Introduction

Diabetes is a metabolic disorder characterized by hyperglycemia and its prevalence has been escalating globally annually1. The administration of hypoglycemic drugs and insulin can alleviate the symptoms of hyperglycemic individuals to an extent, but they are incapable of fundamentally curing diabetes2. The principal factor in the pathogenesis of diabetes is a decline in the quantity and functional impairment of functional pancreatic β cells3. Functional pancreatic β cells in diabetic patients can be reconstructed through three methodologies: pancreas transplantation, islet transplantation, and stem cell transplantation4. Pancreas and islet cell transplantation can replenish damaged islet cells, but due to the deficiency in donor sources and post-transplant immunological rejection issues, their large-scale application is limited5. Stem cell therapy, with its unique biological characteristics, such as robust self-renewal capability, multi-directional differentiation potential, and secretion of diverse bioactive factors, emerges as an ideal approach for controlling blood sugar and achieving islet regeneration6.

The liver is considered an optimal cell source for producing pancreatic cells, as it shares the same origin with the pancreas (both originate from the endoderm), possesses robust proliferation and regeneration capabilities, and has a shared glucose sensing system7. Research has revealed that by using adenovirus, adeno-associated virus, or lentivirus as vectors, transfecting liver cells to alter the expression of key reprogramming genes can induce liver cells to differentiate into IPCs8,9. In addition, by adding small molecule compounds, and suppressing or activating key reprogramming genes, liver cells can also be induced to differentiate into IPCs10,11. However, under current technological conditions, the efficiency of converting liver cells into IPCs remains suboptimal, and the functionality of the induced β cells is relatively immature compared to natural human β cells. Therefore, it is imperative to further elucidate the molecular regulatory mechanism of liver cell reprogramming into IPCs to optimize the induction protocol and enhance the induction efficiency.

The human genome encases less than 2% of DNA that is translated into proteins, with an overwhelming 90% transcribed into non-coding RNA. Long non-coding RNA (lncRNA), typically defined as non-coding RNA transcripts surpassing 200 nt in length, serves as a potential functional molecule within cells, capable of forming intricate structures and interacting with nucleic acids and proteins. A mounting body of evidence suggests that lncRNA serves as a pivotal regulator in gene expression networks by controlling nuclear structure and transcription, and regulating mRNA stability, translation, and post-translational modifications in the cytoplasm12, influencing cell differentiation, stem cell function, and tissue development13. Notably, lncRNA plays a crucial regulatory role in cellular fate determination processes and reprogramming14, regulates pancreatic cell development and function, and is a potential target for enhancing pancreatic cell function and treating type 2 diabetes15. Metastasis Associated Lung Adenocarcinoma Transcript 1 (MALAT1), a lncRNA highly expressed in numerous tissues and regulated during tissue differentiation, can regulate myogenic differentiation and muscle regeneration by modulating MyoD transcriptional activity16. Recent studies have unveiled that lncRNA MALAT1 plays a pivotal role in the differentiation process of induced pluripotent stem cells-derived β cells both in vivo and in vitro. Abnormal overexpression may impair in vitro cell formation and in vivo redifferentiation, while relatively low expression facilitates the transition from stem cell state to differentiation state17. However, the role and mechanism of lncRNA MALAT1 in the process of hepatocyte reprogramming into insulin-secreting cells remain elusive.

MicroRNAs (MicroRNA, miRNA) constitute a category of small non-coding RNA18, which exert pivotal roles in post-transcriptional regulation of gene expression19. They serve as potent regulators in various cellular activities, encompassing cell growth, differentiation, development, and apoptosis20,21. During hepatocyte reprogramming to IPCs, the upregulation or downregulation of miR-124-3p has a demonstrable influence on the efficiency of induction. Suppression of miR-124-3p can augment the expression levels of FoxA2 and Pdx1, thereby facilitating hepatocyte differentiation into IPCs, thereby enhancing the efficiency of induction22. Research indicates that lncRNA MALAT1 mediates proliferation and apoptosis of vascular smooth muscle cells by sponging miR-124-3p positively regulating peroxisome proliferator-activated receptor alpha level23. The PI3K/Akt pathway is one of the most significant pathways identified, extensively participating in cellular physiological activities, and playing a significant regulatory role in diseases such as apoptosis, growth, differentiation, protein translation, etc24. Dysfunction of phosphoinositide 3-kinase family (PI3Ks) may lead to glucose homeostasis imbalance, causing an increase in glucose serum levels, while hyperglycemia is the most critical pathophysiological feature of diabetes25. Studies have found that exosomes derived from pioglitazone-pretreated mesenchymal stem cells activate the PI3K/AKT/eNOS pathway to promote angiogenesis in human umbilical vein endothelial cells, thereby accelerating diabetic wound healing26. miRNAs play a direct or indirect regulatory role in the PI3K/Akt signaling transduction process. By inhibiting specific miRNAs, the activity of PI3K/Akt signal transduction can be enhanced, thereby improving the survival rate of β cells27. This mechanism provides new perspectives and strategies for cell differentiation. Researchers have discovered that kaempferol can promote Sox2 expression, suppress miR-124-3p expression, and activate the PI3K pathway, thereby promoting osteogenic differentiation of bone marrow mesenchymal stem cells28. Research has found that the PI3K signaling transduction pathway plays a crucial regulatory role in the differentiation process of pancreatic β cells29,30. Activation of PI3K/Akt is vital for the differentiation of pancreatic duct cells into insulin-producing cells31.

Therefore, we speculate that during the process of hepatocyte chemical reprogramming to IPCs, lncRNA MALAT1 may influence PI3K expression by regulating miR-124-3p, thereby regulating the induction and differentiation of hepatocytes. However, the specific regulatory mechanism remains to be further explored. A thorough understanding of its regulatory mechanism is essential to enhance the efficiency of hepatocyte induction differentiation and improve the maturity of IPCs.

Materials and methods

Culture and Preparation of primary mouse hepatocytes

Primary mouse hepatocytes were isolated according to the previously described22. The protocol was approved by the Animal Care and Use Committee of Jiangxi Medical College, Nanchang University (Protocol number: 23W008). All experiments were performed in accordance with relevant guidelines and regulations. Cells were seeded onto 100-mm plates containing Dulbecco’s modified Eagle’s medium (DMEM) (Thermo Fisher Scientific, 10567014) with 10% fetal bovine serum (VivaCell, C04001), under 5% CO2 at 37 °C. Primary mouse hepatocytes utilized in the experiments had undergone fewer than two passages.

Differentiation of primary mouse hepatocytes into IPCs and transfection

6 × 105 primary mouse hepatocytes were seeded into 60-mm plates, which were supplemented with knockout-DMEM (Thermo Fisher Scientific, 12660012) containing the main components, 2% knockout-serum(Thermo Fisher Scientific, 10828028), 1 mM β-mercaptoethanol(Thermo Fisher Scientific, 21985023), 1% L-Glutamine(gibco), 1% B-27(Thermo Fisher Scientific, 17504044), 1% N-2(Thermo Fisher Scientific, 17502048), 1% non-essential amino acids(Gibco, 11140050), 20 ng/ml bFGF(PeproTech, 450 − 33), 20 ng/ml EGF(PeproTech, 315-09), and treated with, 5µM Decitabine (Medchemexpress, HY-A0004) for 24 h, then 100ng/ml Trichostatin A (Medchemexpress, HY-15144) for another 48 h, 2µM retinoic acid (Medchemexpress, HY-14649) plus 1X ITS (Gibco, 41400045) for 7 days, and 10mM nicotinamide (Solarbio, IN0150)for 7 days22.

Transfection of primary mouse hepatocytes was performed according to the manufacturer’s protocols of Lipofectamine 3000 (Invitrogen, L3000008). The lncRNA MALAT1 overexpressed plasmid and si-RNA originated from Genechem, and normal induction commenced post-overnight transfection.

RNA preparation, real-time quantitative PCR, and RNA sequencing

All primers were synthesized by Sangon Biotech, according to PrimerBank (https://pga.mgh.harvard.edu/primerbank/, accessed on 30 April 2024) and included in Supplemental Table S1. Total RNA was isolated from cells using Trizol reagents and cDNA was generated using RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific, K1622). Quantitative PCR using SYBR qPCR Mix (Vazyme Biotech, Q711-02)according to the protocols provided by manufacturers. Each sample was amplified in triplicates and normalized with β-Actin signals. For the expression analysis of miR-124-3p, 5 ng of total RNA was reverse-transcribed using miRNA 1 st Strand cDNA Synthesis Kit (Vazyme Biotech, MR101-01) according to the manufacturer’s protocols. Quantification of miR-124-3p was evaluated using the miRNA Universal SYBR qPCR Master Mix(Vazyme Biotech, MQ101-01). U6 was used as a normalization control. Relative expression levels were calculated using the comparative threshold cycle value method (2 − ΔΔ Ct). RNA extracted from the various stages cells and reactions were conducted on the ABI Real-time PCR.

Western blot analysis

Total proteins were extracted from cells in RIPA lysis buffer (Applygen, C1055) containing a 1% protease inhibitor cocktail and 1 mM PMSF. Whole-cell lysates were collected and quantified using the BCA Kit (Applygen, P1511-3) and mixed with 5×SDS-PAGE Protein Sample Loading Buffer (Applygen, B1012-5). Equal amounts of protein were loaded and resolved on 10% SDS polyacrylamide gels followed by transfer onto PVDF membranes, which were then blocked with 5% non-fat milk in Tris-buffered saline and 0.1% tween-20 (TBST) for 1 h at room temperature and then incubated with their corresponding primary antibodies, FoxA2(1:1000, 8186 T, CST), PI3K(1:1000, 4292 S, CST), PDX1(1:1000, TA321081, OriGene), and β-actin(1:2000, TA811000, OriGene), overnight at 4 °C. After washed with TBST and then incubated with their corresponding HRP-linked secondary antibodies for 1 h at room temperature. Protein bands were visualized by ECL and quantified using Image J software (ij154-win-java8, National Institutes of Health, Bethesda, MD, USA; here is the software download address: https://imagej.net/ij/download.html).

Immunofluorescence staining

Primary mouse hepatocytes were plated on glass coverslips, once cell density approximated 90% or differentiated, they were fixed with 4% paraformaldehyde for 10 min at room temperature and washed with PBS. The cells were permeabilized with 0.5% Triton X-100 for 20 min, washed with PBS, and blocked with PBST containing 1% BSA for 40 min. Subsequently, the cells were incubated with primary antibodies (CK18: Origene, TA500015S, 1:300; ALB: CST, 4746 T, 1:300; PDX1: Abcam, EPR22002, 1:500; Insulin: CST, 3014, 1:300; C-peptide: CST, 4593, 1:300) overnight at 4 °C. Fluorescent dyes iFluo 488 or iFluo 594 were applied, slides were mounted with medium with DAPI (Solarbio, S2110), and observations were made under a fluorescence microscope.

Luciferase report assay

TargetScan 7.2 database (http://www.targetscan.org/vert_72/) was used to predict binding sites between miR-124-3p and lncRNA MALAT, PI3K. HEK293T cells were seeded at 3 × 104/well in 48-well plates and cotransfected with 20pmol of miR-124-3p mimic, or miR-NC, and 200 ng of pRL-TK, 1 µg of lncRNA MALAT1-3’UTR-WT, lncRNA MALAT1-3’UTR-Mut, PI3K-3’UTR-WT, PI3K-3’UTR-Mut(all plasmid DNAs were purchased from hanbio) using Lipofectamine 3000 Reagent (Invitrogen, L3000008) and TurboFect-DNA Mix(Thermo Fisher Scientific, R0533) according to the manufacturer’s protocol. After 48 h transfection, luciferase activity was measured using the dual-luciferase reporter assay system (Promega).

In vitro insulin secretion assay

The insulin levels in culture supernatants were measured using a mouse Insulin ELISA (enzyme-linked immunosorbent assay) Kit (Finebio, EH0374). After pre-incubation with Krebs-Ringer Bicarbonate HEPES(KRBH)for 90 min, the differentiated cells were incubated with KRBH with either 5.5 mM or 25 mM glucose for 1 h. Following the collection of conditioned supernatants, analysis was performed.

Streptozotocin-Induced T1D mouse model

A total of 20 C57BL/6 male mice (purchased from gempharmatech) underwent experimentation under approval from Jiangxi Medical College, Nanchang University’s Animal Care and Use Committee (Protocol number: 23W008), and are consistent with ARRIVE guidelines. All experiments were performed in accordance with the relevant guidelines and regulations. The diabetic state was induced through a single intraperitoneal injection of 150 mg/kg STZ in a sodium citrate buffer at pH 4.5 (Yuanye, S17049)32. FBG levels were assessed three days post-STZ administration. Mice with FBG ≥ 16.7 mmol/L were identified as T1D. They were then divided into four groups: Control; T1DM, T1DM + IPCs (500 IEQ); T1DM + IPCs (Si-MALAT1) (500 IEQ). Anesthesia (isoflurane, Sigma-Aldrich, 792632) ensured proper kidney exposure, cells were transplanted into the renal capsule using a screw-drive syringe (26-gauge needle), and the injection site was carefully positioned to avoid vascular damage. Post-transplantation, the site was cauterized, and the fascia and skin were sutured for recovery. The surgical incision was disinfected with povidone-iodine. The mice were placed on a heating pad until they regained consciousness. After awakening, Meloxicam (Medchemexpress, HY-B0261, 4 mg/kg/d) was subcutaneously injected for 2 days.

FBG and serum biochemistry analysis

FBG levels were assessed every 3–4 days simultaneously. Post-centrifugation serum at 4 °C for 15 min at 3500×g was analyzed for fasting insulin via ELISA kit (Meimianbio, MM-0579M1). Lipid profiles including TC, TG, HDL-C, and LDL-C were evaluated with the Tokyo Hitachi automated analyzer.

Data analysis

Quantitative data were analyzed using GraphPad Prism 5 software (San Diego, CA, USA). All results were presented as mean ± SEM and analyzed using one-way ANOVA with the Sidak post hoc test. p < 0.05 was considered a significant difference.

Results

Isolation and characterization of primary mouse hepatocytes

Firstly, we isolated and characterized the primary mouse hepatocytes from the liver of newborn C57 mice within 48 h after birth. Post-isolation, as displayed in Fig. 1 A, it was observed that the majority of cells had a circular or nearly circular shape, with cells seemingly embracing one another, and the nucleus delineated. Simultaneously, glycogen PAS staining identified hepatic cells (Fig. 1B, C). The cytoplasm of the cells displayed a purplish red coloration, while the nucleus exhibited a blue tint, affirming that the isolated cells were hepatic cells laden with glycogen granules. We also utilized immunofluorescence assays to evaluate the characteristics of primary mouse hepatocytes. The results showed that primary mouse hepatocytes expressed high expression of the hepatocyte markers ALB and CK18 (Fig. 1D, E).

Stepwise methodology to induce primary mouse hepatocytes into IPCs

The next step was to differentiate the primary mouse hepatocytes using our step-wise protocol, and a schematic presentation of the protocol is depicted in Fig. 2 A. And as illustrated in Fig. 2B, in fundamental cultivation conditions, primary mouse hepatocytes (S0) exhibit a typical polygonal shape with a large and centralized nucleus. In Stage 1 (S1), liver cells progressively lose their original morphological features, transforming into a flattened spindle shape, accompanied by an increase in cell fragments. Throughout Stage 2 (S2), cells undergo further differentiation and commence clustering to form pancreatic pro-precursor cells. In Stage 3 (S3), the cell population significantly reduces, but cell-to-cell interactions become dense, yielding IPCs.

Successful induction by pancreatic progenitor cells (S2) was indicated by the upregulation of PDX1 expression. In contrast, downregulated expression of Sox9, Nestin, and FoxA2 was detected at the end of the second induction phase, compared to pluripotent progenitor cells (Fig. 2C). After further induction of differentiation in the third stage, significant upregulation of insulin gene expression was observed in IPCs(S3), and immunofluorescence staining showed that the third-stage cells expressed insulin and C-peptide (Fig. 2D).

Fig. 1
figure 1

Isolation and characterization of primary mouse hepatocytes from the liver of newborn C57 mice. (A) primary mouse hepatocytes morphotype under bright-field microscopy. Scale bar = 200 μm. (B, C) PAS staining of glycogen in primary mouse hepatocytes. Scale bar = 100 μm and 50 μm. (D, E) Immunofluorescence staining for ALB and CK18 (hepatocyte markers) expression in primary mouse hepatocytes. Scale bar = 200 μm and 50 μm.

Full size image
Fig. 2
figure 2

Primary mouse hepatocytes differentiation into IPC. (A) Schematic presentation of the step-wise differentiation protocol used to generate IPCs from primary mouse hepatocytes. (B) Photographs of primary mouse hepatocytes differentiated into IPCs. Scale bar = 400 μm and 200 μm. (C) Identifying PDX1 cells using immunofluorescence staining with PDX1 antibody and expression of β-cell development-related genes. (D) Detection of insulin-producing cells using immunofluorescence staining with insulin and C-peptide antibodies and expression of the insulin gene. (n = 3, **P < 0.01, ***P < 0.001).

Full size image
Fig. 3
figure 3

LncRNA sequencing analysis of various stages cells. (A) LncRNAs are significantly differentially expressed during reprogramming. (B) Venn plots of significantly differentially expressed lncRNAs. (C) Volcano plots (S1-vs-S0) demonstrating significantly differentially expressed lncRNAs. (D) Heat maps of lncRNAs significantly differentially expressed during reprogramming. (E) KEGG functional analysis of mRNA target genes with co-expressed lncRNAs.

Full size image
Fig. 4
figure 4

Expression of lncRNA MALAT1/miR-124-3p/PI3K and genes associated with pancreatic development during primary mouse hepatocytes differentiation into IPCs. (A-D) RT-qPCR showing expression of lncRNA MALAT1/miR-124-3p/PI3K and FoxA2 (n = 3, **p < 0.01, ***p < 0.001). (C) Western blot results show significantly higher protein expression levels of PDX1 in primary mouse hepatocytes after induction (n = 3, *p < 0.05, **p < 0.01).

Full size image
Fig. 5
figure 5

Prediction of lncRNA MALAT1/miR-124-3p/PI3K complementary sites and dual luciferase reporter gene assay. (A) Bioinformatics analysis revealing the presence of complementary binding sites for mature miR-124-3p in the 3’-untranslated region (3’UTR) of PI3K mRNA as well as in lncRNA MALAT1 (987–993 bp). (B, C) Luciferase reporter assay showing a significant reduction in luciferase activity in the 293 T cell line overexpressing miRNA-124-3p, regardless of the presence of the luciferase gene 3’UTR carrying lncRNA MALAT1(987–993 bp) or the PI3K mRNA 3’UTR(2866–2872 bp). (**P < 0.01, ***P < 0.001, ns means no statistically significant differences between the two groups).

Full size image
Fig. 6
figure 6

Over-expression of lncRNA MALAT1’s influence on associated gene expression during the process of differentiation induction is examined. (A) qPCR measures the alteration in lncRNA MALAT1 expression following over-expression plasmid transfection. (B, C) qPCR assesses miR-124-3p, PI3K, and pancreatic development-related gene expression in dedifferentiated cells. (D, E) In vitro glucose (5.5mM and 25mM, IPCs stage, per 106 cells) stimulation experiments are conducted. (*P < 0.05, **P < 0.01, ***P < 0.001, ns means no statistically significant differences between the two groups).

Full size image
Fig. 7
figure 7

Downregulating of lncRNA MALAT1’s influence on associated gene expression during the process of differentiation induction is examined. (A) qPCR measures the alteration in lncRNA MALAT1 expression following interfering plasmid transfection. (B, C) qPCR assesses miR-124-3p, PI3K, and pancreatic development-related gene expression in dedifferentiated cells. (D) qPCR was used to evaluate Insulin expression in IPCs cell phase. (E) Western blotting was conducted to detect FoxA2 expression in dedifferentiated cells. (F, G) In vitro glucose (5.5mM and 25mM, IPCs stage, per 106 cells) stimulation experiments are conducted. (*P < 0.05, **P < 0.01, ***P < 0.001, ns means no statistically significant differences between the two groups).

Full size image

LncRNA-seq analysis of various stages cells

We further extracted the RNA from various stages cells, and performed a high throughput RNA-seq analysis, as shown in Fig. 3 A-E, upon dissecting differentially expressed mRNA and lncRNA, an impressive number of 86 significantly different lncRNAs were discovered. A comparison of various gene groups revealed that the S1/S0 group contained the most lncRNA variations. Among all detected lncRNAs, MALTA1 demonstrated the highest expression level, with an average RPKM of 641.3, notable for its upregulation in the S1 phase. Further, a KEGG pathway functional enrichment analysis was conducted on the potential target genes regulated by lncRNAs, revealing that the target genes predominantly enriched in metabolic pathways and signaling pathways such as MAPK, PI3K/Akt, Ras, Hippo, etc33.

Differentiation of primary mouse hepatocytes into IPCs was associated with LncRNA MALAT1/miR-124-3p/PI3K axis

To explore the role played by lncRNA MALAT1 in the generation of IPCs from primary mouse hepatocytes, we examined its mRNA expression along with miR-124-3p and PI3K at various stages. Firstly, as shown in Fig. 4 A-C, we discovered that lncRNA MALAT1, miR-124-3p, and PI3K are significantly upregulated during the de-differentiation stage of hepatic cells. In the process of reprogramming into pancreatic progenitor cells, all three expressions are notably decreased. During the IPCs stage, the expressions of lncRNA MALAT1 and PI3K are notably elevated, whereas the expression of miR-124-3p remains unaltered, compared to the pancreatic progenitor cell stage. Furthermore, as shown in Fig. 4D and E, we noted an apparent upregulation of FoxA2, a gene associated with pancreatic development, during hepatic cell de-differentiation and a dramatic downregulation at the pancreatic progenitor cell stage. PDX1 protein expression exhibits a notable rise during de-differentiation, followed by an even higher expression in the pancreatic progenitor cell stage.

miR-124-3p exerts a negative effect on expression of LncRNA MALAT1 and PI3K

The subsequent step was to gene sequence alignment using bioinformatics software. As shown in Fig. 5 A, it revealed the presence of complementary binding sites for mature miR-124-3p in the 3′-untranslated region (3′-UTR) of PI3K mRNA as well as in the lncRNA MALAT1 sequence. A luciferase reporter assay showed that luciferase activity was reduced significantly in the HEK293T cell line overexpressing miRNA-124-3p, regardless of the presence the 3’UTR of luciferase gene carrying the lncRNA MALAT1 sequence or the PI3K mRNA 3′UTR (Fig. 5B, C). The experimental data suggest targeted silencing of PI3K mRNA and lncRNA MALAT1 expression by miRNA-124-3p. These results also indicate that in the presence of a special sequence structure, there is a competition between lncRNA MALAT1 and PI3K for binding to miR-124-3p.

Effect of overexpression LncRNA MALAT1 on differentiation of hepatocytes into IPCs

To explore the effects of lncRNA MALAT1 overexpression on hepatocytes reprogramming to IPCs via miR-124-3p/PI3K, we first transduced hepatocytes with overexpression plasmids to upregulate lncRNA MALAT1. As shown in Fig. 6 A, The results demonstrated that after 48 h post-transduction of hepatocytes with overexpression plasmids, the expression of lncRNA MALAT1 was approximately 18-fold higher than in the empty vector group, indicating successful transduction of overexpression plasmids into hepatocytes. Subsequently, cells were subjected to the established induction culture protocol for reprogramming to the IPCs stage, and finally verified through qPCR and glucose stimulation insulin release experiments. The qPCR results revealed that during dedifferentiation, compared with the Vector group, the expression of miR-124-3p was significantly upregulated in the MALAT1 group, while the expressions of PI3K and PDX1 were downregulated (Fig. 6B, C). The results of the glucose stimulation insulin release experiment showed that regardless of low or high glucose concentration stimulation, the expression of insulin in the MALAT1 group was significantly lower than that in mouse β cells (Fig. 6D, E).

Effect of inhibiting LncRNA MALAT1 expression on differentiation of hepatocytes into IPCs

By transfecting hepatocytes with a knockdown plasmid to suppress lncRNA MALAT1 expression and reprogram them to the IPCs stage, we elucidate the impact of reducing lncRNA MALAT1 on the process of reprogramming through miR-124-3p/PI3K. Post transfection of a knockdown plasmid into hepatocytes, lncRNA MALAT1 expression was significantly lower than in the control group, indicating successful transfection of the knockdown plasmid into hepatic cells and suppression of lncRNA MALAT1 expression (Fig. 7 A). As shown in Fig. 7B C, the qPCR results revealed that during dedifferentiation, compared to the NC group, the expression of miR-124-3p in the siMALAT1 group was significantly downregulated, while the expressions of PI3K and PDX1 were markedly upregulated, and the expressions of Sox9 and FoxA2 were significantly elevated. WB results also demonstrated that the protein expression of FoxA2 in the siMALAT1 group was higher than in the NC group (Fig. 7E). Moreover, as shown in Fig. 7D, the qPCR results indicated that at the IPCs stage, the insulin expression in the siMALAT1 group was significantly higher than in the NC group. When stimulated by low glucose concentration, the insulin expression in the siMALAT1 group was significantly higher than in the NC control group(Fig. 7 F). When stimulated by high glucose concentration, the insulin expression in the siMALAT1 group was not only significantly higher than in the NC group but also comparable to the level in mouse β cells (Fig. 7G).

Therapeutic effects of IPCs in the T1D animal model

We examined whether IPCs transplantation into a mouse T1D model would improve blood glucose levels. The therapeutic efficacy of transplanted cells was assessed using T1DM models as controls for the IPCs and IPCs (Si-MALAT1). Fasting blood glucose (FBG) levels significantly increased to 16.7 mmol/L in the T1DM model compared with those in normal controls, confirming the T1DM establishment. Further, compared with T1DM mice, T1D mice receiving IPCs and IPCs (Si-MALAT1) showed decreased FBG levels following transplantation (Fig. 8A). Next, Compared to the T1DM group, the IPCs and IPCs(Si-MALAT1) group reduced serum TC, TG, and LDL-C levels, while elevating HDL-C levels (Fig. 8B). Insulin concentration was shown to significantly decrease in the T1DM group compared with that in normal controls. Further, the insulin concentration measured at the end of the experiment indicated that insulin levels were higher in the IPCs (Si-MALAT1) group than in the T1DM group (Fig. 8C). Histological examination of pancreatic tissue from control mice revealed an intact islet, normal pancreatic structure, and orderly arrangement. STZ-induced diabetes caused inflammatory cell invasion with severe degeneration of islet cells, and irregular atrophic acini. Conversely, histological analysis of diabetic mouse pancreas in the IPCs and IPCs (Si-MALAT1) group demonstrated gradual normalization of islet cell morphology (Fig. 8D). Collectively, these results confirmed that IPCs transplantation can help reduce blood glucose levels.

Fig. 8
figure 8

Functional analysis after IPCs transplantation in T1DM mice. (A) Fasting blood glucose levels of mice, glucose levels were monitored using tail-vein blood samples. (B, C) Serum TC, TG, HDL-C, LDL-C, and insulin levels at the end of the experiment. (n = 5, *p < 0.05 versus control group, #p < 0.05 versus T1DM group, respectively). (D) Histological section of pancreas tissue. (E) Immunofluorescence staining of β-cell markers (Insulin) and α-cell markers (Glucagon) in pancreas tissue. (F)Immunofluorescence staining of β-cell markers (Insulin and PDX1) in transplanted kidney grafts, and statistical results of the overlapping fluorescence intensity. All fluorescently stained cells were observed under ×20 objective, and representative images are shown. (n = 5, *p < 0.05 T1DM + IPCs group, respectively).

Full size image

The pancreas tissue of mice was isolated at the end of the experiment, β-cell markers (Insulin) and α-cell markers (Glucagon) were identified using immunofluorescence staining (Fig. 8E). In IPCs mice, the insulin staining was observed, and insulin levels were higher than in IPCs (Si-MALAT1) group mice. Therefore, we concluded that the transplanted IPCs (Si-MALAT1) have better cell function.

Further, the kidneys of mice in the IPCs group and the IPCs (Si-MALAT1) group were isolated. Immunofluorescence staining was employed to identify β-cell markers, namely Insulin and PDX1 (Fig. 8F). In the kidneys of IPCs mice where transplantation had been carried out, IPCs were found, accompanied by low-level insulin staining. IPCs were also detected in the kidneys of IPCs(Si-MALAT1) mice, and the insulin levels were notably higher compared to those in IPCs mice. Based on these findings, we can conclude that the transplanted IPCs (Si-MALAT1) exhibit superior cell survival and functionality.

Discussion

In both type 1 and type 2 diabetes, a plethora of genetic and environmental factors can trigger the progressive loss of β cell quantity and/or function, ultimately manifesting as high blood sugar34,35. Presently, there has been substantial progress in the management of type 1 and type 2 diabetes. Nevertheless, traditional diabetes treatments, primarily based on insulin injection and islet transplantation, present numerous drawbacks. Hence, researchers are exploring novel strategies to reproduce insulin-producing cells in vitro. The most promising approach involves generating β cells derived from stem cells, which can provide an infinite source of insulin-secreting cells36. Compared to other cell sources, hepatic cells exhibit enhanced safety and feasibility, circumventing ethical and immune rejection issues that may arise from embryonic stem cells or allogenic cells7. Genetic reprogramming and chemical reprogramming techniques offer effective methods for hepatic cell transdifferentiation into insulin-secreting cells37. Genetic reprogramming technology, particularly through the expression of specific combinations of transcription factors, has been extensively utilized in the study of cell fate transition38. In the process of hepatic cell transdifferentiation into insulin-secreting cells, researchers have successfully achieved hepatic cell transdifferentiation by introducing a series of transcription factors associated with pancreatic β cell development and function, such as Pdx1, Ngn3, MafA, etc39. These transcription factors can activate the expression of pancreatic β cell-specific genes and suppress the expression of hepatic cell-specific genes, thereby inducing the transformation of hepatic cells into insulin-secreting cells. On the other hand, chemical reprogramming technology regulates cell fate through the use of specific small-molecule chemical compounds40. These compounds can simulate or interfere with intracellular signaling pathways, thereby inducing cell transdifferentiation41. In the study of hepatic cell reprogramming into insulin-secreting cells, researchers have successfully achieved hepatic cell transdifferentiation by screening and optimizing the combination of small-molecule chemical compounds42. This method offers the advantages of simplicity and low cost of operation, and avoids the risks associated with gene manipulation.

This investigation employed a combinatorial approach by incorporating various small molecule compounds to progressively induce liver cells into IPCs. Through the transformation of liver cells into IPCs, we observed an array of noticeable cellular biological alterations. The morphology of liver cells metamorphosed consistently reflecting the attributes of IPCs; concurrently, alterations in the expression levels of specific genes indicated a shift in cell fate. More significantly, these reprogrammed cells exhibited insulin-secreting functions unique to islet cells, further validating their potential as IPCs. The conversion of liver cells into insulin-secreting cells is a complex process that involves the regulation of multiple transcription factors, signaling pathways, and epigenetic mechanisms43. During the chemical reprogramming of liver cells into IPCs, through high-throughput sequencing analysis and qPCR detection, we discovered that lncRNA MALAT1, miR-124-3p, and PI3K displayed differential expression at different stages of reprogramming. LncRNAs and epigenetic modifications are pivotal components of transcriptional mechanisms controlling cell specification and development44. LncRNAs play a crucial role in regulating target gene expression, making them indispensable epigenetic regulators across various cell types45. A growing body of evidence indicates that lncRNAs are potent regulators of β-cell function, and their dysregulation is linked to diabetes46. Prior studies have identified over 1000 lncRNAs in human islets through comprehensive epigenetics analysis, indicating their importance in endocrine and β-cell differentiation programs during pancreatic development47. Notably, lncRNA MALAT1 is considered a diagnostic and therapeutic target for diabetes. LncRNA MALAT1 can suppress pancreatic and duodenal homeobox 1 (PDX1) promoter H3 histone acetylation to reduce PDX1 expression, thereby inducing β cell dysfunction by inhibiting insulin secretion48. The ablation of lncRNA MALAT1 can inhibit the activity of c-Jun amino terminal kinase, activate insulin-induced insulin receptor substrate 1 activity, and upregulate protein kinase B (PKB, also known as Akt) phosphorylation levels, thereby improving insulin sensitivity49. It has been reported that ablation of lncRNA MALAT1 can activate antioxidant gene expression regulated by nuclear factor erythroid 2-related factor 2, reduce ROS accumulation and oxidative stress, thereby reducing inflammation, enhancing sensitivity to insulin signal transduction, and improving β-cell function49. Understanding the regulatory role of lncRNA MALAT1 in pancreatic development is critical for deciphering the regulatory network controlling islet development.

Long non-coding RNA (lncRNA) and microRNA (miRNA) stand as two exemplary classes of non-coding RNA, each known for its pivotal regulatory role in the orchestration of numerous biological processes50. Investigations have revealed that lncRNA MALAT1 orchestrates human umbilical vein endothelial cell (HUVEC) proliferation and apoptosis via the has-miR-124-3p/NR3C2 and/or has-miR-135a-5p/NR3C2 axis51. KEGG pathway functional enrichment analysis indicates that lncRNA target genes are predominantly enriched within metabolic pathways and signaling cascades such as MAPK, PI3K/Akt, Ras, and Hippo, amongst others. A mounting body of evidence suggests that PI3K signal transduction constitutes a pivotal determinant of β cell characteristics and functionality. Modulation of the PI3K pathway, either directly or indirectly (via miRNA interactions), can shape the progression of β cells, ultimately determining the fate of β cell quality27. To elucidate the function and molecular mechanism of the LncRNA MALAT1/miR-124-3p/PI3K regulatory axis during the process of hepatocyte reprogramming into IPCs, this study initially manipulated lncRNA MALAT1 expression in hepatocytes by transfecting overexpression or knockdown plasmids. Subsequently, the hepatocytes overexpressing or silencing lncRNA MALAT1 were chemically reprogrammed to the dedifferentiation stage. The results indicated that silencing lncRNA MALAT1 in the hepatocyte stage could stimulate PI3K expression during the dedifferentiation stage while concurrently suppressing miR-124-3p expression. Conversely, overexpression of lncRNA MALAT1 significantly elevated miR-124-3p expression during the dedifferentiation stage while simultaneously inhibiting PI3K expression. Given the intimate relationship between insulin production and secretion and the PI3K signaling pathway, alterations in PI3K expression may influence insulin levels. This discrepancy from the conventional mechanism of Lnc RNA suppressing micro RNA expression may be due to the extended duration in the process of inducing differentiation, with gene expression exhibiting a dynamic and changing trend. Nevertheless, this trend is relatively consistent during the final IPCs phase.

Through analysis via WB and qPCR, we uncovered the regulatory interplay of the LncRNA MALAT1/miR-124-3p/PI3K axis during reprogramming. This axis appears intimately associated with the expression of pancreatic development-associated genes FoxA2 and PDX1. FoxA2 and PDX1 represent crucial transcription factors in pancreatic development and β-cell function maintenance, and also are endodermal markers during the induction of IPCs52,53. They are not only involved in early pancreatic development but also sustain the characteristics and functions of β cells in mature pancreas54. Research indicates that a decrease in FoxA2 expression during the differentiation of induced pluripotent stem cells into islets can diminish α- and β-cell mass55. FoxA2 is considered to be one of the upstream regulators of PDX1, and activation of PDX1 expression and initiation of pancreatic programming necessitate FoxA256. PDX1 is indispensable for β-cell differentiation, being pivotal for the expansion of pancreatic primordium and the development of endocrine islets57,58. The protein encoded by PDX1 can transcriptionally activate multiple genes, such as insulin, glucose kinase, somatostatin, and islet amyloid polypeptide, all of which are essential for regulating glucose metabolism59,60. Mice lacking PDX1 develop diabetes, and humans harboring PDX1 mutations suffer from monogenic diabetes61,62. Moreover, ablation of the complex conditions of FoxA1 and FoxA2 in pancreatic primordium results in complete loss of PDX1 expression and severe pancreatic developmental deficiency63. Therefore, an increase in their expression may be an essential step in the reprogramming of liver cells into insulin-secreting cells. Furthermore, the IPCs were transplanted into the kidney capsule of mice, we confirmed that IPCs secreted insulin in recipient mice, drastically lowering blood glucose levels compared to control T1DM mice. In addition, we observed significant cell survival and function in the IPCs (Si-MALAT1) group.

Our research revealed that silencing lncRNA MALAT1 enhances the expression of FoxA2 and PDX1 through miR-124-3p/PI3K during dedifferentiation. While an upregulation in FoxA2 and PDX1 expression is a positive signal, we must validate if this indeed elevates the efficacy of hepatocyte reprogramming to IPCs whilst ascertaining the homeostasis of insulin secretory function of these reprogrammed cells. Thus, we followed our predefined small molecule compound-induced culture regimen to chemically reprogram silenced lncRNA MALAT1 liver cells into IPCs, ultimately examining the intracellular insulin expression by qPCR and ex vivo glucose stimulation insulin release experiments. The results unveiled that silencing lncRNA MALAT1 remarkably boosted insulin expression in IPCs, enhancing differentiation efficiency. The experimental results indicate that transfection of lncRNA MALAT1 alters the expression levels of miR-124-3p and PI3K genes in the final IPCs phase, and regulation of primary mouse hepatocytes directed β islet-like cell differentiation efficiency.