Estradiol and Tamoxifen Regulate NRF-1 and Mitochondrial Function in Mouse Mammary Gland and Uterus

M M IVANOVA and others E and 4-OHT regulate NRF-1 51:2 233–246 Research 2

Estradiol and tamoxifen regulate NRF-1 and mitochondrial function in mouse mammary gland and uterus

Margarita M Ivanova, Brandie N Radde, Jieun Son1, Fabiola F Mehta1, Sang-Hyuk Chung1 and Carolyn M Klinge Correspondence Department of Biochemistry and Molecular Biology, Center for Genetics and Molecular Medicine, should be addressed University of Louisville School of Medicine, Louisville, Kentucky 40292, USA to C M Klinge 1Department of Biology and Biochemistry, Center for Nuclear Receptors and Cell Signaling, University of Houston, Email 3605 Cullen Blvd, Houston, Texas 77204, USA [email protected]

Abstract

Nuclear respiratory factor-1 (NRF-1) stimulates the transcription of nuclear-encoded genes Key Words that regulate mitochondrial (mt) genome transcription and biogenesis. We reported that " nuclear respiratory factor-1

estradiol (E2) and 4-hydroxytamoxifen (4-OHT) stimulate NRF-1 transcription in an estrogen " estrogen receptor receptor a (ERa)- and ERb-dependent manner in human breast cancer cells. The aim of " mitochondria

this study was to determine whether E2 and 4-OHT increase NRF-1 in vivo. Here, we report " mouse

that E2 and 4-OHT increase NRF-1 expression in mammary gland (MG) and uterus of

ovariectomized C57BL/6 mice in a time-dependent manner. E2 increased NRF-1 protein in the uterus and MG; however, in MG, 4-OHT increased Nrf1 mRNA but not protein. Chromatin immunoprecipitation assays revealed increased in vivo recruitment of ERa to the Nrf1

promoter and intron 3 in MG and uterus 6 h after E2 and 4-OHT treatment, commensurate Journal of Molecular Endocrinology

with increased NRF-1 expression. E2- and 4-OHT-induced increases in NRF-1 and its target genes Tfam, Tfb1m, and Tfb2m were coordinated in MG but not in uterus due to uterine-

selective inhibition of the expression of the NRF-1 coactivators Ppargc1a and Ppargc1b by E2

and 4-OHT. E2 transiently increased NRF-1 and PGC-1a nuclear staining while reducing

PGC-1a in uterus. E2, not 4-OHT, activates mt biogenesis in MG and uterus in a time-

dependent manner. E2 increased mt outer membrane Tomm40 protein levels in MG and uterus whereas 4-OHT increased Tomm40 only in uterus. These data support the hypothesis Journal of Molecular of tissue-selective regulation of NRF-1 and its downstream targets by E2 and 4-OHT in vivo. Endocrinology (2013) 51, 233–246

Introduction

Uterus and mammary gland (MG) are classical estrogen formation and sexual maturation of the reproductive tract

target tissues in which estrogens, e.g. 17b-estradiol (E2), whereas ERb is essential for ovarian function, but not for bind estrogen receptors a (ERa) and ERb to regulate gene MG or uterus (Couse & Korach 1999). Although many expression and physiological functions (Faulds et al. tissue-speciﬁc phenotypes for aERKO, bERKO, and other 2012). Studies in female ERa knockout (aERKO) and aERKO mice have been reported (Pendaries et al. 2002, bERKO mice demonstrated that ERa is critical for MG Korach et al. 2003, Billon-Gale´s et al. 2009), how estrogens

http://jme.endocrinology-journals.org Ñ 2013 Society for Endocrinology Published by Bioscientiﬁca Ltd. DOI: 10.1530/JME-13-0051 Printed in Great Britain Downloaded from Bioscientifica.com at 10/01/2021 01:32:11AM via free access Research M M IVANOVA and others E2 and 4-OHT regulate NRF-1 51:2 234

regulate mitochondrial (mt) biogenesis and energy pro- how estrogens or antiestrogens regulate the expression or duction by oxidative phosphorylation (OXPHOS) in activity of these coactivators in uterus or MG. One study

uterus and MG is largely undefined. reported that E2 reduced PGC-1a specifically in the uteri of Weight gain by postmenopausal women and ovari- immature and adult mice (Macari et al. 2010). Concor- ectomized rodents and the preservation of muscle mass dantly, ovariectomy increased PGC-1a in mouse brain and inhibition of adipogenesis by hormone replacement endothelial cells (Kemper et al. 2013). therapy (Cooke & Naaz 2004) provide clear evidence for Most studies of the regulation of mt biogenesis by the stimulatory effects of endogenous estrogens on NRF-1 and the PGC-1 family have been performed in metabolism but does not address the effects of endogenous tissues with high oxidative activity: skeletal muscle, liver, estrogens on specific tissues. Estrogens regulate mt brown fat, and heart (Finck & Kelly 2006, 2007, Scarpulla biogenesis and function in normal and cancer cells 2008b). There are no specific studies of how estrogens through nuclear-, mt-, plasma membrane-initiated events regulate mt biogenesis in MG and only one study in mouse that have been reviewed (Yager & Chen 2007, Klinge 2008, uterus (Macari et al. 2010). Although microarray analysis Simpkins et al. 2008, 2010). The specific roles for ERa and of gene expression in mouse uterus identified many genes

ERb in regulating estrogens’ tissue-speciﬁc effects on regulated by E2 and TAM (Moggs et al. 2004, Fong et al. energy, glucose homeostasis, and insulin sensitivity were 2007, 2010), Nrf1 was not included in the microarray recently reviewed (Faulds et al. 2012). In addition to platforms. GEO proﬁle #GDS2208/9902 revealed that Nrf1

central brain actions to regulate appetite and satiety was increased in C57BL/6 mouse uterus by 6 h of E2 (Barros Rodrigo & Gustafsson 2011), ERa maintains treatment (Hewitt et al. 2012). Another study of uterine K K metabolic control in skeletal muscle, white adipose tissue, gene expression revealed lower Nrf1 in Src2 / compared liver, and pancreas by stimulating insulin signaling, while with WT adult mice (Jeong et al. 2007). NCOA2 (SRC-2) is activated ERb results in a diabetogenic/adipogenic pheno- a well-established coactivator for ERa and upregulated by

type, although questions remain (Faulds et al. 2012). The E2 in adult uteri (Klinge 2000, Jeong et al. 2007, O’Malley &

mechanisms and gene targets by which ERa and ERb Kumar 2009). Microarray proﬁling of E2-regulated genes regulate metabolic pathways are largely unknown, cer- in aortas of WT, ERa, and ERb knockout mice revealed tainly tissue-speciﬁc (Barros Rodrigo & Gustafsson 2011), opposite regulation of Nrf1 and many NRF-1-regulated and involve complex cross talk with multiple signaling mitochondrial respiratory chain (MRC) genes by ERa and

pathways including other nuclear receptors, e.g. peroxi- ERb (O’Lone et al. 2007). E2–ERa increased whereas E2–ERb some proliferator-activated receptor a (PPARa) and PPARg reduced expression of Nrf1 and MRC genes in mouse aorta

Journal of Molecular Endocrinology (Papi et al. 2013) and estrogen-related receptors (ERRs; (O’Lone et al. 2007). Nrf1 was not included in a microarray

Deblois & Giguere 2013); coregulators, e.g. PPARg coacti- proﬁling E2-regulated genes in mouse MG (Deroo et al.

vator (PGC-1) family proteins (Scarpulla et al. 2012); and 2009). Hence, how E2 and TAM regulate NRF-1, PGC-1 plasma membrane receptors (Renoir et al. 2013). coactivators, NRF-1 regulated genes, and mt biogenesis in

We reported that E2 and 4-hydroxytamoxifen uterus and MG in vivo is unknown. Similarly, regulation of (4-OHT), an active tamoxifen (TAM) metabolite, stimulate mt biogenesis, by antiestrogens, e.g. TAM, in uterus where transcription of nuclear-encoded nuclear respiratory TAM is an agonist (Sourla et al. 1997) is largely undeﬁned factor 1 (NRF-1) by binding ERa and ERb respectively (Klinge 2008). and increasing ERa,ERb, and RNA polymerase II recruit- Ovariectomy was recently shown to decrease the ment to the promoter of NRF-1 in MCF-7 breast cancer expression of NRF-1, TFAM, and PGC-1b while increasing cells (Mattingly et al. 2008, Ivanova et al. 2011). NRF-1 PGC-1a in isolated brain endothelial cells from C57BL/6 regulates the transcription of nuclear-encoded, mt genes, mice (Kemper et al.2013). Ovariectomy reduced the mt e.g. mt transcription factors Tfam, Tfb1m, and Tfb2m, and biogenesis w20% and decreased the expression of MnSOD nuclear-encoded components of OXPHOS, e.g. cyto- and ATP synthase subunit a, complex 5, suggesting that chrome c (Cycs; Cam et al. 2004, Scarpulla 2006, 2008a, ovarian hormones normally stimulate mt biogenesis but

Richard 2011). In MCF-7 cells, E2-induced NRF-1 resulted suppress reactive oxygen species (ROS) in brain endothelium in an increase in expression of TFAM and increased mt by inducing antioxidant proteins (Kemper et al.2013). This biogenesis (Mattingly et al. 2008). NRF-1’s activity agrees with many reports on the protective effect of depends on its interaction with coactivators PGC-1a, estrogens against brain injury (Simpkins et al.2010). PGC-1b,andPRC(Scarpulla 2008a, 2012, Gleyzer & The goal of this study was to test the hypothesis that

Scarpulla 2011, Richard 2011). Little is known about E2 and 4-OHT stimulate Nrf1 gene expression and

coordinately activate downstream nuclear and mt gene tissues harvested 6, 24, and 72 h after injection. There is no expression and mt biogenesis in vivo in a tissue-selective variation in Nrf1 transcript expression in mouse prefrontal manner. To address this hypothesis, ovariectomized cortex during a 24-h (12 h light:12 h darkness) cycle

(ovex) mice were administered a single dose of E2 (GDS3080), similar to the conditions used in our experi- (100 ng/mouse) or 4-OHT (50 mg/mouse). NRF-1 and its ment. There is no evidence that Nrf1 gene expression varies target genes, both nuclear and mt, mt biogenesis, and with circadian rhythm (Bozek et al.2009, 2010). mt:nuclear DNA ratio were examined 6, 24, and 72 h post injection based on a previous study (O’Brien et al. Mice and treatments for IHC studies 2006). Our results support our hypothesis and reveal novel

tissue-selective actions of E2 and 4-OHT in regulating Female mice were obtained by mating near congenic NRF-1, its coregulators, and its downstream gene targets in C57BL/6 animals (129/C57BL/6 mice were backcrossed to mouse MG and uterus. C57BL/6 for eight generations; i.e. 99.61% of the genome was contributed by C57BL/6). Mice were ovariectomized at 4–6 weeks of age and rested for 2 weeks. They were then

Materials and methods i.p. injected with ethanol vehicle or 1 mgE2 (Sigma– Aldrich) for 24 or 48 h. All procedures were carried out Chemicals according to animal protocols approved by the University

E2,4-OHT,andsesameoilwerepurchasedfrom of Houston Institutional Animal Care and Use Committee. Sigma–Aldrich.

Tissue processing and IHC Antibodies Uteri were harvested, fixed in 4% paraformaldehyde, and Antibodies were purchased as follows: ERa (HC-20) and embedded in paraffin. Serial sections were made long- Tomm40 (H-300), Santa Cruz Biotechnology; NRF-1, itudinally at 5 mm thickness. Sections were deparaffinized, Rockland Immunochemicals, Inc. (Gilbertsville, PA, rehydrated, and microwaved in 10 mM sodium citrate USA); TFAM (DO1P), Abnova (Taipei, Taiwan); nuclear- buffer (pH 6.0) for 20 min as described previously (Balsitis encoded cytochrome c oxidase subunit IV (COX4-1, et al. 2003). Non-specific antibody binding was blocked nuclear-encoded), MitoSciences (Eugene, OR, USA); by incubating sections in 5% donkey serum (Santa Cruz cytochrome c Ab-2, Thermo Fisher Scientific (Fremont, Biotechnology) in PBS for 1 h. Sections were serially

Journal of Molecular Endocrinology CA, USA), and b-actin (Sigma–Aldrich). incubated with rabbit anti-NRF-1 antibody (1:200 in 5% donkey serum, Sigma–Aldrich) and then goat anti- PGC-1a (1:50 in 5% donkey serum, Novus Biologicals, Animals Littleton, CO, USA) at 4 8C overnight; normal rabbit IgG C57BL/6 mice, ovariectomized on postnatal day 20, were and normal goat IgG (Santa Cruz Biotechnology) were purchased from Charles River Laboratories (Wilmington, used as negative controls. Anti-rabbit IgG-Alexa Fluor 488 MA, USA) on postnatal day 25. All animal studies were (Invitrogen) and anti-goat IgG-CFL 594 (Santa Cruz conducted in accordance with the procedures outlined in Biotechnology) were used as secondary antibodies. Nuclei the NIH Guide to the Care and Use of Experimental were stained with Hoechst 33258 (Sigma–Aldrich). Tissues Animals as approved by the AALAC-accredited University were visualized with an Olympus FV1000 laser scanning of Louisville Institutional Animal Care and Use Commit- confocal microscope. tee. After 14-day quarantine, mice were randomized into three groups of ﬁve per group. Mice were given a single RNA isolation and quantitative real-time PCR s.c. injection (between the shoulders) of vehicle control

(sesame oil, 100 ml), E2 (100 ng/mouse), or 4-OHT RNA was extracted using the RNeasy Mini Kit from Qiagen. (50 mg/mouse) and tissues were harvested at 6, 24, and The High Capacity cDNA archive kit (PE Applied Biosys- 72 h after injection. A total of 15 mice were included tems) was used to reverse transcribe RNA using random in each treatment group and time point within each hexamers. Supplementary Table 1, see section on supple- experiment and three separate experiments were per- mentary data given at the end of this article lists gene and formed. All mice were injected with vehicle control at protein names and abbreviations. Murine mRNA transcript

0830 h followed by E2 at 0900 h, and 4-OHT at 1000 h with levels of Nrf1, Tfam, CycS, and Cox4 were analyzed using

K SYBR Green dye (RT2SYBR Green ROX qPCR, Qiagen). formula 2 DDCt and data are presented as relative to Tfb1m and Tfb2m expression levels were analyzed by one- expression in control (sesame oil-injected) mice. step quantitative real-time PCR (qPCR; Power SYBR Green RNA-to-Ct 1-step kit, Applied Biosystems). Relative tran- Protein isolation scription levels were normalized to 18S rRNA or 36b4 mRNA (Macari et al. 2010). 18S rRNA primers and probes Whole tissue extracts were prepared in radioimmuno- were purchased from Assays-on-Demand Gene Expression precipitation buffer and protease inhibitor (Roche; Products (PE Applied Biosystems). Murine Tfb1m and Mattingly et al. 2008) with a mechanical homogenizer Tfb2m primers were from RealTimePrimers (Elkins Park, (Omni International, Kennesaw, GA, USA). Protein PA, USA). The mouse Ppargc1a, Tfam, cytochrome c, and concentrations were determined using the Bio-Rad DC Cox4 primer sequences were reported in Safdar et al. (2011) Protein Assay (Bio-Rad). and were purchased from IDT (Coralville, IA, USA). The mouse Nrf1 primers were also purchased from IDT: Western blot forward, 50-GCACCTTTGGAGAATGTGGT-30;reverse, 50-CTGAGCCTGGGTCATTTTGT-30. Western blotting followed standard procedures (Mat- Transcript expression analysis was determined tingly et al. 2008, Ivanova et al. 2011). In brief, 30–40 mg

between the control and E2- or 4-OHT treated groups protein lysates (see individual ﬁgures) were separated and individual samples were run in triplicate. qPCR was on 10 or 14% SDS–PAGE gels and electroblotted to PVDF performed in the ABI PRISM 7900 SDS 2.1 or ViiA7 real- membranes. Data were captured and analyzed by Care- time PCR systems (PE Applied Biosystems) using relative stream Image Station 4000 R Pro with Carestream quantiﬁcation. Analyses and fold differences were Molecular Imaging Software, version 5.0 (Carestream determined using the comparative Ctmethod.Fold Health, Inc., New Haven, CT, USA). The values from change was calculated from the DDCt values with the regions of interest normalized to the loading control, e.g.

AB 4 MG 4 Uterus * * 3 3 * * * 2 2 * * mRNA expression mRNA expression Nrf1 Nrf1

Journal of Molecular Endocrinology 1 1 *

0 Relative 0 2 2 2 2 2 2 E E E E E E 4-OHT 4-OHT 4-OHT 4-OHT 4-OHT 4-OHT Control Control Control Control Control Control 6 h 24 h 72 h 6 h 24 h 72 h

CDMG Uterus 2.5 5

* * 2.0 4

1.5 3 * * * 1.0 2

0.5 1 Relative NRF-1 protein level Relative Relative NRF-1 protein level 0 0 2 2 2 2 2 2 E E E E E E 4-OHT 4-OHT 4-OHT 4-OHT 4-OHT 4-OHT Control Control Control Control Control Control 6 h 24 h 72 h 6 h 24 h 72 h

Figure 1

E2 and 4-OHT increase Nrf1 mRNA and NRF-1 protein expression in mouse (D) after the indicated time of treatment. Values are the meanGS.E.M. mammary gland (MG) and uterus. Ovariectomized C57BL/6 mice were of 5–10 mice/treatment group in which control was set to 1 for each

given a single s.c. injection of sesame oil (vehicle control), 100 ng E2,or treatment time within each blot for comparison. *P!0.05 vs vehicle 50 mg 4-OHTand were killed 6, 24, or 72 h post injection. (A and B) Q-RT-PCR control (Student’s t-test). Representative western blots are shown in analysis of Nrf1 mRNA level in MG (A) and uterus (B). (C and D) NRF-1 Supplementary Figure 1, see section on supplementary data given at the protein expression was examined relative to b-actin in MG (C) and uterus end of this article.

b-actin or Ponceau S staining, and the normalized value mouse Nrf1 promoter (K962 bp) region with estrogen of the control was set to 1 for comparison between response element (ERE) were sense, 50-GATCTCTGG- separate experiments. GTTGGAGGTC-30 A; antisense, 50-ATAAATGCCCCA- CATGGTGT-30 (see diagram in Fig. 2A) and ‘close to ERE’ primers were sense, 50-CTCTTAACCGCTGAGCCATC-30; Chromatin immunoprecipitation assay antisense, 50-AGGGAAAGGGAGGAGATTCA-30. The pri- Chromatin immunoprecipitation (ChIP) was performed mers containing the Nrf1 intron 3 region with AP-1 and 1 0 with the MAGnify Chromatin Immunoprecipitation ⁄2 ERE binding sites were sense, 5 -TTTGACTTGATA- System (Invitrogen) using the manufacturer’s protocol. CATGTGAATGC-30; antisense, 50-TCTCTGTGAAATAACT- Half of the uterus or the left inguinal MG from one mouse GACAGAAAT-30. Real-time PCR was performed was fixed in 1% formaldehyde at room temperature using 2 ml purified DNA and SYBR Green Master Mix 20 min. Then, the reaction was terminated by addition (SuperArray Bioscience Corp., Frederick, MD, USA). of 0.125 M glycine for 5 min at room temperature Relative promoter enrichment was compared with IgG followed by sedimentation and a wash with ice-cold PBS. for each treatment. ChIP using primers to an intron After aspiration of the PBS, the tissues were frozen at 6regionofNrf1,50-CAGGCAGAGGTGGAGCTAAC-30 K80 8C. The tissues were mechanical homogenized in and 50-CACTGGGCCCTAGCAAATAC-30, were used as a ChIP buffer (Ivanova et al. 2011) and genomic DNA was negative control. The Ct values for ChIP with IgG to any of shared by sonication on ice. Lysed extracts (300 mg) were the primers or ChIP for ERa with primers for intron 6 were incubated with anti-ERa (HC-20) antibody or normal ‘undetectable’ O35G0.5. rabbit IgG (both from Santa Cruz). DNA was eluted from the beads by incubating with 100 ml 10% Chelex (Bio-Rad) mt:nuclear DNA ratios at 95 8C for 10 min, followed by Proteinase K treatment and heat inactivation at 95 8C for 10 min. Following The relative mt content of MG and uterus was determined centrifugation, the supernatant was used for PCR using SYBR green qPCR measuring the ratio of mt encoded amplification. The primers used for qPCR containing the nicotinamide adenine dinucleotide dehydrogenase-5

1/2 ERE and AP1, primer 1 A Mouse Nrf1 promoter C Close to ERE, primer 2 8 14 ERE, primer 3 Intron 3 7 * 12 * Intron 1 Intron 2 12.345 bp 27.964 bp 12.345 bp 6

Journal of Molecular Endocrinology Uterus ChiP primers 2 and 3 ChlP primer 1 10 MG * 5′ UTR mRNA start Protein start 5 ERE ERRBS AP1 ½ ERE 8 * * * E1 E2 E3 E4 4 0 * 310 bp 6 * ½ ERE * 3 –141 GGTCA Fold enrichment 4 Fold enrichment ERE 2 –962 AGGTCAgctTGGTCT AP1 2 –451 TGACT 1 0 0 Control E 4-OHT Control E 4-OHT MG Uterus 2 2 BDEERα ERα α E –ERα E2–ER mRNA NRF1 2 mRNA NRF1 E E E1 E4 1 4 2 2 ERE ½ERE ERE ½ERE Input Control 4-OHT lgG Input Control 4-OHT lgG E E AP1 AP1 1/2 ERE and AP1, primer 1 MG Uterus 4-OHT–ERα 4-OHT–ERα 4-OHT–ERα mRNA NRF1 Close to ERE, primer 2 E E E1 E4 1 4 ERE ½ERE ERE, primer 3 ½ERE ERE AP1 AP1

Figure 2

E2 and 4-OHT increase ERa recruitment to the Nrf1 promoter in vivo in a shown in (A)) for the mouse Nrf1 gene promoter. (C) Immunoprecipitated tissue-dependent manner. (A) Diagram of the mouse Nrf1 gene promoter DNA was normalized to input and IgG control and qPCR data are 1 showing locations of ERE, AP-1, ⁄2 ERE, and ERR-binding sites (ERRBS), expressed relative to IP for control-treated mammary gland (MG, left) and transcription and protein coding start sites, as indicated. Exons (E1, E2, E3, uterus (right). Values are the meanGS.E.M. of ﬁve mice/treatment group. and E4) are indicated as gray boxes. The location of the three primers *P!0.05 from control for each primer. (D and E) ChIP results are

used for ChIP for ERa on Nrf1 are indicated as light gray boxes above the diagrammed for E2–ERa and 4-OHT–ERa recruitment to the Nrf1 promoter

Nrf1 gene. (B) Ovex C57BL/6 mice were treated with vehicle (control), E2, in MG (D) and uterus (E). The bent arrows indicate transcriptional or 4-OHT for 6 h. ChIP with ERa antibody or with rabbit pre-immune upregulation of Nrf1 mRNA as measured by qPCR (see Fig. 1A). serum (IgG) was performed with the three pairs of ChIP primers (locations

(mt-Nd5):nuclear-encoded cystic ﬁbrosis (Cftr) (Bauerly and Supplementary Figure 1). E2 increased NRF-1 protein

et al. 2006). For nuclear DNA and mtDNA quantiﬁcation, 10 in MG only 72 h after E2 treatment while 4-OHT was and 0.1 ng DNA were used as template respectively (Bauerly without effect (Fig. 1C and Supplementary Figure 1). et al.2006). Each sample was analyzed in triplicate. Relative mt copy number to nuclear copy number was ERa binding to mouse Nrf1 promoter in vivo assessed by a comparative Ct method, using the following equation: DCtmitochondria/nuclearZCtmitochondriaKCtnuclear. ChIP assays were used to investigate ERa binding to the The fold-change relative to control animals was calculated mouse Nrf1 promoter. The location of the primers for ChIP using the following equation: 2(KDDCt)mitochondria/nuclear, are shown in Fig. 2A. These regions were chosen based where DDCtmitochondria/nuclearZDCtcontrol mitochondria/nuclear on known or predicted EREs and other elements regulated KDCtmitochondria/nuclear of each animal from different treatment and time groups (Bauerly et al. 2006). Values G A represent mean fold change S.E.M. 4 Tfam MG * Tfb1m * Tfb2m 3 Statistical analysis * * 2 * * Statistical evaluation of the data was performed using

Student’s t-test in Excel and one-way ANOVA followed 1 post-hoc * by Student–Newman–Keuls or Dunn’s tests mRNA expression Relative using GraphPad Prism (San Diego, CA, USA). 0 Control E Control E 2 4-OHT 2 4-OHT Control E2 4-OHT 6 h 24 h 72 h B 3 Uterus Results Tfam Tfb1m E2 and 4-OHT increase Nrf1 expression in both MG Tfb2m * 2 * and uterus

We previously reported that E2 and 4-OHT increase NRF-1 1 * * transcription in MCF-7 and T47D breast cancer cells and * * * * human umbilical vein endothelial cells (Mattingly et al. mRNA expression Relative 0

Journal of Molecular Endocrinology Control E Control E 2008, Ivanova et al. 2011, Mattingly & Klinge 2012), but 2 4-OHT 2 4-OHT Control E2 4-OHT 6 h 24 h 72 h the effect of E2 and 4-OHT on NRF-1 transcription in

estrogen target tissues in vivo is unknown. The expression CD2.0 1.5 MG * Uterus of Nrf1 was measured in the MG and uterus of ovari- * ectomized C57BL/6 mice after treatment with vehicle 1.5 1.0 (sesame oil), E2, or 4-OHT for 6, 24, or 72 h (Fig. 1A and B). 1.0 Nrf1 transcript expression was increased by E2 in MG after 6 and 72 h and in the uterus after 6 and 24 h but not at 0.5 * 72 h. 4-OHT increased Nrf1 expression in both MG and 0.5 Relative TFAM protein level TFAM Relative uterus 6 h after treatment but reduced Nrf1 in the MG, but protein level TFAM Relative not in the uterus, after 24 h. Nrf1 expression returned to 0 0 Control E2 4-OHT Control E2 4-OHT basal in both MG and uterus 72 h after 4-OHT treatment. Basal Nrf1 transcript expressed was unchanged during the Figure 3 time course of treatment (Supplementary Figure 1). E2 and 4-OHT selectively regulate the expression of NRF-1 target genes for mtDNA transcription and replication. (A and B) Q-RT-PCR analysis of the mRNA expression of Nrf1-regulated Tfam, Tfb1m, and Tfb2m expression in Tissue-speciﬁc regulation of NRF-1 protein expression mammary gland (MG; A) and uterus (B) from ovex mice treated with vehicle

control, E2, or 4-OHT for the indicated time. (C and D) TFAM protein by E2 and 4-OHT expression was examined in MG (C) and uterus (D) 72 h after E2 or 4-OHT treatment. The data shown are the meanGS.E.M. of 5–10 mice/treatment Commensurate with E2- and 4-OHT-induced Nrf1 mRNA, group. *P!0.05 vs vehicle control (Student’s t-test). Representative NRF-1 protein increased in uterus 6 and 24 h after western blots are shown in Supplementary Figure 3, see section on treatment with a return to basal levels after 72 h (Fig. 1D supplementary data given at the end of this article.

by ER-tethering mechanisms (Supplementary Figure 2, see treatment while no changes were detected in MG section on supplementary data given at the end of this (Supplementary Figure 4).

article). E2 increased ERa recruitment to the ERE- containing region within the 50 promoter in MG and E2 activated and 4-OHT inhibited mt biogenesis uterus 6 h after treatment (Fig. 2B and C). However, 1 E2–ERa was not recruited to the ⁄2 ERE/AP-1 site located in TFAM is crucial for the initiation of mt transcription and

intron 3 in either MG or uterus (Fig. 2B and C). 4-OHT DNA replication (Campbell et al. 2012). As E2 and 4-OHT increased ERa recruitment to the ERE in both tissues and increased TFAM protein in the MG, we examined the 1 to the ⁄2 ERE/AP-1 region (ChIP primer 1) in the MG. ERa mt/nuclear DNA as an index of mt biogenesis. As recently recruitment corresponds to increased Nrf1 expression reviewed (Kemper et al. 2013), mt:nuclear DNA ratios are a

detected 6 h after treatment with E2 in both the MG and better marker of mt biogenesis than mt staining tech- the uterus (Fig. 1A and B). 4-OHT–ERa was recruited to the niques. The mt:nuclear DNA ratio was determined by 1 ⁄2 ERE/AP-1 binding site in the intron 3 region in the MG, qPCR using primers for mt-Nd5 as a mitochondria gene but not in the uterus. and Cftr (cystic ﬁbrosis) as a nuclear gene (Bauerly et al.

2006). Although E2 increased the mt:nuclear DNA ratio in MG at 6 h, this was not statistically signiﬁcant until 24 h E2 and 4-OHT increase selected NRF-1 target genes in MG (Fig. 4A). 4-OHT decreased mt biogenesis in the MG after

An increase in NRF-1 protein is expected to increase the 72 h. In the uterus, E2 increased mt biogenesis 24 and 72 h

transcription of its target genes, e.g. Tfam, Tfb1m, and after E2 treatment whereas 4-OHT had no effect (Fig. 4B).

Tfb2m involved in regulation of mtDNA transcription and We conclude that E2 increased mt biogenesis in both MG replication. Correspondingly, we detected an increase in and uterus, with a higher increase in uterus, which may

Tfam and Tfb1m in MG 6 h after treatment with E2 and 4-OHT (Fig. 3A). The pattern of Tfam, Tfb1m, and Tfb2m AB 2.5 MG3.5 Uterus 6 h * 6 h gene expression was different in uterus. E2 and 4-OHT 24 h 3.0 24 h 2.0 72 h 72 h repressed Tfb1m in the uterus 6 h after treatment and in 2.5 * * the 4-OHT-treated mice, Tfb1m remained repressed after 1.5 2.0 24 h. 4-OHT reduced uterine Tfam below control levels 1.0 1.5 1.0 * (relative to control) (relative to control)

after 24 h, but expression returned to control levels by mt:nuclear DNA ratio 0.5 mt:nuclear DNA ratio 0.5 72 h (Fig. 3B). While unchanged at 24 h, E increased Journal of Molecular Endocrinology 2 0.0 0.0 Tfam and Tfb1m expression after 72 h in both MG and Control E2 4-OHT Control E2 4-OHT uterus (Fig. 3A and B). No increase in Tfb2m was detected CD 3.0 3.5 MG * in response to either E2 or 4-OHT in either MG or uterus * 2.5 3.0 * (Fig. 3A and B). E2 and 4-OHT reduced Tfb2m in the uterus 2.5 2.0 24 h after treatment and 4-OHT reduced Tfbm2 in the 2.0 1.5 MG 72 h after treatment (Fig. 3A and B). 1.5 1.0 1.0

0.5 0.5 Relative Tomm40 protein level Relative Tomm40 protein level E2 and 4-OHT increase TFAM protein in MG and 4-OHT 0 0 Control E 4-OHT Control E 4-OHT reduces TFAM protein in uterus 2 2

E2 and 4-OHT increased TFAM protein in MG 72 h after Figure 4 treatment (Fig. 3C and Supplementary Figure 3). Reflect- E2 and 4-OHT differently regulate mitochondrial biogenesis in mouse mammary gland (MG) and uterus. (A and B) Total DNA was purified from ing the lack of induction of Tfam mRNA expression by E2 MG and uterus of ovex mice treated with vehicle (control), E2, or 4-OHT or 4-OHT in uterus (Fig. 3B), no increase in TFAM protein for 6, 24, and 72 h. The data are the ratio of the mitochondrial genome encoded gene mt-ND5 normalized to nuclear-encoded gene Ctfr as was detected in uterus 6 or 24 h after treatment or with E2 determined by qPCR. Values are the meanGS.E.M. of 6–15 mice/treatment 72 h after treatment (Fig. 3D and Supplementary Figure 3). group. *P!0.05 vs vehicle control (Student’s t-test). (C and D) Tomm40 Uterine TFAM protein levels were lower than control 72 h protein was examined in MG (C) and uterus (D) after 72 h treatment. The G ! after 4-OHT treatment. Concordantly, expression of mt values are the mean S.E.M. of 4–7 mice/treatment groups. *P 0.05 vs vehicle control (Student’s t-test). Representative western blots are shown in encoded cytochrome c oxidase I (mt-Co1) was significantly Supplementary Figure 4, see section on supplementary data given at the reduced by 4-OHT-treatment in uterus 24 and 72 h after end of this article.

A C COX4 Cox4 2.0 MG E2 and 4-OHT regulate nuclear-encoded NRF-1 target 3 * MG CYSS CycS * * genes Cox4 and CycS * 1.5 2 * 1.0 Cox4 and CycS are nuclear-encoded, NRF-1 target genes for 1 0.5 OXPHOS. We tested the hypothesis that an increase in Relative protein level Relative 0 0.0 2 2 2 2 E E E E NRF-1 protein should result in an increase in the Relative mRNA expression Relative 4-OHT 4-OHT 4-OHT 4-OHT Control Control Control Control expression of its target genes. Cox4 mRNA expression 6 h 24 h 72 h 72 h B D Uterus was increased in MG with E2 treatment at 6 and 24 h 5 3 Uterus COX4 Cox4 4 * CYCS whereas uterine expression was unaffected by E2 and CycS 2 3 * * * inhibited by 4-OHT at all time points (Fig. 5A and B). Cox4 2 1 protein was increased in MG and uterus 72 h after E2 and expression * * * * 1 Relative mRNA Relative

Relative protein level Relative 4-OHT treatment (Fig. 5C and D and Supplementary 0 0 2 2 2 2 E E E E Figure 6A, see section on supplementary data given at the 4-OHT 4-OHT 4-OHT 4-OHT Control Control Control Control end of this article). The reason for the discrepancy between 6 h 24 h 72 h 72 h the absence of an effect of 4-OHT on Cox4 transcript expression and the increase in COX4 protein detected in Figure 5 MG and the inhibition of Cox4 transcript expression but E2 and 4-OHT differently regulate the level of NRF-1 target genes involved in mitochondrial respiration. (A and B) Q-RT-PCR analysis of the mRNA an increase in COX4 protein in uterus at the 72-h time expression of Nrf1-regulated genes Cox4 and CycS in mammary gland (MG) point is unknown. Among the possible explanations are (A) and uterus (B) from ovex mice treated with vehicle (control), E2,or 4-OHT for the indicated time. Values are the meanGS.E.M. of 10–15 increased protein stability and altered mRNA stability. ! mice/treatment group. *P 0.05 vs vehicle control (Student’s t-test). CycS expression was signiﬁcantly increased by E2 in (C) COX4 and (D) cytochrome c protein expression was examined in MG and MG and decreased by 4-OHT in the uterus 72 h after uterus 72 h after the indicated treatment. Values are the meanGS.E.M.of 5–12 mice/treatment group. *P!0.05 vs vehicle control (Student’s t-test). Representative western blots are shown in Supplementary Figure 5,see A section on supplementary data given at the end of this article. MG Ppargc1a 3 Ppargc1b * * reﬂect E2-induced increases in transcription, protein synthesis, and cell proliferation, all metabolically 2 demanding, in the ovex mouse uterus at these time points (Moggs et al. 2004). 1 Journal of Molecular Endocrinology

Relative mRNA expression Relative 0 2 2 2 E E Tissue-speciﬁc regulation of outer mt Tomm40 E 4-OHT 4-OHT 4-OHT Control Control protein expression Control 6 h 24 h 72 h Because E2 increased the mt:nuclear DNA ratio at 24 h in B Uterus MG and uterus, with more of an increase in uterus vs MG, 3 Ppargc1a while 4-OHT increased the mt:nuclear DNA ratio only in Ppargc1b uterus (Fig. 4A), we examined the protein level of one 2 component of the translocase of the outer mitochondrial membrane (TOM) complex that controls the transport of 1 * * * * * nuclear-encoded proteins into mitochondria as a repre- * * * *

sentative outer mt membrane protein (Hoogenraad & mRNA expression Relative 0 2 2 2 E Ryan 2001). Tomm40 is the pore-forming unit of the TOM E E 4-OHT 4-OHT 4-OHT Control complex (Hill et al. 1998). E increased Tomm40 protein Control Control 2 6 h 24 h 72 h levels in MG and uterus 72 h after treatment (Fig. 4C and D and Supplementary Figure 5), a result that correlates with Figure 6 E2-induced mt biogenesis at 24 h in these tissues, which E2 and 4-OHT increase Ppargc1b expression in mammary gland (MG) and would then be detected at the protein level, e.g. 72 h as reduce Ppargc1a and Ppargc1b expression in the uterus. Q-RT-PCR analysis measured here. 4-OHT increased mt biogenesis only in the of Ppargc1a and Ppargc1b expression in MG (A) and uterus (B) from ovex C57BL/6 mice treated with vehicle (control), E2, or 4-OHT for the indicated uterus and not in MG after 24 h and likewise Tomm40 time. Values are the meanGS.E.M. of 5–10 mice/treatment group. *P!0.05 protein was increased only in the uterus after 72 h. vs vehicle control (Student’s t-test).

treatment (Fig. 5A and B). Cytochrome c protein levels did mouse brain endothelial cells (Kemper et al. 2013). Our

not change in MG of mice treated with either E2 or 4-OHT data suggest that Ppargc1a and Ppargc1b are differentially

(Fig. 5D and Supplementary Figure 6B). However, E2 regulated by E2 and 4-OHT in both MG and uterus. increased cytochrome c protein in uterus (Fig. 5C and Supplementary Figure 6B), a result in contrast to the lack Colocalization of NRF-1 and coactivator PGC-1a in of change in CycS mRNA (Fig. 5B). mouse uterus

To address whether NRF-1 and PGC-1a are expressed in the Expression of NRF-1 coactivators PGC-1a and PGC-1b same cell types in mouse uterus, IHC staining was in mouse MG and uterus performed on tissues from ovex mice treated with E2 for PGC1-a and PGC-1b are essential for NRF-1 regulation of 24 or 48 h (Fig. 7). NRF-1 staining was more prominent in gene expression (Scarpulla 2006, 2008a). Ppargc1b was the cytoplasm in vehicle-treated uterine tissues. However,

increasedinMG6hafterE2 and 4-OHT treatment both stromal and epithelial cells showed enhanced

(Fig. 6A). In uterus, Ppargc1a was reduced 6 h after nuclear staining upon E2 treatment for 24 h but not for

4-OHT treatment, and 24 and 72 h after E2 and 4-OHT 48 h. PGC-1a staining was observed in both cytoplasm and treatment (Fig. 6B). Ppargc1b expression was reduced in nucleus of stromal and epithelial cells in vehicle-treated

uterus 24 h after 4-OHT treatment and 72 h after E2 and tissues; however, PGC-1a staining was primarily in the 4-OHT treatment (Fig. 6B). In conclusion, Ppargc1a nucleus in 24 h-treated tissues but not in 48 h-treated

expression was not regulated by E2 or 4-OHT in MG and tissues. Consistent with Q-RT-PCR results (see Fig. 6B),

decreased in uterus (Fig. 6A), consistent with the reports staining was reduced in uteri treated with E2 for 24 and

that E2 repressed PGC-1a in mouse uterus (Macari et al. 48 h compared with vehicle. These results indicate that E2 2010), GeoProﬁle GDS1058 showed Pparg1c in uterus of enhances transient nuclear translocation of NRF-1 and

ovex control, but not E2-treated ovex CD1 mice 4, 8, or PGC-1a in adult uteri. Our results also indicate that E2 24 h after treatment. Ovariectomy increased PGC-1a in regulates NRF-1 and PGC-1a in both immature and mature

E2 48 h IgG control Veh 48 h E2 24 h E2 48 h

NRF-1 Journal of Molecular Endocrinology

PGC-1α

Merge

GE Merge wl GE GE GE LE LE nuclei LE LE S S S

Figure 7

E2 increases NRF-1 and PGC-1a nuclear colocalization in uterus and NRF-1 (green) and PGC-1a (red). Nuclei were stained with Hoechst 33258 decreases PGC-1a protein levels. Ovex mice were treated with vehicle (blue). LE, luminal epithelium; GE, glandular epithelium; S, stroma. White

(veh, EtOH) or 1 mgE2 for 24 or 48 h. Uterine tissue sections were stained for scale bar is 10 mm.

A No change Increase Decrease Not measured mRNA 6 h 24 h 72 h mRNA 6 h 24 h 72 h

MGE2 4-OHT E2 4-OHT E2 4-OHT Uterus E2 4-OHT E2 4-OHT E2 4-OHT Nrf1 Nrf1 Tfam Tfam Tfb1m Tfb1m Tfb2m Tfb2m Cox4 Cox4 CycS CycS Ppargc1a Ppargc1a Ppargc1b Ppargc1b

B Protein 6 h 24 h 72 h Protein 6 h 24 h 72 h

MGE2 4-OHT E2 4-OHT E2 4-OHT Uterus E2 4-OHT E2 4-OHT E2 4-OHT NRF-1 NRF-1 TFAM TFAM Cox4 Cox4 CycS CycS Tomm40 Tomm40

Figure 8

Summary of E2 and 4-OHT regulation of Nrf1, Ppargc1a, Ppargc1b, and Nrf1 and uterus are expressed relative to vehicle control after the indicated target gene expression in mouse mammary gland (MG) and uterus. treatment time (6, 24, or 72 h). Changes in the expression of the indicated genes (A) and proteins (B) in MG

uteri. Most importantly, it was evident that NRF-1 and Regulation of mt biogenesis has been largely deﬁned PGC-1a were expressed in virtually all cells in stroma and by studies in skeletal and cardiac muscle, liver, and brown epithelium and that they colocalized in the nucleus fat (Finck & Kelly 2006, 2007, Scarpulla 2008b). This study

regardless of treatment. demonstrates that E2 and 4-OHT increase Nrf1 mRNA and protein expression in a time- and tissue-speciﬁc manner in MG and uterus. Ppargc1a, Ppargc1b, and NRF-1-regulated Discussion nuclear-encoded genes are also regulated by E2 and 4-OHT in a tissue-speciﬁc manner as summarized Fig. 8. Although

The data presented here are the ﬁrst study of the Nrf1 expression was increased by E2 and 4-OHT 6 h after

Journal of Molecular Endocrinology regulation of endogenous Nrf1 expression, NRF-1 target treatment in both MG and uterus, NRF-1 protein was

gene expression, and mt biogenesis in normal mouse MG higher in uterus and E2 transiently increased nuclear

or uterus by E2 and 4-OHT. This is because Nrf1 was not NRF-1, and PGC-1a staining 4-OHT increased Nrf1 mRNA

included in microarrays previously used to identify E2- and but not NRF-1 protein in MG. The discrepancy between TAM-regulated genes in mouse uterus (Moggs et al. 2004, Nrf1 mRNA and protein may result from differences in Fong et al. 2007, 2010) and MG (Deroo et al. 2009). Earlier, mRNA stability but will require further investigation. Our

we reported that E2 and 4-OHT increase NRF-1 transcrip- results for E2 regulation show an expected inverse tion in breast cancer cells and HUVECs (Mattingly et al. correlation with recent data showing that ovariectomy 2008, Ivanova et al. 2011, Mattingly & Klinge 2012) and decreased NRF-1, TFAM, and Ppargc1b while increasing

that E2 activates NRF-1 transcriptional activity and mt Ppargc1a at both mRNA and protein levels in isolated biogenesis in human breast cancer cells (Mattingly et al. mouse brain endothelial cells (Kemper et al. 2013). 2008, Ivanova et al. 2011). Because NRF-1 is a master Interestingly, our study suggests that the MG appears transcriptional regulator of genes regulating mt function to be more responsive than uterus to upregulation of

(Scarpulla 2008a), we proposed a model that nuclear ER NRF-1 target genes Tfam, Tfb1m, and Cox4 after E2 or upregulation of NRF-1 which, in turn, stimulates the 4-OHT treatment. Serial analysis of gene expression transcription of nuclear-encoded mt genes, thus coordi- (SAGE) identiﬁed four OXPHOS genes Cox6a2, Mt-Co1, nately regulates nuclear–mt transcription and function Mt-Co3, and Mt-Nd3, upregulated in MG 3 h after a single,

(Klinge 2008, Mattingly et al. 2008, Ivanova et al. 2011). In s.c. injection of E2 (50 ng) in aERKO mice, allowing the

normal tissue, E2 increased NRF-1 protein in rat brain authors to conclude that E2 regulates the expression of (Stirone et al. 2005) and reduced Nrf1 mRNA expression in Cox6a2 and the mt encoded genes mainly through ERb mouse brown adipocytes (Rodriguez-Cuenca et al. 2007). (Aboghe et al. 2009). Importantly, and commensurate

with our results that E2 increases TFAM in MG, all three mt To explain the mechanisms involved in regulation

encoded genes upregulated by E2 (MtCo1, MtCo3, and of mouse Nrf1 transcription, we performed ChIP that MtNd3 (Aboghe et al. 2009)) are regulated by TFAM revealed increased ERa occupation of the ERE in the Nrf1

(Scarpulla 2008b). Based on the results summarized in promoter in MG and uterus after E2 and 4-OHT treatment. Fig. 8 and modeled in Fig. 9, we suggest that NRF-1 Similar to ERa ChIP-seq identiﬁcation of ERa binding upregulates its target genes selectively in the MG at least in sites in a whole genome study of mouse uterus (Hewitt part because of the expression of Ppargc1b in MG. In et al. 2012), we also observed ERa bound directly to the contrast, uterine Ppargc1a and Ppargc1b expression are ERE-containing region of the Nrf1 promoter in the uterus

decreased with E2 and 4-OHT treatment. Our results agree of the ovex, control-treated mice. At the same time, our

with other reports showing that E2 repressed Ppargc1a results differ from the apparent lack of ERa recruitment to

transcription and reduced PGC-1a protein in mouse uterus the Nrf1 gene in uterus with 1 h 0.25 mgE2 injection of (Macari et al. 2010) and ovariectomy increased PGC-1a ovex C57BL/6J mice in this ChIP-seq analysis; however,

and reduced PGC-1b in mouse brain endothelial cells the treatment time, route, and dose of E2 are different (Kemper et al. 2013). between our study and that from the Korach Laboratory

Mammary Gland Uterus

E –ERα Nucleus 2 E –ERα Nucleus 2

Ppargc1b Nrf1 Ppargc1a 6, 72 h Ppargc1b Nrf1 6, 24 h 72 h Tfam, Tfb1m 72 h Cox4 Cox4, CycS CycS Tfam, Tfb1m

Cox4 TFAM Cox4, CycS TFAM

Tomm40 Tomm40

mtDNA E2: 24 h E2: mtDNA 24,72 h OXPHOS OXPHOS Mitochondria Mitochondria Journal of Molecular Endocrinology Mammary Gland Uterus

4-OHT–ERα 4-OHT–ERα 6 h 6 h 6 h 6,24 h Ppargc1b Nrf1 Ppargc1a Ppargc1b Nrf1 Nucleus Nucleus 6 h Nrf1 6,24 h 6 h Nrf1 Cox4 Tfam, Tfb1m CycS

Cox4 TFAM 72 h Cox4, CycS TFAM

Tomm40 Tomm40

mtDNA 72 h mtDNA OXPHOS OXPHOS Mitochondria Mitochondria

Figure 9

Model of E2–ERa and 4-OHT–ERa regulation of Nrf1 transcription, NRF-1 MG on the left side and uterus on the right side. Of course, as we used regulation of its target genes, PGC-1 coactivator family expression, whole tissue lysates for the experiments in this study, we cannot attribute Tomm40, and mitochondrial biogenesis in mouse mammary gland (MG) gene/protein changes to a speciﬁc cell type. The time (h) at which changes

and uterus. The E2–ERa-stimulated changes are shown in the top two cell in gene or protein expression were detected is indicated. Bold black models with MG on the left side and uterus on the right side. The 4-OHT– arrows indicate the direction (up- or down-regulation) of the change of ERa-stimulated changes are shown in the bottom two cell models with gene expression.

(Hewitt et al. 2012). Interestingly, ERa was recruited to mouse Tfam promoter (Hewitt et al. 2012). In conclusion,

the Tfam gene with E2 treatment in mouse uterus (Hewitt our data demonstrate that E2 and 4-OHT differentially et al. 2012), commensurate with the increase in Tfam regulate NRF-1 and its downstream gene targets in MG

mRNA expression with E2 in mouse uterus seen 72 h and uterus with a stronger impact of E2 on stimulating mt 1 after treatment (Fig. 3B). The ERa occupation of AP-1/ ⁄2 biogenesis in uterus than MG. ERE site in Nrf1 intron 3 in MG from 4-OHT-treated mice is intriguing because it corresponds to inhibition of NRF-1

expression in our study. Mechanistically, this observation Supplementary data may be related to the tethering interaction of ERa with This is linked to the online version of the paper at http://dx.doi.org/10.1530/ AP-1 (Paech et al. 1997, Cerillo et al. 1998, Heldring et al. JME-13-0051. 2011), but further experiments will be required to dissect the molecular mechanisms involved. ChIP-seq demonstrated ERa bound to the ﬁrst intron of the Wnt4 and Declaration of interest The authors declare that there is no conﬂict of interest that could be Cdkn1 genes in E2-treated mouse uterus and upregulated perceived as prejudicing the impartiality of the research reported. expression (Hewitt et al. 2012), establishing the ability of intron-bound ERa to regulate gene transcription.

Although TFAM, Tfb1m, and Tfb2m are essential for Funding mt gene transcription and DNA replication (Scarpulla This work was supported by NIH R01 DK053220 to C M K.

2008a), their regulation by E2 and 4-OHT did not

completely match effects of E2 and 4-OHT on mt

biogenesis. E2 increased mt biogenesis in the MG and Acknowledgements uterus after 24 h, in agreement with increases in Tfam and The authors thank Dr Numan Al-Rayyan and Kathleen Gajewski for their help in collecting mouse tissues and confocal microscopy respectively. Tfb1m only in the MG. Although 4-OHT increased Tfam They also thank high school student Meghan Hyman whose summer and Pgpargc1b in MG, the mt biogenesis was not increased. research was supported by a grant from the James Graham Brown In fact, after 72 h treatment, 4-OHT reduced MG mt Cancer Center and Maria Lohr, a summer student from Centre College (Danville, KY), for their assistance with some of the experiments. They biogenesis, in agreement with reduced Tfb2m and the thank Dr Barbara J Clark for her thoughtful suggestions for this manuscript. return of Tfb1m to basal 24 h after treatment. This may relate to the fact that TFAM and Tfb1m alone are insufficient to initiate mtDNA replication and Tfb2m References Journal of Molecular Endocrinology promotes mtDNA replication more efficiently than Tfb1m (Cotney et al. 2007). It may also relate to the pro- Aboghe DH, Yoshioka M, Phaneuf D & St-Amand J 2009 Regulation of gene expression by estrogen in mammary gland of wild type apoptotic activity of TAM in MG (Kotoula et al. 1993). and estrogen receptor a knockout mice. Journal of Steroid A summary of our findings is diagramed in Fig. 8 and Biochemistry and Molecular Biology 113 116–126. (doi:10.1016/j.jsbmb. 2008.12.002) modeled in Fig. 9.E2 and 4-OHT increase in vivo ERa Balsitis SJ, Sage J, Duensing S, Mu¨nger K, Jacks T & Lambert PF 2003 recruitment to the Nrf1 gene and regulate NRF-1 signaling Recapitulation of the effects of the human papillomavirus type 16 E7 in time- and tissue-dependent manner in mouse MG and oncogene on mouse epithelium by somatic Rb deletion and detection uterus. NRF-1 activated more of its target genes in MG of pRb-independent effects of E7 in vivo. Molecular and Cellular Biology 23 9094–9103. (doi:10.1128/MCB.23.24.9094-9103.2003) than in uterus after E or 4-OHT treatment, whereas more 2 Barros Rodrigo PA & Gustafsson J-A˚ 2011 Estrogen receptors and the genes were downregulated by E2 and 4-OHT in the uterus. metabolic network. Cell Metabolism 14 289–299. (doi:10.1016/j.cmet. These data are congruent with a report where only ten 2011.08.005) Bauerly KA, Storms DH, HarrisCB, Hajizadeh S,Sun MY, CheungCP, Satre MA, genes were regulated by E2 in mouse MG (Aboghe et al. Fascetti AJ, Tchaparian E & Rucker RB 2006 Pyrroloquinoline quinone 2009) whereas in mouse uterus, E2 altered the expression nutritional status alters lysine metabolism and modulates mitochondrial of 3538 genes (Moggs et al. 2004). The increase in NRF-1 DNA content in the mouse and rat. Biochimica et Biophysica Acta 1760 1741–1748. (doi:10.1016/j.bbagen.2006.07.009) target gene Tfam in MG and uterus 6 h after E2 and 4-OHT Billon-Gale´s A, Fontaine C, Filipe C, Douin-Echinard V, Fouque M-J, treatment was different from the time-delayed increase in Flouriot G, Gourdy P, Lenfant F, Laurell H, Krust A et al. 2009 The TFAM expression detected after E2, not 4-OHT, treatment transactivating function 1 of estrogen receptor a is dispensable for the of ERa-expressing MCF-7 breast cancer cells (Ivanova vasculoprotective actions of 17b-estradiol. PNAS 106 2053–2058. et al. 2011), reflecting a report that in vivo uterine gene (doi:10.1073/pnas.0808742106) Bozek K, Relogio A, Kielbasa SM, Heine M, Dame C, Kramer A & Herzel H regulation by E2–ERa is different from MCF-7 cells and 2009 Regulation of clock-controlled genes in mammals. PLoS ONE 4 may reflect direct recruitment of liganded ERa to the e4882. (doi:10.1371/journal.pone.0004882)

Bozek K, Rosahl AL, Gaub S, Lorenzen S & Herzel H 2010 Circadian Hoogenraad NJ & Ryan MT 2001 Translocation of proteins into transcription in liver. Bio Systems 102 61–69. (doi:10.1016/j.biosystems. mitochondria. IUBMB Life 51 345–350. (doi:10.1080/ 2010.07.010) 152165401753366096) Cam H, Balciunaite E, Blais A, Spektor A, Scarpulla RC, Young R, Kluger Y & Ivanova MM, Luken KH, Zimmer AS, Lenzo FL, Smith RJ, Arteel MW, Dynlacht BD 2004 A common set of gene regulatory networks links Kollenberg TJ, Mattingly KA & Klinge CM 2011 Tamoxifen increases metabolism and growth inhibition. Molecular Cell 16 399–411. nuclear respiratory factor 1 transcription by activating estrogen (doi:10.1016/j.molcel.2004.09.037) receptor b and AP-1 recruitment to adjacent promoter binding sites. Campbell CT, Kolesar JE & Kaufman BA 2012 Mitochondrial transcription FASEB Journal 25 1402–1416. (doi:10.1096/fj.10-169029) factor A regulates mitochondrial transcription initiation, DNA packa- Jeong J-W, Lee KY, Han SJ, Aronow BJ, Lydon JP, O’Malley BW & DeMayo FJ ging, and genome copy number. Biochimica et Biophysica Acta 1819 2007 The p160 steroid receptor coactivator 2, SRC-2, regulates 921–929. (doi:10.1016/j.bbagrm.2012.03.002) murine endometrial function and regulates progesterone-independent Cerillo G, Rees A, Manchanda N, Reilly C, Brogan I, White A & Needham M and -dependent gene expression. Endocrinology 148 4238–4250. 1998 The oestrogen receptor regulates NFkB and AP-1 activity in a cell- (doi:10.1210/en.2007-0122) specific manner. Journal of Steroid Biochemistry and Molecular Biology 67 Kemper MF, Zhao Y, Duckles SP & Krause DN 2013 Endogenous ovarian 79–88. (doi:10.1016/S0960-0760(98)00078-8) hormones affect mitochondrial efficiency in cerebral endothelium Cooke PS & Naaz A 2004 Role of estrogens in adipocyte development and via distinct regulation of PGC-1 isoforms. Journal of Cerebral Blood function. Experimental Biology and Medicine 229 1127–1135. Flow and Metabolism 33 122–128. (doi:10.1038/jcbfm.2012.159) Cotney J, Wang Z & Shadel GS 2007 Relative abundance of the human Klinge CM 2000 Estrogen receptor interaction with co-activators and mitochondrial transcription system and distinct roles for h-mtTFB1 co-repressors. Steroids 65 227–251. (doi:10.1016/S0039-128X and h-mtTFB2 in mitochondrial biogenesis and gene expression. (99)00107-5) Nucleic Acids Research 35 4042–4054. (doi:10.1093/nar/gkm424) Klinge CM 2008 Estrogenic control of mitochondrial function and Couse JF & Korach KS 1999 Estrogen receptor null mice: what have we biogenesis. Journal of Cellular Biochemistry 105 1342–1351. learned and where will they lead us? Endocrine Reviews 20 358–417. (doi:10.1002/jcb.21936) (doi:10.1210/er.20.3.358) Korach KS, Emmen JM, Walker VR, Hewitt SC, Yates M, Hall JM, Swope DL, Harrell JC & Couse JF 2003 Update on animal models developed for Deblois G & Giguere V 2013 Oestrogen-related receptors in breast cancer: analyses of estrogen receptor biological activity. Journal of Steroid control of cellular metabolism and beyond. Nature Reviews. Cancer 13 Biochemistry and Molecular Biology 86 387–391. (doi:10.1016/S0960- 27–36. (doi:10.1038/nrc3396) 0760(03)00348-0) Deroo BJ, Hewitt SC, Collins JB, Grissom SF, Hamilton KJ & Korach KS 2009 Kotoula V, Karkavelas G, Economou L, Sionga A, Boutis L & Kerameos- Profile of estrogen-responsive genes in an estrogen-specific mammary Foroglou C 1993 Effects of tamoxifen and CV 205502 on the gland outgrowth model. Molecular Reproduction and Development 76 morphology and the evolution of the noncancerous mouse mammary 733–750. (doi:10.1002/mrd.21041) gland. Histology and Histopathology 8 627–636. Faulds MH, Zhao C, Dahlman-Wright K & Gustafsson J-A˚ 2012 The Macari C, Teyssier C, Tribollet V, Mouzat K, Forcet C, Horard B, Lobaccaro J-M diversity of sex steroid action: regulation of metabolism by estrogen & Vanacker J-M 2010 Estrogens repress PGC1-a expression in the uterus. signaling. Journal of Endocrinology 212 3–12. (doi:10.1530/JOE-11-0044) Molecular and Cellular Endocrinology 330 33–40. (doi:10.1016/j.mce.2010. Finck BN & Kelly DP 2006 PGC-1 coactivators: inducible regulators of 08.003) energy metabolism in health and disease. Journal of Clinical Investigation Mattingly KA & Klinge CM 2012 Diesel exhaust particulate extracts inhibit 116 615–622. (doi:10.1172/JCI27794) transcription of nuclear respiratory factor-1 and cell viability in human Finck BN & Kelly DP 2007 Peroxisome proliferator-activated receptor umbilical vein endothelial cells. Archives of Toxicology 86 633–642. Journal of Molecular Endocrinology {gamma} coactivator-1 (PGC-1) regulatory cascade in cardiac physi- (doi:10.1007/s00204-011-0778-y) ology and disease. Circulation 115 2540–2548. (doi:10.1161/CIRCU- Mattingly KA, Ivanova MM, Riggs KA, Wickramasinghe NS, Barch MJ & LATIONAHA.107.670588) Klinge CM 2008 Estradiol stimulates transcription of nuclear respir- Fong CJ, Burgoon LD, Williams KJ, Forgacs AL & Zacharewski TR 2007 atory factor-1 and increases mitochondrial biogenesis. Molecular Comparative temporal and dose-dependent morphological and tran- Endocrinology 22 609–622. (doi:10.1210/me.2007-0029) scriptional uterine effects elicited by tamoxifen and ethynylestradiol in Moggs JG, Tinwell H, Spurway T, Chang HS, Pate I, Lim FL, Moore DJ, immature, ovariectomized mice. BMC Genomics 8 151. (doi:10.1186/ Soames A, Stuckey R, Currie R et al. 2004 Phenotypic anchoring of gene 1471-2164-8-151) expression changes during estrogen-induced uterine growth. Environ- Fong CJ, Burgoon LD, Williams KJ, Jones AD, Forgacs AL & Zacharewski TR mental Health Perspectives 112 1589–1606. (doi:10.1289/ehp.7345) 2010 Effects of tamoxifen and ethynylestradiol cotreatment on uterine O’Brien JE, Peterson TJ, Tong MH, Lee E-J, Pfaff LE, Hewitt SC, Korach KS, gene expression in immature, ovariectomized mice. Journal of Molecular Weiss J & Jameson JL 2006 Estrogen-induced proliferation of uterine Endocrinology 45 161–173. (doi:10.1677/JME-09-0158) epithelial cells is independent of estrogen receptor {a} binding to Gleyzer N & Scarpulla RC 2011 PGC-1-related coactivator (PRC), a sensor of classical estrogen response elements. Journal of Biological Chemistry 281 metabolic stress, orchestrates a redox-sensitive program of inflam- 26683–26692. (doi:10.1074/jbc.M601522200) matory gene expression. Journal of Biological Chemistry 286 O’Lone R, Knorr K, Jaffe IZ, Schaffer ME, Martini PGV, Karas RH, 39715–39725. (doi:10.1074/jbc.M111.291575) Bienkowska J, Mendelsohn ME & Hansen U 2007 Estrogen receptors {a} Heldring N, Isaacs GD, Diehl AG, Sun M, Cheung E, Ranish JA & Kraus WL and {b} mediate distinct pathways of vascular gene expression, 2011 Multiple sequence-specific DNA-binding proteins mediate estro- including genes involved in mitochondrial electron transport and gen receptor signaling through a tethering pathway. Molecular generation of reactive oxygen species. Molecular Endocrinology 21 Endocrinology 25 564–574. (doi:10.1210/me.2010-0425) 1281–1296. (doi:10.1210/me.2006-0497) Hewitt SC, Li L, Grimm SA, Chen Y, Liu L, Li Y, Bushel PR, Fargo D & Korach KS O’Malley BW & Kumar R 2009 Nuclear receptor coregulators in cancer 2012 Research resource: whole-genome estrogen receptor a binding in biology. Cancer Research 69 8217–8222. (doi:10.1158/0008-5472.CAN- mouse uterine tissue revealed by ChIP-seq. Molecular Endocrinology 26 09-2223) 887–898. (doi:10.1210/me.2011-1311) Paech K, Webb P, Kuiper GG, Nilsson S, Gustafsson J, Kushner PJ & Scanlan TS Hill K, Model K, Ryan MT, Dietmeier K, Martin F, Wagner R & Pfanner N 1997 Differential ligand activation of estrogen receptors ERa and 1998 Tomm40 forms the hydrophilic channel of the mitochondrial ERb at AP1 sites. Science 277 1508–1510. (doi:10.1126/science.277. import pore for preproteins. Nature 395 516–521. (doi:10.1038/26780) 5331.1508)

Papi A, Storci G, Guarnieri T, De Carolis S, Bertoni S, Avenia N, Scarpulla RC 2008a Nuclear control of respiratory chain expression by Sanguinetti A, Sidoni A, Santini D, Ceccarelli C et al. 2013 nuclear respiratory factors and PGC-1-related coactivator. Annals of Peroxisome proliferator activated receptor-a/hypoxia inducible factor- the New York Academy of Sciences 1147 321–334. (doi:10.1196/annals. 1a interplay sustains carbonic anhydrase IX and apoliprotein E 1427.006) expression in breast cancer stem cells. PLoS ONE 8 e54968. Scarpulla RC 2008b Transcriptional paradigms in mammalian mito- (doi:10.1371/journal.pone.0054968) chondrial biogenesis and function. Physiological Reviews 88 611–638. Pendaries C, Darblade B, Rochaix P, Krust A, Chambon P, Korach KS, (doi:10.1152/physrev.00025.2007) Bayard F & Arnal JF 2002 The AF-1 activation–function of ERa may be Scarpulla RC 2012 Nucleus-encoded regulators of mitochondrial function: dispensable to mediate the effect of estradiol on endothelial NO integration of respiratory chain expression, nutrient sensing and production in mice. PNAS 99 2205–2210. (doi:10.1073/pnas. metabolic stress. Biochimica et Biophysica Acta 1819 1088–1097. 042688499) (doi:10.1016/j.bbagrm.2011.10.011) Scarpulla RC, Vega RB & Kelly DP 2012 Transcriptional integration of Renoir J-M, Marsaud V & Lazennec G 2013 Estrogen receptor signaling as a mitochondrial biogenesis. Trends in Endocrinology and Metabolism 23 target for novel breast cancer therapeutics. Biochemical Pharmacology 85 459–466. (doi:10.1016/j.tem.2012.06.006) 449–465. (doi:10.1016/j.bcp.2012.10.018) Simpkins JW, Yang S-H, Sarkar SN & Pearce V 2008 Estrogen actions on Richard CS 2011 Metabolic control of mitochondrial biogenesis through mitochondria – physiological and pathological implications. Molecular the PGC-1 family regulatory network. Biochimica et Biophysica Acta and Cellular Endocrinology 290 51–59. (doi:10.1016/j.mce.2008.04.013) 1813 1269–1278. (doi:10.1016/j.bbamcr.2010.09.019) Simpkins JW, Yi KD, Yang S-H & Dykens JA 2010 Mitochondrial Rodriguez-Cuenca S, Monjo M, Gianotti M, Proenza AM & Roca P 2007 mechanisms of estrogen neuroprotection. Biochimica et Biophysica Acta Expression of mitochondrial biogenesis-signaling factors in brown 1800 1113–1120. (doi:10.1016/j.bbagen.2009.11.013) adipocytes is influenced specifically by 17b-estradiol, testosterone, and Sourla A, Luo S, Labrie C, Be´langer A & Labrie F 1997 Morphological progesterone. American Journal of Physiology. Endocrinology and Metab- changes induced by 6-month treatment of intact and ovariectomized olism 292 E340–E346. (doi:10.1152/ajpendo.00175.2006) mice with tamoxifen and the pure antiestrogen EM-800. Endocrinology Safdar A, Little JP, Stokl AJ, Hettinga BP, Akhtar M & Tarnopolsky MA 2011 138 5605–5617. (doi:10.1210/en.138.12.5605) Exercise increases mitochondrial PGC-1a content and promotes Stirone C, Duckles SP, Krause DN & Procaccio V 2005 Estrogen increases nuclear–mitochondrial cross-talk to coordinate mitochondrial bio- mitochondrial efficiency and reduces oxidative stress in cerebral blood genesis. Journal of Biological Chemistry 286 10605–10617. (doi:10.1074/ vessels. Molecular Pharmacology 68 959–965. (doi:10.1124/mol.105. jbc.M110.211466) 014662) Scarpulla RC 2006 Nuclear control of respiratory gene expression in Yager JD & Chen JQ 2007 Mitochondrial estrogen receptors – new insights mammalian cells. Journal of Cellular Biochemistry 97 673–683. into specific functions. Trends in Endocrinology and Metabolism 18 (doi:10.1002/jcb.20743) 89–91. (doi:10.1016/j.tem.2007.02.006)

Received in ﬁnal form 17 July 2013 Accepted 26 July 2013 Accepted Preprint published online 26 July 2013 Journal of Molecular Endocrinology