Download physically interact with

CHAPTER 5
ZBF2jGBFt and HY5
physically interact with
each other
79
5.1. Introduction
bZIP proteins are phylogenetically widely distributed transcriptional
regulators, characterised by a basic DNA-binding domain (b) (Landschulz et al,
1988). Although DNA binding of bZIP monomers has been described, they
normally interact with DNA as dimers (Ellenberger et al, 1992; Metallo and
Schepartz, 1997; Cranz et al, 2004). Dimerisation is mediated by the so called
leucine zipper domain, a heptad repeat of leucine or other bulky hydrophobic
amino acids creating an amphipathic helix (Landschulz et al, 1988; Baxevanis
and Vinson, 1993). The formation of bZIP homo- or heterodimers offers a huge
combinatorial
flexibility
heterodimerisation,
to
DNA-binding
regulatory
specificity
transcription
systems.
and
transactivation
affinity,
By
properties and ultimately, cell physiology might be altered (Naar et al, 2001).
For instance, in the animal system, heterodimers of the bZIP proteins Jun and
Fos have been shown to recognise the AP1 motif, which is not targeted by Jun
homodimers
(Kouzarides
and
Ziff,
1988).
As
bZIP
proteins
do
not
heterodimerise promiscuously but specifically (Newman and Keating, 2003), an
important function for gene regulation is anticipated.
In planta, bZIP
heterodimerisation has been shown for closely related bZIPs, for example, the
GBFs (Schindler et al, 1992). In tobacco (Strathmann et al, 2001) and parsley
(Ru··gner et al, 2001), specific heterodimerisation was found between members
of two groups of bZIPs that are related to the Arabidopsis groups S and C.
Interestingly, heterodimerisation between these groups is strongly preferred in
comparison to homotypic dimerisation.
In Arabidopsis ZBF2/GBF1 differentially regulate the photomorphogenic
growth and light regulated gene expression (Mallappa et al., 2006). It has also
been shown that GBF1 protein is less abundant in the dark grown seedlings
and is degraded by a proteasome-mediated pathway independent of COP1 and
SPAl. Furthermore, COP1 physically interacts with GBF1 and is required for
the optimum accumulation of GBF1 protein in light-grown seedlings (Mallappa
et al., 2008). The subcellular localization studies indicate that the lightcontrolled nuclear translocation is one of the important mechanisms for the
activities of GBFs (Terzaghi et al., 1997). gbfl mutants were found to be highly
80
sensitive to ABA-mediated inhibition of seed germination, as compared to wild
type and is epistatic to both cop 1 and spal for ABA sensitivity (Mallapa, PhD
thesis, 2007; JNU, New Delhi). Recently it has been reported that GBF1
reduces CATALASE 2 expression to regulate the onset of leaf senescence in
Arabidopsis thaliana (Smykowski et al., 2010). GBF1, GBF2 and GBF3
heterodimerize and these heterodimers also interact with the G-box, suggesting
a potential mechanism for generating additional diversity from these GBF
proteins (Schindler et al., 1992b). A GBF interacting protein GIP1 has been
reported to enhance the DNA binding activity and reduce the oligomeric state of
GBFs (Sehnke et al., 2005). It has been already deduced that DNA binding
activity of the Arabidopsis G-Box Binding factor GBFl is stimulated by
phosphorylation by casein kinase II (Klimczak et al., 1992; Klimczak et al.,
1995).
HY5
has
been
genetically
defined
as
a
positive
regulator
of
photomorphogenesis based on the light insensitivity of hyS mutants. The
phenotype of hyS seedlings includes defects in light inhibition of hypocotyl
elongation,
light induced chlorophyll accumulation,
and extensive root
abnormalities (Ang and Deng, 1994). hyS mutation enhances cell elongation in
hypocotyl and root hairs, and alters gravitropic and touching response in roots.
This mutation enhances the initiation and elongation of lateral roots. hyS
mutation reduces the secondary thickening of the root and hypocotyl (Oyama
et.al., 1997). HYSlocus (AT5G11260) encodes a protein with a b-ZIP motif. The
longest open reading frame in the HYS eDNA clones encoded a protein of
18.5kD composed of 168 amino acid residues. It contains five heptad repeats of
leucine, and a casein kinase2 phosphorylation site in a highly conserved
region.
COP1, an E3 ligase, interacts with N-terminal of HY5 and with the help
of SPA1 targets HY5 for proteasome-mediated degradation in dark (Ang et al.,
1998; Osterlund et al., 2000b; Saijo et al., 2003). In light, unphosphorylated
HY5 shows stronger interaction with COP1 and has higher affinity to target
promoters and is physiologically more active than phosphorylated version.
Moreover, unphosphorylated HY5 is the preferred substrate for COPl (Hardtke
81
et al., 2000). HYS also integrates both light and hormone signaling pathways as
it promotes expression of IAA7 and IAA14, negative regulators of auxin
signaling (Cluis et al., 2004). Cytokinins and GA are shown to regulate levels of
HYS protein accumulation even in dark (Vandenbussche et al., 2007; Alabadi
et al., 2008). One recent report shows that HYS binds to the promoter of the
transcription factor ABIS gene and is required for the expression of ABIS and
ABIS-targeted LEA genes, thus mediating ABA response in seed germination
(Chen and Xiong, 2008). Many other proteins have been found to interact with
HYS to modulate its activity as CCA 1, a circadian clock associated protein has
been reported to alter the binding activity of HYS to the G-box element present
in LHCB promoters (Andronis et al., 2008). B-box proteins, STH2 and STH3
also interact with HYS (Datta et al., 2008a; Datta et al., 2008b). Recently, it has
been reported that HYS with PIF3 regulates anthocyanin biosynthesis by
simultaneously binding anthocyanin biosynthetic gene promoters at separate
sequence elements (Shin et al., 2007).
HYS acts as heterodimer with its homologue HYH and mediates lightregulated expression of overlapping as well as distinct target genes (Holm et al.,
2002). HYS and HYH were found to be the activators of NIA2, but inhibitors of
NRTI.l
when tested across various light treatments and tissue types
downstream to phytochromes (Jonassen et al., 2008; Jonassen et al., 2009).
Experiments with hyS and hyh mutants reveal that both these factors mediate
responses of the UVR8-dependent pathway, acting with partial or complete
redundancy to stimulate expression of particular genes (Brown and Jenkins,
2008; Brown et al., 2009). Besides these, HYS and HYH, both are required for
the up-regulation of PEXllb, which is involved in peroxysome development in
response to light (Desai and Hu, 2008). Interestingly, overexpression of HYH
rescues lateral root phenotype of hyS mutant, but when both loci have
homozygous null alleles, seedlings show less developed root system suggesting
that HYS and HYH are important negative regulators of auxin signaling
amplitude in embryogenesis and seedling development (Sibout et al., 2006).
However, for most light-modulated genes, the sequence of steps linking
bZIP transcripton factors to changes in gene expression is still largely
82
unknown.
Whether other transcriptional regulator(s)
is/are involved in
amplifying the bZIP-mediated light-signaling pathway remains to be addressed.
In this study, we have investigated the physical interactions of ZBF2/GBF1
with HYS protein in Arabidopsis seedling development.
83
GST
GST-GBF1
GST-HY5
RBCS1A
(A)
GST-GBF1
GST-HY5
anti-GBF1
anti-HY5
(B)
- + - - + - +
+ +
+ + + + +
1 2 3 4 5
+ + · -+++
- ·+++++
-+ - - - + -----+-·+
1 2 3 4 5 6 7 8 9
Figure 7. GBFl and HYS heterodimerize and bind_ together to DNA
in vitro.
(A) Electrophoretic mobility shift assay (EMSA) using recombinant GBFl
and HY5 proteins, binding to RBCSlA minimal promoter.
(B)
Electrophoretic
mobility
shift
assay
(EMSA)
showing further
retardation of DNA-protein complex upon incubation with specific
antibody.
Plus and minus signs in Figure 7 A and B show the presence and
absence of the component in respective lanes.
5.2. Results
5.2.1. GBFl and HYS hetero dimerize and bind to DNA in vitro
HY5 has previously been shown to bind the G-box m otif present in
RBCS-IA promoter and positively regulate it (Osterlund and Deng, 1998). GBF1
has been reported to bind to the G-box of RBCS-IA or Z-box of CABI promoter
(Mallappa et al., 2006). Whereas HY5 positively regulates RBCS- IA promoter,
GBF1/ZBF2 negatively controls the expression of RBCS-IA gene. Considering
the antagonistic regulation of these two bZIP proteins upon the expression of
RBCSIA gene and their comparable affinity to RBCS-IA promoter, we wanted to
investigate whether these two b-ZIP transcription factors heterodimerize and
bind to the G-box element present in RBCS-IA promoter.
To examine the possible heterodimerization of GBF1 and HY5 and
binding to the DNA, we performed electrophoretic mobility shift assays using
GST fusion proteins and 198 base pair minimal promoter fragment of RBCS-IA
promoter (Chattopadhyay et al. , 1998b). Pre-existing homodimeric complexes
were dissociated by incubation of GST-GBF1, GST-HY5 and an equimolar
mixture of both at 50°C for 5 minutes prior to addition to the DNA. GBF1 and
HY5 form discrete protein-DNA complexes those are readily resolved as seen in
lane 3 and 4 (Figure 7 A). When GBF1 and HY5 were mixed together, a complex
with intermediary mobility was detected in lane 5 (Figure 7A). Because the DNA
probe contains only one copy of the G-box motif, our result suggests the
formation of GBF 1/HY5 heterodimers. Furthermore upon mixing the GBF1 and
HY5 proteins, the GBF1 /HY5 heterodimeric complex constitutes the clear
majority, suggesting that the GBF1 /HY5 -DNA complex is preferentially formed
in vitro.
To substantiate the above result, we designed electrophoretic mobility
shift assays using antibodies specific to both the proteins. Upon incubation of
newly formed protein-DNA complex with the antibodies specific to protein,
complexes with fur ther reduced mobility were found on the gel (Figure 7B). In
lanes 4 and 6 (Figure 7B), while incubated with the respective antibodies, DNAprotein-antibody complexes formed by GBF 1 and HY5 were found to move with
84
(A)
PELLET
SUPERNATANT
HY5
(B)
~--------------------~~4--GBFl
..-ACTIN
Figure 8. HYS interacts with GBFl in vitro as well as in vivo.
(A) Anti HYS blot showing in vitro binding of GBFl and HYS. Signal in
supernatant is serving as loading control.
(B) Anti GBFl blot showing binding of recombinant HY5-6His p rotein
with GBFl present in vivo .
Anti ACTIN immunoblot is serving as loading control.
Arrowheads show the bands respective to protein detected in
immunoblot
further decreased mobility comparative to only protein-DNA complexes in lane
3 and 5 respectively. In lane 8 and lane 9 , GBFl/HYS heterodimer complex
was found moving with lower mobility when incubated with either antibody,
comparative to the lane 7, where no antibody was present. Surprisingly, in lane
8 where the complex was incubated with a -GBFl, no homodimeric complex of
HYS/DNA with h igher mobility was observed. Similarly, in lane 9, no
homodimeric complex of GBFl/DNA with higher mobility was observed when
reaction was incubated with a-HYS. This supports our previous observation
and strengthens the preferred formatio n of GBFl/HYS heterodimer complex
with DNA.
5.2.2. GBFl and HYS interact physically
The b-ZIP proteins are likely to bind as dimers to their palind romic DNA
targets. The spatial disposition resulting from leucine zipper dimerization leads
to the two opposed protein monomers to interact symmetrically with each half
site of the DNA palindrome (Siberil et al., 2001) . To further subs tantiate the
results obtained from the DNA-protein interaction studies , we examined the
physical interaction of GBFl and HYS in in vitro pull down assays. We used
poly-His and GST fusion proteins for in vitro pull down assays. HYS-6His
protein was pass ed through a glutathione sepharose 4B beads bound to GST,
GST-GBFl and GST-COPl fusion proteins. a -HYS immunoblot of the pulled
down proteins s h owed that GST-GBFl in teracted with HY5-6His comparable to
GST-COPl (used as a positive control) (Figure 8A). The amount of HYS-6His
retained by GST-GBFl was significantly higher than the background level of
HY5-6His retained by only GST. These results indicate that GBFl interacts
physically with HYS.
We then carried out similar experiments using the total plant protein
extracts . In this experiment we bound HYS-6His recombinant fu s ion protein to
the Ni-NTA aga rose beads and incub a t ed with crude extracts from different
lines i.e. WT-WS, GBF10E1, gbf1-1 , gbf1-2 and gbf1-1 hy5. As shown Figure
8B, when crude extract from WT-WS and GBFlOEl lines were p assed through
HYS-6HIS/Ni- NTA column, GBFl protein was retained in the column (Figure
8B). The Western blot analyses using a GBFl did not show any b and in lanes 3
85
(A)
0.7
0.6
f 0.5
i 0.4
'f'
'!>: 0.3
'il
'iii
'i 0.2
a::
0.1
0
(C)
(B)
...__GBF1
*
~HYS
*
Figure 9. GBFl interacts with HYS in a yeast two-hybrid assay.
(A) Relative 15-galactosidase activities, represented by graph.The value is
avarage of three individual yeast colonies, and error bars represent
standard deviations.
(B)-(C) Immunoblots showing expression of GBFl (B), and HYS (C),fused
with GAL4 activation and binding domain, respectively, in yeast cells.
Cross reacting band is serving as loading control. Arrowheads show the
band respective to fusion proteins in immunoblot.
to 5 where the crude extracts from gbfl-1 , gbfl-2, and gbfl hyS passed through
the HY5-6His/Ni-NTA column, excluding the possibility of any non specific
binding to the HY5-6His/Ni-NTA column. Taken together these results suggest
that GBFl and HYS proteins physically interact with each other.
5.2.3. GBFl and HY5 interact ex vivo
To further substantiate the physical interactions between GBFl and
HYS, we carried out yeast two-hybrid assays.
We inserted GBFl and HYS
eDNA into pGADT7 and pGBKT7 yeast expression vectors and co-expressed
these constructs in yeast strain Y187. Interaction of GBFl a n d HYS was
assayed
by
liquid
galactopyranoside as
0-galactosidase
substrate.
assays
using
The AD-GBFl
chlorophenol
red-0-D-
protein, expressed from
pGADT7-GBF1 construct did not activate transcription together with the Gal4
DNA binding domain (GAL4-DBD) vector control. However, AD -GBFl protein
resulted in 3 to 4-fold increase in 0-galactosidase activity over the vector
control when expressed together with GAL4-DBD-HY5 (Figure 9A). Figure 9B
and C show the Western blot analyses of the expression of GBFl and HYS,
fused to GAL4 activation and binding domain, respectively.
5.2.4. GBFl co-localizes with HY5 in plant cells
HYS gives a diffused nuclear fluorescence when expressed in onion
epidermal cells (Ang et al., 1998) . In order to determine sub cellular localization
of interaction event, we prepared a Cyan Fluorescent Protein fusion of GBFl
and Yellow Fluorescent Protein fusion of HYS followed by co-expression in
onion epidermal cells. We found that GBFl , like HYS localizes uniformly
throughout the nucleus (Figure 3B) . Since both the proteins interact in vitro
and both the proteins give a diffused nuclear fluorescence , we superimposed
HYS-YFP and GBFl-CFP images to see the overlapping of pixels and found an
obvious color change in fluorescence (Figure lOB). Superimposition of only CFP
and YFP images did not show any color change in fluorescence at the same
calibration s (Figure lOA) . These results strongly suggest that GBFl and HYS
are co-localized in the nucleus of onion cells.
86
(A)
(B)
(C)
Figure 10. GBFl colocalizes with HYS in nucleus of onion epidermal
cells.
(A)-(B), An onion epidermal cell co transformed with pCHA-CFP and
pCHA-YFP vector controls (A), pCHACFP-GBFl and pCHA-YFP-HYS
constructs (B).
(C)-(D), An onion epidermal cell co transformed with pCAMBIA1302,
pUC-SPYNE and pUC-SPYCE vector-effector control (C), pCAMBIA 1302,
pUC-SPYNE-GBFl and pUC-SPYCE-HYS constructs (D).
To further confirm this result, we used Bimolecular Fluorescence
Complementation assay using pUC-SPYNE and pUC-SPYCE (Walter et al.,
2004). Here, we constructed C-terminal fusions of GBFl and HYS. GBFl FL
eDNA was fused toN terminal of YFP and HYS FL eDNA to C-terminal of YFP.
These constructs were co-bombarded with the appropriate effector DNA
(pCAMBIA1302T) onto the onion epidermal cells. Co-expression of GBFl-YFPN
and HYS-YFPC induced strong fluorescence in the nucleus of bombarded onion
epidermal cells (Figure 1OD), whereas control pairs gave no fluorescence
(Figure
lOC).
A
cytoplasmic
green
fluorescence
generated
by
empty
pCAMBIA1302T in GFP channel seen in FigurelOC and D worked as a control
for tranformation through bombardment. Bright field image and merged with
fluorescence have been shown along with the laser scan to show the location of
the cells.
87
5.3. Discussion
Both positive and negative transcriptional regulation of transcription
factors by light has been documented. Among these transcription factors b-ZIP
proteins play an important role as those are early response factors and can
bind to light regulated elements such as Z-, and G- box present in light
regulated promoters (Chattopadhyay et al., 1998b; Mallappa et al., 2006; Lee et
al., 2007). Several plant bZIP proteins have been isolated from a variety of plant
species. Based on their DNA recognition sequences, these proteins appear to
fall into two somewhat overlapping classes: the TGACGT JC and the G-box
(CCACGTGG)
binding
proteins.
The
G-box
motif (CCACGTGG)
in
the
Arabidopsis RBCS-lA promoter is bound by the nuclear factor GBF (Schindler
et al., 1992a). The b-ZIP factors form dimers, involving leucine zipper domain
and bind to DNA with their basic domain as dimer (Siberil et al., 2001). The bZIP proteins can form homo- or hetero- dimers and thus they provide a flexible
regulatory system depending upon their overlapping expression pattern
temporally as well as spatially (Schindler et al., 1992b).
Our results reveal that GBFl and HYS proteins physically interact in vivo
and that they readily form G-box binding heterodimers and thereby suggesting
that GBFl
and HYS act together.
Our data suggest that GBFl/HYS
heterodimer is the predominant form, which binds to G-box present in RBCSlA promoter. However, in vivo GBFl/GBFl and HYS/HYS homodimers should
also be present to finely tune the regulation of gene expression and seedling
development. For example, in the absence of GBFl (in a null mutant
background) HYS/HYS homodimer pool predominates and binds to the LRE (Gbox), and thus increases the level of RBCS-lA. Similarly, in hyS mutant
background GBFl/GBFl homodimer level goes up, downregulating RBCS-lA.
Instead in wild type seedlings, presence of both the factors leads to the
formation of GBFl/HYS heterodimer in an appropriate proportion, which
sponsors fine tuning of RBCS-lA expression.
In previous studies GBFl and HYS both have been shown to accumulate
in early hours of light signal perception and to be degraded via 26S proteasome
mediated pathway in dark (Osterlund et al., 2000b; Mallappa et al., 2008).
88
From this perspective we wish to emphasize the interaction between GBF1 and
COP1 reported recently (Mallappa et al., 2008), which reveals possible role of
COP1 in light dependent transcription. Results from this study have indicated
that COP1 and SPA1 stabilize GBF1 in light grown seedlings, however it is well
established that COP1 destabilizes HYS with the help of SPA1 (Saijo et al.,
2003) in dark. Unambiguously GBF1 and HYS are directly related to the
subcellular shuttling of COP1 from nucleus to cytoplasm in response to the
onset of light signal (von Arnim and Deng, 1994; Osterlund and Deng, 1998).
Light might also regulate the subcellular localization of transcription factors
through phosphorylation (Harter et al., 1994). Possibly COP1 stabilizes GBF1
in cytoplasm of light grown seedlings and nuclear translocation of this pool of
GBF1 may be followed by GBF1/HY5 heterodimer formation based upon the
light conditions and physiological necessities. This supports the notion that in
absence of functional COP1, GBF1 gets destabilized and reduced pool of GBF1
molecules in nucleus may give rise to HYS/HYS homodimer formation and up
regulation of RBCSlA promoter in copl seedlings grown in light. Positive
signals from the photoreceptors received by downstream transcription factors
are thus balanced by heterodimer formation and binding to the cis- elements
resulting to a fine tuned gene expression.
Heterodimerization
of C/S
b-ZIP
transcription
factors
has
been
addressed in recent years vigorously (Ehlert et al., 2006) for regulation of
proline dehydrogenase (Weltmeier et al., 2006) and seed maturation genes
(Alonso et al., 2009). Intra family heterodimerization is also well known
(Schindler et al., 1992b; Holm et al., 2002). We first time report here the
heterodimerization of G/H family of b-ZIPs to regulate light stimulated gene
expression.
Heterodimerization is postulated to modulate
DNA binding
specificity in a study with the rice b-ZIP factor LIP19 /OBF1 in regulation of low
temperature induced genes (Shimizu et al., 2005). In context with light signal
this possibility needs to be addressed in assays like ChiP on chip in absence of
either GBF1 or HYS.
The co-localization experiments reflected that GBF1 is co-localized with HYS in
nucleus and no cytoplasmic pool of GBF1 was seen, which is however contrary
89
to the previous finding which confers nuclear as well as cytoplasmic subcellular localization of GBF1 in protoplast cultures grown in dark, blue and red
light conditions (Terzaghi et al., 1997). In this study, G BF's DNA binding
activity primarily found to be cytoplasmic which creates serious doubts upon
interpretation about the regulation of nuclear genes by DNA binding activity
GBFs. Alternatively it can be explained in other way that bombardment of the
exogenous
constructs
on onion
epidermal cells
may
lead
to
nuclear
accumulation in absence of the posttranslational modifications responsible for
the cytoplasmic retention. Another explanation can be given as the cytoplasmic
retention of GBF1 in protoplast might be seen during a specific phase of cell
division in dividing protoplast culture and protein extraction had been done at
the same stage when GBFs were localized in cytoplasm.
In spite of protein-protein interaction between GBF1 and HYS a recent
ChiP on chip study has reported GBF1 upstream region in immunoprecipitate
against HYS that put GBF1 downstream to HYS (Lee et al., 2007), which
supports our finding that GBF1 is epistatic to HYS for inhibition of hypocotyls
elongation in lower intensities of white light. Moderate intensities of cWL result
in antagonistic interaction of both the factors at the protein level. Another
possibility is the presence of two parallel pathways, which act antagonistically,
remains to be elucidated. It is not surprising if the phosphorylation and
dephosphorylation is involved. According to an older study, DNA binding
activity of GBF1 increases upon phosphorylation by a casein kinase II activity
(Klimczak et al.,
1992). Unlike GBF1,
HYS has been reported to be
physiologically more active, when unphosphorylated (Hardtke et al., 2000).
Phosphorylation status upon different amino acid moiety in a protein can affect
not only the interaction with other protein, but also the regulation. Binding to
DNA either as homodimer or heterodimer with other molecule might also be
largely affected, which can lead to the altered functional relationship seen in
gbfl hyS seedlings for different morphological and physiological responses, as
GBF1 interacts with HYS and both are important regulators in blue light signal
transduction.
90
It could be envisioned that heteridimerization of GBFl and HYH might be
a potential mechanism to generate positive or negative regulatory responses
specifically downstream to cryptochromes. Besides this, COPl and SPAl are
required for accumulation of GBFl protein in light, however COPliSPAl
machinary
helps
m
degradation
of HY5.
Another
negative
regulator
ZBFliAtMYC2IMYC2IJIN1 is known to interact genetically and physically
with COPl. It would be interesting to know how these interactions affect the
accumulation of ZBFl I AtMYC2 protein, but unfortunately antibodies raised
against three different epitopes of ZBFl I AtMYC2 were tried and found
producing detectable signal.
91