Abstract

Background

   Bisphosphonate-related osteonecrosis of the jaw (BRONJ) is a serious
   complication associated with bisphosphonate (BP) therapy. Enhancement
   of the osteogenic differentiation of human jaw bone marrow mesenchymal
   stem cells (JBMMSCs) is a key issue in the treatment of BRONJ. In this
   study, we investigated the role and mechanism of LINC01013 in
   regulating osteogenic differentiation of JBMMSCs.

Methods

   Osteogenic differentiation of JBMMSCs was assessed in vitro using
   alkaline phosphatase (ALP), alizarin red staining (ARS), and western
   blotting. JBMMSCs transplanted into the backs of nude mice were used to
   detect JBMMSCs osteogenesis in vivo. Molecular mechanisms involved in
   JBMMSCs osteogenesis were evaluated using real-time fluorescence
   quantitative polymerase chain reaction, western blotting, fluorescence
   in situ hybridization, RNA pull-down, and RNA-seq.

Results

   Homeobox C8 (HOXC8) knockdown enhanced ALP activity, ARS, and
   expression of bone sialoprotein and osteocalcin in JBMMSCs under normal
   and BP stimulation conditions. HOXC8 negatively regulated LINC01013
   expression. LINC01013 enhanced JBMMSCs osteogenic differentiation
   impaired by BP stimulation. Furthermore, LINC01013 regulated the
   expression of inflammation-related genes in JBMMSCs under BP
   conditions. LINC01013 formed a complex with ILF3. Two isoforms of ILF3
   (NF90 and NF110) promoted the osteogenic differentiation of JBMMSCs
   under normal and BP conditions, depending on their nuclear
   localization. Additionally, NF90, which is located in the nucleus,
   inhibited the expression of NLR family pyrin domain containing 3
   (NLRP3).

Conclusions

   In summary, HOXC8 negatively regulates LINC01013 to inhibit osteogenic
   differentiation of JBMMSCs under BP conditions. We also further
   clarified that LINC01013 binding to ILF3 affects ILF3 nuclear
   localization to regulate JBMMSCs osteogenic differentiation and
   regulates NLRP3/Caspase-1 pathway to affect JBMMSCs function under BP
   stimulation.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s13287-025-04467-3.

   Keywords: LINC01013, Human jaw bone marrow mesenchymal stem cells,
   Bisphosphonate-related osteonecrosis of the jaws, Osteogenesis, ILF3

Introduction

   Bisphosphonates (BPs) are widely used in the treatment of bone-related
   conditions, however, BPs are associated with side effects. BP-related
   osteonecrosis of the jaw (BRONJ) is a serious complication of BP
   treatment that often occurs after dental procedures and is commonly
   associated with inflammatory dental diseases [[34]1]. Currently, BRONJ
   is difficult to cure with no optimal treatment options except
   nonsurgical treatments such as local irrigation or surgical removal
   [[35]2]. Several hypotheses, including the inhibition of bone
   remodeling, angiogenesis, and immunosuppression, underlying BRONJ
   pathogenesis have been proposed [[36]1].

   Early BP research focused on inhibiting osteoclast differentiation;
   however, recent studies suggest that BP affects bone marrow mesenchymal
   stem cells (BMSCs) function [[37]3, [38]4]. BP has a
   concentration-dependent dual effect on BMSCs; lower concentrations of
   BP promote osteogenesis in BMSCs, and higher concentrations exert
   inhibitory effect [[39]5–[40]8]. Innovative strategies, including the
   introduction of vascular endothelial growth factor and geranylgeraniol,
   have been explored to counteract BP-induced impairments in cellular
   differentiation [[41]9, [42]10]. However, the specific molecular
   mechanism by which BP inhibits jaw bone marrow mesenchymal stem cells
   (JBMMSCs) osteogenic differentiation remains unknown.

   BRONJ exclusively affects the jaws and not the long bones [[43]11]. The
   osteogenic differentiation potential of axial and jaw bone BMSCs
   differ, with jaw BMSCs having a higher osteogenic differentiation
   potential [[44]11]. Axial and jaw bone BMSCs are also differentially
   sensitive to BP due to their different origins [[45]12]. Identifying
   site-specific gene expression differences is pivotal for understanding
   the pathology of BRONJ. The HOX gene family is differentially regulated
   in jaw-derived and long-bone-derived BMSCs [[46]13]. Homeobox C8
   (HOXC8) inhibits the osteogenic differentiation of odontogenic stem
   cells [[47]14]. Its expression in jaw-derived BMSCs is lower than that
   in long-bone-derived BMSCs [[48]13]. However, the role of HOXC8 in
   JBMMSCs differentiation and function in BRONJ has not yet been
   investigated.

   The human genome is extensively transcribed into RNA, with < 2% of the
   transcriptome translated into proteins [[49]15–[50]17]. This highlights
   the prevalence of non-coding RNAs in the cellular landscape. Long
   non-coding RNAs (lncRNAs) are central players in cell biology and play
   essential regulatory roles in various processes including cell
   differentiation and development [[51]18]. Altered lincRNAs expression
   contributes to the development of bone metabolic diseases [[52]19,
   [53]20]. In our previous study, we demonstrated that HOXC8 directly
   targets the LINC01013 promoter and regulates stem cell differentiation
   by modulating LINC01013 expression [[54]21]. In addition, LINC01013 is
   involved in diverse regulatory roles in various biological processes
   and diseases: LINC01013 promotes cell invasiveness in anaplastic
   large-cell lymphoma [[55]22], regulates cell stemness in liver cancer
   cells [[56]23], improves DNA repair to alleviate endothelial cell
   dysfunction [[57]24], and regulates the development of fibrosis in
   calcific aortic valve disease [[58]25]. However, the role of LINC01013
   as a direct downstream target of HOXC8 in the regulation of JBMMSCs
   function remains unclear.

   With the development of proteomic and epigenetic technologies, the
   interactions between lncRNAs and RNA-binding proteins (RBPs) have
   gained considerable attention. Interleukin enhancer binding factor 3
   (ILF3) is a double-stranded RNA-binding protein that has a 90 kDa
   isoform (NF90) and a 110 kDa isoform (NF110) [[59]26]. ILF3 is
   implicated in a wide array of physiological functions, including viral
   infection response, mRNA transport, translation control, transcription,
   and cell cycle regulation [[60]27]. Despite these insights, the
   complete spectrum of ILF3 involvement in osteogenesis, particularly in
   pathological conditions such as osteoporosis or BRONJ, remains poorly
   understood. Therefore, this study aimed to determine the role of ILF3
   in BRONJ.

   This study aims to bridge these gaps by exploring the roles of HOXC8
   and LINC01013 in the osteogenic differentiation of JBMMSCs within a
   BP-stimulated environment. We sought to decipher the intricate
   regulatory relationship between these molecules under conditions
   mimicking BRONJ and, to understand how they influence JBMMSCs behavior
   and osteogenic potential in response to BP treatment. By focusing on
   this BRONJ-simulated scenario, this study elucidated the molecular
   dynamics of osteogenic differentiation under pathological conditions,
   offering insights into potential therapeutic targets and strategies to
   mitigate or reverse the detrimental effects of BP treatment on bone
   health.

Materials and methods

JBMMSCs culture

   The experiments using human JBMMSCs were performed according to the
   ISSCR Guidelines for the Conduct of Human Embryonic Stem Cell Research.
   Alveolar bone tissue was obtained from the Beijing Stomatological
   Hospital, Capital Medical University, and the study was approved by the
   Ethics Committee of the China Rehabilitation Research Center
   (CMUSH-IRB-KJ-PJ-2023-47). JBMMSCs were cultured in mesenchymal stem
   cell medium (ScienCell, Carlsbad, California) at 37 °C and 5% CO[2], as
   previously described [[61]28]. JBMMSCs at passages 3–5 was used for
   subsequent experiments. JBMMSCs were treated with 5 µM zoledronic acid
   (Zol, Sigma). JBMMSCs were inoculated into osteogenic medium at a cell
   density of 2.0 × 10^5 cells/well as previously described [[62]14].

Viral infection

   The lentiviruses were obtained from GenePharma (Suzhou, China). The
   complementary DNA of human LINC01013, NF90, and NF110 was subcloned
   into an LV5 lentiviral vecto. Overexpression vectors for NF90 and NF110
   nuclear localization signal (NLS) deletion mutants (NF90/NF110ΔNLS)
   were constructed. Short hairpin RNA (shRNA) of HOXC8, LINC01013, and
   ILF3 were cloned into the LV2 lentiviral vector. After overnight
   incubation of JBMMSCs in 100 mm Petri dishes, the cells were exposed to
   lentivirus-containing medium for 12 h. The transfected cells were
   selected using puromycin (2 µg/mL) for 3 days. The shRNA target
   sequences are listed in Table [63]S1.

Western blotting analysis

   Protein was extracted and separated following previously published
   protocols [[64]29]. The primary antibodies used were as follows: ILF3
   (1:2,000, 19887-1-AP, Proteintech, Wuhan, China), ILF3 (1:2,000, A2496,
   Abclonal, Wuhan, China), HOXC8 (1:1,000, Wuhan, Proteintech), NLR
   family pyrin domain containing 3 (NLRP3, 1:1000, A5652, Abclonal,
   Wuhan, China), interleukin-1β (IL-1β, 1:1000, 12703 S, Cell Signaling
   Technology, Danvers, USA), Caspase-1 (1:1000, 22915-1-AP, Proteintech,
   Wuhan, China), osteocalcin (OCN; 1:1,000, A20800, Abclonal, Wuhan,
   China), bone sialoprotein (BSP; 1:1,000, bs-2668R, Bioss, Beijing,
   China), and GAPDH (1:5,000, Sigma-Aldrich, USA).

Real-time fluorescence quantitative polymerase chain reaction (RT-qPCR)

   JBMMSCs were treated with TRIzol reagent (Invitrogen). Total mRNA was
   isolated using an Ultrapure RNA Kit (CW0581, CWBIO, Taizhou, China) and
   reverse-transcribed into cDNA, following the manufacturer’s
   instructions. RT-qPCR was performed using the MagicSYBR Mixture PCR kit
   (CWBIO, Taizhou, China), following the manufacturer’s instructions. The
   PCR primer sequences are listed in Table [65]S2.

Alkaline phosphatase (ALP) and Alizarin red staining (ARS) analysis

   Following five days of osteogenic induction, the collected cells were
   analyzed for ALP activity using an ALP activity kit (Sigma-Aldrich),
   following the manufacturer’s instructions. JBMMSCs were fixed with 70%
   ethanol for 30 min and stained with 2% alizarin red (Sigma-Aldrich) two
   weeks after osteogenic induction.

RNA pull-down assays

   RNA pull-down assays were performed as previously described [[66]30].
   Briefly, the collected JBMMSCs were extracted and transferred into
   centrifuge tubes. The cell lysates were mixed with in vitro-transcribed
   RNAs to form RNA-protein complexes. The complexes were purified using
   beads. Finally, bead-bound proteins were detected by western blotting
   after protein isolation from the beads.

Animal models

   The nude mice (female, 8-week-old) were purchased from the Bejing
   Stomatological Hospital Animal Laboratory (Beijing, China). All animals
   were maintained on a normal light/dark cycle and were free to eat and
   drink. The nude mice were randomly divided into 7 groups with Vector
   group, LINC01013 group, Consh group, LINC01013sh group, Vector group,
   Vector + Zol group and LINC01013 + Zol group, with 5 mice in each
   group. JBMMSCs were incubated with hydroxyapatite-tricalcium phosphate
   (HA/TCP) for 3 h at 37 °C as previously described [[67]14]. The nude
   mice were intraperitoneally anesthetized using 1% pentobarbital sodium
   (40 mg/kg). Subsequently, the resulting mixture was transplanted
   subcutaneously into nude mice. Three months after transplantation, the
   nude mice were euthanized by cervical dislocation after isoflurane
   (RWD, China) inhalation. The transplanted tissue was removed for
   hematoxylin and eosin (HE), Masson’s trichrome and immunofluorescence
   staining. All animal experiments followed the guidelines of the Animal
   Care and Use Committee of the Capital Medical University (approval
   number: KQYY-202309-006, Beijing, China). The work has been reported in
   line with the ARRIVE guidelines 2.0.

Histological and Immunofluorescence staining

   The transplanted tissue was carefully removed after three months and
   fixed in 4% paraformaldehyde. Afterwards, the tissues were decalcified
   for four weeks in a solution containing 10% EDTA, followed by
   dehydration and paraffin embedding. Section (5 μm) were stained with
   hematoxylin and eosin (HE) and Masson’s trichrome. For
   immunofluorescence staining, after dewaxing and rehydration, hydrated
   tissue sections were subjected to citrate antigen retrieval.
   Subsequently, the tissues were blocked with goat serum at 37 °C for
   30 min. The tissue was incubated with primary antibodies against OCN
   (1:200; bs-0470R, Bioss, Beijing, China), NLRP3 (1:200; A5652;
   Abclonal, Wuhan, China), ILF3 (1:200; 19887-1-AP; Proteintech, Wuhan,
   China), and Caspase-1 (1:200; 22915-1-AP; Proteintech, Wuhan, China)
   overnight. The secondary antibody used was Alexa Fluor 594 goat
   anti-rabbit IgG (1:500, Life Technologies, USA). The nuclei were
   stained with DAPI.

RNA-seq and bioinformatic analysis

   Total RNA was isolated from JBMMSCs-LINC01013sh and JBMMSCs-Consh.
   RNA-seq and bioinformatics analyses were performed at SHANGHAI BIOTREE
   BIOTECH Co., LTD (Shanghai, China). The negative logarithm of the p
   value (− LgP) was calculated to represent the correlation between gene
   expression and Gene Ontology (GO) functions or pathways.

Fluorescence in situ hybridization (FISH)

   The FISH probe of LINC01013 was obtained from GenePharma. The FISH
   assay was performed according to the guidelines provided in the FISH
   Kit Manual (GenePharma). Briefly, cells were fixed in 4%
   paraformaldehyde and hybridized with probe mixture solution (4 µM).
   Nuclei were stained with DAPI.

Statistical analyses

   Statistics were calculated using GraphPad Prism 9.0 software.

   Comparisons between groups were performed using Student’s t-test or
   one-way analysis of variance. Results are presented as mean ± standard
   deviation, and p-values less than 0.05 were considered statistically
   significant.

Results

HOXC8 Inhibition promoted JBMMSCs osteogenic differentiation under normal and
Zol-stimulated conditions

   To evaluate the effects of HOXC8 on osteogenesis, HOXC8 shRNA
   lentivirus was used to inhibit HOXC8 expression in JBMMSCs. Western
   blotting confirmed HOXC8 knockdown in JBMMSCs (Fig. [68]1A).
   Correspondingly, HOXC8 downregulation enhanced ALP activity, indicating
   increased osteoblastic function in JBMMSCs (Fig. [69]1B). This was
   further validated by ARS analysis, which demonstrated that HOXC8
   knockdown significantly enhanced JBMMSCs mineralization (Fig. [70]1C).
   Additionally, western blotting analysis revealed an increased
   expression of BSP and OCN upon HOXC8 knockdown, indicative of
   osteoblastic marker upregulation (Fig. [71]1D). RT-qPCR and western
   blotting analysis confirmed HOXC8 upregulation in JBMMSCs treated with
   Zol (Fig. [72]1E, F). Under Zol treatment, ALP activity and ARS
   increased osteogenic activity and mineralization in JBMMSCs with HOXC8
   knockdown (Fig. [73]1G, H). These results indicate that HOXC8
   suppression under Zol stimulation significantly promotes JBMMSCs
   osteogenic differentiation.

Fig. 1.

   [74]Fig. 1
   [75]Open in a new tab

   HOXC8 inhibited Zol-stimulated JBMMSCs osteoblastic function. (A) HOXC8
   protein expression level in JBMMSCs. (B) HOXC8 knockdown enhanced ALP
   activity in JBMMSCs. (C) ARS demonstrating that HOXC8 knockdown
   elevated mineralization in JBMMSCs. (D) Western blotting analysis of
   BSP and OCN expression. (E) RT-qPCR and (F) western blotting showing
   increased HOXC8 expression following Zol-treatment. (G) HOXC8 knockdown
   promoted ALP activity in JBMMSCs during Zol-treatment. (H) ARS
   demonstrating that HOXC8 knockdown enhanced mineralization in
   Zol-treated JBMMSCs. Data are represented as mean ± standard deviation
   (n = 3), and *P < 0.05, **P < 0.01(Student’s t-test or one-way analysis
   of variance). Full-length blots/gels are presented in Figure [76]S1

LINC01013 enhanced JBMMSCs osteogenic differentiation

   Our previous results suggest that LINC01013 is a downstream target gene
   of HOXC8 [[77]21]. RT-qPCR revealed a significant increase in LINC01013
   mRNA levels in HOXC8 silenced JBMMSCs, suggesting that HOXC8 negatively
   regulates LINC01013 expression (Fig. [78]2A). A time-dependent study
   showed a consistent increase in LINC01013 expression over 7 days after
   osteogenic induction (Fig. [79]2B), suggests that LINC01013 may
   regulate JBMMSCs osteogenic differentiation. Therefore, we verified the
   effects of LINC01013 on the osteogenic differentiation potential of
   JBMMSCs. The overexpression of LINC01013 was confirmed by RT-qPCR
   (Fig. [80]2C). LINC01013 knockdown was confirmed using RT-qPCR
   (Fig. [81]2D). LINC01013 overexpression markedly increased ALP activity
   and mineralization in JBMMSCs (Fig. [82]2E, G). Furthermore, ALP and
   ARS activities indicated a significant decrease in osteogenic activity
   and mineralization in JBMMSCs with LINC01013 knockdown (Fig. [83]2F,
   H). We also detected the expression of BSP and OCN in LICN01013
   overexpressed and knockdown JBMMSCs. Western blotting analysis showed a
   significant increase in BSP and OCN expression in JBMMSCs
   overexpressing LINC01013, indicating enhanced osteogenic
   differentiation one week after induction (Fig. [84]2I). A reduction was
   observed in BSP and OCN expression in JBMMSCs with LINC01013 knockdown
   (Fig. [85]2J).

Fig. 2.

   [86]Fig. 2
   [87]Open in a new tab

   LINC01013 enhanced JBMMSCs osteogenic differentiation. (A) RT-qPCR
   demonstrating the enhancement of LINC01013 in HOXC8-silenced JBMMSCs
   compared to controls. (B) A time-dependent study revealed a consistent
   augmentation in LINC01013 expression over a 7-day period under
   osteogenic inducing conditions. (C) LINC01013 overexpression efficiency
   validated with RT-qPCR. (D) RT-qPCR confirming LINC01013 knockdown. (E)
   ALP activity assays and (G) ARS exhibit a notable increase in
   mineralization of LINC01013 overexpressing JBMMSCs. (F) ALP activity
   assays and (H) ARS results demonstrated a notable decrease in
   mineralization in LINC01013 knockdown JBMMSCs. (I, J) Western blotting
   analysis of BSP and OCN expression. (K) HE and (L) Masson staining. B
   points to the formation of bone-like tissue. HA points to
   hydroxyapatite. Arrows represent newly formed bone-like tissue. (M)
   Immunofluorescence staining for OCN (red) indicates enhanced osteogenic
   activity in LINC01013-overexpressing JBMMSCs, with weaker signals in
   LINC01013-knockdown groups. (N, P) Quantitative analysis of mineralized
   areas and (O, Q) OCN expression levels in vivo. Data are represented as
   mean ± standard deviation (n = 3), and *P < 0.05, **P < 0.01,
   ***P < 0.001(Student’s t-test or one-way analysis of variance).
   Full-length blots/gels are presented in Figure [88]S2

   Next, we determined whether LINC01013 affects the osteogenesis of
   JBMMSCs in vivo. HE (Fig. [89]2K, N, P) and Masson staining
   (Fig. [90]2L) revealed that bone mineralized tissue regeneration in the
   JBMMSCs-LINC01013 transplanted tissues was greater than that in the
   control group. In addition, the degree of bone mineralized tissue
   regeneration in the JBMMSCs-LINC01013sh group was lower than that in
   the control group. These findings suggests that bone formation is
   reduced in the absence of LINC01013. Moreover, immunofluorescent
   staining (Fig. [91]2M, O, and Q) showed that LINC01013 overexpression
   enhanced OCN expression, whereas LINC01013 knockdown reduced OCN
   expression. These results suggest that LINC01013 plays an essential
   role in osteogenic differentiation promotion and bone formation in
   JBMMSCs.

Restorative effects of LINC01013 overexpression on Zol-treated JBMMSCs

   To investigate whether LINC01013 could rescue the impaired function in
   Zol-treated JBMMSCs, LINC01013-overexpressing JBMMSCs were exposed to
   Zol. Zol treatment decreased the LINC01013 mRNA levels (Fig. [92]3A).
   We then evaluated the effect of LINC01013 on the osteogenic
   differentiation potential of Zol-induced JBMMSCs. ALP activity and ARS
   analyses revealed that LINC01013 overexpression partially restored the
   osteogenic potential that was reduced by Zol stimulation (Fig. [93]3B,
   C), confirming the role of LINC01013 in osteogenic differentiation
   under Zol treatment.

Fig. 3.

   [94]Fig. 3
   [95]Open in a new tab

   LINC01013 rescued osteogenic differentiation of Zol-stimulated JBMMSCs.
   (A) RT-qPCR analysis showing Zol decreased LINC01013 expression in
   JBMMSCs. (B) ALP activity assays showing that overexpression of
   LINC01013 partially restored ALP activity in Zol-stimulated cells. (C)
   ARS staining illustrating reduced mineralized nodule formation in
   JBMMSCs treated with Zol, with partial recovery observed in the
   LINC01013-overexpression group. (D) H&E and (F) Masson staining
   demonstrating that LINC01013 overexpression under Zol stimulation could
   restore the osteogenic ability of JBMMSCs. B points to the formation of
   bone-like tissue. HA points to hydroxyapatite. Arrows represent newly
   formed bone-like tissue. (H) Immunofluorescence staining of OCN (red)
   showing significantly reduced osteogenic activity in Zol-treated
   JBMMSCs, with increased signals in the LINC01013 + Zol group. (E, G)
   Quantitative analysis of mineralized areas and (I) OCN expression
   levels in vivo. Data are represented as mean ± standard deviation
   (n = 3), and *P < 0.05, **P < 0.01, ***P < 0.001 (Student’s t-test or
   one-way analysis of variance)

   Furthermore, HE and Masson staining provided a clear view of the
   restored tissue architecture in JBMMSCs overexpressing LINC01013 under
   Zol treatment (Fig. [96]3D-G). The cells displayed a more organized,
   denser, and structured extracellular matrix compared to the Zol-treated
   controls. Additionally, immunofluorescence staining revealed increased
   OCN expression in JBMMSCs overexpressing LINC01013, even after Zol
   stimulation (Fig. [97]3H, I). These results indicated that LINC01013
   may counteract the adverse effects of Zol on JBMMSCs and support the
   restoration of osteogenic potential in bone tissue.

Identification of differentially expressed mRNAs in JBMMSCs-LINC01013sh and
bioinformatic analysis of RNA-seq data

   To further investigate the mechanism by which LINC01013 regulates the
   osteogenic differentiation of JBMMSCs, RNA-seq and bioinformatic
   analyses were performed on LINC01013 knockdown JBMMSCs. RNA-seq results
   showed that 1,438 mRNAs were differentially expressed (∣log2FC∣>1,
   FDR < 0.05) in the JBMMSCs-LINC01013sh and consh groups; 1,124 mRNAs
   were upregulated and 314 were downregulated (Table [98]S3). Five
   differentially expressed genes, including mesenteric estrogen dependent
   adipogenesis (MEDAG), vitrin (VIT), matrix Gla protein (MGP), gasdermin
   C (GSDMC), and iodothyronine deiodinase 2 (DIO2), were selected from
   the sequencing results to validate the accuracy and credibility of the
   RNA-seq results. RT-qPCR results (Fig. [99]4A, B) showed MEDAG, VIT,
   MGP, GSDMC, and DIO2 upregulation in the JBMMSCs-LINC01013sh group.
   Conversely, the overexpression group exhibited MEDAG, VIT, MGP, GSDMC,
   and DIO2 downregulation. These results are consistent with those of
   RNA-seq. To further clarify the potential functions of the
   differentially expressed genes, GO enrichment and KEGG pathway analyses
   were performed. The top 20 signaling pathways enriched in the set of
   genes included the calcium and PI3K-AKT signaling pathways
   (Fig. [100]4C). GO functional analysis revealed that the differentially
   expressed genes correlated with binding proteins, inflammatory
   response, immune response, and I collagen-containing extracellular
   matrix (Fig. [101]4D).

Fig. 4.

   [102]Fig. 4
   [103]Open in a new tab

   RNA-seq data analysis of LINC01013 knockdown JBMMSCs. (A) RT-qPCR
   analysis showing significantly decreased mRNA levels of MEDAG, VIT,
   MGP, GSDMC, and DIO2 in LINC01013-overexpressing JBMMSCs compared to
   the vector control. (B) Increased expression of mRNA levels of MEDAG,
   VIT, MGP, GSDMC and DIO2 in LINC01013 knockdown JBMMSCs. (C) KEGG
   pathway enrichment analysis of differentially expressed genes
   highlights pathways significantly affected by LINC01013 knockdown. (D)
   GO enrichment analysis visualizing biological processes influenced by
   LINC01013. Data are represented as mean ± standard deviation (n = 3),
   and *P < 0.05, **P < 0.01 (Student’s t-test)

LINC01013 overexpression inhibited inflammation-related genes expression in
JBMMSCs

   Further bioinformatic analysis of the differentially expressed genes
   focused on understanding the specific inflammatory pathways influenced
   by LINC01013. To investigate the anti-inflammatory mechanism of
   LINC01013, we used RT-qPCR and western blotting to quantify
   transcriptional and translational changes in key inflammasome
   components (Caspase-1, NLRP3, and IL-1β). The RT-qPCR and western
   blotting results showed that LINC01013 overexpression led to a decrease
   in inflammation-related genes including Caspase-1, NLRP3, and IL-1β,
   suggesting that LINC01013 potentially exerts anti-inflammatory effects
   (Fig. [104]5A-D). LINC01013 overexpression in Zol-treated JBMMSCs
   resulted in decreased levels of these markers, further indicating the
   suppressive effect of LINC01013 on inflammasome-related genes.

Fig. 5.

   [105]Fig. 5
   [106]Open in a new tab

   LINC01013 regulated inflammation-related genes expression in
   Zol-treated JBMMSCs. (A-C) RT-qPCR analysis showing significantly
   increased mRNA levels of NLRP3 (A), Caspase-1 (B), and IL-1β (C) in
   Zol-treated JBMMSCs, while these mRNA levels decreased in LINC01013
   overexpressing JBMMSCs. (D) Western blotting analysis demonstrating
   elevated protein levels of NLRP3, Caspase-1, and IL-1β in Zol-treated
   JBMMSCs, which were reversed in the LINC01013 expression group.
   Immunofluorescence staining of NLRP3 (red) (E) and Caspase-1 (red) (F)
   in JBMMSCs overexpressing LINC01013 and the control JBMMSCs under Zol
   treatment. Nuclei (DAPI) are in blue. Data are represented as
   mean ± standard deviation (n = 3), and **P < 0.01 (one-way analysis of
   variance). Full-length blots/gels are presented in Figure [107]S3

   Immunofluorescence staining was performed to visualize NLRP3 and
   Caspase-1 in JBMMSCs after Zol treatment. Following Zol treatment,
   NLRP3 and Caspase-1 levels increased, indicative of inflammasome
   activation. Notably, JBMMSCs overexpressing LINC01013 subjected to Zol
   exposure exhibited a marked reduction in NLRP3 and Caspase-1 levels,
   indicating the negative regulatory influence of LINC01013 on the
   inflammasome machinery (Fig. [108]5E, F).

LINC01013 formed an RNA-protein complex with ILF3

   The functions of lncRNAs are closely associated with their subcellular
   localization [[109]31]. Thus, we employed RNA-FISH to detect the
   subcellular localization of LINC01013 in JBMMSCs. The results revealed
   the predominant nuclear presence of LINC01013 in JBMMSCs
   (Fig. [110]6A). After identifying the subcellular localization of
   LINC01013, predicting the proteins interacting with it was essential
   for gaining a deeper understanding of its molecular mechanisms. Through
   bioinformatics prediction of potential LINC01013-binding proteins, the
   results indicated that LINC01013 may form a protein complex with ILF3
   (Fig. [111]6B). To validate this prediction, we used RNA pull-down
   assay. The results demonstrated the interaction between LINC01013 and
   ILF3 (Fig. [112]6C). NF90 and NF110 are primarily DNA- and RNA-binding
   proteins that play significant roles in the regulation of various RNA
   metabolism processes [[113]27]. Notably, Zol stimulation did not
   influence the transcription and translation of NF90 or NF110, as
   demonstrated by RT-qPCR and western blotting (Fig. [114]6D, E).

Fig. 6.

   [115]Fig. 6
   [116]Open in a new tab

   LINC01013 formed a complex with ILF3. (A) RNA FISH was utilized to
   determine LINC01013 RNA molecules within the cells. The localization of
   18s ribosomal RNA is shown as a positive control for the FISH
   technique. Nuclei (DAPI) are in blue. (B) Predictive analyses showing
   an interaction between LINC01013 and ILF3. (C) Western blotting
   validated the presence of ILF3 protein in cell lysates treated with an
   antisense probe against LINC01013, confirming the interaction at the
   protein level. RT-qPCR (D) and western blotting (E) revealed that Zol
   treatment did not change the levels of ILF3 in JBMMSCs after Zol
   treatment. Data are represented as mean ± standard deviation (n = 3),
   and ns, not significant (Student’s t-test). Full-length blots/gels are
   presented in Figure [117]S4

NF90 and NF110 promoted JBMMSCs osteogenic differentiation

   To detect the effect of NF90 and NF110 on osteogenic differentiation in
   JBMMSCs. ILF3 was silenced using lentiviral ILF3 shRNA in JBMMSCs, and
   the knockdown efficiency was verified by western blotting
   (Fig. [118]7A). The corresponding ALP and ARS activities indicated a
   significant decrease in osteogenic activity and mineralization in ILF3
   knockdown JBMMSCs (Fig. B, C). Western blotting revealed reduced BSP
   and OCN expression in ILF3 knockdown JBMMSCs one week after induction
   (Fig. [119]7D). To test the effect of NF90 and NF110 on the osteogenic
   differentiation of JBMMSCs, we constructed NF90- and
   NF110-overexpressing JBMMSCs, respectively. The efficiency of NF90 and
   NF110 overexpression was validated using western blotting
   (Fig. [120]7E). The cells were then cultured in osteogenic induction
   medium. NF90 and NF110 overexpression markedly increased ALP activity
   and mineralization in JBMMSCs, indicating their positive roles in
   osteogenic differentiation (Fig. [121]7F, G). Western blotting showed
   increased BSP and OCN expression in JBMMSCs overexpressing NF90 and
   NF110, indicating enhanced osteogenic differentiation one week after
   induction (Fig. [122]7H). Then, we evaluated the potential of NF90 and
   NF110 in rescuing the osteogenic differentiation of Zol-stimulated
   JBMMSCs. ALP and ARS activity revealed that NF90 and NF110
   overexpression partially restored the osteogenic potential that was
   reduced by Zol stimulation (Fig. [123]7I, J).

Fig. 7.

   [124]Fig. 7
   [125]Open in a new tab

   NF90 and NF110 enhanced osteogenic differentiation of JBMMSCs under
   normal and Zol conditions. (A) Western blotting confirming the ILF3
   knockdown in JBMMSCs. (B) ILF3 knockdown decreased ALP activity in
   JBMMSCs. (C) ARS results showed that ILF3 knockdown repressed the
   mineralization capacity of JBMMSCs. (D) Western blotting revealing a
   reduction in the levels of BSP and OCN in ILF3 knockdown JBMMSCs. (E)
   Western blotting confirming the overexpression efficiency of ILF3
   variants NF90 and NF110. (F) ALP activity assay showing increased ALP
   activity in JBMMSCs overexpressing NF90 and NF110. (G) ARS results
   indicating enhanced mineralization in NF90 and NF110 overexpressing
   JBMMSCs. (H) Western blotting results showing an increase in the levels
   of BSP and OCN in NF90 and NF110 overexpression JBMMSCs. ALP (I) and
   ARS (J) results showed that NF90 and NF110 overexpression increased the
   mineralization of JBMMSCs under Zol stimulation. Data are presented as
   mean ± standard deviation (n = 3), and *P < 0.05, **P < 0.01 (Student’s
   t-test or one-way analysis of variance). Full-length blots/gels are
   presented in Figure S5

Zol inhibited osteogenesis via NF90 and NF110 nuclear translocation
alteration

   ILF3 function correlates with its intracellular localization [[126]32,
   [127]33]. Under basal conditions, ILF3 was predominantly localized
   within the nucleus, as evidenced by immunofluorescence staining. After
   Zol exposure, ILF3 translocated from the nucleus to the cytoplasm,
   suggesting that Zol treatment alters intracellular distribution of ILF3
   (Fig. [128]8A). Furthermore, we determined the effect of LINC01013 on
   the nuclear-to-cytoplasmic ratio of ILF3. JBMMSCs overexpressing
   LINC01013 exhibited different responses to Zol treatment. The
   persistence of ILF3 within the nuclei of these cells implied that
   LINC01013 overexpression could modulate the intracellular trafficking
   of ILF3 (Fig. [129]8A).

Fig. 8.

   [130]Fig. 8
   [131]Open in a new tab

   NF90 and NF110 regulated JBMMSCs osteogenic differentiation through
   intracellular localization. (A) Immunofluorescence analysis showing the
   localization of ILF3 in JBMMSCs under control, Zol treatment, and
   LINC01013 overexpression conditions. Nuclei were stained with DAPI
   (blue). (B) A schematic representing the NLS within the ILF3 protein
   and its variants, NF90 and NF110. (C) Western blotting showing the
   expression of Flag-tagged ILF3 and their NLS-deleted mutants (NF90ΔNLS,
   NF110ΔNLS) in JBMMSCs. (D) Immunofluorescence staining for Flag (red)
   showing nuclear localization of NF90 and NF110, while the NLS-deleted
   mutants exhibited reduced nuclear retention. (E-G) The ALP activity
   assay (E), ARS (F), and Western blotting (G) showed that NF90
   overexpressing JBMMSCs had a significantly higher osteogenic
   differentiation capacity compared to controls, whereas NF90ΔNLS
   overexpressing JBMMSCs had a reduced osteogenic differentiation
   capacity compared to NF90 overexpressing JBMMSCs. (H-J) ALP activity
   (H), ARS staining (I), and western blotting (J) demonstrated enhanced
   osteogenic differentiation in NF110 overexpressing JBMMSCs compared to
   controls, while NF110ΔNLS reduced differentiation relative to NF110.
   (K) Western blot analysis of NLRP3 expression levels in NF90 and
   NF90ΔNLS overexpressing JBMMSCs. (L) The protein level of NLRP3 in
   NF110 and NF110ΔNLS overexpressing JBMMSCs. (M) Western blot analysis
   showing the levels of NLRP3 in NF90 and NF90ΔNLS overexpressing JBMMSCs
   under Zol treatment. Data are presented as mean ± standard deviation
   (n = 3), and *P < 0.05, **P < 0.01 (one-way analysis of variance).
   Full-length blots/gels are presented in Figure S6-7

   To determine whether the NLS sequences are critical for the function of
   NF90 and NF110 in promoting osteogenesis, a mutant of NF90 and NF110
   lacking the NLS sequences (NF90/NF110ΔNLS) was used [[132]33]
   (Fig. [133]8B). Western blotting confirmed the overexpression efficacy
   of NF90, NF110, and NF90/NF110ΔNLS (Fig. [134]8C). To assess the
   subcellular localization of the NF90/NF110ΔNLS mutant,
   immunofluorescence staining was performed, which confirmed its
   cytoplasmic localization (Fig. [135]8D). Then, we assessed the
   osteogenic differentiation of JBMMSCs overexpressing NF90, NF110 and
   NF90/NF110ΔNLS. ALP activity assay, ARS, and western blotting results
   showed that NF90/NF110ΔNLS overexpression attenuated the potential of
   NF90/NF110 to promote JBMMSCs osteogenic differentiation
   (Fig. [136]8E-J).

   We also investigated the potential regulatory role of ILF3 in NLRP3
   expression in JBMMSCs. Western blotting showed that only NF90, which is
   located in the nucleus of the cell, has a negative regulatory role in
   NLRP3 under normal conditions and Zol stimulation (Fig. [137]8K-M).
   These findings suggest that LINC01013 regulates the nuclear
   translocation of ILF3 and influences the osteogenic differentiation of
   JBMMSCs, with NF90 playing a critical role in modulating NLRP3
   expression after Zol treatment.

   Taken together, the mechanism by which LINC01013 regulated the
   osteogenic differentiation of JBMMSCs was shown in Fig. [138]9. HOXC8
   negatively regulates LINC01013 to inhibit osteogenic differentiation of
   JBMMSCs under BP condition. LINC01013 binding to ILF3 affects the
   nuclear localization of ILF3 and regulates the NLRP3/Caspase-1 pathway
   to affect the function of JBMMSCs under BP stimulation (Fig. [139]9).

Fig. 9.

   [140]Fig. 9
   [141]Open in a new tab

   Mechanism diagram of LINC01013 reversing bisphosphonate-impaired
   osteogenic differentiation of JBMMSCs by regulating intracellular
   translocation of ILF3

Discussion

   In this study, we investigated the role and mechanism of
   HOXC8/LINC01013 during the osteogenic differentiation of JBMMSCs under
   Zol stimulation. We comprehensively elucidated the complex mechanisms
   underlying the mechanisms in which Zol inhibits osteogenic
   differentiation of JBMMSCs. We demonstrated that in response to Zol
   treatment, HOXC8 exerts an inhibitory role and LINC01013 exerts an
   enhanced role in osteogenic differentiation. Furthermore, we highlight
   that ILF3 plays a pivotal role in this differentiation process,
   revealing its potential as a therapeutic target.

   We demonstrated that HOXC8 negatively regulates the osteogenic
   differentiation of JBMMSCs. ALP activity represents the early stages of
   JBMMSCs osteogenic differentiation [[142]34]. HOXC8 knockdown increases
   ALP activity in JBMMSCs. BSP and OCN play critical roles in
   mineralization and are major non-collagenous proteins in the
   extracellular matrix of the bone [[143]35, [144]36]. Consistently,
   HOXC8 knockdown enhanced BSP and OCN expression in JBMMSCs. To confirm
   our observations, we analyzed calcium deposition, an indicator of the
   late stages of BMSCs osteogenic differentiation [[145]37], and found
   that HOXC8 is involved in the Zol-induced negative regulation of
   JBMMSCs osteogenic differentiation.

   Numerous lncRNAs are essential for the pathogenesis of bone diseases.
   We have previously shown that HOXC8 binds to the LINC01013 promoter to
   suppress LINC01013 expression [[146]14]. In contrast to HOXC8,
   LINC01013 expression was decreased in Zol-stimulated JBMMSCs. Both in
   vitro and in vivo results demonstrated that LINC01013 rescued the
   Zol-induced impaired osteogenic differentiation of JBMMSCs. These
   findings indicate that HOXC8/LINC01013 plays an essential role in
   Zol-induced osteogenic differentiation impairment in JBMMSCs.

   We confirmed that LINC01013 was located in the nucleus of JBMMSCs. The
   interaction between lncRNAs and proteins is an important mechanism
   through which lncRNAs exert their regulatory functions [[147]38]. We
   demonstrated an interaction between LINC01013 and ILF3. Previous
   studies have shown that the lncRNA PLXDC2-OT binds to the SIRT7/RBM6
   protein complex to impair its binding and deacetylation function in the
   OSX promoter, thereby enhancing the osteogenic differentiation of MSCs
   [[148]39]. It has also been shown that the lincRNA myocardial
   infarction-associated transcript (MIAT) is a regulator of the early
   stages of Th17 cell differentiation. Mechanistic studies have shown
   that signal transducer and activator of transcription 3 (STAT3) induces
   MIAT expression by binding to the MIAT promoter [[149]40]. Therefore,
   the RNA-protein complex is central to the regulation of stem cell
   differentiation.

   ILF3 shuttles between the nucleus and cytoplasm, and is composed of a
   bipartite nuclear localization domain, an RNA-binding domain with two
   dsRBMs, and an arginine- and glycine-rich C-terminal domain [[150]26].
   NF90/NF110 proteins are primarily involved in regulating gene
   expression by shuttling between the nucleus and cytoplasm [[151]41].
   Previous studies have shown that the regulatory roles of NF90 and NF110
   in modulating bone marrow maturation and IFN responses depend on their
   nuclear localization [[152]33]. Furthermore, ILF3 is involved in
   hyperlipidemia-induced atherosclerotic calcification. The molecular
   mechanism involves ILF3 promotion of calcification in human aortic
   vascular smooth muscle cells by acting on the promoter region that
   regulates bone morphogenetic protein 2 (BMP2) and signal transducer and
   activator of transcription 1 (STAT1), thereby regulating the
   transcription of BMP2 and STAT1 [[153]42]. Accordingly, the function of
   ILF3 is closely associated with its nuclear localization. Consequently,
   we investigated the effect of Zol on the intracellular localization of
   ILF3 in JBMMSCs. We found that LINC01013 regulates ILF3 translocation
   between the nucleus and cytoplasm under Zol stimulation. Previous
   studies have demonstrated that phosphorylation of ILF3 leads to its
   translocation from the nucleus to the cytoplasm [[154]43]. There are
   also studies confirming that lncRNA-BTX prevents ILF3’s migration to
   the cytoplasm from the nucleus by enhancing the interaction between
   ILF3 and ILF2 [[155]32]. However, the molecular mechanism by which
   LINC01013 regulates ILF3 localization requires further study. Moreover,
   our results showed that NF90 and NF110 osteogenic differentiation
   promotion of JBMMSCs depended on their nuclear localization. NF90
   interacts with the 3’-UTR region of various mRNAs such as poly
   (ADP-ribose) polymerase 1and HOXD8 to regulate their mRNA stability
   [[156]44, [157]45]. However, further investigation is required to
   ascertain whether NF90 and NF110 regulate the osteogenic
   differentiation of JBMMSCs by modulating the stability of
   osteogenic-related gene mRNAs. This enables the identification of
   potential target genes.

   GO enrichment analysis of RNA-seq revealed that GO functions were
   associated with inflammatory, immune, and innate immune responses.
   Inflammasomes are innate immune receptors that play a pivotal role in
   inflammation [[158]46]. The NLRP3 inflammasome is the most studied and
   comprises the protein NLRP3, apoptosis-associated speck-like protein,
   and pro-caspase-1 [[159]47]. The increased incidence of BRONJ in mice
   resulting from Zol treatment correlates with NLRP3 inflammasome
   activation [[160]48]. In addition, previous studies have shown that Zol
   induces apoptosis and reduces the osteogenic differentiation of murine
   osteoblast cell lines via the NLRP3/Caspase-1/gasdermin D pathway
   [[161]9]. In addition, NLRP3 influences bone healing after tooth
   extraction. It has been demonstrated that NLRP3 inflammasome activation
   inhibits the osteogenic differentiation of JBMMSCs [[162]49]. To
   investigate the relationship between NLRP3 inflammasome activation and
   BRONJ development, we measured the expression of NLRP3, Caspase-1 and
   IL-1β in JBMMSCs following Zol stimulation. Our results showed that Zol
   increased the expression of NLRP3, Caspase-1, and IL-1β in JBMMSCs.
   Subsequently, the expression of NLRP3 in JBMMSCs overexpressing
   LINC01013, NF90, or NF110 was examined. We found that LINC01013 and
   NF90 reduced NLRP3 expression in Zol-treated JBMMSCs.

Conclusion

   We elucidated the intricate molecular mechanisms underlying osteogenic
   differentiation during BP therapy, with particular focus on the
   involvement of HOXC8, LINC01013, and ILF3. Zol treatment increased
   HOXC8 expression, which suppressed LINC01013 levels in JBMMSCs and
   inhibited their osteogenic differentiation. The reduction in LINC01013
   promoted NF90 and NF110 translocation from the nucleus to the
   cytoplasm, inhibiting the osteogenic differentiation of JBMMSCs.
   Meanwhile, the reduction in LINC01013 and cytoplasmic translocation of
   NF90 led to the increased expression of inflammation-related genes,
   which may contribute to the dysfunction of osteogenic differentiation
   in JBMMSCs. Our results indicate that LINC01013-ILF3 plays a
   significant regulatory role in osteogenic differentiation and
   inflammatory responses in JBMMSCs and serve as potential targets for
   the treatment of BRONJ. Future research should focus on further
   elucidating the molecular mechanisms underlying this pathway and
   exploring therapeutic interventions for BRONJ treatment.

Electronic supplementary material

   Below is the link to the electronic supplementary material.
   [163]Supplementary Material 1^ (14.3KB, docx)
   [164]Supplementary Material 2^ (15.2KB, docx)
   [165]Supplementary Material 3^ (2.2MB, docx)
   [166]Supplementary Material 4^ (13.7MB, xlsx)

Acknowledgements