Plant Physiol. (1998) 118: 1111–1120
Update on Gene Regulation
Transcriptional Regulation in Plants: The Importance of
Combinatorial Control
Karam B. Singh
Department of Molecular, Cell and Developmental Biology, University of California,
Los Angeles, California 90095–1606
GLOSSARY OF TERMS
plant-defense/stress response. A major level at which gene
expression is regulated is the initiation of transcription,
and this is reflected in the percentage of the genome dedicated to transcription factors in plants and other eukaryotes. For example, an analysis of 1.9 Mb of Arabidopsis
genomic sequence from chromosome 4 revealed that about
15% of the genes with predicted or known functions were
involved in transcription, a percentage similar to what has
been found in other eukaryotes (Bevan et al., 1998).
In eukaryotic cells, genomic DNA is complexed with
proteins to form chromatin. One of chromatin’s major roles
is to facilitate the packaging of DNA in the nucleus, but the
structure of chromatin also leads to a general suppression
of gene activity. For gene activation and transcription to
occur, the chromatin in the vicinity of the gene must be
remodeled to allow access for transcription factors and
the recruitment of the RNA polymerase II (Pol II)
transcription-initiation complex. Transcription factors play
important and diverse roles in gene expression, including
chromatin remodeling and recruitment/stabilization of the
Pol II transcription-initiation complex. Transcription factors, which come in many shapes and sizes, can be divided
into a number of functional classes, with some proteins
belonging to more than one class. A major class of transcription factors is activators and repressors. These proteins bind to specific DNA sequences found only in certain
promoters and are instrumental in giving rise to genespecific regulation. A second class of transcription factors
are coactivators or corepressors. These proteins mediate
the transcriptional effects of specific activators/repressors,
in some cases by remodeling chromatin. Whereas this
group of transcription factors are typically not able to bind
to DNA on their own, they can still be promoter-specific as
a result of protein-protein interactions with specific activators and repressors. A third class comprises the general
transcription factors, which are important components of
the Pol II transcription-initiation complex. A fourth class is
architectural transcription factors that are also involved in
remodeling DNA, e.g. by inducing bends that facilitate the
binding of other proteins to the promoter.
In this Update I will address how transcription factors
regulate gene transcription in plants, as well as relying on
Combinatorial control: use of a discrete number of transcription factors in different combinations to give rise to
a wide spectrum of expression patterns.
Enhanceosome: a higher-order nucleoprotein complex that
is formed by the binding of a specific combination of
transcription factors to the transcriptional regulatory sequences of a particular gene.
General transcription factors: components of the Pol II
transcription-initiation complex that are thought to be
common to all Pol II promoters.
Holoenzyme: a large protein complex that is preformed
off of the DNA and contains Pol II and many of the
other components of the Pol II transcription-initiation
complex.
Pol II transcription-initiation complex: a large protein machine that contains dozens of polypeptides and assembles near the transcription start site.
Transcriptional activators: class of regulatory proteins that
are gene specific and function to increase transcription
from target promoters.
Transcriptional synergy: when specific combinations of
transcription factors give rise to significantly higher levels of transcription than the sum of the additive effects
obtained when each factor is assayed individually.
During their development and differentiation, plants
need to integrate a wide range of tissue, developmental,
and environmental signals to regulate complex patterns of
gene expression. The regulation of seed-storage protein
gene expression, resulting in expression during specific
stages of seed development but not in other parts of the
plant, is a striking example of tissue and developmental
control and is of considerable agricultural importance.
Plants also have unique needs and strategies for responding to changes in their environment. When light strikes an
etiolated leaf, numerous genes encoding chloroplastic, mitochondrial, peroxisomal, and cytosolic proteins are activated. Similarly, a number of biotic and abiotic stresses
cause a battery of genes to be activated as part of the
1
This work was supported by grants from the National Institutes of Health and the U.S. Department of Agriculture.
* Corresponding author; e-mail [email protected]; fax 1–310 –
206 –3987.
Abbreviations: ABRC, ABA-responsive complex; ABRE, ABAresponsive element.
1111
1112
Singh
advances in other systems. The focus will be on activators
and the importance of combinatorial control. First, I will
comment on chromatin, chromatin remodeling, and the Pol
II transcription-initiation complex, since it is the recruitment and/or activity of the transcription-initiation complex that is regulated by the gene-specific transcription
factors, and this regulation occurs in the context of
chromatin.
CHROMATIN REMODELING AND
TRANSCRIPTIONAL REGULATION
Chromatin has several levels of structural organization,
with the basic unit being the nucleosome core, which consists of 146 bp of DNA wrapped around a histone octamer.
The presence of nucleosomes affects the accessibility of
DNA to other proteins, including transcription factors and
the Pol II transcription-initiation complex. Higher orders of
chromatin structure are also likely to affect transcription,
e.g. by leading to the organization of chromatin into active
and silent regions. Exciting progress has been made recently on chromatin remodeling, including the involvement of histone acetylation/deacetylation on nucleosome
conformation/stability, and the identification of novel protein complexes that cause nucleosome disruption (for recent reviews, see Cairns, 1998; Struhl, 1998).
Histone acetylation was first reported in 1964 and at the
time was proposed to play a role in the regulation of
transcription (Allfrey et al., 1964). Studies during the next
three decades supported this proposal, although until recently the molecular mechanisms involved were not
known. Acetylation of histones occurs on Lys residues in
the amino-terminal tails that protrude from the surface of
the nucleosome. Acetylation neutralizes the positive charge
of the histone tails and consequently causes a reduction in
their affinity for DNA. This leads to changes in nucleosome
conformation and may lead to unfolding of the nucleosome. Thus, acetylation of histones is normally correlated
with transcriptional activity by facilitating the access of
transcription factors to the DNA, whereas deacetylation of
histones is correlated with transcriptional repression.
A growing number of histone acetylases and histone
deacetylases have been identified recently. Significantly,
many of these proteins had already been associated with
transcription, with some having been identified previously
as components of the Pol II transcription-initiation complex, and others initially identified as coactivators or corepressors. For example, the yeast Gcn5 and mammalian
p300/CBP proteins were initially identified as coactivators
and were subsequently found to be histone acetylases (Bannister and Kouzarides, 1996; Brownell et al., 1996; Ogryzko
et al., 1996). In the case of Gcn5, the histone acetylase
activity has been shown to be required for coactivator
function in vivo (Candau et al., 1997), and overexpression
of Gcn5 leads to increased histone acetylation at promoter
regions of genes regulated by Gcn5 (Kuo et al., 1998).
Histone acetylation has also been observed in plants (Belyaev
et al., 1997) and a maize histone deacetylase has been identified (Lusser et al., 1997).
Plant Physiol. Vol. 118, 1998
Other forms of chromatin remodeling involving multiprotein complexes with ATP-dependent chromatin remodeling activities have also been observed. A good example is
the SWI2/SNF2 complex, which was initially discovered
through genetic studies as a transcriptional regulator of
specific genes in yeast (Stern et al., 1984; Neigeborn and
Carlson, 1984). A number of lines of evidence link the yeast
SWI2/SNF2 complex, which has a size of about 2,000,000
D, to chromatin remodeling (for review, see Cairns, 1998).
Chromatin remodeling does not occur at promoters normally regulated by the SWI2/SNF2 complex in strains with
a defective SWI2/SNF2 complex. In addition, the purified
SWI2/SNF2 complex was able to cause chromatin remodeling to occur in vitro. SWI2/SNF2-related complexes have
also been identified in Drosophila melanogaster and humans.
In D. melangoaster three additional chromatin-remodeling
complexes have been identified, including NURF, which is
able to stimulate the binding of a number of transcription
factors to chromatin templates (Tsukiyama and Wu, 1996).
The identification of distinct multiprotein complexes involved in chromatin remodeling raises important questions
regarding the role each plays in transcriptional regulation,
how they are targeted to specific promoters, and whether
they interact with other types of chromatin-remodeling
activities, such as histone acetylases.
THE POL II TRANSCRIPTION-INITIATION COMPLEX IS
A LARGE PROTEIN MACHINE
The Pol II transcription-initiation complex has a size in
excess of 2,500,000 D (Fig. 1). Pol II itself is a complex
enzyme that in yeast consists of 14 subunits. Whereas Pol II
catalyzes RNA synthesis, numerous other proteins are required for promoter recognition and accurate transcription
initiation. These include the six general transcription factors (Fig. 1), as well as a growing number of accessory
proteins. Some of these accessory proteins may serve as
coactivators or may be involved in chromatin remodeling,
whereas others probably provide regulatory functions that
remain to be elucidated. Some of the general transcription
factors are complex, e.g. TFIID is composed of TATA box
binding protein (TBP) and a number of TBP-associated
factors (TAFs).
The general transcription factors and some of the accessory proteins were first identified biochemically using in
vitro transcription systems. Based on the in vitro studies a
stepwise model was proposed for the assembly of the
transcription-initiation complex starting with the binding
of TFIID to the TATA box (for review, see Roeder, 1996). In
this model a major function of transcriptional activators
was to facilitate the stepwise assembly of the transcriptioninitiation complex. However, some of the Pol II in cells has
been shown to exist in a large protein complex(es) called
the holoenzyme (for review, see Greenblatt, 1997), suggesting that to a large extent the transcription-initiation complex may already be formed in the absence of DNA. A
major role of transcriptional activators may be to recruit
the holoenzyme to specific promoters and/or to stabilize
the transcription-initiation complex once bound at the core
promoter.
Transcriptional Control
1113
Figure 1. A simplified schematic representation of the transcription-initiation complex assembled at a typical Pol II core
promoter containing a TATA box. Pol II, the general transcription factors IIA, IIB, IID (TBP and TAFs) IIE, IIF, and IIH, as well
as a few of the accessory factors (CBP and SRB) are shown. The precise positions of a number of these factors in the
transcription-initiation complex remain to be established. Transcriptional activators (ACT) typically bind at specific sites
upstream of the core promoter. Interactions between activation domains present on the activators and components of the
general transcriptional machinery help to recruit and/or stabilize the transcription-initiation complex.
Until recently, the composition of the transcriptioninitiation complex was not considered to be a major area
of regulation. However, this view is changing with reports
of multiple forms of the holoenzyme and the discovery of
accessory proteins such as the TAFs. TAFs, which were
initially identified as coactivators required by certain activators to function in vitro, also function in core promoter
recognition in vivo (Shen and Green, 1997), and TAF250
has histone acetylase activity (Mizzen et al., 1996). Celltype-specific TAFs have also been identified, e.g. the B-celltype-specific TAF105 (Verrijzer and Tjian, 1996). The potential for TFIID to serve as a major point of regulation has
also been suggested by the discovery of multiple TBPs in
some species, including Arabidopsis. In D. melanogaster, in
addition to a ubiquitous TBP, a second tissue-specific TBP,
TRF (TBP-related factor), has been identified that may be
important for neural-specific patterns of gene regulation
(Hansen et al., 1997). Recently, a TFIID complex that does
not contain TBP has been isolated from human cells (Wieczorek et al., 1998). Surprisingly, this complex, called TFTC
(TBP-free TAF-containing complex), was shown to be able
to substitute for normal TFIID and support transcription
from both TATA-box-containing promoters and TATA-less
promoters, using an in vitro transcription system. Although these exciting results demonstrate that multiple
transcription-initiation complexes exist and offer additional opportunities for the regulation of transcription, it
remains likely that the major level of transcriptional control
is mediated by transcriptional activators and repressors.
PLANTS CONTAIN FAMILIES OF DNA-BINDING
PROTEINS THAT BEHAVE AS
TRANSCRIPTIONAL ACTIVATORS
Whereas there has been relatively little work on the
general transcriptional machinery in plants in comparison
with animals and yeast, an increasing number of transcriptional regulatory proteins have been identified in plants.
Although transcriptional regulators can be activators or
repressors and, in some cases, the same protein can serve
both functions, the focus of this Update is on activators.
Typically, activators have a modular structure consisting of
discrete domains responsible for specific DNA binding,
transcriptional activation, and in some cases dimerization
and/or other forms of protein-protein interactions.
Whereas domain-swap experiments have shown that
DNA-binding domains and activation domains can operate
independently when fused to a heterologous protein, there
is growing evidence that in their native protein context
they may sometimes functionally communicate (for review, see Lefstin and Yamamoto, 1998).
Plant transcription factors contain a variety of structural
motifs that allow for binding to specific DNA sequences.
Most of these DNA-binding domains were first identified
in transcription factors isolated in animal/yeast systems. In
many cases the structure of the DNA-binding domain
bound to DNA has been determined, allowing a good
understanding of how these DNA-binding domains function. In bZIP transcription factors, DNA binding and
dimerization are mediated by the bZIP motif, which consists of a region rich in basic amino acids and an adjacent
bZIP that consists of a 4–3-heptad repeat of hydrophobic
and nonpolar residues (for review, see Hurst, 1995). The
bZIP is required for dimerization of the protein prior to
binding to DNA, whereas the basic region contacts the
DNA-recognition site. Crystal structure of the bZIP motif
of the yeast GCN4 protein complexed with DNA demonstrated that the bZIP motifs resembled a helical forceps as
they gripped the major grove of DNA (Ellenberger et al.,
1992; Fig. 2). The bZIP region of each monomer is packed
together in a coiled coil, whereas the basic region passes
through the major grove and makes a number of contacts
with specific DNA bases and the phosphate backbone.
Many transcription factors have been broadly classified
into a number of families on the basis of the type of
DNA-binding domain, such as bZIP, homeodomain, helixturn-helix, helix-loop-helix, and zinc-finger proteins. In
some cases these can be large protein families. A good
example are MYB family members, which contain a specific
type of the helix-turn-helix motif of about 50 amino acids,
which serves as the DNA-binding domain. In Arabidopsis
more than 80 MYB family members have already been
identified (Romero et al., 1998). Some plant transcription
factors have DNA-binding domains that appear to be
1114
Figure 2. Structure of the bZIP motif of the yeast GCN4 protein
bound to DNA. The two bZIP monomers are depicted in yellow and
the DNA is depicted in red. For further details, see the text and figure
3a from Ellenberger et al. (1992).
unique to plants. For example, the AP2/EREBP family of
plant transcription factors, found in a range of higher
plants, contain a conserved, approximately 60- to 70-amino
acid region required for DNA binding, a part of which has
been predicted to form an amphipathic a-helix (Okamuro
et al., 1997). The large family of viviparous 1 (VP1)-related
proteins, which includes the maize VP1 protein (McCarty
et al., 1991) and the Arabidopsis ARF1 and Monopteros
proteins (Ulmasov et al., 1997; Hardtke and Berleth, 1998),
represents a second class of plant-specific DNA-binding
proteins that play important roles in plants.
Although plant transcription factors typically have only
a single DNA-binding domain, there are examples, such as
the GT-2 and APETALA2 proteins, in which there are two
related DNA-binding domains (Ni et al., 1996; Okamuro et
al., 1997). In the case of GT-2, a rice DNA-binding protein
that interacts with GT-box promoter elements in the rice
phytochrome A gene, the different DNA-binding domains
have been shown to discriminate between closely related
GT-box sequences (see Ni et al., 1996, and refs. therein). For
some transcription factors the ability to form either homodimers and/or heterodimers with related family members is a prerequisite for DNA binding. The ability to form
specific heterodimers is a form of combinatorial control,
and this can expand the number of DNA target sequences
that can be recognized, as well as allow different combinations of activation domains to be recruited to a promoter
element.
Singh
Plant Physiol. Vol. 118, 1998
Several classes of transcriptional activation domains
have been identified and some can be classified on the basis
of their amino acid composition/properties. For example,
there are acidic, Gln-, Pro-, and Ser/Thr-rich activation
domains. Other strong activation domains have been identified that are not particularly rich in any specific amino
acid. Many activators contain more than one activation
domain. Relatively little is known about the structures of
activation domains and how they function, although there
have been many reports of interactions of specific types of
activation domains with one or more components of the
transcription-initiation complex.
The notion that a major function of activation domains is
recruitment of the transcriptional machinery to a given
promoter has gained support from studies of yeast involving activator bypass experiments (for review, see Ptashne
and Gann, 1997). In these experiments fusion of a DNAbinding domain to different components of the holoenzyme was sufficient to confer transcription on genes containing the corresponding DNA-binding site, and
activators were no longer required. However, recruitment
of the transcriptional machinery may not be the only way
that activation domains function. For example, some activation domains may function by altering the conformation
of the transcription-initiation complex assembled at the
core promoter, facilitating promoter escape at some step
during transcription elongation. Controlling both the type
and arrangement of activation domains brought to a specific promoter is a major way that transcriptional control is
manifested, which is discussed below.
COMBINATORIAL CONTROL IS IMPORTANT FOR
TRANSCRIPTIONAL REGULATION AND CAN
LEAD TO FORMATION OF HIGHER-ORDER
NUCLEOPROTEIN COMPLEXES
It is becoming clear that a major mechanism underlying
eukaryotic transcriptional regulation is combinatorial control. Many genes are regulated by multiple transcriptional
activators by virtue of having a specific set of proteinbinding sites in their promoters. At any given time, a
distinct set of transcription factors bind to these different
sites to give rise to higher-order nucleoprotein complexes
that have been called enhanceosomes (for a recent review,
see Carey, 1998). Specific interactions between proteins
that form an enhanceosome, as well as interactions with
components of the general transcriptional machinery, can
then lead to cooperativity in DNA binding and transcriptional synergy. The composition of the enhanceosome assembled at a given promoter may change in response to
environmental, developmental, or other signals. It is possible to form a much larger number of distinct enhanceosomes by using a discrete number of transcription factors
in different combinations. A corollary to this is that,
through combinatorial interactions, a given transcription
factor can play multiple roles and help regulate different
genes whose expression is induced by distinct signals.
There is a growing body of evidence from animal systems,
using both in vivo and in vitro studies, for the existence
and importance of enhanceosomes in mediating transcrip-
Transcriptional Control
tional regulation. In plants there is also evidence for combinatorial control, which is discussed below.
TRANSCRIPTIONAL CONTROL OF ANTHOCYANIN
EXPRESSION BY COMBINATORIAL INTERACTIONS
BETWEEN Myb AND b/HLH PROTEINS
The regulation of pigment production in maize is one of
the best-characterized examples in plants for the importance of combinatorial interactions in gene regulation (for
recent reviews, see Mol et al., 1996, 1998). The biosynthesis
of one class of maize pigments, the anthocyanins, is regulated by both developmental and environmental signals.
UV light induces anthocyanin expression in the epidermis
of leaves and petioles, whereas developmental cues induce
their expression in a number of organs such as kernels,
seedlings, and leaves. The developmental regulation of
anthocyanin expression is the result of combinatorial interactions between two distinct families of plant transcription
factors. The first is the C1 family, which are Myb-related
regulatory proteins and consist of the closely related C1
and P1 proteins (Paz-Ares et al., 1987; Cone et al., 1993).
The second is the R family, which are encoded by the B and
R loci (Ludwig and Wessler, 1990). R family members
contain the b/HLH (basic helix-loop-helix) motif and also
share a high degree of amino acid homology with each
other. The members of each family differ in their tissuespecific expression patterns, which in turn reflect the patterns of pigmentation observed in maize. However, individual family members are not sufficient to induce the
anthocyanin biosynthetic genes, but, rather, a member
from each family must be coexpressed in a particular tissue
for anthocyanin biosynthesis to occur. Thus, genetic studies have demonstrated that the combination of R and C1 is
responsible for pigmentation in the kernel, whereas the
combination of B and P1 is responsible for pigmentation in
mature tissues of the plant, such as the husk leaves (for
review, see Mol et al., 1996). The importance of combinatorial control was also illustrated by transgenic studies
showing that the expression of just C in tobacco or Arabidopsis was not sufficient to increase anthocyanin expression but required the coexpression of R (Lloyd et al., 1992).
The precise mechanism by which combinatorial interactions between the Myb and b/HLH proteins leads to transcriptional activation is not known. Promoter sequences
required by C1 and B/R for activation have been mapped
in the promoters of anthocyanin biosynthetic genes (Sainz
et al., 1997; Lesnick and Chandler, 1998, and refs. therein).
Several possibilities for why C1 requires B or R for transcriptional activation were proposed by Sainz et al. (1997).
One possibility is that B/R acts to increase C1 DNAbinding specificity to promoters in anthocyanin biosynthetic genes, although this seems less likely, since C1 is able
to bind to a specific site in the a1 gene promoter in the
absence of B or R. Another possibility is that the C1 Myb
domain inhibits the C1 activation domain and that this
inhibition is relieved through the interaction with B or R. It
is interesting that, whereas C1 has been shown to have a
strong activation domain, careful analysis of B did not
reveal any activation domains. The B/R proteins may also
1115
assist with the nuclear localization of C1. Whereas C1
absolutely requires either R or B to activate the anthocyanin biosynthetic gene promoters including the a1 gene,
another Myb member, P1, is able to activate a subset of the
anthocyanin biosynthetic genes, including the a1 gene,
without a requirement for either the B or R proteins (Grotewold et al., 1994). Whether P1 is truly acting independently
or is interacting with another b/HLH protein is not known.
COMBINATORIAL CONTROL MEDIATING
ABA-INDUCED GENE EXPRESSION
The plant hormone ABA regulates a number of processes
in plants, including helping to mediate the response to a
number of environmental stresses, as well as the generation of specific expression patterns during seed development. ABA-induced gene expression is an important part
of ABA action (for recent reviews, see Shen and Ho, 1997;
Busk and Pages, 1998). A detailed analysis of the cis-acting
sequences required for ABA-induced gene expression has
been performed on a number of ABA-regulated genes, such
as the wheat EM gene and two barley genes, HVA1 and
HVA22. The first ABRE to be identified was an ACGTcontaining sequence in the EM promoter (Guiltinan et al.,
1990). ABREs have now been found in the promoters of
many ABA-responsive genes, including the rice rab and
barley HVA genes. There are ABA-responsive genes that do
not contain ABREs, and other cis-acting sequences have
been shown to function in ABA-responsive gene expression in some of these promoters.
The ABRE is similar to a family of sequences called the
G-box, which also contain an ACGT core and are present in
a number of gene promoters that respond to different
environmental conditions, such as UV light, anaerobiosis,
and wounding (Menkens et al., 1995). A bZIP protein called
EmBP1 that specifically binds to the ABRE has been identified (Guiltinan et al., 1990). EmBP1 may be part of a larger
protein-DNA complex, which includes VP1 and GF14 proteins (Shultz et al., 1998). In maize VP1 has been shown by
genetic analysis to be important for mediating certain ABA
responses during seed maturation (Carson et al., 1997, and
refs. therein). Although VP1 is unable to directly bind the
ABRE, it is able to transactivate the Em promoter through
the ABRE sequences, presumably via protein-protein interaction with other proteins, such as EmBP1, that are recruited to the ABRE (McCarty et al., 1991; Vasil et al., 1995).
Analyses of truncated forms of VP1 show that the DNAbinding domain (B3) is not required for gene activation
mediated through the ABRE (Carson et al., 1997). It is
interesting that the activation of ABRE-coupled genes requires a second conserved basic domain of VP1, which in
vitro can stimulate the DNA-binding activity of a broad
spectrum of transcription factors, including EmBP1 (Hill et
al., 1996). However, the actual order of events remains
unclear because in vivo footprinting studies did not detect
any major differences in ABRE binding when wild-type
and vp1 mutant embryos were compared (Busk and Pages,
1997). GF14s are plant 14-3-3 proteins that were initially
identified as part of a G-box protein-DNA complex (for
review, see Ferl, 1996). Although their precise functions
1116
remain to be elucidated, they may play a role in promoting
protein-protein interactions. GF14, which is unable to directly bind DNA, was found to interact with both EmBP1
and VP1 and could therefore provide a structural link in
the ABA-responsive protein-DNA complex proposed by
Shultz et al. (1998).
The promoter context can be important for how ABRE/
G-box sequences give rise to specific expression patterns. A
detailed analysis of the barley HVA1 and HVA22 promoters (Shen and Ho, 1995; Shen et al., 1996) demonstrated that
the promoter unit necessary and sufficient to mediate the
ABA response consisted of an ABRE and a closely linked
sequence, called a coupling element (CE), that collectively
constitute an ABRC. As shown in Figure 3, the HVA1 and
HVA22 promoters each had a different ABRC, called
ABRC3 and ABRC1, and distinct CEs, called CE1 and CE3,
respectively. CE1 and CE3 differed both in their nucleotide
sequence and their position relative to the ABRE. Whereas
the ABRE sequences were interchangeable, the CEs were
only partially exchangeable, with CE3 exhibiting flexibility
in terms of its position relative to the ABRE. In contrast,
CE1 could function only when placed distal to the ABRE.
Further underscoring the differences between ABRC1 and
ABRC3 was the finding that VP1 was able to enhance
transcription only through ABRC3 but not ABRC1. It is
interesting that synthetic promoters containing an ABRE
and both CE1 and CE3 were substantially more ABA responsive. Whereas distinct ABRCs may allow for the regulation of ABA-responsive genes in response to different
Singh
Plant Physiol. Vol. 118, 1998
environmental and/or physiological cues, a number of
important questions still need to be addressed. For example, the protein(s) that acts through CE1 and CE3 needs to
be identified and the mechanisms by which ABA leads to
activation of specific bZIP proteins need to be understood.
In rice, de novo protein synthesis may play a role, since the
expression of a bZIP protein called OSBZ8 is rapidly induced by ABA (Nakagawa et al., 1996).
The maize VP1 protein illustrates an important aspect of
combinatorial control whereby a given activator can have
different roles depending on the promoter context. As
mentioned earlier, VP1, acting as an activator, is involved
as part of an ABA-responsive protein-DNA complex to
mediate the ABA-induced expression of the wheat Em
gene. In a different promoter context, VP1 acts as a repressor of a-amylase expression (Hoecker et al., 1995). VP1 also
plays a role in anthocyanin gene expression by transactivating the C1 promoter through a DNA sequence called the
Sph element (Kao et al., 1996, and refs. therein). The Sph
element is quite distinct from the ABRE, which is important for VP1 transactivation of the Em promoter. While
full-length VP1 is unable to bind to the Sph element, the
120-amino acid B3 domain, located at the C terminus of
VP1, is able to bind to the Sph element with high specificity
(Suzuki et al., 1997). An exciting possibility is that other
regions of VP1 inhibit the DNA-binding activity of the B3
domain and that this inhibition may be relieved through
protein-protein interactions between VP1 and other proteins to enable VP1 binding to the C1 promoter.
Dof/bZIP INTERACTIONS MAY REGULATE SPECIFIC
PATTERNS OF PLANT GENE EXPRESSION
Figure 3. A schematic representation of two ABRCs assembled at the
promoters of ABA-responsive genes. A, Promoter sequences from the
barley HVA22 gene that form ABRC1. Sequences of the ABRE (A3)
and the CE (CEI) are shown. bZIP proteins bind to A3 as dimers,
whereas the protein(s) that bind to CEI remains to be identified. B,
Promoter sequences from the barley HVA1 gene that form ABRC3.
The sequences of the ABRE (A2) and the CE (CE3) are shown. bZIP
proteins bind to A2 as dimers, whereas the protein(s) that bind to CE3
remains to be identified. The maize VPI protein, which can transactivate from ABRC3 but not ABRC1, is also shown, although the
precise position of VPI in the complex remains to be established.
There is evidence that in addition to ABA responses,
regulation of some other patterns of plant gene expression
are mediated in part by combinatorial interactions between
bZIP proteins and other types of transcription factors binding to closely linked sites. A good example is the interaction of bZIP and Dof transcription factors. Dof proteins are
a new class of plant transcription factors that contain a
single zinc-finger DNA-binding domain that is highly conserved in plants (for review, see Yanagisawa, 1996). Dof
proteins have been shown to interact specifically with bZIP
proteins, and this interaction results in stimulation of bZIP
binding to DNA target sequences in plant promoters (Chen
et al., 1996). The Arabidopsis glutathione S-transferase-6
gene (GST6) promoter contains a number of Dof-binding
sites closely linked to another promoter sequence called the
ocs element. The ocs elements are a family of 20-bp DNA
promoter sequences that are important for the expression
of a number of pathogen and plant genes and are the
binding sites for bZIP proteins.
The regulation of seed-storage protein gene expression
in maize is another case in which Dof/bZIP interactions
may play an important role. A major class of maize seedstorage proteins are the 22- and 19-kD zein proteins (for
review, see Aukerman and Schmidt, 1994). As shown in
Figure 4, two cis-acting sequences, the O2 (Opaque 2) box
and the prolamin box, have been found to be important for
22-kD zein expression in endosperm tissue. The O2 box is
Transcriptional Control
the binding site for a bZIP protein called O2, and molecular
genetic studies have demonstrated that O2 is required for
the expression of the 22-kD zein promoter in maize endosperm. The prolamin box is found in the promoters of all
zein genes, as well as in storage protein genes in other
cereals. In maize the prolamin box is the binding site for a
Dof protein called PBF (Vicente-Carbajosa et al., 1997). PBF
and O2 are expressed in an identical fashion in maize
endosperm tissue, where they accumulate just before activation of zein gene expression. PBF interacts specifically
with O2 but not another bZIP protein tested (VicenteCarbajosa et al., 1997). The PBF/O2 interaction is likely to
be important, since a functional prolamin box is required
for O2-dependent activation in maize and for activation by
related bZIP proteins in other cereals (Albani et al., 1997;
Vicente-Carbajosa et al., 1997, 1998). It is interesting that
other classes of zein genes lack an O2-binding site but
contain the PBF site. In these cases it is not known whether
PBF interacts with other transcription factors or is able to
act alone.
ENHANCEOSOME FORMATION ON THE IFNb
PROMOTER CAN SERVE AS A USEFUL PARADIGM
FOR PLANT STUDIES
The above examples provide evidence for the importance
of combinatorial control in transcriptional regulation of
plant gene expression. However, the situation in a number
of cases is likely to be more complex with additional proteins playing a role. In animal systems a number of good
examples of combinatorial control have been identified and
shown to lead to the formation of complex enhanceosomes
(Carey, 1998). One of the best-characterized examples
comes from analysis of the IFNb (interferon b) promoter,
which is highly induced following viral infection. A detailed analysis of the promoter region responsible for the
transcriptional enhancement, called the IFNb enhancer,
revealed the organization shown in Figure 5. The IFNb
enhancer contains a number of distinct sequences located
close to each other to give a compact yet complex regulatory region (Thanos and Maniatis, 1995; Kim and Maniatis,
1997, and refs. therein). Six distinct proteins bind to the
IFNb enhancer. The bZIP proteins ATF2 and c-Jun bind as
a heterodimer to a single site, whereas the p50/p65 sub-
Figure 4. A schematic representation of DNA-protein interactions
involved in expression of the maize 22-kD zein gene. The core
sequences of the prolamin (P) box and the O2 box are shown. The
bZIP protein O2 binds to the O2 box as a dimer. The Dof protein PBF
binds to the prolamin box, although at this stage it is not known if
PBF binds as a monomer or as some form of multimer. The dashed
arrows represent potential interactions between PBF and O2.
1117
Figure 5. A schematic representation of the virus-inducible IFNb
enhanceosome. The positions of the four positive regulatory domains
are shown. The cooperative assembly of the enhanceosome leads to
an optimal arrangement of activation domains for interactions with
various components of the transcription-initiation complex. See the
text for further details.
units of NF-kB bind to a second site. An IRF family member
binds to two additional sites and the HMGI (Y) protein
binds to four ATrich sequences located in different regions
of the enhancer.
Analysis of the IFNb enhancer identified four positive
regulatory domains (Fig. 5). Mutations in any of these
domains caused a substantial decrease in the level of induction following viral infection. Each of the positive regulatory domains could also serve as a virus-inducible enhancer if multimerized. However, the pattern of induction
of these artificial enhancers differed significantly from the
wild type. The artificial enhancers had higher basal activity
and were consequently less inducible. In addition, the artificial enhancers were responsive to a number of inducers,
whereas the wild-type enhancer was activated only following viral infection. The helical relationship between the
regulatory elements in the IFNb enhancer is also important
for both the assembly and transcriptional activity of the
enhanceosome, presumably by facilitating cooperative
binding. Thus, insertions of a half-helical turn of DNA
between regulatory elements caused a dramatic decrease in
the level of virus induction. However, insertion of a fullhelical turn of DNA caused no significant change in transcriptional activity.
The ability to reconstruct enhanceosome activity using
an in vitro approach has provided important insight into
the mechanism of transcriptional synergy (Kim and Maniatis, 1997). The in vitro system utilized nuclear extracts
depleted of the endogenous IFNb enhancer-binding proteins. It was demonstrated, using this system and limiting
concentrations of each of the six IFNb enhancer-binding
proteins, that the full synergistic activation of the IFNb
promoter required all of the IFNb enhancer-binding proteins to be present. Under saturating conditions the requirement for HMGI(Y), an architectural protein that does
1118
Singh
not itself activate transcription, was lost. HMGI(Y) appears to function by binding to DNA and inducing conformational changes that facilitate protein-protein interactions and cooperative DNA binding of the other
proteins in the enhanceosome. HMGI(Y) is also able to
directly interact with ATF2 and p50. A competition assay
was used to assay the stability of different complexes
assembled at the IFNb enhancer. These studies demonstrated that there was a good correlation between stability
of the enhanceosome and transcriptional activity. The
stability of the enhanceosome was further increased in the
presence of components of the general transcriptional
machinery, suggesting that the enhanceosome and
transcription-initiation complexes reciprocally stabilize
assembly. The IFNb enhanceosome should serve as a
useful paradigm for studies of transcriptional control in
plants.
CONCLUSIONS
I have reviewed some basic features of eukaryotic transcription with an emphasis on plant activators and the
importance of combinatorial control. The IFNb promoter
illustrates nicely how a high degree of specificity and
transcriptional activity can be generated by the cooperative assembly and stability of a higher-order protein-DNA
complex called an enhanceosome. It is the precise arrangement of the regulatory sequences and their corresponding transcription factors on the DNA that facilitate
the many protein-protein interactions both within the
enhanceosome and between the enhanceosome and components of the transcription-initiation complex. It is these
multiple protein-protein and protein-DNA interactions
that in large part dictate the amount of transcription that
will occur at a given promoter under a specific set of
conditions. The IFNb enhanceosome also demonstrates
the importance of in vivo studies to validate/determine
those combinatorial interactions identified initially at the
molecular/biochemical level that are actually being used
to regulate gene expression in vivo. This is certainly the
case with the regulation of anthocyanin biosynthesis in
maize, in which there is substantial genetic evidence that
the interactions between C and R proteins are critical for
transcriptional control.
The analysis of transcriptional control in plants will
continue to be an exciting field of research. The rapid
progress being made on the isolation of important regulatory proteins, the development of in vitro transcription
systems, and the use of powerful genetic screening approaches for additional mutants using promoter/reporter
gene fusions will facilitate further studies of transcriptional control in plants, which should provide valuable
insight into the mechanisms underlying various aspects of
plant growth and development and lead to agricultural
benefits. The potential for agriculture is illustrated by
studies of the molecular basis of plant tolerance to freezing, an important research area, since freezing temperatures have placed major limitations on agricultural productivity. Through the identification of cold-regulated
Plant Physiol. Vol. 118, 1998
genes and the careful analysis of their expression, an
Arabidopsis transcriptional activator called CBF1 has
been identified (Stockinger et al., 1997), overexpression of
which increased the freezing tolerance of Arabidopsis
plants (Jaglo-Ottosen et al., 1998). If similar mechanisms
are used for freezing tolerance in important crop plants,
the use of CBF1 and/or its orthologs may help make other
plants more cold-tolerant.
ACKNOWLEDGMENTS
I would like to thank Drs. Mike Carey, Maarten Chrispeels, Don
McCarty, and Bob Schmidt for their helpful comments concerning
the manuscript, and Dr. Stephen Harrison and Cell Press for
permission to use Figure 2. I apologize to the many colleagues
whose work or original publications I was unable to cite because of
space constraints.
Received July 13, 1998; accepted August 11, 1998.
Copyright Clearance Center: 0032–0889/98/118/1111/10.
LITERATURE CITED
Albani D, Hammond-Kosack MCU, Smith C, Conlan S, Colot V,
Holdsworth M, Bevan MW (1997) The wheat transcriptional
activator SPA: a seed-specific bZIP protein that recognizes the
GCN4-like motif in the bifactorial endosperm box of prolamin
genes. Plant Cell 9: 171–184
Allfrey V, Faulkner RM, Mirsky AE (1964) Acetylation and methylation of histones and their possible role in the regulation of
RNA synthesis. Proc Natl Acad Sci USA 51: 786–794
Aukerman MJ, Schmidt RJ (1994) Regulation of alpha-zein gene
expression during maize endosperm development. In L Nover,
ed, Results and Problems in Cell Differentiation 20: Plant Promoters and Transcription Factors. Springer-Verlag, Berlin, pp
209–233
Bannister AJ, Kouzarides T (1996) The CBP co-activator is a
histone acetyltransferase. Nature 384: 641–643
Belyaev ND, Houben A, Baranczewski P, Schubert I (1997) Histone H4 acetylation in plant heterochromatin is altered during
the cell cycle. Chromosoma 106: 193–197
Bevan M, Bancroft I, Bent E, Love K, Goodman H, Dean C,
Bergkamp R, Dirkse W, Van Staveren M, Stiekema W, and
others (1998) Analysis of 1.9Mb of contiguous sequence from
chromosome 4 of Arabidopsis thaliana. Nature 391: 485–488
Brownell JE, Zhou J, Ranalli T, Kobayashi R, Edmondson DG,
Roth SY, Allis CD (1996) Tetrahymena histone acetyltransferase
A: a homolog to yeast Gcn5p linking histone acetylation to gene
activation. Cell 84: 843–851
Busk, PK, Pages, M (1997) Protein binding to the abscisic acidresponsive element is independent of VIVIPAROUS1 in vivo.
Plant Cell 9: 2261–2270
Busk PK, Pages M (1998) Regulation of abscisic acid-induced
transcription. Plant Mol Biol 37: 425–435
Cairns BR (1998) Chromatin remodeling machines: similar motors,
ulterior motives. Trends Biochem Sci 23: 20–25
Candau R, Zhou JX, Allis CD, Berger SL (1997) Histone acetyltransferase activity and interaction with ADA2 are critical for
GCN5 function in vivo. EMBO J 16: 555–565
Carey M (1998) The enhanceosome and transcriptional synergy.
Cell 92: 5–8
Carson CB, T Hattori, L Rosenkrans, V Vasil, IK Vasil, PA
Peterson, DR McCarty (1997) The quiescent/colorless alleles of
viviparous1 show that the conserved B3 domain of VP1 is not
essential for ABA regulated gene expression in the seed. Plant J
12: 1231–1240
Transcriptional Control
Chen W, Chao G, Singh KB (1996) The promoter of a H2O2inducible, Arabidopsis glutathione S-transferase gene contains
closely linked OBF- and OBP1-binding sites. Plant J 10: 955–966
Cone KC, Cocciolone SM, Burr FA, Burr B (1993) Maize anthocyanin regulatory gene pl is a duplicate of c1 that functions in the
plant. Plant Cell 5: 1795–1805
Ellenberger TE, Brandl CJ, Struhl K, Harrison SC (1992) The
GCN4 basic region leucine zipper binds DNA as a dimer of
uninterrupted a helices: crystal structure of the protein-DNA
complex. Cell 71: 1223–1237
Ferl RJ (1996) 14–3-3 proteins and signal transduction. Annu Rev
Plant Physiol Plant Mol Biol 47: 49–73
Greenblatt J (1997) RNA polymerase II holoenzyme and transcriptional regulation. Curr Opin Cell Biol 9:310–319
Grotewold E, Drummond BJ, Bowen B, Peterson T (1994) The
myb-homologous P gene controls phlobaphene pigmentation in
maize floral organs by directly activating a flavonoid biosynthetic gene subset. Cell 76: 543–553
Guiltinan MJ, Marcotte Jr WR, Quatrano RS (1990) A plant
leucine zipper protein that recognizes an abscisic acid response
element. Science 250: 267–271
Hansen S, Takada S, Jacobsen RH, Lis JT, Tjian R (1997) Transcription properties of a cell type-specific TATA-binding protein, TRF. Cell 91: 71–83
Hardtke CS, Berleth T (1998) The Arabidopsis gene
MONOPTEROS encodes a transcription factor mediating embryo axis formation and vascular development. EMBO J 17:
1405–1411
Hill A, Nantel A, Rock CD, Quatrano RS (1996) A conserved
domain of the viviparous-1 gene product enhances the DNA
binding activity of the bZIP protein EmBP-1 and other transcription factors. J Biol Chem 271: 3366–3374
Hoecker U, Vasil IK, McCarty DR (1995) Integrated control of
seed maturation and germination programs by activator and
repressor functions of Viviparous1 of maize. Genes Dev 9: 2459–
2469
Hurst HC (1995) Transcription Factors. 1. bZIP Proteins. Protein
Profile 2: 105–168
Jaglo-Ottosen, KR, Gilmour SJ, Zarka DG, Schabenberger O,
Thomashow MF (1998) Arabidopsis CBF1 overexpression induces COR genes and enhances freezing tolerance. Science 280:
104–106
Kao C-Y, Cocciolone SM, Vasil IK, McCarty DR (1996) Localization and interaction of the cis-acting elements for abscisic acid,
VIVIPAROUS1, and light activation of the C1 gene of maize.
Plant Cell 8: 1171–1179
Kim TK, Maniatis T (1997) The mechanism of transcriptional
synergy of an in vitro assembled interferon-b enhanceosome.
Mol Cell 1: 119–129
Kuo MH, Zhou JX, Jambeck P, Churchill MEA, Allis CD (1998)
Histone acetyltransferase activity of yeast Gcn5p is required for
the activation of target genes in vivo. Genes Dev 12: 627–639
Lefstin JA, Yamamoto KR (1998) Allosteric effects of DNA on
transcriptional regulators. Nature 392: 885–888
Lesnick ML, Chandler VL (1998) Activation of the maize anthocyanin gene a2 is mediated by an element conserved in many
anthocyanin promoters. Plant Physiol 117: 437–445
Lloyd AM, Walbolt V, Davis RW (1992) Arabidopsis and Nicotiana
anthocyanin production activated by maize regulators R and C1.
Science 258: 1173–1175
Ludwig SR, Wessler SR (1990) Maize R gene family: tissue specific
helix-loop-helix proteins. Cell 62: 849–851
Lusser JA, Brosch G, Loidl A, Haas H, Loidl P (1997) Identification of maize histone deacetylase HD2 as an acidic nucleolar
phosphoprotein. Science 227: 88–91
McCarty DR, Hattori T, Carson CB, Vasil V, Lazar M, Vasil IK
(1991) The Viviparous1 developmental gene of maize encodes a
novel transcriptional activator. Cell 66: 895–905
Menkens AE, Schindler U, Cashmore AR (1995) The G-box: a
ubiquitous regulatory DNA element in plants bound by the GBF
family of bZIP proteins. Trends Biochem Sci 20: 506–510
Mizzen CA, Yang X-Y, Kokubo T, Brownell JE, Banister AJ,
Owen-Hughes T, Workman J, Wang L, Berger SL, Kouzarides
1119
T, and others (1996) The TAFII250 subunit of TFIID has histone
acetyltransferase activity. Cell 87: 1261–1270
Mol J, Grotewold E, Koes R (1998) How genes paint flowers and
seeds. Trends Plant Sci 3: 212–217
Mol J, Jenkins G, Schafer E, Weiss D (1996) Signal perception,
transduction, and gene expression involved in anthocyanin biosynthesis. Crit Rev Plant Sci 15: 525–557
Nakagawa H, Ohmiya K, Hattori T (1996) A rice bZIP protein
designated OSBZ8 is rapidly induced by abscisic acid. Plant J 9:
217–227
Neigeborn L, Carlson M (1984) Genes affecting the regulation of
SUC2 gene expression by glucose repression in Saccharomyces
cerevisiae. Genetics 108: 845–858
Ni M, Dehesh K, Tepperman JM, Quail PH (1996) GT-2: In vivo
transcriptional activation activity and definition of novel twin
DNA binding domains with reciprocal target sequence selectivity. Plant Cell 8: 1041–1059
Ogryzko VV, Schiltz RL, Russanova V, Howard BH, Nakatani Y
(1996) The transcriptional coactivators p300 and CBP are histone
acetyltransferases. Cell 87: 953–959
Okamuro JK, Caster B, Villarroel R, Van Montagu M, Jofuku KD
(1997) The AP2 domain of APETALA2 defines a large new family
of DNA binding proteins in Arabidopsis. Proc Natl Acad Sci USA
94: 7076–7081
Paz-Ares J, Ghosal D, Wienand U, Peterson PA, Saedler H (1987)
The regulatory c1 locus of Zea mays encodes a protein with
homology to myb proto-oncogene products and with structural
similarities to transcriptional activators. EMBO J 6: 3553–3558
Ptashne M, Gann A (1997) Transcriptional activation by recruitment. Nature 386: 569–577
Roeder RG (1996) The role of general initiation factors in transcription by RNA polymerase II. Trends Biochem Sci 9: 327–335
Romero I, Fuertes A, Benito MJ, Malpica JM, Leyva A, Paz-Ares
J (1998) More than 80R2R3-MYB regulatory genes in the genome
of Arabidopsis thaliana. Plant J 14: 273–284
Sainz MB, Grotewold E, Chandler VL (1997) Evidence for direct
activation of an anthocyanin promoter by the maize C1 protein
and comparison of DNA binding by related Myb domain proteins. Plant Cell 9: 611–625
Shen Q, Ho T-HD (1995) Functional dissection of an abscisic acid
(ABA)-inducible gene reveals two independent ABA-responsive
complexes each containing a G-box and a novel cis-acting element. Plant Cell 7: 295–307
Shen Q, Ho T-HD (1997) Promoter switches specific for abscisic
acid (ABA)-induced gene expression in cereals. Physiol Plant
101: 653–664
Shen Q, Zhang P, Ho T-HD (1996) Modular nature of abscisic acid
(ABA) response complexes: composite promoter units that are
necessary and sufficient for ABA induction of gene expression in
barley. Plant Cell 8: 1107–1119
Shen WC, Green MR (1997) Yeast TAF(II)145 functions as a core
promoter selectivity factor, not a general coactivator. Cell 90:
615–624
Shultz TF, Medina J, Hill A, Quatrano RS (1998) 14–3-3 proteins
are part of an abscisic acid-viviparous1 (VP1) response complex
in the Em promoter and interact with VP1 and EmBP1. Plant
Cell 10: 837–847
Stern MJ, Jensen RE, Herskowitz I (1984) Five SWI genes are
required for expression of the HO gene in yeast. J Mol Biol 178:
853–868
Stockinger EJ, Glimour SJ, Thomashow MF (1997) Arabidopsis
thaliana CBF1 encodes an AP2 domain-containing transcription
activator that binds to the C-repeat/DRE, a cis-acting DNA
regulatory element that stimulates transcription in response to
low temperature and water deficit. Proc Natl Acad Sci USA 94:
1035–1040
Struhl K (1998) Histone acetylation and transcriptional regulatory
mechanisms. Genes Dev 12: 599–606
Suzuki M, Kao CY, McCarty DR (1997) The conserved B3 domain
of VIVIPAROUS1 has a cooperative DNA binding activity. Plant
Cell 9: 799–807
1120
Thanos D, Maniatis T (1995) Virus inductions of human IFNb
gene expression requires the assembly of an enhanceosome. Cell
83: 1091–1100
Tsukiyama T, Wu C (1996) Purification and properties of an
ATP-dependent nucleosome remodeling factor. Cell 83: 1011–
1020
Ulmasov T, Hagen G, Guilfoyle TJ (1997) ARF1, a transcription factor that binds to auxin response elements. Science 276:
1865–1868
Vasil V, Marcotte WR, Jr, Rosenkrans L, Cocciolone SM, Vasil
IK, Quatrano RS, McCarty DR (1995) Overlap of viviparous
(VP1) and abscisic acid response elements in the Em promoter:
G box elements are sufficient but not necessary for VP1 transactivation. Plant Cell 7: 1511–1518
Singh
Plant Physiol. Vol. 118, 1998
Verrijzer CP, Tjian, R (1996) TAFs mediate transcriptional activation and promoter selectivity. Trends Biochem Sci 9: 338–342
Vicente-Carbajosa J, Moose SP Parsons RL, Schmidt RJ (1997)
A maize zinc-finger protein binds the prolamin box in zein gene
promoters and interacts with the basic leucine zipper transcriptional activator Opaque2. Proc Natl Acad Sci USA 94: 7685–7690
Vicente-Carbajosa J, Onate L, Lara P, Diaz I, Carbonero P (1998)
Barley Blz1: a bZIP transcriptional activator that interacts with
endosperm-specific gene promoters. Plant J 13: 629–640
Wieczorek E, Brand M, Jacq X, Tora L (1998) Function of TAFIIcontaining complex without TBP in transcription by RNA polymerase II. Nature 393: 187–191
Yanagisawa S (1996) Dof DNA-binding proteins contain a novel
zinc finger motif. Trends Plant Sci 1: 213