Latchman. Eukariotic Transciption Factors

Подождите немного. Документ загружается.

ACKNOWLEDGEMENTS

I would like to thank all the colleagues, listed below, who have given permis-

sion for material from their papers to be reproduced in this book and have

provided prints suitable for reproduction.

Figures 4.1 and 7.9, photographs kindly provided by Professor W.J. Gehring

from Gehring, Science 236, 1245 (1987) by permission of the American

Association for the Advancement of Science.

Figure 4.14, photograph kindly pr ovided by Dr P. Holland from Holland and

Hogan, Nature 321, 251 (1986) by permission of Macmillan Magazines Ltd.

Figure 4.25 redrawn from Redemann et al., Nature 332, 90 (1988) by kind

permission of Dr H. Jackle and Macmillan Magazines Ltd.

Figures 4.31 and 4.35 redrawn from Schwabe et al., Nature 348, 458 (1990) by

kind permission of Dr D. Rhodes and Macmillan Magazines Ltd.

Figure 4.40 redrawn from Abel and Maniatis, Nature 341, 24 (1989) by kind

permission of Professor T. Maniatis and Macmillan Magazines Ltd.

Figures 7.4 and 7.7, photographs kindly provided by Dr R.L. Davis from Davis

et al., Cell 51, 987 (1987) by permission of Cell Press.

Figures 7.12 and 7.13, photographs kind ly provided by Dr R. Krumlauf from

Graham et al., Cell 57, 367 (1989) by permission of Cell Press.

Figure 8.5, photograph kindly provided by Professor M. Beato from Will mann

and Beato, Nature 324, 688 (1986) by permission of Macmillan Magazines Ltd.

Figure 8.12, photograph kindly provided by Dr C. Wu from Zimarino and

Wu, Nature 327, 727 (1987), by permission of Macmillan Magazines Ltd.

I am also especially grateful to the colleagues who have provided colour

prints of transcription factor structures, allowing us to include this feature.

Plate 1, kindly provided by Dr J. H. Geiger.

Plate 2 kindly provided by Dr T. Li and Professor C. Wolberger.

Plate 3 kindly provided by Professor P. E. Wright from Lee et al., Science 245,

635 (1989) by permission of the American Association for the Advancement

of Scien ce.

Plate 4 kindly provided by Dr R. J. Fletterick.

Plate 5 kindly provided by Professor R. Kaptein from Hard et al., Science 249,

157 (1990) by permission of the American Association for the Advancement

of Science.

Plate 6 kindly provided by Dr D. Rhodes from Schwabe et al., Cell 75, 567

(1993) by kind permission of Cell press.

Plate 7 kindly provided by Professor D. Moras.

xxvi ACKNOWLEDGEMENTS

CHAPTER 1

DNA SEQUENCES, TRANSCRIPTION

FACTORS AND CHROMATIN

STRUCTURE

1.1 THE IMPORTANCE OF TRANSCRIPTION

The fundamental dogma of molecular biology is that DNA produces RNA

which, in turn, produces protein. Hence if the genetic information that

each individual inherits as DNA (the genotype) is to be converted into the

proteins which produce the corresponding characteristics of the individual

(the phenotype), it must first be converted into an RNA product. The process

of transcription, whereby an RNA product is produced from the DNA, is

therefore an essential element in gene expression. The failure of this process

to occur will obviously render redundant all the other steps that follow the

production of the initial RNA transcript in eukaryotes, suc h as RNA splicing,

transport to the cytoplasm or translation into protein (for review of these

stages see Nevins, 1983; Latchman, 2002).

The central role of transcription in the process of gene expression also

renders it an attractive control point for regulating the expression of genes

in particular cell types or in response to a particular signal. Indeed, it is now

clear that, in the vast majority of cases, where a particular protein is produced

only in a particular tissue or in response to a particular signal, this is achieved

by control processes which ensure that its corresponding gene is transcribed

only in that tissue or in response to such a signal (for reviews see Darnell,

1982; Latchma n, 2002). For example, the genes encoding the immunoglobu-

lin heavy and light chains of the antibody molecule are transcribed at high

level only in the antibody-producing B cells while the increase in somatostatin

production in response to treatment of cells with cyclic AMP is mediated by

increased transcription of the corresponding gene. Therefor e, while post-tran-

scriptional regulation affecting for example, RNA splicing or stability plays

some role in the regulation of gene expression (for reviews see Bashirullah

et al., 2001; Graveley, 2001) the major control point lies at the level of

transcription.

1.2 CHROMATIN STRUCTURE AND ITS REMODELLING

1.2.1 CHROMATIN STRUCTURE AND GENE REGULATION

The central role of transcription, both in the basic process of gene expression

and its regulation in particular tissues, has led to considerable study of this

process. Initially such studies focused on the nature of the DNA sequences

within individual genes which were essential for either basal or regulated gene

expression. These sequences will be discussed in section 1.3. It is now clear,

however, that the accessibility of these DNA sequences and hence their ability

to regulate gene expression is controlled by the manner in which they are

packaged in the cell. The packaging of DNA will therefore be discussed in this

section.

It has been known for some time that the DNA in eukaryotic cells is pack-

aged by association with specific proteins, such as the histones, into a struc-

ture known as chromatin (for reviews see Wolffe, 1995; Latchman, 2002;

Felsenfeld and Groudine, 2003) . The fundamental unit of this structure is

the nucleosome in which the DNA is wrapped twice around a unit of eight

histone molecules (two each of histones H2A, H2B, H3 and H4) (for review

see Kornberg and Lorch, 1999). This structure is compacted further into the

so-called solenoid structure in genes which are not transcriptionally active or

about to become active. In contrast, active or potentially active genes exist in

the simple nucleosomal structure. Moreover, in the regulator y regions of

these genes nucleosomes are either removed altogether or undergo a struc-

tural alteration which facilitates the binding of specific transcription factors to

their binding sites in these regions (Fig. 1.1).

Interestingly, the tightly packed solenoid structure can be compacted even

further, by extensi ve looping, to form the chromosomes which are visible

during cell division. These loops are linked at their bases to a protein scaffold

known as the nuclear matrix, with such linkage occurring via specific DNA

sequences, known as matrix attachment regions (MARs) (Fig. 1.2; for review

see Horn and Peterson, 2002).

2 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 1.1

Levels of chromatin

structure in active or

inactive DNA.

Clearly, the access of a transcription factor to its approp riate binding site

will be affected by the manner in which that site is packaged within the

chromatin structure. Evidently, therefore, genes that are about to be tran-

scribed must undergo changes in chromatin structure which facilitate such

transcription by allowing access of activating transcription factors to their

binding sites. Although a detailed discussion of these changes is beyond the

scope of this book (for reviews see Aalfs and Kingston, 2000; Wu and

Grunstein, 2000; Bradbury, 2002; Latchman, 2002; Richards and Elgin,

2002; Felsenfeld and Groudine, 2003), at least two mechanisms which can

alter chromatin structure are of particular importance in terms of transcrip-

tion factor regulation and these will be discussed in turn.

1.2.2 CHROMATIN REMODELLING FACTORS

A number of studies have identified protein complexes which are capable of

binding to DNA, hydrolysing ATP and using the energy generated to disrupt

the nucleosomal structure. The best characterized of these is the SWI/SNF

complex which contains a number of different polypeptides. It was originally

defined in yeast but has now been identified in a range of organisms including

humans (for review see Aalfs and Kingston, 2000; Sudarsanam and Winston,

2000). The critical role of this complex in regulating gene expr ession is indi-

cated by the phenotype of the brahma mutation in Drosophila which inacti-

vates the SWI2 component of the complex. Thus, in this mutant the genes

DNA SEQUENCES, TRANSCRIPTION FACTORS AND CHROMATIN STRUCTURE 3

Figure 1.2

The tightly packed

solenoid structure can be

further compacted by the

formation of loops. These

loops (which contain

approximately 20–

80 000 bases of DNA)

are attached to the

nuclear matrix via specific

DNA elements known as

matrix attachment

regions.

encoding several homeobox-containing genes, which control the correct pat-

terning of the body (see Chapter 4, section 4.2), remain in an inactive chro-

matin structure and are hence not transcribed. This results in a mutant fly

with a grossly abnormal body structure (for review see Simon, 1995).

It is likely that SWI/SNF and other chromatin remodelling complexes can

act by three different methods to alter the accessibility of the DNA. Thus, they

may ac t by altering the association of the histone molecul es wit hin the nucleo-

some so that the nucleosome structure is changed in such a way as to allow

other factors to bind to DNA (nucleosome remodelling: Fig. 1.3a). Secondly,

they may act by causing the nucleosome to move along the DNA, so exposing

a particular DNA sequence (nucleosome sliding: Fig. 1.3b). Finally, they may

act by displacing a nucleosome so that it leaves the target DNA and binds to

another DNA molecule (nucleosome displacement: Fig.1.3c). All these meth-

ods have in common the use of ATP hydrolysis to alter the nucleosome in

some way so as to allow a particular region of DNA to become more accessible

and hence bind specific regulatory factors.

Evidently, these mechanisms beg the question of how the SWI/S NF com-

plex is itself recruited to the genes which need to be activated. This can occur

via its association with the RNA polymerase complex or by its association with

other transcription factors which can bind to their specific DNA binding sites

even in tightly packed, non-remodelled chromatin. These processes are

discussed in subsequent chapters.

Interestingly, it has recently been shown that chromatin remodelling com-

plexes can also be recru ited to the DNA by the SAT BI protein which is

involved in the looping of the chromatin into a highly compact structure

4 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 1.3

The SWI/SNF complex

can allow a regulatory

protein access to its

binding site (X) by (a)

producing an altered

structure of the

nucleosome in a process

known as nucleosome

remodelling; (b) inducing

nucleosome sliding to a

different position on the

DNA; or (c) displacing

the nucleosome onto

another DNA molecule.

(Yasui et al., 2002) (see section 1.2.1). This provides a link between the looping

process and chromatin remodelling/gene regulation and suggests that such

remodelling processes can target the large regions of DNA (20–80 000 bases

of DNA) contained in individual loops.

1.2.3 HISTONE AC ETYLATION

The histone molecules which play a key role in chromatin structure are subject

to a number of post-translational modifications such as phosphorylation, ubi-

quitination or acetylation (f or reviews see Strahl and Allis, 2000; Wu and

Grunstein, 2000; Jenuwein and Allis, 2001; Felsenfeld and Groudine, 2003).

In particular, the addition of an acetyl group to a free amino group in lysine

residues in the histone molecule reduces its net positive charge. Such acety-

lated forms of the histones have been found preferentially in active or poten-

tially active genes where the chromatin is less tightly packed. Moreover,

treatments which enhance histone acetylation, such as addition of sodium

butyrate to cultured cells, result in a less tightly packed chromatin structure

and the activation of previously silent cellular genes. This suggests that hyper-

acetylation of histones could play a causal role in producing the more open

chromatin structure characteristic of active or potentially active genes.

Hence, activati on of gene expression could be achieved by factors with

histone acetyltransferase activity which were able to acetylate histones and

hence open up the chromatin structure, whereas inhibition of gene expres-

sion would be achieved by histone deacetylases which would have the opposite

effect (Fig. 1.4). Most interestingly, rec ent studies have identified both com-

ponents of the basal transcriptional complex and specific activating transcrip-

tion factors with histone acetyltransferase activity as well as specific inhibitory

transcription factors with histone deacetylase activity (for review see Brown et

al., 2000). These findings, which link studies on modulation of chromatin

structure with those on activating and inhibitory transcription factors, are

discussed further in later chapters.

It is clear, therefore, that histone acetylation plays a key role in regulating

chromatin structure. However, in the last few years it has become increasingly

clear that other histone modifications, such as methylation, phosphorylation

or the addition of the small protein, ubiquitin (ubiquitination) are also

involved in this process and that these modifications interact with one another

and with acetylation. Thus, for example, demethylation of the lysine amino

acid at posit ion 9 in histone H3 facilitates phosphorylation of serine 10 and

acetylation of lysine 14 of H3 leading to opening of the chromatin and gene

activation (Paro, 2000) (Fig. 1.5). Such interaction can also occur between

modifications on one his tone molecule and those on another. Thus, ubiqui-

DNA SEQUENCES, TRANSCRIPTION FACTORS AND CHROMATIN STRUCTURE 5

6 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 1.4

(a) An activating molecule

(Act) can direct the

acetylation of histones in

the nucleosome (N)

thereby resulting in a

change in chromatin

structure from a tightly

packed (wavy line) to a

more open (solid line)

configuration. (b) An

inhibitory molecule can

direct the deacetylation of

histones thereby having

the opposite effect on

chromatin structure.

Figure 1.5

Demethylation of the

lysine amino acid at

position 9 in histone H3

facilitates phosphorylation

of serine 10 and

acetylation of lysine 14

leading to a more open

chromatin structure.

tination of histone H2B facilitates subsequent methylation of histone H3

on the lysines at positions 4 and 79 leading to a tightly packed chromatin

structure and gene silencing (Briggs et al., 2002; Sun and Allis, 2002).

This complex pattern of modificat ion has led to the idea of a ‘histone

code’ in which the chromatin structure of a particular gene is specified by

the pattern of different modifications of the histones that package it (for

reviews see Strahl and Allis, 2000; Berger, 2001; Goll and Bestor, 2002;

Turner, 2002) .

Hence, both ATP-dependent chromatin remodelling complexes and

alterations in histone acetylation/modification play a vital role in regulating

the chroma tin structure of specific genes. Although these two processes have

been discussed separately, it is likely that chromatin remodelling and histone

modification enzymes cooperate. Thus, for example, it has been shown that

acetylation of histones can allow recruitment of SWI/SNF to a promoter

(Agalioti et al., 2002) as well as preventing it from dissociating once it has

bound (Hassan et al., 2001). Hence, it appears that these two processes act

together to ensure that the DNA sequences involved in transcription control

become accessible at the correct time in development or in response to

appropriate signals (for review see Wu and Grunstein, 2000; Narlikor et

al., 2002). The nature of these DNA sequences is discussed in the next

section.

1.3 DNA SEQUENCE ELEMENTS

1.3.1 THE GENE PROMOTER

The primary aim of chromatin remodelling processes is to expose specific

DNA sequences so that these can be targeted by transcription factors involved

in the process of gen e transcription. In prokaryotes, such sequences are found

immediately upstream of the start site of transcription and form part of the

promoter directing expression of the genes. Seq uences found at this position

include both elements found in all genes which are involved in the basic

process of tr anscription itself and those found in a more limited number of

genes which mediate their response to a particular signal (for review see

Muller-Hill, 1996).

Early studies of cloned eukaryotic genes, therefore, concentrated on the

region immediately upstream of the tr anscribed region where, by analogy,

sequences involved in transcription and its regulation should be located.

Putative regulatory sequences were identified by comparison between differ-

ent genes and the conclusions reached in this way confirmed either by

DNA SEQUENCES, TRANSCRIPTION FACTORS AND CHROMATIN STRUCTURE 7

destroying these sequences by deletion or mutation or by transferring them to

another gene in an attempt to alter its pattern of regulation.

This work carried out on a number of different genes encoding specific

proteins identified many short sequence elements involved in transcriptional

control (for reviews see Davidson et al., 1983; Jones et al., 1988). The elements

of this type present in two typical examples, the human gene encoding the

70 kd heat inducible (heat shock) protein (Williams et al., 1989) and the human

metallothionein IIA gene (Lee et al., 1987), are illustrated in Figure 1.6.

Comparisons of these and many oth er genes revealed that, as in bacteria,

their upstream regions contain two types of elements. First, sequences found

in very many genes exhibiting distinct patterns of regulation which are likely

to be involved in the basic process of transcription itself and secondly those

found only in gen es transcribed in a particular tissue or in response to a

specific signal which are likely to produce this specific pattern of expression.

These will be discussed in turn.

1.3.2 SEQUENCES INVOLVED IN THE BASIC PROCESS OF

TRANSCRIPTION

Although they are regulated very differently, the hsp70 and metallothionein

genes both contain a TATA box. This is an AT-rich sequence (consensus

TATAA/TAA/T) which is found about thirty base pairs upstream of the

transcriptional start site in very many but not all genes. Mutagenesis or reloca-

tion of this sequence has shown that it plays an essential role in accurately

positioning the start site of transcription (Breathnach and Chambon, 1981).

The region of the gene bracketed by the TATA box and the site of transcrip-

tional initiation (the Cap site) has been operationally defined as the gene

promoter or core promoter (Goodwin et al., 1990). It is likely that this region

8 EUKARYOTIC TRANSCRIPTION FACTORS

Figure 1.6

Transcriptional control

elements upstream of the

transcriptional start site in

the human genes

encoding hsp70 (panel a)

and methallothionein IIA

(panel b). The TATA, Sp1

and CCAAT boxes bind

factors that are involved in

constitutive transcription

while the glucocorticoid

response element (GRE),

metal response element

(MRE), heat shock

element (HSE) and the

AP1 and AP2 sites bind

factors involved in the

induction of gene

expression in response to

specific stimuli.