An Enhancer LncRNA Regulates NFE2 Expression and Proliferation in Human Leukemic K562 Cells*

LU Yan-Fei, QU Song-Ya, ZHU Jing-Jing, LⅠU Chao, WANG Jian,HAN Bing-She, ZHANG Jun-Fang

(1)Key Laboratory of Exploration and Utilization of Aquatic Genetic Resources, Ministry of Education, Shanghai Ocean University,Shanghai 201306, China;2)National Demonstration Center for Experimental Fisheries Science Education, Shanghai Ocean University, Shanghai 201306, China;3)Marine Biomedical Science and Technology Innovation Platform of Lin-gang Special Area, Shanghai 201306, China)

Abstract Objective Transcription factor NFE2 was observed abnormal expression in myeloproliferative neoplasm (MPN)patients. However, how NFE2 is transcriptionally regulated remains ambiguous. This study aims to explore the elements and molecular mechanisms involved in the transcriptional regulation of NFE2. Methods Active enhancers were predicted by public NGS data and conformed experimentally via dual luciferase reporter assay. After that, PRO-seq and GRO-seq data was used to detect enhancer RNAs transcribed from these enhancers. RACE was utilized to clone the full length enhancer RNA (eRNA) transcripts, and RT-qPCR was used to measure their expression in different leukemia cell lines as well as the transcript levels during induced differentiation. Finally, to investigate the molecular function of the eRNA, overexpression and knockdown of the eRNA via lentivirus system was performed in K562 cells. Results We identified three enhancers regulating NFE2 transcription, which located at -3.6k,-6.2k and +6.3k from NFE2 transcription start site (TSS) respectively. At the -3.6k enhancer, we cloned an eRNA transcript and characterized that as a lncRNA which was expressed and located in the nucleus in three types of leukemia cell lines. When this lncRNA was overexpressed, expression of NFE2 was upregulated and decreases of K562 cell proliferation and migration ability were observed. While knocking down of this lncRNA, the level of NFE2 decreases correspondingly and the proliferation ability of K562 cells increases accordingly. Conclusion We identified an enhancer lncRNA that regulates NFE2 transcription positively and suppresses K562 cell proliferation.

Key words NFE2, enhancer, lncRNA, cellular proliferation

Early studies identified two alternative promoters ofNFE2,with the 1F (fetal) promoter used more frequently in the fetal liver and the 1A (adult)promoter used with greater frequency in the adult bone marrow[12]. Thus, all promoters mentioned in this paper refer to the 1A promoter. Expression ofNFE2is regulated by cytokines or growth factors. Ⅰn megakaryocytes, interleukin-1β and platelet-derived growth factor (PDGF) were found to facilitateNFE2expression[13-15]while interleukin-4 had the contrary effect[16].NFE2expression is also controlled by other transcription factors. GATA1[17]as well as RUNX1[18]are the major regulators ofNFE2transcriptional activity and function in maturation of megakaryocytes and platelet generation. NF-κB, as a negative regulator ofNFE2, may be responsible for maintaining low levels ofNFE2in early erythroid progenitors[19]. Ⅰn MPN patients withNFE2overexpression, AML1 has been shown to upregulate NFE2[20]. A typical positive feedback mechanism controlsNFE2in MPN patients, where overexpressed NFE2 leads to an increase of epigenetic enzyme JMJD1C, and more JMJD1C in turn enhancesNFE2transcription[21].

Here, we identifiedcis-regulatory elements associated with theNFE2gene, and we cloned an enhancer RNA (eRNA) transcript of one of the enhancers. We identified this eRNA as a lncRNA and demonstrated its regulatory function onNFE2. Ⅰn addition, we found this lncRNA suppressed the proliferation of K562 cells.

1 Materials and methods

1.1 Cell culture and treatment

K562 (CCL-243, ATCC), HL-60 (CCL-240,ATCC), and U937 (CRL-1593.2, ATCC) cells were cultured in RPMⅠ 1640 medium supplemented with 10% fetal bovine serum (FBS) (10099141, Gibco) and 1% penicillin-streptomycin-glutamine solution(SV30082.01, Hyclone). All cell lines were cultured at 37℃ in a humidified atmosphere containing 5% CO2.For treatment, K562, HL-60, or U937 cells were seeded at a density of 1×108/L, the final concentrations of 30 μmol/L hemin (51280, Sigma-Aldrich), 0.16 μmol/L all-trans-retinoic acid(ATRA) (R2625, Sigma-Aldrich) or 2 μmol/L 12-O-tetradecanoylphorbol 13-acetate (TPA) (S1819,Beyotime) were added to the media to induce erythroid, monocytic, and granulocytic differentiation,respectively, and then cultured for 48 h.

1.2 Chromatin immunoprecipitation （ChIP）

ChⅠP experiments were performed as previously described[22]. Ⅰn brief, 1×107cells were fixed in 1%formaldehyde for 10 min at room temperature and sonicated to shear the chromatin.Ⅰmmunoprecipitation of crosslinked chromatin was performed overnight at 4℃ with H3K27ac and H3K4me1 antibodies (ab4729 and ab8895, Abcam).An equal amount of isotype immunoglobulin G (ⅠgG)was used as background control. Primers for ChⅠP-qPCR are shown in Table S1.

1.3 Dual-luciferase reporter assay

TheNFE2promoter (chr12: 54 300 825-54 301 741, reverse complement, hg38) was amplified, digested withXhoⅠ andBglⅠⅠ, and cloned into the pGL4 luciferase reporter vector (E6651,Promega). The upstream regions -3.6k (chr12:54 303 806-54 305 005, reverse complement, hg38),-6.2k (chr12: 54 306 477-54 307 773, reverse complement, hg38) and downstream +6.3k (chr12: 54 294 275-54 295 160, reverse complement, hg38)were amplified and inserted into pGL4-NFE2-promoter mentioned aboveviaXhoⅠ andSacⅠdigestion. Then these constructs were cotransfected into HEK-293T cells with pRL-TK vector (E2241,Promega) using TurboFect Transfection Reagent(R0531, ThermoFisher) according to the manufacturer’s instructions. Dual-Luciferase Reporter Assay System (E1910, Promega) was used to measured luciferase activity on a FlexStation 3 multimode microplate reader. All assays were performed in triplicate and repeated at least 3 times.

1.4 Rapid-amplification of cDNA ends （RACE）

Total RNA was isolated using TRⅠzol reagent(15596018, Ⅰnvitrogen). Then 1 μg of total RNA was reverse-transcribed and amplified with SMARTer RACE 5'/3' Kit (634858, Takara). Amplified cDNA fragments were inserted into pEASY-Blunt Zero Cloning Vector (CB501, TransGen, China). After that,the constructs were sequenced and blasted to reference genome sequence to determine if it is the required sequence. Primers for RACE are shown in Table S1, and spliced -3.6k-lncRNA sequence was shown in Document S1.

1.5 Quantitative real-time PCR analysis

Extracted total RNA was reverse-transcribed to cDNA using PrimeScriptTMRT reagent kit with gDNA eraser (RR047A, TaKaRa) and then diluted to 1 mg/L.To measure the levels of candidate RNAs, the iTaqTMUniversal SYBR Green Supermix (1725124, Bio-Rad) and a Light Cycler 480ⅠⅠ real-time PCR machine were used. Data was normalized to human GAPDH transcripts. Relative quantitation was carried out by the comparative threshold cycle (CT) method. The primer sequences are listed in Table S1.

The girl watched Gaston on the plate, and she actually didn t like him. He was all ugh, as he had been in the first place. He didn t have a home anymore and he was wandering around on the white plate and he was silly and wrong and ridiculous and useless and all sorts of other things. She cried a little, but only inside, because long ago she had decided17 she didn t like crying because if you ever started to cry it seemed as if there was so much to cry about you almost couldn t stop, and she didn t like that at all. The open halves of the peach seed were wrong, too. They were ugly or something. They weren t clean.

1.6 Nucleus extraction and isolation

About 1×107cells were collected by centrifugation, washed twice with 10 ml of pre-cooled PBS, centrifuged at 1 000gfor 5 min. The supernatant was discarded, the cells were resuspended in 800 μl of hypotonic buffer and incubated on ice for 10 min to swell the cells. The cells were adequately broken up using a Dounce homogenizer after which 10% triton-100 was added and thoroughly mixed.After centrifugation at 1 500gfor 5 min at 4℃, the supernatant and the precipitate were collected separately.

1.7 Lentivirus production and cell transduction

For creation of the overexpression vectors, -3.6klncRNA sequence was cloned and ligated into the pLVX-ⅠRES-Neo vector (632181, Clontech) which had been digested byEcoRⅠ andBamHⅠ. For knockdown vectors, oligonucleotides for sh-3.6klncRNA were annealed and ligated into the pLKO.1-puro vector (8453, Addgene) digested withEcoRⅠ andAgeⅠ. After that, packaging plasmids pCMV-VSVG(8454, Addgene), pCMV-DR8.91 (PVT2323, Life Science Market) and the relevant lentiviral transfer vectors were cotransfected into HEK-293T cells in a mass ratio of 1∶9∶10. After 48 h, the media containing lentivirus particles were collected. To infect K562 cells, hexadimethrine bromide (H9268,Sigma-Aldrich) was added to a final concentration of 8 mg/L. An appropriate number of cells were cultured in this lentivirus-containing medium and collected after 72 h.

1.8 Cell proliferation assay

Collected cells were diluted to 2×107/L with complete culture media. One hundred μl of cell suspension was added to each well of the 96-well plate and three replicates of each sample were set up.The wells without the addition of cells were used as blank controls and transferred to the cell culture incubator for incubation. Three h before each test time point, 10 μl of CCK-8 (C0042, Beyotime) solution was added to each well and continued incubation in the cell incubator. The absorbance was measured at 450 nm and the mean value calculated.

1.9 Cell migration assay

Six hundred μl of complete media was added to the wells of a 24-well plate. Cells were resuspended in serum-free medium to 5×108/L and 100 μl of cell suspension was seeded into the transwell upper chamber, then the chambers were placed into the 24-well plate. After 16 h of incubation, the medium was removed from the chambers and the chambers were removed with forceps and washed twice in a beaker containing PBS. The chambers were placed in a new 24-well plate, 1 ml of 3.7% paraformaldehyde was added and incubated for 20 min at room temperature.Residual formaldehyde was removed from the chambers, after that, the chambers were washed twice with PBS and placed in a new 24-well plate. The plates were incubated for 15 min at room temperature and protected from light with 1 ml of Giemsa staining solution. Wash the chambers with PBS and place in a new 24-well plate, gently wiping away any cells that have not migrated from the upper chamber with a cotton swab. Photographs were taken under an inverted microscope and the number of cells in different fields of view was counted and averaged.

1.10 Bioinformatics and statistical analysis

The ChⅠP-seq datasets downloaded from the ENCODE project were visualized with the ⅠGV. PRO-seq and GRO-seq raw sequencing data were downloaded and aligned to hg38 reference genome.The resulting bam files were converted to bigwig format and then visualized. Data was obtained from at least three independent experiments and are expressed as themeans±standard deviation (SD). Statistical significance (P<0.05) for all the above experiments was assessed by Student’s two-tailedt-test.

2 Results

2.1 Identification of 3 enhancers for the NFE2 promoter

The tissue specificity ofNFE2expression almost exclusively in hematopoietic progenitors and differentiated cells of the erythroid, megakaryocytic,granulocytic and mast cell lineages means that its transcription is tightly regulated[1-2]. To investigate its transcriptional regulation mechanism, we predicted enhancers that might be involved in regulating NFE2 expression. Combining two histone modifications,H3K27ac and H3K4me1, with chromatin accessibility analysis such as DNase Ⅰ hypersensitive sites is commonly used to predict enhancers[23-24]. Besides,transcriptional co-activator CBP and p300 are identified to bind at enhancers[25]. Therefore, we downloaded ChⅠP-seq data of H3K27ac (ENCSR000 AKP), H3K4me1 (ENCSR000AKS), CBP (ENCSR00 0ATT) and p300 (ENCSR000EGE) as well as DNaseseq (ENCSR000EKS) and ATAC-seq (ENCSR868 FGK) data of K562 cell line which displays a high level ofNFE2(Figure 1a). These signals indicate three potential enhancers located at upstream 3.6k,6.2k and downstream 6.3k basepairs from theNFE2transcription start site (TSS). ChⅠP-qPCR was further performed to confirm the H3K27ac (Figure 1b) and H3K4me1 (Figure 1c) enrichment at these regions.

Dual luciferase reporter experiment was performed to further verify the enhancer activity of these three regions. DNA fragments containing the corresponding H3K27ac and H3k4me1 peaks, named-3.6k, -6.2k and +6.3k, respectively, were cloned and inserted upstream of theNFE2promoter controlling a firefly luciferase reporter gene (Figure 1d) and then the constructs were transfected into HEK-293T cells.Compared with the non enhancer group, the -3.6k,-6.2k and +6.3k fragments all showed significantly increased luciferase activity (Figure 1e). Among these, +6.3k fragment displayed the highest enhancer activity. Taken together, these data suggest that these fragments contain enhancer sequences for theNFE2promoter.

Fig. 1 The -3.6k, -6.2k and +6.3k regions of NFE2 are active enhancers

2.2 Two enhancers show bidirectional transcription signals

Enhancer transcription was found to be a widespread occurrence in the genome[26-27].Bidirectional transcription is the main form of enhancer transcription[27-28]. eRNAs have been proved to be involved in the regulation of target genes in a variety of ways, for examples, enhancing the formation or stability of the enhancer-promoter loops[29-31], assisting in the recruitment of transcription factors and co-regulators and regulating their activity[32-34], facilitating RNA Pol ⅠⅠ pause-release to promote transcription elongation[35]. Thus, we explored whether these three enhancer regions have transcriptional signals. PRO-seq (GSM1480327) and GRO-seq (GSM1480325) data were downloaded and used to evaluate the likelihood of transcripts being present in these regions (Figure 2a). These data demonstrate that both -3.6k and -6.2k enhancers have a bidirectional transcriptional signal. Based on these signals, we successfully cloned the eRNA in one direction of the -3.6k region, referred to below as-3.6k-, by the RACE technique (Figure 2b). A negative sign denotes that the sequence of this eRNA is consistent with the minus strand sequence of the human reference genome. Ⅰn brief, this eRNA is 1 859 nt long and extends from 3 779 bp to 1 920 bp upstream ofNFE2TSS. Considering that we used oligo T to enrich the eRNA and that there is a typical polyA signal 16 bp away from the 3' end, this suggests that the eRNA molecule has a polyA structure (Document S1). Subsequently, using the protein coding potential prediction tool CPC2 (coding potential calculator 2)to predict its coding potential, we found that this eRNA do not have coding potential.

Fig. 2 -3.6k and -6.2k enhancers transcribe bidirectionally producing enhancer RNAs

We then explored the expression of -3.6klncRNA in different types of leukemic cells (Figure 2c). Ⅰn K562, HL-60, and U937 cell lines expressing NFE2, -3.6k-lncRNA was present in lower abundance of transcript level. To further clarify the regulatory function of the lncRNA on NFE2, we used hemin,TPA and ATRA to induce differentiation of K562,U937 and HL-60 cells for 48 h, respectively. After induction of erythroid differentiation in K562 cells using hemin,NFE2was significantly increased and-3.6k-lncRNA level had a corresponding variation(Figure 3a). However, after induction of U937 and HL-60 cells, the transcript levels of bothNFE2and-3.6k-lncRNA decreased (Fig 3b, c). These results are consistent with the reported levels of NFE2 at different stages of hematopoietic cell differentiation,where NFE2 is highly expressed in erythrocytes but lower in monocytes and granulocytes. Compared to the adjacent geneCOPZ1, -3.6k-lncRNA maintained a consistent pattern of change withNFE2after induction in different cell lines, providing clues that this lncRNA regulatesNFE2gene transcription.

Depending on the localization and specific interactions with DNA, RNA and proteins, lncRNAs exhibit different functions that ultimately affect gene expression in a variety of biological and physiological environments[37]. Therefore, we then isolated cytoplasmic and nuclear fractions of the above 3 cell lines and determined the content of -3.6k-lncRNA in the different fractions by RT-qPCR as a judgment of the subcellular localization of this lncRNA. GAPDH and 18S rRNA were used as cytoplasmic marker, U1 and MALAT1 were used as cytosolic marker. Our results suggest that in all 3 cell lines -3.6k-lncRNA is localized in the nucleus (Figure 3d-f). Compared to approximately 90% -3.6k-lncRNA localization in the nucleus in K562 cells, this value is slightly lower in HL-60 and U937 cells, which may be due to lower accuracy as a result of lower abundance of this lncRNA in these two cell lines.

Fig. 3 Changes of -3.6k-lncRNA expression after induced differentiation of leukemia cell lines and subcellular localization of this lncRNA

2.3 -3.6k-lncRNA suppresses cell proliferation and migration of K562 cells

To further understand the role of -3.6k-lncRNA in leukemia cells, we separately overexpressed and knocked down the lncRNA in K562 cells by a lentiviral system. RT-qPCR of total RNA shows successful overexpression and knockdown of -3.6klncRNA (Figure 4a, d). Not surprisingly,overexpression of -3.6k-lncRNA increased the transcript level ofNFE2(Figure 4b) and a significant decline ofNFE2was found after knockdown (Figure 4e) providing direct evidence that this lncRNA regulatesNFE2transcription. We then examined the role of -3.6k-lncRNA in the proliferation and migration of K562 cells. The optical density of infected cells, which responds to the number of cells,was measured by the CCK-8 assay every 24 h after seeding. Our results show that overexpression of-3.6k-lncRNA significantly reduces the proliferative capacity of K562 cells (Figure 4c), while the rate of cellular proliferation is accelerated in cells with decreased lncRNA levels (Figure 4f). Ⅰn addition,migratory capacity of -3.6k-lncRNA overexpression K562 cells, which was detectedviatranswell migration assay, was observed having an apparent decrease (Figure 4g, h). However, no significant variation was observed in -3.6k-lncRNA knockdown cells (Figure 4i).

Fig. 4 -3.6K-lncRNA affects NFE2 transcription and K562 cell proliferation

Taken together, these results suggest that -3.6klncRNA is involved in the regulation of NFE2 expression and plays an important role in the proliferation and migration of K562 cell.

3 Discussion

Enhancers are a class ofcis-DNA regulatory elements that are capable of transcription and contain transcription factor recognition sequences. Enhancers are located proximal or distal to the promoter of a target gene and establish physical contact with the promoter of target gene through chromatin interactions, thereby regulating gene expression[38].Enhancers in cancer cells promote the transcriptional expression of oncogenes by interacting with promoters, which in turn leads to tumorigenesis.Therefore, enhancers can be regarded as potential targets for tumor therapy. For instance, PD-L1, a cell surface receptor on tumor cells playing an significant role in immune escape[39], when a distal enhancer ofPD-L1was knocked outviaCRⅠSPR-cas9,PD-L1expression was significantly reduced at both mRNA and protein levels, and immune escape of cancer cells was effectively inhibited[40]. Current studies observe abnormal expression of eRNAs in tumor samples compared to normal tissue[41]. Enhancer overactivation is a unique phenomenon in tumorigenesis, and targeting eRNAs would be a potential new anti-cancer therapeutic strategy[41].The clinical relationship betweenNFE2and tumors has been reported in many papers[42-44]. However, the mechanism that leads to the overexpression ofNFE2in patients with myeloproliferative neoplasms has only been sparsely reported.JMJD1C, encoding a histone demethylase, is one of the target genes of NFE2. Ⅰts protein level is elevated in patients with MPN due to the overexpression ofNFE2[21].Furthermore, JMJD1C, in turn, reduces the level of histone methylation in the promoter and upstream regions ofNFE2, thereby enhancing the expression ofNFE2. Ⅰn this study, we identified 3 enhancer regions in the vicinity of theNFE2promoter. The -3.6k region has been shown to significantly enhance the transcription ofNFE2in the presence of AML1 transcription factor binding[20]. Among enhancers measured by dual-luciferase reporter assay, +6.3k region displays the highest enhancer activity. Ⅰt is worth noting that the +6.3k region is adjacent to theNFE21F promoter, which is almost not used in adult bone marrow cells. Therefore, it is worthwhile to further investigate the role of the +6.3k enhancer in the transcription of adult and fetal types of NFE2.Besides, we found a new transcript of -3.6k enhancer and its regulatory activity forNFE2. The levels of this lncRNA during leukemic cells differentiations vary consistently withNFE2. Thus, in addition to the enhancer we identified, the lncRNA itself could also serve as a novel target for the treatment of MPN.

Ⅰn our K562 cell model overexpressing -3.6klncRNA, our data show that overexpression of this lncRNA has a significant inhibitory effect on the proliferation of tumor cells. K562 cells with -3.6klncRNA knock-down have an enhanced proliferative capacity. These data are consistent with the studies that in the Friend virus-induced mouse model of erythroleukemia, the absence of NFE2 promotes tumor growth by accelerating the rate of cellular proliferation[9], but differ from studies suggested that overexpression of NFE2 promotes the conversion of MPN to AML[20,45]. One possible reason for this is that -3.6k-lncRNA has an additional way of regulating cell proliferation as a lncRNA compared to the NFE2 protein. Ⅰn addition, it is also possible thatNFE2is differentially expressed in different cells at different stages of development, just as NFE2 overexpression causes different phenotypes in early and late erythrocyte maturation. Gene ontology analysis of target genes of NFE2 predicted byNFE2ChⅠP-seq data (ENCSR552YGL) in K562 cells reveals that the functions of many target genes are involved in intercellular communication and cell motility. Ⅰn our study, a significant enhancement in the migratory capacity of K562 cells was observed only upon overexpression of -3.6k-lncRNA, while there appeared to be no effect upon knockdown. Hence, not only the correlation between NFE2 and migration ability needs to be further verified, but the mechanism by which this lncRNA is associated with it also needs to be investigated in greater depth. Ⅰn addition, the failure to clone the remaining transcripts in the two enhancers of upstream with bidirectional transcription suggests that they may not have typical polyA structures. This leads to their rapid degradation after transcription, making them difficult to be captured[38].Further research is needed to confirm their sequences and their role in the transcriptional regulation ofNFE2. Ⅰn conclusion, how -3.6k-lncRNA regulatesNFE2gene expression and whether there are effects on cell proliferation that bypassNFE2demand further investigation.

4 Conclusion

Ⅰn this study, we first identified 3 enhancers that regulateNFE2transcriptional activity, located at-3.6k, -6.2k, and +6.3k relative to theNFE2TSS.Subsequently, we cloned the -3.6k-eRNA and identified it as a lncRNA. We found that this lncRNA mainly localizes in the nucleus and is expressed in three leukemia cell lines (K562, U937, and HL-60).Finally, through overexpression and knockdown experiments, we demonstrated that this lncRNA participates in the regulation of the target geneNFE2,and plays a role in inhibiting the proliferation and migration of K562 cells.

SupplementaryAvailable online (http://www.pibb.ac.cn or http://www.cnki.net):

PⅠBB_20230105_Table_S1.pdf

PⅠBB_20230105_Doc_S1.pdf