Key circular RNAs identified in male osteoporosis patients by whole transcriptome sequencing

Introduction

Osteoporosis (OP) is a systemic disease with osteopenia and microstructural deterioration that increases the risk of fracture susceptibility, especially in the spine, buttocks, and wrists (Kuo & Chen, 2017). Male osteoporosis is a common age-related degenerative disease, characterized by impaired bone formation and low bone turnover. Postmenopausal osteoporosis is related to excessive bone resorption caused by estrogen deficiency. According to the report of the World Health Organization (WHO), the number of OP patients is increasing rapidly (Cabral et al., 2016), and accounts for about 6.6% of the total population in China (Zhang et al., 2018a). Men have a higher mortality rate after fracture than women (Bliuc & Center, 2016). Previous studies have identified some biochemical indexes of OP, which can be used for the diagnosis and monitoring of patients with OP (Wang et al., 2017), such as serum collagen type I N-terminal pre-peptide (PINP) and cross-linked C-terminal peptide (CTX) (Yang et al., 2019). However, conventional biochemical markers are not effective in determining the possible secondary causes of osteoporosis in men (Fink et al., 2016), while currently available biochemical markers cannot detect all risk factors for fractures (Cheng et al., 2017). Therefore, it is identifying new biomarkers to improve the diagnosis and treatment of osteoporosis in men is imperative.

Accumulating evidence shows that ncRNAs are associated with various diseases through indirect or direct regulation of the corresponding gene expression (Yang et al., 2018). ncRNAs, including miRNA and circRNA, play a crucial role in the occurrence, development, and progression of cancer (Dou et al., 2016). Importantly, previous studies have suggested that circulating miRNAs may be used as a critical biomarker for osteoporosis (Fu et al., 2018; Materozzi et al., 2018).

circRNAs are a class of endogenous, abundant, non-polyadenylated RNAs with a covalently closed, continuous loop structure (Zhang, Yang & Xiao, 2018b). These RNAs are associated with various biological processes, and their dysregulated expression are implicated in human diseases, including diabetes, Alzheimer’s disease, tumors, and cardiovascular disease because of their high stability and prevention from RNA exonuclease degradation (Qiao et al., 2018; Zhao et al., 2017). Interestingly, the complex regulatory interactions between different types of ncRNAs have fundamental roles in the development of multiple diseases (Peng et al., 2018; Zhong et al., 2018). circRNAs are well-known as miRNA sponge in inhibiting the function of miRNA via competing endogenous RNA (ceRNA) network (Zhong et al., 2018). For instance, circRNA-ZNF609, containing multiple binding sites for miR-150-5p, regulates ATK3 expression in Hirschsprung’s disease through ceRNA network (Fu et al., 2018). It has also been proved that circRNA MYLK binds competitively to miRNA29a-3p, thereby increasing the expression of the target genes VEGFA, DNMT3B, and ITGB1, involved in the progression of bladder cancer (Fu et al., 2017). Together, these findings suggested that mRNAs, miRNAs, and circRNAs play a major role in various human diseases, such as osteoporosis (Mandourah et al., 2018).

In the present study, whole transcriptome sequencing was carried out on monocytes from male healthy controls and osteoporosis patients. Key circRNAs involved in the pathogenesis of osteoporosis were identified by bioinformatics analysis. Thus, our findings provide a basis for further in-depth study of pathogenic genes and the rapid, simple diagnosis, and treatment of osteoporosis in men.

Material and Methods

Statistical analysis

qRT-PCR experiment was repeated three times. SPSS was utilized for statistical analysis with independent t-test. P < 0.05 was considered as statistically significant.

Results

Differentially expressed circRNAs, miRNAs and mRNAs

Raw data(raw reads) of fastq format were processed using in-house Perlscripts. Clean data (clean reads) were obtained by removing the reads containing adapter, reads containing ploy-N, and low quality reads from raw data. Differential expression analysis of the two groups was performed using the DESeq R package (1.10.1), which further determined the differential expression in digital gene expression data using a model based on the negative binomial distribution. The resulting P-values were adjusted using the Benjamini and Hochberg’s approach for controlling the false discovery rate. Genes with an adjusted P-value identified by DESeq were termed as differentially expressed.

A total of 12,839 circRNAs, including 5,682 circRNAs, were novel. 398 circRNAs were differentially expressed between OP and healthy control. Of those, 195 circRNAs were upregulated while 203 circRNAs were downregulated (Fig. 3 and Table 2) with the cutoff fold-change > 1 and adjusted P-value (padj) < 0.05. We also identified 642 differentially expressed mRNAs (305 upregulated and 337 downregulated) and 51 miRNAs (28 upregulated and 23 downregulated), respectively.

Table 2:

Differentially expressed circRNAs.

ID	log2FC	Pval	Regulation
hsa_circ_0004276	5.1747	1.55E−49	up
hsa_circ_0003060	4.0244	8.44E−27	up
hsa_circ_0005657	1.8979	1.95E−06	up
novel_circ_0020485	1.8896	2.15E−06	up
hsa_circ_0017615	1.4847	0.000134	up
hsa_circ_0004846	1.4061	0.000253	up
novel_circ_0000968	1.3275	0.000491	up
novel_circ_0003426	1.2845	0.000787	up
hsa_circ_0006132	1.1281	0.002859	up
hsa_circ_0042409	1.1104	0.004767	up
novel_circ_0035291	−1.6125	4.20E−05	down
novel_circ_0048949	−1.0447	0.000322	down
novel_circ_0015289	−1.2924	0.000825	down
novel_circ_0006342	−1.2574	0.000944	down
hsa_circ_0000378	−1.1359	0.002834	down
novel_circ_0038918	−1.0059	0.011125	down
novel_circ_0039344	−1.1813	0.001678	down
hsa_circ_0046964	−1.0626	0.003848	down
hsa_circ_0007976	−1.204	0.001402	down
hsa_circ_0003990	−1.0062	0.008878	down

DOI: 10.7717/peerj.11420/table-2

Discussion

Osteoporosis (OP) is a systemic disease with reduction in bone mass and deterioration of microstructure augmenting the risk of fragility and susceptibility to fracture, especially in the spine, hip, and wrists. Increasing evidence has revealed that ncRNAs participated in various diseases by directly or indirectly regulating the corresponding gene expression. Furthermore, ncRNAs are associated with various diseases through indirect or direct regulation of the corresponding gene expression (Yang et al., 2018). ncRNAs, including miRNA, long lncRNA, and circRNA, play a crucial role in the occurrence, development, and progression of cancer (Dou et al., 2016). Importantly, previous studies have suggested that circmiRNAs may be used as vital biomarkers for osteoporosis (Fu et al., 2018; Materozzi et al., 2018). Wei et al. (2012) found that miR-34s inhibit osteoblast proliferation and differentiation in mice by targeting SATB2. Xia et al. (2016) also discovered that miR-31-5p and miR-424-5p were downregulated in cartilage-derived mesenchymal stem cells (CMSCs) from the degraded cartilage. Moreover, Yin et al. (2018) reported previously that circRUNX2 regulated RUNX2 to prevent osteoporosis via hsa-miR-203.

Herein, we performed whole transcriptome sequencing and investigated the regulatory mechanisms and functions of non-coding RNAs, especially circRNAs in the pathogenesis of osteoporosis. A total of 398 circRNAs, 51 miRNAs, and 642 mRNAs were identified to be significantly differentially expressed in osteoporosis compared to healthy controls. Moreover, GO and KEGG enrichment analysis illustrated that the host genes of significantly differentially expressed circRNAs were mainly enriched in the regulation of cell cycle processes, such as BP, organelle part CC, protein binding MF, Toll-like receptor signaling pathway, TNF signaling pathway, and thyroid hormone signaling pathway. Based on the circRNA-miRNA-mRNA regulatory network, some key circRNAs in osteoporosis were discovered further, such as hsa_circ_0042409, hsa_circ_0001924, hsa_circ_0003990, and hsa_circ_0000983.

hsa_circ_0042409 was linked to 8 miRNAs and 26 mRNAs in the ceRNA regulatory network, and was upregulated in osteoporosis. In addition, it regulated the expression of KLC1, RNH1, CPEN1, and STXBP2 by inhibiting hsa-miR-195-5p, hsa-miR-30b-5p, hsa-miR-32b-5p, hsa-miR-378d, hsa-miR-424b-5p, and hsa-miR-6763-5p. Qu, Chu & Wang (2017) demonstrated that miR-195-5p suppresses osteosarcoma cell proliferation and invasion by suppressing naked cuticle homolog 1. Zhang et al. (2019) discovered that DOC2B promoted insulin sensitivity in mice via a novel KLC1-dependent mechanism in skeletal muscle. Importantly, we found that circRNA hsa_circ_0042409 and KLC1 mRNA were significantly increased while hsa-miR-195-5p was significantly decreased in male osteoporosis patients by whole transcriptome sequencing and further validated by qPCR. These findings coincidentally suggested that key circRNA, hsa_circ_0042409, is associated with the development of osteoporosis. The increased expression of hsa_circ_0042409 regulated the expression level of KLC1 by spongy hsa-mir-195-5P, thus promoting the pathogenesis of osteoporosis.

Conclusions

Although further studies might be needed to support these findings, key cicrRNAs (hsa_circ_0042409 et al.) play a major role in the pathogenesis of osteoporosis and could be used as potential biomarkers or targets in the diagnosis and treatment of osteoporosis.

Supplemental Information