Introduction
Osteoporosis (OP), an insidious metabolic skeletal disorder prevalent among the elderly and postmenopausal women, is characterized by diminished bone mass and degradation of bone microstructure, resulting in skeletal fragility and heightened fracture risks [
1]. The pathogenesis of OP is intricate, and current understanding posits that OP primarily arises from an imbalance between bone formation and resorption, driven by dysfunctional osteoblast and osteoclast activities [
2]. As the global population ages, OP emerges as a formidable threat to human health, escalating medical expenses costs, and imposing substantial burdens on OP patients and their families [
3]. Compounding this issue is the absence of specific drugs in clinical practice for a definitive cure, with existing treatments only capable of delaying OP progression and mitigating fracture risks [
4]. Consequently, there is an urgent need to identify novel and effective therapeutic targets for OP.
Circular RNAs (circRNAs) have emerged as a class of endogenous non-coding RNAs characterized by a unique covalent closed loop without a 5' cap or 3' tail. They play a crucial role in modulating various biological activities and pathophysiological responses [
5]. Increasing studies have identified numerous deregulated circRNAs in bone metabolism, with circRNAs serving as potential diagnostic biomarkers for OP [
6]. Circ_HECW2, for instance, exhibits an aberrantly high expression pattern in osteoarthritis patients and lipopolysaccharide (LPS)-exposed chondrocytes, such that Circ_HECW2 overexpression exacerbates LPS-induced chondrocyte apoptosis [
7]. Additionally, Circ_HECW2 has been reported to be upregulated in the serum of coronary artery disease patients and oxidized low-density lipoprotein-induced human cardiac microvascular endothelial cells [
8]. However, the regulatory mechanism of Circ_HECW2 in OP remains elusive.
Accumulating evidence highlights the presence of circRNA-miRNA networks in the intricate regulation of various bone diseases, including OP [
9]. MicroRNAs (miRNAs), as non-coding small RNAs, intricately modulate gene expression post-transcriptionally, contributing significantly to the pathophysiology of diverse musculoskeletal conditions. For instance, specific miRNAs play pivotal roles in regulating cytokine expression, orchestrating the proliferation and differentiation of stromal cell lines involved in the composition of the extracellular matrix, thereby presenting therapeutic opportunities for managing tendon injuries [
10]. miRNAs exhibit distinct expression patterns in osteoarthritis compared to nonosteoarthritic cartilage, and the complex interactions between miRNAs and their target genes are crucial for maintaining homeostatic pathways in osteoarthritis [
11]. Notably, miRNAs are implicated in regulating the expression of OP-related genes and display abnormal expression profiles in OP patients [
12]. Mechanistically, circRNAs function as molecular sponges for miRNAs, disrupting the expression of target messenger RNAs (mRNAs), a phenomenon substantiated in various bone cellular activities [
13]. A deeper comprehension of the pivotal involvement of circRNA-miRNA-mRNA networks in OP may yield novel insights for exploring diagnostic biomarkers and therapeutic targets. Previous miRNome analysis has unveiled a reduction in miR-1224-5p expression in the serum samples of osteoporotic postmenopausal women [
14]. Identified as a crucial osteogenic regulator, miR-1224-5p overexpression notably facilitates fracture healing and ameliorates OP progression resulting from estrogen deficiency and aging [
15]. However, the regulatory network formed by Circ_HECW2 and miR-1224-5p in OP remains unexplored. This study endeavors to elucidate the expression pattern of Circ_HECW2 in OP and investigate its molecular mechanism in lipopolysaccharide (LPS)-induced osteoblast apoptosis, aiming to provide novel insights for the clinical management of OP.
Materials and methods
Ethics statement
This study received approval from the Ethics Committee of The First Hospital of Lanzhou University and adhered to the principles of the Declaration of Helsinki. Informed consent was obtained from all participants and their families.
Study subjects
A total of 65 OP patients (OP group) who were diagnosed and treated at our hospital from July 2021 to July 2022 were enrolled, comprising 26 males and 39 females. Simultaneously, 65 healthy volunteers with normal bone mass were recruited as the control group, including 21 males and 44 females. Inclusion criteria encompassed patients diagnosed with OP by dual-energy x-ray absorptiometry (DEXA; Hologic Inc., Waltham, MA, US) (T-score ≤ − 2.5), capable of self-care in daily life, and able to communicate clearly without cognitive impairment. Exclusion criteria comprised patients with secondary OP caused by tumors, infections, and prolonged use of glucocorticoids; individuals with endocrine system diseases, neuro-muscular system diseases, blood diseases, and severe organ diseases; those who had taken drugs affecting bone metabolism within 3 months before seeking medical attention; and individuals mental disorders. Demographic details including age, gender, smoking history, body mass index (BMI), etc., were recorded. Blood samples were collected at admission and processed in Becton Dickinson (BD) ethylenediaminetetraacetic acid tubes, centrifuged at 3000×g for 15 min, and the serum was stored at − 80 °C for testing. Laboratory indicators such as white blood cells, parathyroid hormone (PTH), inorganic phosphorus, tP1NP, β-CTx, Calcium, 25(OH) D, Lymphocyte count, and Monocyte count were tested in strict accordance with operational procedures and kit instructions.
Bone mineral density (BMD) evaluation
Patients, after discarding metal items, assumed a supine position at the center of the DEXA machine. Special pads were used to lift the lower limbs, aligning the hip and knee joints at a 90° angle. The initial laser point was positioned at the fifth lumbar spine. BMD values of the lumbar spine (L1, L2, L3, L4, and the sum of L1–L4), along with the T value by the system under big data, were collected. BMD values were measured meticulously by the same group of imaging physicians following technical specifications.
Cell culture and treatment
Human osteoblasts HFOB1.19 (ATCC, Manassas, Virginia, USA; cells were identified through short tandem repeats, and Mycoplasma test result was negative) were cultured in medium containing 10% fetal bovine serum and 1% penicillin–streptomycin. The medium was refreshed every three days. Upon reaching 80% confluence, HFOB1.19 cells were subjected to co-culture with 10 µg/mL lipopolysaccharide (LPS; Solarbio, Beijing, China) added for 24 h. Cells without LPS treatment served as controls. Cell transfection was conducted before LPS addition. Briefly, HFOB1.19 cells in the logarithmic phase were transfected with si-Circ_HECW2 (si-NC) and anti-miR-1224-5p (anti-NC) (GenePharma Co., Ltd, Shanghai, China) according to the Lipofectamine 2000 reagent instructions (Thermo Fisher Scientific Inc., Waltham, MA, USA).
Reverse transcription quantitative polymerase chain reaction (RT-qPCR)
Total RNA extracted from serum and osteoblasts using the TRIzol method (Beyotime, Shanghai, China) underwent reverse-transcribed into cDNA with the PrimeScript TM RT Reagent Kit (Takara, Tokyo, Japan). RT-qPCR was performed following the SYBR® Premix EX Taq TM II Kit instructions (Takara) on an ABI7500 PCR instrument (Invitrogen; Thermo Fisher Scientific), with cDNA as the template. The 2
−ΔΔCT method [
16] calculated the relative expression of genes, using U6 as the internal reference of pre-miR-1224 and miR-1224-5p [
17], and GAPDH as the internal reference of CircRNA and mRNA [
18]. Primer sequences are provided in Table
1.
Circ_HECW2 | CCCACCACTTTGAACGCTAC | GGCTGTCAATGCGTGCCT |
HECW2 | TACCACGGCATTAGTGGAGC | ATCCCTTTCTTTAGCCCAACTG |
pre-miR-1224 | TCGGCAGGGTGAGGACTCGGG | CAGCCACAAAAGAGCACAAT |
miR-1224-5p | GCCGAGGTGAGGACTCGGGAGG | CTCAACTGGTGTCGTGGA |
PDK2 | CCCCGTCCCCGTTGTC | TCGCAGGCATTGCTGGAT |
U6 | CTCGCTTCGGCAGCACA | AACGCTTCACGAATTTGCGT |
GAPDH | CGCTCTCTGCTCCTCCTGTTC | ATCCGTTGACTCCGACCTTCAC |
Terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL)
Following transfection and LPS treatment, HFOB1.19 cells were prepared into smears, immersed in 40 g/L paraformaldehyde for 20 min, and then infiltrated with pre-cooled 1% Triton X-100. After sealing for 10 min, the cells were treated with the reaction solution per the TUNEL Cell Apoptosis Test Kit instructions (Beyotime), and finally added with the 4′,6-diamidino-2-phenylindole solution. TUNEL-positive cells were observed under a BX53 fluorescence microscope. The apoptosis index (number of apoptotic cells/number of cells in the field) was calculated based on counting TUNEL-positive cells that were counted in six randomly selected fields of view.
Western blot
Cell protein was extracted using radio-immunoprecipitation assay lysis buffer (Beyotime), and the protein determined was detected using a bicinchoninic acid kit (Beyotime). Equivalent amounts of protein were separated by 12% SDS-PAGE (Beyotime) and transferred onto polyvinylidene fluoride membranes by electrophoresis. After blocking, the membranes were incubated with primary antibodies cleaved-caspase-3 (ab2302, 1:50, Abcam, Cambridge, MA, USA) and GAPDH (ab128915, 1:10,000, Abcam). Following rinsing, the membranes were further incubated with the corresponding secondary antibody (ab205718, 1:2000, Abcam) in the dark for 2 h. After another round of rinsing, the enhanced chemiluminescence assay kit was used for imaging, and Bandscan 5.0 facilitated the quantitative detection of protein bands.
Ribonuclease R (RNase R) resistance assay
Total RNA was extracted from HFOB1.19 cells. Subsequently, 1.5 μg of RNA was digested with or without RNase R (Thermo Fisher Scientific) for 10 min. After reverse transcription, RT-qPCR was performed to detect the mRNA levels of Circ_HECW2 and linear HECW2.
Actinomycin D treatment
HFOB1.19 cells underwent treatment with 1 μg/mL actinomycin D for different time points (0, 8, 16, and 24 h). Total RNA was then extracted for RT-qPCR to assess the mRNA levels of Circ_HECW2 and linear HECW2.
Subcellular fractionation
HFOB1.19 cells were subjected to an ice bath in cytoplasmic lysate for 10 min and then centrifuged to separate the cytoplasmic supernatant from the cell particles containing the nucleus. RNA was subsequently obtained from cytoplasmic and nuclear extracts using the PARIS reagent kit (Invitrogen). RT-qPCR determined the levels of Circ_HECW2, U6 (nuclear control), and GAPDH (cytoplasmic control).
Dual-luciferase reporter gene assay
The binding sites between Circ_HECW2 and pre-miR-1224 and between miR-1224-5p and PDK2 were predicted through the IntaRNA and TargetScan7.2 databases, respectively. Synthesized Circ_HECW2-WT, containing the pre-miR-1224 binding site, and Circ_HECW2-MUT, mutated according to the binding site, as well as the PDK2 3'UTR fragment containing the miR-1224-5p binding site (PDK2-WT) and PDK2-MUT, were individually constructed into the dual-luciferase report vectors (LMAI, shanghai, China). These plasmids were co-transfected with mimic-NC, pre-miR-1224-mimic, and miR-1224-5p-mimic into HFOB1.19 cells using Lipofectamine 2000. After 48 h, the cells were collected and analyzed on the dual-luciferase reporter detection system (Promega, Madison, WI, USA).
RNA pull-down
Biotinylated miR-NC, pre-miR-1224-WT, and pre-miR-1224-MUT (Invitrogen) were transfected into HFOB1.19 cells using Lipofectamine 2000. Following ultrasound treatment, cell lysates were collected and incubated with Dynabeads streptomyces avidin magnetic beads M-280 (Thermo Fisher Scientific). The bound RNA was purified using TRIzol reagent, and the enrichment of Circ_HECW2 was measured using RT-qPCR.
Statistical analysis
Data analysis and graphical representation were performed using the SPSS 22.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism 8.0 (GraphPad Software Inc., San Diego, CA, USA). Counting data were expressed as percentages, and comparisons between two groups were conducted using the χ2 test. Measurement data were presented as mean ± standard deviation. The t test was employed for comparisons between two groups. One-way or two-way analysis of variance (ANOVA) was used for comparisons among multiple groups, followed by Tukey’s multiple comparison test. The receiver operating characteristic curve (ROC) was utilized to analyze the diagnostic efficiency of Circ_HECW2 in OP. Pearson’s correlation coefficient was applied to analyze the correlation of serum Circ_HECW2 and serum pre-miR-1224 with mature miR-1224-5p, as well as the correlation of serum miR-1224-5p with PDK2 mRNA. A significance level of P < 0.05 was considered statistically significant.
Discussion
In the ever-expanding landscape of high-throughput RNA sequencing studies, the aberrant expression of circRNAs in OP has come to the forefront [
6]. In a pioneering contribution, our study underscores the heightened expression of Circ_HECW2 in OP patients, establishing its diagnostic significance for the condition. Specifically, Circ_HECW2’s involvement in accelerating LPS-induced osteoblast apoptosis is delineated. This acceleration is attributed to Circ_HECW2's competitive binding to pre-miR-1224, thereby diminishing the formation of mature miR-1224-5p and orchestrating the upregulation of PDK2 expression.
The intricate interplay between osteoclast-mediated bone resorption and osteoblast-mediated bone formation is a hallmark of OP pathogenesis, disrupting the delicate balance of bone remodeling and precipitating the onset of OP [
24]. Programmed cell death mechanisms, such as apoptosis, ferroptosis, and pyroptosis, play a decisive role in bone cell fate and exert a profound impact on bone metabolism [
25]. Within this context, osteoblast apoptosis emerges as a pivotal event in the initiation and progression of OP, positioning the inhibition of osteoblast apoptosis as a promising avenue for OP treatment [
2]. In our exploration of Circ_HECW2's regulatory mechanism in OP, we opted for LPS induction to simulate osteoblast apoptosis. Subsequent to LPS treatment, both the apoptosis index and cleaved-caspase-3 protein witnessed a significant surge. Notably, there was a conspicuous rise in Circ_HECW2 expression in HFOB1.19 cells following LPS exposure. This aligns with previous findings where Circ_HECW2 exhibited significant upregulation in osteoarthritis patients and LPS-exposed chondrocytes, and its overexpression exacerbated LPS-triggered chondrocyte apoptosis [
7]. Moreover, Circ_HECW2 silencing has been demonstrated to impede LPS-triggered endothelial–mesenchymal transition and cell apoptosis in human brain microvascular endothelial cells [
26]. siRNAs are useful to identify molecular targets and evaluate the efficacy of specific drugs and can themselves be used for therapeutic purposes in tendon healing, rheumatoid arthritis, and OP [
27‐
29]. The transfection of si-Circ_HECW2 in our study corroborated these trends, significantly reducing the apoptosis index and cleaved-caspase-3 protein in HFOB1.19 cells. This lends credence to the notion that Circ_HECW2 downregulation exerts a suppressive effect on osteoblast apoptosis in OP.
Subsequently, our focus shifted to unraveling the intricate molecular mechanism through which Circ_HECW2 regulates osteoblast apoptosis. CircRNAs exert a profound influence on diverse physiological and pathological processes by modulating gene transcription and translation, mediating alternative splicing, or acting as miRNA sponges in specific tissues [
9]. For instance, in a prior study, Circ_HECW2 was identified as a miR-30e-5p sponge, enhancing endothelial–mesenchymal transition triggered by LPS [
26]. Our investigation confirmed that Circ_HECW2 is predominantly expressed in the cytoplasm, suggesting its potential role as a mediator of osteoblast apoptosis by functioning as a miRNA sponge. Dysregulation of miRNA-mediated mechanisms stands out as a prominent pathological factor in OP [
30]. Upon literature review, the significant role of miR-1224-5p in osteoclast responses and proliferation drew our attention [
22]. Exosomal miR-1224 expedites osteoclast differentiation in patients with brain trauma via the Hippo signaling pathway [
31]. Canonical miRNA biogenesis undergoes a two-step process: from primary transcripts (pri-miRNAs) to pre-miRNAs and from pre-miRNAs to mature miRNAs, mediated by DROSHA and DICER, respectively [
32]. IntaRNA website prediction unveiled a binding site between Circ_HECW2 and pre-miR-1224, a relationship validated through dual-luciferase reporter gene assay and RNA pull-down assays. Pre-miR-1224 exhibited heightened expression, while miR-1224-5p was poorly expressed in OP patients and LPS-treated HFOB1.19 cells. Downregulating Circ_HECW2 reduced the expression of pre-miR-1224 but elevated the expression of miR-1224-5p. We postulate that Circ_HECW2 might diminish the formation of mature miR-1224-5p by competitively-binding to pre-miR-1224. Notably, miRNA-1224-5p knockdown impedes osteoblast differentiation, suppresses bone regeneration, and accelerates OP progression in mouse models [
15]. Consistent with these findings, we observed that miR-1224-5p inhibition reversed the inhibitory effect of HECW2 downregulation on osteoblast apoptosis, signifying that Circ_HECW2 promotes osteoblast apoptosis by reducing the formation of mature miR-1224-5p.
Finally, leveraging TargetScan7.2 and miRDB databases, we conducted a screening to identify the downstream genes of miR-1224-5p. Pyruvate dehydrogenase kinases (PDKs), a group of gatekeeper enzymes, play a pivotal role in metabolic regulation by modulating the activity of pyruvate dehydrogenase complex [
33]. Among the four PDK isoenzymes discovered in mammalian tissues, PDK2 inhibition has been shown to prevent ovariectomy-induced bone loss during osteoclastogenesis in mice by suppressing aberrant osteoclast activation [
23]. Our results confirmed the binding relationship between miR-1224-5p and PDK2. Moreover, PDK2 mRNA expression was significantly elevated in the serum of OP patients, and downregulation of Circ_HECW2 reduced PDK2 mRNA expression in LPS-treated HFOB1.19 cells. These findings suggest that Circ_HECW2 might upregulate PDK2 expression by repressing the formation of mature miR-1224-5p, thereby fostering osteoblast apoptosis.
Conclusion
In summary, our findings underscore the diagnostic significance the heightened expression of Circ_HECW2 in osteoporosis (OP), shedding light on its role in fostering osteoclast apoptosis through the intricate miR-1224-5p/PDK2 axis. Despite the valuable insights gained, this study encounters certain limitations. The exploration of Circ_HECW2’s regulatory impact on OP remains confined to the cellular level, warranting future investigations involving animal models to validate its mechanisms. The specificity of Circ_HECW2 in regulating additional targets within the context of OP remains unresolved. Furthermore, the downstream target gene of miR-1224-5p, PDK2, merits further exploration, extending our comprehension of the intricate molecular network involved in OP pathogenesis. Future endeavors will encompass the verification of Circ_HECW2's regulatory mechanism in vivo, coupled with a more in-depth analysis of PDK2, to advance our understanding of the comprehensive regulatory landscape in osteoporosis.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.