Tài liệu Báo cáo khoa học: The heat shock factor family and adaptation to proteotoxic stress pdf

Importance of the expression of the nonclassical heat shock genes was evidenced by the fact that mouse HSF3 and chicken HSF1 play a sub-stantial role in the protection of cells from heat

The heat shock factor family and adaptation to

proteotoxic stress

Mitsuaki Fujimoto and Akira Nakai

Yamaguchi University School of Medicine, Ube, Japan

Introduction

All living organisms respond to elevated temperatures

by producing a set of highly conserved proteins,

known as heat shock proteins (HSP) [1] This response

is called the heat shock response, and is a universal

mechanism of protection against proteotoxic stress,

including heat shock and oxidative stress In

Escheri-chia coli, heat shock genes are under the control of a

specific transcription factor, r32, which directs the

core RNA polymerase to promoters [2] In eukaryotes,

the heat shock response is regulated mainly at the level

of transcription by heat shock factors (HSFs) [3] Heat shock genes, such as HSP110, HSP90, HSP70, HSP40 and HSP27, contain heat shock elements (HSEs) com-posed of at least three inverted repeats of the highly conserved consensus sequence nGAAn in the proximal promoter region [4] Here we call them ‘classical heat shock genes’, which encode major HSPs or molecular chaperones Heat shock triggers the conversion of an HSF1 monomer in a metazoan species that is nega-tively regulated by HSPs into a trimer that binds to

Keywords

evolution; heat shock; protein homeostasis;

protein-misfolding disorder; transcription

factor; vertebrate

Correspondence

Akira Nakai, Department of Biochemistry

and Molecular Biology, Yamaguchi

University School of Medicine,

Minami-Kogushi 1-1-1, Ube 755-8505, Japan

Fax: 81 836 22 2315

Tel: 81 836 22 2214

E-mail: anakai@yamaguchi-u.ac.jp

(Received 10 May 2010, revised 7 July

2010, accepted 23 July 2010)

doi:10.1111/j.1742-4658.2010.07827.x

The heat shock response was originally characterized as the induction of a set of major heat shock proteins encoded by heat shock genes Because heat shock proteins act as molecular chaperones that facilitate protein fold-ing and suppress protein aggregation, this response plays a major role in maintaining protein homeostasis The heat shock response is regulated mainly at the level of transcription by heat shock factors (HSFs) in eukary-otes HSF1 is a master regulator of the heat shock genes in mammalian cells, as is HSF3 in avian cells HSFs play a significant role in suppressing protein misfolding in cells and in ameliorating the progression of Caenor-habditis elegans, Drosophila and mouse models of protein-misfolding disorders, by inducing the expression of heat shock genes Recently, numer-ous HSF target genes were identified, such as the classical heat shock genes and other heat-inducible genes, called nonclassical heat shock genes in this study Importance of the expression of the nonclassical heat shock genes was evidenced by the fact that mouse HSF3 and chicken HSF1 play a sub-stantial role in the protection of cells from heat shock without inducing classical heat shock genes Furthermore, HSF2 and HSF4, as well as HSF1, shown to have roles in development, were also revealed to be neces-sary for the expression of certain nonclassical heat shock genes Thus, the heat shock response regulated by the HSF family should consist of the induction of classical as well as of nonclassical heat shock genes, both of which might be required to maintain protein homeostasis

Abbreviations

BRG1, brahma-related gene 1; DAF-16, abnormal dauer formation 16; HR, hydrophobic heptad repeat; HSE, heat shock element; HSF, heat shock factor; HSP, heat shock protein; MEF, mouse embryonic fibroblast; polyQ, polyglutamine.

the HSE with high affinity, and the bound HSF1

rap-idly induces a robust activation of the classical heat

shock genes [5,6]

There is a single gene encoding HSF in yeast, in

Caenorhabditis elegans and in Drosophila HSF is

required not only for the heat shock response, but also

for cell growth and differentiation in yeast [7] In

verte-brates, there are multiple HSF genes, which encode

members of the HSF family (HSF1–4) In mammals,

as in yeast and Drosophila, the HSF1 is required for

the heat shock response, whereas HSF3 is required for

this response in avian species [8,9] Both mouse HSF1

and chicken HSF3 are necessary for thermotolerance,

at least through the expression of classical heat shock

genes [10,11] In addition to their role in the heat

shock response, mouse HSFs are critical in

develop-mental processes such as gametogenesis and

neuro-genesis, in the maintenance of sensory and ciliated

tissues, and in immune responses [12–14] HSF1- and

HSF3-mediated mechanisms of cellular adaptation to

heat shock have been analyzed in detail in chicken

cells, and were considered specific to chicken cells as it

was believed until recently that HSF3 was an

avian-specific factor In this minireview, we summarize the

evolution of the HSF gene family and HSF-mediated

mechanisms of cellular adaptation to stress in

verte-brates by comparing mammalian and avian cells, and

also review HSF-mediated mechanisms of adaptation

to pathological states related to protein misfolding

Evolution of the vertebrate HSF gene

family

An HSF protein that binds to the HSEs in the HSP

genes was purified from heat shock-induced

Saccharo-myces cerevisiae, Drosophila and human cells [3]

Anti-bodies against HSF were used to isolate a single copy

of the S cerevisiae HSF gene [15–17] Thereafter, a

single HSF gene was isolated from another budding

yeast Kluyveromyces lactis, and from the fission yeast

Schizosaccharomyces pombe by cross-hybridization

[18,19] A single copy of the Drosophila HSF was

iso-lated by screening a library with oligonucleotide

probes derived from HSF peptide sequencing In

mam-mals, human HSF1 and a second HSF gene, HSF2,

were isolated by screening a library with degenerate

oligonucleotide probes [20,21], and the mouse HSF1

and HSF2 genes were isolated by cross-hybridization

with a human HSF1 cDNA probe [22] In chicken,

HSF1, HSF2 and a third HSF gene, HSF3, were

iso-lated by cross-hybridization with a mouse HSF1

cDNA probe [23] Furthermore, another HSF gene,

HSF4, was isolated from human and mouse cells by

the screening of human and mouse cDNA libraries with a chicken HSF3 cDNA probe [24,25], but a mam-malian orthologue of the chicken HSF3 gene was not identified Therefore, HSF3 was considered specific to avian species, and HSF4 was considered specific to mammalian species [5,8,9,12]

Although human and mouse genome sequences have become available [26,27], no HSF3-related sequence was identified in silico from the genome database [28] However, analysis of the chicken genome enabled com-parison of the syntenic regions [29], where the same genes occur in a similar order along the chromosomes

of different organisms [30] For example, HSF2 was flanked by the SERINCI gene in human, mouse and chicken orthologous segments (Fig 1) [31] Likewise, the chicken HSF3 gene was located between Vsig4 and HEPH on chromosome 4, and orthologous segments containing the two genes were found on the human and mouse X chromosome Sequencing of a region between the two genes revealed the mouse HSF3 gene Although sequences related to HSF3 were also observed in an orthologous region of the human gen-ome, this genomic segment is likely to be an HSF3 pseudogene as no transcript was identified [31] Fur-thermore, HSF4 was located in a region between the TRADD–FBXL8 genes and the NoL3 gene in the human and mouse genomes, and chicken HSF4 was identified in an orthologous segment [31]

Comparison of the predicted amino acids of four members of the vertebrate HSF family revealed that sequences of the DNA-binding and trimerization [hydrophobic heptad repeat (HR)-A⁄ B] domains are well conserved (Fig 2) [31] The identity of the amino acid sequence in the DNA-binding domain of mouse HSF1 was much lower for mouse HSF3 (53%) than for mouse HSF2 or HSF4 (70% and 76%, respec-tively) Furthermore, the amino acid sequence of the DNA-binding domain of mouse HSF3 was only 60% homologous to that of chicken HSF3, whereas the sequences of mouse HSF1, HSF2 and HSF4 were much more identical to the corresponding domains of chicken HSF1, HSF2 and HSF4 (92%, 86% and 71% identity, respectively) Moreover, a phylogenetic tree, which was generated from full-length amino acid sequences of the HSF family, showed the relatedness

of mouse HSF3 with chicken HSF3 to be much weaker than that of mouse HSF1 with chicken HSF1, that of mouse HSF2 with chicken HSF2, or even that

of mouse HSF4 with chicken HSF4 (Fig 3) These estimations indicate that the nucleic acid sequences of HSF3 diverged most quickly during evolution, whereas those of HSF1 and HSF2 were similarly conserved The phylogenetic tree also demonstrates the amino

acid sequence of HSF1 to be most closely related with

that of HSF4 among the HSF family, and the amino

acid sequence of HSF2 to be most closely related with

that of HSF3 These findings are consistent with the

assertion that two rounds of whole-genome duplication

occurred in the vertebrate lineage (Fig 4) [32–34]

Alignment of the human and chicken HSF genes with

the mouse HSF gene showed that sequences of the

ex-ons are well cex-onserved, whereas those of intrex-ons are

not [31], suggesting that four duplicated HSF genes

have been conserved during evolution under selective

pressure, except for human HSF3

Expression of classical heat shock

genes induced by two heat-responsive

HSFs

After the identification of mammalian HSF1 and

HSF2 [20–22], and of chicken HSF1, HSF2 and HSF3

[23], research was conducted to reveal the factors

responsible for the heat-inducible HSE-binding activ-ity In mammalian cells, HSF1 remains an inert mono-mer in unstressed cells and forms a trimono-mer that binds

to the HSE in response to heat shock [35,36], whereas the HSE-binding activity of HSF2 is induced during hemin-induced differentiation of erythroleukemia cells and is constitutively high during early mouse development [37–39] In chicken cells, both an HSF1 monomer and an HSF3 dimer were converted to homotrimers that bind to the HSE under heat shock [40] The disruption of HSF genes in mouse embryonic fibroblasts (MEFs) clearly demonstrated that mouse HSF1 is required for the expression of classical heat shock genes [10], whereas mouse HSF2 is not [41] Unexpectedly, disruption of chicken HSF3 in chicken B-lymphocyte DT40 cells resulted in a severe reduction

in the inducible expression of HSP70, and the expres-sion of HSP110, HSP90a, HSP90b and HSP40 were not induced at all in chicken HSF3-null cells [11] These observations imply that duplicated HSF genes

SCRT1

HSF1

10 kb

Human Chr.8

Mouse Chr.15

DGAT1 HSF1 BOP1

SCRT1

HSF1 BOP1

SERINC1

HSF2

HSF2 SERINC1

Human Chr 6

Mouse Chr.10

Chicken Chr.3

HSF2 SERINC1

HSF2 SERINC1

Vsig4

10 kb

20 kb HSF2 SERINC1

Human Chr.6

HSF3

Vsig4

Vsig4

100 kb

100 kb HSF3

HEPH

N L3

HSF4

HSF4

10 kb

Fig 1 Comparative genomic analysis of orthologous segments containing vertebrate HSF genes The location of each segment is

as follows: human Chr.8 q24.3 and mouse Chr.15 D3 for HSF1; human Chr.6 q22.31, mouse Chr.10 B4 and chicken 63.95– 63.98 Mb for HSF2; human Chr X q12, mouse Chr X B4 and chicken 0.252– 0.265 Mb for HSF3; and human Chr.16 q22.1, mouse Chr.8 D3 and chicken 2.44– 2.45 Mb for HSF4 A genomic sequence corresponding to chicken HSF1 cDNA has not yet been identified Arrows indicate the 5¢ to 3¢ orientation of each gene The chicken HSF1 gene is located on chromo-some 2, which is present in three copies in DT40 cells [42] The gray box in human chromosome X is probably an HSF3 pseudogene.

evolved differently in mammalian and avian species

(Fig 4)

As the amino acid sequence of HSF1 is highly

con-served in mammalian and chicken cells, the functional

difference was examined in more detail In fact,

chicken HSF1 is dispensable for the expression of the

classical heat shock genes in DT40 cells [42], and the

ectopic expression of chicken HSF1 in MEF cells

defi-cient in mouse HSF1 does not restore the inducible

expression of the classical heat shock genes [28]

Thus, chicken HSF1 lacks the ability to induce the

expression of classical heat shock genes, whereas

mouse HSF1 is a master regulator of these genes

Interestingly, the amino-terminal region of chicken

HSF1 containing an alanine-rich sequence and the

DNA-binding domain is sufficient to cause the func-tional difference between the two orthologues [28] As chicken HSF1 can bind to the HSE, its amino-terminal domain might inhibit exposure of the activation domain to basal transcriptional machinery Alterna-tively, the corresponding domain of mouse HSF1, but not that of chicken HSF1, could recruit components required for gene activation

Recent identification of mouse HSF3 enabled us to examine the functional difference of HSF3 in mouse and chicken cells [31] In cells exposed to heat shock, mouse HSF3 fused to green fluorescent protein moved into the nucleus, similarly to chicken HSF3 [40], indi-cating that both chicken and mouse HSF3 are heat-responsive factors Furthermore, overexpression of

DBD HR-A/B

HR-C DHR

hHSF1

hHSF2

529

415 536

391422

6 9

0 23 22 25

Human

mHSF1

100 39

6 9

28 79

34 97 24 83

mHSF3

mHSF4

HSF1

16 15 24 33

53

30 31 21 39

76

58 85 67 92

Mouse

cHSF1

cHSF2

cHSF3

58 85 67 92

1

4 46

7

25 38 19

0

Chicken

CeHSF1 DmHSF

53

59 58

31

30

ScHSF

46 32

492

525

411442 535

390421 492

492

491

376408

564

421453 467

395426 510

Fig 2 Members of the HSF superfamily Diagrammatic representation of vertebrate and nonvertebrate HSF family members and of human HSF-related gene products The percentage identity between human HSF1 and each HSF was established using the computer program GENETYX-MAC The number of amino acids of each HSF is shown at the amino-terminal end c, chicken; Ce, Caenorhabditis elegans; DHR, downstream of HR-C; DBD, DNA-binding domain; DHR, downstream of HR-C; Dm, Drosophila melanogaster; h, human; HR, hydrophobic heptad repeat; m, mouse; Region X, a region upstream of the HR-C domain; Region Y, a C-terminal region downstream of the HR-C-like domain; Sc, Saccharomyces cerevisiae hHSF1 (hHSF1-a) [20]; hHSF2 (hHSF2-a) [21]; hHSF4 (hHSF4b) [24]; mHSF1 (hHSF1-a) and mHSF2 (hHSF2-a) [35]; mHSF3 (mHSF3a) [31]; mHSF4 (mHSF4b) [25]; cHSF1, cHSF2 and cHSF3 [23]; cHSF4 (cHSF4b) [31]; DmHSF [Wu 1990]; CeHSF1 (Swiss-P accession no Q9XW45); ScHSF [Pelham; Parker 1988]; hHSFY1 (SP accession no Q96LI6) [106,107]; hHSFX1 ⁄ LW-1 (SP accession no Q9UBD0); hHSF5 (SP accession no Q4G112) The hatched box indicates an HR-C-like domain, in which hydrophobic amino acids are not well conserved The DBD domain of HSF family members is conserved with one region in hHSFY1, hHSFX1 and hHSF5 that may not bind to the HSE.

chicken HSF3 in HSF1-null MEF cells induced the

constitutive and heat-induced expression of classical

heat shock genes In marked contrast, overexpression

of mouse HSF3 in the same cells did not affect the

expression of classical heat shock genes at all, even

after heat shock Therefore, mouse HSF3 lacks the

ability to induce the expression of classical heat shock

genes, whereas chicken HSF3 is a master regulator

Why does mouse HSF3 fail to induce the expression

of classical heat shock genes? It was revealed, by

exam-ining a DNA-binding transcription factor required for

the activation of the GAL genes in response to

galac-tose (GAL4) site-directed luciferase activity, that mouse

HSF3 has strong potential to induce transcription [31]

Deletion analysis showed that the activation domain of

mouse HSF3 is located in its C-terminal region

down-stream of the HR-C-like domain (region Y) whereas

that of chicken HSF3 is located upstream of the HR-C

domain (region X) (Fig 2) The amino acid sequence

of the activation domain of mouse HSF3 is not

con-served in chicken HSF3, which is consistent with a

functional divergence of the activation domain during

evolution Domains of mouse HSF3 were swapped with

the corresponding domains of chicken HSF3, and the

chimeras possessing the chicken HSF3 activation

domain induced the expression of the classical heat

shock genes in response to heat shock In contrast, the

chimeras possessing only the mouse HSF3 activation

domain did not induce their expression Furthermore, the C-terminal activation domain of human HSF1 [43–45] was swapped with the mouse HSF3 activation domain, and the resultant protein did not induce gene expression in response to heat shock These results indicate that the activation domain of mouse HSF3 does not have the potential to activate the classical heat shock genes

Human HSF1 recruits brahma-related gene 1 (BRG1), a component of switch⁄ sucrose nonferment-ing (SWI⁄ SNF) chromatin remodeling complexes, to the HSP70 promoter through direct interaction [46], and expression of an HSF1 mutant, which cannot interact with BRG1, did not restore the induction of HSP70 mRNA expression in HSF1-null MEF cells during heat shock [47,48] It was revealed that mouse HSF3 does not bind to BRG1, or recruit BRG1 to the HSP70 promoter [31], whereas chicken HSF3 does bind to and recruit BRG1 These observations indicate that mouse HSF3 does not induce the expression of the classical heat shock genes, at least in part because

of its inability to interact with BRG1

HSF-mediated adaptation to thermal stress

Heat shock induces both apoptotic and necrotic cell death, but the pathways of cell death and the factors

hHSF5

hHSFY1 hHSFX1

mHSF3 cHSF3

mHSF2 cHSF2

674 991

1000 1000

1000

1000 602

1000

1000 1000

1000 1000

839

908

0.05

684

HSF3

HSF2

HSF4

HSF1

hHSF2 cHSF4 mHSF4 hHSF4

ScHSF

mHSF1 hHSF1 cHSF1

CeHSF1 DmHSF

Fig 3 The phylogenetic tree generated in CLUSTAL W [108] for members of the HSF family Gaps were excluded from all phylogenetic analyses The numerals represent bootstrap values (1000 bootstrap replicates were performed) The unrooted tree was drawn using the pro-gram TREEVIEW [109] The bar represents 0.05 substitutions per site Amino acid sequences of the HSF family were shown previously [28,31].

that are primarily impaired are not clear, as heat

shock causes various types of stress, including

proteo-toxic stress Cells in different states of metabolism and

in different stages of differentiation may be induced to

die by different mechanisms, and there must be

vari-ous targets of extremely high temperatures that induce

cell death Therefore, HSPs should not be the only

proteins that protect against cell death Nevertheless,

HSPs are recognized as major players in the

protec-tion of cells from heat shock, especially from

proteo-toxic stress [1,2] As the expression of a set of HSPs is

regulated by HSFs, HSFs should be involved in the

protection of cells from heat shock or proteotoxic

stress [49]

It is well established that cells pretreated with

suble-thal heat shock can survive lesuble-thal heat shock This

phenomenon is called induced thermotolerance, and is

regulated by mouse HSF1 and chicken HSF3 through

the activation of the heat shock genes [10,11] HSPs

prevent the denaturation and aggregation of cellular

proteins, and support their renaturation when the cells

are recovering [1] At the same time, HSPs inhibit

sev-eral molecules, such as apoptotic peptidase activating

factor 1 (Apaf-1) and cytochrome c, which are

involved in mitochondria-mediated apoptotic pathways

[50]

Both HSF1 and HSF3 complementarily regulate the constitutive expression of some HSPs in normally growing chicken DT40 cells [11,42] In these cells, a lack of the two factors resulted in increased sensitivity

to a single exposure to high temperature because of reduced Hsp90a expression, and the cell cycle is blocked at the G2phase [42] A similar phenotype was observed in yeast S cerevisiae harbouring a mutant HSF [51,52] Mouse HSF1 also regulates the expres-sion of some HSPs, including Hsp90, in various mouse tissues [53–56] Therefore, HSFs could be involved in determining a temperature at which cells can survive

by regulating the constitutive expression of HSPs such

as Hsp90

Curiously, in chicken cells, HSF1 induced only very low levels of expression of the classical heat shock genes, but had a significant effect on the protection of cells from heat shock [28] This effect was not medi-ated through the induction of classical heat shock genes or regulation of the constitutive expression of heat shock genes such as Hsp90 HSF1-null MEF cells, which lack induced expression of the classical heat shock genes, are more sensitive to high temperatures than wild-type cells Remarkably, overexpression of chicken HSF1 in the HSF1-null MEF cells restored resistance to heat shock [28] Moreover, mouse HSF3

HSF1

2RWGD

Mouse cell

HSF1

Non-WGD

HSP induction High expression

HSF2 HSF4

HSF3

Ancestor cell

in the lens

Polyploid

Chicken cell

HSF3

HSF4

HSF4 HSF2

HSF1 HSF1

> 310 Myr ago

High expression

in the lens

HSF3 HSP induction

HSF2

Fig 4 A model to explain the evolution of HSF genes Two rounds of whole-genome duplication (WGD) may have occurred in vertebrate ancestral cells more than 440 million years ago (Ma) [110,111], which resulted in polyploidization Thereafter, avian and mammalian cells evolved differently from an ancestral cell 310 Ma The expression and function of the four HSF genes were conserved or diverged during evolution [112] For example, during mammalian evolution, HSF1 retained the ability to induce the expression of heat shock proteins, whereas it lost this function during avian evolution Instead, avian HSF3 retained the function Expression of the HSF4 gene increased in the lens during both avian and mammalian evolutions (M Fujimoto and A Nakai, unpublished) Diamonds, circles and triangles represent regula-tory regions driving expression in different tissues.

was able to protect HSF1-null MEF cells from heat

shock without inducing the expression of the classical

heat shock genes [31] These observations indicate that

chicken HSF1 and mouse HSF3 protect cells from

thermal stress by regulating the expression of

heat-inducible genes other than classical heat shock genes

Expression of nonclassical heat shock

genes induced by HSFs

Although the heat shock response was originally

char-acterized based on the expression of a limited number

of classical heat shock genes, HSF1 was recently

revealed to regulate the expression of numerous other

genes in the absence or presence of heat shock

Com-prehensive analyses of HSF-binding regions in the

whole genome revealed that  3% of genes are direct

targets in heat-shocked cells in yeast and Drosophila

[57,58], and expression of the majority of the target

genes is induced during heat shock in unicellular yeast

[57] Even in mammalian cells, HSF1 (similarly to

yeast and Drosophila HSF) binds to the promoters of

a great number of genes in the whole genome [59,60],

and about half of the target genes are expressed during

heat shock [59] We now refer to these heat-inducible

genes that are different from the classical heat shock

genes as ‘nonclassical heat shock genes’ As already

discussed, chicken HSF1 and mouse HSF3 play a

sub-stantial role in the protection of cells from heat shock

[28,31], implying significance of the nonclassical heat

shock genes in this process

To establish whether the expression of nonclassical

heat shock genes was induced by mouse HSF3 or

chicken HSF1, nonclassical heat shock genes were

identified in MEF cells [31] Induction of one of the

nonclassical heat shock genes, the gene for the PSD-95⁄

Dlg-A⁄ ZO-1 (PDZ) domain-containing protein

PDZK3⁄ PDZD2 ⁄ PAPIN (plakophilin-related armadillo

repeat protein-interacting protein) [61], decreased, but

was still observed in HSF1-null MEF cells during heat

shock [31] Overexpression of mouse HSF3 or chicken

HSF1 in the HSF1-null MEF cells restored the marked

induction of expression of PDZK3, whereas

knock-down of mouse HSF3 completely abolished the

induc-tion Induction of another nonclassical heat shock

gene, that for a membrane glycoprotein, prominin-2

(PROM2) [62], was abolished in HSF1-null MEF cells,

but was restored when mouse HSF3 or chicken HSF1

was overexpressed [31] These observations clearly

demonstrated that evolutionally conserved mouse

HSF3 and chicken HSF1 uniquely regulate only

non-classical genes, suggesting importance of the regulation

of the nonclassical heat shock genes

It is worth noting that HSF4 also regulates nonclas-sical heat shock genes in lens cells although it is not a heat-responsive factor A set of HSF4-binding regions was identified in lens cells, and the expression of genes located in and near these regions was examined [63] Interestingly, a great number of the genes (33%) were expressed in response to heat shock, and, unexpect-edly, the expression of these genes was not induced in HSF1-null lens cells Surprisingly, HSF4 was required for the expression of half of the genes, in part by facil-itating the binding of HSF1 to the promoters [63] Moreover, the expression of satellite III repeat sequences during heat shock was extensively studied [64,65] HSF1 is required for expression of the satellite III gene during heat shock, but HSF2 also greatly affects its expression, possibly by interacting with HSF1 [66] Taken together, all members of the verte-brate HSF family are involved in the regulation of gene expression during heat shock (Fig 5)

HSF is essential in yeast and human cancer cells

In the budding yeast S cerevisiae, HSF is essential for survival under normal conditions [15,16], consistent with the notion that S cerevisiae HSF is constitutively

Classical heat shock genes Nonclassical heat shock genes

HSF2 HSF1

HSF1

HSE

PROM2 Sat III etc

Adaptation to proteotoxic stress

Fig 5 Adaptation to proteotoxic stress by the HSF family in mice HSF1 remains mostly as an inert monomer in unstressed cells, and

is converted to an active trimer that binds to the HSE located in the proximal promoter region of a limited number of classical heat shock genes during heat shock, which results in induction of the expression of HSPs HSF2 may modulate this process by interact-ing with HSF1 [6,113] Members of the HSF family coordinately bind to the less-conserved HSE located on and near numerous non-classical heat shock genes, and greatly affect heat-induced expres-sion of the genes including PDZK3, PROM2 and satellite III [31,63,66] HSF3 is not expressed in human cells.

a trimer that binds to the HSE However, the binding

of HSF to the promoter of a heat shock gene markedly

increased during heat shock in S cerevisiae in vivo [67]

Furthermore, HSF was essential even in the fission

yeast S pombe, in which HSF forms a trimer only

under stress, similar that observed for vertebrate HSF1

[19] These observations implied that a balance of the

monomer and trimer HSF affects the amount of HSF

bound to the HSE in vivo, but even a small amount of

the trimer could regulate the gene expression, which is

required for survival under normal growth conditions

in unicellular yeasts In fact, HSF binds to many target

genes in vivo, and their products have a broad range of

biological functions, including protein folding and

deg-radation, energy generation and protein trafficking

[57] Human HSF2, but not HSF1, forms a trimer and

functionally complements the viability defect of yeast

cells lacking HSF, and both human HSF1 and HSF2

partially rescue the induction of heat-inducible genes,

which is associated with acquired thermotolerance [68]

These observations suggest that the roles of HSF

under normal growth conditions can be distinguished

from those under stress

Among multicellular organisms, HSF-null Drosophila

was the first to be generated, although a single HSF

was required for oogenesis and early larval

develop-ment, indicating that HSF is dispensable for cell

growth and survival under normal conditions [69]

Subsequently, HSF1-, HSF2- and HSF4-deficient mice

were generated, indicating that these HSFs are also

dispensable [41,70–73] Detailed analysis of these mice

demonstrated that meiosis was impaired in the absence

of HSF1 and HSF2 in female [53,74] and male [71,75–

77] germ cells Neuronal differentiation and migration

were affected in the HSF2-null cerebral cortex [78]

Furthermore, the differentiation and maintenance of

sensory placodes required HSF1 and HSF4

[13,54,72,73] Thus, members of the HSF family exert

essential activities in the absence of stress and are

required for the differentiation of many types of cells

during development [14] However, they are not

neces-sary for cell proliferation and survival under normal

conditions in multicellular organisms

Cancer cells proliferate and survive in different ways

from normal cells, and many of the signalling

path-ways and transcription factors display a striking

dependence on the chaperone machinery [79,80]

Fur-thermore, HSF1 expression is elevated in human

can-cer cells [81,82], suggesting that cancan-cer cells may be

dependent on the heat shock response Therefore, the

effects of loss of HSF1 function on cancer cell

prolifer-ation and survival were examined First, human

cervi-cal epithelial HeLa cells stably expressing short hairpin

RNA for HSF1 were generated (these cells show > 95% reduction in the HSF1 level), and were highly sensitive to combined treatment with both elevated temperature and anticancer reagents [83] Then, decreased lymphomagenesis in a p53-deficient mouse model was shown in the absence of HSF1 [84] Unex-pectedly, in addition to being required for carcinogene-sis in mice, HSF1 is required for proliferation and survival in various human cancer cell lines, including HeLa cells, but not in normal or immortalized cells [85] This observation suggests the possibility of com-mon HSF-mediated mechanisms for cell proliferation and survival in cancer cells and in yeast cells

Is the HSF1 in cancer cells activated? As HSF1 expression, which is correlated with HSE-binding activity [35], is elevated in human cancer cells [81,82], the HSE-binding activity of HSF1 might be higher in cancer cells than in normal cells Even so, the HSE-binding activity is robustly induced in response to heat shock in cancer cells, such as HeLa cells, compared with normal cells [35,36] or in fission yeast cells [19] HSF1 is involved in regulating translation, ribosome biogenesis and glucose metabolism in cancer cells [85], and is also required for the expression of numerous genes, including those for inflammatory cytokines, chemokine-related genes and interferon-related genes, even in normally growing primary cultures of MEF cells [86] Furthermore, the ability of HSF1 to form a trimer is required for the gene expression [87] Taken together, only a little amount of HSF1 trimer regulates the gene expression in normal cells, which is not required for cell growth and survival However, the tri-meric HSF1 is increased in cancer cells and may regu-late the expression of genes, which is indispensable for cell growth and survival, as in fission yeast cells for example [19]

Adaptation to misfolding-related pathological conditions

An imbalance of protein homeostasis is associated with aging and age-related pathological conditions such as neurodegenerative disorders, including Alzheimer’s dis-ease, Parkinson’s disdis-ease, amyotrophic lateral sclerosis, prion disease and polyglutamine diseases These diseases are characterized by conformational changes in disease-causing proteins that results in misfolding and aggrega-tion, and are therefore termed protein-misfolding disorders or protein conformational disorders [49,88] Polyglutamine (polyQ) diseases are caused by an expan-sion of CAG repeats, coding for glutamine, in their respective proteins Misfolding and aggregation of aggregation-prone polyQ proteins results in cellular

toxicity Gain of HSF1 function significantly inhibits

the aggregation of polyQ protein and prolongs life span

in C elegans models of polyQ diseases, whereas loss of

HSF1 function accelerates the aggregation of polyQ and

shortens life span [89,90] It is considered that HSF1

function is mediated through the expression of HSPs

[49,88] Interestingly, a forkhead box (FOXO) family

transcription factor, DAF-16 (abnormal dauer

forma-tion 16), which is a component of the insulin-like

signal-ling pathway that regulates life span, also inhibits polyQ

aggregates, indicating that aging and protein

homeosta-sis are highly related [89,90] In a C elegans model of

Alzheimer’s disease, HSF1 inhibited the formation of

toxic aggregates of an aggregation-prone peptide Ab

(1–42) whereas DAF-16 promoted the formation of

less-toxic high-molecular-weight aggregates [91] Thus,

HSF1 and DAF-16 regulate distinct pathways that

reduce the toxicity of aggregation-prone proteins

Among mouse polyQ models, the R6⁄ 2 polyQ model

has been extensively studied as it is transgenic only for

the 5¢ end of the human huntingtin gene carrying 115–

150 CAG repeat expansions [92] The formation of

polyQ aggregates is observed not only in the brain but

also in nonneuronal tissues, including the skeletal

mus-cle, heart, liver and pancreas, in mice [93] Ubiquitous

overexpression of HSP70 in the R6⁄ 2 Huntington’s

model had no effect on the life span or neuronal

phe-notypes of the mice and delayed aggregation only

slightly [94,95] There is only one HSF1 transgenic

mouse model, in which an actively mutated HSF1 is

expressed in tissues such as the testis, skeletal muscle,

heart and stomach, but not in the brain [75]

Remark-ably, overexpression of an active HSF1 in nonneuronal

tissues in R6⁄ 2 mice crossed with HSF1 transgenic

mice suppressed polyQ aggregates, at least in skeletal

muscle, and markedly extended the life span [96]

Inversely, HSF1 deficiency dramatically shortened the

life span of the prion disease model mice, in which

scrapie prions were inoculated [97], and even resulted

in impaired protein homeostasis of the untreated

neu-ronal cells in some genetic backgrounds [98] These

observations imply significant beneficial effects of the

overexpression of an active HSF1 on the progression

of protein-misfolding disorders in mice Interestingly,

mouse HSF3 and chicken HSF1 suppressed the

forma-tion of aggregates in a cellular polyQ model [28,31],

suggesting that the nonclassical heat shock genes as

well as classical heat shock genes play roles in

amelio-rating disease progression

One therapeutic strategy for protein-misfolding

dis-orders such as polyQ disease would be to elevate the

levels of HSPs that assist normal protein folding and

prevent abnormal folding and aggregation [99] It was

shown that treatment with arimoclomol, a co-inducer

of HSPs through activating HSF1, delays disease pro-gression in the amyotrophic lateral sclerosis mouse model, which overexpresses human mutant Cu⁄ Zn superoxide dismutase-1 [100] HSF1-activating reagents, 17-allylamino-17-demethoxygeldanamycin (17-AAG) and geranylgeranylacetone (GGA), also ameliorated disease progression in a Drosophila model of spinocerebellar ataxias [101] or in a mouse model of spinal and bulbar muscular atrophy [102,103] Furthermore, novel small molecules that activate HSF1 were identified and shown to ameliorate protein misfolding and toxicity of polyQ aggregates [104,105] Thus, activation of HSF1

is actually a promising therapeutic approach for pro-tein-misfolding disorders

Conclusions and perspectives

We have learned from the identification and character-ization of HSF1 and HSF3 in chicken and mouse cells that the function of these two heat-responsive factors has diverged greatly during the evolution of vertebrates, even though the nucleic acid sequence of each has been well conserved Importantly, chicken HSF1 and mouse HSF3 are involved in protecting cells from heat shock and in maintaining protein homeostasis without induc-ing the expression of the classical heat shock genes It was revealed recently that there are numerous nonclas-sical heat shock genes, whose expression is induced dur-ing heat shock in various organisms Remarkably, chicken HSF1 and mouse HSF3, as well as mammalian HSF2 and HSF4, play a role in inducing the expression

of only nonclassical heat shock genes These observa-tions suggest the importance of the regulation and func-tion of the nonclassical heat shock genes Analysis of these new findings will help us to understand why the activation of HSF1 suppresses the progression of pro-tein-misfolding disorders more than HSPs and should

be beneficial in identifying pathways involved in adap-tation to proteotoxic stress Furthermore, these analy-ses would develop our understanding of the biological significance of the heat shock response

Acknowledgements

We thank members of our laboratory for discussions and Naoki Hayashida for comments on the manu-script This work was supported in part by Grants-in-Aid for Scientific Research and on Priority Area-a Nuclear System of DECODE, from the Ministry of Education, Culture, Sports, Science and Technology, Japan, and by the Yamaguchi University Research Project on STRESS

1 Lindquist S (1986) The heat-shock response Ann Rev

Biochem 55, 1151–1191

2 Gross CA, Strauss DB, Erickson JW & Yura T (1990)

The function and regulation of heat shock proteins in

Escherichia coli In Stress Proteins in Biology and

Medi-cine(Morimoro RI, Tissieres A & Georgopoulos C eds),

pp 167–189 Cold Spring Harbor Laboratory Press,

Cold Spring Harbor, NY

3 Wu C (1995) Heat shock transcription factors:

struc-ture and regulation Annu Rev Cell Biol 11, 441–469

4 Fernandes M, O’Breien T & Lis JT (1994) Structure

and regulation of heat shock gene promoters In The

Biology of Heat Shock Proteins and Molecular

Chaper-ones (Morimoro RI, Tissieres A & Georgopoulos C

eds), pp 375–393 Cold Spring Harbor Laboratory

Press, Cold Spring Harbor, NY

5 Morimoto RI (1998) Regulation of the heat shock

transcriptional response: cross talk between a family of

heat shock factors, molecular chaperones, and negative

regulators Genes Dev 12, 3788–3796

6 Bjo¨rk JK & Sistonen L (2010) Regulation of the

mem-bers of the mammalian heat shock factor family FEBS

J 277, 4126–4139

7 Morano KA & Thiele DJ (1999) Heat shock factor

function and regulation in response to cellular stress,

growth, and differentiation signals Gene Expr 7, 271–

282

8 Nakai A (1999) New aspects in the vertebrate heat

shock factor system: Hsf3 and Hsf4 Cell Stress

Chap-erones 4, 86–93

9 Pirkkala L, Nykanen P & Sistonen L (2001) Roles of the

heat shock transcription factors in regulation of the heat

shock response and beyond FASEB J 15, 1118–1131

10 McMillan DR, Xiao X, Shao L, Graves K &

Benjamin IJ (1998) Targeted disruption of heat shock

transcription factor 1 abolishes thermotolerance and

protection against heat-inducible apoptosis J Biol

Chem 273, 7523–7528

11 Tanabe M, Kawazoe Y, Takeda S, Morimoto RI,

Nagata K & Nakai A (1998) Disruption of the HSF3

gene results in the severe reduction of heat shock gene

expression and loss of thermotolerance EMBO J 17,

1750–1758

12 Akerfelt M, Trouillet D, Mezger V & Sistonen L

(2007) Heat shock factors at a crossroad between stress

and development Ann NY Acad Sci 1113, 15–27

13 Nakai A (2009) Heat shock transcription factors and

sensory placode development BMB Rep 42, 631–635

14 Abane R & Mezger V (2010) Roles of heat shock

fac-tors in gametogenesis and development FEBS J 277,

4150–4172

15 Sorger PK & Pelham HRB (1988) Yeast heat shock

factor is an essential DNA-binding protein that

exhibits temperature-dependent phosphorylation Cell

54, 855–864

16 Wiederrecht G, Seto D & Parker C (1988) Isolation of the gene encoding the S cerevisiae heat shock tran-scription factor Cell 54, 841–853

17 Clos J, Westwood JT, Becker PB, Wilson S & Wu C (1990) Molecular cloning and expression of a

hexamer-ic Drosophila heat shock factor subject to a negative regulation Cell 63, 1085–1097

18 Jakobsen BK & Pelham HRB (1991) A conserved heptapeptide restrains the activity of the yeast heat shock transcription factor EMBO J 10, 369–375

19 Gallo GJ, Prentice H & Kingston RE (1993) Heat shock factor is required for growth at normal tempera-ture in the fission yeast Schizosacccharomyces pombe Mol Cell Biol 13, 749–761

20 Rabindran SK, Giorgi G, Clos J & Wu C (1991) Molecular cloning and expression of a human heat shock factor, HSF1 Proc Natl Acad Sci USA 88, 6906–6910

21 Schuetz TJ, Gallo GJ, Sheldon L, Tempst P & Kings-ton RE (1991) Isolation of a cDNA for HSF2: evi-dence for two heat shock factor genes in humans Proc Natl Acad Sci USA 88, 6911–6915

22 Sarge KD, Zimarino V, Holm K, Wu C & Morimoto

RI (1991) Cloning and characterization of two mouse heat shock factors with distinct inducible and constitu-tive DNA-binding ability Genes Dev 5, 1902–1911

23 Nakai A & Morimoto RI (1993) Characterization of a novel chicken heat shock transcription factor, heat shock factor 3, suggests a new regulatory pathway Mol Cell Biol 13, 1983–1997

24 Nakai A, Tanabe M, Kawazoe Y, Inazawa J, Morimoto

RI & Nagata K (1997) HSF4, a new member of the human heat shock factor family which lacks properties

of a transcriptional activator Mol Cell Biol 17, 469–481

25 Tanabe M, Sasai N, Nagata K, Liu XD, Liu PC, Thi-ele DJ & Nakai A (1999) The mammalian HSF4 gene generates both an activator and a repressor of heat shock genes by alternative splicing J Biol Chem 274, 27845–27856

26 International Human Genome Sequencing Consortium (2001) Initial sequencing and analysis of the human genome Nature 409, 860–921

27 Mouse Genome Sequencing Consortium (2002) Initial sequencing and comparative analysis of the mouse gen-ome Nature 420, 520–562

28 Inouye S, Katsuki K, Izu H, Fujimoto M, Sugahara

K, Yamada S, Shinkai Y, Oka Y, Katoh Y & Nakai A (2003) Activation of heat shock genes is not necessary for protection by heat shock transcription factor 1 against cell death due to a single exposure to high tem-peratures Mol Cell Biol 23, 5882–5895

29 International Chicken Genome Sequencing Consortium (2004) Sequence and comparative analysis of the