and its effects on various floral development regulators results in the formation of a single determinate floret on the spikelet. References: Khanday I, Yadav S.R, and Vijayraghavan U. (2013). Plant Physiol 161, 1970–1983. van Mourik S , Kaufmann K, van Dijk AD, Angenent G.C, Merks R.M.H, Molenaar J. (2012). PLOS One 1, e28762 Reinhardt D, Pesce E, Stieger P, Mandel T, Baltensperger K, Bennett M, Traas J, Friml J and Kuhlemeier C. (2003). Nature 426, 255-260 Su Y, Liu Y and Zhang X. (2011) Mol Plant 4, 616–625

Studies on meristem identity regulators in Rice (Oryza sativa), a model plant for cereal crops have revealed how meristem identity and transitions are controlled, bearing implications for crop yield improvement. We were interested in exploring the function of OsULTRAPETALA1 (OsULT1), whose Arabidopsis homolog is a TrxG Factor, in rice inflorescence and spikelet/floret organ development. OsULT1 is a direct downstream target of floret meristem identity and development transcriptional regulator OsMADS1 (Khanday et al., 2016). The aim of this thesis is to functionally characterize OsULT1 by raising transgenic rice plants with ubiquitous knockdown and other lines that are overexpressors of OsULT1. In the dsRNAi-OsULT1 knockdown transgenics, we observed reduced plant height, panicle branching, and delay in flowering time. Interestingly, the pUbi-OsULT1 overexpression transgenics showed a converse phenotype of precocious flowering. Histological studies on young branching inflorescence meristem from the dsRNAi-OsULT1 knockdown transgenics and wild-type plants, was done to understand the onset of developmental abnormalities during panicle development. We observed an increase in the number of lateral organs made from the spikelet meristem; this could be due to a delay in SM to FM transition. These data suggest OsULT1 to be a heterochronic factor regulating meristem progression. Histological analyses of young spikelets showed a homeotic conversion of sterile lemma to a lemma-like organ. In floral meristem of knockdown plants, reduced palea and altered stamen number were noted. We quantified expression levels of some selected well-studied spikelet meristem regulators. qRT-PCR done on RNA from pooled dsRNAi-OsULT1 panicle tissues (Early-stage: 0.1 to 0.3cm and Late-stage: 0.4 to 2cm) and compared to similarly staged wild-type panicles. We found OsMADS1 and OsIDS1 transcript levels to be up-regulated and down-regulated respectively in the dsRNAi-OsULT1 transgenics compared to the wild-type. This could relate to the sterile lemma and rudimentary glume phenotypes observed in the affected knockdown spikelets. Since chromatin modifiers lacking DNA-binding domain recruit TF’s to target genes, we tested the possibility of protein interaction between OsULT1 and OsMADS1 using the yeast two-hybrid assay. This assay confirmed the interaction between OsULT1 and OsMADS1, thus raising the prospect of such a TF–Chromatin factor complex regulating downstream target gene expression by modulation of the histone modification status of the gene loci relevant for SM and FM development. We surveyed the abundance of repressive and activating histone marks in two developmental stages in wild-type panicle tissues as an attempt to correlate histone marks with transcript abundance. We observed as expected an inverse correlation between the expression levels of OsMADS34 and OsMADS22 and the abundance of the H3K27me3 mark at these two loci. We then evaluated the chromatin status at the genes which are differentially expressed in the dsRNAi-OsULT transgenics like OsMADS1, OsIG1, and OsIDS1. The results of the ChIP-qPCR analysis to assess the abundance of histone marks indicate a complex relationship between chromatin marks and transcript abundance. For future in-depth studies of gene targets in specific stages of SM and FM, we have standardized Laser Capture Micro-dissection of specific rice wild-type panicle meristems for transcriptomic studies in these recessed difficult to access tissues. Preliminary data indicate sets of transcripts that could be specific to Primary Branch Meristem (PBM), Secondary Branch Meristem (SBM), and Floral Meristem (FM). Overall, we have used reverse genetics tools to elucidate the functions of a predicted Trithorax-Group factor OsULT1 in spikelet meristem transient maintenance, its lateral organ development, and effects on floret organ numbers. The implications of the studies support the published hypothesis that ancestral rice species had three-fertile floret per spikelet (Ren et al., 2020; Zhang et al., 2017), instead of the one floret per spikelet in seen in extant species. The work in this thesis highlights the important role of chromatin modifiers like the Trithorax factors in rice panicle development.

碩士<br>國立中興大學<br>分子生物學研究所<br>102<br>The flowers of Phalaenopsis orchid have beautiful shape, brightly colors with long flowering period, are mostly used for commercial purposes, such as ornamental layout. Due to the semi-tropical location and climate condition, Taiwan becomes the most appropriate location for planting orchids and this makes orchids the most important export flowers in Taiwan, and consequently makes Taiwan as the "Orchid Kingdom" in the world. The diversities of orchid were mainly crossbred in the past; however many studies using transgenic or mutagenesis methods to change the color and/or appearance of the orchid morphology have been reported recently. Among them, few were used to change the plant architectures. In order to minimize and create novel orchid architectures, a rice OsGA2ox6 gene, involved in GA biosynthesis and when it over-expressed it could produce semi-dwarf plants with multiple shoots and roots, was used in this study. In addition to the long flowering period, the Phalaenopsis orchids have at least 2 to 3 year of growth period to reach the flowering stage, therefore more planting costs were needed. In order to shorten the time need before flowering, a MADS-box gene, MADS14, involved in the regulation of floral initiation and early flowering, was used under the control of a ubiquitin promoter as well. In this study, totally 6550 protocorm-like bodies (PLBs) were transformed by pCAM1301、 pCAM1301-Ubi-GA2ox6 and pCAM1301-Ubi-MADS14 vectors respectively. After hygromycin selection, three regenerated orchid plants from pCAM1301-Ubi-GA2ox6 were obtained and named as GA-1、GA-2 and GA-3, respectively. Comparing to the non-transgenic orchid, these regenerated orchids show dwarfism, multiple shoots and roots, and have smaller and dark-green leaf. All three regenerated orchids show GUS activity in examined leaf and root tissues. Genomic PCR analysis demonstrated that these three regenerated orchids contain the GUS and hptII genes and the GA2ox6 transgene as well. RNA expressions of the GA2ox6 transgene in these three regenerated orchids were detected by RT-PCR and showed the same expression levels among them. Southern blot assay showed these three regenerated orchids were independent transgenic orchid lines with one copy of T-DNA insertion in both GA-1 and GA-3, and two copies of T-DNA insertion in GA-2 line.

Title: Investigating Partners of OsMADS1 Transcription Factor and functions for some associated factors for roles in rice inflorescence and floral development The classic ABCDE model for floral organ identity show that the combinatorial action of MADS domain transcription factors directs floral organ identity and their normal development. While floral MADS-box transcription factors are in general conserved among diverse flowering plants, several studies suggest emergence of distinct functions for some of these conserved factors. Rice E- or SEP-class genes, when compared to Arabidopsis SEP-class, are only partly conserved with rice homologous genes which show some distinct functional diversifications. Rice genes of this class form two clades: LOFSEP- (OsMADS1, OsMADS34, OsMADS5) and SEP3 (OsMADS7, OsMADS8) clades. Unlike Arabidopsis SEP-class genes that are redundant, just loss of function of OsMADS1 alone leads to severe defects in the development of all floral organs. Extensive molecular genetic studies, including from our laboratory, showed OsMADS1 is a master transcription regulator for floral meristem determinacy, floral organ specification and identity. The gene regulatory network controlled by OsMADS1 including other transcription factor targets, and phytohormone auxin and cytokinin pathways. Here we probed for the interacting partners of OsMADS1 to get a better understanding of its function in higher protein complexes during the broad program of floral development. In addition, we took up functional genomic studies of OsMADS1 associated factors for studies on roles for downstream factors controlling rice inflorescence and floral development. Part I: Investigating interacting partners of OsMADS1 transcription factor While interactions between MADS transcription factors are well described, interactions with other classes of factors that can influence its regulatory functions in higher protein complexes are not studied yet. We adopted two methods to identify OsMADS1 interacting factors. First, we predicted interactors of OsMADS1 based on the previous report from our group that identified the co-occurrence of cis DNA binding motifs in the genome-wide OsMADS1ChIP-Seq dataset (Khanday et al., 2016). This was combined with co-expression analysis of transcription factors that could bind these cis-elements based using publicly available rice transcriptome databases. A small set of putative deduced from the above approaches were examined by yeast-two hybrid (Y2H) assays and the interactions noted were further validated by the in-planta Bimolecular fluorescence complementation (BiFC) assay. Y2H assay suggests OsbZIP47 and ERF68 are strong interactors of OsMADS1. Further, we note RFL as a weak interactor, whereas the eudicot Arabidopsis homologs LEAFY and SEP3 have a strong association. In the second approach, we carried out pulldown of endogenous OsMADS1 associated protein complexes from maturing 3-5cm inflorescence tissue lysates using OsMADS1specific antibodies. By mass spectroscopy analysis we identify several classes of proteins that demonstrate association of OsMADS1 function with transcriptional corepressors, meristem transition regulators of floral meristem determinacy, organ identity and also indicate interactions with factors in auxin and Brassinosteroid signaling pathways. Part II: Functional studies on the predicted meristem regulator OsbZIP47, a downstream target and a partner of OsMADS1 OsbZIP47 is a predicted rice homolog for the Arabidopsis and maize factors, PERIANTHIA (PAN) and FASCIATED EAR4 (FEA4) respectively that regulate meristem and floral organs in those model species. Here we aimed to understand functional relevance and pathways controlled by OsbZIP47 for normal rice inflorescence development. Transgenic plants with RNAi based knockdown of OsbZIP47 were raised and they showed pleiotropic effects on both vegetative and reproductive phases of growth including meristem and floral organ defects. Whereas transgenics with overexpression (Ox) of OsbZIP47 did not show any phenotypic defects. Importantly, similar to the PAN protein, we observe that oligomerization of the OsbZIP47 protein is sensitive to redox reagents (diamide, S-glutathionylation modifier and DTT). We also detect robust transcriptional activity for a reporter gene in yeast despite the fact that OsbZIP47 lacks the extended N-ter protein domain seen in PAN which is essential for PAN functions in Arabidopsis. We also show that OsMADS1 can positively regulate expression levels of OsbZIP47 and that of the gene encoding its potential glutaredoxin enzyme, OsGRX19 or MICROSPORELESS1 (OsMIL1). This co-ordinated effect may be to maintain homeostasis and enforce their interaction. To reveal OsbZIP47 downstream pathways, global transcriptome profiling (RNA-Seq) of 0.1-0.5cm panicle tissues from OsbZIP47 KD line was compared with similarly staged wild-type tissues and the data suggest ~2800 deregulated genes. Most important inference is that OsbZIP47 regulates meristem characteristics by controlling the canonical WUS-CLV pathway for stem cell homeostasis and the cytokinin-mediated WUS-KNOX (OSH1/STM/KN1) pathway. Floral defects (particularly of stamens and lodicules) seen in OsbZIP47 KD lines can to an extent be explained by the deregulation of OsMADS16 (B-class gene), OsDL (DROOPING LEAF), APO1, OsMADS68, TGA10, Osnop (Oryza sativa no pollen) that are known regulators for proper development of these organs. Taking leads from Arabidopsis and Maize reports we also tested interaction of OsbZIP47 with auxin response factors ETTIN1 and ETTIN 2, and with meristem regulators, KNOXI-OSH1, RFL using the Y2H assay. We report strong interaction of OsbZIP47 with OSH1 and RFL which further add a layer of complexity in molecular mechanisms by which OsbZIP47 can contribute to meristem characteristics. Part III: B-class OsMADS2 gene, a partner and a downstream factor of OsMADS1 controls gene expression for normal lodicule and stamen development In rice, and all grass florets, lodicules are modified organs equivalent to petal and they have a crucial role for efficient fertilization. Rice floret stamens have functions and structure with greater similarity to their eudicot counterparts. OsMADS1 regulates the normal development and identity of all floral organs; its partnership with B-class factors AP3-like OsMADS16, PI-like OsMADS2, and OsMADS4 is known and is critical for lodicule and stamen developmental control. Noteworthy is that B-class genes are expressed in the lodicule and stamen floral primordia but OsMADS1 transcripts are not detected in these primordia. Prior work from our group showed that OsMADS2, a PI-like factor has evolved to have a greater role in lodicule identity and lesser redundant role for stamen identity. Here, using immunolocalization we detect the presence of OsMADS1 protein in lodicule and stamen primordia, hinting at possibility of either intercellular trafficking or other non-cell-autonomous manner of OsMADS1 action in lodicule and stamen primordia. Further, to delineate gene sets that are coregulated by OsMADS1 and OsMADS2 we determined OsMADS2 genome-wide occupancy by ChIP-Seq using 0.3-2cm panicle tissues and OsMADS2 affinity-purified antibodies. After comparing data acquired in this study with the OsMADS1ChIP-Seq dataset (Khanday et al., 2016) we identify a common set of 280 gene targets. Also, OsMADS2 ChIP-Seq shows that homologs of Arabidopsis AP1 and AP2 class genes are enriched. Thus, as known in Arabidopsis the latter factors can explain OsMADS2 functions in lodicule development. Interestingly, OsMADS2 ChIP-PCR and qRT-PCR transcript measurements in transgenic with partial knockdown of OsMADS2 show that OsMADS1 and OsMADS2 are linked by a positive feedback loop. We confirm HECATE/OsbHLH120 as a common target of OsMADS1 (Dr. Grace Lhainekim Thesis, UVR lab) and OsMADS2 (this study). The homolog of OsbHLH120 in Arabidopsis is a meristem and carpel development regulator. Here we perform functional characterization of OsbHLH120 by RNA interference-mediated (dsRNAi) knockdown (KD). The transgenics had majority of their florets with elongated and deformed lodicules, and a minority showed increased stigma/pistil number. The spikelet organ called the sterile lemma was elongated resembling a lemma-like organ. These phenotypes suggest OsbHLH120 regulates spikelet and floral organ development.<br>IISc

碩士<br>國立中興大學<br>生命科學系所<br>104<br>The early flowering phenotype of a rice T-DNA insertion mutant M52048 had been demonstrated due to the activation of OsMADS14, an Arabidopsis AP1 homologous gene involved in flowering regulation. Overexpressing OsMADS14 with ubiquitin promoter construct Ubi:OsMADS14 in TNG67 revealed early flowering as observed in M52048. In addition, the early expression of two floregin genes Hd3a and RFT1 were observed and enhanced in Ubi:OsMADS14 transgenic rice. However, how the activation and overexpression of OsMADS14 in rice could result in early flowering as AP1 did in the Arabidopsis have not been characterized. The present study aims to understand the molecular mechanisms of early flowering that regulated by over-expressing OsMADS14. The rice homologous to Arabidopsis AP1-regulated genes that lead to early flowering and dwarf, such as floral repressors, RCN 1 (homologous to TFL1), OsTEM1-like (TEM1), OsTOE1-like (TOE1); cytokinin biosynthesis genes, such as OsLOG, OsLOG1-like, OsLOG2-like, OsLOG3-like, OsCKX4, OsCKX5; and GA biosynthesis related genes, such as OsEUI1 and OsGA13ox1, were identified and investigated. Results showed Hd3a and RFT1 expressed much earlier in Ubi:OsMADS14 transgenic rice, and OsLOG and OsCKX5 response similarly to their homologous genes LOG1 and CKX3 in Arabidopsis, which may reduce axillary meristem growth and promote flowering. Surprisingly, the flowering suppressor gene RCN1 was enhanced, response differently from that in Arabidopsis. Therefore, the role of RCN1 in rice required further investigation. Regarding to the GA biosynthesis-related genes, the expression level of OsEUI1and OsGA13ox1 were enhaced that could possibly explain the dwarf phenotype of Ubi:OsMADS14 transgenic rice. The function of rice genes LOC_Os11g05470 (RCN1); LOC_Os05g03040 (OsTOE1-Like); LOC_Os01g49830 (OsTEM1-Like) homologous to Arabidopsis flowering suppressor genes AtTFL1, AtTOE1 and AtTEM1 were investigated. Two T-DNA insertion mutants, M78020 that may activate RCN1 and M89461 that may activate OsTEM1-like gene, were obtained from TRIM database for characterization. The insertion events of these two mutants were confirmed, and the mutant M89461 presente a late flowering phenotype, however mutant M78020 display no phenotype and no different in heading date to that of WT control. The links between the target genes expression and heading date phenotypes are under investigation. Establishing an alcohol inducible transgenic platform is another project of this study. This alcohol inducible system contains an alcohol-sensitive transcription factor AlcR driven by 35S promoter, and a second transcription unit contains AlcA promoter-driven target gene that can be regulated by alcohol-bound AlcR transcription factor. This alcohol inducible system using herbicide resistant bar gene as selection marker which is different from the hygromycin selection system we used routinely. Therefore, the optimization of selection condition was performed. Although several regenerated rice plant were obtained, none of them contain the transgene. In order to set up the alcohol inducible transgenic platform, replacing herbicide with hygromycin selection is suggested.