H19 Is a PERK-Regulated Long Non-Coding RNA That Fine-Tunes UPR Signalling and Inhibits Endoplasmic Reticulum Stress-Induced Cell Death

TL;DR

The lncRNA H19 is regulated by PERK during ER stress and helps control cell survival by modulating UPR signaling.

Contribution

H19 is identified as a novel PERK-regulated lncRNA that modulates UPR signaling and cell fate during ER stress.

Findings

H19 expression is downregulated during ER stress in a PERK-dependent manner.

H19 overexpression enhances ATF6 and PERK signaling while reducing IRE1-XBP1 activity.

High H19 levels correlate with poor prognosis in basal-like breast cancer.

Abstract

The endoplasmic reticulum (ER) responds to stimuli that disrupts its homeostasis by activating a signalling network known as unfolded protein response (UPR), that restores cellular balance and determines cell fate through three key sensors: inositol-requiring enzyme 1α (IRE1α), activating transcription factor 6 (ATF6), and protein kinase RNA-like ER kinase (PERK). Emerging evidence suggests that UPR regulates the expression of numerous long non-coding RNAs (lncRNAs), which play critical roles in maintaining ER homeostasis. Here we show that expression of lncRNA H19 is downregulated in response to ER stress in (MCF7, T47D and 293T) cells. Using genetic and pharmacological approaches, we demonstrate that H19 downregulation is primarily mediated by the PERK arm of the UPR. Specifically, knockdown or chemical inhibition of PERK compromised the ER stress-mediated H19 repression, while PERK activation significantly reduced H19 expression. H19 overexpression promotes the optimal activation of ATF6 and PERK pathways, while it attenuates the signalling by IRE1-XBP1 axis of the UPR. Furthermore, in triple-negative breast cancer (TNBC) cells MDA-MB-231, ectopic H19 provided resistance to ER stress-induced apoptosis. Bioinformatic analyses across multiple breast cancer cohorts revealed that high H19 expression was associated with poor prognosis, particularly in basal-like subtypes. Collectively, our findings show that H19 is downregulated during UPR in a PERK-dependent manner, where H19 in turn modulates UPR signalling and cell fate during conditions of ER stress.

Figures

Full text

1. Introduction

Cells constantly face changes in their microenvironment, such as fluctuations in nutrient availability, oxygen levels, or the accumulation of reactive metabolites, which can disturb protein folding within the endoplasmic reticulum (ER) [1,2]. Accumulation of misfolded proteins in the ER, collectively referred to as ER stress, triggers the unfolded protein response (UPR), an evolutionarily conserved signalling pathway that maintains ER homeostasis. The UPR is mediated by three main ER transmembrane sensors—inositol-requiring enzyme 1α (IRE1α), activating transcription factor 6 (ATF6), and protein kinase RNA-like ER kinase (PERK)—whose activities are normally inhibited through binding to the ER chaperone GRP78/BiP [3]. During conditions of ER stress, GRP78 dissociates from UPR sensors, leading to their activation. Activated PERK phosphorylates eIF2α to reduce global protein synthesis while promoting the translation of ATF4, which upregulates genes for amino acid metabolism, redox balance, and chaperones [4]. Activated IRE1α coordinates unconventional splicing of XBP1 mRNA to produce a spliced XBP1 (XBP1s) which encodes for a transcription factor that upregulates the genes for protein folding, ER-associated degradation, and lipid biosynthesis [5]. During conditions of ER stress ATF6 translocates to the Golgi where it is proteolytically processed to release ATF6p50, which acts as a transcription factor to enhance the expression of ER-resident chaperones [6]. However, under prolonged or severe ER stress, PERK-mediated ATF4 upregulates genes involved in redox homeostasis, including the pro-apoptotic transcription factor C/EBP homologous protein (CHOP) [7,8]. Sustained PERK activation also increases reactive oxygen species, exacerbating cellular damage [9]. Meanwhile, IRE1 can shift from XBP1 splicing to regulated IRE1-dependent decay (RIDD), reducing ER protein load but potentially causing loss of critical proteins and cell death [10]. Together, these mechanisms illustrate UPR’s dual role: promoting survival under mild stress but inducing apoptosis under severe or prolonged stress to maintain tissue homeostasis.

Emerging evidence links long non-coding RNAs (lncRNAs) to UPR signalling, and during the past few years, work from several groups has revealed that all three branches of the UPR regulate specific subsets of ncRNAs. For example, CHOP regulates the expression of Lnc-MGC and lncRNA golgin A2 pseudogene 10 (GOLGA2P10); MALAT1 expression is regulated by IRE1 and PERK pathways [11,12]. The UPR-regulated lncRNAs fine-tune the UPR signalling to modulate cellular adaptation to stress and the regulation of cell fate [13,14,15]. H19 is a highly conserved lncRNA that is highly expressed during embryogenesis but largely silenced in adult tissues [16,17]. Reactivation of H19 expression has been observed in multiple cancers, where it influences cell survival and apoptosis in a context-dependent manner [18,19,20,21,22]. Mechanistically, H19 regulates gene expression through multiple pathways, including acting as a competing endogenous RNA (e.g., miR-200, miR-138, and miR-29a) [23,24,25,26,27,28]; serving as a precursor for miR-675 [18,29,30,31]; and interacting with epigenetic regulators (e.g., EZH2, MBD1) [32,33,34]. H19 also contributes to therapy resistance by inhibiting apoptosis, promoting autophagy, and modulating stress-adaptive signalling pathways such as PI3K/AKT, EMT, NF-κB, and mTOR [35,36,37,38,39,40,41,42,43,44,45]. H19 represents a particularly attractive candidate, as it is reactivated in multiple cancers and has been shown to regulate apoptosis, autophagy, and therapy resistance—hallmark stress-adaptive processes closely regulated by the UPR [3,7,24,36]. However, whether UPR activation directly regulates H19 expression has not been investigated.

In this study, we investigated the role of H19 under conditions of ER stress and UPR. We found that H19 expression is downregulated in a PERK-dependent manner under the condition of ER stress, and that ectopic expression of H19 fine-tunes the activation of UPR sensors. Higher H19 expression levels were associated with poor overall survival (OS) in basal-like breast cancer. Functionally, H19 inhibited ER stress-induced cell death, suggesting that stress-induced downregulation of H19 in the tumour microenvironment may contribute to cancer progression by modulating UPR signalling and cell fate.

2. Results

2.1. Downregulation of H19 During Conditions of ER Stress

Considering XBP1 is an integral arm of UPR, to identify UPR-regulated ncRNAs, transcriptomic data from control and XBP1-targeting shRNA-expressing T47D cells (GSE49955, GEO database) were analyzed. From the differentially expressed genes upon knockdown of XBP1, we found that the XBP1 mRNA was the most downregulated gene among all the repressed genes, verifying the knockdown efficiency. Furthermore, H19 was among the top downregulated ncRNA in XBP1-targeting shRNA-expressing T47D cells. To experimentally determine whether UPR regulates the expression of H19, we treated breast cancer cell lines MCF7 and T47D with classical UPR inducers: brefeldin A (BFA) and thapsigargin (TG). BFA is an inhibitor of anterograde transport from the ER to the Golgi apparatus and TG is an inhibitor of sarcoplasmic/endoplasmic reticulum calcium ATPase [46,47]. Both agents disrupt protein homeostasis by promoting the accumulation of unfolded or misfolded proteins in the ER, thereby triggering ER stress and activating the UPR [48]. After 24 h of treatment with BFA or TG, expression of GRP78—a canonical UPR target gene—was significantly upregulated, confirming the successful activation of UPR. In parallel, H19 expression was significantly repressed in both MCF7 and T47D cells (Figure 1A,B). A time-course analysis further revealed a progressive decrease in H19 expression following BFA treatment in MCF7 cells (Figure 1C). To determine whether this repression of H19 was restricted to breast cancer cells, we examined H19 expression in HEK293T cells upon treatment with BFA. Consistent with our earlier observations, BFA treatment induced GRP78 expression and caused a time-dependent reduction in H19 levels (Figure 1D). Together, these results demonstrate that H19 is consistently downregulated in response to ER stress across multiple cell types and in response to different UPR-inducing agents.

2.2. Downregulation of H19 Expression Is PERK-Dependent

Next, we investigated the role of the UPR signalling pathways—PERK, IRE1, and ATF6—in the downregulation of H19 expression. Next, we used MCF7 control (MCF7 PLKO) and UPR signalling-deficient MCF7 subclones (MCF7-XBP1 KD, MCF7-PERK KD and MCF7-ATF6 KD). The generation and characterization of these subclones has been previously reported [49,50]. The expression of PERK, XBP1, and ATF6 protein were significantly reduced (basal and BFA-treated) in respective knockdown subclones, confirming the knockdown of the target protein (Figure 2A). Further, the quantification of the bands in Figure 2A shows comparable induction of XBP1s in PLKO and PERK-KD samples and a slight reduction in XBP1s in ATF6-KD cells. Our results suggest a crosstalk between UPR arms, as ATF6 knockdown led to a reduction in XBP1s expression (Figure 2A), which is consistent with previous reports showing that ATF6 regulates XBP1 [51]. Notably, RT-qPCR analysis revealed that PERK knockdown partially compromised the downregulation of H19 during UPR, while knockdown of ATF6 and XBP1 had no significant effect (Figure 2B).

Next, we used chemical inhibitors targeting each UPR arm. MCF7 cells were treated with either a vehicle control (DMSO) or BFA in combination with ATF6 inhibitor (Ceapin A7), IRE1α inhibitor (STF083010), or PERK inhibitor (GSK2606414). Consistently, RT-qPCR analysis demonstrated that chemical inhibition of PERK significantly compromised the downregulation of H19 and upregulation of CHOP upon BFA treatment of MCF7, but IRE1α inhibitor and PERK inhibitor had no effect on the repression of H19 (Figure 2C). It was surprising to observe the comparable induction of GRP78 in the presence of Ceapin A7 and BFA alone (Figure 2C) given the fact that GRP78 is a direct transcriptional target of processed ATF6. This suggests that activation of IRE1 and PERK pathways is sufficient to upregulate GRP78 expression in presence of Ceapin A7.

Next, we utilized CCT020312, a selective activator of the PERK pathway, to further establish its role in H19 repression. MCF7 cells were treated with CCT020312, and BFA was used as positive control. First, we assessed the effect of CCT020312 on the induction of CHOP, a downstream target of PERK. As expected, CCT020312 treatment led to a significant upregulation of CHOP, confirming the specific activation of the PERK pathway (Figure 2D). Following PERK activation, we observed a significant repression of H19 in both CCT020312- and BFA-treated cells as compared to vehicle-treated controls (Figure 2D). Collectively, these findings suggest that H19 repression during UPR is PERK-dependent.

2.3. H19 Fine-Tunes the Balance Between UPR Branches

Next, we determined the effect of H19 on activation of three branches of the UPR using reporter plasmids. First, we validated the H19 overexpressing plasmid. HEK293T cells were transiently transfected with either a control vector or an H19-overexpressing plasmid. After 48 h, RT-qPCR analysis showed a significant reduction in the Ct value for H19 in HEK293T cells transfected with the H19-overexpressing plasmid as compared to the control, indicating robust H19 overexpression (Figure 3A). In the same samples the Ct values for the housekeeping gene RPLP0 were comparable (Figure 3A).

To evaluate the impact of H19 on the ATF6 signalling pathway, we used a synthetic luciferase reporter construct with five tandem ATF6-binding elements (Figure 3B) [52]. Luciferase reporter assays showed a significant increase in ATF6 transcriptional activity following TG treatment as compared to DMSO controls, confirming the responsiveness of the ATF6 reporter to ER stress (Figure 3C). Importantly, H19 overexpression further enhanced TG-induced ATF6 reporter activity (Figure 3C), indicating that H19 potentiates ATF6 signalling. To assess the role of H19 in the IRE1-XBP1 signalling, we used an IRE1 activity reporter comprising 410–633 nucleotides sequences of the XBP1 cDNA, having a 26 bp intron that is spliced by RNase activity of activated IRE1 [53]. The splicing of the 26 bp intron leads to a change in reading frame that leads to the translation of HA-tagged mNeonGreen protein fused with the c-myc NLS sequence, allowing the detection of IRE1 pathway activity by quantifying GFP fluorescence and/or Western blot for HA-tagged mNeonGreen protein [53] (Figure 3D). HEK293T cells were co-transfected with IRE1-XBP1 reporter along with either control or H19-overexpressing plasmid. At 16 h post-transfection, cells were treated with TG for 8 or 24 h. Compared to control cells, H19-overexpressing cells exhibited a marked reduction in GFP fluorescence at both time points (Figure 3E). Consistently, Western blot analysis revealed decreased levels of HA-tagged mNeonGreen protein in H19-overexpressing cells (Figure 3F). These results suggest that H19 suppresses the activation of IRE1–XBP1 pathway during ER stress. To examine H19’s effect on the PERK-ATF4 signalling branch we used the PERK activity reporter comprising ATF4 5′ UTR sequence, where ORF1 and ORF2 are used as translation initiation in non-stressful conditions [53]. The phosphorylation of eIF2α by activated PERK stimulates translation initiation at ORF3 during conditions of ER stress. In the PERK activity reporter, the fusion of HA-tagged-mScarlet-I and c-myc NLS coding sequence in frame with the first 84 nucleotides of ORF3 enables the detection of ATF4 translation by quantifying RFP fluorescence and/or Western blot for HA-tagged-mScarlet-I protein [53] (Figure 3G). HEK293T cells were co-transfected with the PERK activity reporter along with either a control or H19-overexpressing plasmid. Following treatment with TG for 8 or 24 h, H19-overexpressing cells displayed increased RFP fluorescence intensity as compared to control cells (Figure 3H). This observation was corroborated by Western blot analysis, which showed elevated levels of HA-tagged ATF4 protein in H19-overexpressing cells (Figure 3I). Together, these findings suggest that H19 enhances activation of the PERK–ATF4 pathway during ER stress.

Next, we generated H19-overexpressing stable clones by transducing HEK293T cells with either lentivirus expressing H19 and their corresponding control followed by puromycin selection. RT-qPCR analysis confirmed H19 overexpression in 293T-H19 subclone (Figure 4A). Next, we assessed both protein and mRNA levels of downstream UPR targets in 293T-H19 and 293T-Ctrl subclones under ER stress conditions. Western blot analysis revealed that the induction of XBP1s was attenuated in 293T-H19 cells as compared to 293T-Ctrl cells, indicating suppression of the IRE1-XBP1 axis of the UPR (Figure 4B). During UPR conditions, full length ATF6 (90 KDa) is transported to the Golgi apparatus to be processed by site 1 and site 2 proteases into cytosolic ATF6 (50 KDa) fragment, as such the decrease in the full length ATF6 protein can be used as a surrogate for its proteolytic processing. We observed that the full-length ATF6p90 level was reduced in both 293T-H19 and 293T-Ctrl cells at 4 hr of TG treatment, indicating proteolytic cleavage and activation of ATF6. However, the extent of full-length ATF6 (90 KDa) was significantly reduced in 293T-H19 at 8 hr of TG treatment as compared to 293T-Ctrl cells, suggesting enhanced activation of ATF6 (Figure 4B). The basal and UPR-induced expression of phosphorylated eIF2α (P-eIF2α) and PERK were significantly higher 293T-H19 compared to the 293T-Ctrl cells (Figure 4B). Next, we examined the levels of UPR target genes (CHOP, HERP, GRP78, DNAJB9 and SEC24D) to determine if H19 modulated their expression. RT-qPCR analysis revealed differential regulation of UPR target genes in response to TG-induced ER stress in 293T-H19 and 293T-Ctrl cells. The expressions of all these genes (CHOP, HERP, GRP78, DNAJB9 and SEC24D) were upregulated in both 293T-H19 cells and 293T-Ctrl cells. The induction of the ATF6 downstream genes (GRP78 and HERP) and the PERK downstream gene (CHOP) was significantly higher in 293T-H19 as compared to 293T-Ctrl cells, whereas induction of the XBP1 downstream gene (DNAJB9) was markedly attenuated in H19-overexpressing clones (Figure 4C).

Importantly, these findings in H19 stable clones mirrored the regulatory effects of H19 observed in the reporter assays. Taken together, these results demonstrate that H19 plays a specific role in modulating ER stress signalling, enhancing ATF6 and PERK activation while repressing IRE1 activity during ER stress.

2.4. H19 Promotes Cell Viability and Cell Survival Under ER Stress

Next, we determined the effect of H19 on the regulation of cell fate under ER stress conditions. For this 293T-H19 and 293T-Ctrl cells were exposed to a wide range of BFA and TG and the percentage of viable cells was measured at 48 h. These experiments revealed that 293T-H19 subclones exhibited significantly higher viability compared to 293T-Ctrl under ER stress (Figure 5A,B), suggesting a protective role for H19. We then directly assessed cell death via flow cytometry following 48 h treatment with BFA and TG, H19-overexpressing clones showed significantly lower levels of cell death as compared to control clones (Figure 5C,D), further supporting the notion that H19 enhances cellular resistance to ER stress-induced cell death.

We reasoned that increased expression of H19 may promote cancer progression by enabling cells to better adapt and survive the stressful conditions within the tumour microenvironment, which can activate UPR and trigger cell death. To further evaluate the effect of H19 on cell fate under conditions of ER stress, we generated stable H19-overexpressing subclones and the corresponding control in the triple-negative breast cancer (TNBC) cell line MDA-MB-231. RT-qPCR analysis confirmed a significant upregulation of H19 expression in H19-overexpressing cells (MDA-H19) as compared to control (MDA-Ctrl) cells (Figure 5E). We then evaluated stress-induced cell viability and cell death in MDA-H19 and MDA-Ctrl cells. These experiments revealed that H19 subclones exhibited significantly higher viability as compared to 293T-Ctrl under ER stress (Figure 5F). Flow cytometry analysis showed that MDA-H19 cells exhibited significantly reduced levels of cell death as compared to MDA-Ctrl cells (Figure 5G,H). Together, these data suggest that H19 overexpression confers resistance to ER stress-induced cell death, most likely through modulating the UPR pathways.

2.5. H19 Expression Is Associated with Poor Outcome in TNBC

Breast is one of the few organs in which H19 is not completely repressed after birth [54,55]. H19 is frequently upregulated in breast cancer tissues, and contributes to tumorigenesis through diverse molecular mechanisms, including epigenetic regulation, microRNA sponging, and modulation of signalling pathways [56,57,58]. H19 has a multifaceted role in the progression of breast cancer, particularly in aggressive subtypes such as TNBC [59]. Bioinformatic analysis revealed that H19 expression levels vary significantly among different molecular subtypes of breast cancer (Figure 6A). Univariate Cox regression analyses across various datasets (TCGA and METABRIC) demonstrated that elevated H19 expression was associated with poor overall survival (OS), specifically in basal-like breast cancer (Figure 6B–E). In the METABRIC dataset, high H19 expression in basal-like patients was linked to poorer survival, with a hazard ratio (HR) of 1.97 (95% CI: 1.29–3.00; p = 0.0017) (Figure 6B,D). Similarly, TCGA data showed an HR of 3.26 (95% CI: 1.19–8.97; p = 0.022) for high H19 expression in basal-like patients (Figure 6C,E). Surprisingly, high H19 expression was associated with better OS in Luminal B subtype in METABRIC cohort and Luminal A subtype in TCGA cohort (Figure 6A,B). The significance of these results is not clear as H19 is an estrogen inducible gene and high expression of H19 is associated with endocrine resistance [60]. Nevertheless, increased H19 expression was associated with poor OS in basal-like breast cancer across multiple independent cohorts.

3. Discussion

The relationship between H19 and the UPR has not been investigated in breast cancer, and our study provides new insights into this unexplored link. In this study, we show that H19 expression is significantly reduced under ER stress in a PERK-dependent manner (Figure 1 and Figure 2). In agreement with our results, it was reported that high-glucose conditions in retinal epithelial cells induced ER stress, which was accompanied by repression of H19 expression [61]. H19 serves as the precursor of miR-675-5p and miR-675-3p, with miR-675-5p being the predominant functional strand (Supplementary Figure S1A,B). Under ER stress, miR-675-5p levels decreased in parallel with H19 (Supplementary Figure S1C), suggesting coordinated regulation. While H19 downregulation is PERK-dependent, the ectopic expression of canonical PERK effectors ATF4, NRF2, and CHOP had no effect on H19 expression, implying that additional, non-canonical PERK-responsive effectors are involved (Supplementary Figure S2). A possible mechanism for H19 repression is ‘squelching effect’ where a regulatory molecule sequesters crucial transcription factors, making them unavailable to regulate the promoters of their target genes [62]. Indeed, activation of a widespread PERK-dependent repression programme, consisting of ER target genes has been reported [63]. Furthermore, we have previously reported PERK-dependent downregulation of miR-106b-25 and miR-17-92 cluster during UPR by indirect transcriptional repression [64,65]. TP53 has been shown to downregulate the expression of up to 415 cell cycle-associated genes through indirect transcriptional repression via p53-DREAM pathway [66]. These observations highlight a preliminary mechanistic insight into how ER stress and UPR may regulate H19 but also underscore the need for more detailed investigation into the pathways involved.

Having established that H19 is downregulated under ER stress, a key question arises: Does H19 influence the ER stress response itself? Feedback regulation is a common feature of lncRNAs, allowing them to fine-tune the pathways that control their expression. Given the central role of the UPR in maintaining ER homeostasis and determining cell fate under ER stress, it is likely that H19 may modulate the activation of one or more UPR branches. In our study, we found that under conditions of ER stress, H19 enhanced the signalling via ATF6 and PERK arms while attenuating the IRE1 arm of the UPR (Figure 3 and Figure 4). Such reciprocal regulation between H19 and UPR signalling is consistent with the dynamic and self-limiting nature of the stress response. Once activated, the UPR strives to restore ER homeostasis while preventing excessive or prolonged signalling that could trigger apoptosis [67]. H19 may differentially impact the signalling via three UPR sensors through distinct mechanisms. H19 is reported to sponge several ER stress related miRNAs such as members of miR-17-92 cluster and miR-106b-25 cluster which can regulate UPR signalling [61,68,69]. Indeed, we have shown that miR-17-92 cluster and miR-106b-25 cluster are repressed during UPR and regulate ER stress-induced apoptosis by targeting BIM [64,65]. Further work is required to identify the downstream effectors of H19 in modulating UPR signalling.

All three sensors of UPR can initiate both protective and proapoptotic signalling during conditions of ER stress. Regulation of cell fate by IRE1-XBP1 axis depends on the balance between its adaptive signalling (production of spliced XBP1) and pro-death signalling (JNK activation and RIDD) [70]. Likewise PERK signalling can confer both protective and proapoptotic effects in response to ER stress. For example, PERK and ATF4 knockout MEFs are hypersensitive to ER stress-induced apoptosis [4,71,72]. Moreover, PERK and ATF4 knockout MEFs fail to induce CHOP during ER stress suggesting that CHOP is not an obligatory requirement for ER stress-induced cell death [4,71,72]. Furthermore, sustained PERK signalling has been shown to impair cell proliferation and promote apoptosis [73]. While the primary role of ATF6 is to restore protein homeostasis and promote cell survival, ATF6 can trigger proapoptotic pathways during prolonged conditions of stress. Pharmacological activation of ATF6 is protective in cardiac, renal and cerebral ischemia/reperfusion models [74]. In the model of chronic pancreatitis, ATF6 has been shown to promote apoptosis of acinar cells by regulating p53 expression [75]. Enhancement of PERK and ATF6 signalling could promote adaptive transcriptional programmes that support protein folding and cell survival [76,77,78], whereas attenuation of the IRE1–XBP1 pathway may help restrain excessive mRNA degradation or pro-apoptotic output [79].

H19 overexpression improved cell survival under ER stress, supporting its role as a pro-adaptive factor that limits terminal UPR activation and apoptosis (Figure 5). Indeed, H19 has been reported to inhibit ER stress-induced apoptosis in different model systems [80]. H19 inhibited ER stress-induced cardiomyocyte apoptosis, which was associated with the activation of PI3K/AKT/mTOR signalling pathway [43]. Knockdown of H19 leads to increased sensitivity of cancer cells to resveratrol, a known inducer of ER stress [81]. H19 inhibition was reported to increase VDAC1 and enhance ER-mitochondria coupling that regulates the exchange of Ca^2+^ at contact sites and determines cell fate [82]. Our bioinformatic analysis and experimental results reveal a pivotal role for H19 in regulating cell fate during the UPR in TNBC cells. While the prognostic impact of H19 was most pronounced in TNBC, we also noted that in the HER2-enriched (HER2-E) subtype, high H19 expression was associated with a higher HR, although this did not reach statistical significance. This trend suggests that H19 may have potential context-dependent relevance beyond TNBC. Notably, previous studies have reported a role for H19 in HER2-E breast cancer, particularly in association with trastuzumab resistance [83]. Therefore, although our current data do not support a statistically significant prognostic role for H19 in the HER2-E subtype, the observed trend may reflect underlying biological heterogeneity or limited statistical power and warrants further investigation.

Based on these findings, we speculate that downregulation of H19 due to ER stress caused by stressful conditions of tumour microenvironment may influence TNBC cell growth and survival through modulation of UPR signalling (Figure 7). Given its ability to fine-tune ER stress-induced apoptosis, targeting the UPR–H19 axis may represent a novel therapeutic strategy. Indeed, approaches aimed at inhibiting cancer-promoting lncRNAs, such as MALAT1 and XIST, in breast cancer have already been reported [84,85]. Disrupting the UPR–H19 interaction could sensitize cancer cells to ER stress-induced apoptosis, potentially limiting tumour growth and overcoming resistance to conventional therapies.

4. Materials and Methods

4.1. Cell Culture and Drug Treatment

Breast cancer cell lines representing different subtypes—MCF7, T47D, BT474, and MDA-MB-231—were obtained from ECACC (Salisbury, Sussex, UK). HEK293T cells were sourced from the Indiana University National Gene Vector Biorepository (Indianapolis, IN, USA). MCF7, BT474, MDA-MB-231, and HEK293T cells were cultured in DMEM supplemented with 10% fetal calf serum, 100 U/mL penicillin, and 100 μg/mL streptomycin at 37 °C with 5% CO_2_. T47D cells were maintained in RPMI-1640 medium. To induce ER stress, cells were treated with brefeldin A (BFA; Tocris, Bristol, UK; Cat #1231) or thapsigargin (TG; Tocris, Bristol, UK; Cat #1138) at the specified concentrations and time points. To inhibit specific UPR arms, cells were treated with Ceapin A7 (ATF6 inhibitor; Bio-Tech, Minneapolis, MN, USA; Cat #6955), STF083010 (IRE1α inhibitor; Merck Millipore, Burlington, MA, USA; Cat #SML0409), or GSK2606414 (PERK inhibitor; Merck Millipore, Burlington, MA, USA; Cat #516535). PERK activation was selectively induced using the EIF2AK3 activator CCT020312 (Merck Millipore, Burlington, MA, USA; Cat #324879).

4.2. RNA Extraction, cDNA Synthesis, and RT-qPCR

Total RNA was extracted using TRIzol™ Reagent (Thermo Fisher Scientific, Waltham, MA, USA; Cat. #15596026) according to the manufacturer’s protocol. Real-time quantitative PCR (RT-qPCR) was performed using a two-step process. The cDNA was synthesized from 1 to 5 µg of total RNA using the HiScript^®^ II 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China; Cat. #R211-01). Reverse transcription was carried out at 25 °C for 5 min, 50 °C for 45 min, and 85 °C for 5 s. Gene expression was then analyzed using the resulting cDNA and predesigned PrimeTime^®^ qPCR assays (Integrated DNA Technologies, Leuven, Belgium). For miRNA-qPCR, hsa-miR-675-5p (Assay ID 478196_mir) and hsa-miR-675-3p (Assay ID 478195_mir) were obtained from Thermo Fisher Scientific. miRNA-specific reverse transcription was performed using the Taq-Man™ MicroRNA Kit (Thermo Fisher Scientific, Waltham, MA, USA; Cat. #4427975) with 100 ng of total RNA per reaction, followed by incubation at 16 °C for 30 min, 42 °C for 30 min, and 85 °C for 5 min. The quantitative PCR for mRNA and miRNA was performed using a 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) under standard thermal cycling conditions: 50 °C for 2 min, 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. RPLP0 and U6 small nuclear RNA were used as endogenous controls for the normalization of mRNA and miRNA expression. RPLP0 is expressed at relatively stable levels across various tissues and conditions, making it a reliable reference for normalizing gene expression data in quantitative PCR. U6 SnRNA is a reliable housekeeping gene for normalization of miRNA RT-qPCR expression analysis. RPLP0 and U6 small nuclear RNA have been used as endogenous controls for mRNA and miRNA quantification during conditions of ER stress, respectively [86,87]. Relative expression levels were calculated using the 2^−ΔΔCT^ method. Details of all gene assays used in this study are summarized in Supplementary Tables S1 and S2.

4.3. Protein Extraction and Western Blot Analysis

Cells were washed once with ice-cold PBS and lysed in RIPA buffer following the indicated treatment times. Protein concentration was determined using the Bradford assay. Equal amounts of protein (40 μg per lane) were loaded per lane for the SDS-polyacrylamide gel for electrophoresis. Proteins were then transferred onto a nitrocellulose membrane and blocked with 5% milk in PBS containing 0.05% Tween-20 (PBST). The membrane was incubated overnight at 4 °C with primary antibodies against ATF6 (Abcam, Cambridge, UK; Cat #ab122897), spliced XBP1 (BioLegend, San Diego, CA, USA; Cat #619502), PERK (Cell Signalling, Danvers, MA, USA; Cat #C33E10), phospho-eIF2α (Cell Signalling Technology, Danvers, MA, USA; Cat #9721S), β-actin (Sigma-Aldrich, St. Louis, MO, USA; Cat #A-5060), HA-Tag (Cell Signaling Technology, Danvers, MA, USA; Cat #MA1-12429), and/or Tri-methyl-histone H3 (K4) (Cell Signaling Technology, Danvers, MA, USA; Cat #9727S). After three washes with PBST, the membrane was incubated with the 1:5000 diluted horseradish peroxidase-conjugated anti-rabbit IgG (Cell Signaling Technology, Danvers, MA, USA; Cat #7074) or anti-mouse IgG (Cell Signaling Technology, Danvers, MA, USA; Cat #7076) secondary antibody for 2 h at room temperature. Signals were detected using the Pico PLUS Chemiluminescent Substrate (Thermo Fisher Scientific, Waltham, MA, USA; Cat #34580).

4.4. Transient Transfection

Human H19 (ENST00000412788) full-length cDNA cloned into the control vector pLenti4/V5 (Life Technologies, Carlsbad, CA, USA) was generously provided by Dr. Reinier from the Institute of Cardiovascular Regeneration, Goethe University, Frankfurt, Germany. The HA-tagged ATF4 (Addgene, Watertown, MA, USA; Cat #115969) and XBP1 (Addgene, Watertown, MA, USA; Cat #115968) reporter plasmids were purchased from Addgene. For transfection in HEK293T cells, 1 µg of plasmid DNA was mixed separately with 6 µL of JetPEI (VWR International, Radnor, PA, USA; Cat #101-10N) in 100 µL of 150 mM NaCl. For transfection in MCF7 cells, 1 µg of plasmid DNA was mixed separately with 2 µL of TurboFect (Thermo Fisher Scientific, Waltham, MA, USA; Cat #15325016) in 100 µL of serum-free Opti-MEM (Thermo Fisher Scientific, Waltham, MA, USA; Cat #11058021). The transfection reagent and plasmid DNA solutions were incubated separately at room temperature for 10 min, then combined and incubated for an additional 20 min. The resulting 200 µL DNA complex was then added to 2 mL of fresh complete DMEM for 293T cells or Opti-MEM for MCF7 cells. After 4 h, Opti-MEM was replaced with complete DMEM.

4.5. Generation of Stable Cell Lines

H19-overexpressing clones were generated via lentiviral transduction. Lentiviral stocks were produced in HEK293T cells using PMD2.G and psPAX as packaging plasmids, along with the H19 overexpression vector (Addgene, Watertown, MA, USA; Cat #200835). MDA-MB-231 Cells were transduced with H19-overexpressing lentiviral or PLKO lentiviral for 24 h, after which positive clones were selected using puromycin-containing media (1 µg/mL) for one week.

4.6. Luciferase Reporter Assay

The 5 × ATF6-pGL3 (ATF6 pathway reporter) contains five copies of ATF6-binding sites (CTCGAGACAGGTGCTGACGTGGCGATTC) cloned into pOFlucGL3, upstream of the c-fos minimal promoter (−53 to +45 of the human c-fos promoter). This reporter plasmid was obtained from Addgene (Cat #11976). Cells were treated with TG for 24 h, 24 h post-transfection. Firefly luciferase and Renilla luciferase activities were measured 48 h after transfection using the Lucetta™ Luminometer (Lonza, Basel, Switzerland). Firefly luciferase activity was then normalized to Renilla luciferase activity for data analysis.

4.7. Flow Cytometry Analysis

Cells were plated in 6-well plates. After 24 h, cells were treated with either vehicle or the required compounds for the indicated time points. The media was collected into a separate 15 mL tube, and cells were harvested by trypsinization. Following centrifugation at 200× g for 5 min at 4 °C, the media was discarded, and cells were washed once with ice-cold PBS. The cells were then resuspended in fluorescent-activated cell sorting buffer. Gating was performed using propidium iodide (PI) unstained cells and positive control cells stained with PI (0.25 µg/mL). The percentage of live or dead cells was determined using the Cytek Northern Lights 2000 (Cytek Biosciences, Fremont, CA, USA), and data analysis was performed using FlowJo v10.9.0.

4.8. MTS Assay

1000 cells were seeded in a 96-well plate at an optimized density in 150 µL of complete culture medium. After 24 h of incubation, cells were treated with compounds for a specific duration or at multiple time-points to track cell viability or proliferation. At each testing time point, 20 µL of MTS solution (2 mg/mL 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) (Promega, Madison, WI, USA; Cat #G3582) and 0.9 mg/mL phenazine methosulfate (PMS), mixed at a 10:1 ratio) was added to 100 µL of medium in each well. The plate was then incubated at 37 °C for 1.5 h. Control wells containing only medium and MTS reagent (without cells) were used for background subtraction.

4.9. Bioinformatic Analysis

Bioinformatic analyses were performed using the Breast Cancer Gene-Expression Miner (bc-GenExMiner) version 5.2. Prognostic analysis for H19 expression was performed by selecting intrinsic molecular subtypes using METABRIC and TCGA datasets integrated within the bc-GenExMiner platform (Supplementary Figure S3). The overall survival associated with H19 expression for molecular subtypes for breast cancer as determined by Sorlie gene expression signature was assessed using parameters shown in Supplementary Figure S3.

4.10. Statistical Analysis

Data was analyzed using GraphPad Prism 10.1. Results are presented as mean ± SD from three independent experiments, unless otherwise stated. The p-value was determined using an unpaired t-test between independent groups, and results with p < 0.05 were considered statistically significant.