Steroid-induced osteonecrosis of the femoral head is a disabling disease, particularly after the 2003 SARS(severe acute respiratory syndrome) outbreak in China, but the mechanism of whom is not completely understood. Approximately 20,000–30,000 people are diagnosed with osteonecrosis in the United States each year; the morbidity rate of non-traumatic femoral head osteonecrosis in Japan was 1.91/100,000 per year. The natural course of femoral head osteonecrosis is necrosis, followed by collapse and osteoarthritis. Once collapse occurs, osteoarthritis and joint replacement is inevitable [4, 5]. Therefore, collapse must be treated surgically for femoral head osteonecrosis, and preventing collapse has been the focus of non-joint replacement therapy [6–14]. Current studies of collapse are typically performed from the biological, biomechanical, or mechanobiological perspective [15–19]; however, the mechanism of collapse is unclear, leading to insufficient effective preventive measures and unsatisfactory efficacy of hip-preserving surgery. As the number of studies of miRNA in orthopedics has increased, specific miRNAs involved in femoral head osteonecrosis have been identified [20–24]. However, the developmental course of femoral head osteonecrosis is a dynamic process. Whether miRNA is differentially expressed in different states of osteonecrotic femoral head (non-collapse vs. collapse) remains unclear. Studies of differential miRNA expression may increase the understanding of the biological mechanisms underlying collapse. This study aimed to identify miRNAs differentially expressed in collapse and non-collapse areas of the osteonecrotic femoral head and validate these miRNAs using reverse transcription-polymerase chain reaction (RT-PCR) assays, and eventually predict the target genes and pathways of these miRNAs.
Quality control and results of differentially expressed miRNAs
The 9 selected samples passed the quality control test. A total of 2085 differentially expressed miRNAs were detected in the 9 samples. Among the detected miRNAs, 433 miRNAs showed more than 1.5-fold differential expression in the T1 group compared to the W1 group (247 miRNAs were up-regulated and 186 miRNA were down-regulated); 344 miRNAs were differentially expressed by more than two-fold in the T2 group compared to the W2 group (225 miRNAs were up-regulated and 119 miRNAs were down-regulated); 107 miRNAs showed differential expression by more than two-fold in the T3 group compared to the W3 group (51 miRNAs were up-regulated and 56 miRNAs were down-regulated) (Figure 1). Combining the data from all three patients, 10 miRNA were differentially expressed by more than two-fold in the collapse area (T1+T2+T3) compared to the non-collapse area (W1+W2+W3) (8 miRNAs were up-regulated: hsa-miR-4472, hsa-miR-4306. hsa-miR-4747-5p, hsa-miR-4441, hsa-miR-4709-3p, ebv-miR-BHRF1-2-3p, hsa-miR-585-3p, and hsa-miR-5572, while 2 miRNAs were down-regulated: hsa-miR-195-5p and hsa-miR-645). Compared to the normal region, has-miR-195-5p showed the most significant down-regulation (Tables 1, 2, 3).
RT-PCR results of hsa-miR-195-5p
The following U6 primers were used for RT-PCR:F: 5′GCTTCGGCAGCACATAT ACTAAAAT3′, R: 5′CGCTTCACGAATTTGCGTGT CAT3′. The primer sequences of hsa-miR-195-5p were: GSP: 5′GGGGTAGCAGCACAGAAAT3′, R: 5′CAGTGCGTGTCGTGGAGT3′(GSP is a specific primer for miRNA, while R is a primer that matches the RT primer). The RT-PCR expression level of hsa-miR-195-5p was significantly lower in the collapse area (T1 T2 T3)than in the non-collapse area (W1 W2 W3) and normal area (Z1 Z2 Z3). (Table 4) Comparison of the results from RT-PCR and the microarray data also indicated that hsa-miR-195-5p expression was downregulated. Amplification and melting curves were obtained (Figure 2).
Target genes, network, GO analysis and KEGG pathway analysis of hsa-miR-195-5p
157 genes were perdicted as the target gene of hsa-miR-195-5p (Figures 3, 4). GO analysis and KEGG Pathway analysis were showed that pathway in cancer is significant (Figures 5, 6).
Previous studies have suggested that collapse occurs because of an imbalance between bone resorption and bone formation after femoral head osteonecrosis, resulting in femoral heads with reduced structural strength that are unable to bear external forces. The studies have also suggested that mechanical support and adjustment of bone cell metabolism can prevent collapse. To comprehensively understand the biological behavior of osteocytes, an increasing number of studies have focused on the differential expression of miRNAs associated with osteonecrosis of the femoral head [23–24, 26–28]. We evaluated whether the decrease in structural strength was associated with the differential expression of relevant miRNAs, which may lead to faster apoptosis and an increased imbalance with bone metabolism. We found that in the osteonecrotic region, hsa-miR-195-5p was down-regulated in the collapse area compared to in the non-collapsed area. This is the first study examining the relationship between miRNA and femoral head collapse.
The anatomical structure of osteonecrotic femoral head has four layers: cartilage, osteonecrotic area, sclerotic area, and normal area. Recent studies have proposed that the decrease in structural strength is correlated with sclerotic area and osteonecrotic edge, enabling collapse prediction [16, 19]. During the pathological process, collapse of the osteonecrotic femoral head occurs because of continuous spreading of trabecular fracture, which is not a sudden process, and thus the osteonecrotic area includes the collapse area and non-collapse area. Femoral head collapse originates from subchondral fracture. Therefore, in this study, subchondral fracture in the osteonecrotic femoral head was defined as the collapse area, while the area in the osteonecrotic femoral head without subchondral fracture was defined as the non-collapse area. Non-collapse and collapse are different pathological stages of osteonecrotic femoral head. We not only developed a method for predicting the risk of collapse, but also provide a new insight into preventing collapse using gene editing techniques.
It has been reported that miR-195-5p is a tumor-suppressor gene that mainly suppresses cell proliferation and invasion and is down regulated in cancer patients [29–34]. This gene inhibits hepatocellular carcinoma by regulating PHF19, regulates rectal cancer through CDK8, regulates osteosarcoma by inhibiting NKD1, regulates prostate cancer through Fra-1 or RPS6KB1. Based on these results, we hypothesized that the mechanism of collapse involves down-regulation of miR-195-5p, which disrupts the proliferation of normal osteoblasts and accelerates cell apoptosis. This leads to a further decrease in structural strength within the femoral head and failure to support external forces, eventually resulting in collapse. Future studies should examine methods of increasing the expression of miR-195-5p to treat cancer and the collapse of femoral head osteonecrosis.
The limitations of this study were as follows: firstly, small sample size, but with reproducibility; secondly, target genes and pathways were only predicted and require further validation; In addition, whether collapse leads to miRNA changes or miRNA changes lead to collapse remains unknown. These factors exhibit a mutually causal relationship. In future studies, we will evaluate the genetic pathways of the collapse-associated differential miRNAs and gene editing technology to prevent and treat collapse.
Three patients were randomly selected from the hospital system in our department. Inclusion criteria were as follows: 1. diagnosed with steroid-induced femoral head osteonecrosis (a clear history of hormone use); 2. diagnosed with ACRO III/IV stage based on X-ray by 3 doctors; and 3. the surgeon and patient decided to perform the total hip joint replacement surgery. Exclusion criteria were as follows: 1. The cause of femoral head osteonecrosis was not clear or involved multiple factors; 2. Severe liver and kidney failure in which the patient could not tolerate surgery; And 3. Surgical strategy in which patient consent was not obtained. Informed consent was obtained from all patients and the ethical review committee of our hospital. According to the inclusion and exclusion criteria, three patients were selected. There were 2 males, one of whom at the age of 30 years used hormones to treat thrombocytopenic purpura, while the other, at the age of 46 years, used hormones to treat skin disease. At the age of 20 years, the female patient used hormones to treat systemic lupus erythematosus.
After surgery, femoral heads were removed and cleaned with saline. Three bone tissues (normal area, collapse area, and non-collapse area) were taken from femoral heads (Figure 7). Two doctors identified these tissues, and a total of 9 samples were obtained (T1 T2 T3, W1 W2 W3, and Z1 Z2 Z3). Bone tissues were immediately stored in liquid nitrogen and ground into powder (≥ 100 mg) in liquid nitrogen. The powder was transferred to an microcentrifuge tube, mixed with 1 mL Trizol, and stored at −70°C until use.
Total RNA extraction and quality control
Total RNA was extracted from tissue cells using Trizol Reagent (Invitrogen life technologies) according to the manufacturer's instructions; total RNA was extracted from whole blood using TRI reagent BD (MRCgene, TB-126). Total RNA was extracted from plasma and serum exosomes using TRIzol LS. RNAsey Mini Kit (QIAGEN) was used for RNA purification; NanoDrop ND-1000 was employed for measurement of RNA concentration after purification. RNA integrity was detected by electrophoresis.
RNA labeling and hybridization using the exiqon method [23R23bibr]
After extraction, RNA samples underwent quality control and miRNA was labeled using miRCURYTM Array Power Labeling kit (Cat #208032-A, Exiqon). The detailed steps are described as follows:One microgram RNA sample was added to 2 μL water, and then 1 μL CIP buffer and CIP enzyme (Exiqon) were added to the RNA sample solution, and the mixture solution was incubated at 37°C for 30 min. The reaction in the sample solution was terminated at 95°C for 5 min, and then 3 μL labeling buffer, 1.5μL fluorescent label (Hy3™), 2.0 μL DMSO, and 2.0 μL labeling enzyme were added and the solution was incubated at 16°C for 1 h. The reaction in the samples was terminated at 65°C for 15 min.
After labeling, the sample was hybridized with miRCURYTM LNA Array (v.18.0) (Exiqon), and the subsequent procedures were conducted according to Exiqon's experimental methods. The hybridization system used was the Nimblegen system (Nimblegen Systems Inc., Addison, WI, USA). The microarray chip was rinsed with Wash buffer kit (Exiqon) after hybridization.
The microarray chip was scanned by the axon GenePix 4000B Microarray Scanner [23R23bibr]
GenePix Pro V6.0 was utilized to read the microarray scan images and the probe signals were selected. The median was taken and combined for the same probes. Probes >= 30.0 for the samples were kept and the median was standardized in all microarray chips. (Figure 8) miRNAs showing differential expressions of more than 1.5-fold were selected.
Verification of significantly differentially expressed miR-195-5p by RT-PCR
cDNA synthesis of RNA samples. 20 μL of RT reaction mixture was prepared, then all cDNA samples were separately set up in the RT-PCR System (Applied Biosystems); The solution was mixed up to 8 μL and centrifuged at 5000 rpm for a short period.
The mixture was added to each well of a 384-well PCR plate, and then 2 μL corresponding cDNA was added to each well. The plate was sealed with sealing film and centrifuged. The 384-well PCR plate was placed in the RT-PCR amplification instrument to carry out the PCR reaction. U6 (internal reference) and all parameters were conducted.
The RT-PCR reaction was performed for the target miRNA in all samples and the internal reference gene (U6) separately. In Table 1, data analysis was performed using 2−ΔΔCT method. SPSS 20.0 (SPSS Inc., Chicago, Ill., USA) was employed for statistical analysis of the data.
Predication of target gene of miR-195-5p
TargetScan, mirBase and miRanda bioinformatics software were used to predict target genes. Gene Ontology(GO, http://www.geneontology.org/) analysis and Kyoto Encyclopedia of Genes and Genomes(KEGG, www.genome.jp/kegg/) pathway analysis were done respectively. The GO project covers three domains: Biological Process, Cellular Component and Molecular Function. Fisher's exact test is used to find if there is more overlap between the differential expression (DE) list and the GO annotation list than would be expected by chance. The p-value denotes the significance of GO terms enrichment in the DE genes. The lower the p-value, the more significant the GO Term (p-value <= 0.05 ). Pathway analysis is a functional analysis mapping genes to KEGG pathways. The p-value denotes the significance of the Pathway correlated to the conditions. Lower the p-value, more significant is the Pathway(p-value cut-off is 0.05).