Functional characterization of giant killer in flower development and meristem regulation in arabidopsis thaliana 2

For instance, study of floral organ identity genes in Arabidopsis thaliana has expedited the process of identifying their conserved counterparts in Dendrobium crumenatum orchid, an ornam

CHAPTER 1: Introduction

1.1 General Introduction

In the course of human civilization, we are constantly exploring ways to harness and domesticate various plant species for our use as food sources, raw materials and bio-energy In view of our current pressing issues on global warming and climate changes, extensive understanding of our sessile co-habitant of Mother Earth has become

increasingly crucial in the development of a sustainable environment to live in

There are plenty of examples showing that plant research has significant impact on

humankind Research studies on our staple crops like oryza sativa (rice) have

contributed in reducing poverty and hunger (beta.irri.org/news/) Since the imminent threat on fossil fuels depletion would have immediate consequences on global

economic and political stability, scientific studies on Elaeis guineensis (oil palm) and Jatropha curcas (Barbados nut) for biodiesel development will certainly help in

soothing the jitters triggered by the global oil crisis (Fairless, 2007; Kennedy, 2007)

In general, scientific findings generated from the plant research will ultimately help in transforming human society and will provide numerous incentives for the future development of the mankind Nevertheless, scientific research that uses economic or agricultural plant species often has a murky outlook It is commonly suffering from various technical glitches, which may be caused by the species undefined genome sequences, their long generation time, and their intractability to molecular and genetic

approaches In light of this, extended and detailed analysis on a model plant like

Arabidopsis thaliana would invariably provide useful insights and lay down a

concrete foundation for the understanding of other plant species For instance, study

of floral organ identity genes in Arabidopsis thaliana has expedited the process of identifying their conserved counterparts in Dendrobium crumenatum (orchid), an ornamental plant that has significant economic implications (Xu et al., 2006)

Arabidopsis thaliana is a small-size flowering plant (also known as mouse-ear cress)

belonging to the mustard family It is amenable to various molecular and genetic manipulations for scientific research (Meyerowitz and Pruitt, 1985) Its genome had been fully sequenced in year 2000 (Arabidopsis Genome Initiative, 2000) and

documented to be at a relatively small size of 120 Mb with about 25, 000 genes sprawling across 5 chromosomes

More than two decades ago, several phenotypically striking floral homeotic mutants

had been identified by Meyerowitz and colleagues (Bowman et al., 1989; Yanofsky et al., 1990; Bowman et al., 1991) Detailed characterization and analysis of these mutant Arabidopsis flowers, together with the works done in Antirrhinum majus, lead

to the formulation of the well-known ABC model in flower development (Coen and

Meyerowitz, 1991) Arabidopsis thaliana has since eased its way through to become

the most popular model in studying plant development, biochemistry, and various molecular pathways including stem cell regulation, some of which were later found to

be conserved even in animal kingdom With its relative short generation time of 7-8

weeks, it has out-competed other plants like Antirrhinum majus, Petunia hybrida, Nicotiana tabacum (tobacco), and oryza sativa in the understanding of plant-specific

events, such as complex hormonal regulation, flowering, pathogenesis, etceteras

Transgenic Arabidopsis plants can be obtained by several transformation methods like Agrobacterium tumefaciens-mediated transformation by the transferred DNA of

the tumor-inducing Ti plasmid (T-DNA) and gene gun bombardment (Clough and

Bent, 1998; Taylor and Fauquet, 2002; Clough, 2005; Zhang et al., 2006; Ueki et al.,

2009) A large collection of T-DNA mutagenesis lines is available from

well-established Arabidopsis databases and stock centers to study the out or down effect of genes (www.arabidopsis.org; http://arabidopsis.info; Raina et al., 2002; Pan et al., 2003) Common gene silencing methods using double stranded RNA and microRNA work well in Arabidopsis and have contributed significantly in the

knock-understanding of numerous gene functions (Baulcombe, 2004; Schwab et al., 2005; Mansoor et al., 2006; Schwab et al., 2006; Ossowski et al., 2008; Bartel, 2009) In addition, it is easy to perform genetic analysis in Arabidopsis thaliana with its diploid

genome and heterosexual reproduction mode

Furthermore, detailed analysis of many cellular and developmental processes is made

possible in Arabidopsis by utilizing inducible systems like heat shock promoter,

ethanol inducible system, and steroid hormone receptors inducible systems In the glucocorticoid receptor (GR) post-translational activation system (Aoyama and Chua, 1997), hormone binding domain of rat glucocorticoid receptor is fused with a gene of

interest, which normally encodes for a DNA binding factor The translated protein fusion is retained in cytosol due to the binding of GR by heat shock protein Upon receiving synthetic steroid dexamethasone (DEX), the fusion protein would then be translocated into the nucleus to provide the gene activity of interest

1.2 Regulation of shoot apical meristem, inflorescence meristem and floral

meristem in Arabidopsis thaliana

In plants, there are two distinct apical stem cell pools governing the growth and differentiation at the shoot tip and root tip The shoot apical meristem (SAM) and root apical meristem (RAM) can give rise to various organs and tissues to sustain

postembryonic growth and differentiation in plants (Veit, 2004) During the

vegetative growth phase, the cells in shoot apical meristem can divide and

differentiate into leaf primordia Upon receiving flowering transition signal (for transition to reproductive phase), the SAM turns into an inflorescence meristem (Figure 1), which divides and gives rise to flower primordia or secondary

inflorescence meristems subtended by bracts Each of these floral primordia is

developed from a floral meristem (FM) and can be further developed into four

specific floral organs, which are perianths (sepals and petals) and floral reproductive organs (stamens and carpels) The FM is terminated after the pre-determined organ

primordia are generated A normal Arabidopsis flower consists of four sepals (calyx),

four petals (corolla), six stamens, and a pistil formed by two fused carpels (Zik and Irish, 2003)

In contrast to floral meristems, shoot apical meristems and inflorescence meristems are indeterminate, which means it can continue to grow and differentiate into

different organs depending on environmental cues it receives The stem cells in shoot apical meristems reside in the L1 and L2 layers of the dome-shaped apical region

termed central zone (CZ) (Carles and Fletcher, 2003; Sharma et al., 2003) The cells

flanking the lateral sides of the central zone form the peripheral zone (PZ) These cells at the peripheral zone differentiate and give rise to new organ primordia The rib zone (RZ) is located beneath the central zone and consists of cells that sustain the growth of stem

The regulation of stem cell niche in shoot apical meristem is a well-coordinated

process in Arabidopsis (Sablowski, 2007) WUSCHEL (WUS) and SHOOT

MERISTEMLESS (STM) are two key regulators for shoot apical meristem

development (Clark et al., 1996; Gallois et al., 2002; Lenhard et al., 2002) WUS is

expressed in a small population of cells at L3 layer beneath the L1 and L2 layers where the stem cells reside It encodes for a homeodomain-containing transcription factor that plays a pivotal role in maintaining the stem cell pools This population of

WUS-expressing cells is commonly regarded as the organizing centre of shoot

meristem In wus mutants, the stem cell pools in shoot apical meristem is depleted precociously, which leads to premature termination of meristem activity (Mayer et al.,

1998) STM is also a homeodomain-containing protein which is expressed in the

meristem It is suggested to play a less prominent role in meristem maintenance in comparison to WUS

WUS expression in the organizing centre is restricted by CLAVATA3 (CLV3) CLV3

is a secreted polypeptide that is expressed in L1 and L2 layers of the central zone

(Rojo et al., 2002) The domain of cells that are expressing CLV3 is commonly

regarded as stem cell pools in the meristem CLV3 moves between cells and serves as

a ligand for a receptor complex consists of CLV1 and CLV2 (Clark et al., 1993; Fletcher et al., 1999) Ectopic CLV3 activity causes instant suppression of WUS

expression and a decrease of meristem activity (Reddy and Meyerowitz, 2005) In

clavata mutants, the WUS expression domain is expanded and the meristem is

enlarged Interestingly, WUS has an instrumental role in the activation of CLV3 expression This WUS-CLV regulatory feedback loop is instrumental in keeping the meristem size in check (Brand et al., 2000; Schoof et al., 2000)

The growth of the indeterminate shoot apical meristem is eventually replaced by the growth of inflorescence meristem upon transition to reproductive development The process of flowering transition (from vegetative meristem to inflorescence meristem)

is accomplished by both internal and external signals, which converge on primary

regulators for floral identity Recently, the protein product of FLOWERING LOCUS

T (FT) gene has been strongly suggested to be the much sought-after flowering signal (Huang et al., 2005; Corbesier et al., 2007) FT is expressed in leaves and its

translated product is found to be transported from leaves to shoot apex though phloem

to trigger flowering upon long-day exposure In addition to FT, phytohormones like

gibberellins (GAs) also play an important role in the induction of flowering transition

(Blazquez et al., 1998) These flowering signals then activate a number of major

regulators that are important for inflorescence meristem identity which include

LEAFY (LFY), APETALA1 (AP1) and CAULIFLOWER (CAL) LFY is a

transcription factor that has an important role in promoting floral determinacy, as well

as inflorescence meristem determinacy (Schultz and Haughn, 1991; Weigel et al., 1992; Weigel and Nilsson, 1995) In lfy mutant, the floral meristems are transformed into inflorescence shoots Conversely, overexpression of LFY transforms the

inflorescence meristem into a terminal flower AP1 belongs to a MADS (MCM-1, AGAMOUS, DEFICIENS, SRF)-domain protein family It is a key determinant for both inflorescence meristem identity and floral organ identity (Irish and Sussex, 1990;

Mandel et al., 1992; Mandel and Yanofsky, 1995) CAL is another MADS-domain protein closely related to AP1 (Kempin et al., 1995) In ap1 cal double mutant, the

floral meristem becomes indeterminate and produces ectopic meristems to form a

cauliflower-like structure at the inflorescence apex In contrary, AP1 overexpression

transforms the inflorescence meristem into a terminal flower

Figure 1: Inflorescence meristem and floral meristem in Arabidopsis thaliana

Upper panel: schematic diagram of inflorescence meristem and floral primordium in

Arabidopsis thaliana L1, L2, and L3 layer of the meristem is shown CLV3

expression domain is highlighted in pink, and WUS expression domain is highlighted

in green Lower panel: a section of the meristem obtained from Ler wild-type plant

1.3 Regulation of flower development in Arabidopsis thaliana

During the processes of floral organ identity specification and floral organ

differentiation, there are well-coordinated interplays of transcription factors The classical ABC model was proposed nearly two decades ago to explain the organ identity determination in flower development (Coen and Meyerowitz, 1991) The ABC model predicts that the combinatorial action from ABC floral homeotic genes is largely responsible for the specification of the floral organ identity The ABC genes,

A class for APETALA1 (AP1) and APETALA2 (AP2), B class for APETALA3 (AP3) and PISTILLATA (PI), and C class for AGAMOUS (AG), have been extensively

studied and have been shown to encode transcription factors (Bowman et al., 1989; Yanofsky et al., 1990; Bowman et al., 1991; Jack et al., 1992; Goto and Meyerowitz, 1994; Gustafson-Brown et al., 1994; Jofuku et al., 1994)

Apart from its role in inflorescence meristem identity, AP1 also serves as an A-class gene for organ identity determination AP1, together with another A-class gene AP2,

is responsible for the specification of the first whorl of floral primordia into sepals

(Irish and Sussex, 1990; Jofuku et al., 1994) AP2 encodes for a member of

AP2/EREBP class of transcription factors In combination with B-class genes activity, AP1 and AP2 promote development of the second whorl of floral primordia into

petals In ap1 mutants, the first whorl organs undergo homeotic conversion, which

turns sepals into bracts subtended by floral meristems In the meantime, the second

whorl organs in ap1 mutant are lost In ap2 mutants, sepals are converted into carpels

or missing, whereas the petals in the second whorl are lost

B-class genes such as AP3 and PI act synergistically with both A-class and C-class

genes to specify the development of second whorl and third whorl into petals and

stamens, respectively (Jack et al., 1992; Goto and Meyerowitz, 1994) In the absence

of B-class genes, the second whorl organs are converted into sepals (in place of petals) and the third whorl organs are transformed into carpels (in place of stamens) Both

AP3 and PI encode for members of MADS domain transcription factors

The C-class gene AG also encodes a member of the MADS domain transcription factors, and AG is necessary for the specification of stamens and carpels, the floral reproductive organs (Bowman et al., 1989; Yanofsky et al., 1990) In ag-1 mutants,

the flower undergoes homeotic conversion to adopt a sepal-petal-petal reiteration instead of the normal sepal-petal-stamen-carpel structure The complete lack of

reproductive organs in ag-1 flowers places AG at the top of hierarchy of genes

controlling reproductive development This conclusion is supported by microarray

expression profiling of wild-type and ag mutant flowers showing that more than 1,

000 genes are expressed downstream of AG (Wellmer et al., 2004) In addition to its function in organ identity specification, AG also plays a prominent role in floral

determinacy In the absence of AG, WUS activity is prolonged in the indeterminate floral meristem (Lenhard et al., 2001; Lohmann et al., 2001) Our recent study has

revealed that AG acts through a zinc finger repressor, KNUCKLES (KNU) to

negatively regulate WUS for floral determinacy (Sun et al., 2009)

Studies performed on SEPALLATA (SEP) genes in Arabidopsis have added an

interesting twist into the classical ABC model for flower development (Pelaz et al., 2000) The three SEP genes, namely SEP1, SEP2 and SEP3 are now widely regarded

as E-class genes in ABCE model of flower development (Figure 2) These E-class genes act as important regulators for the specification of petals (second whorl),

stamens (third whorl) and carpels (fourth whorl) during flower development In sep1 sep2 sep3 triple mutants, the flowers produced consist only of sepals Recently, SEP3

has been shown to act as an upstream activator of the B and C class genes activity

(Kaufmann et al., 2009) Another SEP gene, SEP4 is also involved in the

development of sepals, petals, stamens, and carpels (Ditta et al., 2004) In sep1 sep2 sep3 sep4 quadruple mutans, all flower organs are converted into leaf-like organs

Following the unraveling of SEP proteins function, the ‘quartet model’ of flower organ identity has been proposed to explain the combinatorial effects of ABCE genes (Theissen, 2001; Theissen and Saedler, 2001) In this model, it is suggested that there are four floral homeotic proteins forming protein complexes to regulate and specify the development of floral organs In each whorl, there will be a distinct combination

of the four homeotic proteins, henceforth gives rise to different organ identities For instance, in whorl 2, there might be a combination of AP1, PI, AP3 and SEP proteins

to specify petals Whereas in whorl 4, the protein complex might be a combination of

AG and SEP proteins (two molecules for each protein) in the specification of carpels

Figure 2: The ABCE model of flower development

Schematic diagram showing the ABCE model of flower development Combinatorial action of A, B, and E-class genes specify the development of perianth organ Whereas, the B, C, and E-class genes determine the identity of reproductive organs

1.4 Patterning and differentiation of lateral organs in Arabidopsis thaliana

1.4.1 The patterning of lateral organs

Lateral organs are produced from the flanks of shoot apical meristem During the initiation of lateral organ primordia, it is suggested that auxin efflux carrier like

PINFORMED (PIN) mediates a directional flow of auxin into a localized region that

will mark the site for the emergence of a new primordium (Reinhardt et al., 2000; Benkova et al., 2003) In this site where auxin reaches its maximal concentration, the expression of class-I KNOTTED1-LIKE HOMEOBOX (KNOX) genes are

substantially downregulated The KNOX genes encode for homeobox transcription factors that are involved in the maintenance of shoot apical meristem (Hake et al., 2004) The repression of KNOX genes in the emerging primordia is substantiated by ASSYMETRIC LEAVES1 (AS1) and ASSYMETRIC LEAVES2 (AS2) (Semiarti et al., 2001; Iwakawa et al., 2007; Ueno et al., 2007) AS1 encodes a MYB-domain transcription factor, whereas AS2 encodes an AS2-LOB-domain protein Mutations in AS1 and AS2 genes will lead to the production of asymmetric and lobed rosette leaves (Byrne et al., 2000; Byrne, 2005) Both AS1 and AS2 form a heterodimer to execute their functions AS1/AS2 heterodimer binds directly to promoters of the KNOX genes, BREVIPEDICELLUS (BP) and KNAT1 to downregulate their expression (Guo et al., 2008) In the meantime, the expression of AS1 and AS2 in the shoot apical meristem

is, in turn, negatively regulated by the KNOX gene, STM The boundary specification

between the shoot apical meristem and the new lateral organ primordium is mediated

by a set of redundant, but partially distinct CUP-SHAPED COTYLEDONS (CUC) genes (Takada et al., 2001; Vroemen et al., 2003) CUC genes encode for NAC- family transcription factors Mutations in CUC genes lead to the fusion of cotyledons,

which results in cup-shaped cotyledons

1.4.2 Regulation of abaxial-adaxial polarity

The abaxial-adaxial polarity (Figure 3A) in the lateral organ is promptly established after the emergence of the new primordium There are two sets of genes responsible

for the specification of the abaxial and adaxial cell fates The KANADI (KAN) genes and the YABBY (YAB) genes (Bowman, 2000; Bowman et al., 2002) are required for the establishment of abaxial axis in the lateral organs The KAN genes like KAN1, KAN2, and KAN3 are transcription factors of the GARP family In kan1 kan2 double

mutant plant, the abaxial cell types of most of the lateral organs are replaced by the

adaxial cell types (Eshed et al., 2001) The YAB genes like FILAMENTOUS

FLOWER (FIL), YAB2, YAB3, INNER NO OUTER (INO), and CRABS CLAW (CRC) encode the YAB-family transcription factors FIL, YAB2 and YAB3 are expressed in

the abaxial domain of all of the lateral organs These three genes act redundantly to

promote abaxial cell fate in the lateral organs In fil-5 yab3-1 double mutant, there is a loss of polar development of most of the floral organs (Siegfried et al., 1999) INO expression is restricted to the abaxial side of the ovule’s outer integument In ino mutant, there is a loss of polar differentiation of the outer integument (Baker et al., 1997; Villanueva et al., 1999; Bowman, 2000) CRC is expressed exclusively in

carpels and nectaries (Bowman and Smyth, 1999) In crc mutant, there is only a partial loss of abaxial identity However, in crc kan double mutant, there is an ectopic production of adaxial tissue in the abaxial region (Alvarez and Smyth, 1999; Eshed et al., 1999)

The adaxial axis of the lateral organs in Arabidopsis is primarily established by the genes PHABULOSA (PHB), PHAVOLUTA (PHV), and REVOLUTA (REV) (Talbert

et al., 1995; McConnell et al., 2001; Otsuga et al., 2001; Emery et al., 2003) They

belong to a gene family encoding the class III homeodomain-leucine zipper

(HD-ZipIII) transcription factors Ectopic expression of either PHB or PHV is sufficient for abaxial-to-adaxial conversion In rev mutant, flowers are converted into

filamentous structures, signifying a loss of organ polarity

The interaction between the abaxial- and the adaxial-specifying genes has further contributed to the establishment of polarity in the lateral organs The AP2/ERF-type transcription factor, AINTEGUMENTA (ANT), acts with the YABBY-family protein

FIL, to positively regulate the expression of the adaxial-specifying gene, PHB

(Nole-Wilson and Krizek, 2006) The expression of HD-ZipIII homeobox genes such as

PHB, PHV, and REV are also closely regulated by miRNA165/166 (Mallory et al., 2004; Zhou et al., 2007) The KAN genes repress the expression of REV and PHV at the abaxial site In kan1 kan2 double mutant, the confinement of REV and PHV expression to the adaxial domain is delayed In addition, the expression level of REV and PHV is also higher in the double mutant (Eshed et al., 2001)

In parallel with the KAN genes, ETTIN (ETT, AUXIN RESPONSE FACTOR 3) and AUXIN RESPONSE FACTOR 4 (ARF4), have also been implicated in the regulation

of abaxial cell fate (Pekker et al., 2005) In ett arf4 double mutant, the abaxial tissues

are transformed into the adaxial one, and the phenotype of the double mutant is

closely related to that of KAN mutations The expression of ETT has been found to be closely regulated by trans-acting siRNA (ta-siRNA) (Adenot et al., 2006; Fahlgren et al., 2006; Garcia et al., 2006)

During the development of reproductive organs, NUBBIN (NUB) and JAGGED (JAG)

genes have been proposed to act in parallel to the HD-ZipIII pathway to promote the

adaxial cell fate (Dinneny et al., 2006) NUB and JAG encode C2H2 zinc-finger

transcription factors NUB is expressed in the adaxial side of stamens and carpels, whereas JAG is expressed in a non-polar way in the reproductive organs In jag nub

double mutant, stamens and carpels are abaxialized, and the organ growth is severely impaired

Figure 3: The axes of polarity in lateral organs and the gynoecium

(A) Schematic diagrams showing the abaxial-adaxial axis in lateral organs IM: inflorescence meristem; FM: floral meristem (B) Schematic diagrams showing apical-basal axis in the gynoecium

A

B

1.4.3 Regulation of apical-basal polarity

During the development of female reproductive organ in Arabidopsis, the apical-basal

polarity (Figure 3B) is established and maintained by a number of genes, whose mutations result in impaired development of the gynoecium Major regulators for the differentiation of gynoecium’s apical tissues is a basic helix-loop-helix transcription

factor, SPATULA (SPT) (Heisler et al., 2001) and ETT In spt mutant, there is

reduced growth of the style, stigma, and septum The transmitting tract is also absent

in the spt mutant In addition to its function in abaxial-adaxial patterning, ETT has a

greater role in apical-basal patterning of the gynoecium (Sessions and Zambryski,

1995; Sessions et al., 1997) In ett mutant, there is an expansion of the apical (style

and stigma) and basal (the gynophore) tissues with a reduction of the ovary region It

is suggested that ETT represses the expression of SPT to promote the formation of the ovary STYLISH (STY) genes have also been implicated in the regulation of the

Arabidopsis gynoecium development (Kuusk et al., 2002; Sohlberg et al., 2006) STY1 and STY2 are partially redundant genes encoding proteins with a RING finger- like motif In sty1 mutant, there is aberrant formation of the style of the gynoecium

In sty1 sty2 double mutant, there is an enhanced phenotype of sty1 mutant with

decrease proliferation of stylar xylem, and reduction of stylar and stigmatic tissues

A strong presence of an auxin response factor like ETT in the regulation of the

apical-basal polarity has suggested an instrumental role for auxin in the patterning of the gynoecium Indeed, auxin has been proposed to be the key phytohormone that is

responsible for setting up the apical-basal polarity in the gynoecium (Nemhauser et al., 2000) It is hypothesized that there is a decreasing gradient of auxin from the

apical part of the gynoecial primordium to the basal According to this hypothesis, high levels of auxin induces style and stigma differentiation, intermediate levels of auxin leads to formation of ovary, and low levels of auxin is responsible for the specification of the basal gynophore Recently, STY1 is reported to activate the

expression of YUCCA4, a gene encodes a flavin monooxygenase involved in auxin biosythesis (Sohlberg et al., 2006) Overexpression of YUCCA genes has been shown

to lead to the overproduction of auxin (Cheng et al., 2006)

1.4.4 Downstream target genes of AGAMOUS

Despite AG’s prominent role in reproductive organs development, very few target genes of AG have been characterized in detail AG has been implicated in the

regulation of other MADS-box genes like AP1 and SHATTERPROOF2 (SHP2) (Gustafson-Brown et al., 1994; Savidge et al., 1995) AG prevents the accumulation

of AP1 transcripts in the two inner whorls of the floral organs (whorl 3 and whorl 4) that will give rise to reproductive organs (Gustafson-Brown et al., 1994) A more comprehensive study has pointed out that SHP2 is a direct downstream target of AG (Savidge et al., 1995) SHP2 encodes a MADS-box protein sharing substantial similarity with AG It is formerly known as AGAMOUS-LIKE 5 (AGL5) SHP2 and

its close member, SHP1 have partial redundant function with AG Both SHP1 and

SHP2 are involved in fruit dehiscence (Liljegren et al., 2000) SHP2 is expressed

during early carpel development after the onset of AG expression Ectopic expression

of AG effectively activates the expression SHP2 In ag mutant, there is a loss of SHP

expression Furthermore, AG is shown to bind specifically to a promoter element of

SHP2

AG has been shown to positively regulate the expression of the SPOROCYTELESS (SPL) gene (Ito et al., 2004) SPL is involved in microsporogenesis, a process leading

to pollen formation AG binds to the 3’ region of the SPL gene and activates its

expression In another of our study, AG has been demonstrated to mediate late-stage stamen development (anther morphogenesis and dehiscence, and filament formation

and elongation) (Ito et al., 2007) It does so by controlling the expression of

DEFECTIVE IN ANTHER DEHISCENCE1 (DAD1) DAD1 encodes a gene involved

in the biosynthesis of a lipid-derived phytohormone, jasmonic acid AG binds to the

5’ coding region of the DAD1 gene, and activates its expression at later stages of floral development (days 7-8 after the onset of AG expression)

1.5 Functions of AT-hook DNA binding proteins during development

AT-hook DNA binding proteins may contribute to a functional nuclear architecture

by binding to the nuclear matrix, and may also be structural components that remain inside the nucleus after removal of basic proteins and histones (Strick and Laemmli,

1995; Aravind and Landsman, 1998; Morisawa et al., 2000) AT-hook motifs bind to

the minor grooves of duplex DNA in matrix attachment regions (MARs) of target

DNA sequences (Reeves and Nissen, 1990; Reeves, 2001), an unique property that distinguishes them from common transcription factors that primarily bind to the major grooves of duplex DNA

MARs are stretches of characteristic AT-rich DNA sequences that tend to attain a single-stranded conformation through base unpairing In this regard, MARs are also regarded as base unpairing regions (BURs) This greater tendency of forming

unpairing region in MAR is likely to be the result of sequence-induced torsional stress propagated from the surrounding DNA (Boulikas, 1995) MARs and AT-hook DNA binding proteins are suggested to be the key determinants in anchoring specific DNA sequences to nuclear matrix, a process that helps to generate chromatin loop domain and also to introduce structural changes in the chromatin (Reeves, 2001)

In animals, the MAR binding protein SATB1, which contains an AT-hook DNA binding motif, has been implicated in tissue- or cell type-specific regulation of

multiple genes (Alvarez et al., 2000; Cai et al., 2003; Britanova et al., 2005; Dobreva

et al., 2006; Han et al., 2008) SATB1 is suggested to play a prominent role in

chromatin assembly and histone modification It binds in the minor grooves of duplex

DNA (Nakagomi et al., 1994) and is able to regulate the transcription of multiple genes over long distances (Yasui et al., 2002) It is demonstrated that SATB1 can

recruit histone deacetylase to a specific locus and mediates a large-domain histone

deacetylation within that locus (Yasui et al., 2002)

A close member of SATB1, SATB2 is shown to act as a molecular node in a

transcriptional network regulating skeletogenesis (Dobreva et al., 2006) In Satb2

-/-mice, there are severe defects in craniofacial patterning and osteoblast differentiation SATB2 regulates the expression of multiple genes in osteoblasts, which includes

Bone Sialoprotein (Bsp), Osteocalcin (Ocn), and several Hox genes like Hoxa2

In plants, several AT-hook DNA binding proteins have been identified and are

implicated in various aspects of plant development (Weigel et al., 2000; Lim et al., 2007; Matsushita et al., 2007; Vom Endt et al., 2007; Street et al., 2008) In

Arabidopsis, ESCAROLA (ESC/ORE7) has been isolated from an activation-tagging screen and is suggested to play a role in leaf longevity (Lim et al., 2007) A closely related SOB3 (SUPPRESSOR OF PHYB-4#3) gene is involved in the negative

modulation of hypocotyl growth (Street et al., 2008) AGF1 (AT-HOOK PROTEIN

OF GA FEEDBACK REGULATION 1) has been identified from a yeast one-hybrid screen as a negative regulator of GA signaling through binding to cis-acting sequence

of GA 3-oxidase (Matsushita et al., 2007) In addition, AT-hook proteins are also

likely to participate in the regulation of gene expression in response to jasmonic acids

(JA) in suspension culture of Catharanthus roseus cells (Vom Endt et al., 2007)

Recently, AHL22 (AT-HOOK MOTIF NUCLEAR LOCALIZED PROTEIN22)

has been reported to be involved in the regulation of flowering and hypocotyl

elongation (Xiao et al., 2009) Overexpression of AHL22 effectively delays the flowering process in Arabidopsis In spite of these reports, the function of AT-hook

DNA binding protein has yet to be demonstrated in flower development and plant meristem regulation

1.6 Objective of the study

The aim of this study is to augment the understanding of flower development and

meristem regulation in Arabidopsis thaliana through functional study of a novel

regulator of gene expression, GIANT KILLER Specifically, I wish to address the

following questions in the study: First, is GIK a direct target of AG? Second, what is

the function of GIK in reproductive development? Third, what are the downstream targets of GIK in reproductive development? Fourth, how does GIK mediate gene regulation? Fifth, does GIK have any roles in meristem regulation?

1.7 Significance of this study

In this study, I have presented findings showing that GIK serves as a new way of

control employed by the floral homeotic gene AGAMOUS to regulate multiple genes

expression I have identified several key regulators in reproductive patterning and

differentiation as downstream targets of GIK In particular, I have shown that ETT is negatively modulated by GIK, and that the GIK-mediated ETT regulation is closely

associated with epigenetic changes of the chromatin In addition, my study has also

provided new insight for the understanding of meristem integrity in Arabidopsis thaliana I have shown that ectopic GIK activity effectively disrupts stem cell

regulation, and causes aberrant expression of WUS in inflorescence and floral

meristems

CHAPTER 2: Materials and Methods

2.1 Materials

2.1.1 Plant materials

All Arabidopsis thaliana plants used in this study are in the Landsberg erecta

background unless otherwise stated They were grown at 22 °C under continuous

light condition in plant growth chambers or the plant growth room The 35S::GIK and 35S::GIK-GR-6HA seeds were provided by Dr Toshiro Ito A GIK insertion line was

obtained from the TRAPPER collection (http://genetrap.cshl.edu/TrHome.html)

(stock number, ET14389) The enhancer trap was inserted at the middle of the coding

region, 450 bp downstream from the start codon Homozygous lines were verified by PCR-based genotyping

2.1.2 Bacterial strains

Escherichia coli strains used for the molecular cloning in this study were DH5α and

XL1-BLUE unless otherwise stated The One Shot ccdB Survival-T1 (Invitrogen) was used specifically to amplify and maintain plasmids containing cytotoxic ccdB gene E coli strains were grown in Luria-Bertani (LB) medium at 37 °C with

appropriate selective antibiotics The Agrobacterium tumefaciens strain used for the

plant transformation in this study was C58C1 (GV3101) The C58C1 strain carrying

the pSoup helper plasmid was used specifically to amplify pGreen-based plasmids A tumefaciens were cultured in LB medium at 28 °C with 50 µg/mL rifampicin and 30 µg/mL gentamycin

2.2 Agrobacterium-mediated plant transformation

2.2.1 Preparation of Agrobacterium tumefaciens competent cells for

electroporation-mediated gene transfer

The glycerol stock of C58C1 Agrobacterium was streak onto a LB agar plate and

grown for 48 h at 28 °C Several single colonies were picked for inoculation in a 3

mL LB liquid medium The Agrobacterium culture was then transferred to a 500 mL

LB liquid medium and was grown until it reached a density of 0.4-0.6 at OD600 The culture was then centrifuged at 3, 000 rpm for 20 min at 4 °C The cell pellet was washed twice with ice-cold water and resuspended in 10% glycerol The cells were then aliquoted and frozen in liquid nitrogen, and were stored at -80 °C

medium and transferred into a microfuge tube for incubation at 28 °C for 4 h with continuous shaking The transformed mixture was then centrifuged at 5, 000 rpm for

5 min The pellet was resuspended in 100 µL LB liquid medium and plated onto LB agar plates containing appropriate antibiotics

resuspended in freshly made plant transformation solution containing 5% sucrose and

0.005% Silwet L-77 The whole inflorescence of Arabidopsis was dipped into the

transformation solution for 3-5 min with gentle shaking The transformed plants were then kept in the dark and at high humidity for 24 h at 4 °C The plants were then transferred to a growth chamber and grown under normal condition for seed-setting The seeds were then harvested and dried at room temperature The dried seeds were planted on MS agar plate or directly on soil for germination The T1 plants were selected using appropriate selective agents

2.3 Dexamethasone treatment

Dexamethasone (DEX) treatment was done by submerging Arabidopsis

inflorescences in a solution containing 10 µM DEX and 0.015% Silwet L-77 for 1

min with gentle shaking 35S::GIK-GR-6HA plants were treated with DEX five times

at one-day intervals for phenotypic observation

2.4 Plant observation and photography

Arabidopsis flowers and inflorescences were examined using dissection microscopes

(Olympus SZ51 or Nikon SMZ645) Plant photographs were taken using a Nikon SMZ 1500 stereoscopic microscope attached to the digital camera SIGHT DS-U1

2.5 Scanning electron microscopy

Arabidopsis flowers or inflorescences were collected freshly from plants minutes

before scanning The specimen was mounted on a sample stage and was quick-frozen using liquid nitrogen The frozen specimen was then examined using the scanning electron microscope JEOL JSM-6360LV

2.6 Generation of RNAi silencing lines

To generate the 35S::GIK-RNAi construct, a C-terminal fragment of the GIK coding region (GIK-Cter, 410-808 bp) was amplified using UltraPfu-High-Fidelity DNA polymerase (Stratagene) to produce BamHI-GIK-Cter-ClaI and XhoI-GIK-Cter-KpnI

fragments These fragments were cloned into the pKANNIBAL vector (Wesley et al., 2001) pKANNIBAL-GIK-RNAi was digested by NotI to produce a 35S::GIK-RNAi fragment, which was then cloned into the pMLBART binary vector (Gleave, 1992) GIK-RNAi transgenic plants were selected using the herbicide Basta A few

percentages of the examined flowers showed reproductive defects in the T1 and T2

generations In the T3 generation, there were reduced percentages of the GIK-RNAi flowers showed reproductive defects To generate the 35S::GIK2-RNAi construct, an N-terminal fragment of the GIK2 (AT4g17800; 39-260 bp) coding region was

amplified using UltraPfu-High-Fidelity DNA polymerase (Stratagene) to produce

BamHI-GIK2-Nter-ClaI and XhoI-GIK2-Nter-KpnI fragments These fragments were cloned into the pKANNIBAL vector pKANNIBAL-GIK2-RNAi was digested by NotI

to produce a 35S::GIK2-RNAi fragment, which was then cloned into the pMLBART binary vector One of the T1 plants of 35S::GIK2-RNAi was crossed with gik The GIK2 RNAi gik plants were obtained and confirmed by Basta selection and PCR

genotyping

2.7 Extraction of plant genomic DNA

Plant genomic DNA was extracted using the CTAB extraction method Leaves were ground in an eppendorf tube with pestle The sample was then added with 500 µL of 2× CTAB buffers (3% CTAB, 100 mM Tris-Cl, pH 8.0, 20 mM EDTA, 1.4 M NaCl, 0.2% β-mercaptoethanol) and incubated at 65 °C for 30 min After incubation, the sample was mixed with 500 µL of chloroform, and centrifuged at 9, 000 rpm for 5

min The upper layer of the solution was collected and transferred into a new

microfuge tube The solution was then added with 2/3 volume of isopropanol (330 µL) and left at room temperature for 15 min to precipitate the DNA Subsequently, the solution was centrifuged at 13, 000 rpm for 5 min to collect the precipitated DNA The DNA pellet was rinsed once with 70% ethanol and dissolved in 30 µL of TE buffer containing 20 µg/mL of RNase A

2.8 Expression analysis

2.8.1 RNA isolation

Total RNA was isolated from floral bud clusters at stage 10 or younger (Ito et al.,

2007) using RNeasy plant mini kit (Qiagen) Around 100 mg of the floral sample was transferred into a blue grinding tube and were ground into fine particles in liquid nitrogen The sample was then added with 450 µL of buffer RLT and transferred into

a QIAshredder spin column attached with a collection tube The column was spun at

13, 000 rpm for 2 min The supernatant in the collection tube was carefully

transferred into a new microfuge tube It was then added with 225 µL of 100%

ethanol and mixed by pipetting The mixture was applied onto an RNeasy mini

column and was centrifuged at 10, 000 rpm for 15 s The flow through was discarded and the column was washed once with 350 µL of Buffer RW1 In the meantime, 10

µL of DNase I solution (Qiagen) was pre-mixed with 70 µL of RDD buffer The DNase I-RDD solution was then applied onto the RNeasy silica-gel membrane for

incubation at 37 °C for 30 min After DNase I incubation, the column was rinsed once with 350 µL of buffer RW1 The column was subsequently washed twice with RPE buffer and was centrifuged at 10, 000 rpm for 2 min at the last wash The column was then transferred into a new microfuge tube and the RNA was eluted with 50 µL

RNase-free water

2.8.2 Reverse transcription

The total RNA isolated from plant materials was reverse-transcribed by the

Superscript III RT-PCR system (Invitrogen) Around 2 µg of total RNA was mixed with 0.5 µL of 50 mM Oligo-dT and 0.5 µL of 10 mM dNTPs mix in a 5 µL incubation system The mixture was incubated at 65 °C for 5 min and later placed on ice for 1 min After the cold incubation, the mixture was added with 1 µL of 10× RT buffer, 2 µL of 25 mM MgCl2, 1 µL of 0.1 mM DTT, 0.5 µL of RNase OUT enzyme, and 0.5 µL of Superscript III enzyme in a 10 µL incubation system The solution was incubated at 50 °C for 70 min and then at 85 °C for 5 min Next, the solution was added with 1 µL of RNase H and further incubated at 37 °C for 20 min The cDNA solution was then diluted to 50 µL using TE buffer

pre-2.8.3 Real-time PCR

The cDNA sample was pre-mixed with 2× SYBR Green PCR master mix solution and 0.5 µM of primer mix, and was adjusted with sterile water into a 10 µL reaction

system Quantitative real-time PCR assays were performed in triplicate using 7900HT fast real-time PCR system (Applied Biosystems)

2.9 RNA in situ hybridization

2.9.1 In vitro transcription

The plasmid construct used for GIK in vitro transcription in RNA in situ

hybridization, 2ATH1B-6, was obtained from Dr Toshiro Ito Around 4 µg of the plasmid 2ATH1B-6 was digested with excess amount of XhoI to get a linearized plasmid for an in vitro run-off transcription The digested plasmid was run on an agarose gel to check for complete digestion The digested plasmid was then purified using a gel purification column (Qiagen) Around 2 µg of the purified plasmid DNA was mixed with 2 µL of 10× transcription buffer, 2 µL of 10× NTP-DIG mix, 2 µL of

100 mM DTT, 1 µL of RNase inhibitor, and 40 U of SP6 RNA polymerase in a 20 µL incubation system The reaction mix was incubated at 37 °C for 3 h Next, the in vitro transcribed RNA mix was added with 2 µL of RNase-free DNaseI and incubated at

37 °C for 30 min to remove the DNA template To partially hydrolyze the RNA probe into ~150 bp for improved permeability to tissue sections, the reaction mix was scaled

up to 100 µL with DEPC water for alkaline treatment The reaction mix was added with 100 µL of fragmentation buffer (80 mM NaHCO3 and 120 mM Na2CO3) and incubated at 60 °C for 20 min The reaction mix was then neutralized with 10 µL of 10% acetic acid and precipitated with 22 µL of 3 M NaOAc (pH 5.2), 500 µL of

absolute ethanol, and 2 µL of 10 mg/mL tRNA at -20 °C overnight The precipitated RNA probes were collected by centrifugation at 10, 000 rpm for 18 min and rinse once with 70% ethanol The RNA probes were resuspended in 50 µL of 50%

formamide The probes were used at the final concentration of 0.5 ng/µL/kb in the hybridization solution

2.9.2 Fixation of floral tissues

The fixative solution (4% (w/v) paraformaldehyde in PBS) was prepared freshly and kept at 4 °C Inflorescences were harvested and immediately immersed in the ice-cold fixative solution The immersed samples were kept in vacuum for 30 min to remove trapped air bubbles in the floral tissues The fixative solution was changed once after the vacuum treatment, and the samples were further incubated overnight with the fixative solution at 4 °C with shaking

2.9.3 Dehydration and wax embedding

The fixed floral tissues were washed twice with PBS at 4 °C for 30 min The fixed samples were then dehydrated using a series of ethanol solutions: 30%, 40%, 50%, 60%, 70%, and 85%, each for 2 min at 4 °C The samples were further dehydrated using 95% ethanol added with eosin Eosin was added for the purpose of tissue

staining Subsequently, the samples were treated with 100% ethanol with eosin for four times The fixed dehydrated samples were subjected to tissue infiltration using a

graded series of histoclear in ethanol: 25%, 50%, and 75%, each for 60 min at room temperature Next, the samples were treated twice with 100% histoclear for 60 min at room temperature The samples were then further incubated overnight with 100% histoclear added with 1/4 volume of paraplast chips at room temperature for paraffin infiltration On the second day, the samples were placed at 42 °C for the paraplast chips to melt completely The samples were then added with an additional 1/4 volume

of paraplast chips for melting Later, the samples were transferred to 58 °C for a few hours The histoclear/paraffin wax solution was replaced with an equal volume of freshly melted paraffin wax, and the samples were further incubated overnight at 58

°C Subsequently, the wax solution was changed twice a day for the following three days The wax-embedded samples were allowed to set in a plastic weighing boat at room temperature The wax blocks could be kept at 4 °C for long-term storage

2.9.4 Tissue sectioning

The wax blocks were removed from the plastic weighing boat and further trimmed into small wax cubes for sectioning The samples were sectioned into a series of 7 µm-thick sections using a fully motorized rotary microtome (Leica RM 2165)

Ribbons of sections were placed on ProbeOn Plus slides (Fisher Biotechnology, USA) and were floated with sterile water to flatten the uneven layer of wax The slides were then placed on a slide warmer overnight at 42 °C The sectioned tissues could be kept

at 4 °C for several weeks

2.9.5 Pre-hybridization

Non-radioactive in situ hybridization was performed as previously described (Long and Barton, 1998) On the first day, the sectioned tissues on slides were treated twice with 100% histoclear for 10 min at room temperature for deparaffinization

Subsequently, the slides were rehydrated with a graded series of ethanol: 100%, 95%, 90%, 80%, 60%, and 30%, each for 2 min with occasional shaking The slides were further rehydrated twice with DEPC water and were incubated in 2× SSC for 15-20 min The slides were then treated with Tris-EDTA buffer (0.1M Tris-Cl, and 50 mM EDTA) containing 1 µg/mL Proteinase K at 37 °C for 20 min with constant shaking Next, the slides were treated with PBS solution containing 2 mg/mL of glycine for 2 min to stop the Proteinase K activity The slides were further rinsed twice with PBS, and treated with 4% (w/v) solution of paraformaldehyde in PBS for 10 min Later, the slides were washed twice with PBS for 5 min and were further incubated with

acetylation buffer (0.1 M triethanolamine, and 5% acetic anhydride) for 10 min The slides were further washed twice with PBS for 5 min, and were dehydrated using a graded series of ethanol: 30%, 60%, 80%, 90%, and twice with 100%, each for 30 s The slides were then air-dried on Kimwipe tissue

2.9.6 Hybridization

The slides were incubated overnight with the hybridization buffer containing 0.5 ng/µL/kb RNA probes at 55 °C 240 µL of hybridization buffer was used for every

pair of the slides The hybridization buffer contained 24 µL of 10× in situ salts

solution, 96 µL of formamide, 4.8 µL of 50× Denhardt’s solution, 19.2 µL of 10 mg/mL tRNA, 48 µL of 50% Dextran sulfate solution, and 40 µL of RNA probes at concentration mentioned above The hybridization buffer was adjusted to 240 µL with 50% formamide The RNA probes were pre-incubated at 80 °C for 2 min and were chilled on ice before adding into the hybridization buffer After the RNA probes were added into the hybridization buffer, the buffer was vortex, and incubated at 65

°C for 5 min The hybridization buffer with RNA probes was then carefully applied onto one slide The second slide was slowly sandwiched onto the first slide All the sections were to be in contact with the hybridization buffer The hybridized slides were then placed on a stand in a plastic container that was tightly sealed, and were incubated overnight at 55 °C

2.9.7 Post-hybridization

On the second day, the hybridized slides were washed thrice with 0.2× SSC at 55 °C for 1 h with constant shaking The slides were blocked with 1% Boehringer solution (Boehringer powder dissolved in 100 mM Tris-Cl, pH 7.5, and 150 mM NaCl) for 45 min at room temperature with gently shaking The slides were further blocked with 1% BSA solution (BSA dissolved in 100 mM Tris-Cl, pH 7.5, 150 mM NaCl, and 0.3% Triton-X) for 45 min Anti-DIG was diluted in a ratio of 1:500 in the 1% BSA blocking solution The anti-DIG solution was dispensed into a plastic weighing boat The slides were sandwiched together and submerged into the anti-DIG solution This

would allow the capillary action to soak up the antibody solution throughout the sections The slides were incubated with the anti-DIG solution for 2 h at room

temperature Next, the slides were drained on a Kimwipe tissue and separated The separated slides were washed four times with 1% BSA blocking solution, each for 15 min at room temperature The slides were then incubated with the Tris-MgCl solution (100 mM Tris-Cl, pH 9.5, 100 mM NaCl, and 50 mM MgCl2) for 10 min at room temperature For signal detection, 20 µL of NBT/BCIP stock solution (Roche) was added into 1 mL of Tris-MgCl-PVA solution (10% (w/v) polyvinyl alcohol dissolved

in the Tris-MgCl solution) The mixture was then applied onto one of the slides and the second slide was sandwiched onto the first one The slides were then incubated overnight in total darkness at room temperature On the third day, the slides were drained, separated and were rinsed with tap water thrice to stop the reaction The slides were then dehydrated using 60%, 90%, and twice with 100% ethanol, each for

8 s After dehydration, the slides were air-dried and mounted in 50% glycerol

2.10 ETT promoter analysis

2.10.1 Constructs

The 8.7 kb pETT::GUS construct were obtained from Dr Toshiro Ito The 4.9 kb pETT∆MAR::GUS construct was amplified using UltraPfu-High-Fidelity DNA polymerase (Stratagene) at the extension time of 5 min and then cloned into pENTR

directional TOPO cloning vector (Invitrogen) Clones were sequenced for

confirmation Next, the entry clone was cloned into the pBGWFS7 binary vector (Karimi et al., 2002) by the Gateway cloning method

2.10.2 Transgenic plants and DEX treatments

Ler wild-type plants were transformed with both pETT::GUS and pETT∆MAR::GUS

constructs T1 transgenic plants were selected using herbicide basta Homozygous

transgenic plants with positive GUS reporter expression were crossed with GR-6HA plants to obtained pETT::GUS 35S::GIK-GR-6HA and pETT∆MAR::GUS 35S::GIK-GR-6HA double transgenic plants DEX treatments were performed as

35S::GIK-described above at 2-day intervals Whole inflorescences were collected for GUS staining assay

2.10.3 GUS staining

Whole inflorescences were rinsed and stained to determine GUS activity for GUS

expression analysis as previously described (Ito et al., 2003) Inflorescences were

harvested and fixed with 90% ice-cold acetone on ice for 20 min Later, the acetone solution was removed and the inflorescences were rinsed with the rinse solution (50

mM Phosphate buffer, 0.5 mM K3Fe(CN)6, and 0.5 mM K4Fe(CN)6) The rinse solution was then replaced by the staining solution (50 mM Phosphate buffer, 0.5 mM K3Fe(CN)6, 0.5 mM K4Fe(CN)6, and 2 mM X-GLUC) The samples in the staining

solution were vacuum infiltrated for 1 h and were further incubated overnight at 37

°C The developed GUS signal was examined under the dissecting microscope

2.11 Antigen purification and generation of polyclonal antibodies

Full length GIK cDNA and truncated cDNA with either conserved N-terminal domain

or AT-hook domain were each cloned into the pQE30 vector (QIAGEN) to produce

6xHis-GIK protein Recombinant protein was induced using 1 mM IPTG and purified

on a nickel column (QIAGEN) under denaturing conditions Protein was then

partially refolded through buffer exchange and concentrated using a Centriprep Centrifugal Filter with an Ultracel YM-10 membrane (Millipore) Purified 6xHis-GIK recombinant protein was injected intramuscularly into guinea pigs with Freund’s adjuvant Blood was withdrawn after the fourth and sixth immunizations Whole blood was processed to obtain polyclonal anti-GIK serum

2.12 Chromatin immunoprecipitation

The chromatin immunoprecipitation (ChIP) assay was performed as described

previously (Ito et al., 1997; Ito et al., 2007) using ChIP Assay Kit (Upstate) with

some modifications Inflorescences were ground in liquid nitrogen for 10 min The ground tissues were then added with 250 µL of M1 buffer (10 mM phosphate buffer, 0.1 M NaCl, 10 mM β-mercaptoethanol, and 1 M hexylene glycol) on ice and mixed well with a pestle Another 650 µL of M1 buffer was added to bring up the volume of

the solution The sample was fixed with 24.3 µL of 37% formaldehyde for 10 min at

4 °C Next, the sample was added with 57.6 µL of 2 M glycine solution to stop the fixation The solution was then filtrated by a double-layered miracloth The filtrate was centrifuged at 13, 000 rpm for 1 min at 4 °C The supernatant was discarded, and the pellet was washed four times with M2 buffer (10 mM phosphate buffer, 0.1 M NaCl, 10 mM β-mercaptoethanol, 1 M hexylene glycol, 10 mM MgCl2, and 0.5 % Triton-X) and then once with M3 buffer (10 mM phosphate buffer, 0.1 M NaCl, and

10 mM β-mercaptoethanol) At the last wash, traces of liquid were thoroughly

removed using a small tip The nuclear pellet was then resuspended in 200 µL of SDS lysis buffer (Upstate), and was incubated on ice for 10 min The sample was added with 800 µL of ChIP dilution buffer to bring up the volume to 1 mL and was

subjected to sonication using Sonics VibraCell Ultrasonic Processor Chromatin was sheared into an average DNA length of 500 bp The sonicated sample was centrifuged

at 13, 000 rpm for 10 min at 4 °C The supernatant was transferred into a new tube and was added with 1 mL of ChIP dilution buffer The solubilized chromatin was then precleared with salmon sperm DNA-treated protein A- (in the case of anti-AG, anti-dimethylated H3K9, and normal rabbit IgG) or protein G- (in the case of anti-HA) agarose beads (Upstate) for 2 h at 4 °C After incubation, the sample was centrifuged

at 1, 000 rpm for 1 min at 4 °C 250 µL of the supernatant was kept as an input

control The rest of the supernatant was incubated overnight at 4 °C with anti-AG serum (for AG ChIP experiments), anti-HA (Roche) (for GIK ChIP experiments), anti-modified histone antibodies (Upstate) for dimethylated H3K9, dimethylated H3K4, acetylated histone H3 and trimethylated H3K27 (for histone modification