Latchman. Eukariotic Transciption Factors

Подождите немного. Документ загружается.

continuous elevation of these proteins, suc h as would occur when cells

become infected with a retrovi rus expressing one of them, would result in

cells which exhibited continuous uncontrolled growth and were not subject to

normal growth regulatory signals. Since such uncontrolled growth is one of

the characteristics of cancer cells it is relatively easy to link the rol e of Fos and

Jun in inducing genes required for growth with their ability to cause cancer.

Normally, however, the transformation of a cell to a transformed cancerous

phenotype requires more than simply its conversion to a continuously grow-

ing immortal cell (for review see Land et al., 1983). Since repeated treatments

with phorbol esters can promote tumour formation in immortalized cells, the

prolonged induction of phorbol ester responsive genes by elevated levels of

Fos and Jun may therefore result in the conversion of already continuously

growing cells into the tumorigenic phenotype characteristic of cancer cells

(Fig. 9.13).

Hence the ability of Fos and Jun to cause cancer represents an aspect of

their ability to induce transcription of specific cellular genes. In agreement

with this idea, mutations in Fos which abolish its ability to dimerize with Jun

and hence prevent it from binding to AP1 sites also abolish its ability to

transform cells to a cancerous phenotype. It should be noted, however, that

302 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 9.12

Growth factor stimulation

of cells results in

increased transcription of

the c-fos and c-jun

genes which in turn

stimulates transcription of

genes which are activated

by the Fos-Jun complex.

in addition to their over-expression within a retrovirus, there is also some

evidence that mutational changes render the viral proteins more potent tran-

scriptional activators than the equivalent cellular proteins. Thus the v-Jun

protein appears to activate transcri ption more efficiently than c-Jun due to

a deletion in a region which is involved in targeting the c-Jun protein for

degradation (Treier et al., 1994) and which also mediates its interaction

with a negatively acting cellular factor (Baichwal and Tjian, 1990).

Interestingly, in addition to its central role in the growth response, the Fos,

Jun, AP1 system also appears to represent a target for other oncogenes. Thus,

for example, the ets oncogene which, like fos and jun encodes a cellular tran-

scription factor, acts via a DNA binding site known as PEA3 which is located

adjacent to the AP1 site in a number of TPA-responsive genes such as col-

lagenase and stromelysin. Moreover, the Ets protein cooperates with Fos and

Jun to produce high level activation of these promoters (Wasylyk et al., 1990).

In addition to interacting positively with other factors, the Fos/Jun com-

plex can also inhibit the action of other transcription factors. Thus, as

described in Chapter 6 (section 6.5), the Fos/Jun complex requires the CBP

co-activator in order to activate transcription. It therefore competes with the

TRANSCRIPTION FACTORS AND HUMAN DISEASE 303

Figure 9.13

Effects of Fos and Jun

on cellular growth. In

normal cells (a) a brief

exposure to a growth

signal or phorbol ester

will lead to a brief period

of growth via the

transient induction of Fos

and Jun and hence of

Fos/Jun-dependent

genes. In contrast the

continuous elevation of

Fos and Jun produced

for example by infection

with a retrovirus

expressing Fos or Jun

results in continuous

unlimited growth and

cellular transformation

(b).

activated glucocorticoid receptor for CBP hence preventing the receptor from

activating transcription. Similarly both Fos and Jun can inhibit the activation

of muscle specific promoters by the MyoD transcription factor (see Chap ter 7,

section 7.2.1) thereby preventing cells from differentiating into non-dividing

muscle cells and allowing cellular proliferation to continue (Li et al., 1992).

Hence the Fos and Jun oncogene products play a critical role in the regula-

tion of specific cellular genes in normal cells, interacting with the products of

other transcription factors to produce the controlled activity of their target

genes necessary for normal controlled growth.

9.3.2 v-erbA AND THE THYROID HORMONE RECEPTOR

The v-erbA oncogene is one of two oncogenes carried by avian erythroblastosis

virus (AEV). The cellular equivalent of this oncogene c-erbA, has been shown

to encode the cellular receptor for thyroid hormone (Sap et al., 1986;

Weinberger et al., 1986) which is a member of the steroid/thyroid hormone

receptor super family discussed in Chapter 4 (section 4.4). Following the

binding of thyroid hormone, the receptor/hormone complex binds to its

appropriate recognition site in the DNA of thyroid hormone responsive

genes and activates their transcription (Fig. 9.14).

304 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 9.14

The c-erbA gene

encodes the thyroid

hormone receptor and

activates transcription in

response to thyroid

hormone.

Hence the protein encoded by the c-erbA gene represents a bona fide cellular

transcription factor involved in the acti vation of thyroid hormone responsive

genes. Unlike the case of the fos and jun gene products, which regulate genes

involved in growth, it is not immediately obvi ous how the form of thyroid

hormone receptor encoded by the viral v-erbA gene can transform cells to a

cancerous phenotype.

The solution to this problem is provided by a comparison of the cellular

ErbA protein, which is a functio nal thyroid hormone receptor, and the viral

ErbA protein encoded by AEV. Thus, in addition to being fused to the retro-

viral gag protein at its N terminus, the viral ErbA protein contains several

mutations in the reg ions of the receptor responsible for binding to DNA and

for binding thyroid hormone as well as a small deletion in the hormone

binding domain (Fig. 9.15).

Interestingly, although these changes do not abolish the ability of the viral

ErbA protein to bind to DNA, they do prevent it from binding thyroid hor-

mone and thereby becoming converted to a form which can activate transcrip-

tion (Sap et al., 1986, 1989). However, these changes do not affect the

inhibitory domain which, as discussed in Chapter 6 (section 6.3.2), allows

the thyroid hormone receptor to repress transcription. Hence, the viral v-

ErbA protein can inhibit the induction of thyroid hormone responsive

genes when cells are treated with thyroid hormone by binding to the thyroid

hormone response elements in their promoters and dominantly repressing

their transcription, as well as preventing binding of the activating complex of

thyroid horm one and the cellular ErbA protein (Fig. 9.16). In agreement with

this critical role for repression in producing transformation by v-ErbA, a

mutation in v-ErbA which abolishes its ability to repress transcription by pre-

venting it binding its co-repressor (see Chapter 6, section 6.3.2) also abolishes

its ability to transform cells (Perlmann and Vennstrom, 1995).

Hence the viral ErbA protein acts as a dominant repressor of thyroid

hormone responsive genes being both incapable of activating transcription

itself and able to prevent activation by intact receptor. This mechanism of

TRANSCRIPTION FACTORS AND HUMAN DISEASE 305

Figure 9.15

Relationship of the

cellular ErbA protein and

the viral protein. The

black dots indicate single

amino acid differences

between the two proteins

while the arrow indicates

the region where nine

amino acids are deleted

in the viral protein.

action is clearly similar to the repression of thyroid hormone responsive genes

by the naturally occurring alternatively spliced form of the thyroid hormone

receptor which, as discussed in Chapter 6 (section 6.3.2), lacks the hormone

binding domain and therefore cannot bind hormone. Thus the same mechan-

ism of gene repression by a non-hormone binding receptor is used naturally

in the cell and by an oncogenic virus.

One of the targets for repression by the viral ErbA protein is the erythro-

cyte anion transporter gene (Zenke et al., 1988), which is one of the genes

normally induced when avian erythroblasts differentiate into erythrocytes.

This differentiation process has been known for some time to be inhibited

by the ErbA protein and it is now clear that it achieves this effect by blocking

the induction of the genes needed for differenti ation. In turn such inhibition

will allow continued prolifera tion of thes e cells rendering them susceptible to

transformation into a tumour cell type by the product of the other AEV

oncogene v-erbB which encodes a truncated form of the epidermal growth

factor receptor (Downward et al., 1984) and therefore renders cell growth

independent of external growth factors (Fig. 9.17).

The two cases of Fos/Jun and ErbA therefore represent contrasting exam-

ples of the involvement of transcription factors in oncogenesis both in terms

of the mechanism of transformation and the manner in which the cellular

form of the oncogene becomes an active transforming gene. Thus, in the case

306 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 9.16

Inhibitory effect of the

viral ErbA protein on gene

activation by the cellular

protein, in response to

thyroid hormone. The viral

protein both inhibits

binding by the activated

c-ErbA protein and also

dominantly represses

transcription by means of

its inhibitory domain.

Note the similarity to the

action of the alpha-2

form of the c-ErbA

protein, illustrated in

Figure 6.14.

of Fos and Jun, transformation is achieved by the continuous activation of

genes necessary for growth in normal cell types. Moreover, it occurs, at least

in part, via the natural activity of the cellular oncogene in inducing these

genes being enhanced by their over-expression such that it occurs at an inap-

propriate time or place (Fig. 9.18a). In contrast in the ErbA case transforma-

tion is achieved by inhibiting the expression of genes whose products are

required for the differentiati on of a particular cell type therefore allowing

growth to continue. Moreover, this occurs via the activity of a mutated form

of the transcription factor which, rather than carrying out its normal function

more efficiently, actually interferes with the normal role of the thyroid hor-

mone receptor in inducing thyroid hormone responsive genes required for

differentiation (Fig. 9.18b).

9.3.3 THE myc ONCOGENE

Interestingly, for a considerable period, the techniques of molecular biology

failed in the case of the c-myc oncogene, which was one of the earliest cellular

TRANSCRIPTION FACTORS AND HUMAN DISEASE 307

Figure 9.17

Inhibition of erythrocyte-

specific gene expression

by the v-ErbA protein

prevents erythrocyte

differentiation and allows

transformation by the

v-ErbB protein.

oncogenes to be identified, with its expression being dramatically increased in

a wide variety of transformed cells (for review see Grandori et al., 2000;

Eisenman, 2001). Thus the Myc protein has a number of properties suggesting

that it is a transcription factor, notably nuclear localization, the possess ion of

several motifs characteristic of transcription factors such as the helix-loop-

308 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 9.18

Transformation

mechanisms of Fos/Jun

(panel a) and ErbA (panel

b). Note that Fos/Jun-

induced transformation

occurs because the

proteins induce the

continual activation of

growth regulatory genes

which are normally

expressed only transiently

while v-ErbA-induced

transformation occurs

because the protein

interferes with the action

of its cellular homologue

and hence inhibits the

induction of genes

involved in erythrocyte

differentiation.

helix and leucine zipper elements (see Chapter 4, section 4.5) and the ability

to activate target promoters in co-transfection assays. Despite exhaustive

efforts, however, no DNA sequence to which the Myc protein binds could

be defined, rendering its mechanism of action uncertain.

The solution to this problem was provided by the work of Blackwood and

Eisenman (1991) who identified a novel protein, Max, which can form hetero-

dimers with the Myc protein via the helix-loop-helix motif present in both

proteins. Myc/Max heterodimers can bind to DNA and regulate transcription,

whereas Myc/Myc homodimers cannot do so (for reviews see Grandori et al.,

2000; Baudino and Cleveland, 2001) (Fig. 9.19). This effect evidently parallels

the requirement of the Fos protein for dimerization with Jun in order to bind

with high affinity to AP1 sites (see section 9.3.1).

The Max protein therefore plays a critical role in allowing the DNA binding

of Myc and the structure of a Myc/Max heterodim er bound to DNA has

recently been defined (Nair and Burley, 2003). Moreover, the ability to inter-

act with Max, bind to DNA and modulate gene expression is critical for the

ability of the Myc protein to transform since mutations in Myc which abolish

its ability to heterodimerize with Max also abolish its transf orming ability.

Hence, as was previously speculated, the Myc protein is a transcription factor

whose over-expression causes transformation, presumably via the activation of

genes whose protein products are required for cellular growth (for reviews see

Zornig and Evan, 1996; Grandori and Eisenman, 1997; Levens, 2002).

Interestingly, the Max protein does not appear to represent a passive part-

ner whi ch merely serves to del iver Myc to the DNA of target genes. Rather it

plays a key role in regulating the activity of target gen es containing the appro-

priate bin ding site. Thus, it has been shown that, whereas Myc/Max hetero-

dimers can activate transcription, Max/Max homodimers can bind to the

same site and weakly repress transcription. Moreover, Max can also hetero-

dimerize with another member of the helix-loop-helix family, known as Mad,

to form a strong repressor of transcription (for review see Bernards, 1995)

(Fig. 9.20).

TRANSCRIPTION FACTORS AND HUMAN DISEASE 309

Figure 9.19

Both Myc/Max

heterodimers and Max/

Max homodimers can

bind to DNA whereas

Myc/Myc homodimers

cannot.

The Max/Max homodimer appears to act as a weak repressor simply by

preventing the Myc/Max activator from binding to its appropriate binding

sites and thereby preventing it from activating transcription. In contrast, the

Mad/Max heterodimer appears to act as an active repressor which is capable

of reducing transcription below that which would be observed in the absence

of any activator binding (see Chapter 6, for discussion of the mechanisms of

transcriptional repression). Thus, it has been shown that the Mad protein can

bind the same complex of N-CoR, mSIN-3 and mRPD3, which mediates active

repression by nuclear receptors suc h as the thyroid hormone receptor in the

absence of hormone (see Chapter 6, section 6.3.2) (for review see Wolffe,

1997). As this complex includes the mRPD3 protein, which has histone de-

acetylase activity, it is possible that the Mad/Max heterodimer may repress

transcription, at least in part, by recruiting a complex which deacetylates

histones thereby organizing a more tightly packed chromatin structure

(Fig. 9.21).

In the case of the nuclear receptors, the switch from the repressed state of

target genes to their activation is mediated by the addition of hormone. In

contrast, however, in the case of the My c family it is mediated by signals which

produce a rise in Myc expression and a corre sponding fall in the expression of

Mad. Thus Myc is expressed at very low levels in resting cells and its expres-

sion is induced when cells begin to grow, whereas Max is expressed at similar

high levels in both resting and proliferating cells and Mad is expressed at high

levels only in resting cells and not in proliferating cells. Hence in resting cells

310 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 9.20

Functional effects of

Max/Max homodimers

and of Myc/Max or Mad/

Max heterodimers. Note

that Max/Max

homodimers repress

transcription only weakly

by passively blocking

activator binding, whereas

Mad/Max heterodimers

actively repress

transcription and

therefore have a much

stronger effect.

Figure 9.21

The Max/Mad

heterodimer can recruit a

co-repressor complex

(Co-R) with histone

deacetylase activity which

can produce a more

tightly packed chromatin

structure (compare with

Fig. 6.26).

Mad and not Myc will be expressed and the expression of Myc dependent

genes will be repressed by Mad/Max homodimers. In contrast expression will

be activated by Myc/Max hete rodimers as the cells receive signals to prolifer-

ate resulting in increased Myc expression and decreased Mad expression (Fig.

9.22). Clearly the over-expression of the Myc gene, which is observed in many

cancer cells, would result in a similar production of activating Myc/Max het-

erodimers leading to gene activation. Hence, as in the case of the Fos/Jun

system, transformation by the Myc oncogene appears to depend primarily on

its over-expression resulting in the activation of genes required for cellular

growth.

Interestingly, it has been shown that Myc can also interact with another

transcription factor, Miz-1 (Myc interacting zinc finger protein-1), which is a

zinc finger protein (see Chapter 4, section 4.3 for discussion of this type of

protein). Un like the situation with Max, however, in the absence of Myc, Miz-1

acts as an activator of genes promoting growth arrest. In the presence of Myc,

however, this activity of Miz-1 is inhibited resulting in the repression of these

genes (Peukert et al., 1997). Hence the rise in Myc levels in transformed cells

stimulates the activity of growth promoting genes via Myc/Max-mediated

gene activation and represses growth inhibitory genes via a repres sion of

Miz-1 activity.

As well as regulating growth by altering the tr anscription of specific pro-

tein-coding genes by RNA polymerase II, another means by which Myc can

alter growth has recently been demonstrated. Thus, it has been shown that

Myc interacts with the TFIIIB transcription factor which is essential for tran-

scription by RNA polymerase III (see Chapter 3, section 3.4) and stimulates

the transcription of the genes encoding tRNA and 5S ribosomal RNA

(Gomez-Roman et al., 2003). Since these RNAs are essential for protein synth-

TRANSCRIPTION FACTORS AND HUMAN DISEASE 311

Figure 9.22

In resting cells, Myc-

dependent genes will be

repressed by a Mad/Max

heterodimer. As cells

begin to grow, the

expression of Myc

increases resulting in the

formation of Myc/Max

heterodimers which

activate transcription.