Characterization of HD2C and its interaction with novel ......14-3-3 interaction with HD2C was investigated using Bimolecular Fluorescent Complementation. HD2C bound to both epsilon
Post on 15-Nov-2020
2 Views
Preview:
Transcript
Characterization of HD2C and its interaction with novel partners in
Arabidopsis thaliana
This thesis is presented for the degree of Doctor of Philosophy of the University of Western Australia
Michael Van der Kwast
Biochemistry and Molecular Biology
School of Chemistry and Biochemistry
April 2013
Supervisors: Dr Thomas Martin and Dr Martha Ludwig
Declaration
The work presented in this thesis is my own work except where stated. This work was
carried out in the School of Chemistry and Biochemistry, Faculty of Life and Physical
Sciences, at the University of Western Australia. The material presented in this thesis
has not been presented for any other degree.
Michael Van der Kwast
April 2013
1
Acknowledgment
I would like to express my sincere gratitude and appreciation for the hard work of my
principle supervisor Dr. Thomas Martin. It was only through his guidance and
determination that I achieved my goal of finishing my PhD and accomplishing my
successes in the lab. I gained many insights into the scientific way of thinking through
our frequent meetings (I also enjoyed our many discussions on the merits of various
cricket players).I would also like to thank my second supervisor, Dr. Martha Ludwig,
who offered alternative viewpoints so that I could take my research down unexpected
avenues. Special thanks as well to all of the present and former members of the labs of
Dr. Patrick Finnegan and Martha. I enjoyed our fortnight meetings despite the
occasional disheartened atmosphere as we would share a frustrating or failed
experiment. In particular I must mention the early help that I received from Hung Chi
Lui, who took me under his wing at the start of my PhD.
I have been privileged to work with some talented scientists in Thomas’ lab. In
particular I would like to thank Adrian Sheng Hao Tong, Ruo Han Li, Sally Grasso. A
super, special thanks to Julia Man, for her amazing help when I was writing-up; their
presence kept me motivated (and occasionally entertained) in the lab and I will be sad
to leave such an enriching environment.
I am grateful for the help of Mr John Murphy and Dr Paul Rigby in learning and
operating the confocal microscope. My project was dependent on their skills and their
time was greatly appreciated.
I would like to acknowledge and thank the help of my family and friends. Daniel
Pegdon, I promised to acknowledge you here, but now you have to read EVERY WORD
of this thesis. My parents were always willing to help, and their gift of cooking dinner
when I was stressed was something that I often joked about, but really appreciated.
Lastly I would like to acknowledge and thank my wife, Karen. I am truly grateful for her
unwavering support and help whether I felt that I needed it or not. This was an equally
difficult time for her (although I’m sure that she appreciates all of the new things that
she has learnt about histone deacetylases).
2
Abstract
Histone deacetylases are known to bind a variety of targets and interacting proteins
which are responsible for determining its activity in a temporal and spatial context.
Here, Bimolecular Fluorescent Complementation (BiFC) assays were used to analyse
putative interactions of the plant specific histone deacetylase HD2C isoform from
Arabidopsis thaliana. Specifically, HD2C was shown to form homo- and hetero-dimers
through interaction with other HD2 isoforms; it was shown to interact with both
epsilon and non-epsilon 14-3-3 isoforms; and lastly it was shown to interact with the
transcription factor TGA6.
Localization of HD2C-GFP traced accumulation of HD2C to the nucleus and nucleolus of
Arabidopsis leaf cells. This was evidently dynamic, as salt stress induced the HD2C-GFP
to localize to the nucleolus. This was linked to HD2C dimers, which showed a similarly
nucleolar localization pattern. The nuclear localization was shown to be dependent on
an evolutionarily conserved KKAK motif.
14-3-3 interaction with HD2C was investigated using Bimolecular Fluorescent
Complementation. HD2C bound to both epsilon and non-epsilon classes of 14-3-3
proteins, while the other HD2 isoforms were able to bind to 14-3-3 epsilon. The region
required for this interaction was traced to three distinct areas of the HD2C protein.
Site directed mutagenesis of serine and threonine residues was used to identify critical
potentially phosphorylated binding sites within the C-terminal region of the HD2C. The
study revealed that substitutions of S284, T235 and S239 by alanine abolished 14-3-3
binding to a C-terminal peptide of HD2C.
Lastly, it was attempted to identify a link between HD2C and salicylic acid (SA) and
jasmonic acid (JA) signalling. Interaction between TGA6, a known SA and JA response
transcription factor with HD2C was demonstrated using BiFC. In addition two lines
over-expressing a HD2C-GFP construct in Arabidopsis plants were compared with Col-0
wild type plants to determine if there a SA or JA dependent phenotypes could be
linked to the enzyme. An increased sensitivity of the overexpressing lines to SA and JA
was found during germination. However, when analysing HD2C expression, exogenous
application of salicylic acid and jasmonic acid had no evident effect on HD2C
3
expression. Finally, HD2C was tested for its ability to regulate as-1 like promoter cis-
element containing genes which are known to be regulated by SA and/or JA. A slight
decrease in the expression of PDF1.2, a JA responsive gene, as well as in the expression
of the SA and JA- RLK1 gene was observed, suggesting that HD2C may be somewhat
involved in regulating their expression.
4
List of Abbreviations
ABA abscisic acid
ABRC Arabidopsis Biological Resource Centre
AMP adenosine monophosphate
ATP adenosine triphosphate
Arabidopsis Arabidopsis thaliana
BiFC bimolecular fluorescence complementation
bp base pairs
BSA bovine serum albumin
CaMV cauliflower mosaic virus
cDNA complementary DNA
DAPI 4’, 6-diamidino-2-phenylindole
DNA deoxyribonucleic acid
E. coli Escherichia coli
ECL enhanced chemiluminescence
ER endoplasmic reticulum
EtBr ethidium bromide
GFP green fluorescent protein
H2O2 hydrogen peroxide
HA hemagglutinin
HDACs histone deacetylases
HD2C Arabidopsis histone deacetylase 2C
LB medium Luria-Bertani Broth medium
MCS multiple cloning sites
MeJA methyl jasmonate
5
MQ H2O Milli-Q H2O
MS medium Murashige and Skoog medium
Nicotiana Nicotiana benthamiana
OD600 optical density at 600 nm
ORF open reading frame
PAGE polyacrylamide gel electrophoresis
PCR polymerase chain reaction
PEG polyethylene glycol
RFP red fluorescent protein
PTGS post-transcriptional gene silencing
rpm revolutions per minute
SDS sodium dodecyl sulfate
SDW sterile deionised water
Tris tris (hydroxymethyl) aminomethane
VP1 VIVIPAROUS1
YC (C-YFP) C-terminal fragment of YFP
YFP yellow fluorescent protein
YN (N-YFP) N-terminal fragment of YFP
6
Table of Contents
Chapter 1 ..................................................................................................................................... 10
General introduction ................................................................................................................... 10
1.1 Epigenetics ........................................................................................................................ 11
1.2 DNA methylation ............................................................................................................... 12
1.3 Non-coding RNA ................................................................................................................ 13
1.4 Histone Modifications ....................................................................................................... 14
1.5 Plant specific HD2 family................................................................................................... 15
1.6 Structure of HD2 ............................................................................................................... 15
1.7 HD2 expression patterns ................................................................................................... 18
1.8 HD2 function ..................................................................................................................... 20
1.9 HD2s- functionally redundant? ......................................................................................... 23
1.10 The HD2 complex- determining a higher level protein interactome .............................. 23
1.11 Project hypothesis and aims ........................................................................................... 25
1.11.1 Hypothesis ................................................................................................................ 25
1.11.2 Project aims.............................................................................................................. 27
Chapter 2 ..................................................................................................................................... 29
Materials and Methods ............................................................................................................... 29
2.1 Materials ........................................................................................................................... 30
2.2 Methods ............................................................................................................................ 32
2.2.1 General methods ................................................................................................ 32
2.2 Plant growth and transformations .............................................................................. 32
2.2.1 Sterilizing seeds ................................................................................................... 32
2.2.2 Seed sowing ........................................................................................................ 32
2.3 Transformation of A. thaliana leaves .......................................................................... 33
2.3.1 Floral dip ............................................................................................................. 33
2.5 Bacterial preparations and transformations .............................................................. 34
2.5.1 Competent cell preparation (E.coli) ................................................................... 34
2.5.2 Competent cell transformation (E.coli) ............................................................... 34
2.5.3 Agrobacteria competent cell preparation .......................................................... 35
2.5.4 Agrobacteria competent cell transformation ..................................................... 35
2.5.5 A.tumefaciens infiltration of Nicotiana benthamiana leaves ............................. 36
2.6 Nucleic acid manipulations ......................................................................................... 36
7
2.6.1 Phenol extraction ............................................................................................... 36
2.6.2 Ethanol precipitation .......................................................................................... 37
2.6.3 Isolation of genomic DNA from Arabidopsis thaliana plants .............................. 37
2.6.4 Polymerase Chain Reaction (PCR) ....................................................................... 38
2.6.5.1 High fidelity cloning PCR .................................................................................... 38
2.6.5.2 Standard Taq polymerase, qualitative PCR ............................................................. 38
2.6.6 Miniprep .............................................................................................................. 42
2.6.7 Midiprep .............................................................................................................. 42
2.6.8 Restriction digestion ........................................................................................... 42
2.6.9 Agarose gel staining, excision and purification ................................................... 42
2.6.10 Ligation ............................................................................................................... 43
2.6.11 Semi-quantitative RT-PCR ................................................................................... 43
2.7 Protein assays and procedures ................................................................................... 43
2.7.1 Western Blot ...................................................................................................... 43
2.8 Microscopy .................................................................................................................. 45
2.8.1 Fluorescence microscopy ................................................................................... 45
2.8.2 Confocal microscopy ........................................................................................... 46
2.8.3 Image analysis ..................................................................................................... 46
Chapter 3 ..................................................................................................................................... 47
Characterization of the subcellular localization of HD2C ........................................................... 47
3.1 Introduction ...................................................................................................................... 48
3.1.1 The impact of localization on regulating protein function ................................. 48
3.1.2 Mechanism of Nuclear localization ..................................................................... 49
3.1.3 Nucleolar localization .......................................................................................... 51
3.1.4 HD2-HD2 interactions in tertiary protein complex ............................................. 52
3.1.5 Hypothesis and aims ........................................................................................... 53
3.2 RESULTS....................................................................................................................... 54
3.2.1 Investigating the subcellular localization of the HD2 family of proteins ............ 54
3.2.2 HD2 proteins form nucleolar localized dimers in planta .................................... 55
3.2.3 A nuclear import-related sequence maps to the C-terminus of HD2C ............... 56
3.2.4 The HD2C nuclear localisation signal is dependent on a KKAK motif ................. 58
3.2.5 Sequence alignment of HD2 gene family homologues reveals conservation of
the critical KKAK motif ........................................................................................................ 58
3.2.6 Nuclear localisation is not a pre-requisite for HD2C dimerisation. .................... 59
3.2.7 HD2C localization is altered in response to abiotic stress .................................. 60
8
3.3 Discussion .......................................................................................................................... 60
3.3.1 Summary ............................................................................................................. 60
3.3.2 All combinations of dimers are possible between HD2C and the HD2 family .... 61
3.3.3 HD2C localization state is dynamic and responds to salt stress ......................... 62
3.3.4 HD2C contains a critical KKAK domain necessary for exclusive nuclear
localization .......................................................................................................................... 65
3.3.5 Nucleolar localization is not tied to nuclear localization .................................... 65
3.3.6 Dimerisation does not require nuclear localisation of HD2C. ............................ 66
3.3.7 Conclusion and future ......................................................................................... 67
Chapter 4 ..................................................................................................................................... 75
Characterization of HD2C interaction with 14-3-3 proteins using Bimolecular Fluorescent
Complementation ....................................................................................................................... 75
4.1 Introduction ...................................................................................................................... 76
4.1.1 Possibility of 14-3-3 interaction with HD2C in Arabidopsis thaliana .................. 76
4.1.2 14-3-3 background .............................................................................................. 77
4.1.3 14-3-3 binding site .............................................................................................. 79
4.1.4 Aims and Hypotheses .......................................................................................... 80
4.2 Results .............................................................................................................................. 81
4.2.1 14-3-3 isoforms bind to HD2C in planta ............................................................. 81
4.2.2 HD2 isoforms bind 14-3-3 epsilon....................................................................... 83
4.2.3 Analysis of HD2C 14-3-3 binding domains .......................................................... 83
4.2.4 Determining the site of 14-3-3 binding to a single AA resolution ...................... 85
4.2.5 HD2C-mutants with disrupted C-terminal 14-3-3 binding do not have a clear
shift in localization pattern ................................................................................................. 87
4.3 Discussion ......................................................................................................................... 88
4.3.1 Summary ............................................................................................................. 88
4.3.2 HD2C does not show preference to 14-3-3 isoforms ......................................... 88
4.3.3 HD2C contains multiple 14-3-3 binding sites ...................................................... 90
4.3.4 Identification of 14-3-3 binding sites on HD2C ................................................... 91
4.3.5 The HD2C 14-3-3 binding site does not correspond to any consensus 14-3-3
binding motif ....................................................................................................................... 93
4.3.6 Conclusions and future ....................................................................................... 94
Chapter 5 ................................................................................................................................... 104
Role of HD2C in salicylic acid and jasmonate response in Arabidopsis thaliana ...................... 104
5.1 Introduction .............................................................................................................. 105
5.1.1 Hormone signalling in plants ............................................................................ 105
9
5.1.2 Possible role of HD2C signalling in response to salicylic and jasmonic acids ... 105
5.1.3 Salicylic acid ...................................................................................................... 106
5.1.4 Jasmonic acid and methyl-jasmonate ............................................................... 107
5.1.5 SA/JA crosstalk .................................................................................................. 108
5.1.6 Aims and Hypotheses ........................................................................................ 109
5.2 Results ....................................................................................................................... 110
5.2.1 Expression of HD2 proteins when exposed to SA, INA and MeJA .................... 110
5.2.2 HD2C expression has no impact on root growth when exposed to SA or MeJA
111
5.2.3 35S:HD2C-GFP plants have a delayed germination response to SA and MeJA 112
5.2.4 HD2C binds TGA6 transcription factor .............................................................. 113
5.2.5 Expression of TGA6 overlaps with HD2C in some tissues and developmental
stages 114
5.2.6 Analysis of the expression of genes controlled by TGA6 in HD2C modified plants
114
5.3 Discussion .................................................................................................................. 115
5.3.1 Summary ........................................................................................................... 115
5.3.2 Plants expressing 35S:HD2C-GFP have altered development in response to SA
and MeJA........................................................................................................................... 116
5.3.3 HD2C binds TGA6 transcription factor .............................................................. 118
5.3.4 Conclusions and future work ............................................................................ 120
Chapter 6 ................................................................................................................................... 129
Final Discussion ......................................................................................................................... 129
6.1 Summary ................................................................................................................... 130
6. 2 Results suggest new model of HD2 action ..................................................................... 131
6.3 HD2C for use in genetically modified plants ................................................................... 132
6.4 C-terminal domain appears to be necessary for protein binding ................................... 134
6.5 Future .............................................................................................................................. 136
Chapter 7 ................................................................................................................................... 137
References ................................................................................................................................ 137
Chapter 8 ................................................................................................................................... 148
Appendices ................................................................................................................................ 148
10
Chapter 1 General introduction
11
1.1 Epigenetics
The nature of a cell is determined by specific patterns of gene expression that result in
the production of a correspondingly specific subset of proteins. This capacity to
differentially regulate the composition of proteins in a cell has provided complex
multicellular organisms with a mechanism to evolve discrete tissue types that are
collectively able to perform diverse functions that would otherwise be impossible for a
single cell to perform (Koltunow, Truettner et al. 1990; Atchley and Hall 1991).
Underlying this process are the determinants of gene expression - gene regulators that
control expression via interaction with chromatin in regions neighbouring the gene
region (DeRisi, Iyer et al. 1997). The most overt manifestation of control is exerted
through transcription factors; proteins that bind to promoter regions and enhance or
repress the binding of transcription machinery necessary for RNA production (Pabo
and Sauer 1992; Kadonaga 1998). This system is relatively complex, and given the size
of a genome is an inefficient method of ensuring that only a subset of genes are
expressed at one time. For example, genomic analysis has revealed that there are over
2000 transcription factors in Arabidopsis (Mitsuda and Ohme-Takagi 2009), required
for both specific and endemic expression control. The presence of all genes in a
genome dictates that specific expression is controlled via a global repression
mechanism that has the specificity of transcription factors in targeting gene regions,
but with a fraction of the energy required to maintain this repression. It has become
evident that such global repression is regulated through processes which have come to
be defined under the developing field of ‘epigenetics’ (Holliday 1990; Jaenisch and Bird
2003).
Epigenetics was originally coined by Conrad Woddington in 1942, who defined the
term as ‘the branch of biology which studies the causal interactions between genes
and their products, which bring the phenotype into being’ (Goldberg, Allis et al. 2007).
Early research was somewhat limited by the scope of understanding of modern
genetics, let alone the technology to sequence and characterize DNA, yet nevertheless
recognized that some phenotypes could not be described by Mendelian genetics alone
(Morris 2001). With the advent of DNA characterization and sequencing, a clearer link
was established between genotype and phenotype. Robin Holliday initially suggested
that DNA methylation was responsible for some level of gene regulation, however this
12
speculation was not experimentally proven until his paper "The heritability of
epigenetic defects" which showed that loss of methylation led to heritable
abnormalities in gene expression (Holliday 1987; Holliday 2006). This work has been
widely cited as the first molecular characterization from which all modern epigenetic
research has been based.
More recently, the definition and field of epigenetic research has widened to be a
bridge between genotype and phenotype. This increase in complexity has led to some
level of confusion over the definition of the term ‘epigenetics’ and indeed the utility of
defining such a rapidly expanding field. A consensus definition was published in 2009
to clarify the term, stating that ‘an epigenetic trait is a stably heritable phenotype
resulting in changes in a chromosome without alterations in the DNA sequence’ (Zhang
2008; Berger, Kouzarides et al. 2009). Despite this, even within the timespan of this
thesis the issue of heritability has been argued against because traditionally epigenetic
processes could not conform to the necessity to be truly heritable (Goldberg, Allis et al.
2007). A number of sources suggest that epigenetic pathways involve DNA
methylation, histone modifications or non-coding RNA. These are summarized below.
1.2 DNA methylation
Briefly, DNA methylation is a covalent modification of DNA characterized by the
addition of a methyl group to cytosine at the 5 position of the pyrimidine ring or
adenine at the 6 position of the purine ring (Razin and Riggs 1980). While there is some
evidence that methylation has a role in DNA stability (Eden, Gaudet et al. 2003), the
overwhelming consensus links DNA methylation to regulating gene expression. On a
global level methylated cytosines comprise ~1% of the genome, thus accounting for
70-80% of all CpG dinucleotides in the genome (Saxonov, Berg et al. 2006). A
significant fraction of non-methylated CpG dinucleotides are positioned in the 5’ ends
of many genes, referred to as CpG islands, and which remain transcriptionally active
when unmethylated (Larsen, Gundersen et al. 1992). Similarly, unmethylated and
transcriptionally active genes have been shown to be silenced following methylation of
the promoter region (Brooks, Harkins et al. 2004).
13
The mechanism of this gene repression is of current interest. Two principle
mechanisms of action are currently favoured. The first is that methylation of the DNA
changes the binding surface, thereby preventing the binding of some gene activators
that would otherwise induce gene expression (Watt and Molloy 1988; Mancini, Singh
et al. 1999). Methylation was shown to prevent binding of the mammalian
transcription factor E2F in this way (Watt and Molloy 1988). This mode of action
appears to be the less predominant form of repression. Rather, the second mode of
action is that DNA methylation marks the DNA as a binding surface for specific methyl-
DNA binding proteins (Watt and Molloy 1988). It is these proteins which mediate a
gene repression response. The Arabidopsis genome contains 12 putative methyl-
target specific methylation patterns to repress gene expression (Ito, Koike et al. 2003;
Bogdanović and Veenstra 2009).
1.3 Non-coding RNA
Non-coding RNA refers to transcripts which are not ultimately translated into proteins.
Since its original classification as ‘junk DNA’, it has become clear that non-coding RNA
constitutes a significant molecular class within a cell (Costa 2008). Many proteins were
shown to bind RNA molecules in a process that determined where their activity was
directed onto specific chromosome locations; including histone deacetlyases (Aufsatz,
Mette et al. 2002), transcription factors (Sittka, Lucchini et al. 2008) and DNA
methyltransferases (Imamura, Yamamoto et al. 2004). Additionally, It is thought that
non-coding RNA is essential to direct sequence specific interactions of a limiting
number of chromatin regulatory complexes with target DNA regions so that regulation
can be relevant in the correct cells, in the correct tissues and at the correct time
(Mattick 2001). There are a number of different mechanism by which RNA can exercise
control on gene regulation. These include RNA editing (Covello and Gray 1989), RNA
interference (Hannon 2002) and regulation of chromatin modifying complexes (Hirota,
Miyoshi et al. 2008).
14
1.4 Histone Modifications
Chromatin is the combination of DNA and protein that is found in the nucleus of a cell.
The fundamental unit of chromatin is the nucleosome, which is comprised of an
octamer of four core histone isomers (H2A, H2B, H3 and H4) around which ~147 base
pairs of DNA are wrapped ‘like beads on a string’ (Noll 1974; Margueron and Reinberg
2010). The role of histones has wider implications than simple organization; rather, the
structure and function of chromatin is regulated by a number of epigenetic
mechanisms as well as DNA repair and replication (Berger 2007). This is controlled via
modification to the unstructured, positively charged N-terminal histone ‘tails’ which
extend from the globular protein that itself interfaces with DNA (Luger and Richmond
1998). These modifications are complex both in number and combination. Currently
there are at least nine known post-translational modifications of chromatin
methylation, acetylation, phosphorylation, ubiquitination, sumoylation, ADP
ribosylation, glycosylation, biotinylation, and carbonylation (reviewed in (Loidl 2004;
Margueron, Trojer et al. 2005))
In the context of transcriptional regulation, histone modifications mediate availability
of cis-regulatory elements of genes to transcriptional machinery. The mechanism by
which this occurs is categorized as either class I or class II. Class I include all
modifications that indirectly regulate chromatin structure by recruiting chromatin
remodelling proteins. Class II include all modifications that directly regulate chromatin
structure by chemically interfering with the nucleosome structure (Kouzarides 2007).
Chromatin acetylation occurs on the N-terminal tails of histone H3 at residues K9, K14,
K18 and K24 and histone H4 at residues K8, K12, K16 and K20 (Grunstein 1997). Unlike
other chromatin modifications, acetylation of histones modifies chromatin structure by
both class I and class II modifications. Its class I mechanism is mediated by bromo-
domain proteins that bind specifically to acetylated histones in a similar mechanism to
that described for histone methylation (Zeng and Zhou 2002). Its class II modification
operates through the addition of acetyl groups to neutralize the positive charge of
histone tails and decreases their affinity to DNA (Eberharter and Becker 2002). This
leads to a loose chromatin structure, thereby providing access to the DNA of
transcriptional regulators. It is the competing actions of two classes of enzymes that
15
control this modification, histone deacetylases which remove acetyl groups to histones
and histone acetyl transferases which add acetyl groups to histones.
1.5 Plant specific HD2 family
The HD2 family of enzymes was first characterized as a high molecular weight complex
abundant in maize embryos by Lusser and colleagues in 1997 (Lusser, Brosch et al.
1997). Maize embryos at the time were known to contain four different isoforms of
histone deacetylase, named HD1-A, HD1-B, HD1-BII and HD2 (Brosch, Lusser et al.
1996). HD2 was discriminated as different from the HD1 isoforms based on the
biochemical fractionation and purification techniques at the time and this simple
nomenclature has held for this family of enzymes. Since that time the significance of
the classification has become evident, as bioinformatic analyses have indicated a
separate path of origin from the more ubiquitous HD1 family. Whereas the
RPD3/HDA1 and SIR2 family of enzymes are present in almost all eukaryotes and are
derived from a common enzyme, the HD2s appear to be present only in plants and are
more closely related to the PPIases. It was subsequently suggested that the HD2s and
PPIases share a common ancestral enzyme origin, with the enzymatic histone
deacetylase activity of the HD2s being an example of convergent evolution rather than
diverging from the more common HDACs (Aravind and Koonin 1998).
1.6 Structure of HD2
Currently there is no published x-ray crystallography data for any member of the HD2
family, with structural data so far derived from bioinformatic predictions and analysis
of the primary and secondary structure. Early sequence analysis identified two
important structural domains- the N-terminal catalytic domain and a central acidic
domain, which appear to be present in all HD2 homologues (Wu, Tian et al. 2000). It
was subsequently shown that a C-terminal zinc-finger domain is present in the HD2A
and HD2C isoforms as well as their orthologous proteins in maize and rice (Dangl,
Brosch et al. 2001). The lack of this domain in the HD2B and HD2D isoforms suggests
that there are functional differences between the Arabidopsis isoforms.
16
N-terminal catalytic domain
The HDAC catalytic domain is defined as the minimal region required to catalyze the
cleavage and removal of acetyl groups from histone residues. In assays linking
Arabidopsis HD2A deacetylase activity to the expression of the GUS reporter gene, it
was shown that a region between amino acids 1-162 is both sufficient and necessary
for repression of GUS expression (Wu, Tian et al. 2003). The low resolution of this
deletion assay does not preclude the existence of a more minimal catalytic domain.
From early bioinformatic surveys it was noted that there was a putative critical
histidine 25 and aspartic acid 69 residues which may be involved in the catalytic
process (Aravind and Koonin 1998). A subsequent study tested this hypothesis in the
AtHD2A isoform, using site directed mutagenesis to determine the catalytic
importance of the two residues (Wu, Tian et al. 2003). It determined that while
mutation of the histidine to alanine caused a statistically significant decrease in
deacetylase activity, the aspartic acid mutation had no effect. Moreover, an N-terminal
MEFWG motif is completely evolutionarily conserved and deletion of this motif caused
inactivation of the enzyme’s desacetylase activity. Together this suggests that the N-
terminal of the protein is required for HDAC activity, with the N-terminal MEFWG
motif essential and H25 important for the gene repression activity of AtHD2A.
Central acidic region
The extended central acidic domain is present in all HD2s. It is structured as two
distinct acidic stretches separated by a short cluster of basic amino acids (Dangl,
Brosch et al. 2001). In HD2A the acidic domain stretches between residues 110-206AA
and contains a high density of Aspartate and Glutamate residues which are principally
responsible for the acidic isoelectric point for this protein. In addition, the
concentration of these residues to a short stretch in the middle of the HD2s means
that there is a dense region of highly charged protein at the surface of the protein.
The precise role or function of this region has not yet been elucidated. Similar acidic
domains in other nucleolar proteins such as nucleolin offer possible insight into its
function. Nucleolin is a multifunctional phospho-protein which shifts between the
cytoplasm, nucleus and nucleolus (Ginisty, Sicard et al. 1999). It contains a similarly
acidic charged domain which functions as a highly mobile binding surface for proteins
17
such as histone H1 and various ribosomal proteins (ERARD, Belenguer et al. 1988;
Tajrishi, Tuteja et al. 2011). Given the fact that the positively charged histone tails are
the substrate preference for HDACs, the negatively charged central domain of the
HD2s would logically provide the ideal binding surface. Further clarification of the link
between the central acidic domain and its histone substrate is clearly required.
C-terminal zinc-finger domain
Zinc-fingers are small (~30AA) protein domains common to eukaryotes which have a
well-defined function for binding to DNA, RNA and proteins (Klug and Rhodes 1987;
Laity, Lee et al. 2001). Structurally they are characterised by a series of cysteine and
histidine residues within the protein’s sequence which are coordinated in complex
with a zinc ion (Lee, Gippert et al. 1989). The zinc ion is largely responsible for
mediating and maintaining the protein fold, which itself mediates interaction with
DNA, RNA and proteins. Zinc fingers comprise a large motif family, and are arranged in
classes; separated by the arrangement of cysteine and histidine residues relative to the
complexed zinc ion. This number of cysteine and histidine residues and the
composition of the surrounding protein sequence has a significant impact on the
three-dimensional shape of the motif (Kim and Berg 1996). Because the protein’s fold
is essential for mediating contact between the protein and its target, each fold
typically preferentially binds to one of DNA, RNA or protein as targets (Laity, Lee et al.
2001).
The putative C-terminal zinc-finger domain of AtHD2A and AtHD2C appears to be an
evolutionarily conserved motif specific to the HD2s. Dangl et al compared 8 isoforms of
HD2 from maize, rice and Arabidopsis and showed conservation of this motif in 6 of
the 8 homologues (Dangl, Brosch et al. 2001). The motif was not present in the related
prokaryotic PPIases, which together with other HD2s lacking this moiety, suggests that
it is an evolutionarily recent acquisition adapted for specific HDAC related function.
The zinc-finger motif present in the HD2s conforms to the TFIIIA-type zinc fingers
originally characterized in the species Xenopus with the transcription factor IIIA (Klug
and Rhodes 1987). Throughout eukaryotes it has been identified as a DNA-binding
motif, with each zinc-finger mediating interaction with ~4bp of DNA, centred within
the major groove of the double helix (Theunissen, Rudt et al. 1992). A significant
18
aspect of this DNA:protein interaction is that multiple zinc finger domains are arranged
to stabilize the interaction. This has led to the speculation that HD2s do not directly
interact with DNA through this zinc finger domain, but rather that it mediates various
protein interactions (Dangl, Brosch et al. 2001). However there is some precedent to
suggest that this may not be the case. In Drosophila, the transcription factor GAGA is
the only protein that contains a single C2H2 domain. Together with its surrounding
basic amino acids it is able to provide a stable DNA:protein interaction (Pedone,
Ghirlando et al. 1996; Iuchi 2001). Similarly, the Arabidopsis transcription factor
SUPERMAN binds DNA through its single zinc-finger domain with a similar
arrangement of basic amino acids (Dathan, Zaccaro et al. 2002). Together this indicates
that although rare, the direct interaction of HD2s that contain a zinc finger with DNA is
possible and requires further investigation. Moreover, zinc-finger specificity for
methylated DNA has been reported in human proteins, giving rise to the possibility of
further crosstalk between the DNA methylation and histone acetylation pathway
(Sasai, Nakao et al. 2010).
1.7 HD2 expression patterns
The initial characterization of HD2s was carried out in germinating maize embryos as
the investigators noted that this tissue was ‘a source particularly rich in these enzymes’
(Lusser, Brosch et al. 1997). The presence of these epigenetic effectors being so
prominently expressed in germinating tissue is significant as it suggests that they play a
prominent role in the developmental pathway between seed and seedling. Further
characterization has revealed that HD2 expression is not limited to embryogenesis and
is responsive to various developmental, tissue specific and stress induced pathways
(Hollender and Liu 2008).
Tissue specific and developmental expression patterns
A number of studies have concentrated on the expression profiles of HD2s in an effort
to gain insight into where and when HD2s function. Micro-array data comparing HD2
expression throughout development reveal that HD2s are expressed in all tissues,
although induction is most evident during early and late flowering (Hollender and Liu
2008). Semi quantitative RT-PCR appears to be consistent with these observations,
19
with expression detected in stems, leaves, flower, roots and seedlings for HD2A-C, with
higher expression found in the stem, flower and seedling stages (Zhou, Labbe et al.
2004). It has been suggested that the highly similar expression profiles between the
four Arabidopsis homologues is suggestive of functional redundancy (Hollender and Liu
2008).
In situ hybridization assays were used in the same study to reveal spatial expression
patterns to a higher degree of qualitative resolution (Wu, Tian et al. 2003). It revealed
that expression was evident at all levels in the organs examined, but the highest levels
of accumulation were found in the ovules, embryos, shoot apical meristem and first
leaves. They noted that this pattern of expression is consistent with other genes that
control embryogenesis, such as WUSCHEL and SERK which control flowering through
their action as transcription factors (Jönsson, Heisler et al. 2005; Pérez-Núñez, Souza et
al. 2009). To test the link between HD2s and embryogenesis, they overexpressed a
BBM gene leading to the formation of somatic embryos on transgenic Arabidopsis
cotyledons. This led to an induction in the expression of HD2s, suggesting that there is
a clear link between the HD2 gene expression and somatic and zygotic embryogenesis
(Zhou, Labbe et al. 2004).
In results consistent with this, a specific study in the expression of genes in the ovules
of the plant Solanum chacoense identified the ScHD2a gene in a negative selection
screen to so-called ‘restricted areas’ (Lagace, Chantha et al. 2003). This expression was
triggered by fertilization and appears to correspond with induction of other genes
involved in gene repression such as ScP18 and ScSWlb, suggesting that it played a role
in the transition from fertilized germoplasm to seed development
Environmental induction of HD2 expression
The sessile nature of plants necessitates a dynamic expression pattern that quickly
responds to challenges imposed by the environment. The ability for HD2s to globally
repress the genome in a specific manner suggests that it will be highly responsive to
such challenges, either at the level of transcription or post-translation. Linking the
expression profile of HD2 to various environmental stimuli remains in its infancy with
only a few studies published to date. Typically, environmental response pathways are
20
mediated by various hormones in specific signal transduction cascades that ultimately
change cell expression (Moore 1979; Davies 2010).
In barley the expression of the HD2 isoform HvHD2AC2-1 were measured in response
to treatment with jasmonic acid, abscisic acid and salicylic acid using quantitative RT-
PCR in seven day old seedlings that were either wild type or transgenic lines
overexpressing the HD2 isoform (Demetriou, Kapazoglou et al. 2009). It revealed a
strong induction in expression following JA treatment, a differential response between
the two over-expressing lines in response to ABA treatment while no conclusion could
be drawn following SA treatment. In rice a similar result was obtained, showing a
strong induction following JA treatment and repression following ABA. This led the
author to suggest that there are functional similarities in gene expressions between
monocots.
In Arabidopsis the relationship between HD2C and the ABA pathway have been well
characterized. ABA is a hormone that is primarily involved in drought and salt stress
response (Xiong, Schumaker et al. 2002), inhibition of seed germination (Penfield, Li et
al. 2006) and biotic attack (Hirayama and Shinozaki 2007). In Arabidopsis plants
treatment with ABA was shown to repress HD2C expression using RT-PCR (Sridha and
Wu 2006). In subsequent analyses all HD2 isoforms were shown to have a repression
of HD2 expression in response to ABA and NaCl treatment, suggesting a co-regulatory,
ABA dependent cis-element may be conserved within this family (Luo, Wang et al.
2012).
1.8 HD2 function
The function of HD2s has not yet been fully elucidated either at the molecular, cellular
or whole plant level. The molecular catalysis of histone deacetylation has previously
been reviewed and the lack of knowledge into its enzymatic properties is clearly a
signal for future research to be directed into this field. More recently studies have
focused on forward and reverse genetic screens of the protein’s function, both
because this provides insight into the molecular process of the enzyme and because of
the significance of understanding a critical player in the epigenetic pathway that has a
process endemic to eukaryotes, yet evolutionarily specific to plants (de Ruijter, Van
21
Gennip et al. 2003). From these analyses, critical functions in the context of
development and environmental response have been determined.
Development
It has been well established that HD2 is involved in development of reproductive
tissue. Its initial characterization resulted from the observation of its high level of
expression in maize embryonic tissue, suggesting that it either maintains the seed
state or drives the transition to seed germination of the plant (Brosch, Lusser et al.
1996). Additionally, a number of reverse genetic screens have demonstrated that HD2s
are involved in development, specifically of reproductive tissue. HD2A down-regulation
by RNAi resulted in an aborted seed development phenotype and led the team to
conclude that its function was required for seed to seedling transition (Wu, Tian et al.
2000). Subsequent T-DNA knockout experiments have not shared this specific
phenotype, although it has been suggested that due to sequence redundancy the RNAi
was effectively knocked down expression of all HD2s, leading to a general reduction in
the pool of HD2s in the plant (Colville, Alhattab et al. 2011). The combination of these
results suggest two probable hypotheses; firstly that HD2s are together involved in the
seed formation process and secondly that there is a level of functional redundancy
existing within this family which allows HDAC activity to proceed in the absence of a
single isoform.
A study into Solanum chacoense, a model species for research in fertilization, showed
that an HD2A orthologue was strongly induced in specific ovular regions following
fertilization (Lagace, Chantha et al. 2003). Moreover, this was a highly tissue specific
accumulation only evident in in situ hybridization, whereby ovular expression was
heightened with no difference in HD2 levels in closely surrounding tissue. They
rationalized the formation of siliques with a seedless phenotype which was observed
in the HD2 knockdown plants was related to the absence of this specific surge of
expression in the ovules integument.
Stress response
There is a current concerted effort to identify and manipulate pathways related to
plant stress response because such environmental stresses directly influence the yield
22
of all cropping plants (Berman and DeJong 1996; Peng, Huang et al. 2004; Schijlen, Ric
de Vos et al. 2004; Gill and Tuteja 2010). The capacity for yield gain and loss from year
to year has long been correlated to the environmental stresses faced by plants
annually (Peng, Huang et al. 2004). Traditionally this has been best correlated to
rainfall; however other factors that are increasingly relevant include salinity,
temperature and mechanical stresses that may be caused by wind or precipitation.
In plants there are a number of models showing that the acetylation state of histones
is linked to a plant response to abiotic stresses such as cold, salt and drought (Bae, Cho
et al. 2003; Kim, To et al. 2008; Chen, Luo et al. 2010). The HD2C protein has been
implicated in plant stress response related to salt, drought and ABA (Sridha and Wu
2006). Transgenic lines containing an HD2C-GFP construct conferred a decreased
sensitivity to the inhibition of root growth by NaCl and a corresponding increase in
germination rate when grown on 100mM NaCl conditions. Additionally, over-
expressing lines had an observed partial stomatal closure which correlated to reduced
transpiration and increased drought resistance. This was confirmed by the same group
in subsequent studies where they used T-DNA insertion HD2C knockouts to show a
conversely sensitive phenotype to ABA and NaCl, suggesting that this is a novel
regulator of ABA and salt stress response in Arabidopsis (Luo, Wang et al. 2012).
Recognizing that there was likely to be a link between the ABA resistant phenotype
present in HD2 over-expressing transgenic lines and the ABA response genes that
mediate the hormones signal transduction chain, the researchers analysed the
expression of the response genes in the transgenic plants. Using RT-PCR, they found a
notable up-regulation of LEA class genes which are expressed during seed
development and are implicated in the protection of cellular dehydration.
Contrastingly, there was a down regulation of genes related to water loss such as
ADH1, KAT1-2 and SKOR. While implicating the action of HD2C on these gene regions,
the researchers did not explore further to determine whether HD2C directly mediates
deacetylation of histones at their specific loci (Luo, Wang et al. 2012).
The recognition that HD2 is necessary for the development of reproductive tissue and
that the isoform HD2C is implicated in a stress response role of the plant is significant.
However, it is clear that further identification of the precise regions of genomic DNA
23
that HD2 acts upon is necessary to develop true insight into the mechanism by which it
uses that ultimately results in the observed phenotype.
1.9 HD2s- functionally redundant?
The issue of the level of functional redundancy between isoforms is currently being
determined from several groups. The discrepancy seen between the RNAi and
knockout of AtHD2A previously described has led some to suggest that this is evidence
that redundancy between isoforms is significant (Wu, Tian et al. 2000). Similarly there
does not appear to be a significant difference between the expression patterns of
HD2A-C. However several lines of evidence suggest that this is not true. Firstly,
differences were observed between HD2A and HD2C in the seed germination rates on
glucose, ABA and NaCl (Colville, Alhattab et al. 2011). Secondly, developmentally
abnormal phenotypes seen in the AtHD2A plants were not evident in the AtHD2C
plants, suggesting some deviation in function between the two isoforms. Lastly, a
significant structural difference between isoforms HD2A/C and HD2B/D is the presence
of a zinc-finger motif that may influence a regulatory role between either DNA or
protein binding (Lawrence, Earley et al. 2004). This significant structural moety is likely
to impart a specific subset of processes which suggest a diversification of functions
between isoforms.
1.10 The HD2 complex- determining a higher level protein interactome
‘Interactomics’ describes the study of how molecules are functionally linked in a cell or
organism, and the outcome that these interactions play on the organism’s homeostasis
(Collura and Boissy 2007). Whereas specific functions may be uniquely attributed to a
single protein or protein family; it is inevitably bound by a multitude of other proteins
that control some aspect of its activity. Thus any single protein will itself have an
‘interactome’ defined by the subset of proteins that it binds to, which itself is
dependent on the cell expression state at the time (Marras and Capobianco 2008). The
identification of a protein’s interactome is important when trying to determine its
higher function. Significant insights into the core functions of HD2s have been
determined based on the identification of their interacting partners, either confirming
24
hypotheses based on extrapolations of mammalian HDACs, or by raising new and novel
hypotheses.
A significant feature of mammalian HDACs is that the catalytic activity of the enzyme is
directed to specific substrates by other proteins. A HD2 homologue from longan fruit
was shown to interact with ERF1 using a combination of yeast 2-hybrid assay and
bimolecular fluorescent complementation (Kuang, Chen et al. 2012). ERF1 belongs to
the ethylene response factor (ERF) family of proteins and constitutes one of the largest
transcription factor families in plants. Specifically ERF1 binds to the GCC box of
ethylene regulated promoters. This study focused on the implication of HD2 co-
operation with ERF1 in the context of fruit senescence. However the wider
implications of HD2 interaction with transcription factors is in line with a model
whereby histone deacetylase activity is directed to specific promoters through client
protein led specificity.
Luo and colleagues identified and characterized the interaction between HD2C and
HDA6 using a combination of BiFC, co-imunoprecipitation and pull-down analysis to
suggest that these proteins interact physically with one another (Luo, Wang et al.
2012; Luo, Wang et al. 2012). This interaction was used to rationalize the ABA and NaCl
response seen in HD2C knockout plants also used in this study. The ABA and NaCl
response in plants with altered HD2C expression had previously been identified by this
group, where plants over-expressing an HD2C-GFP construct were resistant to
germination on media with high salt or ABA. They subsequently showed interaction
using the BiFC assay between all HD2 isoforms and both HDA6 and HDA19. Taken
together, these results suggest that there is some level of overlap between the
pathways affecting plant specific and non-plant specific histone deacetylases.
Another important aspect of HDAC activity that has been identified in mammalian
systems is the cross talk between various epigenetic pathways. As previously
mentioned, DNA methylation has been shown to mark specific regions of DNA for
methyl-DNA binding proteins, which themselves mediate the formation of a chromatin
remodelling complex. Several mammalian HDACs have been shown to be involved in
such complexes; indeed the condensation of chromatin is an important step in down-
regulating gene expression via the epigenetic pathway. HD2s have been implicated in
25
this process following the discovery that they bind a DNA methyltransferase in
Arabidopsis (Song, Wu et al. 2010). Here it was shown to bind HD2C, with the binding
shown to affect the DNA methyltransferase activity.
The activity of HDACs has previously been shown to be dependent on post-
translational modifications. Indeed, the HD2 isoform characterized in the initial study it
was shown to have its catalytic activity determined by the phosphorylation state
(Lusser, Brosch et al. 1997). In a pull-down analysis it was identified that putative 14-3-
3 binding proteins, isoforms HD2A-C were identified by mass spectrometry to be co-
precipitated with 14-3-3 epsilon (Paul, Liu et al. 2009). These results are not
confirmation of interaction as they are not performed in planta, and do not match the
requirement of utilizing multiple technologies to identify interaction. 14-3-3 proteins
bind to specifically phosphorylated serine or threonine residues and have been
characterized to recognize specific consensus motifs termed motif 1-3 (DeLille, Sehnke
et al. 2001; Johnson, Crowther et al. 2010). While no HD2 contains such a motif, it
suggests the possibility that it binds to the phospho-HD2 at a novel site and may
mediate HD2 catalytic activity.
The identification of novel HD2 interactions has thus yielded significant insight into the
function and mechanism of HD2 activity that may otherwise not be easily
characterized. It is likely that further core functions of this family of enzymes may be
discovered by uncovering and characterizing other interacting partners. Such novel
interactions would thereby provide valuable insight both into the core function of the
enzyme, as well as providing a biological context in which this function is activated.
1.11 Project hypothesis and aims
1.11.1 Hypothesis
After the initial biochemical analysis of enzyme activity to identify this gene family as a
histone deacetylase enzyme, research has been redirected to molecular biological
studies to identify their biological role in the context of gene regulation. Despite a
growing body of knowledge being made in this field, research has so far concentrated
on developing insight into the role of the proteins using phenotypic marker analyses.
Here changes at the plant level during developmental, temporal, spatial and
26
environmental conditions using wild type and mutant lines reflect the action of HD2
and thus implicate their action in a variety of roles (Wu, Tian et al. 2000). As previously
identified, this has been used to great effect to show that these are down-regulators of
gene transcription with principle roles in development, stress response, and as
negative regulators of elicitor-induced cell death.
From the studies performed so far it is clear that there are a number of aspects still to
be considered relating to the function of HD2 proteins. Most overtly, and the overall
aim of this project, is to identify the mechanism of their regulation in the context of
gene regulation. In mammalian histone deacetylases this has been more widely
researched and it has become clear that as in most biological systems, these enzymes
do not work in isolation. Rather, activity is brought about through the combinatorial
efforts of many different factors. These contribute to the sequestration of the protein
to the subcellular location of its activity, the regulation of this enzymatic activity and
the direction of enzyme activity to specific targets within this location.
While it has been shown that these gene families have arisen from different
evolutionary pathways, there are several lines of evidence supporting the hypothesis
that there are shared regulatory mechanisms involved between mammalian histone
deacetylases and HD2.
1. Phosphorylation has been shown to regulate both nuclear-cytosolic localization
and enzymatic activity in mammalian HDACs. Similarly, it was shown that only
the phosphorylated HD2 is active (Lusser, Brosch et al. 1997), suggesting a
possible role in modulating the enzyme’s activity. Furthermore, mammalian
HDACs have been shown to be regulated at these phosphorylated residues by
14-3-3 proteins. In a recent co-immunoprecipitation screen of 14-3-3 binding
targets in Arabidopsis thaliana, HD2A-C have been listed as putative targets,
suggesting that these phosphorylation sites are also targeted by 14-3-3s.
2. Both homo-dimerization and hetero-dimerization between HDACs in mammals
have been shown to regulate protein activity and the gene targets that they act
on (Fischle, Dequiedt et al. 2002). Such dimerization events have been
speculated on since the enzyme was purified to homogeneity yet ran on an SDS
gel with an approximate size of 400kDa (Lusser, Brosch et al. 1997).
27
3. DNA and histone modification enzymes are rarely directed to their gene targets
in isolation; rather DNA-binding factors take the enzyme to specific sequences
that they recognize and allow modification at these sites. Such a mechanism
has not been proven in the HD2 gene family; however the presence of a single
zinc-finger domain in HD2A and HD2C has led other groups to speculate that
this may be a site of protein binding rather than direct DNA binding. A clear
hypothesis would therefore be to suggest that this is a docking site to allow
general DNA binding proteins to regulate the DNA-histone/HD2 interaction.
1.11.2 Project aims
From the above hypothesis it is clear that the current hole in understanding into the
regulatory mechanism of the HD2s is both present and easily filled. This project
therefore aims to identify the regulation underlying this enzyme at three levels; the
subcellular localization of the protein, control of the enzymatic function and the
prospect of other proteins directing the HD2s to specific genes under defined
biological conditions. The general aims are listed below:
1. Determine the capacity for the four Arabidopsis HD2 proteins to homo- and
hetero-dimerize in planta and subsequently determine the region/s necessary
for dimerization.
2. Analyse the effect that nuclear import may have on the ability of HD2s to
dimerize in the cell.
3. Determine the capacity for the four Arabidopsis HD2 proteins to bind 14-3-3
proteins in planta. Use this same method to determine whether all 14-3-3
isoforms have the potential to bind HD2.
4. Determine the HD2 residue/s necessary for 14-3-3 binding and conduct
enzymatic and phenotypic tests to analyse the impact that destroying this site
may have on the protein’s function.
5. Use bioinformatics, predictions and publications to identify novel proteins that
bind to HD2C. These will then be confirmed using BiFC to test whether this
interaction is conserved across the gene family.
28
By the conclusion of this project there will be a further understanding into the field of
the HD2 proteins, specifically with regard to their regulation. Furthermore, it will
develop an insight into the role that HD2 has on regulating gene expression at both the
cellular and holistic level which has so far been outside of the scope of the previous
studies in this field.
29
Chapter 2 Materials and Methods
30
2.1 Materials
Materials Supplier
Acrylamide Amresco
Agar, Bacteriological Amresco
Agarose, Molecular grade Bioline
Ammonium nitrate Univar
Ammonium persulfate Sigma-Aldrich
Antibodies (HA-probe/ c-myc) Santa Cruz Biotechnologies
Bromophenol Blue (BPB) Bioline
Calcium chloride AnalaR®
D-glucose (anhydrous) Univar
Dithiothreitol (DTT) Bioline
EDTA disodium salt Univar
Ethanol Chem-Supply
Glycerol Univar
Glycine Amresco
HEPES AppliChem
Intercept® 70WG Scotts
Isopropanol Asia Pacific Specialty Chemicals
limited
Magnesium chloride Univar
Magnesium sulfate Univar
Manganese (II) chloride Univar
MES free acid monohydrate Amresco
Methanol Univar
3-(N-morpholino)propanesulfonic acid
(MOPS)
Sigma-Aldrich
Oligonucleotides Integrated DNA technologies
Phenylmethanesulfonylfluoride (PMSF)
Poly (unylpolypyrrolidone) PVPP Sigma-Aldrich
Potassium acetate AnalaR®
31
Potassium chloride Sigma-Aldrich
Potassium dihydrogen phosphate
(Anhydrous)
Amresco
Potassium hydroxide Univar
Restriction endonucleases New England Biolabs
Rubidium chloride Sigma-Aldrich
SDS Amresco
Silwet L-77
Sodium chloride Univar
Sodium hydrogen phosphate (Anhydrous) Amresco
Sucrose Amresco
TEMED Sigma-Aldrich
Tris Amresco
Tryptone Amresco
Tween 20 Sigma-Aldrich
Yeast extract, Bacteriological Amresco
32
2.2 Methods
2.2.1 General methods
2.2.1.1 Centrifugation
Centrifugation steps were performed in a bench-top microcentrifuge (Eppendorf,
Germany) at 16000xg and room temperature unless otherwise specified.
2.2.1.2 Autoclaving
Autoclaving was performed in a Gentinge lab steriliser autoclave, and occurred at a 15
minute incubation at 121 degrees Celsius under pressure of 30psi unless otherwise
specified.
2.2 Plant growth and transformations
2.2.1 Sterilizing seeds
Arabidopsis thaliana seeds used to sow onto MS media were surface sterilized to
ensure that the competing growth of fungi and bacteria did not affect the germination
or growth of seedlings. Arabidopsis thaliana ecotype Columbia 0 seeds were pooled
into 20mg sets and surface sterilized by incubation in 1mL sterilization solution for six
minutes with periodic inversions. The seeds were moved to a laminar flow and allowed
to settle under gravity, with the supernatant subsequently discarded. Seeds were
washed twice with 1mL of 95% ethanol. Following the second wash, the supernatant
was removed by micropipette and the remaining seeds dried at room temperature for
approximately four hours with the microfuge tubes cap off. Surface sterilized seeds
were kept at room temperature and under dark conditions in the sterile
microcentrifuge tube until required.
2.2.2 Seed sowing
Seeds were sown either onto MS agar or into compost, whereas Nicotiana
benthamiana seeds were sown only into compost.
MS agar: MS media was prepared (Murashige and Skoog basal salt mixture, 2.15g/L;
2mM MES; pH 5.8; 0.8% agar) and sterilized seeds scattered on top to a density of
~200 seeds/plate. Plates were then wrapped in tinfoil to maintain darkness and
33
stratified for 2 days at 4 degrees Celsius to facilitate homogeneous germination. After
stratification, seeds were placed in a growth cabinet at 16 hour day/night cycles (~80-
100μE/cm2, cool, fluorescent white light) and 22°C.
Compost mix was prepared and transferred into either 12cm pots (floral dip) or pot-
trays with 5cmx5cm cavities. These were then placed in cat-litter trays filled with a
solution containing three liters of water and 100mL Presept insecticide solution. These
were kept in the solution for two hours to allow the solution to penetrate the
compost. Pools of 20-30 seeds were then scattered onto the compost and pressed
lightly into the soil. In the instance of floral dip, the tops of the pots were covered and
bound in fly screen. Pots were further covered in Saran (plastic) wrap to prevent drying
out of the compost and in tinfoil to keep seeds dark. Potted seeds were stratified for 2
days at 4°C and then moved to a growth cabinet where they were grown under 12
hour day/night cycles at 22°C and ~80-100μE/cm2, cool, fluorescent white light.
2.3 Transformation of A. thaliana leaves
2.3.1 Floral dip
Healthy pots of ecotype Col-0 Arabidopsis thialiana seeds were densely grown in pots.
Fly screen was fastened to the top ensuring that later inversion of the pots would not
cause plants or soil to fall out. After ~5 weeks, the first bolts of influorescence were
removed to encourage a greater subsequent proliferation of flowers.
At this time, A. tumefaciens strains carrying the genes of interest (section ) were grown
in 200 mL of selective LB broth containing Kanamycin (50μg/mL), Rifampicin (50μg/mL)
and Gentamicin (5μg/mL) overnight at 30°C. The overnight bacterial cultures grown
were centrifuged at 5000xg and 4 degrees Celsius and resuspended in 200 mL of 5%
sucrose solution to give a final optical density OD600 of 0.8. Prior to dipping in the case
of Arabidopsis, Silwet L-77 solution was added to the sucrose solution to give a final
concentration of 0.05% (v/v) and mixed thoroughly by repeat inversions. Silwet L-77
allows the lowering of surface tension of aqueous solutions and thus allows for a more
effective infiltration of the protein-vector construct of interest.
34
The prepared Arabidopsis plant was then dipped into the A. tumefaciens sucrose
solution and gently agitated for 10 seconds. Care was taken to ensure all inflorescence
and rosettes were immersed in the solution before returning the plants to a tray in the
growth cabinet covered under commercial ‘sandwich bags’ for 24 hours to maintain
high humidity. Plants were watered normally before a second round of floral dip was
performed six days after the first procedure. Following the second floral dip, loose
bolts were covered with small paper bags and taped to collect the seeds.
2.5 Bacterial preparations and transformations
2.5.1 Competent cell preparation (E.coli)
E.coli DH5 cells were streaked from glycerol stocks on LB agar plates to obtain single
colonies and incubated overnight at 37°C. A single colony was picked and grown
overnight at 37 °C in 4mL LB broth. A 1mL aliquot from this culture was used to spike
100mL of LB broth, which itself was grown to an OD600 of 0.4-0.6. Once reached, cells
were pelleted by centrifugation (Beckman-Coulter, Australia) at 5000g for five minutes
at 4°C. The supernatant was removed and the cell pellet gently resuspended in 40mL
of ice cold TFBI solution at 4°C. Cells were then pelleted by centrifugation at 5000g for
five minutes and 4°C. The supernatant was again removed and the cells resuspended
in 4mL ice cold TFBII solution and left on ice for ten minutes. Cells were then either
used immediately or alternatively snap frozen in liquid nitrogen and stored at -70°C
until needed.
2.5.2 Competent cell transformation (E.coli)
Competent E.coli DH5 cells were obtained either directly following preparation or
from stocks stored at -70 °C which were kept on ice until partially thawed. Further
manipulations of competent cells were performed on ice to ensure maximum
transformation efficiency. A ~5ng sample of plasmid DNA was added to 50μL of
competent cells, mixed gently and incubated on ice for 30 minutes. Following
incubation, the cells were heat shocked at 42 °C for 45 seconds and placed
immediately back on ice. LB media (500μL) pre-warmed to 37 °C was added to the
transformation and gently mixed by inverting three times. Cells in LB medium were
35
then incubated at 37°C for 45 minutes on a shaking incubator at 250rpm for post-shock
recovery. After incubation, 50μL of transformed cell suspension was streaked out on
an LB agar plate supplemented with 50μg/mL kanamycin. The remaining 450μL of the
culture were centrifuged for ten seconds and 400μL of the supernatant was removed.
The cells were resuspended in the remaining 50μL of the supernatant and streaked
onto a second LB plate as described above. Bacteria were allowed to grow overnight at
37°C until colonies had grown to >2mm in size.
2.5.3 Agrobacteria competent cell preparation
A glycerol stock of A.tumefaciens cells was streaked onto selective LB agar plates
containing Rifampicin 50μg/mL and Gentamicin 5μg/mL. These plates were incubated
at 30°C for three days until colonies had grown to >2mm. A single colony was used to
inoculate 3mL of LB medium containing Rifampicin 50μg/mL and Gentamicin 5μg/mL.
The cell suspension was incubated overnight at 30°C with 250rpm overnight shaking.
Of the resulting culture, 100μL was used to inoculate 50mL of selective LB medium
containing Rifampicin 50μg/mL and Gentamicin 5μg/mL. The cell suspension was
incubated overnight at 29°C with 250rpm overnight shaking until the OD600 reached
0.75. The cell culture was then chilled on ice for 10 minutes before centrifugation at
3000 x g for 10 minutes at 4°C. The supernatant was discarded and the pellet of
A.tumefaciens cells was rinsed and gently resuspended in 1mL of ice-cold 20mM CaCl2
solution, followed by another centrifugation to remove the excess antibiotics. The
resulting cell pellet was resuspended in ice-cold 20mM CaCl2 (1 mL per 50 mL culture)
and transferred in 0.2mL aliquots to sterilised microfuge tubes. The tubes were
subsequently snap-frozen in liquid nitrogen and either used immediately or stored at -
70°C.
2.5.4 Agrobacteria competent cell transformation
Transformation of A.tumefaciens competent cells was performed following validation
of plasmid DNA by restriction mapping and DNA sequencing. A 50μL aliquot of
competent cells was used either directly after preparation or from storage at -70°C. To
the cells, ~20ng of plasmid DNA was added and mixed gently by agitation. Plasmid DNA
36
was inserted into the cells by successive 5 minute incubations of the DNA-cell mixture
on ice, in liquid nitrogen and at 37°C respectively. After the final incubation 400μL of
LB broth was added and the cells were recovered by a two hour incubation at 30°C
with orbital shaking at 250rpm. The resulting transformants were plated onto selective
LB agar plates containing Kanamycin (50μg/mL), Rifampicin (50μg/mL) and Gentamicin
(5μg/mL) in 50μL and 200μL aliquots for incubation at 30°C. Successful transformations
should yielded >50 colonies after three days of incubation at 30°C.
2.5.5 A.tumefaciens infiltration of Nicotiana benthamiana leaves
A single colony of A.tumefaciens containing the vector construct of interest (section
2.2.4.4) was inoculated in 5mL of LB medium containing Kanamycin (50μg/mL),
Rifampicin (50μg/mL) and Gentamicin (5μg/mL) and cultured overnight at 30°C with
constant shaking at 250rpm. The resulting culture was centrifuged the following day
and the supernatant removed. The pelleted cells were resuspended in 1.5mL of
infiltration buffer and the OD600 measured. Typical measurements were within the
range of 1.5-3. Ideal infection and subsequent transient transformation is observed
where OD600 is within the range of 0.6-1, thus the cells were diluted to give a
calculated OD600 of 0.75 in 1mL of cell mixture. Acetosyringone was added to a
concentration of 200µM and incubated on ice for an hour. Working suspensions were
prepared by combining the appropriate clones containing the two putatively
interacting proteins, as well as the p19 plasmid at a 1:1:1 ratio. The mixed
A.tumefaciens strains were co-infiltrated into the abaxial air space of N.benthamiana
leaves using a 1mL syringe. Infiltrated plants were returned to the growth cabinet and
grown under long day conditions for 3-4 days before leaf samples were prepared and
observed for protein interaction in the lower epidermal cell layer as indicated by
fluorescence.
2.6 Nucleic acid manipulations
2.6.1 Phenol extraction
DNA was purified from protein using phenol extraction when inactivation of enzymes
such as KpnI was not possible by standard heat inactivation. The DNA solution was
37
mixed with one volume phenol:chloroform:isoamylalcohol (25:24:1) and vortexed for
20 seconds. Subsequent centrifugation was performed for five minutes, allowing
separation of organic and aqueous liquid phases. The aqueous phase containing DNA
was removed and used for downstream purposes whilst the protein remained in the
organic interphase and was discarded.
2.6.2 Ethanol precipitation
Ethanol precipitation was used to purify DNA from a buffer solution. Here, 0.1 volumes
of 0.3M sodium acetate solution was added to the DNA solution and mixed by vortex,
followed by the addition of 2.5 volumes of 100% ethanol. This was mixed by inversion
and incubated on ice for five minutes before the DNA precipitated was collected by
centrifugation for five minutes. The supernatant was discarded while the pellet
containing DNA was washed with 1mL of 70% ethanol, mixed by inversion and
centrifuged for five minutes. The supernatant was again removed and excess ethanol
was evaporated off at room temperature for 15 minutes. The DNA pellet was
resuspended in 30μL SDW for downstream purposes unless otherwise stated.
2.6.3 Isolation of genomic DNA from Arabidopsis thaliana plants
DNA was purified from seedlings of approximately 3 weeks old, grown in compost as
previously described. Whole seedlings were taken, snap-frozen in liquid nitrogen and
ground to a fine powder using a pre-cooled mortar and pestle. To this, 500μL of
2xCTAB solution was added and the suspension was transferred to a pre-cooled
microfuge tube. The suspension was then mixed by vortex for approximately one
minute until the mixture appeared homogenous. Samples were then incubated firstly
at 65° C for two hours in a water bath and then cooled for 5 minutes at room
temperature. A 1uL RNase A (10μg/ml) solution was added and incubated for 30
minutes at 37°C to degrade RNA. Samples were again cooled to room temperature and
DNA was isolated from cellular debris by the addition of 500μL of
chloroform:isoamylalcohol (24:1) which was mixed carefully into the solution by
inversion. The DNA was subsequently isolated by centrifugation for ten minutes,
where phase separation enabled the aqueous supernatant containing genomic DNA to
38
be transferred to a new tube. The chloroform:isoamylalcohol treatment was repeated
with the DNA containing supernatant again transferred to a new tube. To this, 0.8
volumes isopropanol was added and mixed carefully by inversion to precipitate DNA
and centrifugation for ten minutes used to collect the precipitated the DNA. The
supernatant was discarded and the pellet washed with 1mL 70% ethanol followed by a
one minute centrifugation. The supernatant was again removed and the pellet was
dried for 20 minutes at room temperature. The purified genomic DNA was dissolved in
50μL SDW and the solution stored at -20° C.
2.6.4 Polymerase Chain Reaction (PCR)
PCR was used both as the first step of cloning and additionally to detect specific DNA
sequences in bacteria and plants. The difference is derived from the differing type of
DNA polymerase used.
2.6.5.1 High fidelity cloning PCR
This was performed using the Accuzyme (Bioline) premixed solution of polymerase, a
high fidelity polymerase which yields blunt ended amplicons. A standard reaction
contains 25μL of Accuzyme mix, 10pg of forward and reverse primers, 1-5ng of plasmid
DNA and SDW to 50μL. Reaction was mixed by vortex, centrifuged briefly and loaded
onto the thermocycler immediately to prevent degradation of primers by the DNA
polymerase.
2.6.5.2 Standard Taq polymerase, qualitative PCR
Low fidelity PCR was performed as above except, the Accuzyme premix was not used.
In its place, Taq Polymerase (Bioline) was used along with the required buffers
supplied and dNTPs (10mM stock, Bioline). Primer concentration and amount of DNA
and SDW were the same as stated above.
39
Table 2.2: Primers used for PCR amplification of DNA
Primer description Forward primer sequence
shown 5’ to 3’
Reverse primer sequence
shown 5’ to 3’
Full length genes cloned for BiFC and GFP analysis
HD2A PG179NS cloning TTGATTCTTAGCCATGGAGT
TCTGG
TAAGAAACCACTAGTCTTGG
CAGCAGCG
HD2B PG179NS cloning GGAATCTAGAATGGAGTTC
TGGGGAGTTGCGGTG
CGATCTCGAGAGCTCTACCC
TTTCCCTTGCC
HD2C PG179NS cloning CACAACAATGGAGTTCTGG CGACTCTCGAGAGCAGCTG
CACTGTGTTTGGCCTTTG
HD2D PG179NS cloning TTTCACTAGCTATGGAGTTT
TGG
AATATCTCGAGCTTTTTGCA
AGAGGGACCACAAGG
TGA2 PG179NS cloning ATATGGCTGATACCAGTCCG
AGAAC
CGATCTCGAGCTCTCTGGGT
CGTGCAAGCCATAAG
TGA5 PG179NS cloning AATGGGAGATACTAGTCCA
AGAAC
CGATCTCGAGCTCTCTTGGT
CTGGCAAGCCATAG
TGA6 PG179NS cloning CATGGCTGATACCAGTTCAA
GGAC
CGATCTCGAGCTCTCTTGGT
CGTGCAAGCCACAAGGAAC
HD2C deletion primers for BiFC and GFP analysis
HDT3 75 forw: ATGATCTCTCAGGTTGCTTTGGGAG
HDT3 180 forw: CTATG GAGAAGTTTCCTCAGCTGTC TACGG
HDT3 255 forw: ATG AGCGT TTTCTTCTCTGGTTACAAAG
HDT3 360 forw: CTATG GCTGCGAAACAGGTGAACTT TCAG
HDT3 435 forw: ATG GACG GTAGTGAAGAGGATTCTTC
HDT3 540 forw: ATG GAAGAAGATGACTCCTCAGA AGAG
HDT3 615 forw: ATGTCCTCCAAGAACCCTGCGTCCAAC
HDT3 720 forw: CTATGCAGGCGGGTAAGAATTCTGGTGGAG
HDT3 795 forw: ATG GCGTTTGGGTGCAAGTCGTGC
HDT3 180 rev: GGAAACTTCTCGAGAGATAGCGTTCCAATG
HDT3 255 rev: CGATCTCGAGCTTCCAAGTATGAGACAGCG
CAAAG
40
HDT3 360 rev: CGATCTCGAGAGCAGCTTTGAAACCAGCAG
CCTC
HDT3 435 rev: CGATCTCGAGAGCGTCATCATCTTGCTTGGC
TTTG
HDT3 573 rev: CTTCCTCCTCGAGTCCAGAGTTTTC
HDT3 681 rev: GAATCTGTTTTCTCGAGAGTTACAAAC
HDT3 720 rev: CGATCTCGAGCTGTTTTGAGGGATGAGGAG
TTG
HDT3 795 rev: GGTGTCTCGAGCTGCTTCGATGTCTCTCCAG
HD2 Site Directed Mutagenesis Primers
265 SDM FOR
GCTCACGCTAAGGCCAAAC
ACAGTGCAGCTG
GCCTTAGCGTGAGCCTGCA
ATCCCATTTC
234 SDM FOR
GCAGCAGGAGAAGCAGCAA
AGCAGCAGCAGAC
TGCTGCTTCTCCTGCTGCGC
CTCCACCAGAATTC
260 SDM FOR
AAGCACCGAAGGCAGCAGG
AGCGTTTG
CTGCCTTCGGTGCTTGCTGC
TGCTTCGATG
272 SDM FOR
AGGCATGCGCAAGAACCTT
TACTTCGGAAATG
TTCTTGCGCATGCCTTGCAC
CCAAACG
HDT3 KKR>AAA SDM
CCGCAGCAGCATCAGCAGA
ACCCAACTCCTCCAAG
TGATGCTGCTGCGGGTTCTT
CAGGCTTCTTTGG
HDT4 KKA>KKR SDM AAGCGACCAAATGGTGCAT
TTGAGATAGCTAAAG
CACCATTTGGTCGCTTTTTG
CTCGGAGGAG
Primers for cloning into Yeast 2-hybrid vectors
Clon PGAD HDT1 For GGAGGCCAGTGAATTCATG
GAAGTTAAATCAGGAAAGC
CGAGCTCGATGGATCCTTAC
TTGGCAGCAGCGTGC
Clon PGAD HDT2 For GGAGGCCAGTGAATTCGTT CGAGCTCGATGGATCCTTAC
41
GCGGTGACACCAAAAAAC TTTCCCTTGCCCTTGTTAG
Clon PGAD HDT3 For GGAGGCCAGTGAATTCGGT
GTTGAAGTTAAGAATGG
CGAGCTCGATGGATCCTTAA
GCAGCTGCACTGTGTTTG
Clon PGAD HDT4 For GGAGGCCAGTGAATTCGGT
ATCGAGATTAAGCCAGG
CGAGCTCGATGGATCCTTAC
TTTTTGCAAGAGGGACCAC
Clon PGAD Eps For GGAGGCCAGTGAATTCATG
GAGAATGAGAGGGAAAAG
CGAGCTCGATGGATCCTTAG
TTCTCATCTTGAGGC
Clon PGAD TGA6 For GGAGGCCAGTGAATTCGCT
GATACCAGTTCAAGGAC
CGAGCTCGATGGATCCTCAC
TCTCTTGGCCGGGCAAG
HDT3 Clon ZF For
GGAGCAAAGTCGGCCACCA
GAACCTTTACTTCG
GTGGCCGACTTTGCTCCAAA
CGCTCCTGCAGACTTC
Sequence primers
35S-seq forward primer GTAAAGACTGGCGAACAG
NYFP-seq reverse primer ATGAACTTCAGGGTCAGC
CYFP-seq reverse primer AGCTCAGGTAGTGGTTGTC
42
2.6.6 Miniprep
The miniprep procedure was performed using the ‘Wizard® Plus SV Minipreps DNA
Purification System’ as per standard instructions. The column was dried of
contaminating ethanol by a further five minutes of centrifugation at low speed with
the centrifuges lid off, with the DNA finally captured into a microfuge tube through the
addition of 50μL of SDW and a final one minute centrifugation.
2.6.7 Midiprep
Midiprep was performed using a kit supplied from Qiagen following standard protocol
instructions. DNA was air dried at room temperature for 30 minutes and re-dissolved
in 100μL SDW. This was stored at -20°C for downstream applications.
2.6.8 Restriction digestion
Restriction enzymes and associated buffers were purchased from New England Biolabs
(NEB). Plasmid DNA was digested in 1.7mL microfuge tubes containing ~1μg of
plasmid DNA, 2μL of appropriate NEB 10X reaction buffer, 0.2μL of 100X BSA solution
and 10 units of the appropriate NEB restriction enzymes (Table 2.3). These reaction
tubes were vortexed and centrifuged briefly before incubation at 37°C for four hours.
Double digestions were performed where there was an overlap in buffer
compatibilities where the enzyme had >75% activity in a certain buffer. Otherwise,
single digestions were performed in order and separated by a heat inactivation and
ethanol precipitation step.
2.6.9 Agarose gel staining, excision and purification
Digested DNA that required downstream reactions were purified by agarose gel
staining, excision and purification to prevent liberated DNA strands contaminating
further ligation reactions. An agarose gel was prepared in an identical manner as
described in section 2.6.4 with the exception that the addition of ethidium bromide
was omitted. After the DNA samples have been run and separated on the agarose gel,
the gel was added to 200mL of a 0.05% methylene blue solution and incubated for an
43
hour at room temperature with shaking at 50rpm. The methylene stain resolves DNA
at amounts down to 100ng and is facilitated by greater contrast. This was achieved by
removing the methylene blue and destaining with SDW for an hour, at which point the
appropriate DNA strand is clearly in evidence. This was removed by scalpel and at this
point the DNA is purified from the gel following the ‘Wizard® SV Gel and PCR Clean-Up
System’ standard protocol. The resulting solution containing the DNA of interest was
stored at -20°C until required for downstream applications.
2.6.10 Ligation
Unless otherwise mentioned, all ligation reactions were carried out in a reaction mix
containing 1μL each of T4 DNA ligase buffer (NEB), T4 DNA ligase (NEB) and
vector:insert ratio of 1:3. The volume of gene insert to be utilised was determined
based on the amount of DNA present as approximated from an agarose gel. A final
reaction volume of 10μL was apportioned with SDW. The microfuge tubes were briefly
vortexed and centrifuged before being left to incubate either room temperature for
one hour, or overnight at 4°C to allow ligation reaction to occur. Selection for
successfully ligated gene-vector construct was carried out by transformation and
growth of DH5α competent cells (section 2.2.4.2).
2.6.11 Semi-quantitative RT-PCR
RNA was extracted from 100μg plant tissue using the RNeasy Plant Mini Kit (QIAGEN)
following the standard procedure. cDNA was then prepared from the RNA extract by
taking 100ng of RNA and following the standard instructions for the Quantitect
Reverse Transcription Kit (QIAGEN). cDNA was then quantified using a nanodrop.
2.7 Protein assays and procedures
2.7.1 Western Blot
Western blots were performed on both A.tumefaciens infiltrated leaves and the stably
transformed A.thaliana leaves to ensure protein expression and validate protein
identity after co-immunoprecipitation. This is particularly relevant for BiFC as
fluorescence not observed must be attributed to non-interacting proteins rather than
the lack of protein expression in planta.
44
Preparation of samples and extraction of protein
Protein extraction was performed on plant tissue by first grinding in liquid nitrogen
until a fine powder was formed. A minimal aliquot of extraction buffer (50mM Tris-HCl,
pH 7.4, 150 mM NaCl, 2 mM MgCl2, 1 mM DTT, 20% glycerol, 1% Igepal CA-630,
Protease inhibitor cocktail [1:100]), roughly constituting 1mL per 1g of tissue was then
added to the sample and vortexed vigorously for a minimum of two minutes at room
temperature. Further extraction was achieved by homogenising with a hand held
homogeniser for another two minutes. Cellular debris was then cleared by centrifuging
the samples for five minutes, with the supernatant then aliquoted to another tube.
To standardise the amount of protein to be loaded for each sample, a BSA curve was
generated by measuring the optical density OD600 using the Biuret assay. The OD600
reading of each protein extract was also measured by diluting 5μL of protein sample to
795μL of sterile water and 200μL of Biuret solution. The volume of each sample to be
loaded was determined according to the amount of protein loaded for the most
diluted sample. The protein sample was diluted to 1x loading buffer concentration and
boiled for five minutes at 95 degrees Celsius in preparation for SDS-PAGE.
SDS-PAGE and protein transfer
Protein samples in loading buffer were loaded onto a 12% SDS-Polyacrylamide gel and
run at 200mV until the dye front reached the end of the gel. After SDS-PAGE the gel
may either be stained in coomassie blue to visualize the protein bands, or used to
transfer the protein bands onto a membrane for further Western blot.
Antibody incubations
The membrane was removed from the overnight blocking reaction and briefly rinsed
with two changes of wash buffer. Primary antibody, either mouse anti-cmyc or mouse
anti-HA antibody, was diluted 1 in 500 with wash buffer and incubated with the
membrane on an orbital shaker for one hour at room temperature. Thereafter, the
membrane was washed in >4mL/cm2 of washer buffer for 15 min at room
temperature followed by three five minute washes with fresh changes of the wash
buffer.
45
Secondary antibody solution was made up with a 1 in 400 dilution of the goat anti-
mouse HRP-labelled secondary antibody with wash buffer. The washed membrane was
incubated with secondary antibody solution for one hour at room temperature on an
orbital shaker, followed by two five minute washes with wash buffer. A final 50mL of
washer buffer was added and incubated for 15 min at room temperature.
Detection
Detection reagents (Amersham Detection solutions) were equilibrated to room
temperature before mixing in a ratio of 1:1 (0.125mL per cm2 of membrane).
In the dark room, Amersham Hyperfilm™ ECL film was cut to the size of the membrane
and placed vertically onto the membrane in the x-ray cassette. The cassette was closed
and exposed for one minute and subsequently developed in the developing solution
and fixed in the fixer solution. Subsequent films were developed with varying exposure
time based on the appearance of the first film being developed.
2.8 Microscopy
2.8.1 Fluorescence microscopy
Leaf samples to be examined for fluorescence were cut into 0.5cm X 0.5cm pieces and
loaded onto microscope slides using water as a fixative. The Olympus IX-71 inverted
fluorescence microscope was used to observe fluorescence in the lower epidermal cell
layer of Nicotiana benthamiana leaves. Two filters were utilised in this project.
The green fluorescent protein (GFP) observing filter U-MGFPHQ consists of an
excitation filter with a wavelength of 460-480nm, diachronic mirror of 485nm and a
barrier filter of 495-540nm. This filter set was used to visualise fluorescence resulting
from the reconstitution of the yellow fluorescent protein only as there was sufficient
overlap of both excitation and emission spectra between YFP and GFP.
The red fluorescent protein (RFP) observing filter U-MRFPHQ consists of an excitation
filter with a wavelength of 535nm to 555nm, diachronic mirror of 565nm and a barrier
filter of 570nm to 625nm. This filter set was utilised to differentiate autofluorescence
and genuine fluorescence generated from the reconstitution of the YFP protein or
46
whole GFP protein as YFP would not be expected to be sufficiently excited to emit
yellow light at this range.
2.8.2 Confocal microscopy
Leaf material from either transiently transformed N. benthamiana plants or stably
transformed Arabidopsis plants were loaded on glass slides using water as a fixative
solution. These were analysed using the confocal laser scanning microscope (CLSM)
using a TCS SP2 AOBS confocal microscope (Leica) through a 40Xobjective lens.
Confocal images were collected using the ‘Leica confocal softwareTM’ and the following
excitation lasers and emission channels were used:
Protein tag identity Excitation
wavelength (nm)
Emission wavelength (nm)
GFP 488nm 510-540nm
YFP 488nm 530-550nm
Chlorophyll auto- fluorescence 514nm 680-700nm
DAPI 358 nm 461 nm
2.8.3 Image analysis
Quantitative analysis of images was performed using the Image J plugin “confocal
stacks” (Abràmoff, Magalhães et al. 2004). Microscope images were imported into
Photoshop and the contrast and brightness altered.
47
Chapter 3
Characterization of the interaction and subcellular localization of
HD2C
48
3.1 Introduction
3.1.1 The impact of localization on regulating protein function
Identifying the regulation of enzyme activity is essential to building an understanding
of an enzyme’s role in the context of cell response. Knowledge of such regulation is
required as a basis for manipulating this role to divert cellular activity in a specific
direction (Martin 2010). A critical mechanism of enzyme control is by limiting
availability of enzyme to its substrate. In eukaryotes, this level of regulation is
facilitated through the compartmentalization of specific functions to organelles.
Moving an enzyme in or out of an organelle can prevent or enable access to the
enzyme’s substrate (Carmo-Fonseca 2002; Shaffer, Sharma et al. 2005). The advantage
of this is two-fold. Firstly, it ensures that there is a stable pool of enzyme present in the
cell so that signals can quickly be converted to action without the delay caused by gene
expression. Secondly, the removal of enzyme from its substrate is an energy efficient
mechanism of repression that is not dependent on specialized inhibitory proteins that
limit or block substrates from the catalytic domain.
Spatial regulation of a protein was identified in a number of RPD3-like histone
deacetylases (Grozinger and Schreiber 2000; Verdel, Curtet et al. 2000). This was
initially hypothesised when the mammalian RPD3-like HDAC5 was fused to GFP and
expressed in human cells to reveal a mixture of nuclear and cytoplasmic fluorescence
(Wang, Kruhlak et al. 2000). Localisation was subsequently shown to correspond to the
phosphorylation state of the HDAC with phosphorylation marking the protein for
sequestration to the cytoplasm and de-phosphorylation being a requirement for
nuclear import. The plant specific histone deacetylases belonging to the HD2s have not
been analysed to determine whether the regulation of localization is a relevant
mechanism of control for this enzyme. However, it was shown using both
immunofluorescence and direct visualization of GFP tagged HD2 proteins that a
number of HD2 homologues accumulate in the nucleus and nucleolus (Lusser, Brosch
et al. 1997; Zhou, Labbe et al. 2004; Panni, Montecchi‐Palazzi et al. 2011). It will
therefore be important to determine whether nuclear localization is a mechanism of
control used by this class of enzymes and whether their localization can be
manipulated so that access to their histone substrates may be exogenously controlled.
49
3.1.2 Mechanism of Nuclear localization
The nucleus stores chromatin within a double phospholipid membrane perforated with
pores and transporters that facilitate the movement of macromolecules (Fahrenkrog
and Aebi 2003). It was shown that nuclear pores allow the non-specific, passive
diffusion of macromolecules up to 60kDa, although some sources have speculated that
this size exclusion may not apply to dense macromolecules even larger than this
(Perez‐Terzic, Jaconi et al. 1997; Wang and Brattain 2007). Therefore, nuclear
accumulation of proteins under this critical threshold may be achieved by passive
diffusion across the membrane. Subsequent protein-protein interactions may then
amass sufficient size that together trap smaller proteins within the nucleus. For
proteins and complexes larger than this critical size threshold, active transport is
required to mediate passage across the nuclear membrane. HD2s are ~32kDa in size
and therefore are within the size limit that would allow for passive diffusion as a
possible mechanism for nuclear accumulation. However, as functional proteins they
were shown to be part of large complexes which are ~400kDa in size. Such large
complexes would not be able to pass the nuclear membrane. The issue of how their
nuclear and nucleolar localization pattern occurs has not yet been solved.
Nuclear import of proteins by active transport involves three main molecular
components; karyopherins, nuclear pore complexes (NPCs) and Ran guanidine di- and
tri-phosphates (Ran-GDP/GTP). Karyopherins include the protein importin, a molecular
dimer consisting of an alpha and beta subunit which binds to target proteins that
contain a nuclear localization signal (NLS) (Görlich, Kostka et al. 1995). Once bound,
importin is activated and recognized by the NPC where it is transported through the
membrane spanning channel of the pore (Hinshaw, Carragher et al. 1992; Fahrenkrog
and Aebi 2003). In the nuclear matrix, the importin-cargo complex is exposed to Ran-
GTP, which is recognized and bound by importin and results in the release of its cargo.
The importin/Ran-GTP is then recycled to the cytoplasm where Ran GTP itself is
hydrolysed to Ran-GDP by GTPase activating protein (GAP) causing importin to release
Ran-GDP and to become available to form a new importin:protein complex (Moore
1998). Ran-GDP is then recycled by being translocated to the nucleus where a guanine
nucleotide exchange factor (GEF) converts Ran-GDP to Ran-GTP (Lange, Mills et al.
2007). The closely related nuclear export process differs in that the karyopherin
proteins facilitating the transport are exportin proteins. These bind the cargo and Ran-
50
GTP in the nucleus which drives recognition and translocation through the NPC’s.
Conversion of Ran-GTP to Ran-GDP allows the release of cargo and subsequent
recycling of Ran-GDP and importin to the nucleus.
An important element of the nuclear import process is the recognition of a target
protein by importin to specific NLS sequences. For example topoisomerase II was
shown to bind importin alpha in vivo, however this functionality was removed
following mutation of its nuclear localization signal. Localization tags are specific
sequences of amino acids that are recognized by other proteins that facilitate the
transport from one compartment to another (Jans, Xiao et al. 2000). Nuclear
localization sequences can be highly varied, with dozens of consensus motifs shown to
allow active transport of proteins across the nuclear membrane (Cokol, Nair et al.
2000). Based on a number of such sequences identified in plants and animals it was
shown that NLSs consist of clusters of lysine and/or arginine residues that are broadly
classified as being either monopartite, bi-partite or irregular. Simply defined,
monopartite sequences consist of a single string of positive residues such as the SV40
large T antigen NLS [PKKKRRV], whereas bipartite sequences consist of two discrete
clusters of positive residues that are separated by ~10 amino acids such as the
nucleoplasmin NLS [KRPAATKKAGQAKKKK]. Irregular sequences are broadly defined as
anything that does not match the parameters of mono- and bi-partite sequences and
most commonly bind directly to the beta-subunit of the importin hetero-dimer (Lee,
Cansizoglu et al. 2006). These signals are marked by their notable absence of lysine or
arginine residues in the sequence, for instance the human hnRNP A1 protein contains
a novel 38 amino acid transport signal that is required to confer nuclear translocation
(Pollard, Michael et al. 1996).
The importance of nuclear localization sequences are that they provide specific binding
sites for importin, and are therefore required for active import into the nucleus. A
significant regulatory mechanism that determines localization state is therefore
limiting access of the NLS to importin. This is typically achieved by post-translational
modifications such as phosphorylation, or protein-protein interactions such as binding
of 14-3-3 proteins which cause steric changes in protein structure exposing or hiding
the import sequence from the surface of the protein area. This is commonly coupled to
the exposure of nuclear export signals so that the protein may be effectively confined
51
to its designated cellular context. This was shown in the regulation of the Arabidopsis
floral identity transcription factor APETALA3, which requires co-expression of
PISTILLATA to form obligate heterodimers. Interaction with PISTILLATA induces a
structural change in APETALA3 that exposes its nuclear localization signal (McGonigle,
Bouhidel et al. 1996). In this way, despite its cellular presence throughout floral
induction, its role as a transcription factor to induce floral identity is only activated in
response to PISTILLATA expression.
3.1.3 Nucleolar localization
The fact that a noticeable fraction of the HD2C-GFP fluorescence is localized to the
nucleolus is likely to be a significant aspect of its function. The nucleolus is a non-
membrane bound sub-structure of the nucleus which is comprised of DNA, RNA and
protein which accounts for approximately 25% of nuclear volume. The nucleolus is
centrally involved in ribosomal RNA transcription, ribosome assembly, cell cycle
progression, developmental regulation and cell stress response. HD2C has been shown
to have roles in each of these functions (Hollender and Liu 2008), which is consistent
with its nucleolar localization. Therefore determining the protein sequence required
for its nucleolar localization will be significant for future manipulation of its biological
process.
Like nuclear localization, nucleolar localization can be achieved either via direct
nucleolar targeting by a nucleolar localization sequence (NoLS), or by interaction with
proteins that are themselves localized to the nucleolus. NoLSs and NLSs are structurally
similar as they usually contain stretches of positive amino acids, however their
mechanism of action differ. As mentioned previously, NLSs function via active
transport, whereby movement across the nuclear membrane is mediated by importin
which binds the NLS and the nuclear channel which provides the path of movement. In
contrast to this, nucleolar localization is driven by strong associations with nucleolar
core components, suggesting a retentive mode of localization. The similarity in amino
acid compositions is significant, as it has been shown that a significant fraction of NLSs
have NoLS joint targeting. For example, in Saccharomyces cerevisiae, the C-terminus of
UTP20 contains a lysine and arginine rich region which confers both nuclear and
nucleolar localization (Dez, Dlakić et al. 2007). Alternatively NoLSs may be completely
52
separate from any NLS, or indeed contain no NLS and thus rely on passive diffusion to
access its chromatin binding region (Saslowsky, Warek et al. 2005).
Aside from the description of HD2C as a nuclear and nucleolar protein, no further
characterization of the molecular mechanism for its subcellular localization has been
provided. Given the relative importance of its internal sequestration to these cellular
sub-structures, investigation of the process which drives its internal targeting may
provide a valuable insight into its regulatory dynamics.
3.1.4 HD2-HD2 interactions in tertiary protein complex
The HD2 family was first identified by investigating proteins obtained by purification of
chromatin from germinating maize embryos (Lusser, Brosch et al. 1997). Native maize
HD2 was found to elute at a molecular mass of ~400kDa in gel filtration
chromatography, leading to speculation that it functions in a protein complex. A silver-
stained SDS-PAGE revealed this complex to consist of three polypeptides of near
identical size. Subsequently it was shown that these peptides differed only by their
phosphorylation state, with phosphorylation evidently controlling the catalytic activity
(Durek, Schmidt et al. 2010). These data suggested that HD2s operate as a large,
homogenous protein complex that forms as a result of HD2-HDAC interactions. Indeed
the Arabidopsis HD2C protein was previously shown to bind both HDAC6 and 19 in
response to various abiotic stresses (Luo, Wang et al. 2012). It was shown in other
HDAC systems that homologous interactions have both regulatory and enzymatic roles.
HDAC4 and 5 function as a complex with HDAC3 in the nucleus of mammalian cells to
bind and deacetylate their target histone substrate (Grozinger, Hassig et al. 1999).
Aside from its likely regulatory importance, the dimerization or possible
multimerization of HD2 would likely increase its molecular size sufficiently that it is
unable to passively diffuse across the nuclear membrane. Whether this affects its
ability to cross into the nucleus, or instead aids in its nuclear retention is likely
dependent on where the dimer formation occurs. Complex formation of the catalytic
subunit of cyclic AMP dependent protein-kinase (PKA) was shown to control nuclear or
cytoplasmic localization in this way. Complex formation in the cytoplasm forms a high
molecular weight structure that cannot pass into the nucleus and therefore
53
equilibrium is reached where all PKA is cytoplasmic (Harootunian, Adams et al. 1993).
When the complex dissociates, the smaller subunit is released and becomes able to
passively diffuse across the nuclear membrane. HD2s may operate in the opposite way
to this so that complex formation occurs in the nucleus and thereby preventing
diffusion back across the membrane and into the cytosol. This passive approach marks
an alternative strategy to the conventional active transport via NLS targeting
previously described. Alternatively, complex formation could occur in the cytosol and
be a prerequisite for nuclear import via exposure or formation of nuclear import
signals. It would therefore be important to identify where the HD2 complex formation
occurs so that a model of HD2 transport can be proposed.
3.1.5 Hypothesis and aims
The subcellular localizations of HD2A, HD2B and HD2C have been analysed
independently in a number of studies, utilizing Arabidopsis, maize and longan
homologues to demonstrate a conserved nuclear and nucleolar phenotype between
various homologues. However, no study has so far performed a parallel localization
analysis of all HD2 homologues from one organism, in the same tissue type and under
identical experimental conditions to compare subcellular localization between the
proteins. Such an analysis is essential to identify any localization patterns specific to
one homologue which may suggest a diverse function. Furthermore, HD2D was not
previously characterized with regards to its subcellular localization. It is expected that
all Arabidopsis HD2 homologues contain the same nuclear and nucleolar subcellular
localization pattern that was previously identified. It is therefore important to identify
the molecular mechanism of this localization pattern. Furthermore, the possibility of a
dynamic shuttling between cytoplasm and nucleus, or nucleus and nucleolus will be
investigated. Lastly, the prospect of HD2:HD2 interaction will be tested and the
localization of these dimers compared to the HD2-GFP localization pattern. The aims of
this study are therefore:
1. Construct HD2-GFP- fusion proteins to analyse the subcellular localization
patterns of all four HD2 isoforms using the same system and approach.
2. To test for the potential of HD2s to form dimers and to further analyse if all
dimer combinations of HD2C:HD2 can occur in planta. Bimolecular fluorescent
54
complementation (BiFC) will be used here because it offers an in planta
approach to test interaction, and will additionally allow for direct qualitative
evidence of the localization pattern of the dimers.
3. To investigate if HD2C nuclear import is driven by a nuclear localization signal.
To test this, deletion constructs of HD2C will be fused to GFP and visualized in
planta. Site directed mutagenesis will then be performed to determine the
critical residues required for nuclear localization.
4. To test if localization is dynamic by comparing the localization pattern of HD2C-
GFP in stressed and non-stressed plants. Stably transformed Arabidopsis lines
expressing 35S:HD2C-GFP will be made and the movement of HD2C-GFP traced
in response to salt stress. Salt was previously shown to affect HD2C expression
and caused deacetylation of ABA response factors.
3.2 RESULTS
3.2.1 Investigating the subcellular localization of the HD2 family of proteins
HD2 homologues A-D were cloned into the PG179NS-GFP vector so that HD2-GFP
fusion proteins could be expressed. These vectors were used to transform
Agrobacteria, which were themselves used to infect and transiently transform
Nicotiana benthamiana leaves by agro-injection. Three days after injection,
fluorescence was observed using a fluorescent microscope, and recorded using
confocal microscopy. Standard images for each construct are represented in figure 3.1,
showing GFP emission for the protein localization, DAPI emission for staining of the
nuclear area and chloroplast autofluorescence to show cellular context and as a
control for potential autofluorescence in the GFP channel.
GFP fluorescence for the three isoforms HD2A, B and C were very similar. Fluorescence
of these three GFP fusion proteins appeared as a strong, discrete signal surrounded by
a halo of weaker fluorescence that overlayed completely with the fluorescence in the
DAPI channel (figure 3.1). The localisation of the weaker and the stronger fluorescence
suggested that the weaker fluorescence was contained within the nucleus and the
stronger represented subnuclear structures such as the nucleolus. These
interpretations would be in agreement with published work showing that HD2A, B and
C are present in the nucleus and nucleolus. The fluorescence observed here is not
55
caused by autofluorescence as it did not match the fluorescence observed under the
RFP channel used to monitor chloroplastic and other cellular autofluorescence (figure
3.1).
In contrast to HD2A, B and C, HD2D showed a novel, previously not described
localisation pattern with GFP fluorescence observed in both the cytosol and nucleus.
Furthermore, there was no discrete sub-nuclear localization, suggesting that there is
no inherent nucleolar targeting as demonstrated for the isoforms HD2A-C. The HD2D
fluorescent pattern was similar to the pattern observed when expressing GFP alone,
therefore suggesting that there is no targeting mechanism capable of exclusive nuclear
localization in this HD2 homologue.
3.2.2 HD2 proteins form nucleolar localized dimers in planta
A necessity for a large protein complex comprised of only HD2s is the interaction of
HD2s with each other. Here, it was determined if HD2 proteins have the ability to form
HD2 homo- and/or heterodimer complexes. BiFC was used to identify whether
members of the HD2 family have the potential to engage in such interactions in planta.
HD2A-D coding sequences were cloned into PG179NS-YN and PG179NS-YC BiFC
vectors to generate HD2-YN and –YC fusion proteins. Obtained plasmids were
transferred into Agrobacteria and co-injected into Nicotiana benthamiana leaves so
that a HD2-YN and a HD2-YC protein could be co-expressed. First, the ability of HD2C
to form homodimers in planta was investigated. Fluorescence, indicating interaction of
the HD2C-YN and HD2C-YC fusion proteins was observed almost exclusively in the
nucleolus of infiltrated Nicotiana epidermal leave cells (figure 3.2). This was significant
as it revealed a difference from the nuclear and nucleolar pattern of fluorescence
obtained when the HD2C-GFP construct was tested in section 3.1.1. The N-terminal
end of HD2C is the catalytic domain of the protein and therefore unlikely to mediate
dimerization. It was therefore used as a negative control to ensure that interaction
was not dependent on the YFP fragments spontaneously recomplementing. The N-
terminal end (AA1-60) of HD2C was therefore fused to the N-terminal of YFP and
tested with HD2C-YC. This yielded no detectable fluorescence in the YFP channel,
suggesting that the fluorescence reported above was due to HD2C:HD2C interaction
56
(figure 3.2). Western blot detection of each protein was used to ensure that expression
was not the reason for the absence of any fluorescence (appendix 1).
Micro-array data and evidence from semi-quantitative RT-PCR experiments
demonstrated an overlap in physiological expression patterns of HD2s. This gives rise
to the biologically relevant prospect of heteromeric HD2 interaction in the formation
of a functional HD2 complex. To test this hypothesis, HD2A, HD2B and HD2D were
cloned into the PG179NS-YC BiFC vector and tested for interaction with HD2C-YN.
Consistent with HD2C dimers, fluorescence indicating interaction of the HD2C with the
three other HD2 isoforms was almost entirely restricted to the nucleolus, with only
very faint nuclear fluorescence (figure 3.2). It was interesting to observe that the
HD2C:HD2D dimer was similarly nucleolar, which was in contrast to the cytoplasmic
and nuclear localization of the HD2D-GFP fluorescence observed earlier
3.2.3 A nuclear import-related sequence maps to the C-terminus of HD2C
In this thesis and in published work, it was shown that all four Arabidopsis HD2
proteins localize at least partially to the nucleus (Zhou, Labbe et al. 2004). Nuclear
localization can be achieved either by direct migration of the protein if the size of the
protein is beyond the exclusion size of the nuclear pore complex or by an active
transport requiring a nuclear localization sequence. Despite the known nuclear
localisation, no nuclear localisation signal was reported for any of the HD2 proteins so
far. To determine whether HD2 proteins have a NLS, HD2C was chosen for a mutation
analysis coupled with localisation studies of the obtained mutant HD2C variants. As a
beginning for this analysis, N- and C-terminally truncated variants of HD2C were
constructed and fused with a C-terminal GFP. Individual constructs were transformed
into Agrobacteria which were then used to transform N.benthamiana leaf tissue via
injection to assess localization of the encoded HD2C protein mutants. It was hoped
that this analysis would reveal if HD2C contains a region required for nuclear import.
Expression and localisation of the mutant HD2C-GFP fusion proteins were analysed
using confocal and fluorescent microscopy (Appendix 2). Localisation patterns were
assessed as nuclear, nucleolar, cytoplasmic or relevant permutations of these based on
the fluorescent patterns of the various constructs. Shown in figure 3.3A are graphic
representations of the deletion constructs and a description of their localization
57
patterns. Together these data allowed a map of the relative positions of regions
required for nuclear and nucleolar localization to be determined.
N-terminal truncations up to and including amino acid residue 185 did not impact on
the localization of HD2C deletion proteins, with those variants localising to the nucleus
and nucleolus as did the full (figure 3.3A) length protein. Further N-terminal
truncation up to and including residue 205 caused HD2C-GFP fusion proteins to lose
exclusive nuclear/nucleolar localisation and to produce a nuclear, nucleolar and
cytoplasmic expression. Subsequent deletion up to and including residue 240 led to a
complete loss of nucleolar localisation and presence of the protein in the nucleus and
cytoplasm only, which was similar to the fluorescent pattern of GFP alone.
Serial truncations from the C-terminal end up to and including amino acid residue 225
did not impact on the ability of HD2C to localize to the nucleus and nucleolus (figure
3.3A). Further C-terminal truncations, up to and including amino acid 204 resulted in
loss of exclusive nuclear localization as well as nucleolar localisation. These results
suggested that the residues required for exclusively nuclear localization and nucleolar
localization of HD2C were contained between amino acids 185–226. Next the putative
minimal localization domain was tested to determine if it was sufficient to transfer
exclusive nuclear and nucleolar localisation to a GFP protein which is usually found in
the cytosol and the nucleus. Thus, the 185–226AA peptide was cloned into the
PG179NS-GFP vector and expressed via Agrobacteria injection in epidermal leave cells
of Nicotiana benthamiana. The localization of the fusion protein was investigated using
confocal microscopy. Interestingly, despite the clear necessity for this domain to be
present in HD2C to allow exclusive nuclear and nucleolar localization, the putative NLS
domain consisting of residues 185-226 did not confer exclusive nuclear fluorescence to
GFP. Instead the fusion GFP protein was present in the cytosol and the nucleus. A
larger HD2C peptide fragment comprising AA143-226 was therefore tested to
determine the minimal region required to cause exclusive nuclear localization of GFP.
This fragment was sufficient to confer exclusive nuclear localization to GFP, but did not
have any nucleolar presence (figure 3.4B). Instead, a further tested fragment
consisting of amino acids 206-257 conferred a nucleolar presence; however this was
not sufficient to allow exclusive nuclear fluorescence as fluorescence was also
detected in the cytosol (figure 3.3B). Together these results show that overlapping but
58
discrete regions of HD2C are required to impart nuclear localization and nucleolar
localization.
3.2.4 The HD2C nuclear localisation signal is dependent on a KKAK motif
Having shown that residues 143–226 of HD2C are sufficient and required for nuclear
localization, it was determined whether this sequence contains an NLS and if so, to
localize the critical residues. From previous studies it is known that NLSs are comprised
of either a basic amino acid cluster defined as monopartite, or of two basic clusters
separated by ~10 amino acids which constitute a bipartite NLS. The HD2C region that is
critical for nuclear localization contains a number of lysine and arginine residues
(figure 4.4). Four subregions of this peptide sequence contained lysine residues which
may be critical determinants of nuclear localization. To be able to differentiate
between these four subregions, four different full length HD2C constructs with
different permutations of lysine to alanine substitution mutations were constructed;
these were named nuclear localization sequence mutants 1-4 (NLSM1-4), with the
mutated regions shown in figure 3.4A. There was no evident alteration in the nuclear
pattern of localization in NLSM1-3 when compared to the wild type HD2C protein (data
not shown). However the NLSM4 mutation resulted in a nuclear, cytoplasmic and
nucleolar pattern of fluorescence (figure 4B). This indicated that the motif KKAK was a
critical determinant of exclusive nuclear localization.
3.2.5 Sequence alignment of HD2 gene family homologues reveals conservation of
the critical KKAK motif
To determine if the KKAK sequence is evolutionarily conserved in HD2 family members,
a multiple sequence alignment was constructed of the four Arabidopsis HD2s and
related homologues from poplar, rice, wheat and maize (figure 3.5). The sequence
alignment revealed that the three lysine residues in the KKAK motif of HD2C were
absolutely conserved in all of the HD2 homologues, indeed the whole KKAK sequence
was conserved in seven of the eight homologues analysed including in those of
monocot plants. Interestingly, the only difference observed for the KKAK motif was in
HD2D, which had previously been shown to have a nuclear and cytoplasmic
localization pattern. Instead of a KKAK motif, it contained a KKNK motif, which may
59
rationalize the altered localization pattern compared to the other Arabidopsis HD2
homologues.
3.2.6 Nuclear localisation is not a pre-requisite for HD2C dimerisation.
Nuclear retention of HD2C could be mediated either through active transport or
through accumulation of molecular mass, e.g. a higher order protein complex, in the
nucleus which prevents leakage into the cytoplasm. The KKAK motif which is critical for
exclusive nuclear localization conforms to the core [K(/R)KXK] motif present in
monopartite nuclear localization sequences, which suggests that active transport is the
preferred method of translocation. However, this still left two options; firstly that the
import of HD2C occurs via its monomers with multimerisation taking part in the
nucleus or nucleolus or secondly that multimerisation in the cytosol precedes import
of the multimer into the nucleus. As a first step towards answering this question, it
was tested whether dimerisation can occur in the cytosol or if nuclear localisation is an
absolute requirement for dimerisation of HD2C. Co-expression of the NLSM4-YN and
wild type HD2C-YC in N.benthamiana epidermal leaf cells was used to investigate if
dimers could still form and if so, where they appeared (figure 3.6). BiFC analysis
demonstrated firstly that the mutated HD2C is still able to form dimers with the wild
type HD2C protein. Secondly, the dimers were observed in the nucleus and nucleolus
which clearly showed that both isoforms migrated to the nucleus. This distribution was
identical to that of the wild type HD2C:HD2C dimer. As demonstrated earlier in this
thesis, the HD2C mutant protein is not able to move into the nucleus. Thus it can be
concluded that its nuclear and nucleolar localisation required previous dimerisation
with wild type HD2C in the cytosol and subsequent re-localisation in a ‘piggy-back’
mechanism to the nuclear and nucleolar compartment. To further provide evidence
for cytoplasmic dimerisation, subcellular localisation of a NLSM4 homodimer was
tested using BiFC. Fluorescence was observed in the cytosol, nucleus and nucleolus
(figure 3.6). This finally indicated that dimerisation can occur in the cytosol. It
furthermore supported the statement that a complete NLS is not required for
dimerisation. Although not finally conclusive, this result can also be used as evidence
for the hypothesis that dimerisation occurs in the cytoplasm and that dimers or higher
order multimers move into the nucleus with the aid of the NLS.
60
3.2.7 HD2C localization is altered in response to abiotic stress
Although there is a well characterized nuclear and nucleolar localization pattern for
HD2C, there is the possibility that this pattern is only observed in a non-stressed
context. Previously, it was shown that abiotic stresses such as salt displayed various
phenotypes in transgenic Arabidopsis lines with modified HD2C expression. This gives
rise to the possibility that these phenotypes are a result of post-translational
modification to HD2C which may induce a change in protein localization. To test this
hypothesis, transgenic Arabidopsis lines expressing 35S:HD2C-GFP were made using
the floral dip method. Two independent lines were collected which displayed strong
expression as determined by RT-PCR and direct visualization under a fluorescent
microscope (results not shown).
The role of salt stress on HD2C-GFP localization was tested by transferring two week
old seedlings germinated and grown on 0.5MS media onto either 0.5MS media
containing 150mM NaCl, or 0.5MS plates without addition of NaCl (control). After 24
hours, leaf tissue was harvested and investigated under the confocal microscope
(figure 3.7). Results indicated that there was a visual increase in the proportion of
nucleolar fluorescence compared to the rest of the nucleus. This was consistent for
both lines, in three independent experiments. To quantify this observation, images
were compiled and pixel density counts performed using Image J software. Results
showed that in non-treated Arabidopsis cells, the nucleolar fluorescence accounted for
25% of total nuclear fluorescence. In contrast, salt stressed plants had an increased
nucleolar proportion to 39%. A one tailed, independent t-test confirmed that these
results were statistically significant. These results therefore indicate that there is a
dynamic shift of nuclear to nucleolar HD2C localisation in response to salt stress, but
not between the nucleus and cytoplasm.
3.3 Discussion
3.3.1 Summary
Here, the localization of HD2C was analysed using fluorescent protein fusions to trace
the accumulation of HD2C in Nicotiana benthamiana epidermal cells. Consistent with
61
previous results was the observation that HD2C was present in the nucleus and
nucleolus when fused to GFP. However, it appears that this localization may be
dynamic, as stably transformed 35S:HD2C-GFP plants showed increased accumulation
of nucleolar fluorescence after 24 hours of salt stress. The mechanism of this
accumulation remains unknown; however dimerization of HD2s appears to be
involved, as the fluorescent pattern observed when performing a BiFC analysis of
HD2C:HD2 interaction was almost exclusively nucleolar. In addition, the mechanism of
nuclear and nucleolar localization appears to be separate. Mutation of an
evolutionarily conserved KKAK motif removes exclusive nuclear localization but not
nucleolar localization. In addition, a minimal NLS results in the absence of a nucleolar
fluorescent signal. Similarly a minimal nucleolar localization signal resulted in nucleolar
accumulation, but did not confer exclusive nuclear localization.
3.3.2 All combinations of dimers are possible between HD2C and the HD2 family
The prospect of dimerization has been hinted previously from the discovery of the HD2
gene family in maize. Here, using an in planta method, it was shown that HD2C has the
potential to form HD2 protein homo- and heterodimers. The results obtained do not
allow for an interpretation beyond the dimer level. It can be postulated that HD2C is
able to form larger protein complexes involving several HD2 monomers or even other
proteins. Indeed, this postulate would be consistent with the ~400kDa complex
identified via native gel electrophoresis following extraction from maize embryonic
tissue.
The biological and functional relevance of the interactions with other members of the
HD2 family was not explored in this thesis. However, previous work gives an indication
of the possibilities such interactions can have, including enzymatic activity changes, an
influence on subcellular transport and determination of DNA binding specificity. It is
feasible to postulate that dimerization or multimerization of HD2 proteins contributes
to the regulation at the level of the enzyme, whereby protein-protein binding provides
control over the catalytic activity of the enzyme by either switching the activity state of
the enzyme between functional and non-functional, or affecting the substrate
specificity of the enzyme itself. Such an assumption would be in agreement with the
finding that dimerisation is a way in which the activity of a number of mammalian
62
HDACs is regulated. For example, mammalian HDAC1 complex formation precedes
activation of its catalytic activity and additionally mediates interaction with other
transcriptional regulators which are responsible for binding its target chromatin
substrate (de Ruijter, Van Gennip et al. 2003) . Currently, it is not known if complex
formation has a similar impact on HD2 activity and specificity. However, it is known
that HD2 proteins interact with other regulators such as HDAC6, DNA
methyltransferase 2 and ERF7 suggesting a similar kind of regulation for HD2C possibly
also requiring interaction with other HD2s (Chinnusamy, Gong et al. 2008; Song, Wu et
al. 2010). HD2C preferentially binds histone 2B but accept all histones as substrates in
vitro. This could indicate that HD2s are subject to specificity changes which may be
mediated by the ability to combine with other HD2 proteins as demonstrated in this
thesis. The potential for some level of target specificity or changes in histone
deacetylation patterns manifested through different combinations of HD2 isoform
complexes is a logical extension of this work. Indeed, mammalian histone deacetylases
harness the ability to bind various HDAC homologues to recognize specific targets
under various conditions. While not explored in any detail in the case of the HD2s, it is
anticipated that the protein itself is not solely responsible for recognizing and binding
the gene regions that it is likely to function upon. Rather, it is hypothesised that HD2s
act via interaction with other proteins such as transcription factors on specific
chromatin regions and hence genes in a response dependent manner. Similar
mechanisms were shown for other epigenetic models such as DNA and histone
methylation. Thus, it may be hypothesised that the interaction of various HD2
homologues to each other allows HD2s to fine tune their response to triggers through
the combinatorial accumulation of transcriptional regulators which allows recognition
of various specific chromatin regions. Given that HD2C was described as a protein
involved in abiotic stress and ABA responses such triggers could be found in
environmental conditions causing stress (Sridha and Wu 2006).
3.3.3 HD2C localization state is dynamic and responds to salt stress
In this study HD2C localization was compared as a GFP fusion protein with the
localization found in BiFC experiments. Originally, it was hypothesised that cytosolic
and nuclear localization had regulatory implications, given comparable models in
63
plants, mammals and in the context of the transcriptional regulatory machinery where
translocation between cytoplasm and nucleus influenced availability of the protein and
hence its biological activity. Cytosolic localisation was not observed for HD2C-GFP
fusions protein in either stably transformed Arabidopsis plants, or transiently
transformed N.benthamiana. This was consistent with previous studies showing that
HD2C-GFP accumulated in both the nucleus and nucleolus. A shift in localization to the
cytosol was also not observed in stress treatment experiments. This was not seen in
the case of HD2C-GFP expressed in stably transformed plants, either in the different
tissue types dissected and visualized or in the leaf tissue exposed to various under
stresses. When transgenic seedlings expressing HD2C-GFP were kept for 24 hours on
minimal media plus 150mM salt HD2C developed a nucleolar-enriched localization
pattern. Thus, salt treatment caused a shift from the mixed nuclear and nucleolar
pattern observed under control conditions to a more nucleolar appearance. Similar
treatment dependent nucleolar retention has been reported for stress-responsive
factors. For example the Arabidopsis Elf4A-III protein, a putative anchor protein of the
exon junction complex, when expressed as a GFP fusion protein was targeted to the
nucleus during normal growth, but quickly localized to the nucleolus and splicing
speckles in response to hypoxia (Koroleva, Calder et al. 2009). The most widely cited
mechanism for this regulation is via post-translational modifications such as
phosphorylation or ubiquination. Such mechanism was demonstrated for the
mammalian transcription factor RelA, where cytoplasmic and nuclear fractions of the
protein undergo a stress-induced nucleolar translocation which is preceded by
ubiquitination of the protein (Thoms, Loveridge et al. 2010). That post-translational
modifications of HD2 proteins occur was demonstrated for the maize HD2 orthologue
which can be tri-phosphorylated. This gives rise to the possibility that phosphorylation
may be a mechanism determining nucleolar targeting of HD2C. Identification and
mutation of phosphorylation sites in HD2C would be a way forward to test the
hypothesis that phosphorylation of HD2C results in nucleolar retention.
The nucleolus was shown to be involved in the plant stress response because it
contains genes that are responsive to stress response factors. Following salt stress
there is a concerted response that involves the recruitment of proteins to the
nucleolus (Guo, Yang et al. 2012). Similarly the morphology of the nucleolus changes,
64
with the influx and diversity of proteins following stress resulting in an increased size
and activity (Boulon, Westman et al. 2010; Shaw 2013). Another important insight into
the mechanism of nucleolar retention was the nucleolar only localization pattern that
was observed for the HD2 dimers using BiFC. This apparent contradiction between the
HD2C-GFP localization in the nucleus and nucleolus, and the BIFC result showing that
the HD2C dimer localizes to the nucleolus but not the nucleus, may be due to the
differences shown by the two approaches. The HD2C-GFP construct shows the
localization pattern of both monomeric and multimeric forms of HD2C. As in the stably
transformed Arabidospis plants, the fluorescent pattern shows both nuclear and
nucleolar accumulation. In contrast, BiFC analysis monitors only the presence of
dimers. This leads to a model whereby HD2C monomers are localized throughout the
nucleoplasm, hence found in the nucleus and nucleolus and the dimer or multimer
being restricted to the nucleolus only. Additionally, the stability of the reconstituted
YFP ensures that the dimers are locked together; thus even transient associations yield
strong fluorescence. Previously it was implied that the functional enzyme is active as a
large HD2 complex, further implying that the monomeric form of the enzyme is non-
active. Thus the model can be extended by assuming that a non-active HD2C monomer
is present throughout the nucleoplasm until activated, possibly by phosphorylation,
upon which it dimerizes and moves to the nucleolus. In the nucleolus, it could bind to
and deacetylate its specific histone substrate. Abiotic stresses, as shown here, trigger
nucleolar localisation and could hence be activators of HD2C function. While this
hypothesis leads to an attractive model, this study is sufficient only to highlight its
potential, rather than validate its accuracy. To resolve this pathway a number of
further problems need to be investigated. Principle among them is to determine the
dimerization domain of HD2C. A mutant form of the enzyme unable to dimerize will
allow firstly to determine whether the monomer itself can be targeted to the nucleolus
and secondly to investigate if the monomer retains catalytic activity or if dimerisation
is required. Finally it could be addressed whether plants over-expressing a non-
dimerizing HD2C retain the abiotic stress response that was reported for the over-
expression of wild type HD2C in Arabidopsis. Furthermore, the possible combinations
of various HD2 isoforms give rise to the possibility that various permutations yield
specific functions.
65
3.3.4 HD2C contains a critical KKAK domain necessary for exclusive nuclear
localization
A previously unreported HD2D-GFP fusion protein expression demonstrated a novel
HD2 localisation pattern. HD2D, in contrast to the other three HD2 isoforms had a
nuclear and cytoplasmic localization pattern, whereas HD2A-C was instead confined to
the nucleus and nucleolus. This suggests that some integral nuclear transport domain
conferring tight targeting to the nucleus has been lost through divergent evolution of
the HD2D gene sequence. The rationalization of this result was inferred from a HD2C
deletion analysis, which allowed the mapping of a region within the HD2C sequence
which was required for exclusive nuclear localization. A KKAK motif was found to be a
critical determinant of the nuclear localization pattern. The significance of this motif
was further highlighted by the absolute conservation of the sequence across a sample
of monocot and dicots via a multiple sequence alignment, suggesting that each residue
was important for the nuclear localization. HD2D contained a slightly modified KKNK
motif at this position, which may explain its non-nuclear exclusive localization pattern.
The KKAK consensus sequence identified here matches a previously reported core NLS
sequence, designated [K(K/R)XK]. This core sequence relates to both monopartite and
bipartite sequences and appears to be a significant binding site for importin-alpha.
3.3.5 Nucleolar localization is not tied to nuclear localization
The HD2C-GFP deletion analysis indicated that the minimal NLS did not confer
nucleolar localization. Furthermore, mutation of the KKAK motif to AAAA resulted in a
localization pattern that was similar to that observed for the localisation of GFP
expressed on its own. It is known that GFP, when expressed alone, is localised to the
cytosol and due to its small size, can also be found in the nucleus (Grebenok, Pierson
et al. 1997). Thus it could be assumed that the mutant behaved like GFP on its own.
However, the mutant HDAC retained a degree of nucleolar localization which is not
found when expressing GFP alone. This suggests that the nucleolar targeting is not
directly tied to the NLS, but rather exists at a further sequence at the C-terminal site
on the HD2C protein. This is not a novel finding, as there is increasing evidence to
support the notion that NLSs and nucleolar localization sequences (NoLSs) are
66
recognized as separate signals by the cell. An example of this is the accumulation of
small, non-nuclear proteins that accumulate at the nucleolus but do not have a specific
NLS. From the HD2C-GFP deletion analysis it can be postulated that HD2C contains
both a NLS and NoLS which overlap, but are discrete. The functional consequences of
this are also clear, since attempts to challenge the cell with abiotic stress to see if
localization could be shifted from nuclear to cytosolic instead resulted in a shift to
become more nucleolar with no noticeable cytoplasmic leakage. Further
characterization of the NoLS may be a future direction as, similar to the NLS, reverse
genetic approaches can be used to mutate specific amino acids which confer this
localization pattern and test to identify if there is a phenotype. Based on the shift in
localization in response to stress, it would be logical to hypothesise that this would
reduce the salt stress tolerance that has been observed in plants over-expressing
HD2C.
3.3.6 Dimerisation does not require nuclear localisation of HD2C.
Initially it was hypothesised that nuclear retention was maintained either through
active transport or through accumulation of molecular mass, e.g. a higher order
protein complex in the nucleus which prevents leakage into the cytoplasm. The link
between dimerization and nuclear transport has been shown in a number of contexts.
MAPKK, a MAP-kinase kinase involved in cell cycle progression undergoes nuclear
import that is dependent on its dimerization state, where dimerization leads to active
transport of the large >80kDa complex. The monomeric form instead relies on passive
diffusion across the nuclear membrane to account for the proteins nuclear fraction.
Similarly, the effector protein AvrBs3 forms dimers in the cytoplasm prior to its import
to the nucleus following infection from the bacterial phytopathogen Xanthomonas
campestris. In either case a NLS was required for nuclear import of the larger complex,
but was closely linked to its dimerization state.
The identification of a sequence that conforms to a monopartite consensus sequence
suggests that active transport is the preferred method of translocation which, given
the small size of the HD2s (<40kDa), suggests that formation of the complex may occur
in the cytoplasm. This was consistent with the comparison of results between
localisation of HD2D as a GFP fusion protein and as a dimer with HD2C in BiFC studies,
67
with localization shifting from predominantly cytoplasmic to singularly nuclear and
nucleolar.
3.3.7 Conclusion and future
This study originally aimed to identify whether subcellular targeting was employed as a
regulatory measure to limit HD2C access to its histone substrate. It is now clear that
cytoplasmic-nuclear trafficking is not utilized in this manner; rather it appears that the
shift between nucleus and nucleolus may form a dynamic equilibrium which may be
perturbed by various abiotic stresses such as salt, heat and cold. In line with this, the
evidence for HD2C-HD2C association was confirmed using in planta BiFC with this
interaction evidently confined to the nucleolus.
Future work will concentrate on elucidating the molecular cues which orchestrate this
relationship. From these results the regions necessary for nuclear and nucleolar
localization have provided an ideal starting position from which these results can be
pursued. It appears that the active transport of nuclear localization is less significant,
as no evidence could be generated to suggest that this is a target for regulation.
Rather, its identification allowed to test and determine that dimer formation may
occur in the cytoplasm, and hypothesise that complex formation precedes protein
import. More interesting is the identification of the minimal NoLS, whereby nucleolar
localization was conferred. If this is indeed an important regulatory mechanism,
mutation of this region will yield a biologically important phenotype that will develop
some insight to its function when expressed in HD2C knock-out plants.
68
69
70
71
72
73
74
75
Chapter 4 Characterization of HD2C
interaction with 14-3-3 proteins using Bimolecular Fluorescent
Complementation
76
4.1 Introduction
4.1.1 Possibility of 14-3-3 interaction with HD2C in Arabidopsis thaliana
The possibility of interaction between HD2A-C and 14-3-3 proteins has previously been
raised by Paul et al using an antibody affinity purification technique (Paul, Liu et al.
2009). They identified interactors of 14-3-3 proteins with a monoclonal antibody
specific for an exposed domain of multiple Arabidopsis 14-3-3s. This approach enabled
the precipitation of a pool of 14-3-3 isoforms and therefore the co-precipitation of a
greater and more diverse number of 14-3-3 binding proteins. However, this approach
was limited as each potential protein must be further characterized on a case by case
basis to specify which of the individual 14-3-3s the identified protein bound to.
Furthermore, there is a growing body of evidence that suggests that protein
interaction data obtained from high throughput screens such as co-
immunoprecipitation is only biologically relevant in 30-50% of instances (Bader,
Chaudhuri et al. 2003). This is largely due to the disregard of native expression
patterns and the sub-cellular localization of each protein due to the pooling of plant
tissue. Various 14-3-3 isoforms have been shown to have specific expression patterns
and subcellular localizations. For example 14-3-3 omega and phi are widely expressed
in pollen with a cellular presence in both cytoplasm and nucleus. In contrast, 14-3-3 nu
is most highly expressed in the carpel and root tissue and has a predominantly
cytosolic localization pattern which appears to be excluded from the nucleus. As the
study by Paul et al did not discriminate between these isoforms, the data presented
requires further characterization to determine whether each interactor conforms to
these parameters.
The possibility of 14-3-3 binding to HD2C is significant, as mammalian HDACs have
been similarly shown to bind various 14-3-3 isoforms which have a significant impact
on its regulation. The interplay between HDAC4 and 5 with 14-3-3 and their
translocation between nucleus and cytoplasm has previously been discussed in chapter
3. It is likely that an interaction between 14-3-3 and HD2C will have regulatory
implications that can be exploited for the purposes of manipulating a relevant cellular
response. It is therefore of importance to determine if the binding of 14-3-3 to HD2C is
biologically relevant, and then to characterize this interaction to demonstrate what
regulatory role it has on HD2C.
77
4.1.2 14-3-3 background
14-3-3 proteins were first described in a 1967 study to identify the abundant, soluble,
acidic proteins of the mammalian brain. It has since been determined that 14-3-3s
represent a ubiquitous protein family present in all eukaryotes. Their name is derived
from their unique fractionation pattern on DEAE cellulose and starch gel
electrophoresis.
Despite their early discovery, it was only in 1996 that Shaw and colleagues identified
the fundamental function of 14-3-3s; that they bind to target proteins in a
phosphorylation dependent manner (Muslin, Tanner et al. 1996). This discovery led to
a subsequent re-evaluation of the model of phosphorylation dependent signals. It
became clear that in some cases phosphorylation was not sufficient for protein
modification. Rather, it acted as a docking site for proteins such as 14-3-3s, with their
binding sufficient for the steric interaction required for modification of the protein
activity. Structurally 14-3-3 proteins are highly conserved, with sequence homology
between plants and animals ranging from ~40-90%. These proteins consist of an N-
terminal dimerization domain and a C-terminal binding groove which is specific for
phosphorylated serine and threonine binding sites. X-ray crystallography has resolved
the structure for several 14-3-3 isoforms, revealing a dimeric structure which has
diagonal symmetry along two L-shaped monomers to produce a functional W-shaped
protein (Gardino, Smerdon et al. 2006). Each of the L-shaped monomers contains a
functional binding groove, such that the protein dimer is able to bind two
phosphorylated targets simultaneously. Binding of two sites simultaneously was shown
to cause a significant increase in binding stability compared to the binding at a single
site alone. In line with this, a significant number of binding partners have been shown
to contain multiple 14-3-3 binding sites with a spatial separation consistent with the
theoretical length separating the dimers binding grooves. Alternatively, it has been
speculated that the 14-3-3 dimer may function as a molecular bridge to link two
subunits together to stabilize a complex. This has been shown in the H+ ATPase plant
protein, where 14-3-3 binds to two adjacent subunits to stabilize the formation of the
octameric complex.
78
In almost all eukaryotes, the 14-3-3 proteins comprise large multigene families -
humans contain 7 highly conserved isoforms while there are 13 expressed isoforms in
Arabidopsis. Based on structure they are divided into two distinct clades; Epsilon-like
and non-Epsilon like (DeLille, Sehnke et al. 2001). The large gene family which has been
conserved across species led many researchers to speculate that it is functional
specificity of the different isoforms which has driven the evolutionary conservation.
There is some evidence to support this; 14-3-3-GFP fusion proteins constitutively
expressed in Arabidopsis had differing localization patterns, suggesting that this was
being driven by differing binding partners. Similarly, a number of mutant Arabidopsis
plants containing T-DNA knockouts of 14-3-3 genes show a variety of quantifiable
phenotypes which appear to be isoform specific (Purwestri, Ogaki et al. 2009; Tseng,
Whippo et al. 2012). Despite this, there is no clear evidence of a binding partner being
specific to only one 14-3-3 isoform, thus suggesting that there is functional redundancy
between 14-3-3 isoforms.
In plants and animals, various techniques have been used to identify literally hundreds
of putative interaction partners for 14-3-3 proteins. Of these, only a fraction have been
characterized to determine essential information such as their binding site, effect on
protein structure or indeed the regulatory role on their binding partner. What is clear
from the evidence compiled so far is that the binding of 14-3-3 to their target may
cover critical protein regions such as import/export sequences or catalytic domains, or
the binding induces structural changes that induce exposure or burying of these critical
regions. These changes were shown to affect three primary actions; catalytic activity,
localization or degradation state. Briefly, catalytic activity may be either induced such
as the Arabidopsis kinase CPK1, or repressed such as in the case of sucrose synthase
(Camoni, Harper et al. 1998; Toroser, Athwal et al. 1998). Localization may either
suppress import into cellular compartments by blocking a localization sequence, such
as in the case of the human CDC25C kinase (Graves, Lovly et al. 2001). Alternatively its
binding can expose a localization signal such as the mammalian KSRP protein (Díaz-
Moreno, Hollingworth et al. 2009). Structural changes can additionally impact
degradation, as shown in the example of nitrate reductase, where 14-3-3 binding
79
induces protein degradation, thus facilitating loss of enzyme activity in the cell (Weiner
and Kaiser 1999).
4.1.3 14-3-3 binding site
Determination of the role of 14-3-3 regulation is dependent upon the elucidation of its
binding site. Resolving this will allow a reverse genetics approach, whereby gene
mutations result in measurable phenotypes which allow an insight into the function of
this interaction. The search for 14-3-3 binding motifs appears to be somewhat further
progressed in mammals than in plants. Two well conserved binding site motifs have
been characterized for mammalian 14-3-3 binding partners; these are motif 1
[RX(Y/F)Xp(S/T)XP] and motif 2 [RXX(Y/F)Xp(S/T)XP]. From the combination of screens
which have occurred in mammalian systems, it was estimated that >60% of putative
targets contain at least one of these motifs (Fu, Subramanian et al. 2000). However,
other, likely less common motifs, were also identified. One of those is a motif that is
not required to be phosphorylated to allow 14-3-3 binding (Aitken 2006). This is
evident in the minimal binding site WLDLE which has been shown to bind to 14-3-3s
despite the evident lack of a phosphorylated residue (Muslin and Xing 2000; Yaffe
2002). From the limited characterization of non-phosphorylated targets, it appears
that this is the exception rather than the rule.
In plants, the relationship between primary sequence and interaction domain is less
clear. Meek et al identified >100 targets in their immunoprecipitation screen for 14-3-3
targets and noted that only ~40% of these contained the conserved motif 1 or motif 2
which are putative binding sites (Moorhead, Douglas et al. 1999). In part this may be
due to novel, plant specific binding domains. The H+ATPase 2 protein contains one
such domain, with RpSP shown to be the minimal peptide required for 14-3-3 docking
(Fuglsang, Visconti et al. 1999). Due to the obvious limitations imposed on
characterizing each binding site in a protein case-by-case basis, Panni et al used a novel
approach to screen for putative binding domains. They constructed a chip containing
randomized phospho-serine oligomers of 12AA in size and screened for interaction
with 14-3-3 Epsilon (Panni, Montecchi‐Palazzi et al. 2011). Rather than identifying new
discrete binding domains, they published a set of ‘rules’ to characterize the binding
80
sites. From the results it became clear that 14-3-3s preferred to bind to targets
sequences with surrounding positively charged residues, and were less inclined to bind
to peptides containing a proline at the +1 site relative to the phosphorylated residue,
or aspartate or glutamate residues in the sites surrounding the phosphorylated
residue. Together, the conserved binding motifs, requirement of phosphorylated
binding sites and rules that determine the possibility of 14-3-3 binding in the residues
surrounding the phosphorylation site means that a rational approach to determining
14-3-3 binding sites can be used to target specific areas of a 14-3-3 binding protein.
4.1.4 Aims and Hypotheses
Phosphorylation was shown to be necessary for a maize HD2 homologue to exhibit any
catalytic activity on its histone H3 substrate. Given the pull down data which suggested
that HD2C was a target for 14-3-3 binding, together with this evidence of
phosphorylation, two clear hypotheses can be made. The first is that the pull down
data reflects a true, in planta interaction which can be monitored by BiFC. The second
is that 14-3-3 is imposing a functional change on HD2C that is affecting its catalytic
activity. Given the previously characterized C-terminal catalytic domain and N-terminal
NLS and zinc-finger domain, the location of the 14-3-3 binding site may give some
insight into the mechanism of 14-3-3 action. For example, an N-terminal 14-3-3 binding
site would suggest that 14-3-3 binding is directly affecting the catalytic domain of the
enzyme. To characterize the putative 14-3-3 interaction with HD2C, the following aims
were made:
1. Determine the biological relevance of 14-3-3 interaction with HD2C. Firstly it is
important to identify which 14-3-3s are likely to bind to HD2C, and secondly
whether all HD2Cs can bind to 14-3-3s. BiFC will be used to ensure that an in
planta context is assured, and that subcellular localization is a factor that
determines whether there is interaction.
2. Identification of the 14-3-3 binding site on HD2C is necessary so that
manipulations of this site can be used to determine the functional consequence
of 14-3-3 binding. Therefore, HD2C deletion constructs will be constructed and
81
tested using BiFC to determine what HD2C regions are important for 14-3-3
interaction. Site directed mutagenesis may then be employed to determine the
critical residue required for 14-3-3 interaction.
3. A C-terminal 14-3-3 binding site may be critical for catalytic regulation, whereas
an N-terminal site may control nuclear or nucleolar localization. HD2C mutants
unable to bind 14-3-3 will therefore be used to determine what regulatory
function 14-3-3 binding has on HD2C.
4.2 Results
4.2.1 14-3-3 isoforms bind to HD2C in planta
Paul et al identified HD2C as a putative target for 14-3-3s when using the 14-3-3
exposed domain shared by multiple 14-3-3 isoforms (Paul, Liu et al. 2009). It was
necessary to delineate this information to determine which, if any, 14-3-3 isoforms are
specific for the interaction with HD2C. Micro-array data were first investigated by
performing a hierarchal clustering analysis of a subset of 14-3-3 and HD2 isoforms to
determine the context of co-expression in Arabidopsis tissue (Figure 4.1). The micro-
array data suggested that multiple 14-3-3 isoforms have overlapping, developmentally
regulated expression with HD2C. Most evident is that HD2C is most highly expressed in
flower tissue with especially high expression in carpel and ovules. Furthermore, strong
HD2C expression was found in seeds and embryonic tissue and root tips with weaker
expression in the root apical meristem.
Higher HD2C expression clusters most closely with the 14-3-3 isoforms nu, epsilon, psi
and upsilon. The best correlation was found to be between HD2C and 14-3-3 upsilon
with high expression in almost all HD2C high expressing tissues except for root cells,
seedling culture, the stigma of flowers and the embryo. In cell culture,/primary cells
and seedling culture, the abscission zone of flowers and root tips high HD2C expression
correlated with higher expression of 14-3-3 nu and epsilon (Figure 4.1). Expression of
HD2C and 14-3-3 psi was high in cell culture/primary cells, seeds and embryos, roots
and root apical meristem.
Kappa, Phi and Omega were somewhat related with co-expression found for 14-3-3 phi
(root tip), kappa and omega (abscission zone). No correlation with HD2C expression
82
was found for 14-3-3 lambda, mu, pi, iota and omicron. The most significant overlap
between the other HD2 isoforms was in root and root tip, abscission zone and
hypercotyl. This overlapped with expression of 14-3-3 epsilon and omega isoforms.
The data suggested that interaction should be relevant in multiple biological contexts
and can be somewhat 14-3-3 isoform specific. With the knowledge of co-regulation,
four 14-3-3 proteins were chosen for in planta interaction studies using Bimolecular
Fluorescence Complementation (BiFC). The proteins selected were from the 14-3-3
epsilon branch of the 14-3-3 family (epsilon and mu) and from the non-epsilon branch
(nu and psi). Three of the chosen 14-3-3 proteins had good correlations with HD2C
expression (epsilon, nu and psi) and one; mu showed little correlation in the
expression data obtained from Genevestigator (Figure 4.1). Ideally, 14-3-3 upsilon
would have been chosen as it had the best correlation with HD2C expression but the
cDNA encoding for this protein was unavailable at the onset of these tests. The cDNAs
for the four proteins were cloned in frame to the N-terminal of YFP in the PG179NS-YN
vector and were used to transform Agrobacteria in preparation for injection to N.
benthamiana leaves. This work had previously been undertaken by a colleague who
cloned and prepared PG179NS-YN and PG179NS-YC vectors for 12 isoforms of the
Arabidopsis 14-3-3 family (Hung Chi Liu, PhD thesis, UWA 2009). Additionally, the
library also contained N-terminal truncated versions of the 14-3-3 proteins in the two
vectors. Such truncated 14-3-3 proteins are expressed as YN or YC fusion proteins but
cannot dimerise as the first 1-50 amino acids function as a dimerization domain for this
class of proteins. These N-terminal 14-3-3 truncations, when co-expressed with HD2C
fused to the corresponding YN or YC vector, were therefore appropriate controls to
ensure fluorescence was not observed as a result of native reconstitution of the YFP
fluorophore in the absence of protein binding.
The 14-3-3 isoforms were tested with HD2C in a BiFC interaction analysis in N.
benthamiana leaves. Confocal microscopy was used to visualize and capture the
fluorescence in the epidermal layer of the leaves that was caused by the reconstitution
of YFP (figure 4.2). Strong and widely distributed fluorescence was observed in the
nucleus and nucleolus of the plant cells for each of the 14-3-3 HD2C combinations in
each sample, suggesting that interaction was possible between the two tested proteins
in planta. The nuclear and nucleolar fluorescence observed in the cells, was consistent
83
with previous HD2C-GFP analyses. This suggested either that 14-3-3 binding is
required to maintain the localization state of HD2C, or that the interaction did not
influence its localization state.
4.2.2 HD2 isoforms bind 14-3-3 epsilon
Testing all permutations of HD2s and 14-3-3 was superfluous given the evident
redundancy observed between HD2C and the epsilon and non-epsilon 14-3-3
homologues. However, identifying an in planta binding between the remaining HD2
homologues and 14-3-3 epsilon may yield some insight into potential binding regions
given the sequence diversity in some regions of the protein sequences between
homologues, particularly for HD2D. HD2A, B and D were therefore tested for
interaction with 14-3-3 epsilon using BiFC. As previously, HD2-YN and 14-3-3 epsilon-
YC constructs were co-expressed in N.benthamiana leaf epidermal tissue and
interaction of proteins was visualized using confocal microscopy. The N-terminally
truncated 14-3-3 epsilon was again as before used as a control to ensure that
fluorescence was not due to non-14-3-3/HD2 driven YFP reconstitution. As shown in
figure 4.3 all 14-3-3 epsilon/HD2 constructs resulted in strong fluorescence compared
to no evident fluorescence between HD2/deltatruncated-14-3-3 epsilon. As expected
the cellular distribution of HD2A, HD2B and HD2D when bound to 14-3-3 epsilon
mirrored that of their GFP fusion patterns; fluorescence was confined to the nucleus
and nucleolus for HD2A-B and was cytoplasmic and nuclear in the case of HD2D.
Together this data suggests that all HD2 isoforms may bind in vivo to both the epsilon
isoforms of the 14-3-3 proteins. Given the redundancy observed for interactions
between HD2C and 14-3-3 proteins, one can postulate that all other 14-3-3 isoforms
will also interact with all four HD2 isoforms and non-epsilon 14-3-3 proteins.
4.2.3 Analysis of HD2C 14-3-3 binding domains
The reverse genetic analysis to determine the biological significance of 14-3-3 binding
to HD2C requires an elucidation of the binding site. Subsequent mutagenesis to
remove this binding site will allow a direct comparison between the mutated HD2C
unable to bind 14-3-3 and the wild type enzyme which can be compared in the context
84
of the enzyme’s activity, localization and the phenotypic changes to the plant. As
described in chapter 3, HD2C consists of an N-terminal enzymatic domain, a central
acidic region with a neighbouring NLS and a C-terminal zinc finger. Recent phospho-
proteomic screens have identified a C-terminal phosphorylation site that exists
between amino acids 275-287, which may be a relevant site for 14-3-3 binding. A
deletion analysis coupled to BiFC analysis to test for interaction or loss thereof was
used to determine whether the 14-3-3 binding site on HD2C could be narrowed down
to this C-terminal region. Two truncated versions of HD2C corresponding to the N-
terminal domain and the remaining C-terminal of the protein were amplified by PCR,
cloned into the pg179NS-YC vector and transferred to N benthamiana. A deletion
fragment of HD2C consisting of amino acids 226-295 was sufficient to bind 14-3-3
epsilon (figure 4.4) as there was evident fluorescence in both cytoplasm and nucleus
(Appendix 3). However, the remaining N-terminal fragment of HD2C consisting of
amino acids 1-225 also resulted in fluorescence localized to the nucleus and nucleolus.
The differences in localisation between the two deletion constructs can be explained
by the loss of the NLS in the construct 226-295 (see also chapter 3; identification of a
NLS). These results therefore suggest that there are multiple 14-3-3 binding sites in the
HD2C sequence.
The presence of multiple 14-3-3 binding sites within a protein is not unique, as it has
been widely cited both within mammalian HDACs as well other 14-3-3 client proteins.
To obtain a better resolution of the protein regions containing 14-3-3 binding sites
while simultaneously testing for additional binding regions , a so-called overlapping
series of ‘tiling deletions’ was constructed. This series consisted of DNA fragments of
HD2C encoding for short (45-70 amino acids long) peptides. Together, these short
DNA fragments and the encoded peptides cover the length of the HD2C protein twice
with interloping peptides always covering the adjacent borders (figure 4.4). Each ‘tile’
was cloned into the pg179NS-YC vector and tested with 14-3-3 epsilon pg179NS-YN for
interaction in tobacco leaves using BiFC. Fluorescence indicating interaction of the
HD2C fragment with 14-3-3 epsilon are summarized in figure 4.4, with individual
fluorescent images shown in appendix 3. Interestingly, six of the ten peptide
fragments are able to interact with 14-3-3 epsilon. Based on the assumption that any
positive peptide could contain a 14-3-3 binding site at any location within its sequence,
85
this indicated that a minimum of three 14-3-3 binding sites were present in HD2C.
These putative binding sites overlap with the C-terminal zinc finger which contains the
previously annotated phosphorylation site, the NLS sequence and a region between
the central acidic and catalytic domain (figure 4.4). These results are also in agreement
with the analysis using the two larger deletion fragments described above which
suggested that there were both N- and C-terminal 14-3-3 binding sites. Western blot
analysis was used to demonstrate presence of each peptide to ensure that the absence
of fluorescence reflected a lack of protein interaction rather than little or no
expression. As expression was evident in each case, these results suggest there are no
14-3-3 binding sites between amino acids 1-86 and 159-181 of the HD2C sequence
(Appendix 4).
4.2.4 Determining the site of 14-3-3 binding to a single AA resolution
The ultimate aim in identifying binding domains is to generate non-binding mutant
proteins which then allow determination of the minimal binding sequence. In the case
of 14-3-3 proteins, it is known that binding domains for these proteins usually contain
at least one phosphorylated serine or threonine residue. Although non-phosphorylated
motifs were described, the presence of a phosphorylated amino acid appears to be the
most common feature of 14-3-3 binding sites and hence a good target to search for
when trying to identify a 14-3-3 binding domain. HD2C is comprised of ~25% serine
and threonine residues of which 35 serines and 14 threonines are present within the
three potential binding areas identified using the tiling assay (FIGURE 4.4). This large
number allows for a significant number of permutations which would have to be
studied to identify three binding sites. Hence a global mutagenesis of single serine and
threonine residues appeared impractical. Instead mutagenesis was confined to the
most C-terminal tile which corresponds to a peptide between amino acids AA265-294
with the aim to determine a single binding site on the C-terminal end and then to
extend this analysis to eventually encompass the entire protein sequence.
86
4.2.4.1 C-terminal 14-3-3 binding site
The C-terminal tile corresponding to AA265-294 binds to 14-3-3 epsilon to produce a
strong cytoplasmic reconstituted YFP derive fluorescent signal (Figure 4.5). Its
sequence contains eight serine and seven threonine residues. To increase the scope of
binding site disruption and reduce the number of mutants that would need to be
generated, substitution of these residues to alanine was applied to clusters of serine
and threonine residues with a total of five clusters mutated and analysed, designated
HD2C mutation 1-5 (HCM1-5, figure 4.5A).
Each mutant gene fragment was cloned into pg179NS-YC and tested with 14-3-3
epsilon pg179NS-YN in tobacco epidermal cells for interaction using BiFC (figure 4.5B).
Mutant constructs HCM1-4 had no impact on the fluorescence which was generated
by re-complementation of the YFP, suggesting that these are not required for 14-3-3
binding. Rather, modification of S284 and T286 to A284 and A286 (HCM5) significantly
reduced the level of fluorescence in each cell, such that fluorescent was no longer
detected at the same confocal microscope settings as in interaction studies of 14-3-3
epsilon with any of the other four cluster mutation peptides and the non-mutated 241
to 296 peptide. This suggests that this site is required for 14-3-3 binding. To further
resolve the site that is specifically required for 14-3-3 binding, the serine and threonine
residues were individually mutated to alanine residues and the mutant deletion
fragments were again tested using BiFC. Here, mutation of the T286 residue had no
effect on the fluorescent pattern, whereas mutation of S284 reverted to the low
fluorescence indicative of negative binding. These results therefore suggest that S284
is the site that is necessary for 14-3-3 binding to HD2C.
4.2.4.2 Second C-terminal 14-3-3 binding site
The tiling deletions presented in figure 4.4 suggest that a region between AA205-257 is
sufficient for 14-3-3 binding. It was therefore necessary to determine if the inclusion of
this region was sufficient to revert to the positive phenotype despite the HCM5
mutation. A deletion fragment consisting of residues 226-294 containing the H3M5
mutation was therefore constructed and cloned into the pg179NS-YN plasmid. It was
tested for interaction with 14-3-3 epsilon using BiFC in transiently transformed N
benthamiana leaf tissue. As shown in figure 4.6, fluorescence was detected, suggesting
87
that 14-3-3 interaction had been recovered. Candidate serine and threonine residues
were less abundant in this additional region, with only T235 and S239 present. These
sites were therefore the target for site directed mutagenesis, with substitutions to
alanine again used (figure 4.6).
BiFC was again used to trace interaction between the HD2C mutant and 14-3-3 epsilon.
As shown in figure 4.7B, a construct was cloned into the pg179NS-YN plasmid that
consisted of the 226-294 residues which contained alanine substitutions of T235, S239,
S284 and T286; this was designated HCM6. This construct was used to co-transform N
benthamiana leaf tissue via Agrobacteria injection along with 14-3-3 epsilon fused to
the C-terminal YFP fragment. After three days, it was evident from confocal
microscopic imaging that there was a significant reduction in fluorescence compared
to the non-mutated 226-294 HD2C fragment. This indicated that mutation of HCM6
resulted in the removal of a 14-3-3 binding site. Therefore, this suggests that this site is
required for 14-3-3 binding.
4.2.5 HD2C-mutants with disrupted C-terminal 14-3-3 binding do not have a clear
shift in localization pattern
14-3-3 binding affects target proteins function by modifying enzymatic activity,
subcellular localization and degradation state. Enzymatic activity was reported to
centrally involve the N-terminal of the protein, and was therefore unlikely to be
directly affected by these C-terminal sites. In chapter 3, a site central to nucleolar
localization was determined to lie between AA 206-257, thus overlapping with the 235-
TPHPS-239 putative 14-3-3 binding site. The potential role in regulating nucleolar
retention by 14-3-3 binding was therefore investigated by testing the mutated
constructs as GFP fusion proteins and tracing their subcellular distribution in tobacco
leaf tissue.
The HD2C full-length genes were mutated with the same serine- and threonine to
alanine substitutions as in HCM5 and HCM6 above. Mutation was imposed on the full
length HD2C protein so that the resulting localization pattern of the fusion protein
could be compared with the localization pattern of the wild type HD2C fused to GFP.
As shown in figure 4.8, mutating the 14-3-3 binding sites at either of the positions had
88
no evident effect on the subcellular localization of the HD2C protein. From these
results it was therefore assumed that regulation of HD2C by 14-3-3 must involve a
different form of regulation or if localization is affected by 14-3-3 binding, must require
a different 14-3-3 binding site.
4.3 Discussion
4.3.1 Summary
Here, the interaction of 14-3-3 and HD2C was investigated using Bimolecular
Fluorescent Complementation. It was shown that HD2C was able to bind both epsilon
and non-epsilon classes of 14-3-3 proteins in transiently transformed N.benthamiana
leaf tissue. Similarly, all HD2 isoforms were able to bind to the 14-3-3 epsilon isoforms
using the same approach. A deletion analysis of the HD2C protein was then used to
map the interaction of HD2C with 14-3-3 epsilon, revealing three distinct regions that
harbour a 14-3-3 binding site. Site directed mutagenesis of possible serine and
threonine residues that could correspond to the critical phosphorylated binding site on
the C-terminal region of the protein revealed that residue S284 and one of T235 and
S239, when substituted to alanine, abolished 14-3-3 binding. Lastly, the link between
14-3-3 binding and HD2C localization was investigated by comparing the fluorescent
pattern of a full length HD2C-GFP fusion that contained the mutants which perturbed
14-3-3 interaction with a wild type HD2C-GFP protein. No difference in localization was
observed, suggesting that these sites were not involved in controlling the localization
pattern of HD2C.
4.3.2 HD2C does not show preference to 14-3-3 isoforms
14-3-3 proteins function as dimers that bind phosphorylated peptides and coordinate a
regulatory change by modifying the protein structure of their bound target. It is
important to identity the target for each 14-3-3 isoform given the possibility of
functional specificity implied by the varying 14-3-3-GFP subcellular localizations,
varying expression patterns and the evolutionarily conserved large gene family.
Furthermore it is possible that specific heterodimers comprised of co-expressed 14-3-3
homologues can drive the interaction of targets specific for each homologue.
89
In this study, the interaction of HD2C and 14-3-3 epsilon was confirmed using BiFC in
transiently transformed N. benthamiana tissue. Given this result, and the fact that
HD2C contains a very specific nuclear and nucleolar localization pattern which is not
commonly observed in some 14-3-3 homologues, HD2C provided a model protein on
which to test the redundancy of 14-3-3 function. The combination of HD2C with
epsilon (epsilon and mu) and non-epsilon (nu and psi) 14-3-3 proteins in planta
resulted in a fluorescent pattern that was localized in the nucleus and nucleolus. This
result suggests that there is no 14-3-3 specificity between isoforms with regards to
HD2C binding, with both branches of the 14-3-3 family evidently capable of binding to
the HD2C. The significance of this result must be tempered by the obvious issues
relating to expression and the non-quantitative nature of the BiFC assay. Historically,
the debate into 14-3-3 target specificity has centred on the competing ideas of total
functional redundancy vs. recognition and binding of specific targets. However, more
recently studies have focussed on the gradient of target specificity, whereby target
proteins range from ideal to poor binding partners and it is this gradient which drives
specific responses from 14-3-3s within the cell. For example, 14-3-3 isoforms from
barley were tested using quantitative yeast 2-hybrid to measure their affinity to
sucrose phosphate synthase, revealing drastic differences in binding potential. This is
not an inherent 14-3-3 isoform specific binding affinity profile, as the binding affinity
for H+ATPase and nitrate reductase with an overlapping subset of Arabidopsis 14-3-3
isoforms were tested and shown to have unique preferences for 14-3-3 isoforms.
Similarly, the 14-3-3 isoforms chi and epsilon were shown to differentially bind client
proteins from developing Arabidopsis seed, suggesting functional specialization in this
developmental process (Swatek, Graham et al. 2011). These results appear to suggest
that there is no utility in defining functional redundancy as a binary bind/not bind
scenario. Rather, each 14-3-3 client has an intrinsic preference for 14-3-3 which is
defined by unique expression patterns and binding affinities for the 14-3-3s.
Consistent with this observation is that the BiFC assay used in this analysis is not
sufficiently quantitative to differentiate this level of functional specificity. CaMV-35S
promoter driven expression and subsequent translation of 14-3-3-NYFP fusion proteins
result in saturating levels of 14-3-3, ultimately causing fluorescence to be observed in
even weak protein-protein interactions. As such, this study alone can have little input
90
into the issue of functional redundancy within the 14-3-3 gene family. Despite this, it is
clear from the combination of this result, and that obtained by the pull-down data
(Paul, Liu et al. 2009), that the HD2 family is indeed a target for 14-3-3 binding. The
fact that binding was possible with each 14-3-3 isoform is significant, since the
ubiquitous expression of the 14-3-3 family and their localization to the nucleus means
that a phosphorylated HD2 is likely to always have access to a 14-3-3 pool with which it
may bind. This result therefore suggests it is not the 14-3-3 protein itself that is
limiting, but rather it is the phosphorylation which is the critical signalling step
required to induce any association of 14-3-3 and HD2C.
4.3.3 HD2C contains multiple 14-3-3 binding sites
From the tiling deletion assay it was evident that there were multiple 14-3-3 binding
sites on the C-terminus of HD2C. Using site directed mutagenesis it was shown that in
the context of the C-terminal domain, mutation from TPHPS-APHPA and SHT-AHA
resulted in loss of 14-3-3 binding to the 226-294 amino acid C-terminal HD2C
fragment. The occurrence of multiple 14-3-3 binding sites on a protein is not rare and
indeed has been proposed to be a principle mechanism by which 14-3-3 stabilizes non-
native conformations in its ligand (Yaffe 2002). The most overt hypothesis would be
that each serine binds each to one of the monomers of the 14-3-3 dimer to result in a
stable HD2C/14-3-3 complex. Indeed, it was calculated that 14-3-3 dimers that bind to
dual sites may result in cooperative binding that has several orders of magnitude
higher affinity than single sites. The 14-3-3 dimer has diagonal symmetry between its
monomeric parts, with the binding grooves which has been calculated to be separated
by approximately 34.4 Angstroms from the human 14-3-3 zeta crystal structure (Liu,
Bienkowska et al. 1995). Assuming a linear, extended peptide structure, this would be
sufficient to bridge a ~15 amino acid peptide. The two sites identified in HD2C are
more distantly linked, being at least 44 amino acids apart. This may suggest that a
single 14-3-3 dimer does not simultaneously bind these two serine residues. Assuming
that the two binding domains of the 14-3-3 dimer are binding phosphorylated targets,
this may be consistent with the hypothesis that 14-3-3s stabilize the association of two
proteins in a complex by acting as a molecular bridge. Given the findings presented in
91
section 3, the role of 14-3-3s in maintaining a HD2 multimer may be a direction of
future research.
Alternatively, the 14-3-3 binding sites may be within the critical 34.4 Angstroms
proximity in the context of a mature tertiary structure. For example, the mammalian
transcription factor FOXO4 has been shown to carry two 14-3-3 binding sites at Thr28
and Ser193, yet allow simultaneous binding as the two tails containing the binding
sites are occluded from its central forkhead domain (Obsilova, Vecer et al. 2005). No
HD2 structure is currently available from the pfam database, so a detailed analysis of
the structure was not able to be performed. However, the two 14-3-3 binding sites
occurred within and besides a well-defined TFIII-type zinc finger domain. This consists
of two cysteine residues arranged antiparallel with the two histidine residues, together
arranged to complex with a central zinc ion. The antiparallel fold between cys2 and
his2 can be modelled to show that, in isolation, the zinc finger motif may be conducive
to allow a single 14-3-3 motif to contact both serine residues. If this is the case,
simultaneous binding of the 14-3-3 dimer is probably functionally significant and will
require further characterization to elucidate whether binding is an independent event.
4.3.4 Identification of 14-3-3 binding sites on HD2C
Here a site-directed mutagenesis approach tied with BiFC to determine the 14-3-3
binding sites of HD2C. In Johnson et al’s bioinformatics survey of 14-3-3 binding sites,
they suggested that true 14-3-3 binding should be annotated following a defined
procedure that included (1) identification of phosphorylation of the critical residue in
vivo; (2) elimination of the critical serine or threonine residue by mutation or
dephoshorylation; (3) ensuring that in vitro phosphorylation was consistent with in
vivo phosphorylation; (4) structural analysis of 14-3-3 and its target protein binding;
and, (5) where there are multiple 14-3-3 binding sites, to ensure that mutation of one
14-3-3 binding site does not disrupt the binding of another 14-3-3 binding site
(Johnson, Crowther et al. 2010). The work presented in this thesis addressed each of
these criteria with varying degrees of sufficiency with respect to the binding of 14-3-3
to HD2C.
The link between phosphorylation and 14-3-3 binding was only investigated
circumspectly, with preferential investigation of serine and threonine residues as
92
putative 14-3-3 binding sites. Previously it was shown that phosphorylation occurred at
three sites in a maize HD2 homologue on serine residues (Wang, Kruhlak et al. 2000).
Assuming that this has been conserved, this result was consistent with the minimal
three 14-3-3 binding sites described from the tiling deletion analysis. Recently, it has
been shown using phospho-proteomic analyses that HD2C is phosphorylated between
AA270-290, a site that overlaps with the putative 284-SHT-286 14-3-3 binding site
(Durek, Schmidt et al. 2010). Given the fact that 14-3-3 binds to phosphorylated
residues in the vast majority of characterized cases, and that of all the mutated serine
and threonine residues contained on this peptide only the mutated SHT-AHA was not
able to bind, this suggests that the true 14-3-3 binding site has been modified. The
235-TPHPS-239 site has not been similarly characterized. As suggested by Johnson et
al, further characterization by identifying the site of phosphorylation to a single amino
acid resolution is required to lend support to the hypothesis that the two putative 14-
3-3 binding sites raised in this study are functionally significant.
N.benthamiana as opposed to Arabidopsis tissue was used to express the fusion
protein because it ensures a more consistent transformation efficiency, which has
been widely used and documented in literature. BiFC was used to validate protein
interaction and subsequently perturb this interaction by mutation. This ensured that
the protein was expressed in an in planta context rather than the traditional bacterial
expression and pull-down analysis. The advantage of this was that it ensured that
protein binding was biologically relevant with respect to protein co-localization and
that phosphorylation occurred in the correct context because it was a substrate for its
specific kinase. Furthermore, the fluorescence between mutated and wild-type HD2C
could be directly compared and semi-quantified using pixel saturation look-up tables
on the confocal microscope. The comparison between positive and negative
interaction was consistent with the difference in fluorescence observed between
positive and negative controls, with Western blot showing a comparable level of
expression in the two cases.
93
4.3.5 The HD2C 14-3-3 binding site does not correspond to any consensus 14-3-3
binding motif
In the analysis of 14-3-3 binding sites in the C-terminus of HD2C, regions were shown
to affect 14-3-3 binding. This site corresponded to a region of sequence conservation
between the Arabidopsis HD2A-C homologues and other HD2C homologues. Earlier in
their history, 14-3-3 targets were identified and characterized on a case basis which
fostered easy characterization of target proteins with similar binding sites. Mammalian
proteins were shown to bind 14-3-3 proteins in common consensus motifs designated
as motif 1 [RXXpS/pTXP] or motif 2 [RXXXpS/pTXP] (Yaffe 2002). In plants it was
reported that there is a minor modification of this binding motif, as the consensus
sequence [LX(R/K)SX(pS/pT)XP] is instead over-represented in a number of 14-3-3
binding proteins. Additionally the novel binding site on the penultimate
phosphorylated threonine of H+-ATPase AHA2 was shown to be dependent on the
tyrosine-threonine-proline sequence. However, approaches such as yeast-2-hybrid and
co-immunoprecipitation which screen blindly for 14-3-3 binding proteins have resulted
in lists of proteins that bind 14-3-3 proteins, yet contain no corresponding consensus
motif. For example in mammalian targets only ~60% contain a motif 1 or 2, while
Arabidopsis targets only contain the consensus motifs in ~40% of cases. From the large
number of motif 1 and 2 containing targets already resolved, it is clear that these sites
are significant 14-3-3 binding motifs. However, it is also clear that further work is
required to resolve the binding motifs in these non-consensus targets.
As mentioned previously, Panni et al attempted to resolve this using chip-based
approach where random 12-amino acid peptides were washed with purified 14-3-3
protein and assessed to determine whether binding had occurred (Panni, Montecchi‐
Palazzi et al. 2011). Rather than identifying any clear binding motifs, they suggested
that 14-3-3 binding was strongly inhibited by negatively charged amino acids in the -3
and -2 position relative to the phospho-serine, and binding was increasingly favoured
in cases where lysine or arginine occupied the -1 and +2 position relative to the
phospho-serine. Ignoring the biological prerequisite kinase to induce this state in
nature and the fact that it assessed binding of a mammalian 14-3-3 protein, this set of
rules can be applied here in the analysis of the HD2C binding site. The consensus motif
derived from orthologous HD2C protein sequences is [(A/G)LASH(S/T)KAKH] and
94
[H(V/I)ATPHP(S/A)K]. Consistent with the published rules, this motif contains no
inhibitory glutamate or aspartate residues, and it contains lysine residues on the C-
terminal end of the motif.
The accumulation of evidence therefore suggests that these putative 14-3-3 binding
motifs are at least important to 14-3-3 binding, with the evidence weighted towards
the hypothesis that it is the critical phosphorylated residue that binds 14-3-3 directly.
As outlined previously, future research must be directed to fully elucidating this site,
most critically to determine whether the critical residue is phosphorylated.
4.3.6 Conclusions and future
This work has contributed new understanding to both 14-3-3 and HD2C proteins by
characterizing a small aspect of their interaction. Here we show that multiple 14-3-3
isoforms may bind to HD2C at a minimum of three sites along its sequence.
Furthermore one of these sites was characterized as a novel binding site which
overlaps its zinc-finger motif and does not appear to impact on the proteins
localization.
As previously mentioned the C-terminal binding site is, from this data alone, a putative
interaction site that requires further investigation to determine whether it is a true
phosphorylation site that physically interacts with the 14-3-3 protein. Lusser et al’s
original characterization of the HD2 isoform purified active protein from maize
embryonic tissue which was analysed for phosphorylation using phospho-sensitive
antibodies by Western blot (Lusser, Brosch et al. 1997). Technological limitations at the
time prevented a higher resolution analysis which could resolve these phosphorylation
sites. Phosphoproteomics is a possible method of investigation. It relies on the
purification of HD2 protein from plant cells and fragmentation of protein followed by
mass spectrometric analysis to resolve the possible phosphorylation site by comparing
m/z of the various fragments with the expected m/z calculated by the amino acid
sequence alone. This has already been used to identify the phospho-peptide which
overlaps with the putative 14-3-3 binding site highlighted in this study. Assuming that
this approach confirms phosphorylation of the critical 14-3-3 binding site, this would
lend sufficient evidence to suggest that this is a true binding site for physical 14-3-3
95
binding to HD2C. Final confirmation necessitates x-ray crystallography to visualize this
interaction and would provide information on the structural implication of 14-3-3
binding on the zinc-finger domain.
The other unresolved information from this study was that no functional implications
of 14-3-3 binding could be determined. 14-3-3s have been shown to have a number of
functional implications on their target proteins, including regulation of localization,
enzyme activity, degradation state or regulation of tertiary structure. The prospect of
localization change has already been addressed in this study, suggesting that there is
no evident role from the mutation analysis when fused to GFP. The enzymatic
implications would appear to be irrelevant given the previous determination that only
the region from 1-120 amino acids is required for HDAC activity.
96
97
98
99
100
101
102
103
104
Chapter 5
Role of HD2C in salicylic acid and jasmonate response in Arabidopsis
thaliana
105
5.1 Introduction
5.1.1 Hormone signalling in plants
Plants are sessile and therefore require fast recognition and response to the various
environmental conditions that they are exposed to. One important mechanism to
achieve such responses is via plant signalling molecules, the so-called ‘plant
hormones’. Plant hormones are internally-secreted chemicals that occur at low
concentrations and function as signalling molecules to regulate plant growth and other
responses. Unlike the traditional mammalian hormones, plant hormones may be
synthesised in a range of tissues or cell types and may act either in the same tissue or
cell type of synthesis, or be transported throughout the plant for a holistic response to
specific stimuli (Moore 1979; Davies 2010).
The broad definition of plant hormones has led to significant discussion over the
classification of various chemicals as a hormone. The consensus appears to be that
there are five major hormone groups; auxin, abscisic acid, gibberellins, ethylene and
cytokines, while several other chemicals including jasmonates, polyamines, salicylic
acid and brassinosteroids are widely accepted as fulfilling the parameters to be termed
plant hormones(Kende and Zeevaart 1997; Davies 2010). For the purposes of clarity in
this thesis, all the listed chemical classes will be referred to as plant hormones.
5.1.2 Possible role of HD2C signalling in response to salicylic and jasmonic acids
Experimental evidence suggests that HD2s may have a role in salicylic acid (SA) and
jasmonic acid (JA) signalling. In barley the expression of HvHDAC2-1 was shown to
increase in response to exogenously applied SA and JA (Zhou, Labbe et al. 2004).
Furthermore HD2C was shown to interact with HDA19, which itself was shown to be
centrally involved in SA and JA signalling (Choi, Song et al. 2012; Luo, Wang et al.
2012). For example expression of HDA19 corresponded to repression of SA
biosynthesis and SA responsive genes, while conversely leading to an upregulation of
ERF1 and expression of jasmonic acid and ethylene regulated PATHOGENESIS-RELATED
genes, Basic Chitinase and β-1,3-Glucanase (Choi, Song et al. 2012). ERF1 was similarly
shown to bind a HD2 homologue in Longan, suggesting a common mechanism of
control that may exist between the two HDAC families (Kuang, Chen et al. 2012).
106
Identifying firstly whether HD2C is involved in SA and JA signalling and secondly
elucidating the mechanism of this pathway is clearly required.
5.1.3 Salicylic acid
Salicylic acid is a critical signalling molecule whose production is induced by biotrophic
pathogens (pathogens that infect and feed on a living host cell) and modulates the
activation of defence related genes (Métraux, Signer et al. 1990; Cameron, Paiva et al.
1999). Its primary roles are to firstly to enable plants to survive the pathogenic
infection and secondly to stimulate the induction of a longer lasting systemic acquired
resistance (SAR) (Lawton, Weymann et al. 1995). Irrespective of its source of induction,
SAR primes the entire plant for subsequent defence against a broad range of
pathogens including pathogenic bacteria, fungi and viruses (Cameron, Paiva et al.
1999). Exogenously applied SA leads to the onset of SAR, while the removal of SA by
ectopic expression of salicylic hydroxylase prevents the onset of SAR (Lawton,
Weymann et al. 1995). This led to the realization that SA is both sufficient and
necessary for the physiological induction of SAR in a number of SA-dependent
pathways and is therefore a critical component of plant stress response (Shah 2003).
The SA mode of action is fundamentally tied to its activation of defence genes, which
itself is tied directly to the action of the SA response protein NPR1 (Shah and Klessig
1999). T-DNA insertion knockouts of NPR1 accumulate SA but are unable to induce SAR
and are more susceptible to a wide range of pathogens (Schenk, Kazan et al. 2000;
Zhang, Tessaro et al. 2003; Zhang, Francis et al. 2010). Conversely, over-expression of
NPR1 leads to enhanced capacity for induction of SA mediated plant defence and
enhanced disease resistance to a wide range of pathogens (Zhang, Francis et al. 2010).
NPR1 exists as oligomers in the cytoplasm an its structure is stabilised/maintained via
covalent attachment of disulfide bridges (Tada, Spoel et al. 2008). SA accumulation in
the cytoplasm leads to an alteration in the redox state, leading to the monomerization
of NPR1 and subsequent import into the nucleus (Mou, Fan et al. 2003). Here, NPR1
interacts directly with a number of TGA transcription factors (TGA1to 6) and WRKY
transcription factors 18, 53, 54 and 70 (Zhang, Fan et al. 1999; Després, DeLong et al.
2000; Dong 2004; Eulgem and Somssich 2007). Binding by NPR1 appears to activate
107
these factors, leading to the immediate activation of SA-early response genes,
followed by the sustained induction of SA-late response genes.
Identification of the SA-early response genes following SA induction is of current
research interest as it is expected to reveal insight into the early pathogen response of
the plant (Uquillas, Letelier et al. 2004). The best characterized examples include
GLUTATHIONE-S-TRANSFERASE 6 (GST6) and (IEGT) (Chen and Singh 1999; Uquillas,
Letelier et al. 2004). From analyses it is evident that these early response genes are
responsible for mediating changes such as recovery of the cell redox balance,
intracellular stress signalling, improvement of pathogen recognition and metabolic
changes (Schenk, Kazan et al. 2000; Galis and Matsuoka 2007). The SA-late response
genes assist in the latter stages of pathogen response and induction of SAR for the
entire plant. The best characterized example of this are the PATHOGEN RELATED (PR)
gene family, of which PR-1 is commonly used as a marker of SA activated SAR (Eulgem
2005).
5.1.4 Jasmonic acid and methyl-jasmonate
Similar to salicylic acid, jasmonic acid and its jasmonate derivative methyl-jasmonate
(MeJA) are critical signalling molecules that are induced by plant stress (Xu, Chang et
al. 1994). However whereas SA is induced by biotrophic pathogens, JA production is
locally stimulated by plant wound response following mechanical damage or herbivory
(Thaler 1999). Additionally the application of exogenous JA or MeJA induced
expression of defence related genes and a correspondingly increased resistance to
herbivorous challenge. The molecular mechanism of plant defence activation by JA
appears to centrally involve three players: ubiquitin ligase complex, SCFCOI1 and JAZ1/3
(Chini, Fonseca et al. 2007; Thines, Katsir et al. 2007; Fernández-Calvo, Chini et al.
2011). In summary, prior to JA stimulation JAZ1/3 is bound to JIN1 and represses its
ability to activate JA response genes. Upon JA stimulation, SCFCOI1 complex binds
JAZ1/3, leading to the ubiquitination and subsequent degradation of JAZ1/3. This frees
JIN1 and thus allows activation of the expression of JA-responsive genes (Manners,
Penninckx et al. 1998; Thomma, Eggermont et al. 1998) . The most widely studied JA-
responsive gene is PDF1.2, which accumulates rapidly in response to exogenous JA
application (Manners, Penninckx et al. 1998).
108
Aside from the evident role in defence response, the effect of jasmonates on a plant
development and growth are broad. Inhibition of plant growth, seed germination, root
growth and photosynthesis and promotion of senescence, tuber and pollen formation,
tendril coiling and fruit ripening have been correlated to increased endogenous JA
(Creelman and Mullet 1995).
5.1.5 SA/JA crosstalk
SA and JA are centrally involved in the plant defence response but are activated by
differing stimuli; SA by biotrophic pathogens and JA by necrotrophic pathogens and
insect herbivory (Pieterse and van Loon 1999; Glazebrook 2005). Activation of the
defence response by these differing stimuli results in an initial activation of distinct
pathways, although there is significant overlap of factors further down the defence
pathway (Traw and Bergelson 2003). In order to produce a response that is specific to
the pathogen that initially stimulates the response there needs to be cross-talk
between the two pathways to ensure that co-activation is not initiated which may
dilute the defence response and cost unnecessary energy for the plant (Pieterse and
van Loon 1999).
There appears to be a significant amount of crosstalk between the SA and JA response
pathways, with most evidence pointing to an antagonistic imposition between the
pathways. A superficial observation is that increased biotrophic resistance results in
decreased necrotroph resistance; conversely increased necrotroph resistance results in
decreased biotroph resistance (Robert-Seilaniantz, Grant et al. 2011). There is a
molecular basis for this observation. For example in tobacco, JA responsive basic PR1
genes were inhibited by exogenously applied SA while SA responsive acidic PR1 genes
were inhibited by exogenously applied MeJA (Malamy and Klessig 1992). Determining
the players of this crosstalk between each pathway is necessary to rationalize the
subtle defence response that plants use to combat their pathogens.
The SA antagonism towards JA-responsive genes appears to centre on the action of
NPR1 (Spoel, Koornneef et al. 2003). Treatment of Arabidopsis with both SA and MeJA
resulted in activation of SAR, but simultaneous repression of JA-responsive genes. This
phenotype was altered when NPR1 was knocked out by T-DNA insertion, as JA
109
responsive genes where instead activated (Shah, Kachroo et al. 1999). This suggested
that NPR1 was required for SA mediated JA-responsive gene repression. The
mechanism is still being determined; however, it appears that NPR1 can directly block
JA synthesis in the cytoplasm as well as interact with transcription factors in the
nucleus to directly repress JA-responsive genes expression. For example WRKY70 and
WRKY33 were shown to activate the SA-responsive PR1 and PR2 genes while
simultaneously repressing the JA-responsive PDF1.2 gene after activation by NPR1 (Li,
Brader et al. 2004). Additionally, NPR1 interacts with clade II TGA transcription factors
(TGA2, TGA5 and TGA6) which induce expression of some JA-responsive genes, but
repress these same genes when bound to NPR1 (Després, DeLong et al. 2000;
Ndamukong, Abdallat et al. 2007). Because of these dual roles, it has been suggested
that TGA factors intersect with the SA and JA defence pathways as both antagonistic
and synergistic factors to both direct and fine tune the plant defence response.
5.1.6 Aims and Hypotheses
HD2s were shown to be involved in hormone signalling via the ABA abiotic stress
pathway. Expression of HD2C was repressed by ABA and the expression of several
ABA-responsive genes was under-expressed compared to Col-0 in 35S:HD2C-GFP
plants (Luo, Wang et al. 2012). Here it is hypothesised that HD2C may be similarly
involved in the biotic plant defence response, where HD2C mediates the expression of
genes related to JA- or SA- response genes. Underlying changes in gene expression are
evident in related plant phenotypes, therefore it is likely that if HD2C has a role in SA
and/or JA response pathways, there will be a measurable phenotype. Therefore, the
aims of this section are:
1. Determining whether the expression of HD2C is controlled by SA or JA levels in
the cell by treating Arabidopsis plants exogenously with SA or MeJA and
measuring the expression of HD2C by RT-PCR.
2. Use the two constitutively expressing HD2C-GFP lines prepared during the work
described in chapter three to compare the growth of these Arabidopsis lines to
the wild type in response to SA and MeJA.
110
3. Determine the molecular mechanism of HD2 action on the SA and JA pathway
by testing interaction with proteins that are predicted to bind HD2C and which
are related to the plant defence response.
5.2 Results
5.2.1 Expression of HD2 proteins when exposed to SA, INA and MeJA
Accumulation of plant hormones in a cell is associated with the induction or repression
of the associated hormone response genes, allowing a combinatorial response from
the stimuli. Given the possible role in SA and JA responses, it was important to
determine if the exogenous application of SA and JA had an effect on the mRNA
accumulation of HD2 genes. Previously it was shown that 50 μM SA was sufficient to
induce a stress response in Arabidopsis seedlings. The methyl-ester of jasmonic acid,
MeJA, is a more active effector of the jasmonic acid pathway and was also applied at
50 μM concentrations to induce plant stress. Col-0 Arabidopsis seedlings were grown
on 0.5MS for two weeks before transferring to MS plates containing 50µM SA, 50µM
MeJA or no treatment. RNA was extracted and semi-quantitative RT-PCR used to
measure gene expression of HD2C (figure 5.1). As a positive control to ensure that
induction of the plant defence response had occurred, the SA responsive PR1 and
MeJA responsive PDF1.2 were amplified in parallel. The housekeeping gene actin was
used as a control to ensure that any differences were attributed to the effects of
hormone rather than the starting quantity of cDNA.
The actin housekeeping gene expression did change only slightly between individual
samples indicating that starting amounts of RNA were about equivalent between the
treatments (figure 5.1). Hence, any difference observed for the other genes could have
been attributed to the treatment differences. Comparing HD2C expression after
hormone treatment with the MS control treatment showed that there was no obvious
effect on the expression of HD2C in response to either SA or MeJA (figure 5.1).
Expression of PR1 was visibly increased in response to SA treatment, while PDF1.2
appeared to be slightly induced by MeJA treatment. This suggested that despite the
positive control markers indicating that the plant defence response had been
activated, there was no clear evidence to suggest that HD2C expression had been
altered in response to the treatments.
111
5.2.2 HD2C expression has no impact on root growth when exposed to SA or MeJA
Both SA and MeJA have a retarding effect on the root growth of Arabidopsis plants.
Furthermore, the role of HD2C expression was shown to have a statistically significant
effect on root growth in Arabidopsis seedlings when grown on 50µM ABA, with over-
expressing lines having reduced root growth (Sridha and Wu 2006). This therefore
demonstrates that HD2C plays a role in root development and growth. Here, it was
tested if a link exists between the expression of HD2C and the root growth of
Arabidopsis seedlings when exposed to 50 µM SA or50 µM MeJA (Staswick, Tiryaki et
al. 2002). 35S:HD2C-GFP line 1 and line 2, together with Col-0 seeds were germinated
and grown on vertical plates containing half-strength MS until roots had reached ~1cm
in length. Seedlings were then transplanted to vertical plates with half strength MS
containing either 50mM SA, 50mM MeJA or no hormone additive (control). Root
lengths were measured after eight days using the Neural-J plugin from Image-J picture
software (Abràmoff, Magalhães et al. 2004). Both SA and MeJA treatments caused
reduced root growth in each of the Arabidopsis lines. Root lengths were compared
between each transgenic line and the Col-0 wild type for each media by collecting
images of each plate and measuring the distance in pixels between start and end root
length.
As shown in figure 5.2, after eight days the root growth in the wild type on control
media was 845 px, on SA media it was 423 px and on media containing MeJA it was
253px. Compared to this, the roots of 35S:HD2C-GFP line 1 grew 981px on control
media, 438 px on SA media and 254px in the presence of MeJA.The second transgenic
line, 35S:HD2C-GFP line 2, had root growth of 840px on control media, 511px on SA
and 238 px on MeJA media. Thus the relative root growth for the wild type were2:1
(control:SA) and 3.3:1 (control:MeJA). This was similar to the values for the transgenic
line 1 with ratios of 2.2:1 (control:SA) and 3.9:1 (control:MeJA) and for line 2 with
ratios of 1.6:1 (control:SA) and 3.5:1 (control:MeJA). The relative growth rates
between both transgenic lines compared to the wild type were statistically not-
significant at a p-value<5. Together these data suggest that HD2C expression does not
affect root growth in response to SA or MeJA hormone treatment.
112
5.2.3 35S:HD2C-GFP plants have a delayed germination response to SA and MeJA
Germination is the growth of the embryonic seedling and marks the transition
between dormancy and growth (Bradford and Nonogaki 2007). The hormones SA and
MeJA were shown to have an inhibitory effect on this process in concentrations
>0.05mM, manifested by an increase in time after sowing to observe perforation of
the seed coat by the plants radical. To determine the effect that HD2C-GFP
overexpression has on this, Col-0, HD2C-GFP:LI and HD2C-GFP:L2 seeds were each
germinated on half strength MS agar plates supplemented with 0µM, 250µM, 500µM
and 750µM either SA or MeJA. This range of concentrations has been widely used and
shown to repress germination in Col-0 wild type plants (Colville, Alhattab et al. 2011).
Germination was analysed under a light microscope at 5x magnification and scored
based on the presence of the radical protruding from the seed coat. Germination
scores after 60 hours of incubation were accumulated for three separate experiments.
These scores were compared between each transgenic line and its corresponding wild
type for each media tested using a one tailed, independent t-test. This determined if
there was any statistically significant difference between germination rate for Col-0
and the 35S:HD2C-GFP plants on SA or MeJA media relative at each concentration
(figure 5.3). The results suggest that there was no difference between the germination
scores for any of the Arabidopsis lines on 0mM or 0.1mM SA or MeJA growth media
(p<0.05 for all four conditions in relation to MS). However on 0.25mM and 0.5mM for
both SA and MeJA media there was a statistically significant inhibition of germination
at the p<0.05 level when compared to germination rates of the same line on MS media
alone (figure 5.2A). At 0.25mM SA, germination was 95.66% for the Col-0 line
compared to 81.37% for HD2C-L1 and 67.53% for HD2C-L2. A further, more dramatic
reduction of germination was seen for the three lines on 0.5mM SA, germination was
21.13% for the Col-0 line compared to 6.16% for HD2C-L1 and 2% for HD2C-L2. The
effect of MeJA on seed germination was more obvious (figure 5.2B). Germination
scores for Col-0 were 86% for 0.25mM MeJA and 9.1% for 0.5mM MeJA, whereas
HD2C-L1 germinated at a frequency of 55.36% and 4.21% on the two media
respectively and HD2C-L2 germinated with frequencies of 47.21% and 0% respectively.
No germination was evident for any line after 60 hours when 0.75mM MeJA was
present in the media (data not shown). Together these results indicate that 35:HD2C-
113
GFP plants have reduced germination rates; i.e. show increased sensitivity in response
to SA and MeJA when compared to the wild type.
5.2.4 HD2C binds TGA6 transcription factor
The discovery that over-expression of HD2C in Arabidopsis resulted in SA and MeJA
with increased sensitive phenotypes in germination suggest that there is an underlying
molecular mechanism which may be common to the two hormone signalling
pathways. It was hypothesised that transcription factors may bind to HDACs and guide
them to their promoter cis-elements to negatively regulate gene expression(Luo,
Postigo et al. 1998; Doetzlhofer, Rotheneder et al. 1999). A bioinformatics approach
was used to Interrogation of the Predicted Arabidopsis Interactome Resource (PAIR)
database to identify putative HD2C binding proteins which may have a role in SA and
JA. PAIR predicts interactions by integrating indirect evidences for interaction (Lin, Shen et al.
2011). It identified the basic leucine zipper transcription factor TGA6 as a potentially
HD2C interacting protein. The confidence score for this prediction of ~1.3 was greater
than 1 and could therefore represent a true interaction (Lin, Zhou et al. 2011). TGA6 is
involved in the activation of genes related to plant defence and has been shown to
regulate SA and MeJA responsive genes in response to exogenous application of these
hormones (Zhou, Trifa et al. 2000; Zander, La Camera et al. 2010). Interaction of TGA6
with HD2C would rationalize the SA and MeJA response seen in the HD2C-GFP
transgenic plants. It was therefore an obvious candidate to test for in planta
interaction using BiFC.
TGA6 was cloned into the PG179NS-YN vector and transiently expressed with HD2C-YC
in N benthamiana leaf epidermal tissue as previously described (3.2.2). An N-terminal
HD2C fragment (AA1-60) corresponding to the enzymatic portion of HD2C was also
used to test whether the activity domain can interact with TGA6. Strong fluorescence
was detected in the nucleus and nucleolus of transformed cells when investigating
TGA6 interaction with the full length HD2C (figure 5.4) suggesting that interaction
between HD2C and TGA6 was possible. The N-terminal HD2C fragment together with
TGA6 had no significant fluorescence, suggesting that the observed YFP reconstitution
was driven by interaction between the two proteins independent of the activity
domain of HD2C. A Western blot was used to show that, in each case, proteins were
114
expressed so that the lack of fluorescence was not due to lacking or lower expression
levels (Appendix 5)
5.2.5 Expression of TGA6 overlaps with HD2C in some tissues and developmental
stages
The possibility of binding is only biologically relevant if the proteins are present in the
same location and at the same time. TGA6 is a transcription factor that binds NPR1 in
the nucleus, thus its localization overlaps with HD2C (Zhang 2011). To determine the
temporal and tissue specific expression of the gene, a database mining of micro-array
data was used to determine co-expression. As TGA6 and TGA2 and TGA5 are often
associated with the same response and as all three must be knocked out to produce
any phenotype, the expression of these genes were also investigated. Genevestigator
was used to compile the expression data of TGA2, TGA5 and TGA6 together with
HD2A-D in different Arabidopsis tissue (figure 5.5). The expression of HD2C and the
location of its highest expression were previously outlined in a similar analysis in
chapter 4 (4.2.2). TGA6 expression overlapped with HD2C most evidently in lateral root
primordial protoplasts, seedling culture and in seeds. In addition to the tissue specific
expression, developmental stages were also investigated (figure 5.6). Interestingly the
expression of HD2C was highest in senescing tissue, which correlated with highest
expression of TGA2, TGA5 and TGA6. Together these results suggest that the
interaction of TGA6 and HD2C may be most relevant in the seed and senescing tissue.
Considering the possible redundancy of interaction which could exist between other
HD2 isoforms and TGA isoforms, it was interesting to note that HD2A, HD2B, HD2D and
TGA6 shared expression in lateral root tissue (figure 5.5). Similarly, HD2A, HD2B, HD2D
and TGA2 shared expression in the hypocotyl. Together these results suggest that
TGA6 interaction with HD2C should be biologically relevant given the shared sub-
cellular localization pattern and the overlapping gene expression.
5.2.6 Analysis of the expression of genes controlled by TGA6 in HD2C modified
plants
A model for HDAC action suggests that the enzymatic deacetylation activity is directed
to specific chromatin regions by facilitating proteins that recognize specific DNA cis-
115
elements (Luo, Postigo et al. 1998; Doetzlhofer, Rotheneder et al. 1999). TGA6 is
known to bind to the as-1-like motif in the promoter in SA and JA related genes
(Niggeweg, Thurow et al. 2000) and to regulate expression of a number of SA and JA
regulated genes (Zhou, Trifa et al. 2000; Zander, La Camera et al. 2010). Based on the
validated interaction between HD2C and TGA6, it was next necessary to test the
hypothesis that HD2C regulates those same genes. A selection of genes related to the
plant defence response were chosen for an expression analysis using RT-PCR.
Arabidopsis lines Col-0, HD2C-L1 and HD2C L2 were grown for 2 weeks on 0.5MS plates
before RNA was extracted, converted to cDNA which was used to perform a RT-PCR.
Actin and ATPase were amplified in parallel as housekeeping gene controls. Results
indicated that there was no change in the intensity of PCR product in the actin and
ATPase housekeeping controls (figure 5.7) The expression of EDS5, ORA, SID2, EIGT,
UGT and GST6 were not affected by the overexpression of HD2C (figure 5.7). The most
obvious change in intensity indicating a difference in expression was observed for USP,
where expression was much higher in the two 35S:HD2C-GFP lines compared to Col-0.
There was also a visible increase in the expression of PR1 compared to Col-0, although
this was less obvious from the band intensities (figure 5.7). PDF1.2 and RFLK appeared
to be expressed in lower amounts in the 35S:HD2C-GFP lines compared to Col-0.
Together these results indicate that HD2C overexpression alters the expression of USP,
PR1, PDF1.2 and RFLK, but has no effect on EDS5, ORA, SID2, EIGT, UGT or GST6.
5.3 Discussion
5.3.1 Summary
Here, it was investigated what role HD2C may have in the salicylic acid and jasmonate
response pathway. Initially it was shown in Col-0 wild type plants that exogenous
application of salicylic acid was sufficient to induce expression of the salicylic response
gene PR1, but have no evident effect on HD2C expression. Similarly, exogenous
application of methyl-jasmonate was sufficient to induce expression of the jasmonate
response gene PDF1.2, but had no evident effect on HD2C expression. Next, two lines
over-expressing a HD2C-GFP construct in Arabidopsis plants were compared with Col-0
wild type plants to determine if there was any quantifiable phenotypic evidence that
116
HD2C had a role in either SA or JA response. A root length analysis on either media did
not produce any significant difference between the Arabidopsis lines. However a
germination analysis showed that over-expression of HD2C-GFP yielded a sensitive
response phenotype compared to wild type at 0.25mM, 0.5mM and 0.75mM SA and
0.25mM and 0.5mM MeJA. Next, interaction between HD2C and TGA6, a transcription
factor involved in both SA and JA signalling, was shown using BiFC. Micro-array data
suggested that there was some overlap in expression between these two proteins in
seed tissue and in the developmental progression of senescence. To test the
hypothesis that TGA6 directed HD2C to its as-1 like promoter cis-elements, RT-PCR was
used as a semi-quantitative measure of various genes that were shown to be regulated
by TGA6. A slight decrease in the JA responsive PDF1.2 expression, as well as in the SA
and JA-responsive RLK1 expression was observed, suggesting that HD2C may be
involved in regulating their expression.
5.3.2 Plants expressing 35S:HD2C-GFP have altered development in response to SA
and MeJA
It was unknown at the onset of this analysis if HD2C had a role in either the SA or JA
signalling pathways, therefore germination and root length data were obtained for
transgenic lines exposed to SA and MeJA, a biological more active form of the JA
hormone. Despite some overlapping mechanisms with respect to the defence
pathway, each hormone elicits a distinct molecular response and hence the extent of
the response was further evaluated using gene expression studies.
Germination is the process where an embryonic plant contained within its seed
transitions from dormancy to growth. This is mediated by sensitive signal transduction
pathways which induce growth in response to critical conditions such as water
availability, temperature, light and oxygen. Imbibition, where the embryonic tissue
swells sufficiently to break the seed coat, leads to the release of the embryonic root
(radicle) and is an effective marker of germination. The link between HD2 expression
and germination was previously explored in relation to salt and ABA tolerance, where
it was found that overexpression of HD2C leads to an increase in the germination
frequency compared to wild type on 150mM salt as well as in the presence of 15mM
ABA (Sridha and Wu 2006). This is consistent with the very high HD2 expression in
117
barley embryos which led to the suggestion that HD2C is involved in germination.
Germination is controlled principally through the balance between ABA and GA
hormones, whereby dormancy is maintained by ABA until activation of GA biosynthesis
and signalling occur. Aside from the well-defined balance between ABA and GA, other
plant hormones have been shown to impact on seed germination. For example, SA and
MeJA were shown to inhibit germination of Col-0 Arabidopsis seeds at concentrations
>0.1mM.
Given the established link between HD2C expression and germination, the role of SA
and MeJA in delaying germination and the possible role of HD2C in signalling, it was
not surprising to find that HD2C-GFP over-expressing seeds showed an altered
response when germinated on SA or MeJA containing media. Here, in contrast to the
ABA treatment where HD2C:GFP overexpression led to a resistance phenotype, MeJA
and SA treatment of 0.25mM, 0.5mM and 0.75mM resulted in sensitivity phenotype,
as germination was evidently inhibited when compared to the wild type. This
suggested that the germination pathway is affected in different ways by ABA and SA.
This is difficult to rationalize, as previous studies appear to suggest that SA uses an
identical pathway to suppress germination. In monocots, this occurs by suppressing
the GA activated alpha-amylase (Xie, Zhang et al. 2007). The effect on germination by
SA (250µM ) required more than 10 times the molar concentration compared to ABA
(20µM). In the literature, the higher sensitivity to ABA is commonly accounted for by
the fact that SA affects the mRNA accumulation of alpha-amylase through the action
of WRKY38 negative regulation of its gene expression whilst ABA in addition to
negative transcription regulation, it directly decreases amylase secretion and
enzymatic activity, thereby effectively blanketing the cue to initialize germination. In
dicots it appears that salicylic acid inhibition of root growth is mediated by the defence
response. This pathway transfers energy from growth and reallocates it towards
fighting off pathogen attack (Heil, Hilpert et al. 2000).
The question remains therefore as to why the germination data suggest that plants
over-expressing HD2C-GFP are more tolerant to ABA as shown by (Sridha and Wu
2006), yet less tolerant to SA and JA as demonstrated in the study here.
118
The answer may lie in the antagonistic relationship that exists between ABA and SA. It
was shown earlier that pathogenesis by P. syringae is marked by pathogen induced
stimulation of ABA accumulation, leading to a decrease in the SA-mediated defence
(de Torres, Bennett et al. 2009). Similarly exogenous application of small amounts of
SA was shown to null the effect of abiotic stresses such as salt which are controlled by
the ABA pathway (Borsani, Valpuesta et al. 2001). Moreover, the molecular basis of
ABA response and SA response appears consistent with this. 35S:HD2C-GFP plants
showed decreased expression of ABA response genes, therefore the antagonistic
crosstalk with GA is reduced, leading to an increase in expression. Results from the
analysis presented here showed that the SA responsive genes PR1 and EDS5 were
increased in 35S:HD2C-GFP plants. Therefore repression of germination by the
aforementioned inhibition of alpha-amylase is more likely following the alpha-amylase
pathway. This hypothesis clearly requires further investigation, as the data presented
offers only a tentative link between any cross talk of ABA and SA pathways.
5.3.3 HD2C binds TGA6 transcription factor
To rationalize the biological link of SA and MeJA with HD2C, an analysis of putative
binding partners revealed that HD2C interacted with TGA6 in planta. TGA6 belongs to
the TGA family of transcription factors, so-named because they bind to promoter cis-
elements designated as-1 like sequences which are composed of two tandem repeats
of TGACG (Idrovo Espín, Peraza-Echeverria et al. 2012). TGA6 and its redundant
homologues TGA2 and TGA5 were shown to be involved in both the SA and JA induced
plant defence response (Fan and Dong 2002). In the salicylic acid pathway TGA6 binds
NPR1 in the nucleus and activates SA-response genes such as PR1. In the jasmonic acid
pathway, TGA6 activates the jasmonate responsive PDF1.2 and b-chi genes (Zander, La
Camera et al. 2010). However it is also involved in antagonistic crosstalk between the
two pathways, with co-stimulation of SA and MeJA resulting in clade II TGAs binding
ROXY19/GRX480 and repressing the JA-responsive genes, thereby blocking the JA-
defence pathway (Zander, Chen et al. 2012; Gatz 2013).
Sensitive phenotypes in the 35S:HD2C-GFP plants when exposed to SA or MeJA suggest
that HD2C has a role in both of the pathways. Given the binding of a transcription
factor to a negative regulator of gene expression, the possibility that TGA6 was
119
mediating negative regulation of its target through interaction was explored. On the
surface, this hypothesis was not consistent with the data- TGA6 has been shown to
activate SA- and JA-responsive genes, how then did HD2C factor into this relationship?
Indeed, the RT-PCR data was not overwhelmingly supportive of this hypothesis. Of the
TGA targets measured, the most obvious difference in expression was a large induction
of expression for the VSP1 gene, although there was slight induction in expression of
PR1. This clearly suggested that HD2C was not catalytically active on these promoters.
This was surprising given that HD2C was shown to bind HDA19, and that HDA19 was
involved in repressing PR1 expression (Choi, Song et al. 2012). This suggests that HD2C
is involved in a different process than HDA19, and that binding probably mediates a
different process. Similarly there was no obvious change in the expression of ORA,
SID2, EIGT, UGT or GST6; this suggests that HD2C was similarly not active on these
promoters. Given the diverse range of target genes that were targeted by TGA6, this
suggested that HD2C was not binding TGA6 to negatively regulate these targets.
Rather than engaging in the regulation of SA- and JA- responsive genes singularly,
HD2C may be involved in the cross-talk between the two pathways. The mechanism
for this cross talk has not yet been established. Given the repressive role of HD2C on
gene expression, and this unknown negative regulation on JA-responsive genes when
challenged with SA, TGA6 may direct HD2C to JA-responsive genes when stimulated by
SA. Consistent with this, RT-PCR analysis of 35S:HD2C-GFP showed that PDF1.2 and
RLK1 expression was repressed compared to the Col-0 wild type. Although there was
no salicylic induction, this may demonstrate some element of SA-JA crosstalk. These
are JA-responsive factors, with PDF1.2 being the classic example of SA mediated JA-
responsive gene repression. However, this hypothesis is tempered by a study on the
repression of the core JA-responsive element PDF1.2, where it was revealed that its
repression by the SA pathway was not mediated directly by its chromatin acetylation
state (Koornneef, Rindermann et al. 2008). Assuming that HD2C is using TGA6 to
deacetylate the PDF1.2 promoter, this suggests that it is not involved in SA-JA
crosstalk, but is active on the promoter through interaction with another pathway.
A significant amount of effort for this study was directed towards establishing a link
between HD2C action and the plant defence response. Partly this was instigated by the
broad role of TGA19 salicylic and jasmonate plant defence (Zhou, Zhang et al. 2005;
120
Choi, Song et al. 2012). Additionally the already established link between HD2C and the
ABA mediated abiotic stress response provided a tentative link which suggested that
there may be a role linking HD2C with the biotic stress response (Sridha and Wu 2006).
Lastly, there is a significant current interest in linking epigenetic pathways with various
stress responses in an effort to improve plant tolerance to various harsh
environmental conditions. From the data however there is not sufficient evidence to
lead to this conclusion. Although HD2C appears to be involved in SA and JA responses,
the molecular mechanism does not appear to involve binding AS-1 like cis-factors via
TGA interaction. This may partly be due to a confounding effect caused by HD2C
interaction with other factors involved in these pathways. Principally, HD2C interaction
with ERF1 was described in relation to Longan. Assuming that this is conserved in
Arabidopsis, it is likely that this interaction would also produce the SA and MeJA
phenotypes that were observed. Future studies into this interaction may instead look
to the expression data presented in figures 5.5 and 5.6. Interestingly, HD2C and TGA6
overlap most strongly in senescing tissue and seeds. This suggests that interaction is
most likely relevant during these developmental stages.
5.3.4 Conclusions and future work
The most significant aspect of this study was the identification of HD2C interaction
with TGA6, as it provides further evidence that there is significant cross-talk between
general transcriptional regulation elements and the epigenetic pathway. As previously
mentioned, this interaction is evolutionarily practical given the specificity of
transcription factors to significant subsets of genes and the passive but stable ability
for acetylation state to control gene expression. Given the inability to trace any
overwhelming evidence that HD2C is targeting as-1 like cis binding sites to act on,
further characterization into its role is required. The possibility of a role in senescence
was suggested from micro-array data. Both SA and JA have been shown to mediate this
developmental stage.
121
122
123
124
125
126
127
128
TGA6-YC/ HD2C-YN
TGA6-YC/ Δ HD2C-
YN
YFP CHLOROPHYLL DAPI MERGE A
B
Mouse anti-cmyc
Mouse anti-HA
Figure 5.8: BiFC analysis to determine HD2C interaction with TGA6 IN N benthamiana leaf tissue (A) Interaction of HD2C with TGA6 was evident by fluorescence caused by reconstitution of the split YFP fused to each HD2 protein. The YFP emission of each dimer overlaps with the DAPI stained nucleus in a discrete nucleolar compartment. No signal could be detected in the YFP channel for interaction between TGA6 and the 60AA N-terminal region of HD2C. A negative control consisting of the N-terminal (1-60 amino acids) of HD2C was tested for interaction. It yielded no fluorescence. (B) Expression was measured using western blot against the c-myc epitope from the YN vector and HA epitope from the YC vector. Expression was present in both tissues. Scale bar represents 20μm.
129
Chapter 6
Final Discussion
130
6.1 Summary
In this thesis the original aims were to characterize putative interactions to develop
insight into possible regulatory pathways that had not previously been identified.
Initially, localization of HD2C was analysed to trace the accumulation of HD2C in the
cell. HD2C was present in the nucleus and nucleolus when fused to GFP. Next, HD2C
dimers were tested to determine their fluorescent localization pattern. This revealed a
predominantly nucleolar pattern, distinct from HD2C-GP. it appears that this
localization may be dynamic, as stably transformed 35S:HD2C-GFP plants showed
increased accumulation of nucleolar fluorescence after 24 hours of salt stress. The
mechanism of this accumulation remains unknown; however a hypothesis is that
dimerization of HD2s appears to be involved based on the similar localization between
HD2C dimers and the pattern observed in salt treatment. Next two domains of HD2C
were shown to be necessary for nuclear and nucleolar localization. The nuclear
localization was shown to be dependent on an evolutionarily conserved KKAK motif,
while this did not disrupt nucleolar localization.
The interaction of 14-3-3 and HD2C was also investigated using Bimolecular
Fluorescent Complementation. It was shown that HD2C bound to both epsilon and
non-epsilon classes of 14-3-3 proteins, while the other HD2 isoforms were able to bind
to 14-3-3 epsilon. To identify the region required for this interaction, a deletion
analysis of the HD2C protein mapped out the interaction of HD2C with 14-3-3 epsilon.
Three distinct regions were found which enable interaction. Site directed mutagenesis
of possible serine and threonine residues that could correspond to the critical
phosphorylated binding site on the C-terminal region of the protein revealed that
residue S284 and one of T235 and S239, when substituted to alanine, abolished 14-3-3
binding
Lastly, interaction was investigated between TGA6 and HD2C. A link between HD2C
and SA and JA mediated plant defence was tried to be established. However,
exogenous application of salicylic acid and jasmonic acid had no evident effect on
HD2C expression. In addition two lines over-expressing a HD2C-GFP construct in
Arabidopsis plants were compared with Col-0 wild type plants to determine if there
was any phenotypes that suggest that HD2C had a role in either SA or JA response.
There was no statistically significant change in root growth on either media, however a
131
sensitivity phenotype was evident for both hormones in a germination assay. Finally,
HD2C was tested for its ability to regulate as-1 like promoter cis-elements related to
stress response. A slight decrease in the JA responsive PDF1.2 expression, as well as in
the SA and JA-responsive RLK1 expression was observed, suggesting that HD2C may be
involved in regulating their expression.
6. 2 Results suggest new model of HD2 action
Prior to this thesis a HD2 model describing its functionality had not been explicitly
stated, but could be derived by amalgamating several key studies from the past 20
years (figure 6.1A). HD2C is expressed in the cytoplasm as a protein containing an N-
terminal HDAC domain, central acidic domain and C-terminal zinc finger domain. The
functional protein is part of a large, 400kDa protein complex, phosphorylated at
multiple sites and capable of removing acetyl-groups from histone H3 which was
related to down-regulating gene expression. Accumulation in the nucleus and
nucleolus was driven by an unknown mechanism. The enzymatic activity was targeted
to various genes relating to ABA response, rRNA expression and development by an
unknown mechanism. It has been speculated that molecular interactions with different
factors of the transcriptional machinery may mediate interactions of HD2C with
specific histone areas. However only the ethylene response transcription factor ERF7
was experimentally shown to bind a Longan HD2 homologue.
The work of this thesis has expanded this model by focusing on the missing elements
of the regulatory pathway. As shown in figure 6.1B, the nuclear and nucleolar
accumulation were shown to be independent signalling events, where nuclear
accumulation was mediated by a bipartite NLS and the nucleolar accumulation was
driven by a further C-terminal region. Furthermore the nucleolar targeting appeared to
be stress responsive, as nucleolar accumulation was increased in response to salt
stress. It appears that formation of the complex may have a role here, as HD2C-HD2
dimers were shown to accumulate in the nucleolus compared to the nuclear and
nucleolar pattern of the HD2C-GFP. The role of 14-3-3 binding to the C-terminal end of
the protein was shown not be a critical determinant of nuclear or nucleolar
localization, as mutation of these sites resulted in the same localization pattern as the
non-modified version of the enzyme. Lastly, the role of HD2C towards SA and JA
132
signalling was investigated. Results indicated that HD2C expression was linked to SA
and MeJA sensitivity in the context of root growth and germination with
overexpression causing greater sensitivity to these enzymes during germination.
Binding of TGA6 was initially hypothesised to be a critical determinant of this
phenotype; however semi-quantitative analysis of SA and JA-responsive factors that
were shown to be controlled by clade II TGA factors did not show any significant
repression despite heightened HD2C presence in the cell.
Together these results contribute new knowledge to HD2C regulation. Firstly, this
revealed and characterized differences in HD2C localization. Specifically, the difference
between HD2C:HD2 dimers and stress induced HD2C-GFP nucleolar fluorescent
patterns are distinct from the patterns observed in non-stressed HD2C-GFP, HD2C:14-
3-3 and HD2C:TGA6 interaction studies. Together this suggests that HD2C localization
is dynamic and can be manipulated by external conditions. Secondly, characterization
of three different classes of proteins interacting with HD2C extends our current
knowledge of the diversity of the HD2Cs interactome.
6.3 HD2C for use in genetically modified plants
A significant motivation for plant molecular biology research is the potential for its
insights to be harnessed towards manipulating crop plants to yield more food for less
human input. The potential for histone deacetylases to be modified to alter plant
growth and cropping has been documented. For example it was shown that the
acetylation state of the genome is dynamic, with acetylation of specific areas related
to growth and development shifting from acetylated to deacetylated in response to
various stages of plant growth (Kouzarides 1999; Tian, Fong et al. 2005). In Arabidopsis
non-specific blocking of histone deacetylation activity resulted in pleiotropic effects on
growth and development which could be traced to its inability to operate as an
acetylation antagonist (Tian and Chen 2001). Similarly knocking out various individual
HDAC genes results in a number of growth and developmental defects. From these
studies it is clear that acetylation per se is central to growth and development, which
necessitates understanding of the deacetylation component of this system. Precisely
how the HD2 family of histone deacetylases fits into this and how manipulation of their
pathways may lead to a genetically superior crop has not yet been established.
133
Modified expression of HD2s by a constitutive promoter or by gene knockout does not
appear to be the answer. Plants over-expressing HD2C are resistant to salt and drought
stress during germination, survival rate and root growth assays. However the plants
inherent pay-back for this phenotype was evident in a higher proportion of sterility.
Similarly, it was reported that Arabidopsis plants over-expressing HD2A developed
pleiotropic growth defects manifested by altered leaf and flower morphology. Plants
with HD2 knockouts were similarly deficient; HD2C knockouts were correspondingly
sensitive to salt and drought stress and germinated poorly on sucrose media. Together
these results suggest that a simple over- or under-expression of HD2 is not suitable as
a crop modification tool. The results of this thesis provide a starting point for a non-
expression driven alteration in HD2C which has the potential to impact on plant
growth, development and stress response.
This thesis focused on identifying aspects of the enzyme’s regulation that may be
modified to produce a more pathway specific response. For example, this study
showed that it was the nucleolar retention rather than the nuclear translocation which
was the most relevant mechanism in localization control. This was evident both from
the apparent nucleolar targeting which took place following salt stress and the
nucleolar localization of the functional HD2 complex. The region required for nucleolar
retention was narrowed to a C-terminal portion of the protein which did not wholly
overlap with the nuclear targeting sequence. Further analysis of this region yielded the
identification of a 14-3-3 binding site; however mutation of this site did not yield any
evident shift in localization pattern when the mutated HD2C protein was tested as a
GFP-fusion. Together these data suggest that the regulation of nucleolar localization is
a possible site for manipulation. The next step is likely the identification of factors
which mediate this regulation so that the HD2 sequence dependent on these
interactions could be altered to only respond to specific factors. This would achieve an
imposed specificity for nucleolar localization that occurs only in response to
predetermined conditions. This is obviously an ambitious hypothesis given the number
of unknown regulatory factors and the potential diversity of the signal transduction
event. The impact of chemicals on targeting specific interactions in the context of a
whole cell were seen most exquisitely in the context of 14-3-3 binding to H+ATPase in
tomato by fusicoccin. Here fusicoccin locked together the subunits of the H+ATPase.
134
This was shown to cause hyperpolarization of the cell membrane and lead to an
induced pathogen response.
An alternative approach is focussing on the interaction with transcription factors which
mediate contact between HD2 and DNA. This thesis determined that there was
interaction between TGA6 and HD2C. Although it appears that this is not significant in
the context of mediating regulation of SA or JA responsive factors, the developmental
implications of this interaction have yet to be determined. Whatever the biological
consequence of this interaction, it is now clear that interaction with transcription
factors is possible, given the interactions of HD2C with ERF1 and TGA6. The next step is
to determine the HD2C binding site that mediates this contact. Given this information,
modification of the HD2 sequence would enable a direct manipulation on the pathway
that involves the transcription factor. Given the inherent ability for HD2s to regulate
large subsets of genes via its chromatin condensation ability, this would enable a large
scale manipulation on genetic interactions with external stimuli which could be
harnessed to modify plant growth and response to the environment.
6.4 C-terminal domain appears to be necessary for protein binding
Throughout this thesis there were several lines of evidence to suggest that the C-
terminal domain of HD2C was significant in the context of protein-protein interaction.
Most overt was the direct mapping of a 14-3-3 epsilon binding sites, where two
different sites were mapped to the C-terminal end of HD2C. Similarly, the N-terminal
1-60AA fragment of HD2C was used both in dimerization experiments of HD2s and in
the interaction study with TGA6 and failed to interact with either of the two. This
suggests that interaction with other proteins does not reuiqre the N-termianl part of
HD2C. A spin-off of this discovery is the application of the N-terminal fragment as a
negative control for protein interaction experiments. . Furthermore, the tiling deletion
analysis that was performed on HD2C to map 14-3-3 binding sites was performed for
TGA6, mapping a region that it required for binding to a C-terminal site overlapping
the NLS (study not completed, data not shown). The significance of this is two-fold:
Firstly it hints at a fundamental domain structure within the protein, where the N-
terminus contains a conserved enzymatic domain and functional diversity is
135
manifested by the C-terminal end where various subsets of interactors bind. In The
RPD3-like superfamily in Arabidopsis contains a similar domain structure. There, the N-
terminal HDAC site is highly conserved and diversity is most evident in the C-terminal
end.
Secondly, it suggests that there is a co-operation between the two domains. It suggests
that the C-terminal domain is necessary for protein interaction which guides the HDAC
initially into the nucleus and nucleolus, and subsequently to its chromatin targets. Next
the N-terminal end performs its HDAC function. The fact that this relationship is both
synergistic and essential is clear from each study in this thesis. Firstly, the NLS and
nucleolar retention signal are located in the C-terminus which is essential for
colocalization of HD2 with its substrate. Secondly, specific chromatin regions are the
focus of the HDAC activity, which is probably provided via interaction with TGA and
ERF transcription factors and possibly others. Lastly, the enzyme requires
phosphorylation to form an active complex. Initially it was hypothesised that this
suggested N-terminal phosphorylation sites which would stabilize the active domain
and control enzyme activity. However a phosphorylation site was found in the N-
terminus by phosphoproteomics and two 14-3-3 sites which bind phosphorylation sites
were similarly found in this region (Durek, Schmidt et al. 2010). Therefore it is possible
that enzyme regulation is dependent upon secondary mechanisms. Although the
results do not provide clear insight, it may be speculated that phosphorylation
regulates formation of the active HD2 complex or mediates contact with substrate.
Moreover, the presence of the zinc-finger domain in HD2A and HD2C isoforms is a
feature of the C-terminus which remains uncharacterized. It was suggested that the
single zinc finger of HD2C cannot interact with DNA, instead it would provide a domain
for protein or RNA interactions. However, hetero- and Homo- dimers of HD2s raise the
possibility that multiple Zinc-fingers from HD2A and HD2C isoforms make direct DNA
contact possible in such complexes. . Its overlap with a 14-3-3 binding site similarly
suggests that its role may be dynamically regulated by phosphorylation. Whatever its
role, its evolutionary conservation suggest that this domain is a critical feature of this
C-terminus whose function is yet to be determined.
136
6.5 Future
HD2s principle role is to modify gene expression. So far this has been traced by testing
the expression of genes of interest in plants with over- or under-expressing HD2C
genes. For example this approach was used to confirm the impact of HD2C expression
on various ABA responsive genes to rationalize the observation that plants with
modified HD2C expression had different responses to salt, drought and ABA. These
small scope studies shed light on specific pathways that relate to HD2C expression and
function. The risk of this is that the holistic role of HD2C function is diluted by specific
research areas where time is more heavily invested. An alternative approach is the so-
called ‘-omic’ analyses which allow a large scale interpretation of protein activity by
offering a holistic monitoring of responses.
HD2C is an enzyme that affects transcription of genes, and does this by modifying the
chromatin structure of the genome. Two ‘-omic’ approaches are immediately relevent
here.
Micro-array measures mRNA amounts for large subsections of the coded genome (Ball,
Sherlock et al. 2002). In a case relevant to this thesis, it was used to identify novel
functions of a clade II TGA isoform in response to salicylic acid (Thibaud‐Nissen, Wu et
al. 2006). It is anticipated that a similar approach could be used in plants with modified
HD2C expression to identify novel pathways related to HD2C function.
Secondly, various chromatin immune-precipitation (ChIP) techniques such as chip
hybridization (ChIP-chip) (Buck and Lieb 2004; Park 2009) or next generation
sequencing (ChIP-seq) have been developed to screen for modified acetyl-chromatin
states in response to development, stress or mutation. The advantage of this over
micro-array analysis is that it allows the direct visualization of HD2C activity. Thus it
would provide direct evidence relating to the binding targets of HD2C enzyme and
enable an insight to be made into the core function of HD2C.
137
Chapter 7 References
138
Abràmoff, M. D., P. J. Magalhães, et al. (2004). "Image processing with ImageJ." Biophotonics international 11(7): 36-42.
Aitken, A. (2006). 14-3-3 proteins: a historic overview. Seminars in cancer biology, Elsevier. Aravind, L. and E. V. Koonin (1998). "Second Family of Histone Deacetylases." Science
280(5367): 1167. Atchley, W. R. and B. K. Hall (1991). "A model for development and evolution of complex
morphological structures." Biological Reviews 66(2): 101-157. Aufsatz, W., M. F. Mette, et al. (2002). "HDA6, a putative histone deacetylase needed to
enhance DNA methylation induced by double-stranded RNA." The EMBO journal 21(24): 6832.
Bader, J. S., A. Chaudhuri, et al. (2003). "Gaining confidence in high-throughput protein interaction networks." Nature biotechnology 22(1): 78-85.
Bae, M. S., E. J. Cho, et al. (2003). "Analysis of the Arabidopsis nuclear proteome and its response to cold stress." The Plant Journal 36(5): 652-663.
Ball, C. A., G. Sherlock, et al. (2002). "Standards for microarray data." Science (New York, NY) 298(5593): 539.
Berger, S. L. (2007). "The complex language of chromatin regulation during transcription." Nature 447(7143): 407-412.
Berger, S. L., T. Kouzarides, et al. (2009). "An operational definition of epigenetics." Genes & Development 23(7): 781-783.
Berman, M. and T. DeJong (1996). "Water stress and crop load effects on fruit fresh and dry weights in peach (Prunus persica)." Tree Physiology 16(10): 859-864.
Bogdanović, O. and G. J. C. Veenstra (2009). "DNA methylation and methyl-CpG binding proteins: developmental requirements and function." Chromosoma 118(5): 549-565.
Borsani, O., V. Valpuesta, et al. (2001). "Evidence for a role of salicylic acid in the oxidative damage generated by NaCl and osmotic stress in Arabidopsis seedlings." Plant Physiology 126(3): 1024-1030.
Boulon, S., B. J. Westman, et al. (2010). "The nucleolus under stress." Molecular Cell 40(2): 216-227.
Bradford, K. J. and H. Nonogaki (2007). Seed development, dormancy and germination, Blackwell publishing.
Brooks, A. R., R. N. Harkins, et al. (2004). "Transcriptional silencing is associated with extensive methylation of the CMV promoter following adenoviral gene delivery to muscle." The journal of gene medicine 6(4): 395-404.
Brosch, G., A. Lusser, et al. (1996). "Purification and characterization of a high molecular weight histone deacetylase complex (HD2) of maize embryos." Biochemistry 35(49): 15907-15914.
Buck, M. J. and J. D. Lieb (2004). "ChIP-chip: considerations for the design, analysis, and application of genome-wide chromatin immunoprecipitation experiments." Genomics 83(3): 349-360.
Cameron, R. K., N. L. Paiva, et al. (1999). "Accumulation of salicylic acid and PR-1 gene transcripts in relation to the systemic acquired resistance (SAR) response induced by< i> Pseudomonas syringae</i> pv.< i> tomato</i> in< i> Arabidopsis</i>." Physiological and molecular plant pathology 55(2): 121-130.
Camoni, L., J. F. Harper, et al. (1998). "14-3-3 proteins activate a plant calcium-dependent protein kinase (CDPK)." FEBS letters 430(3): 381.
Carmo-Fonseca, M. (2002). "The contribution of nuclear compartmentalization to gene regulation." Cell 108(4): 513-522.
Chen, L.-T., M. Luo, et al. (2010). "Involvement of Arabidopsis histone deacetylase HDA6 in ABA and salt stress response." Journal of experimental botany 61(12): 3345-3353.
Chen, W. and K. B. Singh (1999). "The auxin, hydrogen peroxide and salicylic acid induced expression of the Arabidopsis GST6 promoter is mediated in part by an ocs element." The Plant Journal 19(6): 667-677.
139
Chini, A., S. Fonseca, et al. (2007). "The JAZ family of repressors is the missing link in jasmonate signalling." Nature 448(7154): 666-671.
Chinnusamy, V., Z. Gong, et al. (2008). "Abscisic Acid‐mediated Epigenetic Processes in Plant Development and Stress Responses." Journal of Integrative Plant Biology 50(10): 1187-1195.
Choi, S. M., H. R. Song, et al. (2012). "HDA19 is required for the repression of salicylic acid biosynthesis and salicylic acid‐mediated defense responses in Arabidopsis." The Plant Journal 71(1): 135-146.
Cokol, M., R. Nair, et al. (2000). "Finding nuclear localization signals." EMBO reports 1(5): 411-415.
Collura, V. and G. Boissy (2007). From Protein—Protein Complexes to Interactomics. Subcellular Proteomics, Springer: 135-183.
Colville, A., R. Alhattab, et al. (2011). "Role of HD2 genes in seed germination and early seedling growth in <i>Arabidopsis</i>." Plant Cell Reports 30(10): 1969-1979.
Costa, F. F. (2008). "Non-coding RNAs, epigenetics and complexity." Gene 410(1): 9-17. Covello, P. S. and M. W. Gray (1989). "RNA editing in plant mitochondria." Creelman, R. A. and J. E. Mullet (1995). "Jasmonic acid distribution and action in plants:
regulation during development and response to biotic and abiotic stress." Proceedings of the National Academy of Sciences 92(10): 4114-4119.
Dangl, M., G. Brosch, et al. (2001). "Comparative analysis of HD2 type histone deacetylases in higher plants." Planta 213(2): 280-285.
Dathan, N., L. Zaccaro, et al. (2002). "The Arabidopsis SUPERMAN protein is able to specifically bind DNA through its single Cys2–His2 zinc finger motif." Nucleic Acids Research 30(22): 4945-4951.
Davies, P. J. (2010). Plant hormones: biosynthesis, signal transduction, action!, Springer. Davies, P. J. (2010). The plant hormones: their nature, occurrence, and functions. Plant
hormones, Springer: 1-15. de Ruijter, A. J., A. H. Van Gennip, et al. (2003). "Histone deacetylases (HDACs):
characterization of the classical HDAC family." Biochemical Journal 370(Pt 3): 737. de Torres, Z. M., M. Bennett, et al. (2009). "Antagonism between salicylic and abscisic acid
reflects early host-pathogen conflict and moulds plant defence responses." The Plant journal: for cell and molecular biology 59(3): 375.
DeLille, J. M., P. C. Sehnke, et al. (2001). "The Arabidopsis 14-3-3 family of signaling regulators." Plant Physiology 126(1): 35-38.
Demetriou, K., A. Kapazoglou, et al. (2009). "Epigenetic chromatin modifiers in barley: I. Cloning, mapping and expression analysis of the plant specific HD2 family of histone deacetylases from barley, during seed development and after hormonal treatment." Physiologia Plantarum 136(3): 358-368.
DeRisi, J. L., V. R. Iyer, et al. (1997). "Exploring the metabolic and genetic control of gene expression on a genomic scale." Science 278(5338): 680-686.
Després, C., C. DeLong, et al. (2000). "The Arabidopsis NPR1/NIM1 protein enhances the DNA binding activity of a subgroup of the TGA family of bZIP transcription factors." The Plant Cell Online 12(2): 279-290.
Dez, C., M. Dlakić, et al. (2007). "Roles of the HEAT repeat proteins Utp10 and Utp20 in 40S ribosome maturation." Rna 13(9): 1516-1527.
Díaz-Moreno, I., D. Hollingworth, et al. (2009). "Phosphorylation-mediated unfolding of a KH domain regulates KSRP localization via 14-3-3 binding." Nature Structural & Molecular Biology 16(3): 238-246.
Doetzlhofer, A., H. Rotheneder, et al. (1999). "Histone deacetylase 1 can repress transcription by binding to Sp1." Molecular and cellular biology 19(8): 5504-5511.
Dong, X. (2004). "NPR1, all things considered." Current opinion in plant biology 7(5): 547-552.
140
Durek, P., R. Schmidt, et al. (2010). "PhosPhAt: the Arabidopsis thaliana phosphorylation site database. An update." Nucleic Acids Research 38(suppl 1): D828-D834.
Eberharter, A. and P. B. Becker (2002). "Histone acetylation: a switch between repressive and permissive chromatin." EMBO reports 3(3): 224-229.
Eden, A., F. Gaudet, et al. (2003). "Chromosomal instability and tumors promoted by DNA hypomethylation." Science 300(5618): 455-455.
ERARD, M. S., P. Belenguer, et al. (1988). "A major nucleolar protein, nucleolin, induces chromatin decondensation by binding to histone Hl." European Journal of Biochemistry 175(3): 525-530.
Eulgem, T. (2005). "Regulation of the< i> Arabidopsis</i> defense transcriptome." Trends in Plant Science 10(2): 71-78.
Eulgem, T. and I. E. Somssich (2007). "Networks of WRKY transcription factors in defense signaling." Current opinion in plant biology 10(4): 366-371.
Fahrenkrog, B. and U. Aebi (2003). "The nuclear pore complex: nucleocytoplasmic transport and beyond." Nature Reviews Molecular Cell Biology 4(10): 757-766.
Fan, W. and X. Dong (2002). "In vivo interaction between NPR1 and transcription factor TGA2 leads to salicylic acid–mediated gene activation in Arabidopsis." The Plant Cell Online 14(6): 1377-1389.
Fernández-Calvo, P., A. Chini, et al. (2011). "The Arabidopsis bHLH transcription factors MYC3 and MYC4 are targets of JAZ repressors and act additively with MYC2 in the activation of jasmonate responses." The Plant Cell Online 23(2): 701-715.
Fischle, W., F. Dequiedt, et al. (2002). "Enzymatic activity associated with class II HDACs is dependent on a multiprotein complex containing HDAC3 and SMRT/N-CoR." Molecular Cell 9(1): 45-57.
Fu, H., R. R. Subramanian, et al. (2000). "14-3-3 proteins: structure, function, and regulation." Science Signaling 40(1): 617.
Fuglsang, A. T., S. Visconti, et al. (1999). "Binding of 14-3-3 protein to the plasma membrane H+-ATPase AHA2 involves the three C-terminal residues Tyr946-Thr-Val and requires phosphorylation of Thr947." Journal of Biological Chemistry 274(51): 36774-36780.
Galis, I. and K. Matsuoka (2007). Transcriptomic analysis of salicylic acid-responsive genes in tobacco BY-2 cells. Salicylic Acid: A Plant Hormone, Springer: 371-396.
Gardino, A. K., S. J. Smerdon, et al. (2006). Structural determinants of 14-3-3 binding specificities and regulation of subcellular localization of 14-3-3-ligand complexes: a comparison of the X-ray crystal structures of all human 14-3-3 isoforms. Seminars in cancer biology, Elsevier.
Gatz, C. (2013). "From Pioneers to Team Players: TGA Transcription Factors Provide a Molecular Link Between Different Stress Pathways." Molecular plant-microbe interactions 26(2): 151-159.
Gill, S. S. and N. Tuteja (2010). "Reactive oxygen species and antioxidant machinery in abiotic stress tolerance in crop plants." Plant Physiology and Biochemistry 48(12): 909-930.
Ginisty, H., H. Sicard, et al. (1999). "Structure and functions of nucleolin." Journal of Cell Science 112(6): 761-772.
Glazebrook, J. (2005). "Contrasting mechanisms of defense against biotrophic and necrotrophic pathogens." Annu. Rev. Phytopathol. 43: 205-227.
Goldberg, A. D., C. D. Allis, et al. (2007). "Epigenetics: a landscape takes shape." Cell 128(4): 635.
Görlich, D., S. Kostka, et al. (1995). "Two different subunits of importin cooperate to recognize nuclear localization signals and bind them to the nuclear envelope." Current Biology 5(4): 383-392.
Graves, P. R., C. M. Lovly, et al. (2001). "Localization of human Cdc25C is regulated both by nuclear export and 14-3-3 protein binding." Oncogene 20(15): 1839.
Grebenok, R. J., E. Pierson, et al. (1997). "Green‐fluorescent protein fusions for efficient characterization of nuclear targeting." The Plant Journal 11(3): 573-586.
141
Grozinger, C. M., C. A. Hassig, et al. (1999). "Three proteins define a class of human histone deacetylases related to yeast Hda1p." Proceedings of the National Academy of Sciences 96(9): 4868-4873.
Grozinger, C. M. and S. L. Schreiber (2000). "Regulation of histone deacetylase 4 and 5 and transcriptional activity by 14-3-3-dependent cellular localization." Science Signaling 97(14): 7835.
Grunstein, M. (1997). "Histone acetylation in chromatin structure and transcription." Nature 389(6649): 349-352.
Guo, M., A. Yang, et al. (2012). "The new understanding of Arabidopsis thaliana proteins associated with salinity." Journal of Plant Interactions 7(4): 348-355.
Hannon, G. J. (2002). "RNA interference." Nature 418(6894): 244-251. Harootunian, A., S. Adams, et al. (1993). "Movement of the free catalytic subunit of cAMP-
dependent protein kinase into and out of the nucleus can be explained by diffusion." Molecular biology of the cell 4(10): 993.
Heil, M., A. Hilpert, et al. (2000). "Reduced growth and seed set following chemical induction of pathogen defence: does systemic acquired resistance (SAR) incur allocation costs?" Journal of Ecology 88(4): 645-654.
Hinshaw, J. E., B. O. Carragher, et al. (1992). "Architecture and design of the nuclear pore complex." Cell 69(7): 1133-1141.
Hirayama, T. and K. Shinozaki (2007). "Perception and transduction of abscisic acid signals: keys to the function of the versatile plant hormone ABA." Trends in Plant Science 12(8): 343-351.
Hirota, K., T. Miyoshi, et al. (2008). "Stepwise chromatin remodelling by a cascade of transcription initiation of non-coding RNAs." Nature 456(7218): 130-134.
Hollender, C. and Z. Liu (2008). "Histone deacetylase genes in Arabidopsis development." Journal of Integrative Plant Biology 50(7): 875-885.
Holliday, R. (1987). "The inheritance of epigenetic defects." Science 238(4824): 163-170. Holliday, R. (1990). "Mechanisms for the control of gene activity during development."
Biological Reviews 65(4): 431-471. Holliday, R. (2006). "Epigenetics: a historical overview." Epigenetics 1(2): 76-80. Idrovo Espín, F. M., S. Peraza-Echeverria, et al. (2012). "< i> In silico</i> cloning and
characterization of the TGA (TGACG MOTIF-BINDING FACTOR) transcription factors subfamily in< i> Carica papaya</i>." Plant Physiology and Biochemistry 54: 113-122.
Imamura, T., S. Yamamoto, et al. (2004). "Non-coding RNA directed DNA demethylation of< i> Sphk1</i> CpG island." Biochemical and Biophysical Research Communications 322(2): 593-600.
Ito, M., A. Koike, et al. (2003). "Methylated DNA-binding proteins from Arabidopsis." Plant Physiology 133(4): 1747-1754.
Iuchi, S. (2001). "Three classes of C2H2 zinc finger proteins." Cellular and Molecular Life Sciences CMLS 58(4): 625-635.
Jaenisch, R. and A. Bird (2003). "Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals." Nature genetics 33: 245-254.
Jans, D. A., C. Y. Xiao, et al. (2000). "Nuclear targeting signal recognition: a key control point in nuclear transport?" Bioessays 22(6): 532-544.
Johnson, C., S. Crowther, et al. (2010). "Bioinformatic and experimental survey of 14-3-3-binding sites." Biochemical Journal 427(Pt 1): 69.
Jönsson, H., M. Heisler, et al. (2005). "Modeling the organization of the WUSCHEL expression domain in the shoot apical meristem." Bioinformatics 21(suppl 1): i232-i240.
Kadonaga, J. T. (1998). "Eukaryotic transcription: an interlaced review network of transcription factors and chromatin-modifying machines." Cell 92: 307-313.
Kende, H. and J. Zeevaart (1997). "The Five" Classical" Plant Hormones." The Plant Cell 9(7): 1197.
142
Kim, C. A. and J. M. Berg (1996). "A 2.2 Å resolution crystal structure of a designed zinc finger protein bound to DNA." Nature Structural & Molecular Biology 3(11): 940-945.
Kim, J.-M., T. K. To, et al. (2008). "Alterations of lysine modifications on the histone H3 N-tail under drought stress conditions in Arabidopsis thaliana." Plant and cell physiology 49(10): 1580-1588.
Klug, A. and D. Rhodes (1987). "‘Zinc fingers’: a novel protein motif for nucleic acid recognition." Trends in Biochemical Sciences 12: 464-469.
Koltunow, A. M., J. Truettner, et al. (1990). "Different temporal and spatial gene expression patterns occur during anther development." The Plant Cell 2(12): 1201.
Koornneef, A., K. Rindermann, et al. (2008). "Histone modifications do not play a major role in salicylate-mediated suppression of jasmonate-induced PDF1. 2 gene expression." Communicative & integrative biology 1(2): 143-145.
Koroleva, O., G. Calder, et al. (2009). "Dynamic behavior of Arabidopsis eIF4A-III, putative core protein of exon junction complex: fast relocation to nucleolus and splicing speckles under hypoxia." The Plant Cell Online 21(5): 1592-1606.
Kouzarides, T. (1999). "Histone acetylases and deacetylases in cell proliferation." Current opinion in genetics & development 9(1): 40-48.
Kouzarides, T. (2007). "Chromatin modifications and their function." Cell 128(4): 693-705. Kuang, J.-f., J.-y. Chen, et al. (2012). "Histone deacetylase HD2 interacts with ERF1 and is
involved in longan fruit senescence." Journal of experimental botany 63(1): 441-454. Lagace, M., S. C. Chantha, et al. (2003). "Fertilization induces strong accumulation of a histone
deacetylase (HD2) and of other chromatin-remodeling proteins in restricted areas of the ovules." Plant Molecular Biology 53(6): 759-769.
Laity, J. H., B. M. Lee, et al. (2001). "Zinc finger proteins: new insights into structural and functional diversity." Current opinion in structural biology 11(1): 39-46.
Lange, A., R. E. Mills, et al. (2007). "Classical nuclear localization signals: definition, function, and interaction with importin α." Journal of Biological Chemistry 282(8): 5101-5105.
Larsen, F., G. Gundersen, et al. (1992). "CpG islands as gene markers in the human genome." Genomics 13(4): 1095-1107.
Lawrence, R. J., K. Earley, et al. (2004). "A Concerted DNA Methylation/Histone Methylation Switch Regulates rRNA Gene Dosage Control and Nucleolar Dominance." Molecular Cell 13(4): 599-609.
Lawton, K., K. Weymann, et al. (1995). "Systemic acquired resistance in Arabidopsis requires salicylic acid but not ethylene." MPMI-Molecular Plant Microbe Interactions 8(6): 863-870.
Lee, B. J., A. E. Cansizoglu, et al. (2006). "Rules for nuclear localization sequence recognition by karyopherinβ2." Cell 126(3): 543-558.
Lee, M. S., G. P. Gippert, et al. (1989). "Three-dimensional solution structure of a single zinc finger DNA-binding domain." Science 245(4918): 635-637.
Li, J., G. Brader, et al. (2004). "The WRKY70 transcription factor: a node of convergence for jasmonate-mediated and salicylate-mediated signals in plant defense." Science Signaling 16(2): 319.
Lin, M., X. Shen, et al. (2011). "PAIR: the predicted Arabidopsis interactome resource." Nucleic Acids Research 39(suppl 1): D1134-D1140.
Lin, M., X. Zhou, et al. (2011). "The predicted Arabidopsis interactome resource and network topology-based systems biology analyses." The Plant Cell Online 23(3): 911-922.
Liu, D., J. Bienkowska, et al. (1995). "Crystal structure of the zeta isoform of the 14-3-3 protein."
Loidl, P. (2004). "A plant dialect of the histone language." Trends in Plant Science 9(2): 84-90. Luger, K. and T. J. Richmond (1998). "The histone tails of the nucleosome." Current opinion in
genetics & development 8(2): 140-146. Luo, M., Y.-Y. Wang, et al. (2012). "HD2C interacts with HDA6 and is involved in ABA and salt
stress response in Arabidopsis." Journal of experimental botany 63(8): 3297-3306.
143
Luo, M., Y.-Y. Wang, et al. (2012). "HD2 proteins interact with RPD3-type histone deacetylases." Plant Signaling & Behavior 7(6): 608-610.
Luo, R. X., A. A. Postigo, et al. (1998). "Rb interacts with histone deacetylase to repress transcription." Cell 92(4): 463-473.
Lusser, A., G. Brosch, et al. (1997). "Identification of maize histone deacetylase HD2 as an acidic nucleolar phosphoprotein." Science 277(5322): 88-91.
Malamy, J. and D. F. Klessig (1992). "Salicylic acid and plant disease resistance." The Plant Journal 2(5): 643-654.
Mancini, D. N., S. M. Singh, et al. (1999). "Site-specific DNA methylation in the neurofibromatosis (NF1) promoter interferes with binding of CREB and SP1 transcription factors." Oncogene 18(28): 4108.
Manners, J. M., I. A. Penninckx, et al. (1998). "The promoter of the plant defensin gene PDF1. 2 from Arabidopsis is systemically activated by fungal pathogens and responds to methyl jasmonate but not to salicylic acid." Plant Molecular Biology 38(6): 1071-1080.
Margueron, R. and D. Reinberg (2010). "Chromatin structure and the inheritance of epigenetic information." Nature Reviews Genetics 11(4): 285-296.
Margueron, R., P. Trojer, et al. (2005). "The key to development: interpreting the histone code?" Current opinion in genetics & development 15(2): 163-176.
Marras, E. and E. Capobianco (2008). "A Multiscale Tour in Protein Interactomics." Martin, W. (2010). "Evolutionary origins of metabolic compartmentalization in eukaryotes."
Philosophical Transactions of the Royal Society B: Biological Sciences 365(1541): 847-855.
Mattick, J. S. (2001). "Non-coding RNAs: the architects of eukaryotic complexity." EMBO reports 2(11): 986-991.
McGonigle, B., K. Bouhidel, et al. (1996). "Nuclear localization of the Arabidopsis APETALA3 and PISTILLATA homeotic gene products depends on their simultaneous expression." Genes & Development 10(14): 1812-1821.
Métraux, J., H. Signer, et al. (1990). "Increase in salicylic acid at the onset of systemic acquired resistance in cucumber." Science 250(4983): 1004-1006.
Mitsuda, N. and M. Ohme-Takagi (2009). "Functional analysis of transcription factors in Arabidopsis." Plant and cell physiology 50(7): 1232-1248.
Moore, M. S. (1998). "Ran and nuclear transport." Journal of Biological Chemistry 273(36): 22857-22860.
Moore, T. C. (1979). Biochemistry and physiology of plant hormones, Springer-Verlag. Moorhead, G., P. Douglas, et al. (1999). "Plant proteins containing phosphopeptide motifs that
bind to 14-3-3 proteins." Plant J 18: 1-12. Morris, J. (2001). "Genes, genetics, and epigenetics: a correspondence." Science 293(5532):
1103-1105. Mou, Z., W. Fan, et al. (2003). "Inducers of plant systemic acquired resistance regulate NPR1
function through redox changes." Cell 113(7): 935-944. Muslin, A. J., J. W. Tanner, et al. (1996). "Interaction of 14-3-3 with signaling proteins is
mediated by the recognition of phosphoserine." Cell 84(6): 889-897. Muslin, A. J. and H. Xing (2000). "14-3-3 proteins: regulation of subcellular localization by
molecular interference." Cellular signalling 12(11): 703-709. Ndamukong, I., A. A. Abdallat, et al. (2007). "SA‐inducible Arabidopsis glutaredoxin interacts
with TGA factors and suppresses JA‐responsive PDF1. 2 transcription." The Plant Journal 50(1): 128-139.
Niggeweg, R., C. Thurow, et al. (2000). "Tobacco transcription factor TGA2. 2 is the main component of as-1-binding factor ASF-1 and is involved in salicylic acid-and auxin-inducible expression of as-1-containing target promoters." Journal of Biological Chemistry 275(26): 19897-19905.
Noll, M. (1974). "Subunit structure of chromatin." Nature 251(5472): 249-251.
144
Obsilova, V., J. Vecer, et al. (2005). "14-3-3 Protein interacts with nuclear localization sequence of forkhead transcription factor FoxO4." Biochemistry 44(34): 11608-11617.
Pabo, C. O. and R. T. Sauer (1992). "Transcription factors: structural families and principles of DNA recognition." Annual review of biochemistry 61(1): 1053-1095.
Panni, S., L. Montecchi‐Palazzi, et al. (2011). "Combining peptide recognition specificity and context information for the prediction of the 14‐3‐3‐mediated interactome in S. cerevisiae and H. sapiens." Proteomics 11(1): 128-143.
Park, P. J. (2009). "ChIP–seq: advantages and challenges of a maturing technology." Nature Reviews Genetics 10(10): 669-680.
Paul, A.-L., L. Liu, et al. (2009). "Comparative interactomics: analysis of Arabidopsis 14-3-3 complexes reveals highly conserved 14-3-3 interactions between humans and plants." Journal of proteome research 8(4): 1913-1924.
Pedone, P. V., R. Ghirlando, et al. (1996). "The single Cys2-His2 zinc finger domain of the GAGA protein flanked by basic residues is sufficient for high-affinity specific DNA binding." Proceedings of the National Academy of Sciences 93(7): 2822-2826.
Penfield, S., Y. Li, et al. (2006). "Arabidopsis ABA INSENSITIVE4 regulates lipid mobilization in the embryo and reveals repression of seed germination by the endosperm." The Plant Cell Online 18(8): 1887-1899.
Peng, S., J. Huang, et al. (2004). "Rice yields decline with higher night temperature from global warming." Proceedings of the National Academy of Sciences of the United States of America 101(27): 9971-9975.
Pérez-Núñez, M., R. Souza, et al. (2009). "Detection of a SERK-like gene in coconut and analysis of its expression during the formation of embryogenic callus and somatic embryos." Plant Cell Reports 28(1): 11-19.
Perez‐Terzic, C., M. Jaconi, et al. (1997). "Nuclear calcium and the regulation of the nuclear pore complex." Bioessays 19(9): 787-792.
Pieterse, C. M. and L. C. van Loon (1999). "Salicylic acid-independent plant defence pathways." Trends in Plant Science 4(2): 52-58.
Pollard, V. W., W. M. Michael, et al. (1996). "A novel receptor-mediated nuclear protein import pathway." Cell 86(6): 985.
Purwestri, Y. A., Y. Ogaki, et al. (2009). "The 14-3-3 protein GF14c acts as a negative regulator of flowering in rice by interacting with the florigen Hd3a." Plant and cell physiology 50(3): 429-438.
Razin, A. and A. D. Riggs (1980). "DNA methylation and gene function." Science 210(4470): 604-610.
Robert-Seilaniantz, A., M. Grant, et al. (2011). "Hormone crosstalk in plant disease and defense: more than just jasmonate-salicylate antagonism." Annual review of phytopathology 49: 317-343.
Sasai, N., M. Nakao, et al. (2010). "Sequence-specific recognition of methylated DNA by human zinc-finger proteins." Nucleic Acids Research 38(15): 5015-5022.
Saslowsky, D. E., U. Warek, et al. (2005). "Nuclear localization of flavonoid enzymes in Arabidopsis." Journal of Biological Chemistry 280(25): 23735-23740.
Saxonov, S., P. Berg, et al. (2006). "A genome-wide analysis of CpG dinucleotides in the human genome distinguishes two distinct classes of promoters." Proceedings of the National Academy of Sciences of the United States of America 103(5): 1412-1417.
Schenk, P. M., K. Kazan, et al. (2000). "Coordinated plant defense responses in Arabidopsis revealed by microarray analysis." Proceedings of the National Academy of Sciences 97(21): 11655-11660.
Schijlen, E. G., C. Ric de Vos, et al. (2004). "Modification of flavonoid biosynthesis in crop plants." Phytochemistry 65(19): 2631-2648.
Shaffer, K. L., A. Sharma, et al. (2005). "Regulation of protein compartmentalization expands the diversity of protein function." Developmental cell 9(4): 545-554.
145
Shah, J. (2003). "The salicylic acid loop in plant defense." Current opinion in plant biology 6(4): 365-371.
Shah, J., P. Kachroo, et al. (1999). "The Arabidopsis ssi1 mutation restores pathogenesis-related gene expression in npr1 plants and renders defensin gene expression salicylic acid dependent." The Plant Cell Online 11(2): 191-206.
Shah, J. and D. Klessig (1999). "Salicylic acid: signal perception and transduction." New Comprehensive Biochemistry 33: 513-541.
Shaw, P. (2013). The Plant Nucleolus. Plant Genome Diversity Volume 2, Springer: 65-76. Sittka, A., S. Lucchini, et al. (2008). "Deep sequencing analysis of small noncoding RNA and
mRNA targets of the global post-transcriptional regulator, Hfq." PLoS genetics 4(8): e1000163.
Song, Y., K. Wu, et al. (2010). "< i> Arabidopsis</i> DNA methyltransferase AtDNMT2 associates with histone deacetylase AtHD2s activity." Biochemical and Biophysical Research Communications 396(2): 187-192.
Song, Y. A., K. Q. Wu, et al. (2010). "Arabidopsis DNA methyltransferase AtDNMT2 associates with histone deacetylase AtHD2s activity." Biochemical and Biophysical Research Communications 396(2): 187-192.
Spoel, S. H., A. Koornneef, et al. (2003). "NPR1 modulates cross-talk between salicylate-and jasmonate-dependent defense pathways through a novel function in the cytosol." The Plant Cell Online 15(3): 760-770.
Sridha, S. and K. Q. Wu (2006). "Identification of AtHD2C as a novel regulator of abscisic acid responses in Arabidopsis." Plant Journal 46(1): 124-133.
Staswick, P. E., I. Tiryaki, et al. (2002). "Jasmonate response locus JAR1 and several related Arabidopsis genes encode enzymes of the firefly luciferase superfamily that show activity on jasmonic, salicylic, and indole-3-acetic acids in an assay for adenylation." The Plant Cell Online 14(6): 1405-1415.
Swatek, K. N., K. Graham, et al. (2011). "The 14-3-3 isoforms chi and epsilon differentially bind client proteins from developing Arabidopsis seed." Journal of proteome research 10(9): 4076-4087.
Tada, Y., S. H. Spoel, et al. (2008). "Plant immunity requires conformational charges of NPR1 via S-nitrosylation and thioredoxins." Science 321(5891): 952-956.
Tajrishi, M. M., R. Tuteja, et al. (2011). "Nucleolin: The most abundant multifunctional phosphoprotein of nucleolus." Communicative & integrative biology 4(3): 267-275.
Thaler, J. S. (1999). "Jasmonate-inducible plant defences cause increased parasitism of herbivores." Nature 399(6737): 686-688.
Theunissen, O., F. Rudt, et al. (1992). "RNA and DNA binding zinc fingers in Xenopus TFIIIA." Cell 71(4): 679-690.
Thibaud‐Nissen, F., H. Wu, et al. (2006). "Development of Arabidopsis whole‐genome microarrays and their application to the discovery of binding sites for the TGA2 transcription factor in salicylic acid‐treated plants." The Plant Journal 47(1): 152-162.
Thines, B., L. Katsir, et al. (2007). "JAZ repressor proteins are targets of the SCFCOI1 complex during jasmonate signalling." Nature 448(7154): 661-665.
Thomma, B. P., K. Eggermont, et al. (1998). "Separate jasmonate-dependent and salicylate-dependent defense-response pathways in Arabidopsis are essential for resistance to distinct microbial pathogens." Proceedings of the National Academy of Sciences 95(25): 15107-15111.
Thoms, H. C., C. J. Loveridge, et al. (2010). "Nucleolar targeting of RelA (p65) is regulated by COMMD1-dependent ubiquitination." Cancer research 70(1): 139-149.
Tian, L. and Z. J. Chen (2001). "Blocking histone deacetylation in Arabidopsis induces pleiotropic effects on plant gene regulation and development." Proceedings of the National Academy of Sciences 98(1): 200-205.
146
Tian, L., M. P. Fong, et al. (2005). "Reversible histone acetylation and deacetylation mediate genome-wide, promoter-dependent and locus-specific changes in gene expression during plant development." Genetics 169(1): 337-345.
Toroser, D., G. S. Athwal, et al. (1998). "Site-specific regulatory interaction between spinach leaf sucrose-phosphate synthase and 14-3-3 proteins." FEBS letters 435(1): 110-114.
Traw, M. B. and J. Bergelson (2003). "Interactive effects of jasmonic acid, salicylic acid, and gibberellin on induction of trichomes in Arabidopsis." Plant Physiology 133(3): 1367-1375.
Tseng, T.-S., C. Whippo, et al. (2012). "The role of a 14-3-3 protein in stomatal opening mediated by PHOT2 in Arabidopsis." The Plant Cell Online 24(3): 1114-1126.
Uquillas, C., I. Letelier, et al. (2004). "NPR1-independent activation of immediate early salicylic acid-responsive genes in Arabidopsis." Molecular plant-microbe interactions 17(1): 34-42.
Verdel, A., S. Curtet, et al. (2000). "Active maintenance of mHDA2/mHDAC6 histone-deacetylase in the cytoplasm." Current Biology 10(12): 747-749.
Wang, A. H., M. J. Kruhlak, et al. (2000). "Regulation of histone deacetylase 4 by binding of 14-3-3 proteins." Molecular and cellular biology 20(18): 6904-6912.
Wang, R. and M. G. Brattain (2007). "The maximal size of protein to diffuse through the nuclear pore is larger than 60kDa." FEBS letters 581(17): 3164-3170.
Watt, F. and P. L. Molloy (1988). "Cytosine methylation prevents binding to DNA of a HeLa cell transcription factor required for optimal expression of the adenovirus major late promoter." Genes & Development 2(9): 1136-1143.
Weiner, H. and W. M. Kaiser (1999). "14-3-3 proteins control proteolysis of nitrate reductase in spinach leaves." FEBS letters 455(1): 75-78.
Wu, K. Q., L. N. Tian, et al. (2000). "Functional analysis of HD2 histone deacetylase homologues in Arabidopsis thaliana." Plant Journal 22(1): 19-27.
Wu, K. Q., L. N. Tian, et al. (2003). "Repression of gene expression by Arabidopsis HD2 histone deacetylases." Plant Journal 34(2): 241-247.
Xie, Z., Z.-L. Zhang, et al. (2007). "Salicylic acid inhibits gibberellin-induced alpha-amylase expression and seed germination via a pathway involving an abscisic-acid-inducible WRKY gene." Plant Molecular Biology 64(3): 293-303.
Xiong, L., K. S. Schumaker, et al. (2002). "Cell signaling during cold, drought, and salt stress." The Plant Cell Online 14(suppl 1): S165-S183.
Xu, Y., P.-F. L. Chang, et al. (1994). "Plant defense genes are synergistically induced by ethylene and methyl jasmonate." The Plant Cell Online 6(8): 1077-1085.
Yaffe, M. B. (2002). "How do 14-3-3 proteins work?–Gatekeeper phosphorylation and the molecular anvil hypothesis." FEBS letters 513(1): 53-57.
Zander, M., S. Chen, et al. (2012). "Repression of the Arabidopsis thaliana jasmonic acid/ethylene-induced defense pathway by TGA-interacting glutaredoxins depends on their C-terminal ALWL motif." Molecular plant 5(4): 831-840.
Zander, M., S. La Camera, et al. (2010). "Arabidopsis thaliana class‐II TGA transcription factors are essential activators of jasmonic acid/ethylene‐induced defense responses." The Plant Journal 61(2): 200-210.
Zeng, L. and M.-M. Zhou (2002). "Bromodomain: an acetyl-lysine binding domain." FEBS letters 513(1): 124.
Zhang, D. (2011). "The structure of arabidopsis NPR1: its function as a salicylic acid receptor and a metal-binding protein."
Zhang, X. (2008). "The epigenetic landscape of plants." Science 320(5875): 489-492. Zhang, X., M. I. Francis, et al. (2010). "Over-expression of the Arabidopsis NPR1 gene in citrus
increases resistance to citrus canker." European journal of plant pathology 128(1): 91-100.
147
Zhang, Y., W. Fan, et al. (1999). "Interaction of NPR1 with basic leucine zipper protein transcription factors that bind sequences required for salicylic acid induction of the PR-1 gene." Proceedings of the National Academy of Sciences 96(11): 6523-6528.
Zhang, Y., M. J. Tessaro, et al. (2003). "Knockout analysis of Arabidopsis transcription factors TGA2, TGA5, and TGA6 reveals their redundant and essential roles in systemic acquired resistance." The Plant Cell Online 15(11): 2647-2653.
Zhou, C., H. Labbe, et al. (2004). "Expression and function of HD2‐type histone deacetylases in Arabidopsis development." The Plant Journal 38(5): 715-724.
Zhou, C., L. Zhang, et al. (2005). "HISTONE DEACETYLASE19 is involved in jasmonic acid and ethylene signaling of pathogen response in Arabidopsis." The Plant Cell Online 17(4): 1196-1204.
Zhou, C. H., H. Labbe, et al. (2004). "Expression and function of HD2-type histone deacetylases in Arabidopsis development." Plant Journal 38(5): 715-724.
Zhou, J.-M., Y. Trifa, et al. (2000). "NPR1 differentially interacts with members of the TGA/OBF family of transcription factors that bind an element of the PR-1 gene required for induction by salicylic acid." Molecular plant-microbe interactions 13(2): 191-202.
148
Chapter 8 Appendices
149
150
151
152
153
top related