FEBS Letters 589 (2015) 1150–1155

journal homepage: www.FEBSLetters.org

HSF1 transcriptional activity is modulated by IER5 and PP2A/B55 Yukio Ishikawa, Shotaro Kawabata, Hiroshi Sakurai ⇑ Division of Health Sciences, Kanazawa University Graduate School of Medical Science, 5-11-80 Kodatsuno, Kanazawa, Ishikawa 920-0942, Japan

a r t i c l e

i n f o

Article history: Received 8 December 2014 Revised 16 March 2015 Accepted 19 March 2015 Available online 26 March 2015 Edited by Zhijie Chang Keywords: Chaperone Heat shock factor 1 IER5 Immediate-early gene Protein phosphatase 2A Phosphorylation

a b s t r a c t Heat shock factor 1 (HSF1) is the master transcriptional regulator of chaperone genes. HSF1 regulates the expression of the immediate-early response gene IER5, which encodes a protein that has roles in the stress response and cell proliferation. Here, we have shown that IER5 interacts with protein phosphatase 2A (PP2A) and its B55 regulatory subunits. Expression of IER5 and B55 in cells leads to HSF1 dephosphorylation and activation of HSF1 target genes. The B55 subunits directly bind to HSF1. These results suggest that IER5 functions as a positive feedback regulator of HSF1 and that this process involves PP2A/B55 and HSF1 dephosphorylation. Ó 2015 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

1. Introduction Mammalian heat shock factor 1 (HSF1) plays a central role in the cellular-response to different proteotoxic conditions and regulates both the constitutive and the stress-inducible expression of heat shock protein (HSP) genes. Many HSPs function as molecular chaperones that assist in the correct folding of nascent polypeptide chains and prevent the formation of non-specific protein aggregates [1,2]. Under non-stressful conditions, HSF1 also targets a wide spectrum of genes involved in cell maintenance, cell cycle progression, and differentiation, thereby playing a key role in the processes of aging, carcinogenesis, and development [2–4]. Most HSF1 in the normal cell is present in a latent inactive monomeric state in the cytoplasm, although a small amount of HSF1 engages in transcriptional regulation in the nucleus. In response to proteotoxic stress, such as heat shock, HSF1 forms a trimer, translocates into the nucleus, binds to the heat shock elements (HSEs), and acquires transcriptional ability. The transcriptional activity is modulated by protein–protein interactions and

Author contributions: HS conceived and supervised the study; YI and HS designed experiments; YI, SK and HS performed experiments; YI and HS analyzed data; HS wrote the manuscript. ⇑ Corresponding author. Fax: +81 76 234 4369. E-mail address: [email protected] (H. Sakurai).

post-translational modifications, including phosphorylation, acetylation, and sumoylation [5–8]. HSF1 is a phosphorylated protein even under non-stressful conditions, and this constitutive phosphorylation is implicated in the repression of the transcriptional activity [9–16]. Stress-induced hyperphosphorylation of HSF1 is a prerequisite for acquisition of the activator function [17–20]. IER5 is a member of the growth-related immediate-early gene family and encodes a 327-amino acids protein with putative nuclear localization signals [21]. Expression of IER5 is rapidly and transiently induced after stimulation of resting cells with serum, as other immediate-early genes, and by treatment of cells with ionizing radiation and anti-cancer agents [21–26]. IER5 is a negative regulator of cell proliferation; overexpression of IER5 leads to G2-M cell cycle arrest, whereas IER5 knockdown increases proliferation [24,25]. IER5 reduces the expression of CDC25B, a positive regulator of cell cycle progression, through direct binding to the gene promoter, suggesting a role as a transcription factor [25]. We have recently demonstrated that IER5 is a target of HSF1 and that its expression is induced in response to proteotoxic stress [27]. Expression of IER5 causes an increase in chaperone gene expression and in recovery from thermal stress. It was speculated that IER5 modulates the transcriptional activity of HSF1. In the present study, we first examined the effect of IER5 on the phosphorylation status of HSF1, then investigated the interactions of HSF1, IER5, and the regulatory subunits of protein phosphatase 2A (PP2A).

http://dx.doi.org/10.1016/j.febslet.2015.03.019 0014-5793/Ó 2015 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

Y. Ishikawa et al. / FEBS Letters 589 (2015) 1150–1155

2. Materials and methods 2.1. DNA constructs The coding region of IER5 was cloned into the gene expression vectors pEBMulti-Hyg (Wako Pure Chemicals, Osaka, Japan) and pcDNA3.1(+) (Invitrogen, Carlsbad, CA, USA) to create pEBMultiIER5 and pcDNA-IER5, respectively [27]. For the expression of an influenza hemagglutinin (HA)-tagged protein, the coding region was cloned into pBK266, a derivative of pcDNA3.1(+) containing three copies of the HA-tag sequences (pcDNA-IER5-HA) [28]. Plasmid pK732 is a derivative of pcDNA3.1(+) containing HSF1HA [29]. Nucleotide substitution mutations were introduced by standard methods. The HSF1 fragment was also cloned into the glutathione S-transferase (GST) gene fusion vector pGEX-6P-1 (GE Healthcare, Buckinghamshire, UK). The cDNA of B55 family members was amplified by RT-PCR from the total RNA of HeLa cells and cloned into pcDNA3.1(+), pBK266, and pGEX-6P-1. The firefly reporter construct HSE-SV40-LUC contained HSE4P oligonucleotides upstream of the SV40 promoter of the pGL3-Promoter vector (Promega, Madison, WI, USA) [27]. The firefly reporter construct HSP70-LUC contained the promoter region of HSPA1A in the pGL3-Basic vector (Promega) [27].

1151

Reporter Assay System (Promega) and firefly luciferase activity (arbitrary units) was calculated after normalization to Renilla luciferase values. 2.6. GST pull-down assay GST fusion proteins were immobilized on glutathioneSepharose 4B (GE Healthcare) (Supplementary Fig. S1) [29]. IER5HA, B55b-HA, and HSF1 polypeptides were synthesized in vitro using the TNT T7 Coupled Reticulocyte Lysate System (Promega). The polypeptides were incubated with immobilized proteins in binding buffer [29] at room temperature for 15 min. After three washes with 200 ll of binding buffer, bound polypeptides were subjected to Western blotting. 2.7. Immunoprecipitation and in vitro phosphatase assay

HeLa cells were cultured as described previously [28,29]. Transfection was conducted using HilyMax reagent according to the manufacturer’s recommended protocol (Dojindo Laboratories, Kumamoto, Japan).

Cells were lysed in buffer containing 20 mM Tris–Cl (pH 7.6), 0.1 M NaCl, 0.5 mM EDTA, 0.5% NP-40, 0.5 mM phenylmethylsulfonyl fluoride, and protease inhibitor cocktail. Cell extracts were incubated with an anti-HA antibody at 4 °C overnight and then with Protein A Sepharose CL-4B (GE Healthcare) on a rotating wheel for 90 min. Beads were washed with lysis buffer and bound proteins were subjected to Western blotting. For the phosphatase assay, beads were added in 100 ll of a phosphatase assay buffer containing 50 mM Tris–Cl (pH 7.6), 1 mM MgCl2, and 1 mM 6, 8-difluoro-4-methylumbelliferyl phosphate (DiFMUP, Molecular Probes, Eugene, OR, USA). Fluorometric assay of DiFMUP substrate was conducted using a microplate luminometer (FluoroSkan Ascent, Thermo Scientific) as described [31].

2.3. Western blotting

2.8. Statistical analysis

Cell extracts were prepared and subjected to Western blotting as described previously [27]. For phosphatase treatment and Phos-tag SDS polyacrylamide gel electrophoresis (PAGE), cells were lysed in high-salt buffer containing 50 mM Tris–Cl (pH 7.6), 0.42 M NaCl, 25% glycerol, 0.5 mM phenylmethylsulfonyl fluoride, and protease inhibitor cocktail (Nacarai Tesque, Kyoto, Japan). Cell extracts were diluted 3-fold with distilled water and incubated with 10 U of calf intestinal alkaline phosphatase at room temperature for 10 min. The 6% acrylamide gel contained 20 lM Phos-tag (Wako Pure Chemicals) and 40 lM MnCl2, and Western blotting was conducted according to the manufacturer’s protocol [30]. The following antibodies were used: anti-IER5 [27], anti-HSP70 [27], anti-HSF1 [27], anti-PP2A-Ca/b (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-HA [28], and anti-GAPDH [27].

The data are representative of at least three independent experiments. Significant differences were determined by Student’s t-test.

2.2. Cell culture and transfection

2.4. RT-PCR assay The RT-PCR analysis was performed as described previously [27,28]. The PCR product quantities were compared after normalizing each sample to the amounts of GAPDH PCR products. 2.5. Luciferase assay Cells were transfected with DNA mixtures containing firefly luciferase reporter plasmid, Renilla luciferase control plasmid (pRL-TK), and gene expression plasmid [28,29]. For knockdown experiments, small interfering RNA (siRNA) was co-transfected. The sequences of siRNA are as follows: siIER5, sense 50 -CCUCAUC AGCAUCUUCGGUUU-30 and antisense 50 -ACCGAAGAUGCUGAU GAGGUU-30 ; scrambled siRNA, sense 50 -CCUACGCCACCAAUUUCG UUU-30 and antisense 50 -ACGAAAUUGGUGGCGUAGGUU-30 . The luciferase assay was performed using the Dual-Luciferase

3. Results 3.1. Effect of IER5 on the expression of HSE-containing genes Although IER5 was suggested to contain nuclear localization signal sequences [21], its cellular localization has not been experimentally determined. We prepared cytoplasmic and nuclear extracts from HeLa cells and analyzed for IER5 levels in both the fractions (Fig. 1A). IER5 was present predominantly in the cytoplasmic fraction, and the nuclear fraction contained a small amount of IER5. Heat shock treatment of cells at 42.5 °C for 1 h did not significantly change the cellular distribution of IER5. Therefore, IER5 is mostly located in the cytoplasm. HSF1 was detected in the cytoplasm in non-heat-shocked cells and translocated to the nucleus during heat shock. The reason for slower migration of HSF1 in the heat-shocked samples was hyperphosphorylation (see Fig. 2A) [32]. The chaperone gene HSP70 contains HSEs, and its expression is regulated by HSF1 under both normal physiological conditions and heat shock conditions [4]. We conducted Western blot analysis to examine HSP70 levels in HeLa cells harboring the IER5 expression plasmid (Fig. 1B). Forced expression of IER5 caused a 1.4-fold increase in HSP70 in non-heat-shocked cells. When cells were heat-shocked at 42.5 °C for 1 h and allowed to recover at 37 °C for 5 h, HSP70 levels increased 1.9-fold in control cells and 2.7-fold in cells harboring the IER5 plasmid. RT-PCR analysis showed that the constitutive and heat-induced levels of HSP70 mRNA were increased by IER5 (Fig. 1C). The effect of IER5 on the expression

1152

Y. Ishikawa et al. / FEBS Letters 589 (2015) 1150–1155

NHS

C

B

HS

N

C

IER5

N kDa

kDa 75

IER5

IER5

*

8

*

5

R

R IE

SC

si

si GAPDH

kDa 48 35

0

CD

- HSE + HSE NHS

0

+ HSE HS

siSCR siIER5

4

30

15

0 HSF1-HA

RD

x 2.1

*

HS

HSF1-HA

C2

of the HSE-containing genes was further analyzed using synthetic luciferase reporter constructs (Fig. 1D). Under non-heat shock conditions, IER5 did not affect the luciferase activity of a reporter construct, the expression of which was driven by basal promoter. When HSE oligonucleotides were inserted upstream of the promoter, the luciferase activity increased 1.5-fold in cells harboring the IER5 plasmid. Heat shock led to robust expression of the HSE-containing reporter, and IER5 further stimulated expression. Conversely, silencing IER5 caused a 25% decrease in expression

GAPDH

x 1.4

E

C1 Fig. 1. Transcriptional activity of HSE-containing genes in cells expressing IER5. (A) Western blotting of IER5. HeLa cells cultured at 37 °C (NHS) were heat-shocked at 42.5 °C for 1 h (HS). Cytoplasmic (C) and nuclear (N) extracts were subjected to Western blot analysis using anti-IER5, anti-HSF1, and anti-GAPDH. The positions of molecular mass markers are shown on the right. (B) Western blotting of HSP70. HeLa cells were transfected with the IER5 expression construct pEBMulti-IER5 or control empty vector plasmid pEBMulti-Hyg. The cells were heat-shocked at 42.5 °C for 1 h and allowed to recover at 37 °C for 5 h (HS) or were untreated (NHS). Cell extracts were subjected to Western blot analysis using anti-HSP70, anti-IER5, and anti-GAPDH. The relative levels of HSP70 are indicated. (C) RT-PCR of HSP70 mRNA. Cells transfected as in (B) were heat-shocked at 42.5 °C for 1 h (HS) or were untreated (NHS). The HSP70 mRNA levels were compared with control non-heatshocked cells (cont). Data are expressed as the mean ± S.D. of three experiments. ⁄ P < 0.01. (D) Luciferase assay of the HSE-containing reporter gene. Firefly luciferase reporter constructs HSE-SV40-LUC (+HSE) and SV40-LUC (HSE) were transfected along with pEBMulti-IER5 or pEBMulti-Hyg. Cells were exposed to 42.5 °C for 1 h and then recovered at 37 °C for 4 h (right panel, HS) or were untreated (left panel, NHS). Firefly luciferase activities (AU: arbitrary units) are expressed as the mean ± S.D. of four experiments. ⁄P < 0.01. (E) Luciferase assay of HSE-SV40-LUC in IER5-silenced cells. (Left panel) Cells were transfected with IER5 siRNA or scrambled siRNA (SCR), and extracts were subjected to Western blot analysis using anti-IER5 and anti-GAPDH. (Right panel) Firefly luciferase reporter construct HSESV40-LUC was transfected along with IER5 siRNA or scrambled siRNA. Cells were cultured as in (D). Firefly luciferase activities are expressed as the mean ± S.D. of four experiments. ⁄P < 0.01.

*

x 1.3

*

S121A S303A S307A S363A S363D

D NHS

x 1.7

x 2.0

x 4.5

WT

AD

x 2.4

cont IER5-HA

HSF1-HA

0

529

HR-C

FL

C

8

*

8

16

*

S303 S307 S363 HR-A/B

DB

0.3

HS

E IER5

0.6

*

DBD

T S1 21 S3 A 03 S3 A 07 S3 A 63 A

NHS

cont IER5

S121

HSF1

H3 H2 C1 H1 C2

-P

35

W

0

GAPDH

B Relative luc activity (AU)

16

Relative luc activity (AU)

Relative HSP70 mRNA

cont IER5

kDa HSF1 75

35

D

C

Phos-tag SDS-PAGE

HSF1 48

GAPDH

35

SDS-PAGE

CIAP HS IER5-HA

1.0 1.4 1.9 2.7

75

GAPDH

A

HS

HSP70

48

HSF1

NHS

Relative luc activity (AU)

A

kDa 35

CIAP IER5 C2 C1 -P GAPDH

WT

S363A

C3 C1 C1’ kDa 35

Fig. 2. HSF1 activity in cells expressing IER5. (A) Phosphorylation status of HSF1. Cells transfected with the IER5-HA expression construct pcDNA-IER5-HA or empty vector plasmid pcDNA3.1 were heat-shocked at 42.5 °C for 30 min (HS) or were untreated. Phosphatase (CIAP)-treated and untreated samples were subjected to SDS–PAGE (left panel) and Phos-tag SDS–PAGE (right panel) followed by Western blot analysis using anti-HSF1 and anti-GAPDH. On the SDS–PAGE, the positions of molecular mass markers are shown on the right. On the Phos-tag SDS–PAGE, phosphatase-treated HSF1 bands are marked by P; two HSF1 bands in control cell samples are marked by C1 and C2, and three HSF1 bands in heat-shocked cell samples are marked by H1, H2, and H3. (B) Schematic representation of HSF1. DNAbinding domain (DBD), hydrophobic repeats (HR-A/B and HR-C), conserved negative regulatory domain (CD), and four serine residues are shown. Bars show the regions fused to GST: FL contained full length HSF1; DB (DNA-binding and oligomerization domains) contained residues 1–202 of HSF1; RD (repression domain) contained residues 202–397; and AD (activation domain) contained residues 400–529. (C) Luciferase assay of HSP70-LUC in cells expressing HSF1-HA. Firefly luciferase reporter construct HSP70-LUC was transfected along with the HSF1-HA expression constructs and pcDNA-IER5-HA or pcDNA3.1 (cont). Firefly luciferase activities (arbitrary units) are expressed as the mean ± S.D. of six experiments. The fold increase of luciferase activity by IER5-HA cotransfection is indicated. Asterisks indicate statistically significant upregulation by alanine substitution mutations compared with wild type (⁄P < 0.01). (D) Phos-tag SDS– PAGE of HSF1-HA derivatives. Cells were transfected with the HSF1-HA expression constructs, and extracts were subjected to Phos-tag SDS–PAGE (upper panel) and SDS–PAGE (lower panel) followed by Western blot analysis using anti-HA and antiGAPDH. Two HSF1-HA bands are marked by C1 and C2. Different bands observed in HSF1-HA containing the S303A, S307A, and S363A mutations are marked by closed circles. (E) Phos-tag SDS–PAGE of HSF1(S363A)-HA. Cells were transfected with the HSF1-HA expression constructs and pcDNA-IER5 or pcDNA3.1. Phosphatase (CIAP)treated and untreated samples were subjected to Phos-tag SDS–PAGE (upper panel) and SDS–PAGE (lower panel) followed by Western blot analysis using anti-HA and anti-GAPDH. Three HSF1(S363A)-HA bands are marked by C1, C10 , and C3.

1153

Y. Ishikawa et al. / FEBS Letters 589 (2015) 1150–1155

3.4. Effect of B55 subunits on HSF1 activity The interactions of HSF1 and the B55 subunits were analyzed by a GST pull-down assay (Fig. 4A). HSF1 polypeptides bound to all four GST-B55 fusions. Conversely, B55b-HA polypeptides bound to HSF1 in the GST fusion. The HSF1 regions required for B55b-HA binding were located in the N-terminal DNA-binding and oligomerization domains and the central repression domain. The N-terminal two-thirds of HSF1 were also bound by IER5-HA

A IN

ST

Transcriptional competence of HSF1 is regulated by its phosphorylation status [5,6]. Western blot analysis showed that HSF1 was phosphorylated in cells under normal physiological conditions, as determined by its slower migration on SDS polyacrylamide gels relative to phosphatase-treated samples (Fig. 2A). When cells were heat-shocked at 42.5 °C for 30 min, HSF1 was hyperphosphorylated and migrated more slowly than that of non-heat-shocked cells. Forced expression of HA-tagged IER5 (IER5-HA) did not affect the mobility of HSF1 in this analysis. The addition of Phos-tag, a phosphate-binding molecule, to the gels results in the mobility shift of phosphorylated proteins [30]. On Phos-tag SDS–PAGE, HSF1 from non-heat-shocked cells was detected as two bands C1 and C2, and the HSF1 of heat-shocked cells was detected as at least three bands H1, H2, and H3 (Fig. 2A). Expression of IER5-HA led to a decrease in the intensities of C2 and H3. C2 and H3 bands migrated slowly on the gels and would thus be more extensively phosphorylated than faster migrating bands, implying that IER5 is involved with dephosphorylation of HSF1, regardless of thermal stress. Constitutive phosphorylation of HSF1 on serine residues 121, 303, 307, and 363 is involved in repression of HSF1’s transcriptional activity (Fig. 2B) [9–16]. We constructed HA-tagged HSF1 derivatives containing alanine substitution mutations in these residues and analyzed the effect of IER5 using an HSP70 promoter-driven luciferase reporter construct (HSP70-LUC) (Fig. 2C). Under nonheat shock conditions, the luciferase activity of HSP70-LUC was increased 4.5-fold by the co-transfection of IER5-HA. When HSF1-HA was transfected, the luciferase activity increased 4.2-fold and was further enhanced 2.1-fold by IER5-HA. The alanine substitution mutations of the four residues resulted in a 1.2–1.8-fold higher transcriptional activity than wild-type HSF1-HA, suggesting that phosphorylation of these residues leads to repression of HSF1. In cells harboring HSF1-HA with the S121A, S303A, and S307A mutations, IER5-HA enhanced the reporter expression 1.7–2.4-fold, which was not statistically different from its effect in cells harboring the wild type. However, the luciferase activity of S363A mutant cells was increased only 1.3-fold by IER5-HA. A phospho-mimetic aspartic acid mutation at S363 (S363D) did not significantly change the HSF1-HA activity, and IER5-HA had a small effect (1.4-fold increase) on the expression of the reporter. Because the effect of IER5 was diminished by the S363A and S363D mutations, S363 should be the target residue for IER5 function. On Phos-tag SDS– PAGE, the S303A, S307A, and S363A mutations affected the mobility of HSF1-HA (Fig. 2D). The differences in band patterns reflected defective phosphorylation of S303, S307, and S363. HSF1-HA containing the S363A mutation resolved into at least three bands C10 , C1, and C3 (Fig. 2E). The disappearance of C2 band in S363A suggested the contribution of S363 phosphorylation to the formation of C2 band; however, the band pattern was different from that of wild-type HSF1-HA upon IER5 overexpression. The intensities of C1 and C3 bands of S363A were reduced by IER5. Therefore, IER5 is involved in the dephosphorylation of S363 and other unknown phosphorylated residues.

Previous high throughput protein–protein interaction analysis showed the binding of IER5 to the B55b subunit of PP2A [33]. PP2A is a trimeric holoenzyme of the catalytic C subunit, scaffold A subunit, and regulatory B subunit [34,35]. B55b is a member of the B subunits and constitutes a family together with B55a, B55c, and B55d. The B subunit is important in the regulation of PP2A specificity and activity. We examined the interactions of IER5 and the B55 subunits using in vitro-synthesized IER5-HA polypeptides and GST fusions of B55a, B55b, B55c, and B55d. The results of GST pull-down analysis showed binding of IER5 to all four B55 subunits (Fig. 3A). Coimmunoprecipitation analysis was conducted to analyze the interaction of IER5 and PP2A. As shown in Fig. 3B, immunoprecipitation of IER5-HA by an anti-HA antibody resulted in coprecipitation of the PP2A catalytic subunit (PP2AC). The immunoprecipitated complex contained phosphatase activity, as judged by phosphatase assay using the fluorogenic phosphatase substrate DiFMUP; however, the activity was significantly inhibited by the addition of a PP2A inhibitor, Okadaic acid, at a concentration of 5.0 nM (Fig. 3C). The complex prepared from heatshocked cells had a similar level of phosphatase activity (data not shown). These results indicate the interactions of IER5 with PP2A.

G

3.2. Effect of IER5 on HSF1 phosphorylation

3.3. Interactions of IER5 and PP2A/B55

B

GST-B55

α

β

γ

δ

IER5-HA

PP2AC

IER5-HA

IER5-HA

kDa 35 48

C Phosphatase activity (AU)

of the HSE-containing reporter in heat-shocked cells (Fig. 1E). The constitutive expression was also reduced (10%) by IER5 knockdown, although it was not statistically significant. Therefore, IER5 induces the expression of HSE-containing genes. Because HSE is the recognition sequence of HSF1, it is implied that IER5 stimulates the transcriptional ability of HSF1 under both non-heat shock and heat shock conditions.

150

IER5-HA

100

IER5-HA + OA

50

0

cont

0

15

30

45

60

75

90 (min)

Fig. 3. Interactions of IER5 and PP2A/B55. (A) GST pull-down analysis of IER5 and B55. In vitro-synthesized IER5-HA polypeptides were incubated with GST or GSTB55 (a, b, c, and d) fusions. Input (IN) and bound polypeptides were subjected to Western blot analysis using anti-HA. (B) Co-immunoprecipitation analysis of IER5 and PP2AC. Extracts were prepared from cells transfected with pcDNA-IER5-HA or pcDNA3.1. Immunoprecipitation was conducted using anti-HA and precipitated proteins were subjected to Western blot analysis using anti-PP2AC and anti-HA. The positions of molecular mass markers are shown on the right. (C) The phosphatase activity of IER5-containing complex. Immunoprecipitation was conducted as in (B), and precipitated proteins were subjected to the phosphatase assay using DiFMUP as a substrate. Okadaic acid (OA) was added to the final concentration of 5 nM. Phosphatase activities are given as arbitrary units (AU).

1154

Y. Ishikawa et al. / FEBS Letters 589 (2015) 1150–1155

B

GST-B55

α

β

γ

δ

C HSF1

-P

α

β

γ

δ

G

IN

ST

HSF1

B55βHA IER5HA

GST-HSF1 FL DB RD AD

-P

B55-HA GAPDH

C2 C1 kDa 63 35

D Relative luc activity (AU)

G

IN

ST

A

10

5

0 B55-HA IER5-HA

*

*

* * *

HSF1-HA IER5 B55β C2 C1 GAPDH

α β γ δ

β

kDa 35

Fig. 4. HSF1 activity in cells expressing B55. (A) GST pull-down analysis of B55, HSF1, and IER5. (Upper panel) In vitro-synthesized HSF1 polypeptides were incubated with GST or GST-B55 (a, b, c, and d) fusions. (Lower panel) In vitro-synthesized B55b-HA and IER5-HA were incubated with GST or GST-HSF1 (FL, DB, RD, and AD; see Fig. 2B) fusions. Input (IN) and bound polypeptides were subjected to Western blot analysis using anti-HSF1 and anti-HA. (B) Phosphorylation status of HSF1. Cells were transfected with pcDNA-B55 (a, b, c, and d)-HA or pcDNA3.1 (). Cell extracts were subjected to Phos-tag SDS–PAGE (upper panel) and SDS–PAGE (middle and lower panels) followed by Western blot analysis using anti-HSF1, anti-HA, and anti-GAPDH. Phosphatase-treated sample (P) was co-electrophoresed. Two HSF1 bands are marked by C1 and C2. The positions of molecular mass markers are shown on the right. (C) Luciferase assay of HSP70-LUC. Firefly luciferase reporter construct HSP70-LUC was transfected along with pcDNA-B55 (a, b, c, and d)-HA, pcDNA-IER5-HA, or pcDNA3.1. Firefly luciferase activities (arbitrary units) are expressed as the mean ± S.D. of four experiments. ⁄P < 0.01. (D) Phos-tag SDS–PAGE of HSF1-HA. Cells were transfected with the HSF1-HA expression construct and pcDNA-B55b, pcDNA-IER5, or pcDNA3.1. Cell extracts were subjected to Phos-tag SDS–PAGE (upper panel) and SDS–PAGE (lower panel) followed by Western blot analysis using anti-HA and anti-GAPDH.

polypeptides. These results showed the physical interactions among HSF1, IER5, and the B55 subunits. To investigate the role of the B55 subunits in the regulation of HSF1 transcriptional ability, subunits were expressed in cells as HA-tagged proteins. Western blot analysis with an anti-HA antibody showed the expression of these proteins, although the levels of B55a-HA were slightly lower than those of other subunits (Fig. 4B). When cell extracts were subjected to Phos-tag SDS–PAGE followed by Western blotting with anti-HSF1, the intensities of the HSF1 C2 band were reduced by the expression of each of the four B55 subunits. As shown in Fig. 4C, the four B55 subunits had a stimulatory effect on HSP70-LUC expression, resulting in a 1.5–2.4-fold increase in luciferase activity. B55b-HA and IER5-HA enhanced the reporter expression in an additive manner. On Phos-tag SDS–PAGE, C2 band of HSF1-HA was not observed in the extracts of cells harboring both the B55b and IER5 constructs (Fig. 4D). These results suggest the involvement of the B55 subunits in dephosphorylation and activation of HSF1.

yeast, inhibitory phosphorylation is relieved by Ppt1 phosphatase for the ethanol-specific activation of Hsf1 [37]. Ppt1 is related to human PP5, and PP5 physically interacts with the HSF1–Hsp90 complexes in a Xenopus oocyte model system, in which PP5 is involved in the inhibition, but not the activation, of HSF1, although it has not been tested whether PP5 affects the phosphorylation status of HSF1 [38]. Meanwhile, PP2A is known to indirectly activate HSF1. The PP2A complex containing the B56d subunit dephosphorylates AMP-activated protein kinase (AMPK), and inactive/ dephosphorylated AMPK is unable to phosphorylate the inhibitory phosphorylation site S303 and causes HSF1 activation [39]. Overall, phosphatase is a crucial regulator of HSF1. Our observations implicate IER5 as a positive feedback regulator of HSF1. We propose that IER5 induced by growth factors, heat shock, ionizing radiation, and chemotherapy modulates cell proliferation, the proteotoxic stress response, and the DNA damage response through the regulation of HSF1. Acknowledgements

4. Discussion HSF1 is known to induce the expression of IER5 in response to proteotoxic stress, including heat shock [27]. Expression of IER5 leads to dephosphorylation of HSF1 on serine 363 and to induction of HSF1 target genes, including HSP70. IER5 interacts with HSF1 and protein phosphatase PP2A. The B55 regulatory subunits of PP2A directly bind to HSF1 and enhance HSF1 dephosphorylation and activation. Although we failed to detect an IER5-induced association of PP2A/B55 with HSF1, these results suggest the existence of a positive feedback loop on HSF1 activation dependent on IER5, PP2A/B55, and HSF1 dephosphorylation. Although treatment of cells with phosphatase inhibitors causes activation of HSF1 [36], HSF1 has both stimulatory and inhibitory phosphorylation sites. Constitutive phosphorylation on residues S121, S303, S307, and S363 is involved in the repression of transcriptional activity, whereas stress-induced phosphorylation on other residues is necessary for the activator function [9–20]. Phosphorylation status is determined by the balance between phosphorylating kinases and dephosphorylating phosphatases. Although various protein kinases are suggested to participate in HSF1 phosphorylation, it remains unknown whether HSF1 is dephosphorylated by a specific phosphatase. Our observation that the B55 subunits physically interact with HSF1 and that B55 expression leads to HSF1 dephosphorylation suggests the direct involvement of PP2A/B55 in the dephosphorylation of HSF1. In

We thank E. Kawahara for Western blotting and Y. Ishita, R. Taniguchi, and K. Murai for technical assistance. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.febslet.2015.03. 019. References [1] Hartl, F.U., Bracher, A. and Hayer-Hartl, M. (2011) Molecular chaperones in protein folding and proteostasis. Nature 475, 324–332. [2] Akerfelt, M., Morimoto, R.I. and Sistonen, L. (2010) Heat shock factors: integrators of cell stress, development and lifespan. Nat. Rev. Mol. Cell Biol. 11, 545–555. [3] Neef, D.W., Jaeger, A.M. and Thiele, D.J. (2011) Heat shock transcription factor 1 as a therapeutic target in neurodegenerative diseases. Nat. Rev. Drug Discov. 10, 930–944. [4] Ciocca, D.R., Arrigo, A.P. and Calderwood, S.K. (2013) Heat shock proteins and heat shock factor 1 in carcinogenesis and tumor development: an update. Arch. Toxicol. 87, 19–48. [5] Voellmy, R. (2004) On mechanisms that control heat shock transcription factor activity in metazoan cells. Cell Stress Chaperones 9, 122–133. [6] Björk, J.K. and Sistonen, L. (2010) Regulation of the members of the mammalian heat shock factor family. FEBS J. 277, 4126–4139. [7] Sakurai, H. and Enoki, Y. (2010) Novel aspects of heat shock factors: DNA recognition, chromatin modulation and gene expression. FEBS J. 277, 4140– 4149.

Y. Ishikawa et al. / FEBS Letters 589 (2015) 1150–1155 [8] Xu, Y.M., Huang, D.Y., Chiu, J.F. and Lau, A.T. (2012) Post-translational modification of human heat shock factors and their functions: a recent update by proteomic approach. J. Proteome Res. 11, 2625–2634. [9] Knauf, U., Newton, E.M., Kyriakis, J. and Kingston, R.E. (1996) Repression of human heat shock factor 1 activity at control temperature by phosphorylation. Genes Dev. 10, 2782–2793. [10] Chu, B., Soncin, F., Price, B.D., Stevenson, M.A. and Calderwood, S.K. (1996) Sequential phosphorylation by mitogen-activated protein kinase and glycogen synthase kinase 3 represses transcriptional activation by heat shock factor-1. J. Biol. Chem. 271, 30847–30857. [11] Kline, M.P. and Morimoto, R.I. (1997) Repression of the heat shock factor 1 transcriptional activation domain is modulated by constitutive phosphorylation. Mol. Cell. Biol. 17, 2107–2115. [12] Chu, B., Zhong, R., Soncin, F., Stevenson, M.A. and Calderwood, S.K. (1998) Transcriptional activity of heat shock factor 1 at 37 °C is repressed through phosphorylation on two distinct serine residues by glycogen synthase kinase 3a and protein kinases Ca and Cf. J. Biol. Chem. 273, 18640–18646. [13] Dai, R., Frejtag, W., He, B., Zhang, Y. and Mivechi, N.F. (2000) C-Jun NH2terminal kinase targeting and phosphorylation of heat shock factor-1 suppress its transcriptional activity. J. Biol. Chem. 275, 18210–18218. [14] Hietakangas, V., Ahlskog, J.K., Jakobsson, A.M., Hellesuo, M., Sahlberg, N.M., Holmberg, C.I., Mikhailov, A., Palvimo, J.J., Pirkkala, L. and Sistonen, L. (2003) Phosphorylation of serine 303 is a prerequisite for the stress-inducible SUMO modification of heat shock factor 1. Mol. Cell. Biol. 23, 2953– 2968. [15] Wang, X., Grammatikakis, N., Siganou, A. and Calderwood, S.K. (2003) Regulation of molecular chaperone gene transcription involves the serine phosphorylation, 14-3-3e binding, and cytoplasmic sequestration of heat shock factor 1. Mol. Cell. Biol. 23, 6013–6026. [16] Wang, X., Khaleque, M.A., Zhao, M.J., Zhong, R., Gaestel, M. and Calderwood, S.K. (2006) Phosphorylation of HSF1 by MAPK-activated protein kinase 2 on serine 121, inhibits transcriptional activity and promotes HSP90 binding. J. Biol. Chem. 281, 782–791. [17] Holmberg, C.I., Hietakangas, V., Mikhailov, A., Rantanen, J.O., Kallio, M., Meinander, A., Hellman, J., Morrice, N., MacKintosh, C., Morimoto, R.I., Eriksson, J.E. and Sistonen, L. (2001) Phosphorylation of serine 230 promotes inducible transcriptional activity of heat shock factor 1. EMBO J. 20, 3800– 3810. [18] Soncin, F., Zhang, X., Chu, B., Wang, X., Asea, A., Ann Stevenson, M., Sacks, D.B. and Calderwood, S.K. (2003) Transcriptional activity and DNA binding of heat shock factor-1 involve phosphorylation on threonine 142 by CK2. Biochem. Biophys. Res. Commun. 303, 700–706. [19] Guettouche, T., Boellmann, F., Lane, W.S. and Voellmy, R. (2005) Analysis of phosphorylation of human heat shock factor 1 in cells experiencing a stress. BMC Biochem. 6, 4. [20] Murshid, A., Chou, S.D., Prince, T., Zhang, Y., Bharti, A. and Calderwood, S.K. (2010) Protein kinase A binds and activates heat shock factor 1. PLoS One 5, e13830. [21] Williams, M., Lyu, M.S., Yang, Y.L., Lin, E.P., Dunbrack, R., Birren, B., Cunningham, J. and Hunter, K. (1999) Ier5, a novel member of the slowkinetics immediate-early genes. Genomics 55, 327–334. [22] Okada, A., Kushima, K., Aoki, Y., Bialer, M. and Fujiwara, M. (2005) Identification of early-responsive genes correlated to valproic acid-induced neural tube defects in mice. Birth Defects Res A Clin. Mol. Teratol. 73, 229–238.

1155

[23] Kis, E., Szatmári, T., Keszei, M., Farkas, R., Esik, O., Lumniczky, K., Falus, A. and Sáfrány, G. (2006) Microarray analysis of radiation response genes in primary human fibroblasts. Int. J. Radiat. Oncol. Biol. Phys. 66, 1506–1514. [24] Ding, K.K., Shang, Z.F., Hao, C., Xu, Q.Z., Shen, J.J., Yang, C.J., Xie, Y.H., Qiao, C., Wang, Y., Xu, L.L. and Zhou, P.K. (2009) Induced expression of the IER5 gene by gamma-ray irradiation and its involvement in cell cycle checkpoint control and survival. Radiat. Environ. Biophys. 48, 205–213. [25] Nakamura, S., Nagata, Y., Tan, L., Takemura, T., Shibata, K., Fujie, M., Fujisawa, S., Tanaka, Y., Toda, M., Makita, R., Tsunekawa, K., Yamada, M., Yamaoka, M., Yamashita, J., Ohnishi, K. and Yamashita, M. (2011) Transcriptional repression of Cdc25B by IER5 inhibits the proliferation of leukemic progenitor cells through NF-YB and p300 in acute myeloid leukemia. PLoS One 6, e28011. [26] Wouters, J., Stas, M., Govaere, O., Van den Eynde, K., Vankelecom, H. and van den Oord, J.J. (2012) Gene expression changes in melanoma metastases in response to high-dose chemotherapy during isolated limb perfusion. Pigment Cell Melanoma Res. 25, 454–465. [27] Ishikawa, Y. and Sakurai, H. (2015) Heat-induced expression of the immediate-early gene IER5 and its involvement in the proliferation of heatshocked cells. FEBS J. 282, 332–340. [28] Sakurai, H., Sawai, M., Ishikawa, Y., Ota, A. and Kawahara, E. (2014) Heat shock transcription factor HSF1 regulates the expression of the Huntingtininteracting protein HYPK. Biochim. Biophys. Acta 1840, 1181–1187. [29] Ota, A., Enoki, Y., Yamamoto, N., Sawai, M. and Sakurai, H. (2013) Evolutionarily conserved domain of heat shock transcription factor negatively regulates oligomerization and DNA binding. Biochim. Biophys. Acta 1829, 930–936. [30] Kinoshita, E., Kinoshita-Kikuta, E., Takiyama, K. and Koike, T. (2006) Phosphate-binding tag, a new tool to visualize phosphorylated proteins. Mol. Cell. Proteomics 5, 749–757. [31] Wegner, A.M., McConnell, J.L., Blakely, R.D. and Wadzinski, B.E. (2007) An automated fluorescence-based method for continuous assay of PP2A activity. Methods Mol. Biol. 365, 61–69. [32] Sawai, M., Ishikawa, Y., Ota, A. and Sakurai, H. (2013) The proto-oncogene JUN is a target of the heat shock transcription factor HSF1. FEBS J. 280, 6672–6680. [33] Glatter, T., Wepf, A., Aebersold, R. and Gstaiger, M. (2009) An integrated workflow for charting the human interaction proteome: insights into the PP2A system. Mol. Syst. Biol. 5, 237. [34] Virshup, D.M. and Shenolikar, S. (2009) From promiscuity to precision: protein phosphatases get a makeover. Mol. Cell 33, 537–545. [35] Eichhorn, P.J., Creyghton, M.P. and Bernards, R. (2009) Protein phosphatase 2A regulatory subunits and cancer. Biochim. Biophys. Acta 1795, 1–15. [36] Xia, W. and Voellmy, R. (1997) Hyperphosphorylation of heat shock transcription factor 1 is correlated with transcriptional competence and slow dissociation of active factor trimers. J. Biol. Chem. 272, 4094–4102. [37] Cho, B.R., Lee, P. and Hahn, J.S. (2014) CK2-dependent inhibitory phosphorylation is relieved by Ppt1 phosphatase for the ethanol stressspecific activation of Hsf1 in Saccharomyces cerevisiae. Mol. Microbiol. 93, 306–316. [38] Conde, R., Xavier, J., McLoughlin, C., Chinkers, M. and Ovsenek, N. (2005) Protein phosphatase 5 is a negative modulator of heat shock factor 1. J. Biol. Chem. 280, 28989–28996. [39] Zhu, X.N., Chen, L.P., Bai, Q., Ma, L., Li, D.C., Zhang, J.M., Gao, C., Lei, Z.N., Zhang, Z.B., Xing, X.M., Liu, C.X., He, Z.N., Li, J., Xiao, Y.M., Zhang, A.H., Zeng, X.W. and Chen, W. (2014) PP2A–AMPKa–HSF1 axis regulates the metal-inducible expression of HSPs and ROS clearance. Cell. Signal. 26, 825–832.