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

Подождите немного. Документ загружается.

Uncoupling Proteins

Daniel Ricquier and Fre

ric Bouillaud

Centre National de la Recherche Scientiﬁque, Paris, France

Uncoupling proteins (UCPs) are mitochondrial transporters

present in the inner membrane of mitochondria. They belong

to the family of anion mitochondrial carriers including adenine

nucleotide transporters, phosphate carrier and other transpor-

ters. The term uncoupling protein was originally given to

UCP1 which is uniquely present in mitochondria of brown

adipocytes, the thermogenic cells devoted to maintenance of

body temperature in mammals. In these cells, UCP1 acts as a

proton carrier creating a shunt between complexes of

respiratory chain and the ATP-synthase. Purine nucleotide

inhibit UCP1 whereas fatty acids activate it. Activation of

UCP1 stimulate respiration and the uncoupling process results

in a futile cycle and dissipation of oxidation energy into heat.

UCP2 is ubiquitous and highly expressed in lymphoid system

and macrophages. UCP3 is mainly expressed in skeletal

muscles. In comparison to the established uncoupling and

thermogenic activities of UCP1, UCP2, and UCP3 rather

appear to be involved in the limitation of free radicals levels in

cells than in physiological uncoupling and thermogenesis. The

mechanism of the protonophoric activity of the UCPs is still

controversial since it has been proposed that the UCPs

transport fatty acid anions and catalyse proton transport by

fatty acid cycling. Quinones and superoxide ions may also

activate the UCPs.

The Mitochondria and the

Coupling of Respiration

to ADP Phosphorylation

Cellular respiration, the reactions of the citric acid cycle,

fatty acid oxidation, and several steps of urea synthesis

and gluconeogenesis take place in specialized cellular

organites, the mitochondria.

MITOCHONDRIA

In addition to oxidative phosphorylation and metabolic

pathways, mitochondria are involved in thermogenesis,

radical production, calcium homeostasis, apoptosis, and

protein synthesis. Mitochondria contain two compart-

ments bounded by inner and outer membranes. The

outer membrane is permeable to many small metabolites

whereas the permeability of the inner membrane is

controlled in order to maintain the high electrochemical

gradient created by the mitochondrial respiratory chain

which is necessary for energy conservation and ATP

synthesis in mitochondria. The inner membrane trans-

ports anion substrates such as ADP, ATP, phosphate,

oxoglutarate, citrate, glutamate, and malate.

COUPLING OF RESPIRATION

ATP SYNTHESIS

It has long been known that respiration and mitochon-

drial ATP synthesis are coupled. The observation that

respiration rate increased when mitochondria syn-

thesized more ATP led to the concept of respiratory

control by ADP phosphorylation. In fact, there is a link

between mitochondrial ATP synthesis and cellular ATP

demand by a feed-back mechanism. In agreement with

Mitchell’s theory, it was demonstrated that the mito-

chondrial electrochemical proton gradient, generated as

electrons are passed down the respiratory chain, is the

primary source for cellular ATP synthesis. In this way,

several complexes of the respiratory chain pump protons

outside of the inner membrane during reoxidation of

coenzymes and generate a proton gradient which is

consumed by the reactions of ATP synthesis. Proton leak

represents another mechanism consuming the mito-

chondrial proton gradient. Mitchell’s theory predicted

that any proton leak non coupled to ATP synthesis

would represent an uncoupling of respiration and result

in thermogenesis. An excellent example of such an

uncoupling of respiration to ADP phosphorylation is

represented by the mitochondrial UCP of brown

adipocytes (UCP1) which dissipates energy of substrate

oxidation as heat.

Brown Adipose Tissue and UCP1:

History of a True

Respiration Uncoupling

Maintenance of body temperature in a cold environment

or at birth requires thermogenesis. Another situation

requiring thermogenesis is arousal from hibernation.

The two regulatory thermogenic processes are shivering

and metabolic thermogenesis also referred to as non-

shivering thermogenesis. The largest part of nonshiver-

ing thermogenesis in small mammals is achieved in the

brown adipose tissue (BAT).

BROWN ADIPOSE TISSUE (BAT)

BAT is found in all small mammals and in the newborn

of larger mammals, such as humans. BAT is located in

speciﬁc body areas near large blood vessels and consists

of brown adipocytes which are distinct from white

adipocytes. The brown adipocytes are characterized by

numerous mitochondria containing a highly developed

inner membrane (Figure 1). Activation of thermogenesis

in BAT occurs in newborns, rodents exposed to cold, and

in animals emerging from hibernation. It is commanded

by the central nervous system and the orthosympathetic

ﬁbers innervating each brown adipocyte. The norepi-

nephrine released by these ﬁbers binds to adrenergic

receptors on the surface of the brown adipocytes. The

later steps of the activation of thermogenesis in brown

adipocytes are production of cyclic AMP, activation of

lipolysis, and oxidation of fatty acids by the numerous

mitochondria. Released fatty acids stimulate brown

adipocyte respiration and heat production.

RESPIRATION UNCOUPLING IS THE

THERMOGENIC MECHANISM

BROWN ADIPOCYTES

In most cell types such as muscular ﬁbers, ATP

utilization stimulates ADP phosphorylation and respir-

ation (Figure 2). Original observations made in 1967

revealed that the high respiratory rate in brown

adipocyte mitochondria was not controlled by mito-

chondrial ADP phosphorylation suggesting that energy

from substrate oxidation was dissipated into heat

instead of being converted in ATP (Figure 2). In line

with Mitchell’s chemiosmotic theory, if proton leakage

is activated by a signal such as free fatty acids in the

case of the mitochondria of brown adipocytes, the

mitochondrial membrane potential is decreased and

concomitantly, respiration is activated. Since, in

addition, the activated proton leakage is not linked

to ADP phosphorylation, respiration is uncoupled

from ATP synthesis and oxidation energy is dissipated

in the form of heat.

FIGURE 1 Histology of brown adipocyte. (A) The cytosol of brown

adipocytes is characterized by numerous mitochondria and lipid

droplets. (B) Magniﬁcation of a brown adipocyte mitochondrion

showing parallel cristae and UCP1 detected using antibodies (black

dots). (Figure kindly provided by Dr Saverio Cinti, University of

Ancona.)

FIGURE 2 Comparison of skeletal muscle and brown adipose tissue

mitochondria. (A) Mitochondria are cellular organelles converting the

redox energy liberated during oxidation of organic substrates into ATP,

a molecule containing energy under a form readily usable by most

enzymes working to maintain cell structure and integrity or to perform

work such as mechanical work in muscle. The cellular need for ATP

controls oxidation rate by mitochondria. (B) In brown adipocyte, a

speciﬁc uncoupling protein referred to as UCP (precisely UCP1) is

present in a large amount in the inner membrane. UCP1 is able to fool

mitochondria which accelerates their oxidation rate whereas no ATP is

produced. UCP1 removes the control of energy expenditure by ATP

demand and therefore oxidation rate and energy expenditure increase

under the form of heat. This process is used in newborns, arousal from

hibernation, and small mammals exposed to cold when muscle

thermogenesis would not be sufﬁcient.

314 UNCOUPLING PROTEINS

THE UCP OF BROWN

ADIPOCYTE MITOCHONDRIA

The protein responsible for proton leakage within the

inner mitochondrial membrane and for respiratory

uncoupling was identiﬁed in 1976–77, was puriﬁed,

and its cDNA was cloned. It has an apparent molecular

weight of 33 000 and was originally called UCP. This

protein is known as UCP1 since the discovery of UCP2 in

1997. UCP1 is abundant in the inner membrane of the

mitochondria of brown adipocytes and is speciﬁc to

these cells. It short-circuits the proton circuit between

complexes of the respiratory chain and the mitochon-

drial ATP-synthase which is the main consumer of the

proton gradient (Figure 3). When there is no need for

thermogenesis, purine nucleotides bound to UCP1

inhibit its activity. Upon activation of thermogenesis

by the sympathetic nervous system, fatty acids overcome

the inhibitory effect of nucleotides and activate UCP1.

Heterologous expression of UCP1 in yeasts or mamma-

lian cell lines induces respiration uncoupling. The

protonophoric activity of UCP1 has been reconstituted

in liposomes where it can be inhibited by nucleotides and

activated by free fatty acids. The mechanism of action of

UCP1 is still controversial: some scientists believe that it

is a true proton transporter, whilst others assert that it

returns anionic fatty acids to the intermembrane space,

after they have crossed the membrane in protonated form.

The observation that Ucp12 /2 mice were unable

to maintain body temperature in the cold proved that

the UCP1-induced uncoupling of respiration was

responsible for cold-activated nonshivering thermogen-

esis of rodents. In agreement with this observation,

ectopic expression of UCP1 in skeletal muscles of

transgenic mice promotes substrate oxidation in muscles

and resistance to obesity and type 2 diabetes. UCP1

functions as a dimer and each monomer is made of six

transmembrane fragments (Figure 4). Moreover, UCP1,

the ADP/ATP translocator and other mitochondrial

anion carriers derive from the same ancestral gene and

probably share a similar structure. They all have a

triplicated structure (Figure 4).

The Novel UCPs

Given their speciﬁc role in thermogenesis, it has always

seemed logical that brown adipocytes be equipped with

an original mechanism, partial uncoupling of respir-

ation, brought into play by a speciﬁc protein, UCP1,

which induces proton leakage. In fact, it is known that

mitochondrial respiration is always accompanied by

heat production as it is imperfectly coupled to ADP

phosphorylation in all types of cells. To explain this

incomplete coupling of respiration and the energy loss

mechanism, some authors have invoked slippage of the

respiratory chains, while others referred to proton leaks.

It has been estimated that proton leaks from the

inner membrane of mitochondria of hepatocytes

FIGURE 4 Model of insertion of UCP in the membrane. The

determination of structure of the UCPs remains difﬁcult because

crystallization of such membrane proteins has not been successful yet.

To modelize the organization of such proteins, indirect methods such

as prediction of secondary structure by computers can be used.

Experimentally, restricted domains of the protein in the membrane

(indicated by stars) were explored using speciﬁc antibodies. The UCPs

are made of six transmembraneous segments (probably alpha helices),

and three hydrophilic loops stemming from the matricial side of the

membrane. An internal repetition in the structural arrangement of the

protein suggests that the protein (made of 300 amino acid residues)

derived from the ancient triplication and divergence of a 100 amino

acid domain (made of 2 helices and 1 loop).

FIGURE 3 UCP1-induced proton cycling. Q (coenzyme Q) refers to

complex II which is the succinate dehydrogenase. This enzyme directly

reduces coenzyme Q but has no proton pumping activity contrary to

complexes I, III and IV. UCP is inserted in the mitochondrial inner

membrane where, also present, is a multienzymatic complex called the

respiratory chain made of complexes I to IV. The respiratory chain

reoxidize reduced coenzymes and electrons are driven to oxygen. This

oxido-reduction step liberates energy which is used to generate an

electrochemical gradient of protons across the inner membrane. This

gradient is normally consumed by the ATP-synthase which phosphor-

ylates ADP. UCP1 transports protons passively and makes possible a

futile cycle of protons across the inner membrane leading to increased

energy expenditure. This schema illustrates the situation encountered

in brown adipocytes of mammals where a large amount of respiratory

chains as well as a large amount of UCP1 are present. Activation of the

futile cycling increases considerably energy expenditure and thus heat

production in these thermogenic cells. In other cells where homologues

of the UCP1 are expressed at much lower level, this pathway would

represent a minor contributor to energy expenditure, but might be of

importance to avoid oxidative damage.

UNCOUPLING PROTEINS 315

and myocytes could explain about 20% of the body’s

basal metabolism.

UCP2: A HOMOLOGUE OF THE BROWN

FAT UCP PRESENT IN VARIOUS TISSUES

A clone corresponding to a protein with 59% sequence

identity with the UCP of brown fat was isolated. This

new UCP, called UCP2, was also identiﬁed in humans. In

contrast with UCP1, the mRNA of UCP2 is present in

almost all tissues and many cell types, such as spleen,

thymus, digestive tract, lung, brain, adipose tissues,

skeletal muscle, heart, adipocytes, myocytes, and

macrophages (Figure 5). Expressed in yeasts, murine

UCP2 decreased the potential of the mitochondrial

membrane, raised the respiration rate and reduced

sensitivity to uncouplers: UCP2 therefore was able to

uncouple respiration and appeared to be a second

mitochondrial UCP. The UCP2 gene is located on

chromosome 7 of the mouse and chromosome 11 of

humans, near a region linked to hyperinsulism and

obesity. The expression of UCP2 mRNA was measured

in obesity-prone and obesity-resistant mice given a lipid-

rich diet: the obesity-resistant mice overexpressed UCP2

mRNA in their adipose tissue. These ﬁndings, together

with the function of the protein and the chromosomal

location of its gene, led to propose a role for UCP2 in

diet-induced thermogenesis. However, this hypothesis

was not validated in Ucp22 /2 mice.

UCP3: ANOTHER UCP HOMOLOGUE

PREDOMINANTLY PRESENT

SKELETAL MUSCLE

cDNAs corresponding to UCP3, a protein homologous

to UCP1 and UCP2, were cloned soon after the

discovery of UCP2. The amino acid sequence of UCP3

is 72% identical to that of UCP2 and 57% identical to

that of UCP1. UCP3 mRNA is predominantly expressed

in the skeletal muscles of humans, mice, and rats

(Figure 5). Human and murine UCP2 and UCP3 genes

are juxtaposed on the same chromosome.

PLANT AND BIRD UCPS

Functional studies of isolated plant mitochondria

suggested that plants contain mitochondrial UCPs.

Following this observation, a cDNA from a new UCP

was isolated from a potato cDNA library. The sequence

of the corresponding protein, called stUCP, is 44%

identical to that of UCP1 and 47% identical to that of

UCP2. The mRNA of stUCP is present in most plant

organs. The most surprising result was that stUCP

mRNA was markedly induced in the leaves of plants

exposed to a temperature of 48C. Cold, therefore,

induced stUCP as it induces UCP1 in the brown adipose

tissue of animals. However, plant and animal UCPs may

have different functions. More recently, a chicken UCP

termed avUCP was characterized. This gene is only

expressed in skeletal muscles of birds and ducks and is

induced upon exposure to the cold. Therefore avUCP

seems to be involved in regulatory thermogenesis.

Other proteins referred to as UCP4 and BMCP1/UCP5

were characterized. However their exact relationship

to UCP1, UCP2, and UCP3 will require further work.

ROLE AND FUNCTION OF UCPSOTHER

THAN

UCP1: A ROLE IN CONTROLLING

THE

LEVEL OF REACTIVE

OXYGEN SPECIES

Numerous questions remain unanswered, notably con-

cerning the exact catalytic activity of each UCP, and the

nature of the endogenous ligands of UCP2 and UCP3.

These proteins may simply translocate protons. As said

above for UCP1, it has been proposed that all the UCPs

are active as fatty acid cycler through the membrane.

According to this later hypothesis, protonated fatty

acids cross the membrane and release a proton on the

matricial side; then, the UCP facilitates the translocation

of the anionic form of fatty acids. High loads of UCP2 or

UCP3 can uncouple respiration in yeast or mammalian

cell, but it is not certain whether the low physiological

levels of these proteins may be sufﬁcient to induce a net

uncoupling of respiration in vivo.Thedivergent

observations of the regulation of the activities of UCP2

or UCP3 by fatty acids or nucleotides will require further

studies. Interestingly, it has been proposed that UCP2

and UCP3 could export fatty acid anion under con-

ditions of elevated fatty acid oxidation.

FIGURE 5 Tissular distribution of UCP1, UCP2, and UCP3 RNAs.

The data correspond to mouse tissues. A very similar picture was

obtained when analyzing human tissues. 18S refers to ribosomal RNA.

316 UNCOUPLING PROTEINS

Metabolic Activity of UCP2 and UCP3

Conﬂicting data regarding the association of UCP2 or

UCP3 genetic polymorphisms to body mass index,

susceptibility to obesity, resting metabolic rate, meta-

bolic efﬁciency, fat oxidation, insulin resistance, suscep-

tibility to gain fat with age were obtained. However,

UCP2 appears to act as a negative regulator of insulin

secretion. Moreover, mice overexpressing a large

amount of human UCP3 in skeletal muscle weigh less,

have a decreased amount of adipose tissue, and an

increased resting oxygen consumption.

Reactive Oxygen Species

Guided by the fact that UCP2 is expressed at high level

in the immune system, a role for UCP2 in immunity was

searched for. In fact, Ucp2 2 /2 mice are more resistant

to infection by a parasite. This phenotype was explained

by the fact that disruption of the Ucp2 gene provoked an

elevation of ROS level that facilitated the killing of the

pathogens. Mice lacking the Ucp3 gene also produced

more oxygen radicals in myocytes. Therefore, it appears

that UCP2 and UCP3 contribute to limit the level of

reactive oxygen species in cells. Such an activity may be

explained by a mild uncoupling activity of the proteins

since a small decrease of the mitochondrial membrane

potential opposes production of superoxide ion in

mitochondria (Figure 6). The observation that UCP2

protects against atherosclerosis highlights its ability to

counteract radical synthesis.

UCP1, UCP2, and UCP3,

Conclusions and Perspectives

UCP1 has a well-demonstrated uncoupling activity and

an essential role in maintenance of body temperature in

small rodents exposed to the cold, but the exact

biochemical and physiological roles of UCP2 and

UCP3 remain to be further identiﬁed. Certain data

support a true activity of these UCPs in respiration

uncoupling and substrate oxidation, other data do not

agree with a role for these UCPs in controlling adiposity.

Several reports support a role for UCP2 and UCP3 in

decreasing the level of reactive oxygen species in cells.

FIGURE 6 Role of UCP and superoxide generation by mitochondria.

(A) mitochondria phosphorylating ATP: a proton gradient generator

(the respiratory chain, left square) uses the energy liberated by the

movement of electrons to pump protons through the membrane (grey

zone) towards the intermembrane space. A proton gradient consumer

(the ATP-synthase, right square) phosphorylates ADP coming from

outside into ATP, that is exported afterwards. Therefore, a proton cycle

across the inner membrane couples oxidation to ATP synthesis (and

vice versa). (B) and (C) when oxidation is impaired (B) or slows down

(C), an excess of electrons aiming to travel across the respiratory

complexes cannot reach oxygen (to form water). Single electrons react

with molecular oxygen and form superoxide anion (O2

). It occurs

either when the respiratory chain is poisoned (red arrow in panel B) or

under physiological circumstances when the use of ATP by the cell is

weak, ATP level rises, and ADP level decreases (panel C). In that case,

the membrane potential rises to its highest value which impedes proton

pumping, slows down reoxidation, and therefore increases superoxide

production. (D) in tissues other than the thermogenic brown adipose

tissue, the UCPs may induce a controlled leak of protons across the

inner membrane that allows reoxidation to occur and maintain the

membrane potential below the threshold inducing superoxide gener-

ation. The mechanism of proton transport (proton channeling or fatty

acid cycling), as well as the identity of regulators of UCPs inside the

cells are still controversial. Fatty acids would be cofactors of the proton

transport whereas nucleotides are inhibitors. The membrane potential

itself is a variable inﬂuencing heavily UCP1 activity. It has been

proposed that superoxide anion and quinones directly activate proton

transport by UCPs.

UNCOUPLING PROTEINS 317

Whilst UCP1 is present at a high level, an open question

is that of the exact amount of UCP2 and UCP3 in cells

since whatever is the exact intrinsic ability of these

proteins to uncouple respiration from ATP synthesis, a

certain amount of the proteins will be required to

achieve a physiological uncoupling. Finally, the possible

contribution of a UCP to the hypermetabolic syndrome

(known as Luft’s disease) previously described in

patients remains to be investigated.

FURTHER READING

Boss, O., Hagen, T., and Lowell, B. B. (2000). Uncoupling proteins 2

and 3. Potential regulators of mitochondrial energy metabolism.

Diabetes 49, 143–156.

Cannon, B., and Nedergaard, J. (1985). The biochemistry of an inefﬁ-

cient tissue: Brown adipose tissue. Essays Biochem. 20, 110– 164.

Echtay, K. S., Winkler, E., and Klingenberg, M. (2000). Coenzyme Q is

an obligatory cofactor for uncoupling protein function. Nature

408, 609– 613.

Enerba

ck, S., Jacobsson, A., Simpson, E. M., Guerra, C., Yamashita,

E. H., Harper, M. E., and Kozak, L. P. (1997). Mice lacking

mitochondrial uncoupling protein are cold-sensitive but not obese.

Nature 387, 90 –94.

Garlid, K. D., and Jaburek, M. (1998). The mechanism of proton

transport mediated by mitochondrial uncoupling proteins. FEBS

Lett. 438, 10 –14.

Himms-Hagen, J., and Ricquier, D. (1998). Brown adipose tissue. In

Handbook of Obesity (G. Bray, C. Bouchard and W. P. T. James,

eds.) pp. 415 –441. Marcel Dekker, New York.

Ledesma, M., Garcia

de Lacoba, M., and Rial, E. (2002). The

mitochondrial uncoupling proteins. Genome Biol. 3 (on line

November 29).

Nicholls, D. G., and Locke, R. M. (1984). Thermogenic mechanisms in

brown fat. Physiol. Rev. 64, 1–64.

Ricquier, D., and Bouillaud, F. (2000). The uncoupling protein

homologues: UCP1, UCP2, UCP3, StUCP & AtUCP. Biochem. J.

345(Pt 2), 161–719.

Skulachev, V. P. (1988). Uncoupling: New approaches to an old

problem of bioenergetics. Biochim. Biophys. Acta 1363, 100–124.

Stuart, J. A., Cadenas, S., Jekabsons, M. B., Roussel, D., and Brand,

M. D. (2001). Mitochondrial proton leak and the uncoupling

protein 1 homologues. Biochim. Biophys. Acta 1504, 144– 158.

BIOGRAPHY

Daniel Ricquier is Professor at Neckes-Enfants Malades Faculty of

Medicine and Fre

ric Bouillaud is Main Investigator at the Institut

National de la Sante

et de la Recherche Me

dicale. They work at CNRS

Unit 9078 at Faculte

de Me

decine and Institut de Recherches Necker-

Enfants Malades in Paris. They hold Ph.Ds from Pierre and Marie

Curie University in Paris. Their laboratory ﬁrst cloned UCP1, UCP2,

BMCP1/UCP5, and avUCP. They also collaborated to the character-

ization of the ﬁrst plant UCP. With their collaborators, they are authors

of original reports dealing with biochemistry, physiology, and genetics

of the UCP family.

318 UNCOUPLING PROTEINS

Unfolded Protein Responses

David Ron

New York University School of Medicine, New York, USA

The unfolded protein response (UPR) is a transcriptional and

translational response to the accumulation of unfolded proteins

in the endoplasmic reticulum (ER). The UPR is mediated by

highly conserved signaling pathways that are activated by

imbalance between the load of unfolded (or malfolded) ER

client proteins and the capacity of the organelle to process this

load. Collectively these pathways restore equilibrium to the

protein-folding environment in the organelle by increasing the

expression of genes that enhance nearly all aspects of ER

function and by transiently repressing the biosynthesis of new

client proteins. Interest in the UPR has been stimulated by the

realization that postsynthetic protein processing constitutes an

important step in gene expression and that protein malfolding

plays an important role in human disease.

ER Function and ER Stress

THE ER, A PROTEIN-PROCESSING

MACHINE

Proteins destined for secretion and membrane insertion

are translocated across the ER membrane in an unfolded

state. In the ER lumen they undergo chaperone-assisted

folding, a variety of organelle-speciﬁc post-translational

covalent modiﬁcations and often chaperone-assisted

assembly into oligomeric structures. Once properly

folded and assembled, most ER client proteins are

packaged into vesicles that are transported to more

distal sites in the secretory pathway. Proteins that fail to

attain their proper folded and oligomeric conformation

are retained in the ER by continued binding to

chaperones and are ultimately translocated from the

organelle to the cytoplasm for proteasomal degradation,

in a process known as ER-associated protein degra-

dation (ERAD).

To fulﬁll these various functions the ER is endowed

with a unique complement of proteins. These include

enzymes for post-translational covalent modiﬁcations

(e.g., N-linked glycosylation, disulﬁde bond formation),

chaperones that assist in folding and assembly steps,

translocation channels and transporters involved in

transmembrane trafﬁc, and various components

involved in the postassembly steps of client protein

egress from the organelle. The build-up of these

components deﬁnes the capacity of the organelle to

handle its client proteins. In the late 1980s Sambrook,

Gething, and their colleagues discovered that manipula-

tions that interfere with the function of various aspects

of the ER client protein-handling machinery, and

thereby perturb protein folding in the ER, selectively

up-regulate the expression of genes that encode com-

ponents of that machinery. The extent of this transcrip-

tional response was not fully appreciated at the time,

however its selectivity was noted, hence its name: the ER

unfolded protein response (or UPR).

It was imagined that the cell possessed means to

monitor the load of client proteins presented to its ER

and responded to an increase in such load by up-

regulating the capacity of its ER to process unfolded

client proteins. An additional clue to the workings of this

response was provided by the seminal observation that

forced overexpression of an ER chaperone, BiP (also

known as GRP78) markedly suppressed the activity of

the UPR. Thus, at least one target gene of the UPR (BiP)

is able to exert negative feedback on the entire response.

BiP overexpression does not restore function to a

challenged ER; this requires the coordinate expression

of numerous UPR target genes. However, BiP, which is a

member of the highly conserved HSP70/DnaJ family of

chaperones, promiscuously recognizes hydrophobic

stretches of amino acids. These are normally incorpor-

ated into the cores of properly folded polypeptides and

assembled oligomeric complexes, but remain exposed

on the surface of unfolded, malfolded or unassembled

proteins. The ability of BiP to suppress the UPR

suggested that it might be doing so by masking a stress

signal generated by many different unfolded and

malfolded proteins. This phenomenon was assigned

the heuristic term ER stress and its level reﬂects the

balance between client protein load and the capacity of

the organelle to process that load.

PHYSIOLOGICAL AND

PATHOLOGICAL ER STRESS

To the extent that cell types vary in the load of

secreted proteins that they are called upon to produce,

they are subject to widely different levels of physio-

logical ER stress. This explains enhanced activity of

the UPR in various professional secretory cells, such as

pancreatic cells of vertebrates or intestinal cells of the

nematode, Caenorhabditis elegans.ERstressalso

occurs when a mutation in an abundantly expressed

ER client protein renders that protein especially

difﬁcult to fold. An example is provided by various

degenerative diseases affecting myelinated neurons in

which mutations in a component of the myelin sheath

(an ER client protein) cause the protein to malfold and

induce high levels of ER stress which, over time,

destroys the myelin-producing cell. It is important to

emphasize that most mutations that impede folding of

ER client proteins do not cause measurable ER stress.

However, because they diminish expression of the

properly folded protein, such mutations may deprive

the organism of the latter’s beneﬁcial actions. This

genetic mechanism underlies such serious human

diseases as cystic ﬁbrosis, familial hypercholesterol-

emia, and hemophilia. Nonetheless, the level of

expression of the mutant protein is not enough to

globally challenge ER function in the cell that produces

it and the phenotypic expression of the mutation

reﬂects the lack of an important protein, rather than

the production of a toxic one.

The above constitute client protein-driven ER stress,

however ER stress may also initiate from impaired

function of the organelle, which occurs in cells

deprived of energy sources or oxygen, or in cells

exposed to certain ER-speciﬁc toxins. We do not

understand in detail how ER stress contributes to

further organelle dysfunction and ultimately cell death.

However a framework for thinking about this has

recently emerged with the realization that unfolded

and malfolded proteins present reactive interfaces that

have not been vetted by evolution and may thus

disrupt the cellular machinery by interacting promis-

cuously and illegitimately with essential cellular

components. According to this theory, proteotoxicity

is normally held in check by the chaperones that bind

such potentially toxic protein interfaces and let go only

once the latter have been buried in the hydrophobic

cores of the properly folded client protein. ER stress

(by deﬁnition) challenges the capacity of the chaper-

ones and may permit illegitimate protein interfaces

to emerge. The ability of chaperone overexpression to

suppress ER stress signaling suggests that the need

to prevent proteotoxicity is an important driving force

in evolution of the UPR. The unifying feature of ER

stress need not be the presence of toxic moieties on the

surface of every unfolded or malfolded protein. The

common feature of diseases of protein folding might

instead be the exhaustion of a protective chaperone

reserve, which normally suppresses the potential

proteotoxicity of certain (possibly normal) folding

intermediates of ER client proteins.

The Yeast Unfolded

Protein Response

IRE1, A PROTOTYPE OF

TRANSMEMBRANE STRESS SIGNALING

Early studies on the UPR were carried out in the yeast,

Saccharomyces cerevisiae, an organism that lends itself

well to forward genetic screens. To screen for mutations

affecting the UPR, the regulatory region of yeast BiP

gene (KAR2) was fused to a reporter and mutant yeast

with suppressed activity of this reporter were sought.

The ﬁrst gene thus identiﬁed was inositol requiring 1

(IRE1), so-called because its loss of function had been

previously noted to result in inositol auxotrophy. IRE1

encodes a transmembrane ER resident protein with an

N-terminal domain residing in the ER lumen and a C-

terminal domain that is exposed on the cytoplasmic

side. The membrane topology and subcellular localiz-

ation of IRE1 immediately suggest a mechanism for

transmitting information on the state of the ER

(topologically equivalent to the extracellular space) to

the cell’s interior. The lumenal domain somehow senses

ER stress, conveying the signal across the ER mem-

brane to the cytoplasmic domain, which broadcasts it

to the nucleus, turning “on” UPR target gene

expression (Figure 1).

IRE1’s C-terminal, cytoplasmic, effector-domain, is a

protein kinase and undergoes autophosphorylation

when activated by ER stress. This suggested that IRE1

might function like other transmembrane receptors that

are also protein kinases and convey their signal by

phosphorylating downstream targets, often through a

kinase relay. However, other than IRE1 itself no

substrates for the aforementioned kinase have been

identiﬁed to date.

UNCONVENTIONAL SPLICING

HAC1 MRNA

The clues to understanding propagation of the UPR

signal, downstream of IRE1 again came from yeast

genetics. Mutations in two additional genes were

noted to block the yeast UPR. One of these, HAC1,

encodes a transcription factor, which binds to and

activates the promoters of the primary target genes

of the yeast UPR, the second, more mysteriously

turned out to be RLG1, which encodes for a multi-

functional enzyme involved in tRNA splicing. ER

stress promotes accumulation of HAC1 protein and

this is blocked by mutations in IRE1 and RLG1.

320

UNFOLDED PROTEIN RESPONSES

The underlying mechanisms, which turned out to be

full of surprise, were revealed in a series of

brilliant experiments carried out in the Walter and

Mori laboratories.

The regulated step in HAC1 expression entails the

removal of a 242 base internal segment at the 3

end

of the mature mRNA. The enzyme which precisely

cuts the mRNA at the 5

and 3

ends of this segment

turned out to be IRE1 itself, whose highly sequence-

speciﬁc endoribonucleolytic activity appears to be

subordinate to its kinase activity (though the molecu-

lar details remain to be worked out). The two ends of

the severed HAC1 mRNA are held together by

extensive base pairing and are subsequently religated

by the protein product of RLG1. IRE1 is the master

regulator of this unconventional mRNA splicing

reaction, since its kinase and endoribonuclease

are activated by ER stress. The removal of the

aforementioned internal segment of the HAC1

mRNA de-represses HAC1 translation and the protein

encoded by the processed mRNA is more stable and

transactivates its target genes more potently than that

encoded by the unprocessed mRNA.

Mutations in IRE1 and HAC1 similarly abolish most

signaling in the yeast UPR. This was convincingly

demonstrated by expression proﬁling which showed

that the vast majority of the genes induced by exposing

yeast to toxins that promote ER stress were no longer

activated in IRE1 or HAC1 mutant strains. Thus the

yeast UPR consists of a linear pathway in which HAC1

is an essential target of IRE1 and together the two

proteins are required for the activation of most known

UPR target genes.

Diversiﬁcation in the

Metazoan UPR

CONSERVATION OF THE IRE1 PATHWAY

Systematic genomic sequencing of the metazoan

C. elegans and random cDNA sequencing from several

mammalian species, predicted proteins highly related to

yeast IRE1 in higher eukaryotes. Their common

membrane topology, localization to the ER, and

conserved kinase and endoribonuclease domains

suggested that they were indeed IRE1 homologues.

However, HAC1 homologues could not be identiﬁed by

simple sequence comparison to the yeast. Eventually,

however, a combination of forward genetic screens in

C. elegans and creative guesswork led to the identiﬁ-

cation of XBP-1 as a functional homologue of HAC1 in

higher eukaryotes. It too encodes a transcription factor

whose expression is tightly regulated by IRE1 through

the processing of its mRNA.

Less conserved, however, is the role of the IRE1

pathway in the yeast and metazoan UPR. Whereas in

the former, IRE1 and HAC1 are absolutely required

for activating nearly all UPR target genes, in mam-

mals, IRE1 and XBP-1 are dispensible for activation of

FIGURE 1 IRE1 and unconventional splicing of its target mRNA signal the unfolded protein response. ER stress leads to oligomerization and

trans-autophosphorylation of IRE1 (indicated by the “P” on its cytoplasmic effector domain). Phosphorylation unmasks the effector function of

IRE1 (cartooned by the red bristles on the cytoplasmic domain), which consists of the endoribonucleolytic processing of its target mRNA, XBP-1 in

metazoans, and HAC1 in yeast. The two ends of the cleaved mRNA are joined together by a ligase. The unconventionally spliced XBP-1/HAC1

mRNA is more efﬁciently translated than the unprocessed mRNA and encodes a protein that is more stable and more effective at trans-activation of

UPR target genes in the nucleus.

UNFOLDED PROTEIN RESPONSES 321

all but a very small set of UPR targets. This is all the

more remarkable when one considers that most UPR

target genes are conserved between yeast and mammals,

it is just that the latter have found alternative means

to couple their expression to ER stress. C. elegans

occupies an intermediate position between yeast

and mammals, in that most of its identiﬁable UPR

targets are partially dependent on IRE1 and XBP-1,

but can to some extent also be activated independently

of that pathway.

Despite this redundancy in the mammalian UPR,

both IRE1 and XBP-1 are essential for embryonic

development (mammals have two IRE1 genes, a

nonessential beta isoform restricted in its expression

to the intestinal and bronchial epithelium and a

broadly expressed alpha isoform which is essential).

The reason(s) for the embryonic lethality of IRE1

and

XBP-1 null animals are not fully understood, but it is

hypothesized that in mammals the pathway may have

diverted to specify the capacity for especially high

levels of protein secretion. This hypothesis, which is

clearly in need of further experimental support,

suggests that speciﬁcation of a secretory cell fate

activates a conserved stress pathway that drives a

developmental process, namely acquisition of an

apparatus required by professional secretory cells

for high-capacity protein secretion. This idea is

further supported by the observation that many

UPR target genes function far downstream in the

secretory pathway rendering it unlikely that such genes

merely relieve the stressed ER of its load of unfolded

client proteins.

ATF6 AND INTRAMEMBRANE

PROTEOLYSIS IN THE METAZOAN UPR

The identiﬁcation of ATF6 by Mori and colleagues

conﬁrmed the predicted redundancy in the metazoan

UPR. Like IRE1, ATF6 is also an ER-localized trans-

membrane protein. However, its cytoplasmic effector

domain consists of a transcription factor that activates

UPR target genes directly. In its membrane-bound form,

ATF6 is inert, as it cannot reach the nucleus. Activation

involves regulated intramembrane proteolysis, which

liberates the transcription factor part of ATF6 from the

ER membrane under conditions of ER stress (Figure 2).

The next surprise came when Brown, Goldstein, Prywes,

and colleagues identiﬁed the proteases involved in this

highly regulated event. These turned out to be the same

proteases that process and activate the membrane-

bound transcription factor, SREBP, which regulates

genes involved in sterol and fatty acid biosynthesis

and assimilation.

There are no mammalian gene knockouts reported

for ATF6 (of which at least two isoforms exist).

However, unlike cells lacking IRE1 or XBP-1, cells

lacking the proteases required for ATF6 processing are

severely impaired in UPR target gene expression. It

seems likely therefore that mammalian ATF6 has

picked up some of the role performed by IRE1 and

XBP-1/HAC1 in simpler organisms. The similarities

between ATF6 and the SREBPs highlight another

remarkable feature of the UPR. IRE1 and HAC1

mutant yeast are unable to synthesize adequate

amounts of the membrane phospholipid precursor,

inositol. Though the details remain to be worked out,

insufﬁciency of membrane lipid components also

signals through the yeast UPR and enzymes involved

in phospholipid metabolism are targets of IRE1 and

HAC1. In metazoans this aspect of the UPR apparently

has been split off and relegated to dedicated transcrip-

tion factors, the SREBPs that are activated by insufﬁ-

ciency of lipid components. However the ancient

link between the client protein-based UPR and mem-

brane lipid insufﬁciency signaling has remained in the

form of shared machinery for proteolytic activation of

ATF6 and SREBP.

Translational Control in the UPR

PERK COUPLES ER STRESS TO EIF2

PHOSPHORYLATION AND

TRANSLATIONAL REPRESSION

In addition to activated gene expression ER stress

results in a dramatic reduction in protein synthesis.

Translational repression is an active process limiting

the inﬂux of client proteins into the stressed ER and

thus serves as a counterpart to the gene expression

program, which increases the organelle’s capacity to

process client proteins. Brostrom and colleagues and

Kaufman and colleagues noted, early on, that transla-

tional repression by ER stress is associated with

phosphorylation of the

-subunit of translation

initiation factor 2 (eIF2

) on serine 51. This phos-

phorylation site is conserved in all eukaryotes and

serves to regulate translation initiation in diverse

stressful conditions. Trimeric eIF2, in complex with

guanosine triphosphate (GTP), recruits the amino-

acylated initiator methionyl –tRNA to the small

ribosomal subunit, allowing translation initiation.

Recognition of an AUG initiation codon on the

mRNA leads to hydrolysis of GTP to GDP and

dissociation of the ribosome–eIF2 complex. To par-

ticipate in another round of translation initiation, the

GDP bound to eIF2 must be exchanged to GTP. The

enzyme catalyzing this exchange reaction (eIF2B) is

inhibited by phosphorylated eIF2.

Distinct eIF2

kinases were known to be activated

by distinct stress signals; PKR by double-stranded RNA

322

UNFOLDED PROTEIN RESPONSES