CYP27A1 deficiency promoted osteoclast differentiation

Construction of CYP27A1 knockout mice

CYP27A1 knockout (KO) mice were generated using the CRISPR/Cas9 system. Firstly, one sgRNA targeting the near sequence of the inserted site was constructed and transcribed in vitro. The donor vector with the inserted fragment was designed and constructed in vitro. Subsequently, Cas9 mRNA, sgRNA and donor were co-injected into zygotes, followed by transplantation into the oviduct of pseudopregnant ICR female mice at 0.5 dpc. After 19–21 days of transplantation, F0 mice were birthed, all the offspring of ICR females (F0 mice) were identified by PCR and sequencing of tail DNA. Finally, positive F0 mice were crossed with wild type mice to construct heterozygous mice. Homozygous mice were generated by mating heterozygous mice. Cas9 endonucleases in introns 1–2 and 2–3 of the CYP27A1 gene were directly cleaved by sgRNAs, and double-strand break (DSBS) were generated. Such breaks were repaired by non-homologous end joining (NHEJ), and resulted in the disruption of CYP27A1. The sequences was used as follows: gRNA1, CAGTAGTCTCCCTTGTTGTA; gRNA2, GTCTCTGTCTACAGAACCGC; gRNA3, TACCAACACGTTGTGTCATT; gRNA4, AGAAGAGGCAGTGGCTGTTC.

Animal experiments

Mice were obtained from the Nanjing Institute of Biomedicine, Nanjing University, and were maintained under the specific pathogen-free conditions in the animal facility. The water and food were supplemented every 3 days, and bedding was changed every 6 days. The animals were euthanized in the CO₂ incubator. Eight-week-old male mice were used for all experiments. All experiments were approved by the Laboratory Animal Ethics Committee of Shandong Provincial Hospital (NSFC: No. 2022-050).

Osteoclast differentiation

Eight-week-old wild type (WT) and homozygous C57/BL6 mice were used for the isolation of osteoclastic precursors. After euthanasia, the femur and tibia of mice were obtained and the bone marrow was rinsed with the α-MEM culture medium using a syringe. After filtration, bone marrow-derived monocytes (BMMs) were obtained by centrifugation. BMMs were cultured in the α-MEM complete culture medium in a 5% CO₂ incubator at 37 °C and 100% humidity for 24 h. After lysing erythrocytes, BMMs were cultured in the α-MEM complete culture medium containing 30 ng/ml M-CSF for 48 h. Subsequently, the medium was replaced with the complete medium containing 30 ng/ml M-CSF and 50 ng/ml RANKL every 2 days. Osteoclast formation was observed after 5 days of culture. Cells containing three or more nuclei stained by TRAP were considered as multinucleated osteoclasts.

TRAP staining

BMMs were cultured in 96-well plates, and were differentiated into osteoclasts, followed by immobilization for 20 min using 4% paraformaldehyde. The cells were subsequently incubated with TRAP solution at 37 °C in the dark for 45 min, and were stained with hematoxylin solution for 5 min. The staining cells were captured using an olympus microscope (IX53).

Construction and sequencing process of the mRNA gene library

Osteoclasts were lysed in the trizol for RNA extraction. The integrity and quantification of RNA were determined using the agarose gel electrophoresis and Nanodrop 2000 (Thermo Fisher, Waltham, MA, USA), respectively. Samples were divided into the control and KO groups with three repetitions. The library was obtained through a series of operational modifications.

The original image data obtained from sequencing were converted into sequence data by base calling, which were called raw data or raw reads. In order to ensure data quality, the original data should be filtered before information analysis to reduce the interference brought by invalid data. First, the raw reads were quality controlled using FASTQ (Chen et al., 2018) to filter low-quality data and obtain clean reads. StringTie was used to assemble the reads of RNAseq into transcripts (Pertea et al., 2015). From the alignment results of HISAT2, we reconstructed the transcripts using StringTie and calculated the expression levels of all genes in each sample using RSEM (Li & Dewey, 2011). The expression information was analyzed using DESeq2 software (Love, Huber & Anders, 2014). Based on the results of the differential analysis, significant differential genes were screened (FDR < 0.05 and |log2FC | > 1).

Enrichment analysis of differentially expressed genes (DEGs)

Differential genes were mapped to each term of the GO database (http://www.geneontology.org/), and the number of differential genes for each term was calculated to obtain the number of differential genes in the list of differential genes with a certain GO function (Yu et al., 2012). A hypergeometric test was then applied to find out the GO entries that were significantly enriched in the differential genes as compared to the background. The KEGG is the main public database about the pathways. Pathway significance enrichment analysis was performed in KEGG pathways units using a hypergeometric test to identify pathway significantly enriched in differential genes compared to the entire background (Ogata et al., 1999). The most important biochemical metabolic pathways and signal transduction pathways were identified by pathway significance enrichment.

Analysis of the protein interaction networks

The STRING protein interaction database (http://string-db.org) was mainly used to analyze the interaction network among differential genes (Szklarczyk et al., 2015). For the species included in the database, the differential gene set was extracted from the database to build the interaction network map using cytoscape (Shannon et al., 2003). For the species not included in the database, the sequences in the target gene set were first blastx aligned to the protein sequence of the reference species included in the string database, and the protein interaction relationship of the reference species on the alignment was used to build the interaction network.

Determination of the key genes

GeneMANIA (http://www.genemania.org) is an online analysis tool, and can search many large, publicly available biological datasets to find related genes, which include protein-protein, protein-DNA and genetic interactions, pathways, reactions, gene and protein expression data, protein domains and phenotypic screening profiles (Franz et al., 2018). Data is regularly updated. Using the gene tool, the interaction of the targeted genes with other genes was found.

Quantitative real-time PCR

Total RNA was extracted from cells in the CYP27A1 WT and KO groups (n = 3 per group) using Trizol reagent. The concentration and purity of total RNA were detected using a Nanodrop 2000 spectrophotometer (Thermo Fisher, Waltham, MA, USA). Reverse transcription was preformed to synthesize cDNA using Hiscript®iii Rt Supermix for qRT-PCR (RC323-01; Nanjing Vazyme Biotech Co, Nanjing, China). The cDNA was used as the template to amplify targeted sequences with specific primers using the Chamq Universal SYBR qRT-PCR Master Mix (Q711-02; Nanjing Vazyme Biotech Co, Nanjing, China). The results were analyzed by the 2^−ΔΔCT method (Livak & Schmittgen, 2001). The sequences of the eight primers were shown in Table 1.

Western blot analysis

The total protein was extracted using RIPA lysis buffer. The protein concentration was measured using BCA method. Equal amount of protein was subjected to 10% SDS-PAGE electrophoresis, and was then transferred to PVDF membrane, followed by blocking with 5% nonfat milk for 2 h at room temperature. After rinsed in TBST buffer, the protein was labeled by the corresponding primary antibody at 4 °C overnight and HRP-coupled secondary antibody at room temperature for 1 h. The protein bands were visualized using the enhanced chemiluminescence method. The antibodies used were as follows: anti-MMP9 (1:1,000, 502095; Zen-Bio Inc, Durham, NC, USA); anti-NFATc1 (1:1,000, 251865; Zen-Bio Inc, Durham, NC, USA); anti-c-Fos (1:1,000, A0236; ABclonal Technology, Woburn, MA, USA); anti-TRAP (1:1,000, 382344; Zen-Bio Inc, Durham, NC, USA); anti-CTSK (1:1,000, A1782; ABclonal Technology, Woburn, MA, USA); anti-PPAR-γ (1:1,000, 340844; Zen-Bio Inc, Durham, NC, USA); anti-AKT (1:1,000, R23412; Zen-Bio Inc, Durham, NC, USA); anti-IL7 (1:1,000, A1650; ABclonal Technology, Woburn, MA, USA); anti-COL1A2 (1:1,000, 380760; Zen-Bio Inc, Durham, NC, USA); anti-BGN (1:1,000, 16409-1-AP, ptgcn); anti-SPARC (1:1,000, 121259; Zen-Bio Inc, Durham, NC, USA); anti-CYP27A1 (1:1,000, A1982; ABclonal Technology, Woburn, MA, USA); anti-ELANE (1:1,000, A13015; ABclonal Technology, Woburn, MA, USA); anti-S100A9 (1:1,000, R25648; Zen-Bio Inc, Durham, NC, USA); anti-LY6C2 (1:1,000, ET1702-78; HUABIO, New Boston, MI, USA); anti-GAPDH (1:1,000, 10494-1-AP, ptgcn); anti-actin (1:1,000, 81115-1-RR, ptgcn).

Statistical analysis

The statistical analysis was performed using GraphPad Prism software (version 8.3.0, San Diego, CA). Data are presented as the mean ± standard deviation (SD). The difference between two groups was analyzed using Student’s t-test. p < 0.05 was considered statistically significant.

CYP27A1 KO promoted osteoclast differentiation

Since our previous findings have confirmed that 27HC-stimulated conditional medium of lung adenocarcinoma cells promotes osteoclast differentiation, and CYP27A1 is a synthetase of 27HC, we speculated that CYP27A1 KO might affect osteoclast differentiation. However, contrary to expectations, CYP27A1 KO also significantly promoted osteoclast formation (Fig. 1A). The in vivo imaging of bones from CYP27A1 WT and KO mice was performed using micro-CT. The results showed that CYP27A1 KO led to bone loss (Fig. 1B). Further investigation demonstrated that CYP27A1 KO promoted the expression of osteoclast-related genes, including MMP9, NFATc1, c-Fos, TRAP and CTSK (Figs. 1C and 1D).

Identification of DEGs between CYP27A1 WT and KO osteoclasts

To explore the molecular mechanism of CYP27A1 KO in osteoclast differentiation, RNA sequencing was performed to identify the DEGs between the CYP27A1 WT and KO groups. The DEGs between the two cohorts were shown in the form of volcano plot (Fig. 2A). Figure 2B illustrated eight top DEGs with four up-regulated and four down-regulated genes, including ELANE, LY6C2, S100A9, GM20708, BGN, SPARC, and COL1A2. The expression patterns of eight DEGs were hierarchically clustered, which was presented in the heatmap (Fig. 2C). These genes with similar expression patterns might have common functions or participate in common metabolic pathways and signaling pathways. The expression of hub genes was verified by qRT–PCR and Western blot. As shown in Fig. 2, all eight hub genes exhibited a similar tendency among the RNA-sequencing analysis, qRT–PCR and Western blot analysis. The mRNA and protein expression of COL1A2, BGN, SPARC and GM20708 was up-regulated in the osteoclasts from CYP27A1 KO group (Figs. 2D–2F), while the mRNA and protein levels of CYP27A1, ELANE, S100A9 and LY6C2 were down-regulated in CYP27A1 KO group (Figs. 2G and 2H), which indicated the accuracy and reliability of the RNA-sequencing results. The detailed descriptions of the eight genes were listed in Table 2. Transcriptome data have been uploaded to the NCBI database (BioProject ID, PRJNA847887).

Functional enrichment analysis of the DEGs

The function of DEGs was explored by GO and KEGG enrichment analysis. The GO enrichment analysis showed that the DEGs were significantly associated with cholesterol biosynthetic process, cholesterol metabolic process, and response to organic substance, especially with skeletal system development and regulation of cell differentiation, suggesting the key role of CYP27A1 in osteoclast differentiation (Fig. 3A). The KEGG enrichment analysis revealed that the DEGs were mainly involved in metabolic pathways, PPAR signaling pathway, cholesterol metabolism, IL-17 signaling pathway and the PI3K/AKT signaling pathway, especially osteoclast-related pathways, including rheumatoid arthritis, osteoclast differentiation and the TNF signaling pathway (Fig. 3B). Also, the expression of major enrichment pathway related genes was verified by Western blot analysis (Figs. 3C and 3D).

PPI protein network and gene network

To further investigate the interaction of DEGs, a visual PPI network was created using cytoscape (Fig. S1). These key genes interacted and influenced each other, forming a protein interaction network centered on BGN that is closely linked to the osteoclasts differentiation. To further investigate the interaction and function of these core genes, GeneMANIA was applied to construct the gene network that used BGN, S100A9, CYP27A1, SPARC, COL1A2 and ELANE as the central hub with 19 associated genes (Fig. 4). Among these genes, most genes were closely linked to the osteoclasts differentiation. These results suggested the correlation of CYP27A1 to the osteoclast differentiation.