Remodeling Ca²⁺ dynamics by targeting a promising E-box containing G-quadruplex at ORAI1 promoter in triple-negative breast cancer

Abstract

ORAI1 is an intrinsic component of store-operated calcium entry (SOCE) that strictly regulates Ca²⁺ influx in most non-excitable cells. ORAI1 is overexpressed in a wide variety of cancers, and its signal transduction has been associated with chemotherapy resistance. There is extensive proteomic interaction of ORAI1 with other channels and effectors, resulting in various altered phenotypes. However, the transcription regulation of ORAI1 is not well understood. We have found a putative G-quadruplex (G4) motif, ORAI1-Pu, in the upstream promoter region of the gene, having regulatory functions. High-resolution 3-D NMR structure elucidation suggests that ORAI1-Pu is a stable parallel-stranded G4, having a long 8-nt loop imparting dynamics without affecting the structural stability. The protruded loop further houses an E-box motif that provides a docking site for transcription factors like Zeb1. The G4 structure was also endogenously observed using Chromatin Immunoprecipitation (ChIP) with anti-G4 antibody (BG4) in the MDA-MB-231 cell line overexpressing ORAI1. Ligand-mediated stabilization suggested that the stabilized G4 represses transcription in cancer cell line MDA-MB-231. Downregulation of transcription further led to decreased Ca²⁺ entry by the SOCE pathway, as observed by live-cell Fura-2 Ca²⁺ imaging.

Graphical Abstract.

ORAI1 expression is downregulated in the presence of a stable G4 structure at the predicted ORAI1-Pu motif, which is formed at a regulatory docking site for transcription factors and RNA polymerase II. The G4 structure might act as a dynamic switch, as it is a transient structure and is regulated by various factors within a cell. Ligand-mediated stabilization of ORAI1-Pu via TMPyP4/ BRACO-19 indicated gain-of-function in ORAI1 expression. In MDA-MB-231, we observed greater fold endogenous binding of Zeb1 at the ORAI-Pu motif, compared to BG4, which binds to stable G4 structures. In MDA-MB-231, the ORAI1-Pu locus might be in a partially unfolded state or unfolded state, leading to over-expression of ORAI1. However, further studies can reveal the physiological dynamics of the structure and its functional effects on ORAI expression. Overexpression of ORAI1 increases its transport to the plasma membrane and increases SOCE on interaction with activated STIM at physiological conditions where there is loss of Ca²⁺ in the endoplasmic reticulum (ER) calcium store. A spike in intracellular Ca²⁺ cascades its effect on deregulated cellular activity and enhances various hallmarks of cancer, epithelialmesenchymal transition (EMT), drug resistance, and immune evasion. Etc.

Introduction

Metastasis continues to remain the leading cause of death for cancer patients, accounting for 90% of cases of the disease, despite notable advancements in diagnostic and therapeutic techniques. Among the most crucial regulators of proliferation and metastasis is the ubiquitous second messenger Calcium ions (Ca²⁺).[1] Ca²⁺ intricately orchestrates a myriad of cellular processes pivotal to normal physiological functioning, spanning from cell motility and proliferation to apoptosis and gene transcription.[2] The delicate equilibrium of Ca²⁺ homeostasis is perturbed in a variety of cancer types and models, suggesting its essential role in oncogenic transformation and metastasis.[3] The orchestration of Ca²⁺ signals is facilitated by a wide repertoire of channels, pumps, and exchangers, many of which show notable overexpression in particular cancer phenotypes.[2] [3] [4] [5] [6] The upheaval of the Ca²⁺ signaling system is a critical factor that makes cancer cells resistant to treatment interventions, including new molecular and immune-based strategies.[3] The resistance of chemotherapeutic drugs is significantly impacted by elevated Ca²⁺ levels and a modified state of Ca²⁺ homeostasis.[4] Given this, several Ca²⁺ release-activated Ca²⁺ (CRAC) transporter blockers, including carboxyamidotrizol and RP4010, has undergone clinical studies, either alone or in combination with other therapies, to sensitize tumors to treatment.[7]

Store-operated Ca²⁺ entry (SOCE) stands as an omnipresent and indispensable pathway governing the delicate equilibrium of intracellular Ca²⁺ levels.[8] This intricate mechanism of Ca²⁺ influx finds its canonical mediation through ORAI1, a Ca²⁺ channel localized on the plasma membrane, with its activity finely modulated by the dynamic Ca²⁺ reservoirs of the endoplasmic reticulum (ER). [8] [9] [10] Upon depletion of ER Ca²⁺ stores, a cascade is triggered wherein the ER Ca²⁺ sensor, stromal interacting molecule 1 (STIM1), undergoes oligomerization and translocation to specialized ER-plasma membrane junctions. Here, at these pivotal cellular crossroads, STIM1 engages with ORAI1, orchestrating a localized surge of Ca²⁺ influx that underpins myriad cellular processes.[8] [11] The proliferation and migration of several basal breast cancer cells are well-regulated by ORAI1. Interestingly, the heightened expression of the ORAI1 calcium channel emerges as a hallmark feature within the context of triple-negative breast cancer (TNBC), underscoring its indispensable role in fueling the migratory potential and metastatic dissemination of malignant breast cancer cells. Indeed, targeting ORAI1 holds significant promise as a novel therapeutic strategy for combating cancer metastasis and overcoming chemoresistance.[11] Research endeavors to clarify the molecular mechanisms of ORAI1-mediated signaling in cancer are ongoing; nonetheless, the transcriptional control of this gene is still mysterious.

We aimed to explore the vicinity of the transcription start site (TSS) in the promoter region to understand the transcriptional regulation and detect the existence of druggable targets such as G4, which may be subject to selective regulation in cancer treatment. A putative promoter G4 motif, ORAI1-Pu, was identified through bioinformatic analysis of the ORAI1 promoter. Along the line of discovery, G4 structures are among the most significant non-canonical structures in our genome, which are involved in the regulation of various composite biochemical processes such regulation of replication, transcription & translation. G4s are formed from guanine-rich repeat motifs; their physicochemical characteristics and structural diversity suggest they are effective therapeutic targets and highly selective.[12] They consist of π-π stacked G-quartets with Hoogsteen hydrogen bonding and are stabilized by cations like Na⁺ and K⁺. [12, p. 4] [13] G4 can have various structural conformations depending on their strand orientations such as parallel, anti-parallel and mixed, further variability in the G4 structure can arise by the formation of inter or intra molecular topologies. (Figure S1A.) Guanine-rich repeats have been discovered in several genomic loci, including telomeres, ribosomal DNAs, immunoglobulin heavy chain switch regions, microsatellite and minisatellite repeats, and the promoter regions of various oncogenes. G4s have been identified as critical agents in the complex regulation of gene expression, influencing several genes at the transcriptional and translational levels.[12] [14] [15] [16] Stabilization of these structures offers a promising new area for the development of innovative anti-cancer therapeutics.[17] [18] The nexus between these G4 structures and the six primary hallmarks of cancer has already been established.[19] Meanwhile, metabolic reprogramming has emerged as a critical hallmark of cancer in recent years. But as of yet, no G4 has been reported to be in the promoter of a membrane transporter gene, which is crucial for metabolic homeostasis and signal transmission during oncogenesis.

Since ORAI1 actively contributes to the promotion of drug resistance in cancer cells, further investigation into the ORAI1-Pu G4 motif can emerge as a new therapeutic target for the manipulation of further combination therapies. The ORAI1-Pu G4 sequence had a propensity to form a G4 structure with parallel topology and an uncommon dynamic 8-nt loop, according to an extensive biophysical characterization and structural determination. Additional research on ligand interaction and incellulo investigations revealed the endogenous existence of the G4, which, upon probable ligand-induced stabilization, reduced promoter activity. Due to the presence of the E-box motif, the dynamic loop region may also be a likely docking site for different transcription factors, such as Zeb1,[20, p. 1] c-Myc,[21] etc. Chromatin Immunoprecipitation (ChIP) data validated the recruitment of Zeb1 within the ORAI1-Pu G4 region. It has been found that Zeb1 is a highly expressed oncogene and facilitator of epithelial-to-mesenchymal transition (EMT) in TNBC, and the upregulation of ORAI1 might be a function of binding to E-box presented as a docking site at the ORAI1-Pu G4 region. Finally, live-cell ratiometric Ca²⁺ imaging analysis validated our hypothesis regarding the functional consequences of stabilizing ORAI1-Pu with G4 stabilizing agents, TMPyP4 and BRACO-19. TMPyP4 and BRACO-19 are well-known and commercially available G4 binders. (Figure S1.B.) These are cell-permeable molecules and are used extensively to study G4 structures. [22] [23] [24] Fura-2 AM-based detection of Ca²⁺ uptake revealed a decrease in SOCE activity in both human MDA-MB-231 and mouse B16 cells, indicating that this effect might not be species-specific. Overall, our study reports the first-ever proof-of-concept therapeutic module of G4-based anticancer therapeutics targeting Ca²⁺ dynamics via acting on the ORAI1 promoter.

Materials and methods

Bioinformatics analysis

The promoter region of the ORAI1 gene sequence (NCBI database Gene ID: 84876) was analyzed and endorsed by the online Eukaryotic Promoter Database (EPD).[25] The putative G4 motifs were identified, and their potential score to form a quadruplex was calculated using the QGRS G4 predictor software tool.[26] Further, the Encode database and UCSF browser were studied to analyze the ChIP-Seq elements near the putative G4 sequence ORAI1-Pu with respect to the Transcription Start Site (TSS).[27] [28] The EPD transcription factor motif search tool used the JASPER CORE Database 2018[29] to bioinformatically predict the transcription factors with putative interaction at the ORAI1-Pu G4 motif (p-value cutoff set at < 0.01). The 100 vertebrates’ conservation by PhastCons Alignment tool in the UCSF browser was used to study conversed regions within the predicted sequence.

Sample preparation

All DNA oligonucleotides for structure calculations were purchased from TIB MOLBIOL (Berlin, Germany) and purified by precipitation with potassium acetate and ethanol. The concentration of DNA oligonucleotides was determined by measuring the UV absorbance in water at 260 nm at 80 °C. All DNA oligonucleotides were dissolved in 10 mM potassium phosphate buffer, pH 7. Samples were annealed by heating to 90°C for 5 minutes, followed by slow cooling to room temperature. Final sample concentrations were 5μM for UV and CD measurements and ranged from 0.3 to 0.4 mM for the NMR studies.

The cationic porphyrin derivative 5,10,15,20-tetrakis (1-methyl-4-pyridinio) porphyrin tetra (p-toluenesulfonate) (TMPyP4) was obtained from Sigma-Aldrich (catalog no. 613560). The synthetic acridine analogue BRACO-19 was obtained from Sigma-Aldrich (catalog no. SML0560) and dissolved in ultrapure water (Invitrogen) at a high stock concentration of 20 mM and stored in the dark at 20 °C.

Thermal stability determined by UV melting experiments

UV melting experiments for the G4s were performed using a Jasco V-650 spectrophotometer (Jasco, Tokyo, Japan) equipped with a Peltier thermostat. The absorbance of the oligonucleotides (~5μM) was recorded at 295 nm from 10 to 90 °C at a heating and cooling rate of 0.2 °C/min and a bandwidth of 1nm using quartz cuvettes of 1 cm path length. Melting temperatures T_m were determined from the minimum of the first derivative of the heating curve. All experiments were performed in triplicate.

Topology prediction by CD spectroscopy

CD spectra were acquired using a Jasco J-810 spectropolarimeter (Jasco, Tokyo, Japan) equipped with a thermoelectrically controlled cell holder. CD spectra were obtained by accumulating five scans recorded at a rate of 50 nm/min over a range of 220-320 nm. The bandwidth was 1 nm and the response time was 4 seconds. Oligonucleotides (~5μM) were measured in 1-cm quartz cuvettes at 20 °C in 10 mM potassium phosphate buffer, pH 7. All spectra were blank-corrected by subtraction of the buffer spectrum.

Structural determination by NMR spectroscopy

All NMR spectra were acquired on a Bruker Avance NEO 600 MHz NMR spectrometer equipped with an inverse ¹H/¹³C/¹⁵N/¹⁹F quadruple resonance cryoprobe head and z-field gradients. Oligonucleotides (0.3-0.4 mM) were dissolved in 10 mM potassium phosphate buffer, pH 7.0, with 90% H₂O/10% D₂O. Spectra were either acquired at 25 °C or 40 °C. Topspin 4.0.7 and CcpNmr Analysis 2.4.2 were used for spectral processing and analysis.[30] The chemical shift of protons was indirectly referenced to sodium trimethylsilyl propionate (TSP) through the temperature-dependent water chemical shift at pH 7.0. Chemical shifts of carbon were referenced to sodium trimethylsilylpro-panesulfonate (DSS) through an indirect referencing method. For details on the NMR experiments and acquisition parameters used see the Supporting Information.

NMR structure calculations

Initially, 400 structures were calculated using a simulated annealing protocol in XPLOR-NIH 3.0.3 and 100 lowest-energy starting structures were used for further refinement.[31] The refinement was carried out in vacuum using AMBER18[32] and was followed by simulated annealing to obtain twenty converged structures. For an additional refinement in water, ten lowest-energy conformations were selected. After neutralizing the system by the addition of potassium ions, two additional potassium ions were placed within the inner channel of the G4 core and the system hydrated with TIP3P water.[33] The trajectory of a final simulation of 4 ns at 1 atm and 300 K was averaged over the last 500 ps and minimized in vacuum to obtain the final ten lowest-energy structures. The calculated structures were analyzed by using the VMD 1.9.2 software. Pymol 1.8.4 was used for the three-dimensional structural representations. For a detailed protocol of the simulations including settings for the NMR restraints see the Supporting Information.

Mammalian cell culture

MDA-MB-231 and A549 cell lines were grown in Dulbec-co’s Modified Eagle Medium (DMEM) media (Gibco Catalog No. 11995-065) supplemented with 10% (v/v) FBS (Gibco Cat. No.- 10082147), antibiotics Gentamicin (75 μg/ml), 1% Pen-Strep, and Amphotericin B (0.375 μg/ml). They were maintained and passaged in tissue culture-treated T-25 flasks and all subsequent experiments were done after 2-3 cell passages. Cells were kept under humidified conditions at 37 ° C and 5% CO₂.

Promoter activity by luciferase assay

The upstream promoter region of the human ORAI1 gene with the putative G4-forming sequence, ORAI1-Pu, was amplified by PCR using human genomic DNA as a template and then cloned into the pGL4.72[hRlucCP] vector (Promega; Madison, USA; catalog no. E6901) at the KpnI/HindIII site. The construct, named ORAI1-G4-WT, contained a 300 bp (-259 to +41 from TSS) sequence with the proximally placed ORAI1-Pu G4 sequence. Its mutant variant, named ORAI1-G4-Null, was synthesized by deleting the ORAI1-Pu sequence. Both the constructs were outsourced from GenScript, USA. For the reporter assay, 0.5 X 10⁴ cells were seeded on a 96-well culture plate. After 24 hours, each pGL4.72[hRlucCP] construct (wild-type or mutated, 100 ng) was transiently transfected into cells in Opti-MEM media (HiMedia) with Lipofectamine 2000 (Invitrogen). Each well was also transfected with a pGL3 vector (10 ng, Promega) as a transfection control. After 4 hours of transfection, the reduced serum medium used for transfection was replaced with a fresh complete one, and appropriate doses of TMPyP4 or PBS (for control cells) were added to the cells, which were further incubated for the next 24 hours. On the following day, cells were harvested and lysed with 1× Passive Lysis Buffer (PLB; Promega) to determine firefly and Renilla luciferase activities using a Dual-Luciferase Reporter Assay System (Promega; Catalog number-E1910) in a Luminoskan luminometer (ThermoFisher Scientific). Renilla luciferase activity was normalized with firefly luciferase activity for each sample and considered as relative luciferase activity. The assay was repeated three times, each set as triplicates. The average fold change of wildtype was calculated to the G4 null mutant values and the p-value (by paired t-test) was determined. Graphical representation was done using GraphPad Prism 8.

RNA expression analysis

Cells were grown in 6-well plates to 60% confluency, followed by TMPyP4 (25 μM and 50 μM) and BRACO-19 (25 μM and 50 μM) treatment for 24 hours. The cells were lysed and extracted in a Trizol RNA-stabilizing solution. RNA was extracted subsequently using chloroform and isopropanol following a standard protocol. Estimation of RNA was done using nanodrop. First-strand cDNA synthesis was done with Verso cDNA synthesis kit (ThermoFisher Scientific) followed by a quantitative real-time PCR assay by using PowerUp SYBR Green Master Mix (Applied Biosystems-ThermoFisher Scientific). As a basal constitutive expression control, the 18S gene expression was quantified. The PCR primers used to quantify ORAI1 and 18S rRNA are given in Table S1. The annealing temperature used was 54°C. Quantification expression was done by the fold enrichment method; the equations used are:

$$\begin{array}{c}\text{ΔΔCt}=\text{Ct}(\text{target})-\text{ Ct }(\text{IgG});\\ \text{Fold enrichment}:2^{\wedge }(-\text{ΔΔCt})\end{array}$$

The RNA expression was quantified three times in triplicates. Their statistical significance p-value (by paired t-test) and graphical representation were computed in GraphPad Prism 8.

Chromatin immunoprecipitation (ChIP)

The chromatin immunoprecipitation assay (ChIP) was conducted to monitor the occupancy level of RNA polymerase and various E-box binding oncogenic transcription factors (Myc22, Zeb1, SNAI2) at the ORAI1-Pu promoter region of the ORAI1 gene promoter.[34] We also monitored the binding of BG4, which is an anti-G4 antibody, a probable cellular predictor of G4 formation at ORAI1-Pu. A negative control set with an anti-IgG antibody was also performed to eliminate effects due to nonspecific binding. We have used the ThermoFisher Scientific™ Pierce™ Magnetic ChIP kit (Catalog number # 26157) to perform the ChIP experiment. MDA-MB-231 cells were seeded at a density of 1 × 10⁶ cells per 100 mm culture dish and harvested after 24 hours of incubation in a CO₂ incubator. Cells were processed, and immunoprecipitation was carried forward as per the directed protocol in the Pierce™ magnetic ChIP kit product sheet. The details of the antibodies used for the co-immunoprecipitation and the concentrations used are given in Table S2. We further proceeded to real-time qPCR amplification with the purified ChIP genomic fragments. Real-time qPCR was performed in a QuantStudio 5 (Applied Biosystems) thermal cycler to amplify the collected purified DNA using PowerUp™ SYBR™ Green Master Mix (Catalog number # A25742) as per the manufacturer's protocol. Designed PCR primers encompassing the ORAI1-Pu sequence were used to amplify the ORAI1 promoter region ( Table S1). The thermal cycler was set at the following conditions: initial denaturation step by heating at 95 °C for 10 minutes followed by 45 cycles of 30 seconds of initial denaturation at 95 °C, 30 seconds at 55 °C annealing temperature, and 30 seconds of extension at 72 °C. The experiment was repeated thrice, and the average fold change with respect to IgG (negative control) was considered to calculate statistical significance. Statistical analysis was performed using a paired t-test in GraphPad Prism version 8 software.

Calcium imaging

Calcium imaging was performed as reported earlier [35] [36]. Cells were cultured on confocal dishes (SPL life sciences) to attain 60-80% confluency. Cells were pre-treated with either 50μM TMPyP4/BRACO-19 or vehicle control for 48-72 hours. Cells were then incubated in a culture medium containing 4μM Fura-2AM for 30 minutes at 37°C, 5% CO₂. Post-incubations, cells were washed 3 times and bathed in HEPES-buffered saline solution (2 mM CaCl₂, 1.13 mM MgCl₂, 140 mM NaCl, 10 mM D-glucose, 4.7 mM KCl, and 10 mM HEPES, pH 7.4) for 5 minutes. Further, 3 washes were given and cells were bathed in HEPES-buffered saline solution without 2 mM CaCl₂ to ensure the removal of extracellular Ca²⁺ before starting the measurements. A digital fluorescence imaging system (Nikon Eclipse Ti2 microscope coupled with a high-speed PCO camera) was used and fluorescence images of several cells were recorded and analyzed. Excitation wavelengths of 340 nm and 380 nm were alternately employed for Fura-2AM and the emission signal was recorded at 510 nm.

Results

Identification of a putative G4 motif at the promoter region of ORAI1

A putative G4 motif, ORAI1-Pu, was predicted by the QGRS mapper (G score: 37) at the promoter region of the human ORAI1 gene (NCBI database Gene ID: 84876), approximately 100 bp upstream of the TSS in the sense strand, as denoted by the Eukaryotic Promoter Database (EPD) (Supplementary Fig1). Further, UCSC Genome Browser data on Homo sapiens (GRCh38/hg38) from chromosome 12: 12626399-122626500, encompassing the ORAI1-Pu region, which starts at Ch12: 121626426 position gives an in-depth role of the putative G4-motif in the regulation of transcription of our gene of interest ORAI1. ORAI1-Pu: 5’ TGGGCGGGGCACAGGTGGGCGGGG 3’, was seen to be a region of H3K4me3 methylation site as indicated by ChIP-seq data (Barski et al.),[37] which has a role in RNA polymerase II promoter-proximal pause-release,[38] also, this region is indicated to be a probable binding site to RNA polymerase subunit 1 which has a DNA binding domain. The sequence of our interest, ORAI1-Pu, also incorporates the hexanucleotide consensus E-box motif 5’-CANNTG-3’,[39] [40] making it a potential transelement with an effective role in basic helix-loop-helix-like transcription factor binding.[41] Therefore, ORAI1-Pu presented as a putative G4 forming region, with an indicative role as a promoter regulatory element with a loop region containing an E-box. The G4 motif was further seen to be conserved in primates, and except for variations in the loop regions, it was mostly conserved in other mammals, indicating its evolutionary significance in the regulation of the ORAI1 gene. (Figure S1C.)

The ORAI1-Pu sequence forms a parallel G4 structure

The circular dichroism (CD) spectral signature of the wild-type ORAI1-Pu sequence (5’-TGGGCGGGGCACAGGTGGGCGGGG-3’) features negative and positive amplitudes at around 243 and 263 nm, respectively, typical of a parallel G4 with exclusive homopolar tetrad stacking as a result of three connecting propeller loops (Figure 1A, B). In addition, its imino proton spectral region shows twelve well-resolved Hoogsteen-type guanine imino proton resonances between 11.0 and 12.0 ppm, indicating the formation of a single G4 structure with three G-tetrad layers (Figure 1C top).

Figure 1. Structure determination of the ORAI1-Pu G4.

(A) Schematic representation of the ORAI1-Pu topology with residue numbering; anti-G residues of the G-core are colored grey, loop and overhang residues are represented by circles colored in blue and red, respectively. (B) CD spectrum of the ORAI1-Pu sequence. (C-E) 2D NOESY spectral regions of ORAI1-Pu (mixing time 300 ms). (C) H1-H1 spectral region with corresponding 1D imino proton spectrum shown on top with resonance assignments. (D) H8/H6(ω₂)-H1'(ω₁) spectral region. (E) H8(ω₂)-H1(ω₁) spectral region. (F) 1D ¹⁵N-filtered HMQC spectra of ORAI1-Pu sequences site-specifically labelled at G2, G6, G17, G21, G23, G9, G14, and G15 (10% ¹⁵N enrichment). Assignments of guanine H1 (left) and H8 resonances (right) with G imino and G H8 spectral regions of unlabelled ORAI1-Pu shown on top. NMR spectra were acquired in a 10 mM potassium phosphate buffer, pH 7.0, at 40 °C.

NMR structure determination of the ORAI1-Pu sequence

A detailed structural characterization of the ORAI1-Pu sequence was carried out employing standard NMR methodologies. Full resonance assignments of the folded ORAI1-Pu sequence were obtained by the analysis of NOESY, DQF-COSY, ¹H-¹³C HSQC, and ¹H-¹³C HMBC experiments. In addition, unambiguous identification of G imino and H8 protons was achieved by site-selective incorporation of ¹⁵N-labelled guanine residues (Figure 1C-F, for assignment strategies and additional spectra, see the Supplementary Information and Figure S2-S4, Table S3). NMR-derived distance and torsion angle restraints were used to determine the three-dimensional structure of ORAI1-Pu by restrained molecular dynamics calculations in explicit water. (Table 1) ORAI1-Pu features a parallel topology with a first 1-nt followed by a long 8-nt and another 1-nt propeller loop running in a counter-clockwise direction (Figure 2). Such long propeller loops have previously been reported for the parallel G4 in a mutated c-MYC and KRAS sequence.[42],[43] A superposition of the ten lowest-energy structures shows that the G-core is well-defined. The resulting structures show significant flexibility along the second 8-nt propeller loop (Figure S5). The 5’-terminal residue T1 stacks below G2 and G21 of the 5’-tetrad in agreement with experimental NOE data (Figure 2C). 3’-Terminal G24 stacks onto G23 of the 3’-tetrad (Figure 2D).

Table 1. NMR restraints and structural statistics of calculated structures.

Sequence	ORAI1_Pu
NOE distance restraints
Intra-residual	97
Inter-residual	134
exchangeable	42
Other restraints
Hydrogen bonds	48
Dihedral angle	40
planarity	3
chirality	120
Structural statistics
Pairwise heavy atom RMSD value (Å)
All residues	4.1±1.6
G-tetrad core	0.80±0.2
NOE violations:
Maximum violation (Å)	0.22
Mean NOE violation (Å)	0.0021±0.001
Deviations from idealized geometry
Bond lengths (Å)	0.01±0.0001
Bond angles (degree)	2.2±0.04

Figure 2. Representative structure of the ORAI1-Pu G4.

(A) Top view and (B) side view; anti-G residues are colored grey, loop residues are colored blue and overhang residues are colored red. (C) View onto the 5'-tetrad with stacked T1 and (D) view onto the 3'-tetrad with stacked G24.

Impact of mutations on the stability of the ORAI1-Pu structure

ORAI1-Pu mutants were subjected to a CD and NMR spectral analysis (Figures S6 and S7). All mutated sequences exhibit a signature typical of a parallel fold. Additionally, UV melting experiments were performed on all sequences to assess the effect of substituted nucleobases on the stability of the G4 structures. Initially, each residue within the 8-nt propeller loop containing an E-box motif has individually been replaced by a thymidine (Table S4). Melting temperatures for all these ORAI1 single mutants were found to be ~61 ºC and thus comparable to the melting temperature of the wild-type ORAI1-Pu sequence determined to be 61.4°C. This demonstrates that a single mutation in the long propeller loop does not affect the stability of the ORAI1-Pu structure and suggests the absence of any critical additional interactions of particular loop residues. With all bases in the long propeller loop replaced by thymine in the sequence ORAI1_T8, the melting temperature more noticeably decreased by 5 ºC when compared to the wild-type sequence. This can be attributed to reduced short-lived stacking interactions within the long loop when substituting purine with pyrimidine bases. With 3’-terminal G24 replaced by a thymidine in the ORAI1_G24T sequence, the melting temperature decreased by 3 ºC. Such a destabilization can be attributed to the favorable stacking of G24 onto the 3’-tetrad in wild-type ORAI1 as revealed by its high-resolution structure (Figure 2D). This indicates that G24 plays a more significant role in the stability of the structure. An additive destabilization with a decrease in melting temperature by 8.3 ºC was found for the ORAI1_T8_G24T sequence, exclusively featuring T residues within the long propeller loop as well as a 3’-terminal G24-to-T replacement. Likewise, single and triple G-to-T substitutions in the long propeller loop in addition to a G24-to-T replacement in ORAI1_G9T_G24T and ORAI1_G9G14G15G24T sequences resulted in a decreased melting temperature by 3.8 ºC and 6.6 ºC, respectively. Taken together, 3' flanking G24 contributes to the stability of the ORAI1-Pu quadruplex by stacking interactions with the outer tetrad, whereas the 8-nt sequence motif in the ORAI1-Pu propeller loop results in only rather unspecific and modest stabilizing effects when compared to an all-thymidine loop with the same length.

Binding of ORAI1-Pu motif with various G4 binding ligands-TMPyP4 and BRACO-19

TMPyP4 and BRACO-19 are some common and widely used ligands that show high affinity binding to G4-like structures.[44],[45] Braco-19 is a 3,6,9-trisubstituted acridine derivative, and TMPyP4 is a cationic porphyrin derivative that can end-stack with the G4 tetrads and provide stability to a G4 structure. Both TMPyP4 and BRACO-19 show strong binding in the sub-micromolar range to the ORAI1-Pu motif and K_d values of about 0.1 µM and 0.6 µM were determined by isothermal titration calorimetry (ITC) (Figure S8 and Table S5). Strong binding of BRACO-19 is also reflected in the increased melting temperature (80 ± 2 °C) upon its addition to ORAI1-Pu as observed from the temperature-dependent ellipticity at 263 nm (Figure S9). In addition, stacking of BRACO-19 to an outer tetrad of the G-quadruplex is suggested by the appearance of new, albeit broadened, imino signals upon ligand addition that are upfield-shifted compared to the imino resonances of the free quadruplex (Figure S9). For TMPyP4, binding seems to follow a more complex behavior possibly comprising various binding modes. Thus, a very broad CD melting transition with an increased temperature (70.4 ± 2 °C) for half-dissociation is observed for a TMPyP4 complex. On the other hand, NMR titrations with TMPyP4 show a gradual loss of free quadruplex imino resonances and the appearance of considerably upfield-shifted signals. Such a behavior may indicate exchange between different ligand binding sites but some minor but also major structural rearrangements of the quadruplex cannot be excluded.

Effects of the ORAI1-Pu motif on promoter activity

The effect of ligand binding on the promoter activity was studied by performing a Dual-Luciferase Reporter assay with a wildtype (with ORAI1-Pu motif) and a control mutant (without ORAI1-Pu motif) construct. Increasing concentration of TMPyP4 treatment (5, 10, 20, 40, 60 μM) post-transfection in the wildtype construct showed an initial non-significant change at lower concentrations but at higher concentrations ~30% loss in normalized reporter expression with respect to untreated control. However, for the G4 null mutant construct, there were non-significant changes in reporter activity. In biophysical studies, we have observed that TMPyP4 has a stabilizing effect on the ORAI1-Pu structure. Conceivably, TMPyP4 stabilizes the promoter G4 element, i.e., ORAI1-Pu, and decreases the promoter activity. The absence of the G4 promoter element in the mutated construct, i.e., G4-null, and the non-significant change in the reporter activity, indicate that TMPyP4 does not reduce promoter activity. TMPyP4 may interact with other probable promoter elements within the -300 bp promoter sequence cloned in the pGL4.72[hRlucCP] vector (Figure 3A & 3B). To understand the role of ORAI1-Pu element in the milieu of all genomic elements, the transcription regulation was further examined in cancer cell-line MDA-MB-231 by quantification of ORAI1 mRNA level by q-PCR assay upon treatment with two G4 binding ligand TMPyP4 and BRACO-19. The results corroborated with the promoter activity studies done by luciferase assay as treatment with TMPyP4 and BRACO-19 both decreased expression of the ORAI1 gene. 25μM TMPyP4 and BRACO-19 treatment reduced ORAI1 expression by ~55% and ~45%, respectively, which was further reduced by greater than 90% (p < 0.05) at 50μM (Figure 3C). The decrease observed in the mRNA expression profile is greater than the significant change observed in the luciferase assay since it could not incorporate the effects of all genomic elements within the selected region of promoter cloned within the pGL4.72[hRlucCP] vector.

Figure 3. ORAI1-Pu G4 as a potential promoter element.

(A) The representative schematic diagram for the reporter luciferase construct. The promoter sequence of ORAI1 (249 bp upstream and 41 bp downstream sequence of ORAI1 from TSS) was cloned with or without the wild-type G4 (ORAI1-Pu) scaffold into the Kpn-I and Hind-III restriction sites. Kpn-I and Hind-III restriction sites are upstream of the hRluc gene, and the sequence is cloned at the promoter region involved in the expression of the reporter gene hRluc. The deletion clone without the ORAI1-Pu sequence is called ORAI1-Pu-null. Abbreviations include ampicillin resistance gene (AmpR), Pause (RNA pol pause signal), oriC (origin of replication), hCL1 and hPEST (protein-destabilizing sequences), and hRluc (Renilla luciferase gene). (B) In MDA-MB-231, the dual luciferase assay was run in increasing concentrations of TMPyP4 using the wildtype and ORAI1-Pu-Null constructs. After comparing the activity of firefly luciferase to that of Renilla luciferase, the relative luciferase activity was computed and displayed as a fold change. (C) Representation of effects of G4 ligand. There is stabilization of ORAI1-Pu. q-PCR of mRNA post-treatment and quantitative analysis by ΔΔCt method shows a loss in ORAI1 expression with TMPyP4 (25μM & 50μM) or BRACO-19 (25μM & 50μM) treatment. (* represents p-value < 0.05, ** is p-value <0.01, *** is p-value <0.001, **** is p-value <0.0001).

Endogenous study of the ORAI1-Pu motif and its probable role as cis-regulatory element (CRE)

In order to measure the frequency of the recently discovered G4 formation in chromatin within the cell, we employed the chromatin immunoprecipitation assay (ChIP) with the well-characterized G4 structure-specific antibody (BG4).[46] [47] The PCR primer used encompassed the near promoter region -125 bp to +25 bp of the TSS, which includes our sequence of interest ORAI1-Pu (Figure 4A). Pull-down of the region with BG4, compared to IgG, showed significant fold change, indicating the formation of endogenous G4 at the near promoter site of the ORAI1 gene (Figure 4B). This region also showed a positive interaction with RNA polymerase II on pull-down with RNA polymerase II antibody, evidencing its active role in transcription. The presence of a well-characterized E-box motif 5’-CAGGTG-3’ within the sequence, instigated the investigation of endogenous binding of E-box binding transcription factors at this element. Among the transcription factors (Slug, Myc22, Zeb1) studied, Zeb1 showed the highest proficiency for interaction (Figure 4C). We also studied E2F1, which has been characterized to have a major role as a metabolic regulator[48] and also has a function in the regulation of Zeb1.[49] ORAI1-Pu also encompasses the probable site for E2F1 binding as studied bioinformatically with the EPD transcription factor search tool.[50] ChIP pull-down assay reveals the plausibility for binding; however, it was lower than that of Zeb1. The results represent that the region has a high likelihood of being a regulatory element by interaction with one or more transcription factors that might be interacting with the non-canonical G4 endogenous structure.

Figure 4. Transcriptional landscape of ORAI1 promoter mediated by G4.

(A) Primer design region for ChIP assay encompassing the ORAI1-Pu sequence at the ORAI1 promoter. (B) Endogenous characterization of G4 formation by BG4 at ORAI1-Pu and RNA polymerase II binding affinity by subsequent fold enrichment on chromatin pull-down assay, with IgG as negative control. (C) Representation of the region with E-box motif. Transcriptional regulation of ORAI1-Pu by acting as a docking platform for various transcription factors (SNAI2, Zeb1, E2F1, and c-Myc) as observed by fold enrichment in ChIP assay. IgG was used as a negative control. (* represents p-value < 0.05, ** is p-value <0.01, *** is p-value <0.001, **** is p-value <0.0001).

Decrease in SOCE upon ligand-mediated destabilization of ORAI1-Pu

To examine the role of G4 binding ligands TMPyP4 and BRACO-19 on ORAI1 functioning, we performed classical SOCE measurements. ORAI1 is a well-established store-operated Ca²⁺ entry channel in a variety of cell types, including MDA-MB-231 human breast cancer cells [11] [51]. Therefore, we pre-treated MDA-MB-231 cells with TMPyP4 and BRACO-19 for 72 hours and performed live cell calcium imaging experiments. Cells were stimulated with 2μM Thapsigargin (Tg) in the absence of extracellular Ca²⁺. Tg inhibits ER-localized sarcoplasmic/endoplasmic reticulum Ca²⁺ ATPase (SERCA) pump, thereby releasing ER Ca²⁺ stores. Then, we added 1 mM CaCl₂ in the bath solution and this resulted in Ca²⁺ entry via SOCE channels present on the plasma membrane, i.e., ORAI1. In the first set of experiments, we treated cells with either BRACO-19 or vehicle control nuclease-free water (NFW) and performed live cell imaging. As evident in the traces, BRACO-19 treatment led to a significant decrease in SOCE in MDA-MB-231 cells (Figure 5A). We repeated the experiment in multiple independent runs and the data from hundreds of cells showed about 50% reduction in ORAI1-mediated SOCE (Figure 5B). Next, we performed similar imaging experiments with pre-treatment of TMPyP4. We observed the same trend upon treatment with TMPyP4, i.e., SOCE is clearly decreased upon TMPyP4 treatment (Figure 5C). After performing multiple independent runs, the data from over three hundred cells suggest over 40% abrogation of SOCE (Figure 5D). Taken together, these data demonstrate that modulation of G4 with TMPyP4 and BRACO-19 results in a substantial reduction in ORAI1 functioning in MDA-MB-231.

Figure 5. Downregulation of SOCE activity by G4 binding ligands at ORAI1-Pu.

(A) and (C) Representative Ca²⁺ imaging traces for control nuclease free water i.e. NFW (n=94) and treatment (A) BRACO-19 (n=84) or (C) TMPyP4 (n=75) where ‘’n” denotes the no. of cells in that trace. Cells were stimulated with 2 μM thapsigargin (Tg) in Ca²⁺ free buffer and restored with 1 mM extracellular Ca²⁺. In (B) and (D), the extent of SOCE was calculated from 350 NFW and (B) 229 BRACO-19 or (D) 333 TMPyP4 treated MDA-MB-231 cells, which were imaged from 4 independent experiments (“n = x, y” where “x” denotes a total number of cells imaged and “y” denotes the number of traces recorded). Data presented are mean ± S.E.M. For statistical analysis, an unpaired student’s t-test was performed. Here, **** means p< 0.0001.

To further substantiate the effect of G4 modulation on ORAI1 functioning, we measured SOCE in B16 mouse melanoma cells. We have already reported that ORAI1 constitutes a functional SOCE channel in these cells [52] [53]. We pre-treated B16 cells with TMPyP4 and carried out live cell Ca²⁺ imaging at 48 and 72 hours post-treatment. As observed in MDA-MB-231 cells, TMPyP4 treatment results in a significant decrease in ORAI1-mediated SOCE in B16 cells as well (Figure S10A). The quantitation of SOCE from multiple independent runs in over two hundred cells demonstrates that TMPyP4 treatment time-dependently reduces SOCE in B16 cells with above 40% decrease at 48 hours and around 60% abrogation at 72 hours (Figure S10B). Interestingly, in Ca²⁺ imaging experiments with TMPyP4 and BRACO-19 treatment, the ER Ca²⁺ release does not come back to baseline as quickly as seen in the control NFW condition. Indeed, when we performed longer duration Ca²⁺ imaging experiments with TMPyP4, we observed that in the control vehicle treatment cells Ca²⁺ comes back to baseline within 10 minutes (Figure S11A) while it takes almost 90 minutes in the drug-treated cells (Figure S11B). Importantly, even when we waited for ER Ca²⁺ release to reach back to baseline and then added Ca²⁺ in the bath, the amplitude of SOCE was significantly lower in drug-treated cells in comparison to control vehicle-treated cells (Figure S11C). Collectively, these data elegantly corroborate that the G4 modulation results in a robust decrease in ORAI1 function in B16 cells. Notably, data from human MDA-MB-231 cells and mouse B16 cells reveal that the effect of G4 modulation on ORAI1 function may be similar in various cell types and species. It is rather a conserved phenomenon observed in cell lines originating from different tissues and species.

Another interesting observation from the Ca²⁺ imaging experiments is that the treatment with TMPyP4 and BRACO-19 results in ER Ca²⁺ release, which does not come back to baseline as quickly as seen in the control NFW condition. This suggests that apart from ORAI1, TMPyP4 and BRACO-19 may also be acting on G4 close to other Ca²⁺ handling proteins such as plasma membrane Ca²⁺ ATPase (PMCA) and/or ER Ca²⁺ leak channels. However, future studies would be required to decipher the comprehensive effect of G4 modulators on cellular Ca²⁺ signaling via their action on other channels/transporters.

Discussion

In our study, we found evidence that the ORAI1 gene promoter had the propensity for the formation of regulatory non-canonical G4 DNA structure. The G4 forming motif was evolutionarily conserved and shows twelve well-resolved Hoogsteen-type G imino proton resonances between 11.0 and 12.0 ppm, indicating the formation of a single G4 species with three G-tetrad layers. Determination of the high-resolution G4 structure revealed that it constitutes a parallel-stranded G4 with three loops and exclusive homopolar tetrad stacking. The first loop was 1-nt long, followed by a long 8-nt loop and another 1-nt propeller loop running in a counter-clockwise direction. The presence of the long 8-nt loop is unusual in parallel G4s (Figure 2). However, this region is a highly dynamic single-stranded region and has an E-box-like motif 3’-CAGGTG-5’ that might be a probable binding site for transcription factors (Figure S5).[54] Mutation studies with this sequence revealed that single-loop nucleotide mutations did not change the T_m of the G4 significantly, suggesting the absence of any critical additional interactions at particular loop residues for the stability of the structure (Table S5). TMPyP4 and BRACO-19, two commonly used G4 binders, were used to study the effects of stabilization of the G4 on promoter activity. Both the ligands showed an affinity to bind to the putative G4 and stabilized it, as seen by an increase in thermal stability by CD spectroscopic melting experiments. (Figure S8 and S9) It was found by Dual-Luciferase reporter assay that ORAI1-Pu acts as a transcriptional repressor and treatment with the stabilizing ligand TMPyP4 reinforces its regulatory activity. The treatment of these ligands in breast cancer cell line MDA-MB-231 further corroborated our finding as treatment and possible stabilization of ORAI1-Pu significantly decreased the mRNA levels of the ORAI1 gene, which is commonly overexpressed in TNBC subtypes (Figure 3).[9] [55] However, the exact effects of structural dynamics occurring at this locus are ambiguous and cannot clarify which conformer of the unfolded or partially folded structure regulates transcription. The endogenous presence of the G4 structure was noted by ChIP with a widely used anti-G4 antibody, BG4, as there was significant interaction in the pull-down region, which was amplified with a primer specific to the ORAI1-Pu region. Further ChIP investigation revealed that the region was an RNA polymerase II binding site and possibly provided a docking platform for various transcription factors. The highest propensity of binding was with the E-box binding transcription factor Zeb1 (Figure 4). Zeb1 has been recognized as an oncogene over-expressed in several cancer cell lines, including MDA-MB-231. It has been associated with regulating various genes to initiate EMT in cancer cells. ORAI1 is a metabolic Ca²⁺ signaling regulator and has a major role in cancer progression. Zeb1 docking to the ORAI1-Pu region might give us new insight into the complex regulation of these oncogenic factors. Since there is higher fold binding of Zeb1 compared to BG4 in MDA-MB-231, the G4 structure is probably partially destabilized resulting in overexpression of ORAI1. However, further experiments need to be carried out to understand the partial folding dynamics of the G4 at this site. ORAI1 mRNA deregulation by TMPyP4/BRACO-19 ligands led to the downstream effects resulting in the abrogation of store-operated calcium entry by ~50% (Figure 5). This observation was evident from data in both human MDA-MB-231 cells and mouse B16 cells (Figure S10), suggesting that the effect of G4 modulation on ORAI1 function is consistent across various cell types or species. As the ORAI1-Pu sequence is conserved in primates and nearly conserved in mammals, it might be a conserved phenomenon observed in cell lines originating from different tissues and species.

Conclusion

ORAI1 is a plasma membrane Ca²⁺ channel that mediates SOCE and thereby regulates a wide array of cellular functions.[10] The Ca²⁺ homeostasis that is maintained by ORAI1 is in a complex signaling cascade involving a lot of other cellular effectors.[56] ORAI1 might mediate its function through store-dependent or independent pathways by interaction with other factors overexpressed in cancer, like SPCA2.[57] Although much has been studied about the ORAI1-mediated pathways and its subsequent role in cancer,[9] the transcription regulation of ORAI1 is still not extensively characterized. The presence of the ORAI1-Pu motif and the propensity of regulation of the gene through this element directs us to some insights into the cisregulatory element of ORAI1 (Figure 6). The structure of the non–canonical G4 at ORAI1 promoter can be studied further and presented as a druggable target to regulate the transcription of ORAI1.[18] In our study, we have elucidated the role of G4 using one cancer sub-phenotype (TNBC) of breast cancer using MDA-MB-231 cells. G4s are transient structures, and stabilization/destabilization might be regulated temporally within a cell and spatially within various tissues and subtypes depending on physiological factors.[14] Therefore, further study of the G4 present at the regulatory element of ORAI1 can provide us with insights into the regulation pattern of this transporter gene in various alternate signaling pathways. In summary, our study marks a significant milestone by introducing the pioneering proof-of-concept therapeutic framework of G4-based anticancer treatments, specifically tailored to modulate ORAI1-mediated Ca²⁺ dynamics.

Figure 6. Overview/TOC.

ORAI1 expression is downregulated in the presence of a stable G4 structure at the predicted ORAI1-Pu motif, which is formed at a regulatory docking site for transcription factors and RNA polymerase II. The G4 structure might act as a dynamic switch, as it is a transient structure and is regulated by various factors within a cell. Ligand-mediated stabilization of ORAI1-Pu via TMPyP4/ BRACO-19 indicated gain-of-function in ORAI1 expression. In MDA-MB-231, we observed greater fold endogenous binding of Zeb1 at the ORAI-Pu motif, compared to BG4, which binds to stable G4 structures. In MDA-MB-231, the ORAI1-Pu locus might be in a partially unfolded state or unfolded state, leading to over-expression of ORAI1. However, further studies can reveal the physiological dynamics of the structure and its functional effects on ORAI expression. Overexpression of ORAI1 increases its transport to the plasma membrane and increases SOCE on interaction with activated STIM, at physiological conditions, where there is a loss of Ca²⁺ in the endoplasmic reticulum (ER) calcium store. A spike in intracellular Ca²⁺ cascades its effect on deregulated cellular activity and enhances various hallmarks of cancer, epithelial-mesenchymal transition (EMT), drug resistance, and immune evasion.

Supplementary Material

Supplementary information is available online.

Highlights.

• G4 motif regulating the expression profile of the ORAI1 transporter.

• Structural analysis revealed a parallel G4 structure with high thermal stability.

• Stable G4 formation alters Store Operated Calcium Entry in TNBC.