Chap. 7 Transcriptional Control of Gene
Expression (Part C)
Topics
• Molecular Mechanisms of Transcription Repression and Activation
• Regulation of Transcription Factor Activity
Goals
• Learn how activators &
repressors control chromatin
structure and pre-initiation
complex assembly.
• Learn how the activities of
activators and repressors are
controlled.
Transcriptionally active polytene
chromosomes
Overview of Activator
& Repressor Function
Activators and repressors
control gene expression in
three general ways. First,
these TFs modulate chromatin
condensation state, and
thereby the access of general
TFs and RNA Pol II to
promoters (Fig. 7.2, top).
Second, these TFs interact
with co-activators such as
mediator complex, which helps
RNA Pol II bind to promoters
(Fig. 7.2, middle). Third, TFs
interact with elongation
factors, stimulating Pol II
elongation away from the
promoter region (Fig. 7.2,
bottom).
Co-activators & Co-repressors
The distinction between an activator/co-activator and
repressor/co-repressor is based on whether or not the protein
binds specifically to DNA. Namely, activators/repressors have
DNA binding domains that allow them to bind to DNA. Coactivators/co-repressors typically don’t bind to specific
sequences in DNA. They typically exert their effects on
transcription initiation via protein-protein interactions within
transcription initiation complexes at promoters, or by
modifying histone tails. Most of our discussion about activator
and repressor functions will center on yeast as less currently
is known about their functions in higher eukaryotes.
Interphase Chromatin
Interphase chromatin exists in two
different condensation states (Fig.
6.33a). Heterochromatin is a
condensed form that has a
condensation state similar to
chromatin found in metaphase
chromosomes. Euchromatin is
considerably less condensed.
Heterochromatin typically is found
at centromere and telomere
regions, which remain relatively
condensed during interphase. The
inactivated copy of the Xchromosome (Barr body) that
occurs in cells in females also
occurs as heterochromatin. In
contrast, most transcribed genes
are located in regions of
euchromatin. Common modifications
occurring in histone H3 in heteroand euchromatin are illustrated in
Fig. 6.33b.
Formation of
Heterochromatin
The trimethylation of histone H3 at
lysine 9 (H3K9Me3) plays an
important role in promoting chromatin
condensation to heterochromatin (Fig.
6.34a). Trimethylated sites are
bound by heterochromatin protein 1
(HP1) which self-associates and
oligomerizes resulting in
heterochromatin. Heterochromatin
condensation is thought to spread
laterally between “boundary
elements” that mark the ends of
transcriptionally active euchromatin
(Fig. 6.34b). Recruitment of the
H3K9 histone methyl transferase
(HMT) to HP1 sites promotes
heterochromatin spreading by
catalyzing H3 methylation.
Gene Silencing at Yeast Mating-type Loci
The analysis of gene expression at the yeast mating-type loci has
been useful in identifying repressor TFs that cause gene silencing
by promoting chromatin condensation. All yeast, regardless of
mating type have two gene clusters specifying proteins that confer
the a (HMLa) and a (HMRa) mating types (Fig. 7.34). What
determines mating type is whether a copy of the HMLa or HMRa
genes is transferred to the MAT locus where these genes can be
expressed. Silencer sequences flanking the HMLa and HMRa loci
keep the DNA in these regions condensed until at cell division, one
loci or the other is decondensed and becomes accessible for
transfer to the MAT locus.
The Yeast Telomere Silencing Complex
The functions of several proteins
in chromatin condensation at
telomeres (and mating-type
silencing loci) is illustrated in Fig.
7.35d. First, a repressor called
RAP1 binds to the telomere DNA
repeat (and to silencer sequences
near the mating-type loci). RAP1
binding nucleates the assembly of
a co-repressor complex containing
the SIR2, SIR3, and SIR4
proteins. Sir3 and 4 also bind to
hypoacetylated histone H3 and H4
N-terminal tails in adjacent
chromatin. SIR2, a histone
deacetylase, catalyzes further
hypoacetylation. As a result, ~4
kb of DNA at the telomere is
condensed. Higher order
complexes between several
condensed telomeres subsequently
form. Similar condensed regions
are thought to occur at the mating-type loci. As discussed in
Chap. 6, heterochromatin protein 1 (HP1) and H3K9 histone
methyl transferase (HMT) also cause chromatin condensation
(Fig. 6.34).
Repressor-directed Histone Deacetylation
Another protein complex involved in repressor-directed histone
deacetylation and chromatin condensation of yeast genes is shown
in Fig. 7.36a. The UME6 repressor binds to URS1 control
elements and recruits a co-repressor complex containing SIN3 and
RPD3 to these sites. RPD3 is a histone deacetylase, and this
enzyme removes acetyl groups from histones in the vicinity of the
URS1 sequence. The nucleosomes bound to DNA in this region
(which contains a TATA box promoter) subsequently condense, and
expression of the gene is repressed.
Activator-directed Histone Hyperacetylation
Yeast genes can be turned on by histone acetylation and chromatin
decondensation (Fig. 7.36b). In the example shown, the GCN4
activator first binds to its UAS upstream of the TATA box of a
regulated gene. GCN4-UAS binding recruits a co-activator complex
containing the GCN5 histone acetylase to the site. Through the
activity of GCN5, histone N-terminal tails are hyperacetylated
leading to chromatin decondensation and formation of beads-on-a
string type chromatin. General TFs and RNA Pol II are then able
to interact with the promoter, and the gene is transcribed.
Multisubunit proteins called chromatin remodeling complexes also
are important in decondensing chromatin.
Experimental Demonstration of Chromatin
Decondensation in vivo
An experiment demonstrating that some TF activation domains
regulate chromatin condensation is shown in Fig. 7.37. In this
experiment, DNA consisting of a tandemly repeated lac operator
sequence was incorporated into a yeast chromosome. When a
fluorescently tagged wild-type Lac repressor is introduced into the
cell, the DNA is shown to be confined to a small region of the
nucleus (left). However, when the Lac repressor is fused to a yeast
activation domain that interacts with a histone acetylase, staining
spreads throughout a larger volume of the nucleus indicating the
DNA has been decondensed (right).
Mediator Complex
Mediator complex is a
multisubunit co-activator
that helps load RNA Pol II
onto promoters. Some
mediator subunits bind to
RNA Pol II and others bind
to TF activation domains.
The structure of yeast
mediator bound to RNA Pol
II is shown in Fig. 7.38a.
The subunit composition of
human mediator complex is
shown in Fig. 7.38c. Some
mediator subunits exhibit
histone acetylase activity.
Mediator Interactions at Promoters
As illustrated in Fig. 7.39, some mediator subunits bind to RNA
Pol II and TAF subunits in TFIID. Other subunits bind to the
activation domains of TFs bound to promoter proximal elements
and enhancer sequences (not shown). It is not uncommon that a
collection of ~100 proteins is assembled at an active eukaryotic
promoter.
Modular Structure of Activators II
Functional domains in activators are joined by flexible protein
linker sequences (Fig. 7.27). Due to the presence of linkers,
the spacing and location of DNA control elements often can be
shifted without interfering with DNA binding and regulation of
promoters. The evolution of gene control regions through
shuffling of DNA binding sequences between genes may have
been favored due to the lack of strong requirements for control
element spacing and location. Furthermore, the evolution of new
activator protein genes through domain swapping has probably
also been facilitated by linker sequences.
Yeast Two-hybrid System (I)
The flexibility of activatormediated gene transcription in
eukaryotes is exploited in the
widely used yeast two-hybrid
system, which is an in vivo system
for identification of interacting
pairs of proteins. In this method,
genes encoding hybrid proteins
containing yeast DNA-binding and
activation domains are first
constructed (Fig. 7.40a). The
"bait hybrid" contains the GAL4
DNA binding domain fused to the
"bait domain" which is one of the
proteins being tested. The "fish
hybrid" contains the second
protein, the "fish domain”, fused
to a GAL4 activation domain. If
the bait and fish domains can
interact, then the two hybrid
proteins can activate transcription
of a reporter gene (e.g., HIS)
(Fig. 7.40b).
Yeast Two-hybrid System (II)
An entire cDNA
library of genes
can be cloned into
the fish vector to
screen for all
proteins that can
interact with the
bait protein (Fig.
7.40c). Proteins
used as the bait
and fish domains
can come from any
organism.
Regulation of TF Activity
TF activity is controlled in several ways. Methods of control
include 1) regulation of the synthesis level of the TF in tissues
and cells, 2) regulation of TF activity via a signal transduction
pathway coupled to a receptor for a hormone or growth factor,
and 3) direct binding of the TF to certain small molecules, e.g.,
steroid hormones, in the case of nuclear receptors. Examples of
hormones that regulate nuclear receptor activity are shown in
Fig. 7.41.
RAR, retinoic
acid receptor
GR, glucocorticoid
receptor
TR, thyroxine receptor
Modular Structure of Nuclear Receptors
Like most TFs, nuclear receptors have a modular structure
(7.42). They all contain highly conserved DNA-binding domains,
consisting of C4 zinc finger motifs, located near the middle of
the polypeptide chain. The N-terminal region serves as an
activation domain in some receptors. The ligand-binding domain is
located in the C-terminal sequence region. This region serves as
a hormone-dependent activation domain in some receptors, and in
other receptors, as a repression domain in the absence of ligand.
Nuclear Receptor DNA Response Elements
The DNA sequence elements bound by some nuclear receptors are
shown in Fig. 7.43. The conservation of these sequences is
reflected in the structural conservation of the DNA binding
domains. The GR and ER
receptors bind DNA as homodimers
with a two-fold rotational
symmetry. Thus, their binding sites
are organized as inverted repeats
on the two DNA strands. The VDR,
TR, and RAR receptors bind DNA
as heterodimers in which one
subunit is always the RXR TF. The
DNA sequence elements recognized
by these heterodimers is a AGGTCA
direct repeat. The spacing of the
repeats is important in establishing
binding specificity.
Homodimeric Nuclear Receptor Action
In the absence of ligand, the ER and GR receptors are localized
to the cytoplasm where they form a complex with an inhibitor
protein related to the Hsp70 heat-shock chaperone. The
binding of the hormone to the ligand-binding domain releases
the inhibitor and monomeric receptors move to the nucleus (Fig.
7.44). In the nucleus, they dimerize and bind to response
elements in the vicinity of their target genes. The activation
domains of these
receptors promote
chromatin decondensation
via activation of
chromatin remodeling and
histone acetylase
complexes. They further
interact with mediator to
help load RNA Pol II
onto promoters.
Heterodimeric Nuclear Receptor Action
Heterodimeric nuclear receptors, e.g., RXR-VDR, RXR-TR,
and RXR-RAR, are localized exclusively in the nucleus. They
bind to DNA in the absence of ligand, and when bound, cause
chromatin to condense via activation of histone deacetylases.
Upon ligand binding, the ligand-binding domain undergoes a
conformational change. The ligand-bound conformation
stimulates the activity of histone acetylases and also can bind
mediator, stimulating pre-initiation complex assembly at
promoters.