Breast cancer cells utilize T3 to trigger proliferation through cellular Ca²⁺ modulation

Abstract

High levels of thyroid hormones are linked to increased risk and advanced stages of breast cancer. Our previous work demonstrated that the biologically active triiodothyronine (T3) facilitates mitochondrial ATP production by upregulating Ca²⁺ handling proteins, thereby boosting mitochondrial Ca²⁺ uptake and Krebs cycle activity. In this study, different cell types were utilized to investigate whether T3 activates a Ca²⁺-induced signaling pathway to boost cancer cell proliferation. Using live-cell imaging, biochemical assays, and molecular profiling, differences in intracellular signaling among MCF7 and MDA-MB-468 breast cancer cells, non-cancerous breast cells hTERT-HME1, and PC3 prostate carcinoma cells, previously found to be insensitive to thyroid hormones in terms of proliferation, were investigated. Our findings revealed that T3 upregulates 1,4,5-trisphosphate receptor 3 via thyroid hormone receptor α. This boosts mitochondrial Ca²⁺ uptake, reduction equivalent yield, and mitochondrial ATP production, supporting the viability and proliferation of breast cancer cells without affecting non-cancerous hTERT-HME1 or PC3 prostate carcinoma cells. Understanding the interplay between T3 signaling, organellar interaction, and breast cancer metabolism could lead to targeted therapies that exploit cancer cell vulnerabilities. Our findings highlight T3 as a crucial regulator of cancer metabolism, reinforcing its potential as a therapeutic target in breast cancer.

Graphical Abstract

Introduction

The metabolism of cancer cells is marked by dynamic alterations that promote their survival and proliferation [1], presenting a significant challenge in understanding and combating malignancies [2]. In this context, recent studies have shed light on the intricate interplay between cellular signaling pathways and metabolic reprogramming, highlighting the pivotal role of thyroid hormones (THs) as master regulators of cellular metabolism [3].

The release of THs is controlled by the hypothalamic-pituitary-thyroid axis. Thereby, the hypothalamus secretes the thyrotropin‐releasing hormone (TRH), which binds to the TRH receptor of the pituitary gland and stimulates the thyroid‐stimulating hormone (TSH) secretion. When TSH binds to the TSH receptor in the thyroid gland, TH synthesis is initiated. 3,5,3′,5′-tetraiodo-l-thyronine (T4), also known as thyroxin, serves as a precursor for the biologically active 3,5,3′-L-triiodothyronine (T3) and the inactive reverse T3 (rT3), produced by cleavage of one iodide through different deiodinases. T3 reveals its function via binding to nuclear thyroid hormone receptors α (THRα) and β (THRβ), which act as transcription factors and cause alterations in gene expression. Besides this so-called canonical mechanism, the function of T3 can also be initiated outside the nucleus to subsequently activate intracellular second messenger signaling pathways [4].

In epidemiological studies, low TSH level were associated with an increased risk for colon, lung, prostate, and breast cancer, while high TH levels were further associated with advanced clinical stages of breast cancer [5, 6]. Moreover, a study in female mice showed a correlation between increased TH levels and the incidence and aggressiveness of breast cancer [7]. In contrast, hypothyroid patients with glioblastoma and head and neck cancer had longer survival than euthyroid patients [5, 6].

Previous studies linked the effects on different cancer cells to the TH interaction with the surface receptor integrin αvβ3 and subsequent modulation of signaling cascades linked to phosphatidylinositol-3-kinase or mitogen-activated protein kinase that further modulation of estrogen receptor α [8]. For instance, T4 was shown to promote the proliferation of the human breast cancer cell line MCF7 by inducing the mitogen-acted protein kinase pathway, resulting in the phosphorylation and activation of the estrogen receptor α (ERα) [9]. Moreover, 10–100 nM T3 was shown to induce proliferation of MCF7 and T47D breast cancer cell lines via the activation of the ERα [10]. Besides, TH also increased proliferation in human non-small cell lung carcinoma and small cell lung cancer cells [11] and colorectal cancer cells [12] through integrin αvβ3 activation. Notably, a concentration of 1 µM T3 was found to repress the anti-apoptotic senescence-marker protein-30 and thereby facilitate apoptosis in MCF7 cells [13]. Moreover, T3 was reported to enhance the proliferation of human androgen-dependent prostate carcinoma cells LNCaP but not in PC3 cells [14]. These reports suggest that THs have not just a dose- but also cell-type-dependent effect.

Although breast cancer presents as a heterogeneous disease, various malignant breast tumors appear to adopt a common strategy for metabolic reprogramming. This involves the enhancement of fatty acid metabolism, accompanied by upregulation of electron transport chain (ETC) proteins and ATP synthase [15]. The function of the ETC and subsequent ATP generation by oxidative phosphorylation is largely reliable in terms of Krebs cycle activity. Within the Krebs cycle, pyruvate, isocitrate, and α-ketoglutarate dehydrogenases rely on mitochondrial Ca²⁺ levels. Consequently, different cancer cells also dynamically modulate the interaction between mitochondria and the biggest internal Ca²⁺ store, the endoplasmic reticulum (ER), and rely on a different set of proteins involved in mitochondrial Ca²⁺ homeostasis [16]. While the ER takes up cytosolic Ca²⁺ via sarcoplasmic-endoplasmic type ATPases [17], stored Ca²⁺ can be rapidly provided by the Ca²⁺ release channel inositol 1,4,5-trisphosphate receptors (IP3Rs), varying in expression profiles, activation, and function [18, 19]. The second messenger inositol 1,4,5-trisphosphate (IP3) is generated in response to various extracellular stimuli, which activate plasma membrane receptors and, subsequently, phospholipase C. This stimulation results in the hydrolysis of phosphatidylinositol 4,5-bisphosphate to inositol 1,4,5-trisphosphate and diacylglycerol. IP3 then binds to receptors of the IP3R family, which comprises three isoforms (IP3R1, IP3R2, and IP3R3) encoded by different genes and are differently expressed in various cell types [20]. Ca²⁺ flux between ER and mitochondria predominantly happens in regions of mitochondria-associated ER membranes, specialized ER membranes that reversibly bind to proteins located in the outer mitochondrial membrane. The voltage-dependent anion channel facilitates Ca²⁺ influx through the outer mitochondrial membrane, while the mitochondrial Ca²⁺ uniporter (MCU) regulates the Ca²⁺ uptake to the mitochondrial matrix, a mechanism further refined by the gatekeeper mitochondrial Ca²⁺ uptake 1 (MICU1). In a previous study, we found that the protein arginine methyl transferase 1 (PRMT1) causes hampered mitochondrial Ca²⁺ uptake by methylation of MICU1. Under these conditions, uncoupling proteins 2 and 3 (UCP2, UCP3) re-establish proper mitochondrial Ca²⁺ uptake [21]. This mode of mitochondrial Ca²⁺ uptake dependent on PRMT1 and UCP2 was found to be utilized by cancer cells to ensure a proper mitochondrial Ca²⁺ and, thus, energy supply [22]. Recently, we revealed that T3 facilitates mitochondrial ATP production through upregulation of the UCP2 and, thus, facilitation of mitochondrial Ca²⁺ uptake via the pore-forming MCU [3]. This finding hinted at a potential impact of THs on proliferation and migration of cancer cells through Ca²⁺ handling modulation.

Since previous clinical [5, 6] and preclinical studies [7] have pointed to a tumor-promoting effect of THs in breast cancer, we investigated whether T3 induces a Ca²⁺-induced signaling pathway that is specific to breast cancer cells. Besides different breast cancer cell lines, including MCF7 and MDA-MB-468, we utilized hTERT-immortalized human mammary epithelial cells hTERT-HME1 as a non-cancerous breast cell line and the prostate carcinoma cell line PC3, previously revealed to be insensitive to THs regarding proliferation [14]. Through a comprehensive approach involving cellular imaging, biochemical assays, and molecular profiling, we revealed that breast cancer cells utilize THs to fine-tune cellular Ca²⁺ handling to meet their increased energy demand, ensuring enhanced proliferation activity.

Methods

Cell culture and handling

Four different cell lines, including MCF7 (human breast cancer cells with estrogen, progesterone, and HER2 receptors) (RRID:CVCL_0031), PC3 (human prostatic adenocarcinoma cells) (RRID:CVCL_0035), and hTERT-HME1 (immortalized human mammary epithelial cells) (RRID:CVCL_3383), and MDA-MB-468 (ER-negative human breast cancer cells) (RRID:CVCL_0419) were obtained from American Type Culture Collection (ATCC) and cultured in a 37 °C (5% CO₂) incubator using cell culture media, consisting of 1% Dulbecco's Modified Eagle Medium (DMEM, D5523) (Sigma-Aldrich, Vienna, Austria) with the additives 10% fetal calf serum (Gibco-Thermo Fisher Scientific, Vienna, Austria), 1% streptomycin (Gibco-Thermo Fisher Scientific, Vienna, Austria), and 1% penicillin (Gibco-Thermo Fisher Scientific, Vienna, Austria). Cells were treated with 100 nM T3 (Sigma-Aldrich, Vienna, Austria; #T6379) in media for either 1 or 3 h before measurements were performed. T3 was dissolved in ethanol and 1M HCl (4:1), as previously described [3]. Control cells were treated for the same time with the respective T3-free solvent solution.

For passaging or harvesting, cells were washed with phosphate-buffered saline (PBS) and subsequently detached with 0.05% trypsin–EDTA solution at 37 °C. Afterward, cells were resuspended in cell culture media and centrifuged at 1500 rpm. Subsequently, the cell pellet was resuspended in the cell culture medium to determine the number of cells/ml of medium using a Neubauer chamber. Finally, the cells were seeded according to the desired cell number.

Transfection

Cells were transiently transfected at a confluence of 60 to 70%. For overexpression of organelle-targeted biosensors, transfection was performed with 1 µg plasmid DNA and 3 µl of PolyJet reagent (SignaGen Laboratories, Maryland, USA) in 1 ml of serum- and antibiotic-free medium. To achieve a knockdown (KD) transfection was performed with 1 µg plasmid DNA and 100 µM of respective siRNA using 2.5 µl of TransFast™ transfection reagent (Promega, Mannheim, Germany) in 1 ml of serum- and antibiotic-free medium. siRNAs were obtained from Microsynth (Balgach, Switzerland) and included human MCU siRNA-1: 5′-GCC AGA GAC AGA CAA UAC U dTdT-3′, human MCU siRNA-2: 5′-GGA AAG GGA GCU UAU UGA A dTdT-3′, human UCP2 siRNA: 5′- GCA CCG UCA AUG CCU ACA A dTdT-3′, human UCP3 5′- GGA ACU UUG CCC AAC AUC A dTdT-3′, human IP3R1 5′- GAA AUU CAG AGA CUG CCU CTT dTdT -3′, human IP3R3 5′- CA GAC UAA GCA GGA dTdT -3′, human THR-alpha 5′- CAA ACA CAA CAU UCC GCA UUT T dTdT -3′, human THRß 5′- GCC UGU GUU GAG AGA AUA GAA TT dTdT -3′, and human control siRNA: 5′- UUC UCC GAA CGU GUC ACG U dTdT-3′. After 4—6 h at 37°C, the transfection mixture was replaced with 2 ml of fresh cell culture medium.

Live-cell imaging

For live-cell imaging experiments, cells were seeded on 30 mm glass coverslips in 6-well plates and accordingly transfected (mentioned above) and maintained at 37 C until a confluence of 80–90% was obtained, as described [3]. Before the measurement, cells were kept in an EH-loading buffer (135 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 10 mM HEPES, 2.6 mM NaHCO₃, 440 μM KH₂PO₄, 340 μM Na₂HPO₄, 10 mM D-glucose, 0.1% vitamins, 0.2% essential amino acids, and 1% penicillin–streptomycin, pH adjusted to 7.4) at room temperature. During the measurement, the cells were continuously perfused by a perfusion system (PS9D, NGFI, Graz, Austria; www.ngfi.eu), providing different buffers to the cells. For all live cell imaging experiments, a Ca²⁺-containing buffer (145 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 10 mM D-glucose, and 10 mM HEPES, pH adjusted to 7.4) was used.

Assessment of mitochondrial Ca2+, ATP, and membrane potential

Alterations in mitochondrial Ca²⁺ homeostasis were examined in MCF7 and PC3 cells, expressing the mitochondrial-targeted biosensor 4mtD3cpv as described [3, 21]. Cells were stimulated with 100 µM of the IP3-generating agonist ATP (Sigma Aldrich, Vienna, Austria; #34,369–07-8) to evoke the mitochondrial Ca²⁺ uptake [18]. Basal mitochondrial ATP changes were measured using the FRET-based biosensor mtAT1.03 [23, 24]. To determine mitochondrial membrane potential, cells were incubated in 20 nM of the fluorescent indicator TMRM (Invitrogen™; Vienna, Austria; #T668) in EH-loading buffer for 20 min at room temperature, as described [25]. A full disruption of the mitochondrial membrane potential by application of carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP) (Abcam, Cambridge, UK; #370–86-5) was used to determine the minimum and thus normalize the TMRM_mito/ TMRM_nuc ratio. Measurements were performed using an inverted wide-field microscope (Observer.A1, Carl Zeiss GmbH; Vienna, Austria) with a 40 × immersion oil objective (Plan Apochromat 1,3 NA Oil DIC (UV) VISIR, Carl Zeiss GmbH) and a CFP/YFP filter cube as described [21, 23, 25]. 4mtD3cpv and mtAT1.03 were excited with a wavelength of 440 nm (440AF21, Omega Optical, Brattleboro, VT, USA), and the emission was detected at 480 nm and 535 nm (480AF30 and 535AF26, Omega Optical, Brattleboro, VT, USA) to determine the ratio F₅₃₅/F₄₈₀. TMRM was excited at 550 nm, and the emission was collected at 600 nm. A region of interest containing the mitochondrial TMRM fluorescence (TMRM_mito) and the cytosolic TMRM (generally the area of the nucleus) fluorescence (TMRM_nuc) was selected, and a ratio TMRM_mito/TMRM_nuc was calculated over time. Measurements were recorded with the NIS-Elements AR software (Nikon, Vienna, Austria) and analyzed using GraphPad Prism (GraphPad Software, San Diego, CA). For data evaluation, both channels were background subtracted by selecting a background region of interest. Moreover, fluorescence bleaching correction was applied using an exponential decay fit of the basal fluorescence extrapolated to the whole measurement.

Assessment of cytosolic Ca2+

To observe cytosolic Ca²⁺ levels in MCF7 and PC3, cells were incubated with 2 µM of the fluorescent cytosolic Ca²⁺ indicator Fura 2-acetoxymethyl ester (Fura-2AM) (Abcam, Cambridge, UK; #AB120873) for 30 min in EHL-Buffer, as previously described [3, 26]. After incubation with Fura-2AM, the cells were washed twice with EH-loading buffer. Fura-2AM was illuminated with a wavelength of 340 nm and 380 nm, and the emission was detected at 515 nm (495dcxru; Omega Optical, Brattleboro, VT, USA) to determine the ratio F₃₄₀/F₃₈₀. Cells were excited with a wavelength of 480 nm and 430 nm, and the emission was detected at 530 nm in order to determine the ratio F₄₈₀/F₄₃₀. Measurements in Fura-2AM stained cells were performed on an NGFI AnglerFish C-Y7G imager equipped with a Nikon CFI Super Fluor 40 × oil NA 1.3 objective equipped with the NGFI PS9D perfusion system (NGFI, Graz, Austria) was used. The results of the measurements were recorded with live-acquisition software and analyzed using GraphPad Prism (GraphPad Software, San Diego, CA). The measurement was background-subtracted using a background region of interest and corrected for bleaching using an exponential decay fit. Results are shown as the ratio of F₃₄₀/F₃₈₀.

ER/Mitochondria colocalization and mitochondrial structure

To determine the colocalization between ER and mitochondria, MCF7 cells were transfected with the ER-targeted biosensor ERAT1.03 [27]. Therefore, cells were incubated with 100 nM TMRM in EH-loading buffer for 30 min and 3D imaged on a confocal spinning disc microscope (Axio Observer.Z1 from Zeiss, Gottingen, Germany) with Z increments of 0.2 µm as previously described [16, 28]. The microscope is equipped with a Nipkow-based confocal scanner unit (CSU-X1, Yokogawa Electric Corporation), a motorized filter wheel (CSUX1FW, Yokogawa Electric Corporation), a 100 × objective lens (Planapochromat 100x/1.4 Oil, Zeiss, Germany) and an AOTF-based laser merge module for laser lines 405, 445, 473, 488, 514, and 561 nm (Visitron Systems, Puchheim, Germany). All images were taken with a CoolSnap HQ2-CCD camera (Photometrics, Tucson, Arizona, USA). Data acquisition and software control was achieved by VisiView Premier Acquisition software (2.0.8, Visitron Systems, Puchheim, Germany). Image stacks were deconvoluted through blind deconvolution (NIS-Elements 5.20.02, Nikon, Austria). Morphology parameters were measured automatically via a custom-made ImageJ macro using the following procedure. An additional background subtraction based on the rolling ball method was introduced to enhance contrast for later analysis. A global auto Otsu threshold using a stack histogram as well as a local auto Otsu threshold (radius of 640 nm) based on a single slice histogram were applied to the stack and merged. ERAT1.03 NA staining was Otsu thresholded, dilated, filled, and subsequently used as a mask for TMRM-stained mitochondria to exclude non-transfected cells. Binarized mitochondria were segmented via the ImageJ plugin 3D manager. Mitochondrial volume and surface were determined through the plugin 3D Geometrical Measure. The Plugin 3D Ellipsoid Fitting generated an ellipsoid fit of mitochondria to measure mitochondrial elongation. To analyze the colocalization between ER and mitochondria, TMRM and ERAT1.03 NA were determined on a single-cell level using ImageJ and plugin coloc2. The Pearson's coefficient was calculated as previously described [16].

Mitochondrial NADH autofluorescence measurement

Mitochondrial NADH + H⁺ autofluorescence was acquired by excitation of the different cell lines at 340 nm, as described [29]. In brief, cells were perfused with Ca2⁺ buffer for 2 min to establish a baseline. Afterward, 1 µM FCCP was added until the signal reached a persistent minimum. Following this, 100 mM NaN₃ was applied to inhibit mitochondrial respiration and a stable plateau was attained. Finally, by setting the minimal NADH + H+ autofluorescence after FCCP as 0% and maximum NADH + H⁺ autofluorescence after NaN3 as 100%, the NADH + H⁺ redox index was calculated in relation to the baseline.

NAD⁺/NADH + H⁺ ratio

NAD⁺/NADH + H⁺ ratios and NADH + H⁺ levels were measured with the Amplite TM Flourimetric NAD⁺/NADH + H⁺ Ratio Assay kit from (AAT Bioquest, California, US; #15,263). The reaction was performed following the manufacturer's protocol. One day before the experiment, respectively treated MCF7 and PC3 cells were plated on 96 polystyrene cell culture microplates at a density of 35,000 cells/well. The protein concentration of each treatment group was determined using the PierceTM BCA Protein Assay Kit (Fisher Scientific, Vienna, Austria).

Detection of mRNA expression levels

For RNA extraction, the peqGOLD Total RNA Kit (VWR Life Science, Vienna, Austria) was used according to the manufacturer's instructions. The concentration was measured, and the purity was controlled using the spectrophotometer Nanodrop ND-1000 Spectrophotometer (Peqlab, Erlangen, Germany), which was used to subsequently transcribe the isolated RNA into complementary DNA (cDNA). cDNA transcription was performed using the Applied Biosystems cDNA reverse transcription kit (Thermno Fischer, Foster City, CA; #4,368,814). QuantiFast SYBR Green RT-PCR kit (Qiagen, Hilden, Germany) was used for the quantitative real-time PCR (qRT-PCR) and performed on the LightCycler 480 (Roche Diagnostics, Vienna, Austria). As a housekeeping gene, the human hypoxanthine phosphoribosyltransferase 1 (hHPRT1) from Qiagen QuantiTect was used to normalize expression levels of candidate genes. The primers used were obtained from Invitrogen (Vienna, Austria) and were as follows:

• UCP2: forward: 5’-TCCTGAAAGCCAACCTCATG-3‘, reverse 5’GGCAGAGTTCATGTATCTCGTC-3;

• PRMT1: forward: 5'-TGCTCAACACCGTGCTCTATGC-3', reverse: 5'-TCCTCGATGGCCGTCACATACA-3'; MCU: forward: 5'AGAGATAGGCTTGAGTGTGAAC-3', reverse: 5'-TTCCTGGCAGAATTTGGGAG-3';

• MCUb forward: 5'-TATAGTACCGTGGTGCCACCTGATG-3', reverse: 5'-TTGTAGGTCCTGAAGGAATGAACCA-MICU1: forward: 5'-CAGGTTCAGAGCATCATTCG-3’, reverse: 5'-GAACACAAGCCAGACTTGAG-3';

• IP3R1: forward: 5'- GGAGTTTCAGCCCTCAGTGG -3, reverse: 5'- CTTCAGGCACAGAGACCAGG -3;

• IP3R2: forward: 5'- GGGTGCAGGACAGTGGAACA -3, reverse: 5'- TTGGAGGGACTCCATCTCTATTTTT -3; IP3R3: forward: 5'- TTCATCAGCACTTTGGGGT-3, reverse: 5'- TGAAGAGGCAGTCACGGAAC -3;

• THR-alpha: forward: 5'- CAGAGGAGAACAGTGCCAGG -3, reverse: 5'- CTGCTCGTCTTTGTCCAGGT -3; THRß: forward: 5'- AAAAGAGACCTCCTGCTCCG -3, reverse: 5'- AAGGAAATCGCAGATCCCGC -3.

Protein isolation, quantification, and Western blot analysis

Cells were washed with PBS 48 h after transfection, scraped from 10 cm dishes, and lysed with RIPA buffer (Tris–HCl pH 7.6; 150 mM NaCl; 5 mM EDTA; 1% NP-40; 1% Triton X-100; 1% sodium deoxycholate; 0,1% SDS) supplemented with 20 µl of protease inhibitor cocktail (Sigma Aldrich, Vienna, Austria; P8340-1ML) per 1 ml RIPA buffer and transferred into pre-cooled tubes. Cell lysates were shock frozen in liquid nitrogen, and afterward, cell lysates were centrifuged at 18,000 g for 15 min at 4 °C. The supernatant was collected for protein quantification. The Pierce BCA Protein Assay Kit was used according to the manufacturer's protocol. Absorbance was finally measured on a spectrophotometer (UviLine 9400, SCHOTT Instruments). Western blots were performed following standard protocols. Samples were heat denatured in 1 × Laemmli sample buffer and subjected to a 10% SDS–PAGE gel in parallel to PageRulerTM Plus Prestained Protein Ladder (Fisher Scientific, Vienna, Austria) and then transferred onto a polyvinylidene difluoride membrane (Millipore, Vienna, Austria). Proteins were detected using the following antibodies: MCU (1:1000) (Cell Signaling Technology Cat# 14,997, RRID:AB_2721812), UCP2 ( 1:1000) (Cell Signaling Technology Cat# 89,326, RRID:AB_2721818), IP3R3 (1:1000) (Thermo Fisher Scientific Cat# A302-159A, RRID:AB_1720366) and β-actin (1:1000) (Cell Signaling Technology Cat# 8457, RRID:AB_10950489). The secondary horseradish peroxidase antibodies, anti-mouse (1:1000) (Vector Laboratories Cat# PI-2000, RRID:AB_2336177), anti-rabbit (1:1000) (Santa Cruz Biotechnology Cat# sc-2054, RRID:AB_631748), were used and visualized with super signal west pico luminol/enhancer developing solution (Fisher Scientific, Vienna, Austria).

Determination of cell proliferation

To determine cell proliferation rates, 100,000 cells were seeded per well in 6-well plates. For KD experiments, cells were transfected accordingly (mentioned above) with the respective siRNA 48 h before passaging. Cells were treated with 100 nM T3 in full DMEM, while control cells were treated with the respective T3-free stock solution. After five days, the cells were harvested, as mentioned above, and the number of cells/ml of medium using a Neubauer chamber was determined. At least three replicates were counted for each condition, and the averages were calculated. Subsequently, proliferation rates were calculated as a percentage of control.

Statistics

The data were analyzed using Microsoft Excel Professional Plus 2019 (Microsoft Corporation, Albuquerque, NM, USA). Statistical analyses and design of graphs were carried out with GraphPad Prism Version 9.3.0 for Windows (GraphPad Software, San Diego, CA, USA) using Student's t-test or one-way ANOVA, if applicable. Data are expressed as MEAN ± SEM unless otherwise indicated. Results were obtained from at least three independent experimental days. Differences were considered statistically significant at p < 0.05 and presented as specific p-values (* = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001).

Results

T3 augments mitochondrial Ca²⁺ uptake in MCF7 breast cancer cells

Fluorescence microscopy was conducted on MCF7 breast cancer cells and PC3 prostate cancer cells expressing the FRET-based mitochondrial Ca²⁺ sensor 4mtD3cpv, treated with 100 nM T3 for 3 h or the respective solvent control. Thereby, the basal Ca²⁺ levels and Ca²⁺ uptake upon IP3-generating agonist-induced ER Ca²⁺ depletion through ATP application were recorded in a live-cell imaging experiment (Fig. 1A). While basal mitochondrial Ca²⁺ levels in MCF7 and PC3 cells remained unaffected by T3 (Fig. 1B), mitochondrial Ca²⁺ uptake upon ATP treatment was significantly enhanced in T3-treated MCF7 cells (Fig. 1C). To explore whether alterations in mitochondrial Ca²⁺ dynamics were attributed to changes in cytosolic Ca²⁺, cells were treated with the cytosolic Ca²⁺ dye Fura-2AM, and both basal and IP3-induced elevation in cytosolic Ca²⁺ were measured (Fig. 1D). T3 treatment elicited no significant changes in either basal (Fig. 1E) or peak cytosolic Ca²⁺ levels (Fig. 1F) in MCF7 and PC3 cells. The mitochondrial membrane potential is the main driving force of Ca²⁺ uptake into the organelle. Interestingly, 3 h treatment with T3 had no impact on mitochondrial membrane potential in either MCF7 or PC3 cells (Fig. 1G). Given that mitochondrial Ca²⁺ is heavily reliant on the interplay between mitochondria and the primary internal Ca²⁺ reservoir, the ER, the interaction between mitochondria and ER was probed in MCF7 cells expressing the ER-targeted sensor ERAT1.03 and stained with tetramethylrhodamine (TMRM) using confocal microscopy (Fig. 1H). Notably, there was no discernible difference in the physical interaction between mitochondria and ER between treated and untreated MCF7 cells, reflected by Pearson's correlation coefficient (Fig. 1I). Furthermore, mitochondrial volume (Fig. 1J) and surface area (Fig. 1K) remained unaltered upon T3 treatment in MCF7 cells. Consequently, the enhanced mitochondrial Ca²⁺ uptake observed in T3-treated MCF7 cells was hypothesized to be linked to direct modulation of proteins involved in ER-mitochondrial Ca²⁺ flux.

Fig. 1

Treatment with T3 [100 nM] for 3 h triggers enhanced mitochondrial Ca²⁺ uptake. Representative curves showing [Ca²⁺]_mito for MCF7 control (black) and MCF7 cells treated with T3 (blue) (A). Bar graphs represent basal [Ca²⁺]_mito levels in MCF7 and PC3, transfected with the FRET sensor 4mtD3cpv, comparing control cells to cells incubated with T3 (B). Bar graphs represent [Ca²⁺]_mito uptake in MCF7 and PC3, comparing control cells to cells incubated with T3 (C). Representative curves showing [Ca²⁺]_cyto for MCF7 control (black) and MCF7 cells treated with T3 (blue) (D). Bar graphs represent basal [Ca²⁺]_cyto levels in MCF7 and PC3, incubated with FURA-2AM, comparing control cells to cells incubated with T3 (E). Bar graphs represent [Ca.²⁺]_cyto uptake in MCF7 and PC3, comparing control cells to cells incubated with T3 (F). Bar graphs reveal mitochondrial membrane potential normalized to fully disrupted mitochondrial membrane potential in response to FCCP [10 μM] in MCF7 and PC3, control cells, and cells treated with T3, incubated and perfused with TMRM (20 nM) (G). Images show colocalization between ER and mitochondria with and without T3 incubation. Cells expressing ERAT1.03 were incubated with TMRM (20 nM) (H). Bar graphs represent the Pearson correlation coefficient in MCF7 control cells and MCF7 cells treated with T3 (I). Bar graphs represent mitochondrial volume in MCF7 control cells and MCF7 cells treated with T3 (J). Bar graphs represent the mitochondrial surface in MCF7 control and MCF7 cells treated with T3 (K). Results were obtained in at least 3 independent experiments. All bar graphs show mean ± SEM. If applicable, significant differences were assessed via one-way ANOVA or unpaired t-test and presented as specific p-values (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001)

T3 promotes ER-mitochondrial Ca2+ flux through IP3R3 overexpression

First, we investigated mRNA expression differences of genes involved in Ca²⁺ handling between T3-sensitive MCF7 cells and T3-insensitive PC3 cells. MCF7 cells exhibited significantly enhanced mRNA expression levels of MCU, UCP2, PRMT1, and IP3R3, while the dominant negative subunit MCUb [30], as well as MICU1, remained unchanged (Fig. 2A). Prior research had already hinted at a cancer cell-specific, MCU-dependent mitochondrial Ca²⁺ uptake modulated by UCP2 and PRMT1 [21], a phenomenon potentially active in MCF7 breast cancer cells. Subsequently, we explored whether T3 further potentiated the expression of these pivotal mitochondrial Ca²⁺ handling genes. This examination revealed a significant increase in the mRNA expression of IP3R3 in MCF7 cells (Fig. 2B), which was also found to be the most abundant IP3R isoform in MCF7 cells (Additional File 1A). PC3 cells exhibited no significant alterations in the mRNA expression of Ca²⁺ handling proteins (Fig. 2C). These results pointed to a potentially enhanced ER-mitochondrial Ca²⁺ flux in MCF7 cells through increased IP3R3 expression. To test whether heightened mRNA expression levels of IP3R3 translated into augmented protein levels in MCF7 cells, we conducted Western blot analysis, and densitometric evaluation (Fig. 2D) of immunoreactive bands (Fig. 2E) revealed a significant rise in IP3R3 protein level, without any change in the protein expression of MCU and UCP2 (Fig. 2E, Additional File 1B). Functional assessment was subsequently performed by the application of a siRNA to induce a specific KD of IP3R3 (Fig. 2F). MCF7 cells with respective IP3R3 KD, as well as with MCU and UCP2 KD, were utilized for live-cell imaging experiments testing whether blockage of mitochondrial Ca²⁺ uptake could prevent the T3-induced Ca²⁺ rewiring. We, therefore, measured mitochondrial Ca²⁺ levels under basal conditions and following ATP addition (Fig. 2G). Remarkably, T3 treatment failed to induce an amplified mitochondrial Ca²⁺ uptake in cells with IP3R3 KD (Fig. 2H). Besides, siRNA-induced KD of MCU (Fig. 2I) and UCP2 (Fig. 2J) also prevented T3-induced increase in mitochondrial Ca²⁺ uptake, underscoring the critical role of these proteins in orchestrating Ca²⁺ translocation into mitochondria subsequent to ER Ca²⁺ depletion via IP3R3.

Fig. 2

T3 [100 nM] facilitates IP3R3 expression and ER-mitochondrial Ca²⁺ flux. Bar graphs represent the absolute mRNA expression levels of genes in MCF7 and PC3 control cells (A). Bar graphs represent the mRNA expression ratio, shown as a change in percentage, of respective genes in MCF7 (B) and PC3 (C) cells treated with T3 for 3 h, compared to the corresponding control cells (set to 100%). Bar graphs represent the protein expression ratio of MCU, UCP2, and IP3R3 in MCF7 cells treated with T3 for 3 h as a percentage, compared to the corresponding control (set to 100%) (D). The image shows a representative Western blot of IP3R3 and the corresponding housekeeping gene ß-actin in control cells or cells after T3 treatment (E). Bar graphs represent the mRNA expression ratio, shown as a change in percentage, of IP3R1 and IP3R3 in MCF7 transfected with siRNA against IP3R3, compared to control cells (set to 100% in percentage) (F). Representative curves showing [Ca²⁺]_mito for MCF7 + siRNA against IP3R3 (black) and MCF7 cells with siRNA against IP3R3 treated with T3 for 3 h (blue) (G). Bar graphs represent [Ca²⁺]_mito uptake in MCF7 transfected with siRNA against IP3R3 and 4mtD3cpv, comparing control cells to cells incubated with T3 for 3 h (H). Bar graphs represent [Ca²⁺]_mito uptake in MCF7 transfected with siRNA against MCU and 4mtD3cpv, comparing control cells to cells incubated with T3 for 3 h (I). Bar graphs represent [Ca²⁺]_mito uptake in MCF7 transfected with siRNA against UCP2 and 4mtD3cvpv, comparing control cells to cells incubated with T3 for 3 h (J). Results were obtained in at least 3 independent experiments. All bar graphs show mean ± SEM. If applicable, significant differences were assessed via one-way ANOVA or unpaired t-test and presented as specific p-values (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001)

Enhanced mitochondrial Ca2+ uptake fuels mitochondrial energy metabolism

Given the pivotal role of mitochondrial Ca²⁺ in driving the activity of Krebs' cycle dehydrogenases, an elevation in mitochondrial Ca²⁺ levels potentially enhances mitochondrial ATP production. To validate whether a rise in mitochondrial Ca²⁺ indeed influences mitochondrial ATP levels, we employed MCF7 cells expressing the mitochondrial ATP sensor mtAT1.03. Remarkably, after 3 h of T3 treatment, basal mitochondrial ATP levels increased in MCF7 cells but not in PC3 cells (Fig. 3A). As for mitochondrial Ca²⁺ uptake, T3 treatment prevented an increase in mitochondrial ATP levels in cells with siRNA-induced KD of IP3R3 (Fig. 3B). Besides, MCU KD (Fig. 3C) and UCP2 KD (Fig. 3D) prevented the impact of T3 on mitochondrial ATP levels in MCF7 cells, underscoring the potential of T3-induced rewiring of mitochondrial Ca²⁺ uptake by modulating ER-mitochondrial Ca²⁺ flux via IP3R3. Next, we determined mitochondrial ATP levels after IP3-induced ER-Ca²⁺ depletion, leading to a rise in mitochondrial Ca²⁺, which promptly caused an increase in mitochondrial ATP levels (Fig. 3E). Notably, treatment of MCF7 cells with 100 nM T3 for 3 h significantly facilitated the mitochondrial ATP production linked to the mitochondrial Ca²⁺ rise, while PC3 cells remained unaffected by T3 treatment (Fig. 3F). Also, ATP levels in MCF7 cells treated with siRNA against MCU did not show any alterations (Fig. 3G). Given that Ca²⁺ amplifies the activity of Krebs cycle dehydrogenases, we evaluated the yield of reduction equivalents by measuring NADH + H⁺ autofluorescence in MCF7 and PC3 cells. We induced maximal ETC activity by applying 1 µM FCCP and inhibited the ETC with 100 mM sodium azide (NaN₃) to get maximal NADH + H⁺ levels (Fig. 3H). Enhanced NADH + H⁺ levels were observed in MCF7 cells following treatment with 100 nM T3 for 3 h, whereas PC3 cells remained not significantly different (Fig. 3I). The finding of increased NADH + H⁺ levels was further confirmed by an fluorimetric assay analyzing reduction equivalents in the supernatant of cells (Fig. 3J). Moreover, the NAD + /NADH + H⁺ was found to be significantly reduced by T3 treatment in MCF7 cells, while PC3 remained unaffected (Fig. 3K). These results pinpoint that mitochondrial Ca²⁺ uptake facilitated by T3 causes enhanced reduction equivalent yield, which boosts ETC and, thereby, the mitochondrial ATP production.

Fig. 3

Enhanced mitochondrial Ca²⁺ uptake triggers mitochondrial ATP generation after T3 [100 nM] treatment for 3 h. Bar graphs represent basal [ATP]_mito in MCF7 and PC3 cells, transfected with mitochondrial-targeted mtAT1.03, comparing control cells to cells incubated with T3 (A). Bar graphs represent basal [ATP]_mito in MCF7 transfected with siRNA against IP3R3 and mtAT1.03, comparing control cells to cells incubated with T3 (B). Bar graphs represent [ATP]_mito in MCF7 transfected with siRNA against MCU and mtAT1.03, comparing control cells to cells incubated with T3 (C). Bar graphs represent [ATP]_mito in MCF7 transfected with siRNA against UCP2 and mtAT1.03, comparing control cells to cells incubated with T3 (D). Representative curves showing [ATP]_mito for MCF7 control cells (black) and MCF7 cells treated with T3 (blue) after stimulation with 100 µM ATP (E). Bar graphs represent [ATP]_mito levels in MCF7 and PC3 cells, transfected with mtAT1.03, comparing control cells to cells incubated with T3 (F). Bar graphs that represent [ATP]_mito uptake levels in MCF7, transfected with mtAT1.03 and siRNAi against MCU, comparing control cells to cells incubated with T3 (G). Representative curves showing NADH + H⁺ autofluorescence for MCF7 (blue) and PC3 cells (gray), after perfusing the cells with FCCP [1 µM] and NaN₃ [100 mM] (H). Bar graphs represent NADH + H⁺ redox index in percentage in MCF7 and PC3, comparing control cells to cells incubated with T3 (I). Bar graphs represent the absorbance of NADH + H⁺ measurement in MCF7 and PC3 cells, comparing control cells and cells treated with T3 (J). Bar graphs represent NAD + /NADH + H.⁺ ratio in MCF7 and PC3 cells, comparing control cells and cells treated with T3 (K). Results were obtained in at least 3 independent experiments. All bar graphs show sean ± SEM. If applicable, significant differences were assessed via one-way ANOVA or unpaired t-test and presented as specific p-values (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001)

THRα mediates T3-induced mitochondrial Ca2+ and energy metabolism remodeling

Given the structural similarity between rT3 and biologically active T3, we conducted live-cell imaging experiments to examine whether rT3 elicited any effects. Remarkably, we observed no discernible impact on basal mitochondrial Ca²⁺ levels (Fig. 4A), mitochondrial Ca²⁺ uptake (Fig. 4B), or mitochondrial ATP levels (Fig. 4C) in MCF7 cells, suggesting that T3 induces changes via its canonical function. To explore whether expression level changes in IP3R3 after 3 h are mediated through THRα or THRβ, we assessed whether specific siRNAs targeting THRα or THRβ (Fig. 4D) impeded T3-induced increase in IP3R3 after 3 h of T3 treatment. Notably, T3 treatment caused no increase in the mRNA expression level of IP3R3 in cells with THRα KD, whereas the IP3R3 expression level remained significantly elevated after T3 treatment in the case of THRβ KD (Fig. 4E). Interestingly, THRα was found to be the more abundant isoform of THR in MCF7 and PC3 cells (Additional File 2A), pointing to a major role of THRα in these cancer cells. Live-cell imaging in MCF7 cells with siRNA against THRα failed to detect any changes in basal mitochondrial Ca²⁺ levels (Fig. 4F) or an increase in mitochondrial Ca²⁺ uptake upon T3 treatment (Fig. 4G) as present in control cells (Fig. 1C). The KD of THRβ did not alter basal mitochondrial Ca²⁺ levels (Fig. 4H) and did also not prevent the T3-induced increase in mitochondrial Ca²⁺ uptake (Fig. 4I). Additionally, THRα KD prevented a T3-induced elevation in mitochondrial ATP levels (Fig. 4J), while an increase in mitochondrial ATP was detected in cells with THRβ KD (Fig. 4K). Consequently, we assumed that T3-induced rewiring of mitochondrial Ca²⁺ handling is predominantly conveyed via THRα-associated signaling. In addition, we delved into the time-dependent aspect of T3-induced changes on IP3R3. Following 1 h of treatment with 100 nM T3, mRNA expression levels of MCU, UCP2, and IP3R3 remained unchanged (Fig. 4L). Additionally, live-cell imaging failed to detect any shifts in basal mitochondrial Ca²⁺ levels (Fig. 4M), mitochondrial Ca²⁺ uptake (Fig. 4N), or mitochondrial ATP levels (Fig. 4O). These findings suggest that T3-induced alterations in IP3R3 expression levels require more time to manifest and are conveyed via THRα.

Fig. 4

T3 [100 nM] modulates mitochondrial Ca²⁺ levels via THRα in MCF7 cells. Bar graphs represent basal [Ca²⁺]_mito (A), [Ca²⁺]_mito uptake (B), and basal [ATP]_mito (C), comparing control cells to cells incubated with T3 or rT3 for 3 h. Bar graphs represent the mRNA expression ratio of THRα and THRß in cells transfected with siRNA against THRα (light grey) and THRß (dark grey), compared to control cells (set to 100%) (D). Bar graphs represent the mRNA expression ratio of IP3R3 in cells with a specific knockdown of the THRα (dark blue) and THRß (light blue)and treatment with T3 for 3 h, compared to the control cells (set to 100%) (E). Bar graphs represent basal [Ca²⁺]_mito (F) and [Ca²⁺]_mito uptake (G) in THRα knockdown cells, comparing control cells to cells incubated with T3 for 3 h. Bar graphs represent basal [Ca²⁺]_mito (H) and [Ca²⁺]_mito uptake (I) in THRß knockdown cells, comparing control cells to cells incubated with T3 for 3 h. Basal [ATP]_mito levels with siRNA against THRα (J) or THRß (K) are presented, comparing control cells to cells incubated with T3 for 3 h. Bar graphs represent the mRNA expression ratio of MCU, UCP2, and IP3R3 in cells treated with T3 for 1 h compared to the control cells (set to 100%) (L). Bar graphs represent [Ca²⁺]_mito basal (M) levels and [Ca²⁺]_mito uptake (N), comparing control cells to cells incubated with T3 for 1 h. Bar graphs represent basal [ATP]_mito levels, comparing control cells to cells incubated with T3 for 1 h (O). For [Ca.²⁺]_mito measurements, cells were transfected with 4mtD3cpv, and for [ATP]_mito with the mtAT1.03 sensor. Results were obtained in at least 3 independent experiments. All bar graphs show mean ± SEM. If applicable, significant differences were assessed via one-way ANOVA or unpaired t-test and presented as specific p-values (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001)

T3-driven mitochondrial Ca²⁺ enhancement sustains cancer cell viability

Given that ATP is crucial for the viability, proliferation, and migration of cancer cells, we investigated whether MCF7 cells sensitive to T3-induced mitochondrial Ca²⁺ rewiring also exhibited increased proliferation. Thereby, we further analyzed corresponding non-cancerous breast tissue cells, hTERT-HME1 (HME1), as well as PC3 cells, insensitive to T3-induced mitochondrial Ca²⁺ rewiring. Within 5 days of treatment with 100 nM T3, the cell number of MCF7 significantly increased in comparison to MCF7 cells without T3 treatment. In contrast, T3 treatment did not affect the cell number of HME1 or PC3 cells (Fig. 5A). Notably, MCF7 cells reached a doubling time of approximately 38 h, HME1 cells of 55 h, and PC3 cells of 59 h (Additional File 3A). Consequently, doubling times of all cell types should have been sufficient to observe potential proliferative effects within 5 days. Importantly, T3 failed to induce significant changes in the mRNA expression levels of the main proteins associated with mitochondrial Ca²⁺ uptake, including MCU, UCP2, IP3R1, and IP3R3, in HME1 cells (Fig. 5B), as already shown for PC3 cells (Fig. 2C). In MCF7 cells, the KD of IP3R3, MCU, and UCP2 prevented the effect of T3 on the proliferation rate (Fig. 5C), underscoring the critical role of ER-mitochondrial Ca²⁺ flux in MCF7 cancer cell viability and proliferation. Subsequently, we also examined compounds modulating mitochondrial Ca²⁺ uptake, including the MCU inhibitor mitoxantrone and the UCP2 inhibitor genipin. Both compounds prevented the proliferative effect of T3 on MCF7 cells (Fig. 5D), highlighting the potential of these compounds.

Fig. 5

Mitochondrial Ca.²⁺ is essential to the proliferation boost induced by 100 nM T3. Bar graphs represent the normalized cell count of MCF7, HME1, and PC3 cells treated with T3 compared to the corresponding control after five days of T3 incubation, presented as percentages with respective untreated control set to 100% (marked with a line) (A). Bar graphs represent the mRNA expression ratio, shown as a change in percentage, of genes in HME1 cells treated with T3 for 3 h, compared to the corresponding control cells (set to 100%, marked with a line) (B). For MCF7, cell proliferation (± T3) was determined for cells transfected with siRNA against IP3R3, MCU, or UCP2 compared to the absolute control (set to 100%) (C). Bar graphs represent the percentage of MCF7 cells (± T3) treated with 100 nM mitoxantrone and 10 µM genipin (D). Results were obtained in at least 3 independent experiments. All bar graphs show mean ± SEM. If applicable, significant differences were assessed via one-way ANOVA or unpaired t-test and presented as specific p-values (* = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001)

Discussion

In the present study, we demonstrate that T3 upregulates IP3R3 via THRα in MCF7 breast cancer cells, resulting in enhanced mitochondrial Ca²⁺ uptake. The increase in mitochondrial Ca²⁺ uptake stimulates the Krebs cycle, leading to higher electron equivalent yield and elevated mitochondrial ATP production. Importantly, this reprogramming of mitochondrial Ca²⁺ dynamics by T3 is exclusively observed in MCF7 cells, where it promotes viability and proliferation while leaving PC3 prostate cancer cells and non-cancerous HME1 breast cells unaffected. These findings underscore the cell-type-specific sensitivity of breast cancer cells to T3-mediated mitochondrial regulation.

To mimic hyperthyroidism, we used 100 nM T3, a concentration also used in previous studies [3, 31]. We opted against transitioning to T3-depleted fetal calf serum, as our prior investigations demonstrated a concentration of 0.3 nM free T3 in the serum used [3]. This concentration was deemed insufficient to simulate a hyperthyroid state. Since research unveiled a circadian pattern of free T3, which lags behind the peak of TSH, occurring once daily during the early morning hours [32], we administered treatments to our cells only once in all experiments to prevent potential unphysiological interference from repeated treatments.

We previously reported that the incubation with 100 nM T3 for 3 h results in the upregulation of UCP2 in cervix carcinoma cells, HeLa, resulting in enhanced basal mitochondrial Ca²⁺ levels and mitochondrial Ca²⁺ uptake associated with facilitated mitochondrial ATP production [3]. Now, we aimed to investigate whether mitochondrial Ca²⁺ rewiring distinguishes T3-sensitive MCF7 breast cancer cells from T3-insensitive PC3 prostate carcinoma cells. Furthermore, we focused on the downstream signaling cascade leading to T3-induced adaptational mechanisms in mitochondrial Ca²⁺. 100 nM T3 induced an increase in mitochondrial Ca²⁺ uptake (Fig. 1C) via the upregulation of IP3R3 (Fig. 2B) in MCF7, which points to a potentially breast cancer-specific effect of T3. Notably, the basal mRNA expression levels of IP3R3, MCU, UPC2, and PRMT1 (Fig. 2A) were found to be significantly higher in MCF7 breast cancer cells than in PC3 prostate carcinoma cells. These differences in the basic setup of Ca²⁺ handling proteins might point to a specific mode of mitochondrial Ca²⁺ uptake [22], which potentially links breast cancer cells’ susceptibility to T3-induced rewiring of mitochondrial Ca²⁺ and metabolic activity.

Previous reports already presented increased IP3R3 expression levels in different cancer types. IP3R3 was positively correlated with tumor size, node invasion, and histologic grade, and high IP3R3 levels negatively correlated with patient survival [33, 34]. Besides, increased IP3R3 expression was also discussed as a potential biomarker in human pancreatic cancer [35] and colorectal cancer [36], suggesting a crucial role of IP3R3 in the development and progression of several cancer types. A recent study associated the capture of IP3R at ER-mitochondrial contact sites with stimulation of local Ca²⁺ transfer and oxidative metabolism in HEK cells [37]. We showed that the T3-induced mitochondrial Ca²⁺ boost is linked to an increase in the yield of reduction equivalents and mitochondrial ATP (Fig. 3). Thereby, the Ca²⁺ rewiring might be essential to an altered mitochondrial oxidative metabolism central to the malignant phenotype in breast cancer patients, preferentially utilizing lipids to fuel mitochondrial metabolism [15].

The lack of rT3's impact on IP3R3 mRNA expression levels, as well as on mitochondrial Ca²⁺ and ATP levels (Fig. 4A-C), ruled out a nonspecific binding or interference and pointed to a specific cellular response to T3. Thereby, THRα was found to be the critical mediator of IP3R3 upregulation. Notably, a previous study also found that THRα expression predicted a reduced five-year survival in breast cancer gene 1-associated breast cancer cases, while THRβ expression turned out to be a positive prognostic factor [38], pointing to a critical role of THRα in tumor cell progression.

Notably, in former publications, activation of the ERα was speculated to cause T3-induced proliferation of breast cancer cells [9, 10]. Therefore, we investigated the impact of T3 on the mitochondrial Ca²⁺ and ATP in an ERα-negative breast cancer cell line, MDA-MB-468. Notably, we found enhanced basal mitochondrial Ca²⁺ levels (Additional File 4A), increased mitochondrial Ca²⁺ uptake (Additional File 4B), and increased basal mitochondrial ATP levels (Additional File 4C) after T3 treatment, highlighting that T3-induced rewiring of mitochondrial Ca²⁺ and ATP homeostasis in breast cancer cells is independent of ERα.

A different gene expression pattern of T3-sensitive genes upon TH treatment was reported, including a biphasic expression upregulation within 6 h, followed by a drop and a second increase versus a 2-phase increase in the expression with a rise in the first 6 h and after the following 48 h [39]. Based on these findings, the detected upregulation of IP3R3 mRNA and protein expression after 3 h is assumed to have a lasting effect. We found an enhanced proliferation rate of MCF7 cells within a duration of 5 days after a one-time treatment with 100 nM T3, while PC3 prostate cancer cells, as well as non-cancerous HME1 breast cells, remained unaffected (Fig. 5A). Doubling times from our proliferation experiments (Additional File 3A) differed from previously published doubling times of 35 h for MCF7 cells [40], 30 to 40 h for HME1 cells [41, 42], and 33 to 40 h for PC3 cells [43]. While this effect was potentially due to different culturing conditions, measured and published doubling times allow the detection of potential proliferative effects of T3 across all tested cell lines. Interestingly, the KD of the Ca²⁺ handling proteins IP3R3, MCU, and UCP2 abolished the pro-proliferative effect of T3, pointing to an effect strongly dependent on mitochondrial Ca²⁺ rewiring and the subsequent boost in mitochondrial metabolic activity. In line with these findings, also the MCU inhibitor mitoxantrone [44], as well as the UCP2 inhibitor genipin [45], inhibited the proliferative effect of T3 (Fig. 5D). Notably, mitoxantrone has been in long-term use for the treatment of advanced breast cancer as a DNA topoisomerase II inhibitor [46], while genipin was shown to induce apoptosis in different breast cancer cells, including MCF7 cells [47, 48]. Based on the recent findings, these compounds' cytotoxic effects on cancer cells might also be linked to mitochondrial Ca²⁺ alterations. Contrary to Warburg's hypothesis implicating dysfunctional mitochondria in driving glycolysis-dependent cancer cells, research revealed pyruvate kinase M2's pivotal role as a bottleneck in generating metabolic intermediates crucial for cellular building blocks via glycolysis. The resulting limited ATP output from glycolysis causes the indispensability of mitochondria in cellular metabolism [49] and might also be the reason for cancer cells’ vulnerability to mitochondrial Ca²⁺ alterations.

Conclusion

In summary, we show that T3-induced amplification of ER-mitochondrial Ca²⁺ flux energizes the metabolic requirements and fosters the aggressive phenotype of breast cancer cells. Understanding the intricate interplay between T3 signaling, ER-mitochondrial dynamics, and breast cancer metabolism holds promise for the development of targeted therapies aimed at disrupting the metabolic vulnerabilities of cancer cells. Moreover, our findings contribute to the growing body of evidence implicating T3 as a key regulator of cancer metabolism, underscoring its potential as a therapeutic target in breast cancer and beyond.