Long Non-Coding RNA H19 Leads to Upregulation of γ-Globin Gene Expression during Erythroid Differentiation

Abstract

Long noncoding RNAs (lncRNAs) are important because they are involved in a variety of life activities and have many downstream targets. Moreover, there is also increasing evidence that some lncRNAs play important roles in the expression and regulation of γ-globin genes. In our previous study, we analyzed genetic material from nucleated red blood cells (NRBCs) extracted from premature and full-term umbilical cord blood samples. Through RNA sequencing (RNA-Seq) analysis, lncRNA H19 emerged as a differentially expressed transcript between the two blood types. While this discovery provided insight into H19, previous studies had not investigated its effect on the γ-globin gene. Therefore, the focus of our study was to explore the impact of H19 on the γ-globin gene. In this study, we discovered that overexpressing H19 led to a decrease in HBG mRNA levels during erythroid differentiation in K562 cells. Conversely, in CD34+ hematopoietic stem cells and human umbilical cord blood-derived erythroid progenitor (HUDEP-2) cells, HBG expression increased. Additionally, we observed that H19 was primarily located in the nucleus of K562 cells, while in HUDEP-2 cells, H19 was present predominantly in the cytoplasm. These findings suggest a significant upregulation of HBG due to H19 overexpression. Notably, cytoplasmic localization in HUDEP-2 cells hints at its potential role as a competing endogenous RNA (ceRNA), regulating γ-globin expression by targeting microRNA/mRNA interactions.

Keywords:

• lncRNA H19
• γ-globin gene
• β-thalassemia
• erythroid differentiation

Introduction

In southern China, the incidence of β-thalassemia remains high. Consequently, efforts to combat the disease have primarily concentrated on prevention through prenatal diagnosis, aiming to prevent the birth of children with genetic defects. However, individuals living with β-thalassemia, especially those who are transfusion-dependent, are in great need of treatment options to reduce their anemia and transfusion dependence. Fetal hemoglobin (HbF) is a hemoglobin synthesized during fetal life that functions similarly to adult hemoglobin (HbA). Numerous clinical studies have shown that elevated HbF production is associated with reduced morbidity and mortality in sickle cell disease (SCD) and β-thalassemia [1,2]. For example, Esrick et al. used shmiR vector (BCH-BB694 BCL11A) for erythroid-specific knockdown of BCL11A to increase HbF levels [3]; in addition, Fu et al. used gene editing technology for the first time to treat children with β0/β0 severe thalassemia, and achieved important results. The principle of this approach is to reactivate the expression of the γ-globin gene [4]. Therefore, reactivating the shut-off γ-globin gene and increasing HbF synthesis have emerged as the preferred clinical treatment for β-thalassemia. This process involves a critical γ → β switching process, which occurs around birth; the mechanism is complex and regulated by numerous trans-regulatory proteins acting in combination with cis-regulatory elements.

Long noncoding RNAs (LncRNAs) are non-coding RNAs >200 nucleotides in length that are closely involved in the regulation of life activities such as embryonic development, muscle growth, fat metabolism, the immune response, and cancer development [5,6]. Furthermore, recent studies have found that lncRNAs play an important role in γ→β switching and erythroid differentiation. For example, lncRNACCDC26 was significantly up-regulated in childhood acute myeloid leukemia (AML), and after knocking down the expression of CCDC26, the expression of globin gene at the transcriptional level was shown to switch from fetal to embryonic [7]. Moreover, when lncRNAHBBP1 was overexpressed, the expression of γ-globin at the transcriptional and protein levels increased in HUDEP-2 cells, suggesting a role of HBBP1 in globin regulation [8]. Finally, it has also been reported that lncRNABGLT3 is a developmental stage-specific lncRNA that can positively regulate the γ-globin gene [9]. Based on these studies, it is clear that lncRNAs play a pivotal role in regulating the γ-globin gene.

In previous studies, we found that lncRNA H19 was among the differentially expressed transcripts between premature and full-term groups. Additionally, a total of 75 umbilical cord blood samples from healthy newborns were collected to verify the differential expression of H19. Among the 17 genes selected for RT-qPCR, H19 was also consistent with the results of RNA-Seq [10]. Moreover, a large number of studies have shown that H19 acts as ceRNA in tumors (breast cancer, nasopharyngeal carcinoma, tongue squamous cell carcinoma, ovarian cancer, pancreatic cancer, gastric cancer, esophageal cancer, etc.) to regulate the expression of downstream target genes [11–17]. However, no studies have been published on H19 in the γ-globin gene. Because H19 is highly expressed in many organs from early embryonic development to fetal life, it is scarcely expressed after birth, except in muscle and heart tissue, and has distinct spatiotemporal expression patterns [18,19]. Similar to γ-globin, H19 expression abundance decreases with developmental maturity, but whether there is an association between these two genes, and the mechanism through which H19 may regulate HBG expression, has not yet been reported. Notably, the cellular localization of lncRNAs determines their function. The majority of lncRNAs in the cytoplasm act as ceRNAs, while nuclear lncRNAs are closely related to gene expression. Their nuclear functions include chromosome scaffolding, chromatin remodeling, alternative splicing, and epigenetic control of transcription [20]. We investigated the effect of H19 on HBG expression during erythroid differentiation. After confirming the impact of H19 HBG expression, we conducted localization experiments of H19 to lay the foundation for subsequent mechanistic studies.

Materials and methods

Patient sample collection

CD34+ cells were derived from healthy adult blood donors in Guizhou Blood Center; the use of these specimens was approved through the signing of a contract with Guizhou Blood Center.

CD34+ cells were isolated from adult peripheral blood

A blood sample was transferred into a centrifuge tube and an equal volume of pre-prepared DPBS (containing 2% FBS and 1 mM EDTA) was added to the sample. Next, a 50 ml centrifuge tube was prepared and lymphocyte separation medium (Solarbio, Beijing, China) was added to the bottom. The diluted blood was gently layered onto the lymphocyte separation medium to ensure a clear interface. After centrifugation, the supernatant was removed, and the remaining liquid was transferred to a new centrifuge tube. An appropriate volume of DPBS was added, and the cell pellet was obtained after centrifugation. The pellet was then resuspended and transferred into a flow cytometry tube. Selection cocktail (Stem Cell Technologies, Vancouver, Canada) was added to the tube and incubated at room temperature for 10 min. RapidSphere (Stem Cell Technologies, Vancouver, Canada) was mixed and added to this tube, followed by incubation at room temperature for 5 min. DPBS was added to bring the volume up to 2.5 ml. The tube was incubated in an EasySep™ Magnet (Stem Cell Technologies, Vancouver, Canada) at room temperature for 3 min., then the supernatant was poured out. This step was repeated 3–4 times. Finally, the tube was removed from the EasySep™ Magnet and the cells were cultured at a concentration of 5 × 10⁴ cells/mL in a 6-well culture plate.

Cell lines and culture conditions

In this study, K562 cells, CD34+ cells and HUDEP-2 cells were used as models of γ-globin regulation.

The K562 cells were maintained in RPMI-1640 medium (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) after adding 10% fetal bovine serum (FBS) (Zhejiang Tianhang Biotechnology Co., Ltd, Zhejiang, China) and 1% penicillin-streptomycin. Hemin (Solarbio, Beijing, China) was used to induce erythroid differentiation in K562 cells, and the concentration of it was 50 μmol/L.

CD34+ cells were expanded in SFEM II medium (Stem Cell Technologies, Vancouver, Canada) containing 1% penicillin-streptomycin, 50 ng/mL human stem cell factor (SCF), 50 ng/mLFlt3, and 50 ng/mL TPO. For erythroid differentiation, CD34+ cells were cultured in SFEM II medium supplemented with 10 ng/ml interleukin-3 (IL-3, days 7–14), 50 ng/ml human SCF (days 7–18), and 3 U/mL erythropoietin (days 7–21), 1% penicillin-streptomycin.

HUDEP-2 cells were expanded in SFEM II medium, 1 μM dexamethasone, 1 μg/mL doxycycline, 50 ng/mL SCF, 3 U/mL erythropoietin, and 1% penicillin-streptomycin served as supplements for the medium. During erythroid differentiation, the cells were cultured in IMDM (Gibco) supplemented with 2% penicillin-streptomycin, 2% FBS, 10 μg/mL insulin, 3% human serum, 500 μg/mL holo-transferrin, 3 U/mL heparin, 50 ng/mL human SCF (days 0–4), 3 U/mL erythropoietin (days 0–7), and 1 μg/mL doxycycline (days 0–7).

At different stages of CD34+ cells differentiation, cells were collected for Wright-Giemsa staining (Baso Diagnostics Inc., Zhuhai, China), and the morphology of cells and cytoplasmic staining were observed to determine the erythroid differentiation stage.

Cell transfection

According to the results of preliminary experiments, the MOI (multiplicity of infection) for K562 and CD34+ cells was set at 30, while the MOI for HUDEP-2 cells was 3. Lentivirus infection was performed during the logarithmic growth phase of the cells, in a 24-well plate without the addition of penicillin-streptomycin to the infection system. The required volume of virus was calculated based on the MOI value and the number of cells per well. After 24 h, fresh culture medium containing penicillin-streptomycin was replaced. Cell infection was observed under a fluorescence microscope after 2–3 days of infection. To identify cells with stable overexpression of H19, the cells were treated with polybrene (HANBIO) after the cell growth status had become stable. The HBLV-h-H19-Null-ZsGreen-PURO vector used for overexpressing H19 was purchased from HANBIO (Shanghai, China).

Flow cytometry analysis

Erythroid differentiation and maturation were analyzed and compared between overexpression and empty vector groups using flow cytometry. Briefly, transfected cells were induced to differentiate on different days, incubated with fluorescent antibody, and subjected to flow cytometry analysis. PE-conjugated anti-CD71 and APC-conjugated anti-CD235a antibodies were procured from BD Biosciences (Franklin Lakes, NJ, USA).

RNA extraction and reverse transcription quantitative real-time polymerase chain reaction (RT- qPCR)

Total RNA was extracted from K562, CD34+ and HUDEP-2 cells according to the manufacturer’s instructions using the TRIzol method. The primers (H19, BCL11A, HBG, HBB, LIN28B, and GAPDH) were from Shanghai Sangon Biotech Co., Ltd. (Shanghai, China) (Table 1). The extracted RNA was reverse-transcribed into cDNA using the PrimeScript RT reagent kit with gDNA Eraser (Takara). Then, the cDNA samples were amplified by RT-qPCR, performed using TB Green® Premix Ex Taq™ II (Takara) to measure the expression of genes, and GAPDH was employed as an endogenous control to normalize the expression of lncRNA and mRNA. Lastly, qPCR was performed on a CFX Connect^TM Real-Time System (Bio-Rad).

Table 1. Information of the primers.

Gene	Sequences (5′-3′)
H19	F: GGAGTCTGGCAGGAGTGATG R: ACCAAAAGTGACCGGGATGA
BCL11A	F: TATGCCCCGCAGGGTATTTG R: CTTCCGTGTTCGCTTTCTAAGTA
HBG	F: TTCACAGAGGAGGACAAGGCT AC R: GCAGAGGCAGAGGACAGGTT
HBB	F: AGGAGAAGTCTGCCGTTACTG R: CCGAGCACTTTCTTGCCATGA
LIN28B	F: TTGAGTCAATACGGGTAACAG GA R: TGACAGTAATGGCACTTCTTTGG
GAPDH	F: GCACCGTCAAGGCTGAGAAC R: TGGTGAAGACGCCAGTGGA

Western blot analysis

Western blot assay was performed to detect the expression levels of γ-globin and β-globin. Total protein was extracted from cells using 4% SDS (Solarbio) lysate. Equal amounts of protein samples were placed in 15% separation gel (Yazhi Biotechnology Co., Ltd., Shanghai, China) and transferred to a 0.2 µm polyvinylidene fluoride membrane. After that, the membranes were incubated with specific primary antibodies at 4 °C overnight. The primary antibodies used in the study were as follows: β-globin (Abcam), γ-globin (Abcam), and GAPDH (Proteintech Group, Inc., Wuhan, China). Lastly, ECL system (Thermo Fisher Scientific) was used to detect the immunoreactive bands. We obtained the gray value of the target protein bands by analyzing the pictures with ImageJ analysis software (NIH, Bethesda, MD, USA).

Fluorescence in situ hybridization (FISH)

The subcellular localization of H19 in K562 and HUDEP-2 cells was detected using a FISH kit (RiboBio Co., LTD, Guangzhou, China). Briefly, cells were fixed in 4% paraformaldehyde. Pre-hybridization buffer was added to fixed cells after fixation, and cells were hybridized overnight at 37◦C. At last, DAPI was utilized for counterstaining the nuclei of K562 or HUDEP-2 cells. For nuclear fractions and cytoplasmic fractions, U6 probe and 18s probe were used as reference controls. The images were captured using a Fluorescence microscope.

Subcellular fractionation assay

Isolation and extraction of nuclear and cytoplasmic RNA were performed using the PARISTM Kit from Thermo Fisher Scientific. Freshly cultured cells were collected and ice-cold Cell Fractionation Buffer was added for centrifugation. The cytoplasmic fraction was then carefully aspirated away from the nuclear fraction, and ice-cold Cell Disruption Buffer was added to the nuclear pellet. Next, the lysate was mixed with an equal volume of 2X Lysis/Binding Solution and ACS grade 100% ethanol. Finally, Wash Solution and Wash Solution 2/3 were applied to wash the lysate. Expression patterns of H19, GAPDH, and U6 in the nuclear and cytoplasmic fractions were identified using RT-qPCR.

Statistical analysis

The experimental data were analyzed using GraphPad Prism, version 9.3.1 (GraphPad Software, La Jolla, CA). The independent-sample t-test was used for comparison between the term and preterm groups, and the overexpression H19 and empty vector groups. All experiments were repeated more than three times. Data are presented as mean ± standard deviation (x ± SD), and P < 0.05 was considered statistically significant.

Results

The infection rate of cells after overexpression of H19

The overexpression of H19 was achieved by packaging the H19 overexpression lentivirus, which carried the green fluorescent protein (GFP) gene on the viral vector. Therefore, the introduction of the target gene can be indirectly assessed by detecting the expression of the fluorescent protein in the cells, which indicates the percentage of cell infection. The lentivirus was separately used to infect K562 cells (Figure 1A), CD34+ cells (Figure 1B), and HUDEP-2 cells (Figure 1C). After 48–72 h post-infection, the percentage of infected live cells was observed under a fluorescence microscope. From the figures, it can be seen that the infection rate is high and the cells are in good growth condition.

Figure 1. (A-1) (B-1) (C-1) Bright-field sight. (A-2) (B-2) (C-2) Fluorescence sight. (A-3) (B-3) (C-3) Merge of bright-field sight and fluorescence sight.

Effect of overexpression of H19 in K562 cells on HBG and erythroid differentiation-related genes

During erythroid differentiation of K562 cells, HBG and erythroid-related gene expression were detected. Before erythroid differentiation was induced, a portion of cells in the overexpression (OE) and empty vector (CK) groups were collected (day 0). K562 cell differentiation was then induced by hemin, and the cells were collected on days 1–4. As shown in Figure 2, the OE group showed significant up-regulation of H19 expression on days 0–4 compared with the CK group (P < 0.001), suggesting that H19 overexpression was successful. On days 0 and 4, HBG expression was significantly down-regulated in the OE group (P < 0.01); on days 1–3 of erythroid differentiation, HBG expression was down-regulated in the OE group, but not statistically significantly (P ≥ 0.05). Lee et al. demonstrated that LIN28B is involved in the regulation of HBG expression [21]. Thus, following overexpression of H19, we examined the expression level of LIN28B. Contrary to the HBG expression results, LIN28B expression was up-regulated in the OE group compared with the CK group on days 0–3. However, LIN28B expression was significantly down-regulated in the OE group on day 4 (P < 0.01).

Figure 2. Changes in gene expression at the transcriptional level after overexpression of H19 in K562. (A)–(E) show the changes of H19, HBG and LIN28B transcriptome levels on days 0–4 of erythroid differentiation in the CK and H19 OE groups (ns: P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001).

In the process of erythroid differentiation of K562, the expression of various genes is tightly regulated. Within the vast and complex erythroid regulatory network, the expression levels of H19, HBG and LIN28B also undergo changes. By integrating the expression levels of different erythroid differentiation stages for each gene, the longitudinal trends of gene expression during erythroid differentiation can be observed. For H19 (Figure 3A), the expression levels of H19 in both groups decrease overall as the cells undergo erythroid differentiation. Compared to the CK group, the OE group shows a more significant upregulation of H19 after entering erythroid differentiation (days 1–4). For HBG (Figure 3B), the overexpression of H19 in comparison to the CK group generally suppresses the expression of HBG in K562 cells. In the undifferentiated state (day 0), the overexpression of H19 is associated with low HBG expression. From days 1–2 of erythroid differentiation, the inhibitory effect of H19 overexpression on HBG expression is not significant. However, starting from day 3 of erythroid differentiation, the inhibitory effect intensifies and continues until day 4. These results indicate that the overexpression of H19 suppresses the expression of HBG in K562 cells. For LIN28B (Figure 3C), the overexpression of H19 promotes LIN28B expression during erythroid differentiation days 0–3. However, on day 4 of erythroid differentiation, the overexpression group shows a significant downregulation of LIN28B expression. Therefore, during erythroid differentiation, the overexpression of H19 initially promotes LIN28B expression (days 0–3) and then inhibits it (day 4).

Figure 3. Trends In the transcriptional expression of related genes in K562 cells. (A) Trends of H19 in K562 erythroid differentiation day 0–4; (B) Trends of HBG in K562 erythroid differentiation day 0–4; (C) Trends of LIN28B in K562 erythroid differentiation day 0–4.

Effect of overexpression of H19 on HBG in CD34+ cells

CD34+ cells were cultured in two stages: stem cell proliferation from day 0 to day 7 and erythroid differentiation from day 7 to 21(As shown in Figure 4A). On day 18 of CD34+ cell culture, cells from the OE and CK groups were collected to detect the expression of H19, HBG, HBB, and BCL11A using RT-qPCR. As shown in Figure 5A, H19 expression was significantly upregulated in the OE group compared with the CK group (P < 0.001), illustrating the successful overexpression of H19 in CD34+ cells. As shown in Figure 5B, overexpression of H19 significantly upregulated HBG expression (P < 0.001). However, notably, the expression of HBB was also up-regulated in the OE group (P < 0.01) (Figure 5C). In addition, BCL11A expression was down-regulated in the OE group compared with the CK group (P < 0.05) (Figure 5D). Future studies will also consider, besides BCL11A, the expression of other repressors of γ-globin gene transcription.

Figure 4. (A)The culture process of CD34+ cells. (B) The erythroid differentiation process of HUDEP-2 cells.

Figure 5. Transcriptional changes in CD34+ cell culture on day 18. (A)–(D) show the transcriptional changes of H19, HBG, HBB and BCL11A on cell culture day 18 in the CK and H19 OE groups (*P < 0.05; **P < 0.01; ***P < 0.001).

Western blotting was also used to detect the expression of HBG and HBB. As shown in Figure 6, both HBB and HBG were significantly up-regulated in CD34+ cells in the OE group on day 11 (P < 0.05), consistent with the mRNA expression levels. Cells were collected on day 16 and stained for cytomorphological observation; the CD34+ cells in the OE group were slightly larger in size than those in the CK group (Figure 7).

Figure 6. Detection of β-globin and γ-globin expression in CD34+ cells on culture day 11 by Western blotting; analysis of relative expression in the OE-H19 and CK groups (*P < 0.05).

Figure 7. Wright-Giemsa staining of CD34+ cells on day 16 (×400).

Transferrin receptor (CD71) and glycophorin A (CD235a) are two important cell surface markers in the differentiation process of hematopoietic stem/progenitor cells into the erythroid lineage. During the CD34+ cell erythroid-directed differentiation process, the changes in the two membrane surface markers, CD71 and CD235a are as follows: (a) initially, both markers are not expressed, and the cells are CD71 negative/CD235a negative; (b) as erythroid differentiation progresses, CD71 begins to be gradually expressed on the cell membrane, and the cells become CD71 positive/CD235a negative; (c) as differentiation and maturation deepen, the expression of CD235a is activated, resulting in cells becoming simultaneously positive for CD71 and CD235a(Q2); (d) in the terminal differentiation stage of the erythroid lineage, CD71 expression is turned off, and only CD235a is expressed. The cells appear as CD71 negative/CD235a positive, and nucleated cells can be detected. Protein level detection is typically performed on the 14th day of erythroid differentiation, which corresponds to the 21st day of CD34+ cell culture. The expression of CD71 and CD235a was detected by flow cytometry, and the erythroid differentiation and maturity of the OE and CK groups were analyzed and compared. As can be seen in the Figure 8, overexpression of H19 resulted in a lower proportion of cells at the terminal differentiation stage.

Figure 8. Flow cytometry analysis of CD71 and CD235a erythroid markers in CD34+ cells on day 21. The horizontal coordinate represents CD71 and the ordinate represents CD235a.

Effect of H19 overexpression on HBG in HUDEP-2 cells

The erythroid differentiation process of HUDEP-2 cells is divided into two stages: days 0–4 and days 4–7 (Figure 4B). As shown in Figure 9 A, H19 was significantly up-regulated in the OE group at both day 4 and day 7 (P < 0.001). This result indicates successful overexpression of H19 in HUDEP-2 cells. Moreover, overexpression of H19 upregulated HBG expression at the transcriptional level (Figure 9B).

Figure 9. Transcriptional changes in HUDEP-2 cell lines during erythroid differentiation. (A) Changes in H19 expression at days 4 and 7 of erythroid differentiation. (B) Changes in HBG expression at days 4 and 7 of erythroid differentiation (***P < 0.001).

Localization of H19 in K562 and HUDEP-2 cells

Considering that the potential molecular mechanism and biological function of lncRNAs is closely related to their intracellular localization [22], we then investigated the distribution of H19 in K562 and HUDEP-2 cells. After H19 were divided into nucleus and cytoplasm in K562 cells, and the results of which demonstrated that H19 localized preferentially in the nucleus (Figure 10C). However, opposite condition was observed in HUDEP-2, that is, H19 was mainly existed in the cytoplasm (Figure 10D). Similar results were confirmed with the FISH assay (Figure 10A,B). What’s more, LncRNAs located in the cytoplasm have been widely recognized to serve as ceRNAs [23]. Thus, we hypothesize that H19 functions as a mainly cytoplasmic lncRNA and may play a role as a ceRNA to regulate gene expression at the post-transcriptional level.

Figure 10. Subcellular localization of H19. FISH was used to confirm H19 location in K562 (A) and HUDEP-2 (B) cells, using Cy3 probes for H19, DAPI for nuclear staining. RNA was extracted from the nuclear and the cytoplasmic fractions of K562(C) and HUDEP-2(D) cells, and H19 expression of the nuclear and the cytoplasmic fraction was measured by RT-PCR. GAPDH and U6 served as markers for cytoplasm and nucleus, respectively.

Discussion

In this study, H19 overexpression downregulated γ-globin expression in K562 cells during erythroid differentiation but significantly upregulated it in CD34+ and HUDEP-2 cells. Mechanistic studies revealed distinct distributions of H19 in K562 HUDEP-2 cells, shedding light on how H19 impacts γ-globin expression.

H19 is a maternally expressed, paternally imprinted carcinoembryonic gene whose known for its long non-coding RNA product. This lncRNA is crucial in the development of various diseases, including cancer, osteoporosis, cardiovascular disease, digestive system disorders, and organ fibrosis. It also plays a role in maintaining cancer stem cells and promoting osteogenic differentiation of stem cells [11,24–29]. In higher species, the vast majority of lncRNA expression is strongly developmental stage- and tissue-specific, which indicates that lncRNAs have important biological functions [30,31]. In this experiment, H19 was significantly down-regulated in the term group compared with the preterm group. This result suggests that there may be a link between H19 and γ-globin gene expression. In addition to H19, RT-qPCR validation confirmed that LIN28B expression was also significantly down-regulated in the term group. The expression of levels of H19 and LIN28B were simultaneously down-regulated in the term group, indicating that their expression levels are interrelated. LIN28B is an mRNA binding protein. Peng et al. found that there is bidirectional negative feedback between H19 and let-7, and between let-7 and LIN28B, in breast cancer stem cells (BCSCs), where this regulatory lncRNA-miRNA-mRNA loop plays an important role in maintaining the stemness of BCSCs [11]. Furthermore, Ren et al. observed a positive correlation between the abundance of H19 and LIN28B in clinical lung cancer samples; related mechanistic studies showed that H19 up-regulated the expression of LIN28B by sponge adsorption of miR-196b [32]. The regulatory relationship between H19 and LIN28B follows the “lncRNA-miRNA-mRNA” sequence. However, Helsmoortel et al. revealed a regulatory effect of LIN28B on H19 in non-tumor cells for the first time [33]. Regarding β-thalassemia, Yang F et al. reported that hsa-circRNA-100466 is at the center of a ceRNA network that suppressed the transcription of SOX6 by sponging miR-19b-3p, thereby increasing HbF levels [34]. If the well-established regulatory mechanisms of H19 and LIN28B in tumor cells are equally applicable to erythroid cells, such as CD34+ cells, H19 may up-regulate γ-globin expression by up-regulating LIN28B. For this part of the mechanistic study, we completed localization experiments at this stage, and in HUDEP-2 cells, H19 was mainly located in the cytoplasm, in line with our expected results, further suggesting the possibility that H19 participated in the ceRNA network. However, this needs to be confirmed by further studies.

Regarding our three erythroid differentiation models, as there was only a small number of CD34+ cells, and because HUDEP-2 cells show unstable growth and were sometimes difficult to collect in sufficient numbers, we performed LIN28B detection in K562 cells only. The results showed that overexpression of H19 resulted in up-regulation of LIN28B (except on day 4), which is largely consistent with previous findings [33,35]. In the CD34+ model, BCL11A was significantly down-regulated in the OE group compared with the CK group, consistent with a phenotype of up-regulated γ-globin expression; this is also in agreement with existing studies [36]. CD34+ and HUDEP-2 cells show inherent differences from K562 cells: CD34+ cells derived from adult peripheral blood show the greatest similarity to the globin expression pattern in adults, while a major limitation of K562 as an erythroleukemia cell line is that, when induced to differentiate into erythroid cells, hemoglobin express the embryonic/fetal pattern rather than HbA [37,38]. Moreover, in K562 cells, H19 may have a different regulatory pathway from non-tumor CD34+ cells. Additionally, lentiviral multiplicity of infection (MOI) has a toxic effect on cells and disrupts the regulation of their gene expression; thus, we cannot exclude the possibility that lentiviral MOI inhibits HBG expression in K562 cells [39]. There are also reports in the literature of different responses to the same (or different) treatments between CD34+ and K562 [40, 41]. Because CD34+ cells are the closest cell model to physiological erythroid differentiation, we focused on detecting erythroid differentiation of CD34+ cells. The results showed that, in the same differentiation stage, the cell volume was larger in the OE group, and overexpression of H19 slowed down the erythroid differentiation and maturation processes of CD34+ cells. Remarkably, in CD34+ cells, the expression of HBB was also up-regulated in the OE group. Similar results have been reported in the literature on the simultaneous increase of HBG and HBB expression in CD34+ cells; however, the differentiation status of CD34+ cells did not change significantly in the literature studies, which is different from the results of this study [36, 42]. Lastly, the observed phenomenon indicates that the distribution of H19 in K562 and HUDEP-2 cells is distinct, possibly owing to the cellular localization of lncRNAs, which plays a significant role in determining their function. In other words, H19 appears to play opposite roles in these two cell types, suggesting that H19 may have different localizations.

However, our study still has some limitations. For instance, we only used hemin as the inducer for K562 cells and did not consider other γ-globin inducers. Existing references support its effectiveness in inducing γ-globin expression [7, 43, 44], and our laboratory has accumulated experience in using hemin as an inducer to induce erythroid differentiation in K562 cells. There are many different types of γ-globin inducers available, and our choice was primarily based on the experience accumulated by our research group. This is a limitation of our study. Second limitation is the use of only one reference sequence (GAPDH). Although GAPDH is widely used in this field, it is a drawback of our study that we only selected GAPDH as the housekeeping reference sequence. This decision was based on the fact that in our preliminary experiments, the selected housekeeping gene was not regulated. However, this limitation should be acknowledged. The final limitation is the insufficient number of cells collected, which prevented us from conducting protein validation for γ-globin and Lin28B in K562 cells, γ-globin in HUDEP-2 cells, and BCL11A in CD34+ cells. Therefore, we were unable to validate the protein expression levels of these targets.

In summary, despite the limitations of our study, it undeniably demonstrates the effect of H19 overexpression on γ-globin gene expression and the possible regulatory mechanisms based on RT-qPCR validation of the transcriptome sequencing results [10]. This study provides a basis for further study of γ-globin gene expression regulation and the underlying mechanisms. Future research directions may focus on the interaction between H19 and other regulatory factors, as well as the precise regulatory mechanisms during erythroid differentiation. Additionally, exploring the functions and regulatory networks of H19 in other cell types, such as cord blood CD34+ cells, or animal models, may provide a more comprehensive understanding of its role in γ-globin expression regulation.

Acknowledgements

We are grateful to all patients who provided their blood samples.

Disclosure statement

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

Additional information

Funding

This work was supported by the [National Natural Science Foundation of China] under Grant [number 81960040]; [Guizhou Provincial Science and Technology Projects] under Grant [numbers 20165670, 20192808, 20191206, 20205011].The authors declare no competing interests.