Targeted replacement of HSF1 phosphorylation sites at S303/S307 with alanine residues in mice increases cell proliferation and drug resistance

i. Summary/Abstract

Mammalian heat shock factor HSF1 transcriptional activity is controlled by multitude of phosphorylations that occurs under physiological conditions or following exposure of cells to a variety of stresses. One set of HSF1 phosphorylation is on serine 303 and serine 307 (S303/S307). These HSF1 phosphorylation sites are known to repress its transcriptional activity. Here, we describe a knock-in mouse model where these two serine residues were replaced by alanine residues and have determined the impact of these mutations on cellular proliferation and drug resistance. Our previous study using this mouse model indicated the susceptibility of the mutant mice to become obese with age due to an increase in basal levels of heat shock proteins (HSPs) and chronic inflammation. Since HSF1 transcriptional activity is increased in many tumor types, this mouse model may be a useful tool for studies related to cellular transformation and cancer.

ii. Introduction

Heat shock transcription factor (HSF1) is an evolutionarily conserved gene and is widely expressed in most tissues in human and other mammals. HSF1 responds to a variety of stressors and regulates the expression of a wide range of genes at transcriptional level to maintain cellular proteostasis and support cell survival [1-4]. Under non-stressed physiological conditions, HSF1 exists as an inactive monomeric form and is distributed mainly in the cytoplasm. Upon exposure of cells to heat shock or other stresses, HSF1 is activated through a multistep processes, including trimerization, nuclear translocation, binding to the heat shock element (HSE) on promoter of target genes, and initiating gene transcription [1, 4-6]. HSF1 activation and inactivation are regulated by multiple factors. For example, heat shock proteins (HSPs), whose transcription are upregulated following HSF1 activation in stressed cells, provide negative feedback to inactivate HSF1. Studies show that both HSP70 and HSP90 bind to HSF1 and inhibit its DNA binding capacity [7-9]. In addition, the posttranslational modification of HSF1 protein, including phosphorylation, acetylation and sumoylation, are important mechanisms for regulation of its activation, inactivation and stability.

Phosphorylation of HSF1 plays an important role in regulation of its activity and has been widely investigated. Phosphorylation of HSF1 is strongly induced following exposure of cells to heat shock, and an increase in its phosphorylation correlates with its transcriptional activity. So thus far, 15 serine residues (S121, S216, S230, S292, S303, S307, S314, S319, S320, S326, S344, S363, S368, S419, S444,) and 4 threonine residues (T142, T323, T367, and T369) of HSF1 protein, which can be phosphorylated under stress and/or unstress physiological conditions, has been reported [6, 10-12]. Notably, while phosphorylation of most of these serine or threonine residues do not affect HSF1 trans-activation capacity, it has been demonstrated that several serine residues (S230, S362, S363, S303 and S307) contribute to HSF1 activity. Phosphorylation of S326 play a positive stimulatory role in HSF1 activation by facilitating the association between HSF1 and the coactivator Daxx and has been widely used as marker for HSF1 activation [13]. Additionally, it has been reported that phosphorylation of S230 and S419 of HSF1 confers stimulatory function toward HSF1 activation [14, 15]. On the other hand, constitutive phosphorylation of S303, S307, and S363 represses the activation capacity of HSF1 [16-18]. It has been proposed that phosphorylation of S303 and S307 promotes the binding of the scaffolding protein 14-3-3ε to the HSF1 and subsequently sequesters HSF1 in the cytoplasm [19, 20]. Along this line, phosphorylation of S303/S307 residues is required for HSF1 degradation by facilitating its binding to FBXW7 ubiquitin ligase [21].

The functions of the endogenous phosphorylation of S303/S307 residues on cellular processes under physiological and stressed conditions in vivo remains elusive. Here, we generated a knock-in mouse model in which the S303 and S307 of HSF1 were mutated to alanine residues (named Hsf1^303A/307A). Our previous studies show that lack of S303 and S307 phosphorylation reduces the threshold of HSF1 activation, and that HSF1 is mildly activated in mouse embryo fibroblasts (MEFs), as well as in multiple tissues in vivo under non-stressed physiological conditions [22]. In the current study, we describe in detail the strategy and method for the generation of Hsf1^303A/307A mouse line. We further provide evidence on the effects of HSF1 activation on cellular proliferation, tumorigenesis and drug sensitivity using MEFs and Hsf1^303A/307A mouse line. The Hsf1^303A/307A knock-in MEFs and mice are ideal cellular and animal models for studying the role of S303 and S307 on HSF1 function in vitro and in vivo.

iii. Materials

Reagents:

Molecular biology:

LB Broth (Fisher Scientific, cat no.AAH2667636)

Agar powder (Fisher Scientific, cat no.AAA1075236)

Chloroform (Fisher Scientific, cat no. J67241.AP)

Isopropanol (Fisher Scientific, cat no. T036181000)

Ethanol (Fisher Scientific, cat no. T038181000)

Agarose (Invitrogen, cat no. 16500500)

Sodium chloride (Fisher Scientific, cat no. 447302500)

dNTP (Fisher Scientific, cat no. FERR0181)

Tris (Fisher Scientific, cat no. 17926)

EDTA (Fisher Scientific, cat no. 17892)

Sodium Dodecyl Sulfate (SDS) (Fisher Scientific, cat no. 15525017)

Polyvinylidene fluoride membrane (Fisher Scientific, cat no. 88518)

Sodium Deoxycholate (Fisher Scientific, cat no. 89905)

NP-40 (Fisher Scientific, cat no. J61055.AE)

Phosphatase Inhibitor Cocktail (Fisher Scientific, cat no. 78420)

Trizol (Invitrogen, cat no. 15596026)

Diethylnitroseamine (Sigma, cat no. 73861)

Southern blot:

SSPE (20X) (Fisher Scientific, cat no. AM9767)

Denhardt’s Solution (Fisher Scientific, cat no. 750018)

Sodium hydroxide (Fisher Scientific, cat no. A16037.36)

³²P-ATP (Perkin Elmer, NEG002A100UC)

Nylon Membrane (for Southern blot)( Fisher Scientific, cat no.P177016)

Cell culture and histology:

Dulbecco's Minimal Essential Medium (Corning, cat no. MT10102CV)

Fetal Bovine Serum (FBS)(Gibco, cat no. 26140079)

Crystal violet (Fisher Scientific, cat no. 405835000)

37% Formaldehyde (Fisher Scientific, cat no. 119690250)

Hematoxylin & Eosin staining kit (Abcam, cat no. ab245880)

Drug for selection of clones and treatment:

Ganciclovir (Fisher Scientific, cat no.AC461710250)

G418 Sulfate (Corning, cat no.30234CR)

Puromycin (MP Biomedicals, cat no. 0219453910)

Blasticidin (MP Biomedicals, cat no. MP215047783)

Doxorubicin (Fisher Scientific, cat no. J64000.MF)

Etoposide (Fisher Scientific, cat no. J63651.MC)

Enzymes:

T4 DNA ligase (Fisher Scientific, cat no. 15224017)

Taq DNA polymerase (Fisher Scientific, cat no. FEREP0406)

pfu Tubor DNA polymerase (Agilent, cat no. 600250)

XhoI Restriction Endonuclease (New England Biolabs, cat no. R0146L)

ClaI Restriction Endonuclease (New England Biolabs, cat no. R0197L)

SacI Restriction Endonuclease (New England Biolabs, cat no. R3156L)

BglII Restriction Endonuclease (New England Biolabs, cat no. R0144L)

NotI Restriction Endonuclease (New England Biolabs, cat no. R0189L)

Plasmids:

BAC clone PR23-266H9 (BACPAC Resource Center)

pBluescript II KS plasmid (pBS) (Agilent, cat no. 212207)

pKT1 plasmid (gift from Dr. Gail Martin, (27))

phage DNA vector λDASHII-254-2TK (gift from Dr. N. R Manley, University of Georgia, Athens, GA)

Commercial kits:

PCR-based site-directed mutagenesis (Agilent, cat no. 200523)

Gigapack II Packaging Extract (Agilent, cat no. 200201,)

Decalabel DNA labeling kit (Fisher Scientific, cat no. K0622)

DH5alpha competent bacteria (Fisher Scientific, cat no. 18265017)

iScript cDNA synthesis kit (Bio-Rad, cat no. 170-8891)

iQ SYBR green Supermix (Bio-Rad, cat no. 708880)

iv. Methods

1. Generation of Hsf1^303A/307A knock-in mice

To replace the HSF1 wild-type (WT) allele S303/S307 with 303A/307A mutant allele, we constructed a targeting vector in which a 3007 bp proximal, 1252 bp neomycin-Lox P site (for positive selection of embryonic stem (ES) cells)), and a 3809 bp distal DNA fragments were ligated into the targeting vector (Figure 1A). The proximal and distal DNA fragments were amplified by PCR with high fidelity DNA polymerase (pfu Tubor DNA polymerase (Agilent) and using a BAC clone PR23-266H9 (BACPAC Resource Center), which contained the entire HSF1 gene as template. The sequences of the primers used for the PCR reactions were as follow: (i) For the proximal DNA fragment, forward primer: 5’-CCGCTCGAGCAAGGAGCTGAAGGGATCTGC-3’ (contains a Xho I site), reverse primer: 5’-CCGGAGCTCGCTCTGCCCAAGGTCATAAAGG-3’ (contains a Sac I site); (ii) for the distal fragment, forward primer: 5’-CCGATCGATTGAAAATGGGTGGAACTAACTT-3’ (contains a Cla I site), reverse primer: 5’-CCGCTCGAGGCACACTCCTCTTAGTCTGGG-3’ (contains a Xho I site). The PCR fragments were digested with the indicated restriction enzymes and inserted into pBluescript II KS plasmid (pBS). The plasmids were introduced into DH5alpha competent bacteria (18265017, Fisher Scientific) and a single bacterial clone was selected and amplified. The DNA sequence of the inserted HSF1 fragment was verified by DNA sequencing. To generate the neomycin-Lox P fragment, we cloned the neomycin cassette and Lox P (sequence: 5’-ATAACTTCGTATAGCATACATTATACGAAGTTAT-3’) fragment into pKT1 plasmid (gift from Dr. Gail Martin [23]). The neomycin-Lox P fragment was released by Sac I and Cla I restriction enzymes. The 303A and 307A mutation were introduced by PCR-based site-directed mutagenesis (200523, Agilent) using primers containing point mutations (underlined), forward primer: 5’-AAGCAAGAGCCCCCCGCCCCACCTCACGCCCCTCGGGTACTGGAG-3’, reverse primer: 5’-CTCCAGTACCCGAGGGGCGTGAGGTGGGGCGGGGGGCT-CTTGCTT-3’ (using the pBluescript plasmid containing the distal arm fragment as a substrate).

Figure 1. Generation of HSF1 (303A/307A) knock-in mice.

(A) Schematic illustrating the targeting strategy for generation of HSF1 (303A/307A) (Hsf1^303A/307A) mutant mice (modified from published Figure (Mol Cell Biol. 2018 Aug 28;38(18). pii: MCB.00095-18) [22]. The genomic HSF1 locus, targeting vector, targeted locus (with mutant HSF1 (303A/307A) and the neomycin (neo) gene in the genome of ES cells and mouse), and the targeted locus with mutant HSF1 (303A/307A) (without neo gene) in the genome of mutant mice is presented. The neo gene is presented in blue and exons are in red. Locations of S303 and S307 sites in the genomic HSF1 locus, probes for Southern blot analysis, and genotyping primers (P1 to P6) for PCR are indicated. Notably, Neo is flanked by Lox P sites (L) and can be removed after crosses with the Cre recombinase transgenic mice. Note; distances are not in proportion. B, Bgl II; S, Sac I; X, Xho I; N, Not I; C, Cla I restriction enzyme sites. TK is thymidine kinase, pBS is pBluescript KS plasmid.

(B) Southern blot analysis to select correctly targeted ES cell clones. The genomic DNA was isolated from ES cell clones (obtained after double drug selection, see Materials and Methods), digested with Bgl II restriction enzyme, fractionated on agarose gel, and transferred to the nitrocellulose membrane. The fractionated genomic DNA was hybridized with the external probe to yield a 5.7kb (for WT) or 6.9kb (for mutant) DNA band. Lanes 1-5, 7 and 9 represent WT ES cell clones; Lanes 6 and 8 represent heterozygous ES cell clones.

(C) Southern blot analysis to verify WT and homozygous mutant mice. Southern blot analysis was perform as described in (B) using genomic DNA isolated from tails of WT and homozygous mutant mice.

(D) PCR-based genotyping and verification of WT (W/W), Hsf1^303A/307A/WT (W/M), and Hsf1^303A/307A (M/M) mice containing neo gene cassette in the HSF1 allele. The PCR analysis was performed with primer combination of P1, P2 and P3 (location of the primers are indicated in panel A) using tail genomic DNA as templates. A 630 bp fragment for mutant allele and 330 bp fragment for the WT allele were amplified by above primers.

(E) DNA sequencing analysis to confirm the 303A/307A mutation in the HSF1 gene. Sequence of DNA fragment including the S303 and S307 coding sequence was amplified by PCR using genomic DNA from WT and homozygous Hsf1^303A/307A mice. The coding substitutions replacing serine with alanine (AGC to GCC) are indicated in the mutant genomic locus.

(F) PCR-based genotyping of WT (W/W), heterozygous Hsf1^303A/307A/WT (W/M), and homozygous Hsf1^303A/307A (M/M) mice following crossing Hsf1^303A/307A homozygous mice with transgenic mice expressing Cre recombinase to remove the neo cassette. The PCR analysis was performed with primer combination of P5 and P6 (location of the primers are indicated in panel A) using tail genomic DNA as templates. A 286 bp fragment for the mutant allele and 246 bp fragment for the WT allele were amplified by these primers.

(G) Relative mRNA levels of HSF1 in WT and Hsf1^303A/307A MEFs determined by RT-PCR. Bars are means +/− SD (n = 5 per group).

(H) Immunoblot analysis using antibody to detect total HSF1 or the phosphorylated form of HSF1 at S303/S307 in MEF cell lysates prepared from WT and Hsf1^303A/307A (M/M). β-Actin represents loading control.

The proximal, distal, and neomycin-Lox P DNA fragments were released from relevant plasmids using indicated restriction enzymes. In the following step, these DNA fragments were introduced into the Xho I site of the phage DNA vector λDASHII-254-2TK (gift from Dr. N. R Manley, University of Georgia, Athens, GA) that was flanked by two thymidine kinase (TK) genes for drug negative selection (ganciclovir) of the ES cell clones. The phage DNA vectors were packaged using Gigapack II Packaging Extract (). The single phage clone whose genome contained all 3 DNA fragments in correct orientation were identified and verified by PCR and DNA sequencing. The final targeting vector, containing 2 TK genes and proximal, distal, and neomycin-Lox P fragments were released from the phage DNA by Not I restriction enzyme digestion.

2. Electroporation of targeting vector into ES cells and identification of positively targeted ES cell clones

The targeting vector (plasmid) was amplified and linearized at Not I site to allow transfection into ES cells. The ES cells were electroporated with the targeting vector. The ES cell clones doubly selected using G418 (200μg/ml) (positive selection) and ganciclovir (2μM) (negative selection) were subjected to Southern blotting analysis to identify correctly targeted ES cell clones.

To identify the clones containing homologous recombination of mutant HSF1, an external probe was designed. The external probe hybridizes to the DNA region that is close to, but not included in the targeting vector. According to the targeting strategy, we designed a 600 bp probe that predictably hybridizes to the 5′ region of the targeting vector as indicated in Figure 1A. The external probe for Southern blotting was prepared by PCR using specific primers, forward primer: 5’-GGGAAGAATGGGGACTAAAC-3’, reverse primer: 5’-TACATGTGGCACTCACATAT-3’. The DNA probe was labeled with ³²P-ATP (NEG002A100UC, Perkin Elmer) using commercially available kits (Decalabel DNA labeling kit). The genomic DNA isolated from ES cells were digested with restriction enzyme Bgl II, fractionated by agarose gel electrophoresis, transferred to nylon membrane, and hybridized with ³²P-ATP labeled external probe. The targeted (mutant) allele yielded a 6.9kb band, while WT allele shows a 5.7kb band in Southern blotting analysis (Figure 1B). Three ES cell clones (out of 129 clones obtained after double drug selection) containing homologous recombination (HR) of HSF1 mutant allele were identified by Southern blotting. The representative Southern blot result is presented in Figure 1B.

3. Microinjection of positive ES cell clones into blastocysts, generation of chimeras and selection of mice with germline transmission

Two correctly targeted ES cell clones were microinjected into mouse blastocysts to generate germline-transmitting chimeric mice. The chimeric mice were intercrossed with C57BL/6 mice. The heterozygous mice (WT/Hsf1^303A/307A) with germline transmission of mutant allele were identified by PCR and Southern blot analysis.

The heterozygous mice with germline transmission of the mutant allele were intercrossed to generate homozygous (Hsf1^303A/307A) and WT mice. The genotype of the mice was detected by Southern blotting as described above and by PCR. The Southern blotting result of DNA isolated from homozygous mutant and WT mice are presented in Figure 1C. When the genotype of HSF1 knock-in mutant mice was confirmed by Southern blot, thereon mice were routinely genotyped using PCR of tail DNA. To identify the mutant and WT HSF1 allele, the following primers were used in the PCR reactions: Primer P1 (5’- CCATGGGACTGCCAGTAAGT-3’) and P2 (5’- CTTGTCTAGGCAGGCTACGC-3’) were used to generate a 630 bp DNA fragment for WT allele, and primer P2 and P3 (5’- CTCGACATCGGAAGATCCAT-3’) were used to produce 330 bp DNA fragment for mutant allele (Figure 1A and 1D). To remove the neomycin (neo) gene from the genome of Hsf1^303A/307A mice, the homozygous mice were crossed with Cre transgenic female mice (Jackson Laboratory). After the neo gene cassette was removed, a 40 bp Lox P sequence remained in the intron 8 of the HSF1 gene and thereby in Hsf1^303A/307A cells, which was used in genotyping using primers P5 and P6, resulting the mutant band to being 40bp longer. The removal of the neo gene and genotype of mice were identified by PCR using primer P5 (5’-GCAGGACCTTTATCCCTCCT-3’) and P6 (5’-GTTGGGGACAAAGGGGTATC-3’) and DNA sequencing. The P5 and P6 primers produced a 246 bp band for WT allele and a 286 bp band for mutant (Figure 1F).

Note that all the work using mice was approved by the Institutional Animal Use and Care Committee of Augusta University.

4. Characterization of HSF1 expression in Hsf1^303A/307A knocked-in mice.

To investigate whether our knock-in strategy affects the transcription of HSF1, we performed RT-PCR to determine the HSF1 mRNA level using the following primers: Forward primer: 5’-TCTCACTGGTGCAGTCGAAC-3’ and Reverse primer: 5’-GTAGGCTGGAGATGGAGCTG-3’. The data indicate that the mRNA level of HSF1 in Hsf1^303A/307A cells is comparable to that of WT cells (Figure 1G). The DNA point mutation of S303/S307 residues in HSF1 were verified by DNA sequencing using DNA isolated from homozygous Hsf1^303A/307A mouse tissue (Figure 1E). The elimination of phosphorylation at residues S303/S307 in HSF1 were further confirmed by immunoblot analyses using S303/S307 phospho-specific antibody (gift from Dr. L. Sistonen, Abo Akademi University, Finland) (Figure 1H).

5. Generation and transformation of mouse embryo fibroblasts (MEF)

Primary MEFs were prepared from embryonic day 13.5 (E13.5) following timed pregnancies as described previously [24, 25]. MEFs were immortalized following transfection with plasmids encoding SV40 large T antigen or transformed by stably expressing oncogenes (E1A and RAS) using retroviral vector system. Transformed MEFs were selected using puromycin (2 μg/ml for E1A) and blasticidin (3 μg/ml, for RAS) and cultured in Dulbecco's Minimal Essential Medium (DMEM) supplemented with 10% heat inactivated fetal calf serum (FCS).

6. Drug treatment and colony formation assay

Cellular survival using colony formation assays was determined as previously described [24, 25]. Briefly, cells were treated with chemotherapeutic agents (doxorubicin at 0.5 μg/ml or etoposide 5 μg/ml concentrations for 16 h at 37°C). An appropriate number of cells were plated and incubated for 10 days at 37°C and 5% CO2. Colonies were stained with crystal violet and those containing more than 50 cells were counted. Plating efficiency (PE) of untreated cells was also determined. The surviving fraction was determined as number of colonies for treated cells divided by the number of cells plated and divided by PE for each group.

7. Hematoxylin and Eosin (H&E) staining

Liver and tumor tissues were fixed in 10% formalin, embedded in paraffin and 7 μm tissue sections were prepared as described previously [26, 27]. For the H&E staining, the tissue sections were deparaffinized in xylene and rehydrated in a series of alcohol/water mixtures before staining with H&E sequentially.

8. Western blotting

Cell lysates were prepared from cultured cells using RIPA buffer (150 mM NaCl, 1% NP-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris, pH 7.5) containing protease and phosphatase inhibitor cocktail (Thermo Scientific, IL). The cellular protein lysates were fractionated on SDS-PAGE gels, transferred onto a polyvinylidene fluoride membrane (PVDF), and probed with the indicated primary followed by secondary antibody [26, 27].

9. Quantitative RT-PCR

Quantitative RT-PCR was performed as described previously [26, 27]. Briefly, total RNA was isolated from tissues or cultured cells using Trizol and reverse transcribed using the Iscript cDNA synthesis kit. Quantitative RT-PCR was performed using iQ SYBR green Supermix on an CFX Opus 96 Real-time PCR System (Bio-Rad Laboratories).

10. Diethylnitroseamine (DEN)-induced hepatocellular carcinoma (HCC)

DEN-induced HCC was generated using Hsf1^303A/307A and WT mice as described previously [26, 27]. A single dose of DEN (25 mg/kg body weight) was injected into 14-day-old male WT and Hsf1^303A/307A mice intraperitoneally (i.p.) to initiate tumor formation. DEN-treated mice were observed for evidence of sickness twice a week. At 10 months post-DEN injection, mice were euthanized and livers were removed and analyzed for the presence of HCCs using a dissecting microscope. The number of all tumors and the number of tumors larger than 0.2 cm in diameter were quantified. The maximum size of the tumors in each liver was measured using caliper. Tumor volume was calculated using the following formula: V (volume) = W² x L/2, where W represents width of the tumor, L represents length of the tumor.

v. Physiological effects of HSF1 following S303/S307 mutations to alanine residues

v.1. Loss of HSF1 phosphorylation at S303/S307 promotes cell proliferation, drug resistance and tumorigenesis

A number of studies have shown that HSF1 is required for malignant transformation and is essential for tumor cell survival. It has been demonstrated that deletion or inhibition of HSF1 leads to cell death in various tumor types [27-30]. In vivo, we have shown that loss of HSF1 blocks DEN-induced HCC in an animal model [27]. As mentioned above, loss of HSF1 S303/S307 phosphorylation sites increases its transcriptional activity and the expression levels of its downstream target genes, such as HSP90, HSP70 and HSP25 [22]. To examine whether increasing HSF1 activity facilitates process of tumorigenesis, we first determined cellular proliferation of transformed Hsf1^303A/307A MEFs. WT and Hsf1^303A/307A primary MEFs were transformed by overexpression of oncogene E1A and RAS. The data presented in Figure 2A show significantly higher growth rate of transformed Hsf1^303A/307A MEFs compared to WT MEFs. The doubling time of transformed Hsf1^303A/307A MEFs was reduced by 35% compared to that of WT MEFs (1.42 days in WT versus 0.94 day in Hsf1^303A/307A). Consistently, colony formation assay of Hsf1^303A/307A MEFs show significantly higher number of colonies observed than for WT MEFs (Figure 2B). Lack of HSF1 phosphorylation at S303/S307 significantly increased cell survival when cells were exposed to chemotherapeutic reagents (Figure 2C-D). Taken together, above results indicate that activation of HSF1 by mutations of S303/S307 phosphorylation sites promotes cell proliferation and their response to drug treatment.

Figure 2. Lack of phosphorylation of HSF1 at 303A/307A increases cell proliferation and promotes survival of cells exposed to chemotherapeutic drugs.

The Hsf1^303A/307A and WT primary MEFs were transformed by transfection of plasmids encoding E1A and RAS as described in Materials and Methods.

(A) Growth rate of E1A and RAS transformed Hsf1^303A/307A and WT MEFs. The proliferation rate of Hsf1^303A/307A MEFs was significantly higher than that observed in WT MEFs.

(B) Plating efficiency (PE) using colony formation assay. Untreated WT and Hsf1^303A/307A E1A and RAS transformed MEFs were plated and cultured for 10 days. Number of colonies were counted and the PE was calculated as described in Materials and Methods. Left panel shows representative image of the plates containing colonies. Right panel is the quantification of the PE.

(C-D) The E1A and RAS transformed MEFs were treated with etopotoside (C) or doxorubicin (D) at the indicated concentrations for 16 hours. Determination of surviving fraction using colony formation assay was performed as described in Materials and Methods. Left panels are representative images of plates containing colonies. Right panels are quantification of the surviving fractions.

In all relevant panels, statistical significance is indicated (*p < 0.05, **p< 0.01, ***p< 0.001). Note that all experiments repeated at least two times with multiple samples.

Our previous study show that loss of HSF1 S303/S307 phosphorylation in mice leads to age-associated obesity and obesity-related chronic inflammation [22]. It is well established that this metabolic state (obesity) and chronic inflammation are risk factors for cancer initiation and development [31, 32]. To investigate whether Hsf1^303A/307A mice respond differently to tumor induction, we generated DEN-induced HCC model using Hsf1^303A/307A and WT mice (Figure 3A-F). The total number of tumors at 10-month post-DEN treatment in Hsf1^303A/307A mice slightly increased compared to that of WT mice (Figure 3C). However, Hsf1^303A/307A mice show significantly larger tumors (> 0.2 cm) compared to WT mice. In addition, the maximum tumor size and the tumor area covered in Hsf1^303A/307A mice were also greater than that observed in WT mice (Figure 3E-F). Taken together, above experiments indicate that endogenous activation of HSF1 through lack of S303/S307 phosphorylation promotes tumorigenesis.

Figure 3. Mice harboring HSF1 303A/307A exhibit accelerated tumorigenesis in vivo.

(A) 2 sets of representative macroscopic images of DEN-induced HCC in WT and Hsf1^303A/307A mice. HCC were induced by DEN as described in the Materials and Methods section. Arrows indicate tumor nodules.

(B) Histological analysis of HCC (H&E staining).

(C-F) Quantification of DEN-induced liver tumors observed in WT and Hsf1^303A/307A mice. Total tumor number (C), tumor number larger than 0.2 cm in size (D), tumor volume (E), and tumor area (F) in livers were quantified. Scale bars are mean ± SD (n = 10 mice per group). For all relevant panels, statistical significance is indicated (*p < 0.05, **p < 0.01, ***p < 0.001). ns-not significant.

Conclusions.

The above data indicate that mutations of only two of HSF1 serine phosphorylation residues to alanine leads to significant physiological consequences to the organism in vivo and to cells in vitro. Since most tumors exhibit higher levels of HSF1 activity, the above mouse line is a good model to study cellular proliferation, mechanisms of drug resistance and cancer.