Downregulation of heat shock factor 4 transcription activity via MAPKinase phosphorylation at Serine 299

Accepted Manuscript Title: Downregulation of heat shock factor 4 transcription activity via MAPKinase phosphorylation at Serine 299 Authors: Xiukun Cui, Jiuli Han, Jing Li, Wen-Wen Cui, Dan-Dan Wu, Shiyuan Liu, Wenxian Xue, Xiang-Xiang Wang, Yuanfang Ma, Jing Zhang, Jun Zhang, Hongmei Mu, Fengyan Zhang, Yanzhong Hu PII: DOI: Reference:

S1357-2725(18)30218-8 https://doi.org/10.1016/j.biocel.2018.10.003 BC 5435

To appear in:

The International Journal of Biochemistry & Cell Biology

Received date: Revised date: Accepted date:

31-5-2018 7-10-2018 8-10-2018

Please cite this article as: Cui X, Han J, Li J, Cui W-Wen, Wu D-Dan, Liu S, Xue W, Wang X-Xiang, Ma Y, Zhang J, Zhang J, Mu H, Zhang F, Hu Y, Downregulation of heat shock factor 4 transcription activity via MAPKinase phosphorylation at Serine 299, International Journal of Biochemistry and Cell Biology (2018), https://doi.org/10.1016/j.biocel.2018.10.003 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Downregulation of heat shock factor 4 transcription activity via MAPKinase

SC RI PT

phosphorylation at Serine 299

Xiukun Cui1, Jiuli Han1, Jing Li1, Wen-Wen Cui1, Dan-Dan Wu1 ，Shiyuan Liu1，

Wenxian Xue1, Xiang-Xiang Wang1, Yuanfang Ma1, Jing Zhang1, Jun Zhang1, Hongmei

National Laboratory for Antibody Drug Engineering, Henan International Union

M

1Joint

A

N

U

Mu2 ,Fengyan Zhang3 and Yanzhong Hu1, 2*

Lab of Antibody Medicine, Department of Cell Biology and Genetics, Henan University

Key Lab of Myopia and Cataract, Institute of Eye Disease, Kaifeng Central

TE

2 Kaifeng

D

School of Basic Medical Sciences. Kaifeng, China.

Hospital, Kaifeng, China.

of ophthalmology, The First Affiliate Hospital of Zhengzhou University.

EP

3Department

A

CC

Running title: Regulation of Phosphorylation of HSF4/S299

Highlights

Proper tuning of HSF4 transcription activity is important for lens development.



Identification of HSF4/S299 is one of MEK-ERK1/2 targets.



phosphorylation of HSF4 at S299 by ERK1/2 induces HSF4 sumoylation and low

SC RI PT



expression of alpha B-crystallin. 

Phosphorylation of HSF4/S299 is responsive of turning off HSF4 transcription activity during postnatal lens development.

Abstract

U

Dysfunction of HSF4 is associated with congenital cataracts. HSF4 transcription activity

N

is turned on and regulated by phosphorylation during early postnatal lens development.

A

Our previous data suggested that mutation HSF4b/S299A can upregulate HSF4

M

transcription activity in vitro, but the biological significance of posttranslational

D

modification on HSF4/S299 during lens development remains unclear. Here, we found

TE

that the mutation HSF4/S299A can upregulate the expression of HSP25 and alpha Bcrystallin at both protein and mRNA levels in mouse the lens epithelial cell line, but

EP

HSF4/S299D does not. Using the rabbit polyclonal antibody against phospho-S299 of HSF4, we found that EGF and ectopic expression of MEK1 can increase the

CC

phosphorylation of HSF4/S299 and induce HSF4 sumoylation, and these effects are inhibited by U0126. ERK1/2 can phosphorylate the S299 in HSF4/wt but not in

A

HSF4/S299A in the in vitro kinase assay. Functionally, ectopic MEK1 can inhibit HSF4controled alpha B-crystallin expression but has less effect on HSF4/S299A. EGF can upregulate phospho-HSF4/S299 and downregulate alpha B-crystallin expression in P3 mouse lens, and this downregulation is suppressed by U0126. During mouse lens

development, phosphorylation of HSF4/S299 is downregulated in P3 lens and upregulated in P7 and P14 lens. However, in 2 months old lens, both phosphorylation of HSF4/S299 and total HSF4 protein are decreased. Interestingly, ERK1/2 activity is lower

SC RI PT

in P3 lens than in P7 and P14 lens, which is in line with the phosphorylation of HSF4/S299. Taken together, our data demonstrate that HSF4/299 is a phosphorylation target of MEK1-ERK1/2, and phosphorylation of S299 is responsible for tuning down HSF4 transcription activity during postnatal lens development.

A

N

U

Keywords: HSF4, phosphorylation of S299, PDSM motif, MEK, ERK1/2

D

M

1. Introduction

TE

Lens development is regulated temporospatially by turning on or off a series of

EP

transcription factors. Genetic variations in many transcription factors have been associated with congenital cataracts(Graw, 2004), which affect about 0.015-0.03% of

CC

children worldwide. Heat shock transcription factor 4 (HSF4) is a critical regulator of lens development(Bu et al., 2002; Min et al., 2004). HSF4 is turned on in E15.5

A

embryonic lens, and its transcription activity is important for postnatal lens development(Fujimoto et al., 2004). Mutations in HSF4’s DNA binding domain or other coding sequences are associated with hereditary autosomal dominant and recessive cataracts both in humans and canine(Berry et al., 2017; Bu et al., 2002; Mellersh et al., 2006). Loss of function studies indicate that deficiency of HSF4 leads to nuclear

cataracts in early neonatal age mice(Fujimoto et al., 2004; Min et al., 2004). These data suggest proper regulation of HSF4 transcription activity is important for lens homeostasis.

SC RI PT

HSF4 belongs to a family of heat shock factors that controls the transcription of heat shock proteins in response to proteotoxic stress(Nakai et al., 1997). In mammalian tissue, HSF4 exists as two splice variants, HSF4a and HSF4b (Tanabe et al., 1999). HSF4b contains 42 amino acids more than HSF4a in the regulatory domain due to

alternative splicing between exon 8 and 9. This results in their highly divergent functions.

U

HSF4a can inhibit the expression of HSP90 and HSP70 in HeLa cells by associating

N

with TFIIF(Frejtag et al., 2001). HSF4b can replace yeast HSF and maintain yeast

A

survival(Tanabe et al., 1999). It can also upregulate the Gal4-reporter system in Cos7

M

cells after heat shock. HSF4b is the predominant variant in lens. Its expression level and

D

transcription activity are developmentally regulated in lens tissue(Min et al., 2004;

TE

Morange, 2006). HSF4b is more active in postnatal lens, and its expression and activity decline with increasing age(Min et al., 2004). HSF4b is expressed in lens epithelial and

EP

cortical fiber cells and shows differential regulatory effects on the expression of HSP25, alpha B-crystallin, γ-crystallin, vimentin and FGFs (Cui et al., 2016; Fujimoto et al., 2004;

CC

Zhang et al., 2014). For example, HSF4b can upregulate the expression of HSP25,

A

alpha B-crystallin, γ-crystallin, and simultaneously downregulate the expression of vimentin and FGFs. This suggests that HSF4 transcription activity may be regulated by complicated signaling mechanisms. Phosphorylation, sumoylation and acetylation are three common posttranslational modifications that most of the transcription factors undergo. Accumulating evidence

suggest HSF4b is regulated by phosphorylation(Hu and Mivechi, 2006; Tanabe et al., 1999). HSF4b protein contains a phosphorylation-dependent sumoylation motif (ΨKxExxSP, PDSM) flanking S299(Hietakangas et al., 2006). The PDSM motif is

SC RI PT

conserved in HSF1, HSF2 and other transcription factors but not HSF4a. Mutation HSF4b/S299A or K294R in the PDSM motif can upregulate the transcription activity of

Gal-HSF4 fusion protein in the Gal4-reporter system in Cos 7 cells and K526 tumor cells in vitro(Hietakangas et al., 2006). Phosphorylation of S299 increases the sumoylation of K294 and enhances the interaction between HSF4 and the transcription inhibitor Daxx

U

(Zhang et al., 2010). Our previous data show that mutation HSF4b/T472A can decrease

N

the association of HSF4 with nuclear pore transporter protein importin beta-1, blocking

A

HSF4 nuclear translocation(Zhang et al., 2014). FGF2 can upregulate HSF4b protein

M

stability and transcription activity by activating ERK1/2–mediated phosphorylation(Hu et al., 2013). HSF4b recruits ERK1/2 and DUSP26 into one complex, where DUSP26

D

attenuates ERK1/2-mediated HSF4b protein stability(Hu and Mivechi, 2006). However, it

TE

is unclear whether HSF4/S299 is the target of ERK1/2 kinase.

EP

In this paper, we studied the biological significance of the phosphorylation of S299/HSF4 during lens development and the associated protein kinase for this site.

CC

Consistent with previous data, mutation of S299 to A upregulates the expression of

A

HSP25 and alpha B-crystallin in lens epithelial cells, but mutation of S299 to D does not. Using antibody against phospho-S299, we found that EGF or ectopic MEK1 can upregulate HSF4/S299 phosphorylation and HSF4/K294 sumoylation, resulting in the downregulation of alpha B-crystallin expression in lens epithelial cells in vitro and in P3 mouse lens ex vivo. ERK1/2 can directly phosphorylate HSF4b/S299. Moreover, we

found that phosphorylation of HSF4/S299 is associated with ERK1/2 activity during lens development in the postnatal age. These results suggest that phosphorylation of S299 by ERK1/2 decreases HSF4 transcription activity during lens development.

SC RI PT

2. Materials and Methods 2.1 Cell lines and plasmids: mLEC/Hsf4-/- and HLE-B3 were cultured in DMEM media

containing 10% FBS and 1x streptomycin and penicillin(Hu et al., 2013). Cells were passed every two days. For the plasmids, pWZL-HA-Hsf4b: the cDNA encoding human HSF4b

U

was amplified from pcDNA-Flag-HSF4b and inserted into the retrovirus vector pWZL-Blast

N

vector at the EcoRI restriction site to generate the pWZL-HA-HSF4b construct. pWZL-HA-

A

Hsf4b/S299A and pWZL-HA-Hsf4b/S299D: The site-mutation at the Hsf4b/S299A or

M

Hsf4b/S299D was performed on the pWZL-HA-Hsf4b plasmid template following the kit protocol. pcDNA-Flag-Hsf4b/S299A, pcDNA-Flag-Hsf4/S299D, pMyc-Hsf4b/S299A or

D

pMyc-Hsf4b/S299D: the cDNAs of HSF4/S299A or S299D were amplified from pWZL-

TE

Hsf4b/S299A or pWZL-Blast-Hsf4b/S299D, and subcloned into the pcDNA3-Flag or pMyc vector at the EcoRI restriction site. pcDNA-HA-MEK1: the cDNA encoding human MEK1

EP

was amplified with RT-PCR and subcloned into the pcDNA plasmid between the BamHI

CC

and XhoI restriction sites. pEBG-sumo-1: the cDNA encoding sumo-1 was generated from the pcDNA-sumo-1 construct and subcloned into pEBG vector at the BamH1 and XbaI

A

sites fusing to downstream cDNA of GST. All of the constructs were confirmed by DNAsequencing. 2.2. Antibodies：Rabbit polyclonal anti-phospho-S299/HSF4 and anti-HSF4 antibodies were generated by immunizing the rabbit with the peptide NH2-CKEEPAp[Ser]PGGD-

CONH2 and NH2-CKEEPASPGGD-CONH2 coupled with adjuvants (Abgent Biotechnology, Suzhou). The antibodies were purified from the serum by affinity purification. Rabbit polyclonal antibodies against Hsp25, beta-actin and GAPDH were from

SC RI PT

Sigma (St. Louise, USA). Rabbit antibodies against alpha B-crystallin, Flag-Tag, and HATag were from Santa Cruz Biotechnol (Dallas, USA).

2.3. Reconstitution of Hsf4b and Hsf4b mutants into the mLEC/Hsf4-/- cells: The retroviral vectors that express the empty vector, HA-Hsf4b, HA-Hsf4b/S299A or HA-Hsf4/S299D mutants were transiently transfected into 239-phoenix cells for 36 hours. The cell

U

supernatants were collected and mixed with 2 g/ml of polybrane. The viral supernatants

N

were subjected to infect the mLEC/Hsf4-/- cell line. After 24 hours infection, the cells were

A

selected in complete media containing 4  g/ml of blasticidin for 3-4 days. The cells were

M

pooled together. Four stable cell lines were generated (including mLEC/Hsf4-/-, mLEC/HA-

D

Hsf4b, mLEC/HA-Hsf4b/S299A and mLEC/HA-Hsf4b/S299).

TE

2.4 Immunoblotting and immunoprecipitation. The procedures of immunoblotting and

EP

immunoprecipitation assays have been described in a previous paper (Cui et al., 2015). 2.5 Immunoprecipitation kinase assay in vitro. The HLE-B3 cells were transfected with HA-

CC

MEK1 to activate the endogenous ERK1/2. The cell lysates were immunoprecipitated with anti-ERK1/2 antibody or IgG control. The co-precipitated ERK1/2 was incubated with

A

bacterial GST protein alone, GST-HSF4(282-318) and GST-HSF4(282-318)/S299A in the buffer containing ATP, and the proteins were immunoblotted with antibodies against phospho-S299 and phosphor-ERK1/2. The bacterial GST and GST-HSF4 truncates used in the kinase assay were shown by coommasie blue staining.

2.6 Quantitative RT-PCR: Total RNA was extracted with Trizol buffer following the kit protocol. 1  g of total RNA was used to synthesize the first strand of cDNA with the kit (promega, USA). The primers for amplifying the expression of Hsp25 and B-crystallin are:

SC RI PT

forward: 5’-CAGGACGAACATGGCTACA-3’, reverse: 5’-AGAGCGCACAGATTGACAG3(for Hsp25) and forward 5’-AAGAACGCCAGGACGAACAT-3’, reverse 5’-

GAGAGGATCCACATCGGCTG-3’ (for B-crystallin). The 18S RNA was used as the

internal control. The samples were mixed with the SYBR green mixture and run on the

qPCR machine. The sample values were generated against the standard curve created by

U

the same gene primer pair and normalized with the values of the 18S RNA. The data is

N

presented as relative fold change with respect to the control and represent three

M

A

independent experiments. The Student’s t-tests was used for statistically analysis. 2.7 Immunofluorescence assay: For lens, the mouse lens, which were isolated under the

D

microscope, were fixed in 4% paraformaldehyde /PBS for 2 days and then embedded in

TE

paraffin. The samples were sliced into 4 m slices with the microtome (Leica, Germany). After antigen retrieval, the slices were blocked in 3% BSA and 1% normal rabbit serum

EP

(Santa Cruz Biotechnol, Shanghai, China) for 1 hr, and then incubated with antibodies

CC

against Rabbit anti-phospho-Hsf4/S299. After washing with TBS buffer containing 0.025% NP-40, the tissue sections were incubated with Alexa Fluor 488-conjugated secondary

A

antibody for 1 hour. For cultured cells, the lens epithelial cells were grown on coverslips and fixed in 3.7% paraformaldehyde for 20 minutes. The cells were washed with PBST buffer and permeabilized with 0.2% NP-40 buffer for 5 minutes. After being blocked in 5% BSA buffer for 1hr, the cells were incubated with primary antibody for 1 hr and secondary antibody conjugated with Alexa Fluor 488 for 1hr. The nuclei were stained with Dapi. The

fluorescent signals were photographed with the fluorescence microscope Zeiss 540 under the 1000 index. 2.8 Nuclear-Cytoplasm fraction: the nuclear and cytosolic compartments were isolated

SC RI PT

following the protocol provided with the kit (Thermo Scientific TM, USA). Briefly, the epithelial and fiber tissues of the P3 mouse lens were separated with fine forceps under the microscope. The tissues were homogenized in cytosolic buffer (50mM Tris.HCl pH.7.4,

150mM NaCl, 1mM EDTA, 1x cocktail protease and phosphatase inhibitor) and incubated on ice for 10 minutes. The concentrated lysis buffer was added to the suspended cells and

U

incubated on ice for 1 minute. The lysates were centrifuged. The supernatants were

N

collected as cytosolic fraction. The pellets were re-suspended in nuclear lysis buffer for

A

nuclear fraction. The protein concentration was determined using the BCA kit. Equal

M

amount of proteins from the two fractions were used to immunoblot [15].

D

3. Results

TE

3.1 Mutation HSF4b/S299A upregulates the expression of HSP25 and alpha B-crystallin

EP

in lens epithelial cells

CC

S299 and P300 are in the phosphorylation-dependent sumoylation motif (ΨK294xExxS299P300, PDSM ) of HSF4b, which consists of a MAP kinase target site that

A

is conserved among different species (Fig. 1A). Mutation of S299A was reported to upregulate the HSF4 transcriptional activity in the Gal4-reporter system (Hietakangas et al., 2006), but its biological significance has not been completely studied. Our previous data indicated that HSF4b controls the expression of HSP25 and alpha B-crystallin in lens epithelial cells(Hu et al., 2013). To determine whether S299 is involved in regulating

HSF4 transcription in lens epithelial cells, we mutated S299 of HSF4b into Alanine (A) and phospho-mimetic Aspartic acid (D). We generated four stable cell lines (mLEC/Hsf4/- vector, mLEC/HA-Hsf4b, mLEC/HA-Hsf4b/S299A and mLEC/HA-Hsf4b/S299D) by

SC RI PT

infecting mLEC/Hsf4-/- cells with recombinant retroviruses expressing empty vector, HAHsf4b (wild-type), HA-Hsf4b/S299A and HA-Hsf4b/S299D (Fig.1B, lanes 1, 2, 3 and 4). The immunoblot results show that mutation HSF4/S299A significantly upregulates the protein expression of HSP25 and Alpha B-crystallin compared to wild-type HA-HSF4

(Fig.1B, comparing lanes 3 to 2), while mutation of HSF4/S299D did not (Fig.1B, lane 4).

U

Consistently, the quantitative RT-PCR results indicate that the expression of HSP25 and

N

alpha B-crystallin mRNAs increased 1.9 or 2.3 fold more in mLEC/HA-Hsf4b cells, and

A

3.9 and 5.3 fold more in mLEC/HA-Hsf4b/S299A cells than in mLEC/Hsf4-/-/vector (Fig.

M

1C). HSF4b/S299D does not change the mRNA levels of HSP25 and alpha B-crystallin compared to wild-type HA-HSF4b (Fig. 1C). These results suggest that the modification

D

of S299 negatively regulates HSF4b transcription activity. Our previous data showed

TE

that HSF4b binds to the HSP25 promoter to initiate transcription. We tested the ability of

EP

HSF4b/S299A and HSF4b/S299D to bind the HSP25 promoter using the ChIP assay. As the results indicate in Figure 1D, mutation S299A does not alter the ability of HSF4b

CC

to bind the HSP25 promoter, while mutation HSF4b/S299D increases this binding 2 fold when compared to HSF4b/wt (Fig. 1D, upper panel and bar graph). The

A

immunofluorescence assay shows that neither mutation S299A nor S299D affects HSF4 nuclear localization (Fig. 1E). Taken together, these results show that mutation of HSF4/S299 to alanine upregulates HSF4-mediated HSP25 and alpha B-crystallin

expression, which implies that posttranslational modification of S299 in wildtype HSF4 plays a critical role in regulating HSF4 activity. 3.2 MAP kinase upregulates the phosphorylation of HSF4b/S299

SC RI PT

To investigate whether HSF4b/S299 is regulated by phosphorylation, we made a rabbit polyclonal antibody by using a peptide antigen containing phospho-S299. The

results of the immunoprecipitation assay showed that this antibody recognizes the coimmunoprecipitated Flag-HSF4b, but not Flag-HSF4b/S299A (Fig. 2A, upper panel

U

lanes 2 and 3). Equal amounts of Flag-HSF4b and Flag-HSF4b/S299A were used for

N

the immunoprecipitation assay (Fig. 2A, low panel). The Myc-HSF4b proteins that co-

A

precipitated with anti-Myc antibody (Fig.2B, lane 3) or anti-Phospho-S299 antibody (Fig.

M

2B, lanes 4 and 5) are recognized by both anti-HSF4 and anti-Myc antibodies (Fig. 2B, lanes 3-5). To determine whether this antibody could directly detect the phospho-299 of

D

HSF4 in the cell lysates, we transiently transfected the constructs expressing Flag-

TE

HSF4b, Flag-HSF4b/S299A, Flag-HSF4/S299D and Flag-HSF4/T222A into HLE-B3 cells. The immunoblotting results showed that this rabbit polyclonal antibody recognizes

EP

the wild-type Flag-HSF4b and Flag-HSF4b/T222A mutant proteins (Fig. 2C, lanes 2 and

CC

5), but not the Flag-HSF4b/S299A and Flag-HSF4b/S299D mutants (Fig.2C, lanes 3 and 4). Furthermore, we treated the co-immunoprecipitated Flag-HSF4b and Flag-

A

HSF4b/S299A with calf intestine alkali phosphatase (CIAP) for 30 minutes to remove to the phosphate residue (Hogue, 1995) and tested whether this could block the antiphospho-S299 antibody from binding to the HSF4b protein. As the results indicate in Fig. 2D, CIAP reduces the ability of the rabbit anti-phospho-S299 antibody to bind the co-precipitated Flag-HSF4b protein (Fig. 2D lane 4) and the size of both Flag-HSF4b

and Flag-HSF4b/S299A proteins (Fig. 2D, middle panels, lanes 3 and 4). Anti-phosphoS299 antibody cannot react with the co-precipitated Flag-HSF4/S299A regardless of CIAP treatment (Fig. 2D lanes 5 and 6). Taken together, these results suggest that this

SC RI PT

rabbit polyclonal antibody can specifically recognize the phosphate residue on Serine 299. CIAP could not completely block the antibody from binding HSF4b (Fig. 2D, lane

4 ). We hypothesize that the dephosphorylation by CIAP might change the structure of HSF4, which in part impact the ability of the antibody to bind the dephosphorylated HSF4b.

U

Previous reports suggested that S299 and its flanking amino acids constitute the

N

PDSM motif (ΨK294xExxS299P) (Hietakangas et al., 2006), of which the SP motif is a

A

conserved target of MAP kinase (Xue et al., 2008). To determine whether S299 is

M

phosphorylated by MAPKinase, HLE-B3 lens epithelial cells were transfected with empty

D

vector, Flag-HSF4b, Flag-HSF4b/S299A, Flag-HSF4b plus HA-MEK1 or Flag-HSF4b

TE

+HA-MEK1+ U0126 (Fig.3A, lanes 1, 2, 3, 4 and 5). The immunoblotting results show that the phosphorylation of S299 is detected on Flag-HSF4b (Fig.3 lane 2) but not on the

EP

Flag-HSF4b/S299A mutant (Fig.3A, lane 3). MEK1 significantly increases the phosphorylation of S299 of HSF4b (Fig. 3A, compared lanes 2 to 4 and bar graph), and

CC

this increased phosphorylation is inhibited by U0126 (Fig.3A, lane 5 and bargraph). The

A

ectopic expression of Flag-HSF4b, Flag-HSF4b/S299A, MEK1 and phospho-ERK1/2 are shown in Fig.3A. Moreover, EGF is a common growth factor that activates the MAPKinase pathway. We treated the HLE-B3 cells expressing Flag-HSF4b with EGF (Fig. 3B, lane 4) or EGF plus U0126 (Fig.3B, lane 5) for 30 minutes, and the phosphorylation of HSF4/S299 was measured with anti-phospho-S299 antibody. The

results showed that the phosphorylation of Flag-HSF4/S299 is upregulated by EGF (compared lanes 4 to 3), and U0126 reduced this S299 phosphorylation. No phosphorylation signal is detected from the Flag-HSF4/S299A protein (Fig. 3B, lane 2).

SC RI PT

The expression of Flag-HSF4b and Flag-HSF4b/S299A proteins are indicated in Fig. 3B middle panel. These results suggest that the EGF-MEK1-ERK1/2 pathway is

responsible the phosphorylation of HSF4b/S299. To further prove this, we performed an in vitro kinase assay. The cells that express HA-MEK1 were immunoprecipitated with anti-ERK1/2 antibody (Fig. 3C, lanes 1-3) or IgG control (Fig. 3C, lane 4). The co-

U

precipitated ERK1/2 or IgG control were incubated with bacterial GST protein alone, or

N

GST-HSF4(282-318) or GST-HSF4(282-318)/S299A in the kinase buffer containing

A

ATP. The phosphorylation of S299 was detected by immunoblotting with the anti-

M

phospho-S299 antibody. The results showed that the co-precipitated ERK1/2 can phosphorylate S299 of GST-HSF4(282-318), but not of GST-HSF4(282-318) or GST

D

protein alone (Fig. 3C, comparing lanes 2 to 1 and 3). No phosphorylation of GST-

TE

HSF4(282-318) by IgG control was observed (Fig. 3C, lane 4). The co-precipitated

EP

ERK1/2 is detected by anti-phosphoERK1/2 antibody (Fig. 3C, middle panel). The bacterial GST, GST-HSF4(282-318) and GST-HSF4(282-318) proteins are shown by

CC

Coomassie blue staining (Fig. 3C, low panel). The result of the kinase assay confirmed

A

that S299 in HSF4b is a ERK1/2 phosphorylation site. 3.3 MEK1-ERK1/2 increases HSF4b sumoylation and downregulates HSF4b-mediated alpha B-crystallin Published reports show that the HSF4/S299A mutation downregulates HSF4b sumoylation in the tested Cos 7 cells(Hietakangas et al., 2006). To determine whether

the sumoylation of HSF4 is regulated by MEK1-ERK1/2 mediated phosphorylation of S299, HLE-B3 cells were transiently transfected with GST-sumo-1 alone, or GST-sumo1 together with Flag-HSF4b, Flag-HSF4b/S299A, Flag-HSF4b/S299A/K294R or with

SC RI PT

Flag-HSF4b+MEK1. The immunoblotting results showed that MEK1/2 increases FlagHSF4 sumoylation (Fig. 4A, lane 5) and U0126 inhibits MEK1-mediated HSF4

sumoylation (Fig. 4A, lane 6). No sumoylation is observed in Flag-HSf4b, Flag-

HSF4/S299A or Flag-HSF4/S299A/K294R ( Fig. 4A, lanes 2, 3 and 4) without MEK1.

Using the same condition in Fig. 4A, we performed the immunoprecipitation with anti-

U

Flag antibody followed by immunoblotting with anti-GST-sumo1 antibody. The results

N

showed that the slightly sumoylated Flag-HSF4b was immunoprecipitated with Flag-

A

HSF4b (Fig. 4B, lane 2), and MEK1 increased the amount of co-precipitated sumo-

M

HSF4b (Fig. 4B, lane 5). No sumo-HSF4/S299A or sumo-HSF4/S2991/K293R were coprecipitated (Fig. 4B, lanes 3 and 4) by anti-Flag antibody. These results suggest that

D

MEK1-ERK1/2 induced phosphorylation of S299 can upregulate HSF4 sumoylation. One

TE

function of sumoylation is to regulate protein stability. We tested stability of HSF4b,

EP

HSF4b/S299A and HSF4b/S299A/K293R protein, and found that the half-life of the HSF4b/S299A/K293R mutant, which has an impaired sumoylation motif in HSF4b

CC

(Hietakangas et al., 2006), is significantly shorter than that of HSF4b wild-type (Fig. 4C), while HSF4b/S299A is slightly short compared to HSF4b/wt. These results suggest that

A

phosphorylation of S299 regulates the stability of HSF4 in lens cells. To determine the regulatory effect of MEK1-mediated phosphorylation on HSF4 transcription activity, HLE-B3 cells were transiently transfected with empty vector, Flag-HSF4b, FlagHSF4b+MEK1, Flag-HSF4b/S299A and Flag-HSF4/S299A+MEK1(Fig.4D, lanes 1, 2, 3,

4 and 5). The immunoblotting results showed that alpha B-crystallin is induced by both Flag-HSF4b and Flag-HSF4/S299A, and the induction of alpha B-crystallin by FlagHSF4b/S299A is more than by Flag-HSF4b (Fig.4D, lanes 2 and 4). MEK1 inhibits

SC RI PT

HSF4-mediated alpha B-crystallin expression (Fig. 4D, lane 3 upper panel) but has less effect on HSF4/S299A-controlled alpha B-crystallin expression (Fig.4D, lane 5). These results suggest the phosphorylation of S299 by MEK1-ERK1/2 can inhibit HSF4controlled alpha B-crystallin expression in HLE-B3 cell line.

3.4 Phosphorylation of HSF4/S299 downregulates alpha B-crystallin expression in

N

U

ocular lens

A

Turning on or off HSF4 activity is important for postnatal lens development. The

M

above results showed that phosphorylation of HSF4/S299 mediates the fine tuning of HSF4 activity in lens epithelial cells (Fig.2, 3 and 4). To determine whether the

D

phosphorylation of S299/HSF4 is regulated during lens development, we measured the

TE

phosphorylation of HSF4/S299 in lens tissue at P3, P7, P14, 2 and 12 months old lens by immunoblotting with anti-phospho-S299/HSF4 antibody. As the results indicate in

EP

Figure 5A, phosphorylation of S299 is low in P3 lens compared to that in P7 and P14

CC

lens, while the level of total HSF4 protein did not change (Fig.5A lanes 1-3). Both phosphorylated S299 and total HSF4 protein decreased in 2 and 12 months old lens

A

(Fig.5A lanes 4 and 5). Interestingly, ERK2 activity is lower in P2 lens than in P7, P14, 2 and 12 months old lens (Fig.5A middle panel). Accordingly, we proposed that ERK1/2 might participate in phosphorylating HSF4/S299 and tuning HSF4 transcription in early postnatal lens. To prove this, mouse P3 lens were cultured ex vivo in the media containing 10ng/ml EGF or EGF plus U0126. We found that EGF upregulates the

phosphorylation of the lens HSF4/ S299 and inhibits the expression of alpha B-crystallin (Fig. 5B lane 2). U0126 inhibits EGF-mediated HSF4/S299 phosphorylation and the downregulation of alpha B-crystallin expression (Fig. 5B lane 3). These results suggest

SC RI PT

that activation of ERK1/2 correlates with HSF4/S299 phosphorylation and the downregulation of HSF4-mediated alpha B-crystallin expression in early postnatal lens. Meanwhile, we studied the localization of phospho-S299/HSF4 in P3 lens. We observed phosphorylated S299/HSF4 in anterior epithelial cells and equatorial primary and

secondary nucleated fiber cells (Fig. 6A) in P3 lens. In the anterior epithelia, phospho-

U

S299 is detected mainly in the cytoplasm (Fig.6B), while in the fiber cells, it is

N

predominantly in the nucleus (Fig.6 C). We confirmed the localization of phospho-

A

S299/HSF4 in P3 lens with a cytoplasm-nuclear fraction assay. Similar to the

M

immunofluorescence results, we found that both HSF4 and its phospho-S299/HSF4 are express predominantly in cytoplasm of the epithelial cells and in the nucleus of fiber

D

cells (Fig. 6D lanes 1, 2, 3 and 4). Even so, the levels of both HSF4 and its

TE

phosphorylated S299 are much higher in the fiber cells compared to that in the epithelial

EP

cells (Fig.6D). There is much less phospho-S299/HSF4 than total HSF4 protein in both the epithelia and fiber tissue. HSP25 and alpha B-crystallin, which are downstream

CC

targets of HSF4 during lens development (Fujimoto et al., 2004), are highly expressed in fiber tissue compared to that in epithelium. BRG1 is used as nuclear marker while

A

GAPDH is used for the cytoplasm. Accordingly, we propose that HSF4 is more active in the fiber cells than in the anterior epithelium. Increasing the amount of phosphorylated HSF4 in P3 lens fiber should be responsible for tuning down HSF4 transcription activity. 4. Discussion

Proper tuning of HSF4 transcription activity is essential for postnatal lens development. Mutation S299A was reported to upregulate HSF4 transcription activity in the Gal4-reporter assay in Cos7 cell line (Hietakangas et al., 2006; Tanabe et al., 1999),

SC RI PT

which suggested that posttranslational modification of S299 is involved in modulating HSF4 transcription activity. In this paper, we generated a rabbit polyclonal antibody that specifically recognizes phospho-S299/HSF4. Using this antibody, we proved that

HSF4/S299 is modified by phosphorylation (Fig. 2). We found that phosphorylation of HSF4 at S299 has at least two regulatory effects. One is to tune down HSF4’s

U

transcription activity (Fig. 4C and 5B), and the other is to increase HSF4 sumoylation

N

and protein stability (Fig. 4). Moreover, we found that the EGF-MEK1-ERK1/2 pathway

A

is responsible for S299 phosphorylation. Ectopic MEK1 or EGF upregulated the

M

phosphorylation of HSF4b at S299, and downregulated HSF4-controlled alpha Bcrystallin in lens epithelial cells as well as in P3 lens ex vivo (Fig.3 and 4). The level of

D

phospho-S299/HSF4 is lower in P3 lens than in P7 and P14 lens, which is concomitant

TE

with ERK1/2 activity (Fig. 5A). These results suggest that the transcription activity of

EP

HSF4 is modulated in part by ERK1/2-mediated phosphorylation of S299 during early postnatal lens development. The phosphorylation of S299 may be a marker to reflect

CC

the transcription activity of HSF4 during lens development.

A

S299 is in the regulatory domain of the HSF4b protein and is a critical amino acid

in the PDSM motif (ΨKxExxSP) , which is conserved in most of transcription factors (Hietakangas et al., 2006). The PDSM motif is known to downregulate HSF4b transcription activity in the Gal4-reporter assay. However, the biological function of the posttranslational modification of S299 is not well characterized. Since HSF4b is the

unique splice variant expressed in lens and controls the transcription of small heat shock proteins HSP25 and alpha B-crystallin in lens epithelial cell lines(Hu et al., 2013; Nakai et al., 1997; Zhang et al., 2010), measuring the expression of HSP25 and alpha B-

SC RI PT

crystallin can reflect the transcription activity of HSF4. As the data indicates in Figure 1, HSF4b/S299A significantly upregulates the expression of HSP25 and alpha B-crystallin, which agrees with previous reports suggesting that the posttranslational modification of S299 negatively modulates HSF4 transcription (Hietakangas et al., 2006; Zhang et al., 2010). However, we did not observe a difference between HSF4 and HSF4/S299A in

U

their ability to bind the HSP25 promoter (Fig. 1D), suggesting that mutation S299A does

N

not impact the ability of HSF4 to bind HSEs. Our previous data showed that mutation

A

S299A impairs the association between HSF4 and transcription inhibitor Daxx (Zhang et

M

al., 2010), which implies that phosphorylation of S299 may regulate HSF4’s ability to associate with the basic components of the transcription machinery (Frejtag et al.,

D

2001). The underlying mechanisms of HSF4b/S299A regulation is still under

TE

investigation. Interestingly, in contrast to the positive regulatory effect of S299A,

EP

mutation HSF4b/S299D, which is usually used to imitate the phosphorylated amino acid, does not change the expression of HSP25 and alpha B-crystallin when compared to wild

CC

type HSF4 (Fig. 1B), although HSF4b/S299D more active binds to the HSP25 promoter than wild type HSF4b (Fig. 1D). This suggests that either mutation of HSF4/S229 to

A

aspartic acid cannot completely imitate phosphorylated serine, and this situation has been reported in other factors (Peters et al., 2006), or phosphorylation of S299 is a transient event and is just responsible for turning down the high transcription activity of HSF4. The data in Fig. 4 suggest that phosphorylation of S299 induces HSF4

sumoylation, which is known to repress HSF1 and HSF2 transcription activity. Sumoylation may be another mechanism by which phosphorylation of S299 turns down HSF4 transcritpion activity(Anckar et al., 2006; Xu et al., 2012).

SC RI PT

MAP Kinase is the central signaling hub linking diverse signaling pathways to the transcription factors. S299 and its downstream P300 of HSF4 make up the MAP kinase phosphorylation motif (S/T)P (Xue et al., 2008). To study whether S299 is a substrate of MAP kinase, we first made a rabbit polyclonal antibody against phospho-S299. The data in Figure 2 proved that this antibody can specifically recognize phospho-S299 of HSF4.

U

Using this specific anti-phospho-S299 antibody, we found that MEK1 or EGF can

N

upregulate the phosphorylation of S299, and the phosphorylation is inhibited by ERK1/2

A

inhibitor U0126. These results suggest that S299 in HSF4 is a phosphorylation target of

M

ERK1/2, and this is further confirmed by the result of the in vitro kinase assay (Fig. 3C).

D

Phosphorylation by ERK1/2 can initiate many signaling cascades and lead to

TE

diverse outcomes that are specific to individual transcription factors. Phosphorylation of HSF1 at S303 by ERK1/2 can induce sumoylation and downregulation of HSF1

EP

transcription activity(Hietakangas et al., 2003). In Figures 4 and 5, the MEK1 or EGF

CC

signaling pathway can downregulate HSF4-mediated alpha B-crystallin expression in the HLE cell line in vitro (Fig.4D) and in P3 lens ex vivo (Fig. 5B), and the downregulation

A

can be reversed by ERK1/2 inhibitor U0126. These results suggest that ERK1/2 phosphorylation negatively regulates HSF4 transcription activity in part by phosphorylating S299. We show that this phosphorylation leads to increased HSF4 sumoylation. However, it is still unclear how sumoylation regulates HSF4 transcription activity.

Collectively, the above results suggest that phospho-S299 can be a used as a marker for the transcription activity of HSF4. With this in mind, we studied the phosphorylation of S299/HSF4 during mouse lens development. Phospho-S299 is lower

SC RI PT

in P3 lens than in P7 and P14 lens, and the level of phosphorylation of S299 is negative correlated with ERK1/2 activity, suggesting that ERK1/2 participates in modulating the

transcription activity of HSF4 during postnatal lens development. Moreover, in P3 lens, we found that HSF4 is predominantly in the nucleus of fiber cells and the cytoplasm of epithelium cells (Fig. 6D). Concomitantly, the expression of HSP25 and alpha B-

U

crystallin proteins is more in fiber tissue than in the epithelium (Fig. 6D), indicating that

N

HSF4 is more active in the fiber tissue than in epithelium. Interestingly, phospho-

A

S299/HSF4 proteins are also upregulated in the nucleus of the fiber cells compared to

M

that of epithelial cells (Fig. 6D). We propose that the increased phosphorylation of S299 might be responsible for turning down the high HSF4 activity to a proper level during

TE

5. Conclusion

D

lens development.

EP

We made a rabbit polyclonal antibody against phospho-S299/HSF4, and with this

CC

antibody we determined that HSF4/S299 is a target of ERK1/2 phosphorylation. Hyperphosphorylation of HSF4/S299 by the EGF-MEK1-ERK1/2 pathway increased

A

HSF4 sumoylation and downregulated alpha B-crystallin expression. Phosphorylation of S299 may be used as an indicator reflecting HSF4 transcription activity in lens development.

Conflict of interest: authors in this paper declare no conflict interest in this paper

SC RI PT

Acknowledgement: We thank Muhan Hu, a student in the MD/PhD program at the University of Alabama at

Birmingham, Birmingham, AL, USA for English editing. This work is support by the Chinese

U

NSFC grant foundations 31571451, 81570825, 81400387, U1604171, U1404810 and 81770911

A

N

References

A

CC

EP

TE

D

M

Anckar, J., Hietakangas, V., Denessiouk, K., Thiele, D.J., Johnson, M.S., Sistonen, L., 2006. Inhibition of DNA binding by differential sumoylation of heat shock factors. Mol Cell Biol 26(3), 955-964. Berry, V., Pontikos, N., Moore, A., Ionides, A.C.W., Plagnol, V., Cheetham, M.E., Michaelides, M., 2017. A novel missense mutation in HSF4 causes autosomal-dominant congenital lamellar cataract in a British family. Eye (Lond). Bu, L., Jin, Y., Shi, Y., Chu, R., Ban, A., Eiberg, H., Andres, L., Jiang, H., Zheng, G., Qian, M., Cui, B., Xia, Y., Liu, J., Hu, L., Zhao, G., Hayden, M.R., Kong, X., 2002. Mutant DNA-binding domain of HSF4 is associated with autosomal dominant lamellar and Marner cataract. Nat Genet 31(3), 276-278. Cui, X., Liu, H., Li, J., Guo, K., Han, W., Dong, Y., Wan, S., Wang, X., Jia, P., Li, S., Ma, Y., Zhang, J., Mu, H., Hu, Y., 2016. Heat shock factor 4 regulates lens epithelial cell homeostasis by working with lysosome and anti-apoptosis pathways. Int J Biochem Cell Biol 79, 118-127. Cui, X., Xie, P.P., Jia, P.P., Lou, Q., Dun, G., Li, S., Liu, G., Zhang, J., Dong, Z., Ma, Y., Hu, Y., 2015. Hsf4 counteracts Hsf1 transcription activities and increases lens epithelial cell survival in vitro. Biochim Biophys Acta 1853(3), 746-755. Frejtag, W., Zhang, Y., Dai, R., Anderson, M.G., Mivechi, N.F., 2001. Heat shock factor-4 (HSF4a) represses basal transcription through interaction with TFIIF. J Biol Chem 276(18), 14685-14694. Fujimoto, M., Izu, H., Seki, K., Fukuda, K., Nishida, T., Yamada, S., Kato, K., Yonemura, S., Inouye, S., Nakai, A., 2004. HSF4 is required for normal cell growth and differentiation during mouse lens development. EMBO J 23(21), 4297-4306. Graw, J., 2004. Congenital hereditary cataracts. Int J Dev Biol 48(8-9), 1031-1044.

A

CC

EP

TE

D

M

A

N

U

SC RI PT

Hietakangas, V., Ahlskog, J.K., Jakobsson, A.M., Hellesuo, M., Sahlberg, N.M., Holmberg, C.I., Mikhailov, A., Palvimo, J.J., Pirkkala, L., 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(8), 2953-2968. Hietakangas, V., Anckar, J., Blomster, H.A., Fujimoto, M., Palvimo, J.J., Nakai, A., Sistonen, L., 2006. PDSM, a motif for phosphorylation-dependent SUMO modification. Proc Natl Acad Sci U S A 103(1), 45-50. Hogue, B.G., 1995. Bovine coronavirus nucleocapsid protein processing and assembly. Adv Exp Med Biol 380, 259-263. Hu, Y., Mivechi, N.F., 2006. Association and regulation of heat shock transcription factor 4b with both extracellular signal-regulated kinase mitogen-activated protein kinase and dualspecificity tyrosine phosphatase DUSP26. Mol Cell Biol 26(8), 3282-3294. Hu, Y.Z., Zhang, J., Li, S., Wang, C., Chu, L., Zhang, Z., Ma, Z., Wang, M., Jiang, Q., Liu, G., Qi, Y., Ma, Y., 2013. The transcription activity of heat shock factor 4b is regulated by FGF2. Int J Biochem Cell Biol 45(2), 317-325. Mellersh, C.S., Pettitt, L., Forman, O.P., Vaudin, M., Barnett, K.C., 2006. Identification of mutations in HSF4 in dogs of three different breeds with hereditary cataracts. Vet Ophthalmol 9(5), 369-378. Min, J.N., Zhang, Y., Moskophidis, D., Mivechi, N.F., 2004. Unique contribution of heat shock transcription factor 4 in ocular lens development and fiber cell differentiation. Genesis 40(4), 205-217. Morange, M., 2006. HSFs in development. Handb Exp Pharmacol(172), 153-169. Nakai, A., Tanabe, M., Kawazoe, Y., Inazawa, J., Morimoto, R.I., 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(1), 469-481. Peters, G.A., Li, S., Sen, G.C., 2006. Phosphorylation of specific serine residues in the PKR activation domain of PACT is essential for its ability to mediate apoptosis. J Biol Chem 281(46), 35129-35136. Tanabe, M., Sasai, N., Nagata, K., Liu, X.D., Liu, P.C., Thiele, D.J., 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(39), 27845-27856. Xu, Y.M., Huang, D.Y., Chiu, J.F., 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(5), 2625-2634. Xue, Y., Ren, J., Gao, X., Jin, C., Wen, L., Yao, X., 2008. GPS 2.0, a tool to predict kinase-specific phosphorylation sites in hierarchy. Mol Cell Proteomics 7(9), 1598-1608. Zhang, J., Hu, Y.Z., Xueli, L., Li, S., Wang, M., Kong, X., Li, T., Shen, P., Ma, Y., 2010. The inhibition of CMV promoter by heat shock factor 4b is regulated by Daxx. Int J Biochem Cell Biol 42(10), 1698-1707. Zhang, J., Ma, Z., Wang, J., Li, S., Zhang, Y., Wang, Y., Wang, M., Feng, X., Liu, X., Liu, G., Lou, Q., Cui, X., Ma, Y., Dong, Z., Hu, Y.Z., 2014. Regulation of Hsf4b nuclear translocation and transcription activity by phosphorylation at threonine 472. Biochim Biophys Acta 1843(3), 580-589.

Figure legends

Figure 1. The regulation of HSF4 transcription activity by mutation HSF4/S299A and

SC RI PT

S299D. A, the conserved PDSM motif of HSF4b among different species. B,

Immunoblotting assay to determine the protein expression of alpha B-crystallin

(CRYAB), HSP25, HA-HSF4 and -actin in the cell lines mLEC/Hsf4-/- (lane 1),

mLEC/HA-Hsf4b (lane 2), mLEC/HA-Hsf4b/S299A (lane 3) and mLEC/HA-Hsf4b/S299D

U

( lane 4) ; B, Quantitative RT-PCR to measure the mRNA expression of alpha B-

N

crystallin and HSP25 in the four cell lines used in (A); C, ChIP assay to measure the

A

binding of HSF4b (lane 2), HSF4/S299A (lane 3) and HSF4/S299D (lane 4) to HSE

M

elements in the HSP25 promoter. Lanes 5-7 are input control. ns: unspecific band. The bar graph is the densitometry quantification of bands in (C) from three repeated

D

experiments; D, Immunofluorescence assay. The cell lines mLEC/ HA-Hsf4b,

TE

mLEC/Hsf4/S299A and mLEC/Hsf4/S299D were incubated with anti-HA antibody

DAPI.

EP

followed by Alexa Fluor 488 conjugated secondary antibody. The nucleus is stained with

CC

Figure 2. Rabbit anti-phospho-S299 antibody can specifically recognize HSF4b/S299. A, immunoprecipitation assay, the cells express Flag-HSF4b (lane 2) or Flag-

A

HSF4b/S299A (lane 3) were immunoprecipitated with anti-Flag antibody or IgG control (lane 1) following by immunoblotting with antibodies against phospho-S299 and Flag (upper and middle panels). The ectopic expression of Flag-HSF4b, Flag-HSF4b/S299A and endogenous -Actin in the cell lysates were indicated in low panels; B,

Immunoprecipitation assay. The cells expressing myc-Hsf4b were used for the immunoprecipitation with antibodies again Myc-tag (lane 3), phospho-S299 antibody clone numbers 1 and 2 (lanes 4 and 5) or control IgG (lane 2). The co-

SC RI PT

immunoprecipitates were immunoblotted with antibody against HSF4 or Myc-tag. Lane 1 is cell lysate input; C, Immunoblot of the cell lysates containing empty vector (lane 1), Flag-Hsf4b/wt, (lane 2), Flag-Hsf4b/S299A (lane 3), Flag-Hsf4b/S299D (lane 4) and

Flag-Hsf4/T22A (lane 5) with antibodies against phospho-S299, Flag-tag and -Actin; D, Calf intestine alkali phosphatase (CIAP) reduces phosphorylation of S299 in the co-

U

immunoprecipitation assay. The co-immunoprecipitated Flag-HSF4b and Flag-

N

HSF4b/S299A were treated with CIAP for 30 minutes and then immunoblotted with

A

antibodies against phospho-S299 and Flag (the top two panels). The cell lysates used

M

for above immunoprecipitation were immunoblotted with anti--actin antibody (low

D

panel).

TE

Figure 3. Phosphorylation of S299 is regulated by EGF or MEK1. A, Ectopic expression of MEK1 upregulates HSF4b/S299 phosphorylation. HLE-B3 cells were transiently

EP

transfected with empty vector (lane 1), Flag-Hsf4b (lane 2), Flag-HSF4b/S299A(lane 3),

CC

Flag-HSF4b and HA-MEK1 (lanes 4 and 5). The cells expressing Flag-Hsf4b + HAMEK1 were treated with U0126 for 6hrs (lane 5). The immunoblotting was performed

A

with antibodies against phospho-S299, Flag, phospho-44/42 ERK1/2, HA-MEK1 and actin. The bar graph represents the ratio of phosphor-S299 verse Flag-HSF4 that were quantified by densitometry in (A); B, EGF upregulates the S299 phosphorylation. The HLE-B3 cells were transiently transfected with empty vector (lane 1), Flag-Hsf4b/S299A (lane 2), Flag-HSF4b (lanes 3-5). The cells expressing Flag-Hsf4b were treated with

EGF (10ng/ml) for 20 minutes (lane 4) or with EGF plus U0126 (lane 5). The cell lysates were immunoblotted with antibodies against phospho-S299, Flag and -actin; C, ERK1/2 phosphorylates S299/HSF4 in immunoprecipitation kinase assay in vitro. The HLE-B3

SC RI PT

cells, which were transfected with HA-MEK, were immunoprecipitated with anti-ERK1/2 antibody. The immunoprecipitated ERK1/2 was incubated with GST (lane 1), GST-

HSF4b (282-318) (lane 2) or GST-HSF4b(282-318)/S299A (lane 3) in kinase buffer

containing ATP following by immunoblotting with antibodies against phospho-S299 and phospho-ERK1/2. The bacterial GST and GST-HSF4 truncates were stained in

U

coomassie blue.

N

Figure 4. Phosphorylation of HSF4/S299 by MEK1 induces HSF4 sumoylation and

A

downregulates alpha B-crystallin expression in HLE-B3 cells. A, MEK1 induces HSF4

M

sumoylation. The HLE cells were transiently co-transfected with GST-sumo-1 together

D

with empty vector (lane 1) or with plasmids expressing Flag-Hsf4b (lane 2), Flag-

TE

Hsf4b/S299A (lane 3), Flag-Hsf4b/S299A/K294R (lane 4), or Flag-Hsf4b plus HA-MEK1 (lanes 5 and 6). The cells co-expressing GST-sumo-1, Flag-Hsf4b and HA-MEK1 were

EP

treated with U0126 (lane 6). The cell lysates were immunoblotted with HSF4 antibody;

CC

B, MEK1 increases HSF4 sumoylation in immunoprecipitation assay, the cells lysates used in A (lanes 1-5) were immunoprecipitated by anti-Flag antibody. The co-

A

precipitates were immunoblotted with antibodies against GST-sumo-1 and Flag; C, mutation of HSF4/S299A/K294R shortens the HSF4 half-life. The cells expressing FlagHSF4b (lanes 1-3), Flag-HSF4b/S299A (lanes 4-6) and Flag-HSF4b/S299A/K294R (lanes 7-9). The cells were treated cyclohexamide for 0, 6 and 12 hrs and then were immunoblotted with antibodies against Flag and -actin; The densitometry of the Flag

band at 6 and 12 hrs that were normalized by densitometry of -actin band were divided by that at 0hr was accounted for the fold change; D, MEK1 inhibits alpha B-crystallin expression. The HLE-B3 cells that express empty vector (lane1), Flag-Hsf4b (lane 2),

SC RI PT

FlagHsf4b+ HA-MEK1 (lane 3), Flag-Hsf4b/S299A (lane 4) and Flag-Hsf4b/S299A+HAMEK1(lane 5) were immunoblotted with antibodies against alpha B-crystallin (CRYAB), phospho-S299, Flag-tag, phospho-ERK1/2, HA-MEK1 and -Actin.

Figure 5. Phosphorylation of S299/HSF4 is regulated in postnatal lens. A, Immunoblot of

U

the expression of phospho-S299, HSF4, phospho-ERK1/2, ERK1/2 and -Actin in P3,

N

P7, P14, P28 and P60 days old mouse lens; B, EGF upregulates phospho-S299/HSF4

A

and downregulates alpha B-crystallin expression in P3 lens ex vivo. P3 lens were

M

cultured in M199 media containing EGF (lane 2) or EGF plus U0126 (lane 2). The lens lysates were immunoblotted with antibodies against alpha B-crystallin, phospho-

D

S299/HSF4, HSF4, phosphor-ERK1/2, ERK1/2 and -Actin.

TE

Figure 6. Localization of phosphorylated S299/HSF4 in P3 lens. P3 lens were subjected

EP

to immunofluorescence staining with antibody against phospho-S299/HSF4. The nuclei were stained with DAPI. A, Image under low magnification; B and C, Images under high

CC

magnification, Bar scale: 100px; D, Cytosol-nucleus fraction assay. P3 lens were used to separate the anterior epithelia from the fiber cells (n=8). The nuclear and cytosolic

A

fractions were separated with kits following the protocol. The fractions were immunoblotted with antibodies against phosphor-S299/HSF4, endogenous HSF4, HSP25, alpha B-crystallin. BRG1 was used for nuclear marker, while GAPDH was for a cytosolic loading control.

D

TE

EP

CC

A

SC RI PT

U

N

A

M

D

TE

EP

CC

A

SC RI PT

U

N

A

M

D

TE

EP

CC

A

SC RI PT

U

N

A

M

D

TE

EP

CC

A

SC RI PT

U

N

A

M

D

TE

EP

CC

A

SC RI PT

U

N

A

M

D

TE

EP

CC

A

SC RI PT

U

N

A

M