Effect of EGR1-Mediated LncRNA HOTAIR on MRP1 Gene Expression and MDR in Lung Cancer

Background: Multi-drug resistance-associated protein 1 (MRP1) plays critical roles in the multi-drug resistance (MDR) of cancer cells, and its expression is regulated by several transcription factors. However, the effects of EGR1 on MRP1 expression and MDR in lung cancer cells remain unknown. Methods: The expression of Transcription factors EGR1 and lncRNA HOTAIR in Non-small cell lung cancer were detected by using qRT-PCR, at the same time, Western Blot was used to detect the expression of transcription factor EGR1. MTT assay, Flow Cytometry and Observation with laser confocal microscope assay were used to explore the role of EGR1 and lncRNA HOTAIR in Non-small cell lung cancer cells. ChIP PCR assay and Luciferase reporter assays were used to demonstrate the molecular biological mechanism of EGR1 and lncRNA HOTAIR in Non-small cell lung cancer. Results: We found that EGR1 could bind to the MRP1 promoter at site -53/-42 bp and thereby regulate MRP1 expression. EGR1 knock-down reduced MRP1 expression, while EGR1 overexpression increased it. Knockdown of EGR1 increased the drug sensitivity to 5-Fluorouracil (5-FU), Toosendanin (TSN), and cisplatin (DDP), reduced the eux ability of Rho-123, and induced apoptosis of drug-resistant lung cancer cells A549/DDP, while EGR1 overexpression had the opposite effects. Further, we demonstrated that lncRNA HOTAIR mediated the effects of EGR1 on MRP1 and MDR via sponging of miR-6807-3p. Moreover, we showed that miR-6807-3p exerts its function by targeting the EGR1 3'UTR. Conclusions: We revealed the role and molecular mechanisms of the novel HOTAIR/miR-6807-3p/EGR1 axis in the regulation of MRP1 expression and MDR in lung cancer cells. Our ndings identify EGR1 as a potential biomarker and therapeutic target for MDR in human lung cancer. aim study to elucidate the role of EGR1 and HOTAIR in regulation of MDR. We show that EGR1 can bind to the promoter region of the MRP1 gene and can be regulated by lncRNA HOTAIR through miR-6807-3p, thereby modulating the expression of MRP1 and affecting the MDR of human lung cancer. CGCACCGCCGCCTGGTT. from PCRs 1.5% agarose electrophoresis. cell incubator. cytometry uorescence and of cells subtraction of the uorescence retained from the total uorescence is the uorescence index of Rho-123. repeated three times and average obtained to ability of Rho-123 to ow-out. HOTAIR these data demonstrate that Egr1 mediates lncRNA HOTAIR regulation of MRP1 expression and plays a critical role in the development of MDR during chemotherapy.


Construction of vectors
Construction of EGR1, lncRNA HOTAIR, and miR-6807-3p overexpression vectors The full-length EGR1 cDNA was ampli ed from A549/DDP total cDNA using PCR with primers containing the EcoRI/HindIII restriction sites. The ampli ed product was puri ed and cloned into the pcDNA3.1 vector (Addgene, China) to generate the pcDNA3.1-EGR1 construct. The lncRNA HOTAIR sequence (NR: 003716.3) was searched according to the known HOTAIR gene sequence (GeneID: 100124700) in NCBI GenBank, and the full length lncRNA HOTAIR gene was chemically synthesized and then cloned into the pcDNA3.1 vector to construct pcDNA3.1-lncRNA HOTAIR. Full-length miR-6807-3p fragment was ampli ed by PCR from the total DNA of A549/DDP cells. The ampli ed product was puri ed and cloned into pcDNA3.1 vector to construct pcDNA3.1-miR-6807-3p. The empty pcDNA3.1 vector was used as a negative control. The primers used for plasmid construction are shown in Table 1.
The predicted interaction between lncRNA HOTAIR and miR-6807-3p was analyzed with the help of the DIANA tool on the biometric prediction website. Dual luciferase reporter assays were performed to determine if miR-6807-3p is a direct target of lncRNA HOTAIR. The binding region of lncRNA HOTAIR was cloned and chemically synthesized. The primers used for plasmid construction are shown in Table 3. All recombinant vectors were veri ed by sequencing after construction in Sangon Biotech Co., Ltd. (Shanghai, China).

Real-time RT-PCR analysis
Total RNA was extracted from A549 and A549/DDP cells using EZ-10 RNA Mini-Preps KitS (BIO BASIC INC, China). Total RNA was reverse transcribed to cDNA using Bioteke super RT Kit (Bioteke Corporation, China). The real-time PCR reactions were carried out in a 96-well microtiter plate using the power 2×SYBR Real-Time PCR Premixture kit (Bioteke Corporation, China) and Real-Time PCR (Eppendorf, USA). All samples were analyzed in triplicate in three independent experiments. The uorescence of the PCR products was detected by the same apparatus. The number of cycles for the ampli cation plot to reach the threshold limit (Ct value) was used for quanti cation. β-actin and U6 were used as endogenous controls. Using the Ct values obtained by the analysis software, the relative expression levels of EGR1 and MRP1 mRNAs, lncRNA HOTAIR, and miR-6807-3p were calculated through the 2 -ΔΔCt algorithm [26][27][28]. The primers used in PCR are shown in Table 4.
Then, the samples were immunoprecipitated with control IgG (Bioss, China), anti-RNA polymerase II antibody, and anti-EGR1 antibody (Bioss, China). Afterwards, the protected DNA fragments were collected, and PCRs were performed using the following speci c primer set: Forward Primer: GCCGAGAGGTGGCTGGTCC; Reverse Primer: CGCACCGCCGCCTGGTT. Finally, the products from PCRs were analyzed by 1.5% agarose gel electrophoresis.
Luciferasereportergeneassay HEK-293T cells were grown in 24-well plates and co-transfected with 2 μL polyethylenimine (PEI; Sigma), 0.5 μg EGR1 overexpression vector, 0.5 μg of MRP1 promoter-luciferase plasmid DNA, and 0.01 μg of Renilla luciferase reporter plasmid DNA. After 4 h of incubation, the medium without fetal calf serum was replaced with 1640 complete medium. A549/DDP cells were grown in 24-well plates and co-transfected with 2 μL PEI (Sigma), 0.5 μg miR-6807-3p overexpression vector, and 0.5 μg of HOTAIR luciferase plasmid DNA. After 4 hours of incubation, the serumfree medium was replaced with 1640 complete medium. Firefly and Renilla luciferase activities were measured 48 hours after transfection, using a dual-luciferase reporter system (Promega, WI, USA) following the manufacturer's protocol [27]. The firefly luciferase activity was normalized to the Renilla luciferase activity. Each experiment was repeated at least three times.
Flow cytometry to determine the drug e ux of Rho-123 Fluorescence intensity of Rho-123 (99%, Shanghai Yuanyi Biotechnology Co., Ltd.) in A549/DDP transfected with the aforementioned expression vectors was detected using ow cytometry (BECKMAN COULTER, USA) to assess the change in MRP1 protein e ux capacity. The cells were collect in EP tubes containing 10 μg/mL Rho-123 and incubated for 30 minutes at 37 °C in a 5% CO 2 cell incubator. Flow cytometry was used to measure the uorescence intensity and emission wavelength of Rho-123 in cells with excitation wavelengths of 488 nm and 530 nm [26]. The subtraction of the uorescence retained from the total uorescence is the uorescence index of Rho-123. This process was repeated three times and an average value was obtained to calculate the ability of Rho-123 to ow-out.

Observation of cell morphological changes
According to the results of MTT assay, three anticancer drugs (DDP, TSN, and 5-FU) were added to A549/DDP cells transfected with different expression vectors. After 48 hours, acridine orange (AO) staining was performed. Cell morphology was observed under laser confocal microscope (Leica, Germany).

Statistical analysis
All results are expressed as means ± SD. Representative bands were selected from independent western blotting experiments. When data distributions approximated normality and two groups were compared, a Student's t-test was performed to evaluate the signi cance of differences. Differences were regarded as signi cant at a level of P<0.05. All statistical tests were performed using Prism software (GraphPad).

EGR1 is overexpressed in multidrug-resistant lung cancer cells A459/DDP
To uncover the role of EGR1 in the MDR of human lung cancer, its expression level was assessed in lung cancer cell lines A549 and multidrugresistant A549/DDP. qRT-PCR and western blot showed that the expression of EGR1 mRNA and protein was 3.15 and 1.92-fold higher, respectively, in A549/DDP cells than in A549 cells (Fig 1).
EGR1 upregulates the expression of MRP1 gene in A549/DDP cells To investigate the role of EGR1 on the expression of MRP1 in A549/DDP cells, four interference vectors targeting EGR1 mRNA were transfected into A549/DDP cells for 24, 48, and 72 hours. The most e cient interference vector was sh-EGR1-3 in 48 hours, which decreased the expression of EGR1 mRNA and protein by 75% and 62%, respectively, and was thus chosen for use in the subsequent experiments (Fig 2A, 2B).
The mRNA levels of MRP1 decreased signi cantly in EGR1 knock-down cells. By contrast, EGR1 and MRP1 mRNA levels increased by 7-fold and 2.6-fold, respectively, while protein levels increased by 2.13-fold and 1.87-fold at 48 hours, respectively, in EGR1-overexpressing cells.
Rescue experiments, where sh-EGR1-3 and pcDNA3.1-EGR1 Δ were co-transfected into A549/DDP cells, showed that the expression of MRP1 gene was restored to a level consistent with that of the control group (Fig 2C, 2D). These results suggest that expression of EGR1 could positively regulate the expression of MRP1 gene. EGR1 knockdown reduces MDR and promotes apoptosis in A549/DDP cells Next, uorescence intensity of Rho-123 was measured in A549/DDP cells transfected with sh-EGR1-3, pcDNA3.1-EGR1 Δ , pcDNA3.1-EGR1, or control vectors for 48 hours to investigate the effect of EGR1 on e ux capacity of MRP1. The uorescence intensity of Rho-123 was 2.23-fold higher in EGR1-silenced cells and 0.63 times lower in EGR1-overexpressing cells than in the control group, while there was no signi cant difference between the rescue experimental and control groups (Fig 3A-H).
To detect the effect of EGR1 on the drug sensitivity and cell morphology of A549/DDP cells, the cells were transfected with sh-EGR1-3, pcDNA3.1-EGR1 Δ , pcDNA3.1-EGR1, or control vectors for 48 hours, then treated with three anticancer drugs, 5-FU, TSN, and DDP. The IC50 values of these drugs in A549/DDP cells decreased signi cantly upon knock-down of EGR1 and increased signi cantly upon over-expression of EGR1, but there was no signi cant change in the rescue experimental group compared with the control group (Table 5). A549/DDP cells transfected with control vectors showed no signi cant difference in cell morphology and only the early stage of apoptotic body. Most EGR1 knock-down cells showed apoptosis and chromatin contraction, and some of the cells ruptured. Instead, EGR1-overexpressing cells showed no obvious sign of apoptosis and were relatively intact. The rescue experimental group had the same morphology as the control group and underwent early apoptosis (Fig 3I). These results indicate that silencing EGR1 gene expression can increase the sensitivity of cells to anticancer drugs DDP, TSN, and 5-FU, and promote apoptosis.

EGR1 is a transcription factor of MRP1
To explore the mechanisms by which EGR1 regulates the expression of MRP1 in A549/DDP cells, we analyzed the MRP1 promoter in search of putative EGR1 binding sites. The region between -68 bp and -27 bp (relative to the transcription start site) revealed four putative EGR1-binding sites through bioinformatic analysis (Fig 4A). ChIP-PCR was performed to detect whether EGR1 binds directly within the promoter region of MRP1. As shown in Figure 4B, EGR1 bound to the MRP1 promoter region at the putative binding sites, whereas the control IgG did not. Dual luciferase assays were then carried out to further con rm the speci city of EGR1 binding. A549/DDP cells were co-transfected with pGL3-MRP1-600 (containing the promoter region of MRP1 from -560 to +176 bp) or pGL3-600-ΔE1~ΔE4 (the EGR1-binding site-mutated vectors) and pcDNA3.1-EGR1 and phRL-TK vector (known as control reporter vector). The results showed higher luciferase activity in cells transfected with pGL3-MRP1-600 than in those transfected with pGL3-basic. Instead, luciferase activity was lower in pGL3-600-ΔE1~ΔE4-transfected than in pGL3-MRP1-600-transfected cells, but still higher than in pGL3-basic-transfected cells. The relative activity of luciferase in cells transfected with the deletion mutant pGL3-600-ΔE2 was signi cantly reduced by approximately 98% (Fig 4C). Taken together, these experiments show that EGR1 interacts with the MRP1 promoter region -53/-42 bp.
LncRNA HOTAIR regulates MRP1 gene expression and MDR Next, we explored the role of lncRNA HOTAIR in regulating MRP1 expression and MDR. First, we determined that the expression of HOTAIR was 1.81-fold higher in A549/DDP cells than in A549 cells (Fig S1). To investigate the relationship between HOTAIR and MRP1, we constructed HOTAIR overexpression plasmids and three HOTAIR-targeting shRNA-expressing plasmids. Of these, sh-LncRNA-H1 showed the highest knockdown e ciency ( Fig 5A). As expected, MRP1 expression was greatly increased after HOTAIR overexpression and reduced after HOTAIR knockdown and was almost restored to the control level through rescue with an RNAi-resistant form of HOTAIR (Fig 5B). Interestingly, similar changes occurred in the expression of EGR1. Further, consistent with the direct regulation of MRP1 expression by HOTAIR, intracellular Rho-123 retention was dramatically decreased after HOTAIR overexpression, increased after HOTAIR knock-down, and fully restored after HOTAIR rescue (Fig 5C). These results reveal that lncRNA HOTAIR regulates MRP1 and contributes signi cantly to development of MDR and sensitivity of cells to chemotherapeutic drugs.

EGR1 is critical for HOTAIR-mediated MRP1 upregulation and MDR
To determine whether EGR1 is required for HOTAIR-mediated regulation of MRP1 expression and MDR, we tested the effect of EGR1 knockdown on MRP1 expression and MDR after overexpression of HOTAIR in A549/DDP cells. Expression of both HOTAIR and MRP1 showed a dramatic increase in cells transfected with pcDNA3.1-HOTAIR. Interestingly, the increased transcription and expression of MRP1 remained unaffected by the scrambled shRNA control but were abolished by EGR1 knock-down (Fig 6A, 6B). These results con rm the hypothesis that EGR1 expression and activity are critical for lncRNA HOTAIR to modulate MRP1 expression. In addition, retention of Rho-123, which was reduced by HOTAIR overexpression, was mostly restored by EGR1 knock-down (Fig 6C). These ndings implicate that EGR1 is essential for HOTAIR to regulate MRP1 expression and MDR development.
HOTAIR regulates the expression of EGR1 by sponging miR-6807-3p Next, we explored the mechanisms by which HOTAIR regulates the expression of EGR1. Using DIANA tools and Targetscan, we predicted that miR-6807-3p could bind to HOTAIR and EGR1 (Fig 7A). Dual luciferase assays were carried out to con rm the speci city of miR-6807-3p interaction with the putative binding sites. Compared to the cells transfected with control, cells co-transfected with pcDNA3.1-miR-6807-3p and psiCHECK2-HOTAIR-wt or psiCHECK2-EGR1-3'UTR-wt showed reduced luciferase activity (Fig 7B). Instead, the mutant plasmid had no effect.

Discussion
Chemotherapy-based treatment is currently one of the most important approaches for several types of cancers but is hampered by the occurrence of MDR during the administration of chemotherapeutic agents. MRP1 upregulation is regarded as one of the main causes of MDR in cancer cells, but the mechanisms underlying this change in its expression remain obscure. EGR1 is an immediate-early gene that is induced by a wide variety of stimuli and has important roles in the responses of different types of cells [31][32][33]. EGR1 can promote both growth and apoptosis of tumor cells [34,35]. In this study, we found that the expression of EGR1 in A549/DDP cells was higher than that in A549 cells (Fig. 1). Moreover, the expression of MRP1 and the e ciency of cell e ux in A549/DDP were increased upon EGR1 over-expression, and the opposite trend was observed upon EGR1 knock-down (Fig. 2, 3). These ndings reveal that EGR1 may be involved in the regulation of MRP1 gene expression. As a nuclear transcription factor, EGR1 binds to a GC-rich motif (5′GCG/TGGGCG3′) through its three zinc-nger DNA-binding domains [36] to modulate transcription of its target genes [19]. We identi ed putative EGR1 binding sites in the MRP1 upstream promoter sequence, and con rmed by ChIP analyses and dual luciferase assays that EGR1 could bind to a speci c site that spans from nucleotides − 53 to -42 of the promoter (Fig. 4). So, we identi ed EGR1 as a transcription factor of MRP1 and demonstrated that it plays critical roles in MRP1 upregulation and MDR.
Recently, lncRNAs have been widely identi ed and annotated [37][38][39]. LncRNAs play important roles in a variety of cellular processes via modulating gene expression at different levels, including RNA transcription, editing, and transport, and protein translation [40][41][42][43][44][45][46]. Although increasing evidence highlights the important role of lncRNAs in human disease, reports about the role of lncRNAs in MDR at the molecular level are scarce. The HOTAIR gene is located on human chromosome 12, and the region between HOXC11 and HOXC12 in the HOXC cluster is composed of 6 exons [30]. In this study, we found that lncRNA HOTAIR was upregulated in A549/DDP cells (Fig S1), and overexpression of HOTAIR promoted MRP1 expression and MDR development. The opposite trend was observed when HOTAIR was silenced in A549/DDP cells.
Importantly, the expression of EGR1 was upregulated after HOTAIR overexpression and downregulated after RNAi-based HOTAIR knock-down in A549/DDP cells (Fig. 5). Moreover, EGR1 knock-down substantially abolished MRP1 upregulation and MDR mediated by HOTAIR overexpression in A549/DDP cells (Fig. 6). Collectively, these data demonstrate that Egr1 mediates lncRNA HOTAIR regulation of MRP1 expression and plays a critical role in the development of MDR during chemotherapy.
Recent studies demonstrated that lncRNAs regulate gene expression via several mechanisms [37,47,48]. Increasing evidence suggests that competing endogenous RNAs (ceRNAs) might be implicated in lncRNA-mediated target regulation [49][50][51]. The ceRNA hypothesis de nes lncRNAs, pseudogenes, or mRNAs, as molecular sponges that can competitively bind to miRNAs and thereby modulate the activity and expression of their targets. Here, we used DIANA tools and Targetscan to predict the binding of HOTAIR and EGR1 to miR-6807-3p. Dual luciferase assays con rmed the speci city of this binding. Moreover, miR-6807-3p could regulate endogenous EGR1 expression (Fig. 7).

Conclusion
In summary, our study has revealed a new lncRNA HOTAIR/miR-6807-3p/EGR1 regulatory axis for MRP1 upregulation and de ned EGR1 as a novel therapeutic target for the reversal of MDR and the e cient treatment of drug-resistant cancers.