Lennarz W.J., Lane M.D. (eds.) Encyclopedia of Biological Chemistry. Four-Volume Set . V. 4

Подождите немного. Документ загружается.

cytokines. Four family members of IRAK have been

reported in mammals (IRAK-1, IRAK-2, IRAK-4, and

IRAK-M). IRAK-1 and IRAK-4 are ubiquitously

expressed active kinases, whereas IRAK-M is preferen-

tially expressed in macrophage and monocytes. Stimu-

lation of IRAK-4-deﬁcient macrophages with ligands of

TLR2, TLR3, and TLR4 induced almost no production

of inﬂammatory cytokines. Inﬂammatory cytokine pro-

duction was also signiﬁcantly diminished in IRAK-1-

deﬁcient macrophages stimulated with LPS (the TLR4

ligand). Thus, IRAK-1 and IRAK-4 are involved in the

TLR-mediated signaling pathways. On the other hand,

IRAK-M-deﬁcient macrophages showed overproduction

of inﬂammatory cytokines in response to TLR ligands,

indicating that this molecule is involved in the negative

control of cytokine production. Stimulation of TRAF6-

deﬁcient macrophages with LPS led to impaired produc-

tion of inﬂammatory cytokines, suggesting that TRAF6 is

also important in the TLR-mediated signaling pathway.

In addition to MyD88, TIR domain-containing

adaptor protein (TIRAP)/MyD88-adaptor-like (Mal)

has been identiﬁed as the second TIR domain-containing

molecule. Stimulation of TIRAP/Mal-deﬁcient mice

with LPS (the TLR4 ligand) and lipopeptide (the TLR2

ligand) induced no production of inﬂammatory cyto-

kines. Thus, TIRAP/Mal is essential for the MyD88-

dependent pathway of TLR2 and TLR4.

MYD88-INDEPENDENT

SIGNALING PATHWAY

As described, stimulation of MyD88-deﬁcient macro-

phages with TLR ligands did not lead to inﬂammatory

cytokine production. However stimulation with LPS

induced delayed activation of JNK and NF

B in these

cells, indicating that there is another bypass pathway

which transduces signals in the absence of MyD88.

Beside LPS, dsRNA (the TLR3 ligand) is also capable of

activating NF

B in MyD88-deﬁcient macrophages. In

addition, it has been shown that LPS and dsRNA both

induce the activation of IRF-3 transactivator and lead

to the production of IFN-

in a MyD88-independent

fashion.

Recently, a third TIR domain-containing adaptor

protein called TIR domain-containing adaptor inducing

IFN-

(TRIF) was identiﬁed. TRIF-deﬁcient mice

showed no activation of IRF-3 or production of IFN-

in response to the TLR3 and TLR4 ligands. Therefore,

TRIF is an essential molecule for the activation of IRF-3

transactivator through TLR3 and TLR4. Furthermore,

the lack of LPS-induced activation of NF

B and JNK

in MyD88/TRIF double deﬁcient cells indicates that

TRIF is essential for the TLR4-mediated MyD88-

independent pathway.

Structural and Functional

Homologues of TLRs

RP105, a molecule containing an ectodomain with

LRRs and one cysteine-rich region but lacking a

cytoplasmic TIR domain, has been identiﬁed. RP105 is

preferentially expressed in mature B cells. Mice deﬁcient

in RP105 or TLR4 does not show proliferation in

TABLE II

Distribution of TLRs

Types of Toll-like receptor

Cell/tissue type TLR1 TLR2 TLR3 TLR4 TLR5 TLR6 TLR7 TLR8 TLR9 TLR10

1 Monocytes þþ2 þ þþþ þ þ þ

2 Macrophage þþ2 þ þþþ þ þ þ

3 Plasmacytoid dendritic cell þþ2 þþ2 þþ/ 2 M(þ )H(þ)NK

Myeloid dendritic cell NK þ NK þ NK þ H(þ/2 )M(þ) þ M(þ)H(2 )NK

4 Immature dendritic cell

þþNK þþNK NK NK NK NK

Mature dendritic cell

NK NK þ NK NK NK NK NK NK NK

5 Mast cell NK þ NK þ 2 þ NK þ NK NK

6 Intestinal cell (Apical side) NK NK NK þ(low) NK NK NK NK NK NK

Intestinal cell (Basolateral side) NK NK NK NK þ NK NK NK NK NK

7 Renal epithelia NK þ NK þ NK NK NK NK NK NK

8 Pulmonary epithelia NK NK NK þ NK NK NK NK NK NK

9 Corneal epithelia NK NK NK þ NK NK NK NK NK NK

Initially immature dendritic cells (DCs) express TLR1, 2, 4, and 5 as soon as immature DC encounter with antigen the expression of TLR1, 2,

4, and 5 regresses and expression of TLR progresses.

Abbreviations: M, mouse; H, human; NK, not known.

192 TOLL-LIKE RECEPTORS

FIGURE 1 Intracellular signaling pathway through TLR. Binding of TLR ligands initiates TLR-mediated signaling pathways from the

cytoplasmic TIR domain. The TIR domain-containing adaptor MyD88 is essential for the induction of inﬂammatory cytokines via all TLRs.

Another TIR domain-containing adaptor TIRAP is involved in the MyD88-depdendent pathway of TLR2 and TLR4. In TLR3 and TLR4 signaling,

the IRF-3 transcription factor is activated in a MyD88-independent fashion. The third TIR domain-containing adaptor TRIF mediates the MyD88-

independent signaling, thereby inducing IFN-

TOLL-LIKE RECEPTORS 193

response to LPS. This shows that RP105 is required for

LPS-induced responses in B cells.

Besides mammals, plants also have more than 100

genes which encode proteins with the TIR domain. Some

of these proteins also have LRRs and nucleotide-binding

site (NBS). These proteins play an important role in

defense against plant virus.

Future Perspective

Pathogens are dynamic entities, which contain many

complex biomolecules such as LPS and lipoprotein. In

recent years, intensive research has been conducted to

elucidate the different TLR signaling pathways such as

the MyD88-dependent and MyD88-independent path-

way by using puriﬁed pathogen-derived biomolecules.

However, it is still not known how all these pathogen-

derived complex molecules interact with macrophages,

what the net or ﬁnal signals for clearance of pathogens

are, or how different pathways communicate with each

other in vivo. In the near future, all these questions will

be clariﬁed using several mutant mouse strains lacking

the different components of TLR signaling pathways.

FURTHER READING

Akira, S. (2003). Mammalian toll-like receptors. Curr. Opin.

Immunol. 15, 5 –11.

Akira, S., Takeda, K., and Kaisho, T. (2001). Toll-like receptor: Critical

proteins linking innate and acquired immunity. Nature 2,

675–680.

Medzhitov, R., and Kopp, E. B. (1999). The Toll receptor family and

control of innate immunity. Curr. Opin. Immunol. 11, 13–18.

Takeda, K., and Akira, S. (2003). Toll receptors and pathogen

resistance. Cell. Microbiol. 5, 143–153.

Takeda, K., Kaisho, T., and Akira, S. (2003). Toll-like receptors. Annu.

Rev. Immunol. 21, 335– 376.

Yamamoto, M., Sato, S., Hemmi, H., Hoshino, K., Kaisho, T.,

Sanjo, H., Takeuchi, O., Sugiyama, M., Okabe, M., Takeda, K.,

and Akira, S. (2003). Role of adaptor TRIF in the MyD88-

independent toll-like receptor signaling pathway. Science 301,

640–642.

BIOGRAPHY

Shizuo Akira is a Professor in the Department of Host Defense at

Research Institute for Microbial Diseases, Osaka University in Osaka,

Japan. His principal research interests are in mammalian innate

immunity. He holds a Ph.D. from Osaka University and has received

postdoctoral training at the University of California, Berkeley.

He analyzes the molecular mechanism by which the innate

immunity senses microbial invasion in mammals through generation

of knockout mice.

Kiyoshi Takeda was an Assistant Professor in the Department of Host

Defense at Research Institute for Microbial Diseases, Osaka University

in Osaka, Japan. He received Ph.D. from Osaka University. Presently,

he is a Professor in Department of Molecular Genetics at Medical

Institute of Bioregulation, Kyusha University in Fukuoka, Japan.

His research focuses on the mechanism for regulation of macrophage

activity and the involvement of innate immunity in inﬂammatory and

infectious diseases.

Himanshu Kumar is a graduate student, pursuing research in the

Department of Host Defense at Research Institute for Microbial

Diseases, Osaka University in Osaka, Japan. His research interest is

elucidation of new molecules involve in TLRs signaling.

194 TOLL-LIKE RECEPTORS

Transcription Termination

Thomas J. Santangelo and Jeffrey W. Roberts

Cornell University, Ithaca, New York, USA

Transcription termination is the process by which RNA

synthesis by RNA polymerase (RNAP) is stopped and both

the RNA and enzyme are released from the DNA template.

Since most regulation occurs at the initiation of RNA synthesis,

failure to terminate transcription at the ends of operons would

lead to bypass of downstream gene regulation due to rogue

elongation complexes continuing transcription beyond the end

of the upstream operon. Two types of transcription termin-

ation in bacteria have been characterized in both biochemical

reactions with puriﬁed RNA polymerase and in genetic

experiments: (1) “intrinsic termination,” which requires only

special template sequences, and (2) “factor-dependent termin-

ation,” which requires an additional enzyme that interacts

with RNA polymerase. The RNA polymerase transcription

elongation complex is extremely stable – it must persist over

time and distance in order for gene expression to succeed – but

it also must be disrupted efﬁciently at termination sites.

Models of terminator function have been strongly

informed by the recently determined atomic structures

of RNAP and the transcription elongation complex,

which reveal the molecular sources of stability that are

overcome by termination mechanisms.

Figure 1 shows a general elongation complex, high-

lighting the major contacts that provide stabilizing

energy to the complex. The RNAP enzyme surrounds

an 8 or 9 bp DNA–RNA hybrid that is held within the

main channel of the enzyme. The RNA emerges from the

enzyme through an exit channel where additional

contacts are made between single-stranded RNA

(ssRNA) and the protein. By far, the greatest sources of

stabilizing energy to the complex are the hydrogen-

bonding within the hybrid itself and the interactions

between the hybrid and the surrounding protein.

Terminators function by weakening these contacts,

leading to dissociation of the hybrid and disruption of

the complex.

Intrinsic Termination

Intrinsic terminators are composed of two critical

elements: a short region of dyad symmetry in the DNA

which encodes an RNA sequence capable of forming an

RNA–RNA duplex or RNA “hairpin,” and a sequence

encoding a uridine-rich sequence in the RNA that

immediately follows the hairpin (Figure 1). Transcript

release occurs 8–10 bp downstream of the base of the

RNA hairpin. Correct base pairing within the stem of

the hairpin, the presence of the uridine-rich segment,

and the correct spacing of these elements are critical to

termination efﬁciency. Terminator efﬁciency also can be

modulated by the DNA sequence immediately upstream

and downstream of the termination position.

Current research is focused on the way these sequences

and structures facilitate release of the transcript. The

complex has been weakened just by sequence context

alone; the ﬁnal 8 or 9 bases of a terminated transcript

occupy the hybrid-binding site and a dA:rU hybrid is the

weakest possible combination of nucleic acids. In the

absence of a terminator hairpin structure, the weak

hybrid alone is sufﬁcient to induce RNAP to pause at the

end of the uridine-rich sequence. The role of the uridine-

rich sequence in termination is likely twofold: (1) to

pause the elongation complex immediately downstream

of the region capable of hairpin formation; and (2) to

minimize the energetic contribution of the hybrid to the

overall stability of the complex.

The role of the hairpin structure and its effects on the

elongation complex remain debated. Although their

details differ, most models predict that hairpin formation

disrupts the upstream bases of the hybrid, triggering

collapse of the elongation complex. One model proposes

that hairpin formation immediately upstream of the

complex pushes RNAP downstream in the absence of

further synthesis. Forward translocation of several base

pairs would reduce the length of the weak RNA/DNA

hybrid, and the loss of hybrid interactions would result in

transcript release. An alternative model of termination

evokes a relatively large conformational change in the

enzyme upon interaction of the hairpin with RNAP. The

hairpin structure formed in the RNA could physically

interact with the complex, opening the enzyme and

freeing the RNA transcript from its enclosure within the

elongation complex, a reaction that again would be pro-

moted by the weak hybrid. It is also possible that a large

conformational change accompanies a forward translo-

cation to result in both transcript and enzyme release.

All of the multisubunit RNAPs are predicted (and

some are known) to share a common three-dimensional

structure and thus are similar in terms of the contacts

made between the enzyme and nucleic acid framework.

Thus, termination of transcription by the specialized

RNA polymerases in eukarya (RNA pol I, II, and III) and

the archaeal RNAPs likely also involves mechanisms to

initiate disruption of the RNA/DNA hybrid. Termina-

tion by pol I and pol III in eukarya, and the RNAP from

archaea, does involve oligo-T tracts at or near the

termination positions, although the mechanisms are

unknown.

Factor-Dependent Termination

RHO-MEDIATED TERMINATION

Rho factor is a hexameric RNA-dependent ATPase that

is found in most eubacteria, but not in archaea or

eukarya. Its general mechanism of action is to bind the

nascent transcript of an elongation complex, wrapping

, 60 nucleotides of RNA, translocating toward RNAP

in an ATP-dependent reaction, and ﬁnally dissociating

the complex by an unknown mechanism. An atomic

structure of the Rho hexamer shows a hexameric ring of

monomers, each with two domains. The structural

model proposes that a ring of RNA-binding domains

engages most of the RNA, which then projects through

the center of the ring to contact the ATPase domains that

effect translocation. A break in the hexameric ring

(“lock-washer” model) is proposed to allow RNA to

penetrate the hexamer.

Rho function is largely determined by the very tight

coordination between transcription and translation,

because it requires free RNA immediately upstream of

RNA polymerase in order to act. Translation in bacteria

occurs as the transcript is being made, and the ribosome

travels immediately behind the elongation complex,

FIGURE 1 (A) The transcription elongation complex is shown, highlighting the protein–nucleic acid contacts that provide stabilizing energy to

the complex. Subunits are colored according to the key at the left;

has been made partially transparent in order to reveal the main channel of RNA

polymerase and the nucleic acid framework inside. NT, non-template DNA strand; T, template DNA strand. (B) Sequence of an intrinsic terminator

(the E. coli trpA terminator) is shown in black with the corresponding RNA sequence and hairpin shown in red. The region of dyad symmetry in the

DNA is shown in bold and is underlined and codes for the RNA hairpin.

196 TRANSCRIPTION TERMINATION

so that free RNA is generally unavailable to Rho.

However, when the ribosome stalls are released, as at a

translation stop codon, transcription and translation

become uncoupled. As RNA polymerase continues to

transcribe, more RNA is exposed to solution, allowing

Rho access to the RNA. Since RNAP elongates at 50–

100 nucleotides per second, Rho provides a rapid

mechanism to ensure that RNAP does not transcribe

much beyond the point of uncoupling.

Although a major function of Rho is to abort futile

transcription as described above, Rho-dependent ter-

mination also is the natural mechanism at certain sites,

likely serving as the sole mechanism to terminate

transcription of many operons in E. coli. Well-charac-

terized examples of such sites have properties that favor

Rho action during transcription: absence of translation,

the absence of any secondary structure that would

prevent Rho wrapping RNA, and a cytidine-rich

transcript that favors Rho binding; furthermore, natural

Rho-dependent terminators are transcription pause sites,

a property that presumably gives Rho time to translocate

to RNAP and carry out release. It is likely that these

characteristics are sufﬁciently common that Rho could

generally act after translation fails in most operons.

MFD AND TRANSCRIPTION-COUPLED

DNA REPAIR

DNA damage is repaired more rapidly in actively

transcribed than in rarely transcribed genes, and the

template strand is preferentially repaired over the non-

template strand. These facts led to the discovery that an

RNAP molecule stalled at a DNA lesion can serve as a

signal that the DNA is damaged and lead to its rapid

repair. This activity contributes to an essential function,

because a lesion in a single copy gene would block

production of an essential protein. In bacteria, RNAP

stalled at a template strand DNA lesion is the target of

the Mfd protein, also termed the Transcription Repair

Coupling Factor. Mfd is an ATPase and a DNA

translocase that mediates both removal of RNAP stalled

at the lesion and recruitment of the nucleotide excision

repair machinery to the site of the lesion. Similar to Rho

activity, Mfd can terminate transcription of almost any

obstructed RNAP. Mfd binds to the DNA immediately

upstream of the stalled elongation complex and makes

direct contact with RNAP; using its two translocase

domains, Mfd then pushes RNAP downstream along the

DNA in an ATP-dependent reaction, collapsing the

upstream edge of the transcription bubble as it moves. In

the case where elongation is blocked by a template

strand DNA lesion, the rewinding and forward translo-

cation are proposed to force the RNA from the

elongation complex and drive RNAP downstream to

mediate termination. The rewinding of upstream

sequences also can rescue complexes that are backtra-

cked, forcing RNAP into an active, forward conﬁgura-

tion in the case where continued elongation is possible.

General Elongation Factors – NusA

and NusG

NusA and NusG are nearly universally conserved

eubacterial proteins – both essential in E. coli – that

affect the rate of elongation by RNA polymerase and

modulate the efﬁciency of termination. Both NusA and

NusG bind to the RNAP core enzyme and inﬂuence the

activity of RNAP alone, in combination with each other,

or in combination with speciﬁc regulatory factors. In a

puriﬁed system, NusA generally slows elongation,

enhances pausing, and increases intrinsic termination

efﬁciency. NusA may interact with the RNA hairpin and

enhance the proposed interactions between RNA hair-

pins and RNAP. NusG appears, in many ways, to func-

tion oppositely: in general, it enhances elongation and

limits pausing. The molecular details of the interac-

tions between these factors and RNAP are unknown.

Both NusA and NusG also inﬂuence Rho activity, and

although strong Rho-dependent terminators are pause

sites, the effects of NusA and NusG on Rho function are

opposite to the prediction based on their effects on

pausing: thus, NusA promotes pausing but inhibits Rho

function, and NusG inhibits pausing and promotes Rho

function. It has been suggested that the NusG effect may

result from NusG inhibiting backtracking; this could

stimulate Rho function if Rho is unable to release a

backtracked complex. There may be a distinction

between direct effects of the factors on RNA polymerase

at a site where Rho-dependent termination occurs, and

indirect effects due to NusA and NusG modulation of

the overall rate of transcription. Thus, NusA inhibition

could in fact favor termination by slowing RNAP so that

Rho can catch it through translocation along the RNA.

An important relation between the activities of NusA

and Rho in vivo is shown by the ability of a NusA

defective cell to survive if Rho is also defective, but not if

Rho is normally functional.

Regulated Termination

and Antitermination

Termination can serve as a means to regulate gene

expression in two fundamentally different modes that

involve antitermination. First, site-speciﬁc attenuator-

based systems act at particular termination sites in

operons to allow or disallow continued transcription,

thus determining whether downstream genes are

expressed. These terminators are usually intrinsic, but

TRANSCRIPTION TERMINATION 197

are sometimes Rho-dependent. Second, generalized

antiterminators modify RNA polymerase so that it

ignores terminators that otherwise stop transcription,

again allowing expression of downstream genes. Finally,

a few examples are known of generalized modiﬁcations

of RNA polymerase that cause termination.

SITE-SPECIFIC ANTITERMINATION:

TTENUATORS AND RIBOSWITCHES

Intrinsic termination is used to control expression of

many operons through devices that alternatively permit

or prevent termination in a leader sequence or initially

translated region of a set of genes. Such operons contain

alternative RNA structures in the beginning of the

transcribed region; competition between alternative

folding patterns determines whether or not intrinsic

terminators are formed, and in turn, whether transcrip-

tion of the remaining portion of the operon or gene

continues. The competition between structures in the

leader sequence is inﬂuenced in different examples by

coordinated translation and ribosome positioning on the

RNA, or by small molecules, including tRNAs, that bind

to the RNA transcript and affect global folding.

Regulation of the trp operon is the classic example of

translation-coupled attenuation. Stalling of the ribo-

some at tryptophan codons, because of limitation in

charged-aminoacyl tRNAs, determines which sequences

of the transcript can fold, and thus whether a terminator

hairpin can be formed. Attenuation can also be mediated

by interactions between uncharged tRNAs and leader

sequences containing binding sites for these tRNAs,

termed T-boxes. Riboswitches consist of the leader and

initially translated regions of regulated operons, which

undergo alternative folding in response to small regu-

latory molecules; both positive and negative ribo-

switches have been described.

TERMINATION AND

ANTITERMINATION FACTORS

Speciﬁc regulatory factors, such as the

N and Q

proteins, the bacterial RfaH protein, and the HK022

Nun, regulate expression of speciﬁc genes or operons by

changing the termination capacity of RNA polymerase.

In general, the terminator/antiterminator is recruited to

particular transcribing RNAPs via cis-acting nucleic acid

sequences. Details of the exact modiﬁcations these

factors make to the complex and the effects of these

modiﬁcations on the stability of the elongation complex

are just beginning to be understood at a molecular level.

N binds a speciﬁc site (nut) in the nascent RNA, and

in a complex with the general elongation factors NusA

and NusG, and the bacterial proteins NusB and S10,

interacts with RNAP to stimulate elongation through

downstream terminators. Both

Q and RfaH require cis-

acting DNA sequences to pause RNAP at discrete sites

and present a novel complex for interaction.

Q protein

interacts with RNAP at a

-dependent promoter

proximal pause and becomes a subunit of the elongating

RNAP, modifying it to prevent recognition of down-

stream terminators. Q likely remodels the active site and

strengthens interactions between the enzyme and the

hybrid to resist the action of a terminator. RfaH regulates

expression of particular operons by interacting with

RNAP at a regulatory pause (ops sequence) to stimulate

expression of distal genes. HK022 Nun protein stimu-

lates transcription termination after binding the same

RNA sequences that N protein binds, thus blocking

superinfection of its host by another lambdoid phage that

requires N function for growth. Elongation complexes

are halted by the action of Nun but not disrupted. Thus,

Nun is not a release factor; Nun-arrested complexes

require the activity of the Mfd protein to efﬁciently

release the transcript and disrupt the elongation complex.

FURTHER READING

Burns, C. M., Richardson, L. V., and Richardson, J. P. (1998).

Combinatorial effects of NusA and NusG on transcription

elongation and Rho-dependent termination in Escherichia coli.

J. Mol. Biol. 278, 307–316.

Gusarov, I., and Nudler, E. (1999). The mechanism of intrinsic

transcription termination. Mol. Cell 3, 495–504.

Komissarova, N., Becker, J., Solter, S., Kireeva, M., and Kashlev, M.

(2002). Shortening of RNA: DNA hybrid in the elongation

198 TRANSCRIPTION TERMINATION

complex of RNA polymerase is a prerequisite for transcription

termination. Mol. Cell 10, 1151–1162.

Martin, F. H., and Tinoco, I. Jr. (1980). DNA– RNA hybrid duplexes

containing oligo (dA:rU) sequences are exceptionally unstable and

may facilitate termination of transcription. Nucleic Acids Res. 8,

2295–2299.

Park, J. S., Marr, M. T., and Roberts, J. W. (2002). E. coli transcription

repair coupling factor (Mfd protein) rescues arrested complexes by

promoting forward translocation. Cell 109, 757–767.

Selby, C. P., and Sancar, A. (1993). Transcription-repair coupling and

mutation frequency decline. J. Bacteriol. 175, 7509– 7514.

Skordalakes, E., and Berger, J. M. (2003). Structure of the Rho

transcription terminator: Mechanism of mRNA recognition and

helicase loading. Cell 114, 135–146.

Toulokhonov, I., Artsimovitch, I., and Landick, R. (2001). Allosteric

control of RNA polymerase by a site that contacts nascent RNA

hairpins. Science 292, 730–733.

Weisberg, R. A., and Gottesman, M. E. (1999). Processive antitermi-

nation. J. Bacteriol. 181, 359– 367.

Yarnell, W. S., and Roberts, J. W. (1999). Mechanism of intrinsic

transcription termination and antitermination. Science 284,

611–615.

BIOGRAPHY

Jeffrey Roberts is the Robert J. Appel Professor of Molecular and

Cell Biology in the Department of Molecular Biology and Genetics

at Cornell University. His principal research interest is the

mechanism of termination and antitermination in bacteria. He

obtained his Ph.D. from Harvard University, and he is a member of

the National Academy of Sciences and the American Academy of

Arts and Sciences, and a Fellow of the American Society for the

Advancement of Science.

TRANSCRIPTION TERMINATION 199

Transcriptional Silencing

Ann Sutton and Rolf Sternglanz

Stony Brook University, Stony Brook, New York, USA

Transcriptional silencing is the process by which large regions of

a eukaryotic genome are rendered transcriptionally inactive due

to a change in chromatin structure. The role of chromatin

structure in the regulation of gene expression has become an

exciting area of study in recent years. In all eukaryotes, DNA is

packaged into nucleosomes to form a protein-DNA structure

called chromatin. Each nucleosome consists of 146 base pairs of

DNA wrapped around a protein octamer consisting of the four

core histones, H2A, H2B, H3, and H4. Linker DNA separates

adjacent nucleosomes. Microscopic analysis of chromatin from

higher eukaryotic cells has revealed the presence of two types of

chromatin, a highly condensed form called heterochromatin,

and a less condensed form called euchromatin. In general, genes

within heterochromatin are transcriptionally silenced, while

those within euchromatin are transcriptionally active. The

transcriptional silencing results from the inaccessibility of the

DNA to components of the transcriptional machinery or

from an inability of RNA polymerase to elongate through

heterochromatin. Silenced domains are also less accessible to

recombination, replication, and repair machinery than are

euchromatic regions. Silencing differs from gene-speciﬁc

repression in a number of ways. For example, silenced domains

extend over large regions of DNA, while gene repression is more

local and results from protein interactions within the promoter

of a single gene. Furthermore, silenced chromatin can be

maintained through many generations. The mechanism of this

inheritance is not yet well understood.

Heterochromatic or silenced chromatin domains exist in all

eukaryotes examined to date. Regions of the DNA that are

known to exist as heterochromatin include telomeres and

centromeres in ﬁssion yeast and metazoans, chromosome 4 in

Drosophila, and one of the X chromosomes in female mammals.

Transcriptional Silencing in Yeast

Transcriptional silencing is best understood in two

genetically tractable organisms, the budding yeast,

Saccharomyces cerevisiae, and the ﬁssion yeast, Schizo-

saccharomyces pombe.

BUDDING YEAST, S. CEREVISIAE

The mechanism and regulation of formation of silent

chromatin domains have been studied most extensively

in the budding yeast, S. cerevisiae, which has four major

silenced regions: the two silent mating loci, telomeres,

and ribosomal DNA (rDNA). Haploid budding yeast

exists as one of two mating types, a or

, determined by

whether a-or

-speciﬁc alleles are expressed at the MAT

locus. Yeast also have identical a and

alleles at two

additional loci, HML (which usually carries the

genes), and HMR (which usually carries the a genes).

However, the genes at HML and HMR are not expressed

because of transcriptional silencing. If silencing is lost at

these loci as a consequence of a mutation in a silencing

factor, cells express both a and

mating type

information and lose the ability to mate. Largely

through the use of genetic screens for mutants defective

in mating, most of factors required for silencing at the

HM loci have been identiﬁed.

Silencing requires cis-acting regions that ﬂank the

mating type genes as well as trans-acting factors, in

particular the four silent information regulator proteins

(Sir1–Sir4). The cis-acting regions, termed silencers,

contain binding sites for at least two of the following

three proteins: Rap1, Abf1, and the origin replication

complex (ORC) (Figure 1). All three proteins have

essential functions in yeast in addition to their roles in

gene silencing; Rap1 and Abf1 are transcription factors,

and ORC is required for DNA replication. On the other

hand, at the silent mating loci, these proteins function to

recruit the Sir proteins to the silencer. Orc1, one of the

six subunits of the ORCs, binds Sir1, which then recruits

Sir4, which most likely binds as a Sir4–Sir2 complex.

Rap1 and Abf1 recruit Sir3 and Sir4. Once the Sir

proteins are recruited to the silencer, multiple protein-

protein interactions between Sir2, Sir3, and Sir4, and

between Sir3 and Sir4 and the amino terminal tails of

histone H3 and H4, cause the spreading of these Sir

proteins, and consequently silenced chromatin, from the

silencers to the regions to be silenced, in this case the

HM loci (Figure 1). The extent of the repressed domains

can be determined by monitoring the expression of a

reporter gene placed at increasing distances from the

silencer. Silencing at HMR spans about 3.5 kbp of DNA,

and silencing at HML also extends over several kbp.

Sir3 and Sir4 are believed to play a structural role in

silent chromatin. Sir2, on the other hand, has recently

been shown to be an enzyme. It is an NAD

-dependent

histone deacetylase that removes acetyl groups from

lysines in the amino terminal tails of histone H3 and H4.

While this deacetylase activity is not required for the

binding of the Sir proteins to the silencers, it is required

for the spreading of the Sir proteins and the formation of

silent chromatin. Hypoacetylation of the lysines in the

histone tails has long been known to be a characteristic

of heterochromatin and recent studies suggest that Sir3

preferentially binds to histone tails that are not

acetylated. A model to explain the spreading of silent

chromatin from the silencers is that Sir2, once recruited

to a silencer, deacetylates the lysines on the tails of

histones H3 and H4 on adjacent nucleosomes. Sir3 and

Sir2–Sir4 complexes then bind to these histones, Sir2

deacetylates the histone tails on adjacent nucleosomes,

Sir proteins then bind to them, and so on. The unusual

NAD

requirement of Sir2 for its deacetylase activity

may link silencing to cellular metabolism, and indeed,

mutations in genes involved in a salvage pathway for

NAD

biosynthesis have been shown to affect silencing.

The mechanism of transcriptional silencing at telo-

meres in S. cerevisiae is quite similar to that at the HM

loci. Telomeres contain multiple binding sites for Rap1,

the same protein found at the HM silencers. Rap1 is able

to recruit Sir3 and Sir4–Sir2 and these proteins spread

several kbp from the telomeres into adjacent nucleo-

somes, causing silencing of genes in these regions

(Figure 2). Sir1 is not involved in telomeric silencing,

because there are no ORC-binding sites at telomeres and

Sir1 is recruited by ORC. Silencing within the rDNA

genes requires Sir2, but none of the other Sir proteins.

At these loci, Sir2 is a component of the RENT complex

that also contains Net1 and Cdc14. The role of these

proteins in the establishment and spreading of silencing

at rDNA remains to be elucidated. Since Sir2 is the only

Sir protein required for rDNA silencing, the mechanism

of silencing at this locus is considered fundamentally

different from that at the HM loci and telomeres.

However, the deacetylation activity of Sir2 is required

for rDNA silencing, and in that sense the mechanism is

similar at all silent loci in budding yeast.

While silent chromatin can extend over several kbp of

DNA, mechanisms exist to stop the spreading. Assays

have been developed to look for sequences that can

function as boundary elements between euchromatin

and heterochromatin. In some cases, boundary elements

appear to consist of strong transcriptional promoter

Rap1

ORC

Nucleosomes

HMR

2 22 2

Abf1

a2 a1

93 bp 2 kbp 75 bp

ORC

Rap1

BBA

BAE

Abf1

2 22 2

Nucleosomes

HML

31 bp 3 kbp 112 bp

FIGURE 1 The silent mating loci in budding yeast, S. cerevisiae. The proteins that bind to the silencers are shown, as well as the four Sir proteins,

Sir1 (cross hatched), Sir2 (orange), Sir3 (yellow), and Sir4 (purple). Nucleosomes are in green with diagonal hatch marks. Although the Sir proteins

are depicted only between the two silencers, actually the proteins spread bidirectionally from the silencers. (A) The HMR locus. (B) The HML locus.

TRANSCRIPTIONAL SILENCING 201