Latchman. Eukariotic Transciption Factors

Подождите немного. Документ загружается.

were able to activate transcription when the DNA binding site was placed

close to the start site of transcription in the promoter region. In contrast,

the glutamine-rich domain was unable to activate transcription when the

binding site was located downstream of the transcription unit mimicking a

position within an enhancer element (see Chapter 1, section 1.3.4). The

acidic domain was strongly active from this enh ancer position while the

proline-rich domain could also activate transcription from this position but

only weakly.

These findings indicate therefore that clear differences exist in the abilities

of the different activation domains to activate transcription when bound to

the DNA at different positions relative to the promoter. Such differences are

likely to be important in determining the functional activity of different fac-

tors. In addition, such differences in the activity of different activation

domains are likely to reflect differences in the mechanisms by which these

factors act. In agreement with this idea, acid ic or proline-rich activation

domains derived from mammalian factors can also activate transcription

when introd uced into yeast cells, whereas glutamine-rich domains cannot

do so (Kinzler et al., 1994).

In the next sections we will consider the mechanisms by which activation

domains act, focusing particularly on the acidic domains where most informa-

tion is available. Similarities and differences in the mode of action of the other

activation domains will be discussed where this information is available.

ACTIVATION OF GENE EXPRESSION BY TRANSCRIPTION FACTORS 141

Figure 5.5

An acidic activation

domain (A) can stimulate

transcription when bound

to DNA in the promoter

(P) close to the

transcriptional start site

or when bound at a

distant enhancer (E). In

contrast, a proline-rich

domain (P) stimulates

only weakly from an

enhancer position and a

glutamine-rich domain

(Q) does not stimulate at

all from this position.

142 EUKARYOTIC TRANSCRIPTION FACTORS

5.3 INTERACTION OF ACTIVATION DOMAINS WITH THE

BASAL TRANSCRIPTIONAL COMPLEX

5.3.1 ACTIVATORS AND THE BASAL TRANS CRIPTIONAL

COMPLEX

The widespread interchangeability of acidic activation domains from yeast ,

Drosophila and mammalian transcription factors discussed above, strongly

suggests that a single common mechanism may mediate transcriptional activa-

tion by acidic activation domains in a wide range of organisms. This idea is

supported by the finding noted above that mammalian transcription factor s

carrying such domains, such as the glucocorticoid receptor, can activate a

gene carrying their appropriate DNA binding site in yeast cells while the

yeast GAL4 factor can do so in cells of Drosophila, tobacco plants and

mammals (reviewed by Guarente, 1988; Ptashne, 1988).

These considerations suggest that the target factor or factors with which

these activators interact is likely to be highly conserved in evolution. A num-

ber of experiments have indicated that in many cases this target factor is likely

to be required for the transcription of a number of different genes and not

solely for that of the activated gene. Thus the over-expression of the yeast

GAL4 protein which contains a strong activation domain results in the down

regulation of genes which lack GAL4-binding sites such as the CYC1 gene as

well as activating genes which do contain GAL4-binding sites. This phenom-

enon, which has been noted for a number of transcription factors with strong

activation domains, is known as squelching (for review see Ptashne, 1988).

Although the degree of squelching by any given factor is proportional to the

strength of its activation domain, squelching differs from activation in that it

does not require DNA binding and can be achieved with truncated factors

containing only the activation domain and lacking the DNA binding domain.

This phenomenon can therefore be explained on the basis that a transcrip-

tional activator, when present in high concentration, can interact with its

target factor in solution as well as on the DNA. If this target factor is present

at limiting concentrations it will therefore be sequestered away from other

genes that require it for transcription, resulting in their inhibition (Fig. 5.6).

The existence of squelching indicates therefore that in many cases the

target factor for activation domains is likely to be a component that is

required for the transcription of a wide range of genes and which is conserved

from yeast to mammals allowing yeast activators to work in mammalian cells

and vice versa. Obviously, such a common component could be part of the

basal transcriptional complex required for transcription of a wide range of

genes in different organisms. Clearly, an activating factor could act by stimu-

lating the binding of such a component so that the basal complex assembled

more efficiently. Alternatively, it could act by intera cting with a factor which

had already bound so that the activity or stability of the assembled complex

was stim ulated. It appears that both these mechanisms are used and they will

be discussed in turn.

5.3.2 STIMULATION OF FACTOR BINDING

As described in Chapter 3 (section 3.5.1) the basal transcriptional complex

can assemble in a stepwise manner with the binding of TFIID being followed

by the binding of TFIIB and then the binding of RNA polymerase in associa-

tion with TFIIF. Clearly an activator could increase the rate of complex assem-

bly by enhancing any one of these assembly steps. Indeed, there is evidence

that activators target several of these steps in the assembly process (Fig. 5.7).

Thus, for example, it appears that acidic activators interact directly with TFIID

(see Chapter 3, section 3.5.1) to stimulate the binding of TFIID to the pro-

moter (Fig. 5.7a). Interestingly, this enhanced recruitment of TFIID, induced

by transcriptional activators, which was initially observed in the test tube, has

been confirmed in intact cells using the ChIP assay described in Chapter 2

(section 2.4.3) (Kuras and Struhl, 1999; Li et al., 1999).

ACTIVATION OF GENE EXPRESSION BY TRANSCRIPTION FACTORS 143

Figure 5.6

The process of

squelching. In the normal

case, illustrated in panel

(a), two distinct activator

molecules A and B

involved in the activation

of genes one and two

respectively, both act by

interacting with the

general transcription

factor T and both genes

are transcribed. In

squelching, illustrated in

panel (b), factor A is

present at high

concentration and hence

interacts with T both on

gene one and in solution.

Hence factor T is not

available for transcription

of gene two and

therefore only gene one

is transcribed while

transcription of gene two

is squelched.

Although increased binding of TFIID to the promoter will directly enhance

the assembly of the complex by allowing TFIIB to bind, there is evidence that

activators can also act directly to improve the recruitment of TFIIB indepen-

dent of their effect on TFIID (Fig. 5.7b). Thus it has been shown that both an

acidic activator and glutamine or proline-rich activators can greatly stimulate

the binding of TFIIB to the promoter (Choy and Green, 1993). Hence activa-

tors can enhance the assembly of the basal transcriptional complex by inde-

pendently enhancing the binding of both TFIID and TFIIB. This ability of

activators to act at these two independent steps results in a strong synergistic

activation of transcription in the presence of different activators targeting

either TFIID or TFIIB (Gonzalez-Couto et al., 1997).

As with TFIID, it has been shown that TFIIB interacts directly with activat-

ing molecules. Thus TFIIB can be purified on a column containing a bound

144 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 5.7

The binding of an

activating molecule (A) to

its binding site (ABS) can

enhance both the binding

of TFIID to the TATA box

(a) and the recruitment of

the TFIIB factor (b) so

enhancing the rate of

basal complex assembly

and of transcription.

acidic activator and interactions of TFIIB with non-acidic activators have also

been reported. Moreover, mutations in the activator that abolish this interac-

tion with TFIIB prevent it from activating transcription (for review see Hahn,

1993b). Thus the effect of activators on TFIIB is mediated via a di rect protein–

protein interaction which is essential for their ability to stimulate transcrip-

tion.

In addition to the stepwise pathway of complex assembly, it has also been

proposed that the basal transcriptional complex can assemble in a much

simpler manner with binding of TFIID being followed by binding of the

RNA polymerase holoenzyme which contains the polymerase itself, TFIIB,

TFIIF and TFIIH as well as a number of other proteins (see Chapter 3, section

3.5.2). There is evidence that activators can also act in this pathway not only by

enhancing the rec ruitment of TFIID as described above but also by directly

enhancing the binding of the RNA polymerase holoenzyme itself (Fig. 5.8a).

Thus, for example, if a DNA binding domain is linked to the yeast protein

Gal11, which is a component of the RNA polymerase holoenzyme, the holoen -

zyme is recru ited to the DNA via this DNA binding domain and transcription

is activated (Fig. 5.8b) (Barberis et al., 1995). Hence the need for activators can

be bypassed by recruiting the RNA polymerase holoenzyme to DNA via an

artificial DNA binding domain.

ACTIVATION OF GENE EXPRESSION BY TRANSCRIPTION FACTORS 145

Figure 5.8

(a) Activators can act by

enhancing the binding of

the RNA polymerase

holoenzyme complex

following binding of

TFIID. (b) In agreement

with this idea, the need

for an activator can be

bypassed by artificially

attaching a DNA binding

domain to the Gal 11

component of the

holoenzyme so enhancing

holoenzyme recruitment

by allowing it to bind to

DNA directly.

Indeed, on the basis of experiments of this type, Ptashne and Gann (1997)

have argued that the sole role of activators is to enhance the assembly of the

basal complex by interacting with one or other of its specific components so

facilitating their recruitment to the DNA. However, while suc h enhanced

recruitment of specific components of the complex clearly plays a major

part in the action of transcriptional activators and operates in both pathways

of complex assembl y, it is likely that other effects are also involved in the

action of transcriptional activators. These effects are di scussed in the next

section.

5.3.3 STIMULATION OF FACTOR ACTIVITY

In addition to their effects on complex assembly, it is clear that activators can

also stimulate transcription at a subsequent step following assembly of the

complex, resulting in its enhanced stability or increased activity (Choy and

Green, 1993) (Fig. 5.9).

146 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 5.9

An activator can stimulate

transcription both by

promoting the assembly

of the basal transcription

complex and by

stimulating its activity

following assembly.

An obvious m echanism for activation would be for activating domains to

interact directly with the RNA polymerase itself to increase its activity (Fig.

5.10a). As discussed in Chapter 3 (section 3.1), the largest subunit of RNA

polymerase II contains at its C terminus multiple copies of a sequence whose

consensus is Tyr-Ser-Pro-Thr-Ser-Pro-Ser, which is highly conserved in evolu-

tion and is essential for its function. This motif is very rich in hydroxyl groups

and lack s negatively charged acidic residues.

It has therefore been suggested that this motif could interact directly either

with a negatively charged acidic domain or via hydrogen bonding with amide

groups in a glutamine-rich activation domain. This woul d provide a mechan-

ism for direct interaction between activating domains and RNA polymerase

itself while the evol utionary conservation of the target regio n within the poly-

merase would explain why yeast activators work in mammalian cells and vice

versa. In agreement with this idea it has been shown that yeast mutants

containing a reduced number of copies of the heptapeptide repeat in RNA

polymerase II are defective in their response to activators such as GAL4.

Although these results are consistent with a direct interaction between

transcriptional activators and the RNA polymerase, they are equally consistent

with an indirect interaction in which the activator contacts another compo-

nent of the transcriptional machinery which then interacts with the repeated

motif in the polymerase. Indeed, despite the attractiveness of a model invol-

ving direct interaction between activating factors and the polymerase itself,

ACTIVATION OF GENE EXPRESSION BY TRANSCRIPTION FACTORS 147

Figure 5.10

Two possible

mechanisms by which an

activating factor (A) could

stimulate the activity of

the basal transcriptional

complex. This could

occur via direct

interaction with the RNA

polymerase itself (a) or

by interaction with

another transcription

factor such as TFIID

which in turn interacts

with the polymerase (b).

it is unlikely to be correct and it appears that activators interact with the

polymerase indirectly via other factors (Fig. 5.10b).

TFIID is one potential candidate for the component with which activating

factors interact since this factor is both required for the transcription of a

wide variety of genes both with and withou t TATA boxes (see Chapter 3,

section 3.6) and is highly conserved in evolution, with the yeast factor being

able to promote transcription in mammalian cell extracts and vice versa.

Evidence for an effect of activating factors on TFIID has been obtained in

the case of the yeast acidic activating factor GAL4 (Horikoshi et al., 1988).

Thus, in the absence of GAL4, TFIID was shown to be bound only at the

TATA box of a promoter containing both a TATA box and GAL4 binding

sites. In contrast, in the presence of GAL4 bound to its upstream binding sites

in the promoter, the conformation of TFIID was altered such that it now

covered both the TATA box and the start site for transcription (Fig. 5.11).

Moreover, no change in TFIID conformation was observed in the presence of

a truncated GAL4 molecule which can bind to DNA but lacks the acidic

activation domain. Hence, an acidic activator can produce a change in

TFIID conformation resulting in its binding to the start site for transcription

and this effect correlates with the ability of GAL4 to activate transcription

rather than being a consequence of its binding to DNA. It is clear therefore

that activating molecules can alter the configuration of TFIID bound to the

promoter by interacting with it.

As well as interacting with TFIID to change its configuration, activators can

also interact with TFIIB changing its conformation and enhancing its ability to

recruit the complex of RNA polymerase II and TFIIF (Roberts and Green,

148 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 5.11

Effect of GAL4 binding

on the binding of TFIID.

1994; Fig. 5.12). Hence activators appear to target both TFIID and TFIIB in

two ways. First, as described in the previous section, they enhance their bind-

ing to the promoter and secondly, they alter their conformation so as to

enhance their activity (Fig. 5.13).

Together with TFIIB, TFIID constitutes a major target for transcriptional

activators. Interestingly, however, other components of the basal complex

such as TFIIA (Ozer et al., 1994), TFIIF (Joliot et al., 1995) and TFIIH (Xiao

et al., 1994) have also been shown to interact with transcriptional activators.

Hence a number of different factors within the basal transcriptional complex

serve as targets for direct interactions with transcriptional activators. It is

clear, however, that in many cases, acti vators interact with the basal complex

only indirectly via other factors and suc h interactions are discussed in the next

section.

5.4 INTERACTION OF ACTIVATION DOMAINS WITH OTHER

REGULATORY PROTEINS

5.4.1 THE MEDIATOR COMPLEX

As noted in section 5.3.1, the existence of the squelching phenomenon indi-

cates that activators act by contacting a fact or which is involved in the tran-

scription of a wide range of genes. Although this could be a component of the

basal transcriptional complex (see section 5.3) studi es in yeast resulted in the

purification of a multi-protein complex (distinct from the basal transcriptional

ACTIVATION OF GENE EXPRESSION BY TRANSCRIPTION FACTORS 149

Figure 5.12

The binding of an acidic

activator (A) to TFIIB

produces a

conformational change

which enhances the

ability of TFIIB to interact

with the RNA

polymerase/TFIIF

complex, thereby

enhancing its recruitment

to the promoter.

complex) which could prevent squelching when added in excess. This so-

called ‘mediator’ complex therefore represents a target for transcriptional

activators which is present in limiting amounts so that activators compete

for it. Hence, its addition in excess relieves this competition and prevents

squelching.

The mediator complex consists of over twenty proteins and, following its

original identification in yeast , has now been found in a wide range of multi-

cellular organisms including humans. It therefore appears to be a conserved

component of the transcriptional machinery involved in activation of a wide

range of genes (for reviews see Malik and Roeder, 2000; Myers and Kornberg,

2000; Boube et al., 2002).

As well as interacting with activators, the mediator also interacts with RNA

polymerase II itself. Indeed, the mediator is part of the RNA polymerase

150 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 5.13

Activators can stimulate

both the binding of TFIIB

and TFIID and enhance

their activity by altering

their conformation

(square to circle).