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

Подождите немного. Документ загружается.

Tubulin and its Isoforms

Eva Nogales

Howard Hughes Medical Institute, University of California, Berkeley, California, USA

Tubulin is an

dimeric protein that self-assembles into

microtubules and is present in all eukaryotes. Tubulin is highly

conserved across species, reﬂecting the sequence constraints

imposed by microtubule structure and function. Both

- and

-subunits exist in numerous isotypic forms and undergo a

variety of posttranslational modiﬁcations. Tubulin assembly

and disassembly, which are linked to GTP hydrolysis, make the

microtubule network dynamic both in time and space to

accommodate the needs of the cell during the cell cycle. Puriﬁed

tubulin retains its self-assembling capabilities, allowing the

biochemical and biophysical characterization of the micro-

tubule polymerization and depolymerization processes.

A variety of proteins interact with tubulin in the cell, affecting

the stability of microtubules and their function. Numerous

ligands bind to tubulin and inﬂuence its assembly properties,

among them several drugs that have proven to have anticancer

properties.

-tubulin is a rarer tubulin isoform involved in the

nucleation of microtubules at microtubule organizing centers.

Recently, additional tubulin isoforms of yet ill-deﬁned function

have been identiﬁed.

-Tubulin

PHYSICAL PROPERTIES

General

Tubulin

- and

-subunits have molecular weights of

, 50 kDa and are 36–42% identical and 63% homolo-

gous. Both tubulin subunits bind guanine nucleotides.

The binding to

-tubulin at the N-site is nonexchange-

able, while the binding to

-tubulin at the E-site is

exchangeable. Magnesium increases the afﬁnity of the

-subunit for GTP with respect to GDP. Nucleotide in

oligomeric tubulin or in microtubules does not exchange

with the solution, except for terminal subunits at

microtubule ends.

Neuronal cells are particularly rich sources of tubulin

because microtubules are required for axonal transport.

Neural tissue contains sufﬁcient tubulin to allow tubulin

puriﬁcation by repeated cycles of 378-induced assembly

and 08-induced disassembly, with intervening centrifu-

gation to alternately pellet microtubules or impurities.

The yield of tubulin from 1 kg of brain and yeast is

, 150 and , 5 mg, respectively. The assembly of puriﬁed

tubulin can be assayed by light scattering, X-ray

scattering, centrifugation, and electron microscopy.

Isotypes and Posttranslational Modiﬁcations

Tubulin exists in different isotypic forms, the biological

signiﬁcance of which is still a matter of debate. The

number of tubulin isotypes increases with the organism

complexity. While yeast has only two

- and one

-isotypes, higher eukaryotes have up to seven

- and

six

-tubulin isotypes. Certain isotypes have been found

to be tissue speciﬁc, and differential expression of

-tubulin isotypes has been observed during the cell

cycle. Some of these isotypes have been shown in vitro to

have different relative stabilities, and such differences

seem important for the response of the cell to anti-

tubulin, anticancer drugs. The majority of differences

between isotypes localize within the last 15 residues of

the sequences, a region that has been identiﬁed as

important in the interaction of microtubules with micro-

tubule associated proteins (MAPs), pointing to a possible

relevance for the functionality of microtubules in the cell.

Both tubulin subunits can be extensively altered by

posttranslational modiﬁcation, including detyrosina-

tion/tyrosination, acetylation/deacetylation, phos-

phorylation, polyglutamylation, and polyglycylation.

All of these modiﬁcations, except for the acetylation of

-tubulin at Lys 40, occur at the divergent, highly charged

C-terminal end of

-and

-tubulin. As for the different

tubulin isotypes, the functionality of the posttransla-

tional modiﬁcations is still a matter of debate. Certain

modiﬁcations were initially identiﬁed as causing micro-

tubule stability, and had been later described as the

effect, not the cause of microtubule stability.

Structure

The structure of tubulin was solved by electron crystal-

lography of zinc-induced two-dimensional tubulin

sheets stabilized with the anticancer drug taxol.

The

- and

-tubulin have basically the same secondary

structure, each being made of a core of two

-sheets

surrounded by helices as shown in Figure 1.

The N-terminal domain forms a Rossmann fold with

the nucleotide-binding site at the C-terminal end of six

parallel strands that are surrounded by ﬁve helices.

An intermediate domain containing the binding site of

taxol and colchicine is formed by a mixed

-sheet of four

strands and ﬁve surrounding helices. The C-terminal

domain contains two long helices that overlap the

previous two domains and constitute the outside crest of

the protoﬁlaments in the microtubule to which motor

molecules bind. The nucleotide sits at the interface

between subunits along the protoﬁlament. The N-site in

-tubulin is buried within the dimer, while the E-site in

-tubulin is partially exposed in the dimer but occluded

in microtubules. FtsZ, a ubiquitous protein in eubacteria

and archebacteria essential for cytokinesis, is the only

known structural homologue of tubulin.

Synthesis and Folding

Tubulin synthesis in cells is regulated by a process in

which an increased subunit concentration leads to

speciﬁc degradation of

-tubulin mRNA. In animal

cells there is a mechanism to assure equivalent synthesis

- and

-subunits. For its correct folding tubulin

requires the citosolic chaperonin CCT (also referred to

as TriC, TCP1, or Ct-cpn60), a hetero-oligomer formed

by eight different subunits assembled into a hexadeca-

mer of two double rings. Folding by CCT requires cycles

of binding and full release, each cycle consuming one

ATP by the chaperonin. In addition tubulin requires

additional chaperonin cofactors that bind sequentially

-and

-monomers and are necessary for the

formation of the tubulin heterodimer. Folding cofactors

are important also in regulating

tubulin ratios.

TUBULIN POLYMERIZATION

Microtubule Assembly and Structure

The tubulin sequence and structure contains the

information required for its self-assembly into polar,

dynamic microtubules, which in turn interact with a

variety of cellular factors. Tubulin dimers bind head

to tail making linear protoﬁlaments, which associate

in a parallel fashion giving rise to a polar microtubule.

In a cell the so-called minus-end of microtubule, capped

-subunits, is attached to the centrosome where

-tubulin and related proteins nucleate microtubules.

The more dynamic plus-end, capped by

-subunits,

binds the kinetochore in mitosis.

The orientation of the tubulin subunits in the

microtubule is such that the C-terminal helices form

the crest of the protoﬁlaments on the outside surface,

making them an essential part of the binding site

for motor proteins (kinesins and dyneins) and MAPs.

The taxol-binding site is on the inside surface of the

microtubule, and close to lateral interactions. The lateral

contact between protoﬁlaments is dominated by the

interaction of the so-called M-loop in the second domain

with loop H1–S2 and helix H3 in the N-terminal,

nucleotide-binding domain.

The plus end of the microtubule is crowned by

-tubulin subunits exposing their nucleotide end to the

solution, while the minus end is crowned by

-subunits

exposing their catalytic end. When a dimer is added to a

plus end, its catalytic end contacts the E-site nucleotide

of the previous subunit forming the interface that should

bring about hydrolysis.

FIGURE 1 Ribbon diagram for the structure of

-tubulin

corresponding to a view from the inside of the microtubule with the

plus end at the top.

-Tubulin (top) is bound to GDP while

-tubulin

(bottom) is bound to GTP. The color scheme highlights the three

domains in the structure of each monomer: green for N-terminal,

nucleotide-binding domain (helix H7 or core helix is shown in lime);

second or intermediate domain in blue, C-terminal domain in purple.

TUBULIN AND ITS ISOFORMS 273

Dynamic Instability and GTP Cap Model

Microtubules are highly dynamic and can switch

stochastically between growing and shrinking phases,

both in vivo and in vitro. This nonequilibrium behavior,

known as dynamic instability, is based on the binding and

hydrolysis of GTP by tubulin subunits. Only dimers with

GTP in their E-site can polymerize, but following

polymerization this nucleotide is hydrolyzed by inter-

actions with the previous tubulin dimer and then becomes

nonexchangeable. In the GTP cap model the unstable

body of the microtubule made of GDP–tubulin is stabil-

ized by a layer of tubulin subunits at the ends that still

retain their GTP. When this cap is stochastically lost, the

microtubule rapidly depolymerizes. Depolymerization

may occur by weakening of lateral contacts at the ends,

and the consequent release of the constrained GDP sub-

units into a curved, lower energy, conformational state.

Microtubule assembly and stability are further

modiﬁed in the cell by interaction with cellular factors

that stabilize or destabilize microtubules at different

points in the cell or at different stages in the cell cycle.

Measurement of microtubule dynamics in vitro and

in vivo by DIC or ﬂuorescence microscopy yields the

rate constants for addition of tubulin–GTP and loss

of tubulin-GDP subunits at the two microtubule ends,

as well as the rates of catastrophe (switch from growth

to shrinkage) and rescue (switch from shrinkage

to growth). A variety of drugs bind to tubulin and affect

its polymerization. Microtubule-depolymerizing drugs,

such as colchicine, nocodazole, vinca alkaloids, or

podophyllotoxin, have a much higher afﬁnity for the

dimer than for tubulin in microtubules, so that

disassembly is caused by their mass action effect. At

substoichiometric concentrations tubulin-drug subunits

bind to the microtubule ends and form caps that

dramatically modify microtubule dynamics. Micro-

tubule-stabilizing drugs such as taxoids, epothilones,

or discodermolide act by binding preferentially to the

polymerized form of tubulin.

An alternative microtubule behavior is treadmilling, a

net ﬂow of subunits from the plus to the minus end

without a signiﬁcant change in microtubule length.

Anti-Mitotic Tubulin Ligands

Recent years have seen the discovery of numerous

tubulin ligands with anti-mitotic properties and anti-

cancer potential. Concerning their binding site these

agents can be classiﬁed into three main groups: those that

bind tubulin at the colchicine binding site; those who

bind it at the vinblastine site; and those who bind it at

the taxol site. Functionally these antimitotic ligands

can be separated into two classes: those that inhibit

microtubule assembly (e.g., the colchicine and vinblas-

tine families), and those that promote microtubule

assembly and stabilization (e.g., the taxol family). In

spite of these differences, the main action of all these

agents seems to be to cause mitotic arrest by inhibiting

normal dynamic instability at very low concentrations.

The colchicine-binding site is located at the

monomer–monomer interface within the dimer, in

agreement with a model of colchicine action in which

binding a distortion of the dimer structure that

inhibits its polymerization. Vinblastine-like compounds

are thought to bind at longitudinal polymerization

contacts, resulting in a distorted protoﬁlament

structure. Finally, there is direct information on the

binding site of taxol from the crystal structure of

tubulin. Taxol binds to

-tubulin, near the M-loop

and the site of lateral interactions, in a hydrophobic

pocket that in

-tubulin is occupied by eight extra

residues (see Figure 2). New microtubule-stabilizing

agents with promise for cancer treatment, such as

epothilones, discodermolide, eleutherobin, and the

sarcodictins, while very different in structure, all seem

to compete with taxol for the binding to a common site.

It has been proposed that taxol stabilizes lateral

contacts, or alternatively, that it acts as a bridge

that holds the N-terminal and second domains in a

relative orientation that favors longitudinal contacts

between subunits.

-Tubulin and Microtubule

Nucleation

Crucial to their dynamic behavior and function is the

nucleation of microtubules in the cell at microtubule

FIGURE 2 Anticancer drug taxol at its tubulin-binding site. Tubulin

is shown on a surface representation with the same color scheme as in

Figure 1. Both the N-terminal and second domain are part of the taxol-

binding pocket.

274 TUBULIN AND ITS ISOFORMS

organizing centers (MTOCs), to which they are attached

by their minus ends. An essential role in microtubule

nucleation is played by

-tubulin, a protein with high

homology to

-tubulins that localizes at MTOCs.

-Tubulin forms ring structures that serve as templates

for microtubule growth. The direct involvement of

-tubulin in microtubule nucleation has been demon-

strated in vitro using puriﬁed

-tubulin-containing ring

complex (

TuRC) containing at least seven different

proteins. There are two main models in the literature.

A model based on the shape and size of the

-TuRC is

that the ring forms the ﬁrst helical turn of the growing

microtubule, serving as a template for longitudinal

interaction with the tubulin

dimers. An alternative

model, based on the similarity of rings structures formed

-tubulin,

-tubulin, and FtsZ, is that

-tubulin

forms a protoﬁlament-like structure by longitudinal self-

association that then serves as a template for lateral

interaction with

-tubulin protoﬁlaments.

RARER TUBULIN ISOFORMS:

-, z-,

AND

-TUBULINS

Four new tubulin isoforms have recently been discovered.

The

-tubulin was identiﬁed in Chlamydomonas mutants

having abnormal basal bodies (these are microtubule

structures at the base of cilia and ﬂagella structurally

similar to the centrioles in the centrosome at MTOCs).

Human

-tubulin was subsequently found in the human

genome database, and shown to localize to the centro-

some, where it partially colocalizes with

-tubulin.

The 1-tubulin was identiﬁed from the human genome

database on the basis of sequence similarity to other

tubulins. Like

-tubulin, 1-tubulin localizes to the

centrosome, but in a cell-cycle-dependent manner: in

cells with duplicated centrosomes 1-tubulin localizes

only with the old centrosome.

Even rarer,

-tubulin has so far only been found on

kinetoplastid protozoa where it localizes to the basal

body, while

-tubulin has been found in paramecium

where it may interact with

TuRC.

TUBULINS AND THE CELL

The involvement of the microtubule cytoskeleton in a

large number of essential and diverse functions requires

both reliability and ﬂexibility from the system at the

expense of biochemical and structural complexity.

Dynamic instability is an inherent property of micro-

tubules, built into the

-tubulin structure. The spatial

and temporal organization of the microtubule network

in the cell is obtained through the regulation of dynamic

instability by an increasing number of factors that

ﬁne-tune the behavior of the microtubule system

to accommodate the requirements of the cell.

Regulation may happen at many different stages,

via transcription of different tubulin isotypes, the

control of tubulin monomer folding, the formation of

functional dimers, the posttranslational modiﬁcation

of tubulin subunits, the nucleation of microtubules, or

the interaction of microtubules with numerous stabil-

izers and destabilizers. Tubulin-binding drugs can

dramatically disrupt the ﬁnely tuned behavior of

microtubules. Finally, while

-tubulin is known to

be essential for microtubule nucleation, additional

tubulin isoforms, only recently discovered, have yet ill-

deﬁned functions.

FURTHER READING

Desai, A., and Mitchison, T. J. (1997). Microtubule polymerization

dynamics. Ann. Rev. Dev. Biol. 13, 83– 117.

Downing, K. H. (2000). Structural basis for the interaction of tubulin

with proteins and drugs that affect microtubule dynamics. Ann.

Rev. Cell Dev. Biol. 16, 89–111.

Lewis, S. A., Tian, G. and Cowan, N. J. (1997). The

- and

-tubulin

folding pathways. Trend. Cell Biol. 7, 479– 484.

Lowe, J., Li, H., Downing, K. H., and Nogales, E. (2001). Reﬁned

structure of alpha beta-tubulin at 3.5 A resolution. J. Mol. Biol.

313, 1045– 1057.

Mitchison, T., and Krischner, M. (1984). Dynamic instability of

microtubule growth. Nature 312, 237 –242.

Nogales, E. (2000). Structural insights into microtubule function. Ann.

Rev. Biochem. 69, 277–302.

Nogales, E., Wolf, S. G., and Downing, K. H. (1998). Structure of the

tubulin dimer by electron crystallography. Nature 391,

199–203.

Nogales, E., Whittaker, M., Milligan, R. A., and Downing, K. H.

(1999). High resolution structure of the microtuble. Cell 96,

79–88.

TUBULIN AND ITS ISOFORMS 275

BIOGRAPHY

Eva Nogales is an Assistant Professor in the Department of

Molecular and Cell Biology at UC Berkeley, an Assistant Investigator

at the Howard Hughes Medical Institute, and a Staff Scientist at

Lawrence Berkeley National Laboratory. She was trained as a

physicist in Spain, her country of origin, and obtained a Ph.D. in

biophysics from Keele University for her work at the Synchrotron

Radiation Source at Daresbury, UK. During her postdoctoral

studies with Dr. Kenneth H. Downing at Lawrence Berkeley

National Laboratory she obtained the structure of

-tubulin using

electron crystallography. Her present research centers around the

structural bases of microtubule dynamics and the structural

characterization of protein complexes involved in eukaryotic

transcription.

276 TUBULIN AND ITS ISOFORMS

Tumor Necrosis Factor Receptors

Karen G. Potter and Carl F. Ware

La Jolla Institute for Allergy and Immunology, San Diego, California, USA

Tumor necrosis factor receptors (TNFRs) are a family of

structurally similar cytokine receptors that act as transducers

of cell death and induce the expression of genes involved in

cellular differentiation and survival. Binding of speciﬁc ligands

to their cognate TNFR initiates the recruitment of adaptor

proteins, either death domain (DD)-containing or TNFR-

associated factor (TRAF) family of adaptor proteins, to the

cytosolic signaling domain of the receptor to initiate diverse

effector functions. The most well-known function of the TNF

superfamily is in immune regulation and development with

speciﬁc roles in host defense, inﬂammation, cellular homeo-

stasis, and lymphoid organogenesis. A critical role for TNFR in

immunobiology is evidenced by the linkage of naturally

occurring mutations in TNF family genes to human disease,

as well as by the targeting of TNF family members by viruses as

a mechanism of immune evasion. However, some TNF family

members also act outside the immune system by regulating

the development of hair follicles, sensory neurons, or bone-

resorbing osteoclasts.

Features of Tumor Necrosis

Factor Receptors

STRUCTURE

Tumor necrosis factor receptors (TNFRs) are identiﬁed

by a highly conserved, cysteine-rich domain (CRD) in

the extracellular portion of the protein that binds ligand

(Figure 1). The CRD generally contains six cysteines that

form three disulﬁde bonds typically recognized by the

signature sequence motif CxxCxxC. Currently, there are

29 members of the cellular TNFR family in mammals,

and several variants of TNFRs are found in herpes-

viruses and poxviruses. The cellular receptors are

primarily type I transmembrane proteins (extracellular

N terminus). Some receptors in this family lack

transmembrane and cytoplasmic domains and are

secreted, functioning as decoy receptors for the ligand.

The TNFR family can be divided into two general

groups – those that contain a death domain (DD) and

those with a peptide motif that binds TNFR-associated

factors (TRAF) adaptors – based on the structure of

their cytoplasmic tails and the signaling adaptors they

recruit to propagate signals to the cells.

Given the predominant role of the TNFR family in

regulating immunity, this suggests that the evolution of

the receptors in this family arose coincident with the

evolution of adaptive immunity, also found exclusively

in vertebrates. The size of the TNFR superfamily

appears to have grown in a large part by gene

duplication as many of the TNFR genes are linked to

discrete loci reﬂecting their evolutionary derivation.

Perhaps most obvious are those TNFRs found on

chromosome (Ch) 12p13 (TNFR1, LT

R, and CD27),

which likely underwent duplication and translocation

events giving rise to the larger locus of TNFR on Ch

1p36 (TNFR2, HVEM, Ox40, CD30, AITR, 4-1BB, and

DR3) (Figure 2A). Strikingly, NGFR is the only receptor

that binds ligands structurally unrelated to TNF,

representing a clear functional demarcation from the

typical TNFR (Figure 2B). Another subtle branch in the

TNFR family tree are those receptors that engage BAFF,

the B cell survival factor, and related ligands APRIL and

TWEAK (Figure 2C). These receptors have a single CRD

and in the case of BAFFR only two disulﬁde bonds.

Functional divergence is evident in the role some

TNFRs play in bone (Osteoprotegerin-RANK-RANKL/

TRANCE) and ectodermal (Ectodermal dysplasin EDA-

EDAR), and angiogenesis (TWEAK-Fn14).

EXPRESSION

Expression patterns of the receptors are complex.

Several members of the TNFR superfamily are expressed

on cells of the immune system; for example, BAFFR

expression is exclusive to B lymphocytes. Other recep-

tors are found in hair follicles (e.g., EDAR and TROY)

or the nervous system (e.g., NGFR), yet others have

very broad tissue expression patterns, such as TNFR1.

The expression of some TNFR is inducible and regulated

and others are constitutive. For example, TNFR2 is

absent on naı

ve T cells but is rapidly upregulated upon

activation of T cells. In contrast, HVEM is highly

expressedonnaı

ve T cells, but shows diminished

expression upon T-cell activation. Regulation of recep-

tor expression can also be achieved through the

generation of soluble receptors; soluble TNFRs can be

generated by proteolytic processing (CD27, CD30,

CD40, TNFR1, and TNFR2), alternative splicing of

membrane forms (Fas, 4-1BB), or encoded in the

genome without a transmembrane region (OPG,

DcR3, TRAIL-R3).

LIGAND BINDING

TNF ligands are characterized as type II (intracellular N

terminus) transmembrane proteins containing a “TNF

homology domain” (THD) in their C terminus

(Figure 1). The THDs of each ligand associate to form

trimeric proteins; in most cases, the ligands form

homotrimers, but LT

and LT

form heterotrimers,

and so do BAFF and APRIL, as has been shown recently.

The TNF ligands are active in their membrane forms

where cell-to-cell contact is required to initiate signaling;

however, some ligands can be proteolytically cleaved

from the surface creating soluble cytokines that affect

cells distally.

The TNF ligand family currently includes 20 proteins

that are able to pair off with one or more receptors.

In fact, some ligands display overlapping receptor

recognition; for example, LIGHT and LT

2 bind

R, and LT

and LIGHT bind HVEM. The ligand-

receptor pairing seems promiscuous, although compara-

tive studies using mice deﬁcient in a speciﬁc TNF ligand

or receptor suggest that each cytokine-receptor system

has a unique role in immune physiology.

TNFR Signaling

Each TNF-related ligand has three receptor binding sites

that can cluster together two or three cell surface

receptors, juxtapositioning the cytoplasmic tails of the

receptors to initiate signal transduction. Recruitment of

specialized signaling molecules (adaptors) to the cyto-

plasmic domain occurs following receptor clustering.

Propagation of TNFR signals occurs through two

distinct classes of cytoplasmic adaptor proteins:

TRAFs or DD molecules.

TRAF SIGNALING

TRAF adaptor proteins are a small family of RING

ﬁnger proteins that play a critical role in propagating

signal transduction leading to the activation of latent

transcription factors (Figure 1). There are six numeri-

cally named TRAF proteins (TRAF1–6) that function

predominantly in TNFR-induced signaling, although

TRAF6 is also a key player in signal transduc-

tion initiated by the interleukin-1 receptor and the

Toll-like receptor (TLR) superfamily. Several TRAFs

can bind directly to the cytoplasmic tails of most of

the TNFR. The binding site found in CD40 or the

R is a short peptide sequence, PXQXT/S or

IPEEGD, respectively. The binding site in TRAF for

receptor binding is ﬂexible, thus accommodating a

variety of motifs. TNFRs with a DD bind TRAFs

indirectly via other adaptor proteins, such as TRADD

(TNFR-associated DD). Each TRAF interacts with

several different receptors such that TNFRs display

distinct interaction patterns with multiple TRAFs.

Since each TRAF is believed to have distinct biological

effects, the variation in TRAF binding by the TNFR

may be able to direct the signaling pathway to distinct

biological outcomes.

Functionally, binding of TRAFs to TNFR culminates

in the activation of transcription factors that act to

regulate gene expression. For example, nuclear factor

B (NF

B) is a small family of latent transcription

factors that induces the expression of a large variety of

genes involved in inﬂammatory and immune responses

(Figure 3). Two distinct forms of NF

B are recognized:

B1 (RelA, p65) and NF

B2 (p100/p52), and each

pair with itself or other proteins (e.g., p50, RelB, and

cRel) that form active transcription factors that bind

FIGURE 1 Structural features of TNFR family. Space ﬁlling model of

a trimeric TNFR1 ligand, LT

, as it would exist as a membrane

anchored ligand, and a single TNFR1 rotated 1808 to reveal contact

residues. Also shown are crystal structures of the cytoplasmic DD

found in some TNFR and the TRAF that interacts with TNFR-

containing TRAF binding sites.

278 TUMOR NECROSIS FACTOR RECEPTORS

FIGURE 2 The TNF–TNFR superfamily grouped on the basis of chromosome (Ch) localization of TNFR. (A) TNFR and TNF ligands are shown

with arrows connecting ligand– receptor pairs. CRDs are shown as small ovals. DD is denoted as a rectangular box in the cytoplasmic tail of

appropriate receptors. The signaling and effector functions induced by either DD-containing or TRAF-binding TNFR is shown. (B and C) Same as

(A), with remaining TNFR grouped on the basis of chromosome localization and ligand binding.

TUMOR NECROSIS FACTOR RECEPTORS 279

DNA. NF

B1 is held in the cytosol by an inhibitor

protein (I

B) while an inhibitory domain controls the

cytosolic localization of NF

B2. The inhibitory pro-

teins are phosphorylated and degraded in response to

activation signals coming from diverse sources that all

target the inhibitor

B protein kinase complex (IKK).

Degradation of the inhibitor allows NF

B to localize

to the nucleus and activate transcription. The NF

and B2 pathways induce distinct sets of genes: NF

regulates the expression of proinﬂammatory chemo-

kines and adhesion molecules, while the NF

pathway controls the expression of distinct chemokines

and cytokines involved in lymphoid organogenesis.

BAFFR is only able to activate the NF

B2 pathway,

suggesting that this pathway is crucial for the

expression of survival genes speciﬁcally important for

B lymphocytes. By contrast, TNFR1 is unable to

activate NF

B2, which may account for its strong

proinﬂammatory action.

TRAF2 propagates signals to mitogen-activated

protein (MAP) kinases, including JNKs/SAPKs, ERKs,

and p38s, to activate the transcription factor AP-1 that

plays roles in stress responses and cellular homeostasis.

Like TRAF2, TRAF5 cooperates to activate NF

Band

AP-1 transcription factors, while TRAF3 negatively

regulates NF

B activation and thus may play a role

in promoting cell death. TRAF6 also activates NF

and AP1.

DEATH RECEPTOR SIGNALING

A number of TNFR, including TNFR1 and Fas, are

also termed “death” receptors because they regulate

apoptotic cell death. Their cytoplasmic tails contain a

region of , 80 amino acids that fold into six

-helices,

termed the DD. The DD serves as a protein interaction

motif to recruit signaling molecules to the inert

cytoplasmic domain of TNFRs. As such, the DD of

TNFR can self-associate or associate with other DD-

containing adaptor proteins such as FADD, TRADD,

and RIP.

In the simplest scheme known to activate the cell

death machinery, the adaptor FADD is recruited to Fas

initiating formation of the death-inducing signaling

complex (DISC). Procaspase 8 is recruited to the DISC

through a second interaction motif contained in FADD,

termed the death effector domain (DED), activating

downstream effector caspases (e.g., caspase 3) resulting

in the cleavage of critical cellular substrates and the

eventual collapse of the cells and death (apoptosis).

Similarly, TNFR1 activates apoptosis although it

requires the adaptor protein TRADD to facilitate

FIGURE 3 Mechanisms of signaling by TNFR. Upon ligation of LT

R, two NF

b pathways are induced. The ﬁrst leads to induction of the

B1 pathway and the activation of IKK

and RelA, which control expression of inﬂammatory genes. The second pathway results in the

activation of NF

B2 and the processing of p100 to p52 following the activation of NIK and IKK

, leading to the transcription of genes implicated in

secondary lymphoid organogenesis and homeostasis. BAFFR also activates the NF

B2 pathway, whereas TNFR1 only activates the NF

pathway. TNFR1 also contains a DD to signal caspase activation and apoptosis.

280 TUMOR NECROSIS FACTOR RECEPTORS

FADD recruitment, and is also able to recruit another

adaptor protein RIP that also plays a role in procaspase

8 activation. The DISC is regulated by two forms of

another protein known as FLIP, which can switch the

DISC between activating caspases or NF

B, which in

turn regulates genes that block apoptosis, such as the

inhibitors of apoptosis (IAP) that promote cell survival.

Effector Functions of

the TNFR Family

Despite similarities in structure and signaling mechani-

sms, TNFRs function in diverse, and often opposing,

roles in immune physiology. As discussed above, several

reasons that explain their functional diversity include

differences in ligand binding, tissue and cellular

expression, regulation of expression, and association

with signaling adaptor proteins.

INFLAMMATION

A role for TNF family members in host defense and

inﬂammation was ﬁrst appreciated when it was realized

that TNF is identical to a factor termed cachectin, a

protein known to cause fever and wasting. Inﬂammatory

cells, such as macrophages, induce TNF when special-

ized innate immune receptors called TLRs recognize

nonself patterns on microbial pathogens. Binding of

TNF to its receptors, TNFR1 or TNFR2, induces a

variety of molecules that are crucial for initiating the

acute inﬂammatory response. For example, TNF signal-

ing induces increased expression of adhesion molecules

on endothelial cells and secretion of chemokines to

promote inﬂammatory cell migration to the site of

infection. TNF can also activate macrophages and

neutrophils to increase their phagocytic capacity and

their ability to secrete tissue-degrading enzymes into

infected tissues. Although inﬂammation is crucial to

prevent infection, it must also be minimized after the

pathogenic challenge subsides in order to prevent

uncontrolled damage to fragile host tissues. As such,

uncontrolled TNF signaling can cause chronic inﬂam-

mation and wasting, or acutely released, septic shock.

Drugs that block TNF signaling, such as soluble TNFR

decoy receptors, have been developed to prevent

inﬂammatory diseases such as rheumatoid arthritis and

inﬂammatory bowel disease.

APOPTOSIS

Cell death by apoptosis is one of the major functions of

DD-containing TNFR. Fas, for example, mediates cell

death for varied purposes. First, Fas, and its ligand

FasL, cooperate as major effectors, along with the

perforin/granzyme pathway, to trigger killing of patho-

gen-infected cells by cytotoxic T-lymphoctyes (CTLs)

or by natural killer (NK) cells. In a second context, Fas is

required to eliminate excessive CTLs and other effector

cells. Fas is induced on long-term antigen-activated

lymphocytes to elicit their removal and minimize the

excessive expansion of T lymphocytes that can occur in

response to T-cell activation in the space-restricted

lymphoid environment. In fact, two naturally occurring

mutations in mice encoding the gene lymphoprolifera-

tive disorder (lpr), lpr,orlpr

that affect the expres-

sion or the biological activity of the Fas receptor,

respectively, demonstrate the important role of Fas in

lymphocyte homeostasis. Both of these mutant mice

develop an autoimmune phenotype characterized

by massive lymphadenopathy, splenomegaly, and auto-

antibody production.

CELL SURVIVAL

BAFFR, BCMA, and TACI are some of the most

recently discovered members of the TNFR family that

are predominantly expressed on B cells and represent a

good example of TNFRs that play a role in cell

survival. All three receptors bind the TNF-related

ligand BAFF, produced by macrophages and dendritic

cells, while BCMA and TACI can also bind the related

ligand APRIL. BAFF is required in the maintenance and

differentiation of mature B cells in peripheral tissues by

supporting B-cell survival and maturation presumably

through activation of anti-apoptotic genes. The mech-

anism for enhanced B-cell survival by BAFF may be

partly due to the stimulation of the pro-survival protein

bcl-2 and a decrease in the expression of the pro-

apoptotic factors Bak and Blk. A crucial role for

BAFFR in BAFF-mediated B-cell survival is evidenced

by the fact that neither BCMA- nor TACI-deﬁcient mice

have defects in B-cell maturation, whereas A/WySnJ

mice, containing a naturally occurring mutation in

BAFFR, do show defects in B-cell maturation. Over-

expression of BAFF in mice can result in autoimmune

symptoms, and similarly patients with systemic lupus

erythematosus (SLE), rheumatoid arthritis, and Sjo-

gren’s syndrome show elevated levels of BAFF in the

blood. Other TNFRs also play roles in the survival and

development of various tissues including the TRANCE

system in bone, EDAR in ectodermal tissues, and

NGFR in neurons.

COSTIMULATION

T lymphocyte activation is also regulated by TNFRs

where they act as costimulatory signals to positively or

negatively inﬂuence the proliferation of antigen-stimu-

lated T cells. OX40, CD27, 4-1BB, and HVEM,

expressed on CD4

and CD8

T cells, promote T-cell

TUMOR NECROSIS FACTOR RECEPTORS 281