1. Introduction
Osteoporosis (OP) refers to a clinically common metabolic osteopathy characterized by decreased overall bone mass, bone tissue microstructure destruction, and increased bone fragility and fracture susceptibility.[1] Osteoporotic fracture occurs in the hip (proximal femur), thoracolumbar vertebra, and other parts of the body, which has great harm and is one of the main causes of disability and death in elderly patients. Epidemiological studies have shown that nearly 200 million individuals worldwide are diagnosed as OP annually.[2] Moreover, approximately 9 million individuals suffer from osteoporotic fractures. It is estimated the number is expected to double by 2040.[3] However, only 29% of patients recovered their previous activity after hip fracture.[4] Thus, OP is one of the major public health problems in the world, and the medical treatment and nursing of OP require a large amount of manpower and financial resources, which will impose a heavy family and social burden on societies all over the world.
OP is an imbalance in bone remodeling, as a result of bone resorption by osteoclasts outweighing bone formation by osteoblasts (OBs).[5] Several previous studies utilized the peripheral blood mononuclear cells (PBMs) to explore the pathogenic mechanism of OP by microarray analysis[6–9] and found that circulating monocytes are involved in the process of OP. For example, monocytes can produce several cytokines and growth factors which affect bone metabolism,[10] such as macrophage colony-stimulating factor, interleukin 1, IL-6, tumor necrosis factor (TNF) and transforming growth factor β. In addition, monocytes are also precursors of osteoclasts with bone resorption activity. Under certain conditions, osteoclasts are generated from monocytes in vitro.[11,12]
Bone contains OBs, osteoclasts, osteocytes and so on, which is the most direct site of the pathogenesis of OP and the good material for studying OP. Liu found that SMAD4, CACNG1 and TRIM63 are involved in the molecular mechanisms of OP through silicon analysis of bone biopsy samples, and pointed out that MIR-331 may be a novel potential biomarker of OP.[13] However, there is interaction between different tissues, and the results of a single tissue are not enough to explain changes in the body. Considering the relationship between PBMs and OP, there are few comparative studies on gene expression profiles between PBMs and bone tissue. Therefore, exploring the similarities and differences in the gene expression profiles between PBMs and bone tissue is of the utmost importance for better understanding the mechanisms underlying OP.
In this study, we compared the gene expression profiles of PBMs and bone tissue from patients with OP. Based on the differentially expressed genes (DEGs), Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analysis were performed in the 2 tissues. Then, overlapped pathways in the 2 tissues were selected, followed by an investigation of all DEGs involved in overlapped pathways. Furthermore, according to the above DEGs, a protein-protein interaction (PPI) network and a transcription factors (TFs)-DEGs regulatory network was established to evaluate the potential genes and TFs involved in OP. This study may provide new clues for better understanding the pathogenesis and novel insight for developing diagnostic biomarkers of OP and preventing the progression of OP in the future.
2. Materials and Methods
2.1. Subjects
This study was examined and approved by the Ethics Committee of Clinical Research of Traditional Chinese Medicine of Fujian Academy of Chinese Medical Sciences (Number: NCT01834105). Trials were conducted in compliance with the International Code of Medical Ethics of the World Medical Association.[14] Each participant provided written informed consent.
The data were obtained from OP women who were examined by the special department of OP and OP women who were expected to undergo hip replacements in Fuzhou Second Hospital of Fujian Province. These patients were randomly selected. The 10 patients from the special department of OP of Fujian Academy of Traditional Chinese Medicine were divided into 2 groups: 7 patients with primary OP and 3 healthy women served as normal controls (NC); Age between 52 and 68 years (average 60.6 ± 4.742 years). The 15 patients from Fuzhou Second Hospital of Fujian Province were divided into 12 patients with primary OP and 3 healthy women as NC; Ages were between 52 and 68 years (average 62.35 ± 4.64 years).
Inclusion criteria:
- Those who met the diagnostic criteria of OP, according to “Recommended Diagnostic Criteria for OP in Chinese patients (second draft)”[15];
- All subjects had informed consent and signed the informed consent voluntarily.
Exclusion criteria:
- Those who did not meet the diagnosis of OP;
- Rheumatoid arthritis, diabetes mellitus, hyperthyroidism, and other secondary OP;
- Those with serious cardiovascular and cerebrovascular diseases;
- Those who showed abnormal liver and/or kidney function tests;
- Those who use drugs that may affect BMD, such as vitamin D, bisphosphonate, estrogen, androgen, and thyroid hormone.
All subjects were surveyed by trained doctors using standardized questionnaires to collect data on lifestyle, health status, disease history, and medication history, and recorded the general information of age, height, and weight. The BMD of the lumbar spine (L2–4), and left femoral neck, rotor, ward’s triangle were measured by Medilink Osteocore Dual-Energy X-Ray Ababsorptiometer (Montpellier, France). All BMD measurements were performed by the same trained senior technologist. In addition, the femur (1g) was collected from each patient during hip arthroplasty and stored in liquid nitrogen as soon as possible, and peripheral blood (5 mL) was collected from each patient after a night of fasting (≥12 hours).
2.2. Microarray assay
Mononuclear cells were isolated via lymphocyte lysis and added 1 mL of Trizol reagent. While the femur stored in liquid nitrogen was ground into a powder and added 1 mL of Trizol lysate. Total RNA was isolated using Trizol (Invitrogen, Carlsbad, USA) and an RNeasy Kit (Qiagen, Hilden, Germany), according to the manufacturer’s instructions, including a DNase digestion step. After quantifying the RNA on the Nanodrop-2000 (Thermo Fisher Scientific, Waltham, USA), followed by denaturing gel electrophoresis.
All samples were sent to Shanghai Kangcheng Biology Co. Ltd. for microarray assay. RNA samples were amplified and labeled using the Quick Amp Labeling Kit (Agilent, SantaClara, USA) and hybridized with the Agilent whole-genome oligo microarray in Agilent’s SureHyb Hybridization Chambers. Post hybridization and washing, the processed slides were scanned using Axon Gene Pix 4000B microarray scanner (Molecular Devices, Sunnyvale, USA). The resulting text files were extracted from the Agilent Feature Extraction Software (version 10.5.1.1) and imported into the Agilent GeneSpring GX software (version11.0) for further analysis. Microarray data sets were normalized in GeneSpring GX using the Agilent FE 1-color scenario (mainly median normalization). The positive effect of this median normalization is illustrated in Box-plot, and genes marked present or marginal were chosen for data analysis. The microarray data are publicly available with accession numbers GSE56116 and GSE230665 at the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo/).
2.3. Identification of DEGs in PBMs and bone tissue
DEGs were identified based on the limma package of R (version: 3.6, http://www.bioconductor.org/packages/release/bioc/html/limma.html), with the cutoff criterion of fold change > 2.0 and P < .05.
2.4. Functional enrichment analysis
The GO and KEGG enrichment analysis were performed on DAVID (version: 6.8, https://david.ncifcrf.gov/), with the cutoff criterion of P < .05.
2.5. PPI network construction
The STRING online database (version:10.5, https://string-db.org/) was used to predict and analyze the interactions of DEGs involved in pathways. The parameter score was set at 0.9. PPI networks were visualized using Cytoscape software (version: 3.6.1, https://cytoscape.org/).
2.6. TF-DEGs regulation networks
TFs were predicted with Animal TFDB: (version:3.0, http://bioinfo.life.hust.edu.cn/AnimalTFDB/), with the cutoff criterion of P < .05. TF-DEGs regulation networks were visualized using Cytoscape software (version: 3.6.1, https://cytoscape.org/).
2.7. Reverse transcription-polymerase chain reaction (RT-PCR)
After RNA extraction, 1 ug RNA was used for reverse transcription cDNA. TB GreenTM Premix ExTaqTMII (Takara, Dalian, China) was used to conduct a PCR experiment on real-time ABI 7500 FAST PCR system (Applied Biosystems, Foster city, USA). Primers for the qRT-PCR assay were as follows: RNF215 (forward 5′-GCG GTC GTC TGC TGC TGA TG-3′, reverse 5′-CTG CTC CTT GCC CAC GTA TGC-3′), SULT1B1 (forward 5′-AAG GGA CGG CTG GTG ACT GG-3′, reverse 5′-TGC GGA ATT GAA GTG CAG TTT TGG-3′), ALCAM (forward 5′-GGA AAC TAT GTC TGC GAA ACT G-3′, reverse 5′-CCT TCC ACA TGG CAG ATT ATT G-3′), KRAS (forward 5′-TGT GGA CGA ATA TGA TCC AAC A-3′, reverse 5′-GCA AAT ACA CAA AGA AAG CCC T-3′). The PCR conditions were as follows: Initial denaturation at 95°C for 5 seconds; for 40 cycles 95°C for 30 seconds, 60°C for 30 seconds, followed by a final step of 72°C for 5 minutes. Data were conducted by the relative quantification (ΔΔCq) method. GAPDH expression level was used for normalization.
2.8. Statistical methods
The relative expression level of mRNA was presented as the mean deviation. The Student test was used for analysis group comparison. All statistical hypothesis tests were 2-sided and performed at the 0.05 significance level.
3. Results
3.1. Identification of DEGs in PBMs and bone tissue
Microarray analysis revealed that 226 DEGs were identified between OP and NC in the PBMs of patients. Meanwhile, a total of 2295 DEGs were identified between OP and NC in the bone tissue of patients. After clustering analysis, the expression heatmap of DEGs in PBMs and bone were constructed, respectively (Fig. 1A and B). The volcano maps of DEGs in PBMs and bone between OP and control groups were shown respectively in Figure 1C and D.
;)
Figure 1.: DEGs and overlaps. (A) Heatmaps of DEGs in PBMs between OP and control groups. Red represents upregulated genes, whereas blue represents downregulated genes. (B) Heatmaps of DEGs in bone between OP and control groups. Red represents upregulated genes, whereas blue represents downregulated genes. (C)Volcano maps of DEGs in PBMs between OP and control groups. Red represents upregulated genes, whereas blue represents downregulated genes. (D)Volcano maps of DEGs in bone between OP and control groups. Red represents upregulated genes, whereas blue represents downregulated genes. (E) Diagram of overlapped DEGs. Blue circular represents DEGs in PBMs, yellow circular represents DEGs in bone. DEGs = differentially expressed genes, OP = osteoporosis, PBM = peripheral mononuclear blood cell.
To explore the similarities and differences between the gene expression profiles in PBMs and bone tissue, we compared the expression profiles of the 2 tissues. As shown in Figure 1E, a total of 14 overlaps were obtained by comparing the 226 DEGs with the 2295 DEGs. The table indicated the expression of 14 overlaps were upregulated in the 2 tissues, whereas the expression tendency of DUSP22 was inconsistent in the 2 tissues. Therefore, we obtained 13 common DEGs from the 2 tissues (Table 1).
Table 1 -
Fourteen overlaps in PBMs and bone tissue.
| Gene symbol |
Fold change (PBMs) |
Fold change (bone) |
| RNF215 |
4.41560893 |
9.02646677 |
| BAG4 |
4.23397683 |
2.06496526 |
| TSPAN9 |
3.68805182 |
3.21487311 |
| SULT1B1 |
3.62856558 |
4.76538239 |
| GALNT4 |
3.50655594 |
2.33160087 |
| WDR52 |
3.14810479 |
2.0242468 |
| LONP2 |
2.70739027 |
2.98832884 |
| ZNF347 |
2.56799034 |
2.75936594 |
| HIST2H3A |
2.3699037 |
2.39668176 |
| ALCAM |
2.33186278 |
4.18894662 |
| KRAS |
2.08385357 |
7.2496971 |
| ZAK |
2.07874256 |
2.42119944 |
| TSPAN4 |
2.01288036 |
2.06266851 |
| DUSP22 |
2.38442913 |
0.44783622 |
PBMs = peripheral blood mononuclear cells.
3.2. Functional enrichment analysis
To obtain the biological function of the above DEGs, GO and KEGG pathway analysis were performed in the 2 tissues. As shown in Figure 2A, the top 10 of GO terms were involved in regulation of T cell activation, adaptive immune response based on somatic recombination, leukocyte mediated immunity in PBMs. However, in the bone tissue, the top 10 of GO terms were involved in urea transmembrane transport, renin secretion into blood stream, renal response to blood flow circulatory and renin-angiotensin regulation of systemic arterial blood pressure (Fig. 2B). From the result, we can see DEGs in PBMs were more involved in immune response, while DEGs in bone were more involved in renal response and urea transmembrane transport.
Figure 2.: GO and KEGG pathway analysis of DEGs and overlapped pathways. (A) BP of DEGs in PBMs between OP and control groups. (B) BP of DEGs in bone between OP and control groups. (C) KEGG pathway of DEGs in PBMs between OP and control groups. (D) KEGG pathway of DEGs in bone between OP and control groups. (E) Blue circular represents pathways in PBMs. Yellow circular represents pathways in bone. DEGs = differentially expressed genes, GO = Gene Ontology, KEGG = Kyoto Encyclopedia of Genes and Genomes, OP = osteoporosis, PBM = peripheral mononuclear blood cell.
As with the GO terms, KEGG pathways were identified in the 2 tissues. The top 10 of KEGG pathways were mainly involved in cell adhesion molecules, asthma, allograft rejection, intestinal immune network for IgA production in PBMs (Fig. 2C). And the top 10 of KEGG pathways were mainly involved in ecm-receptor interaction, asthma, allograft rejection, renin-angiotensin system in the bone tissue (Fig. 2D).
Many overlapped pathways were identified in the 2 tissues. So, analysis of common pathways in the 2 tissues was conducted, and 90 common pathways were obtained in the 2 tissues. As shown in Figure 2E, almost all of the pathways in PBMs were overlapped with those in the bone tissue.
3.3. PPI network construction
According to the results of pathway analysis, we selected all DEGs involved in 90 common pathways in PBMs and bone tissue. A total of 352 DEGs were obtained by combining the DEGs set of the 2 tissues. The above 352 DEGs were constructed into PPI network by using the STRING app in Cytoscape software. By removing the separated and separately connected nodes, a complex network of DEGs (combined score > 0.9) was constructed and presented (Fig. 3). The results identified 225 nodes and 1820 interactions. Cytoscape software was applied to show the connectivity of each node in the PPI network. The degrees (>15) of connectivity are shown in Table 2. In this study, the nodes degree (>25) were defined as “hub proteins,” including phosphoinositide-3-kinase regulatory subunit 1 (PI3K1), amyloid beta precursor protein (APP), G protein subunit beta 5 (GNB5), formyl peptide receptor 2 (FPR2), G protein subunit gamma 13 (GNG13), and phospholipase C gamma 1 (PLCG1).
Table 2 -
Connectivity degree of each node in the PPI network.
| Gene |
Degree |
Gene |
Degree |
Gene |
Degree |
| PIK3R1 |
54 |
ITGAV |
18 |
CXCR6 |
15 |
| APP |
38 |
TGFB1 |
18 |
P2RY4 |
15 |
| GNB5 |
32 |
ITGB3 |
17 |
CCL27 |
15 |
| FPR2 |
32 |
HGF |
17 |
LAMC2 |
15 |
| GNG13 |
32 |
CCR2 |
17 |
LTB4R2 |
15 |
| PLCG1 |
27 |
GNA14 |
17 |
HTR1E |
15 |
| EGF |
24 |
STAT3 |
16 |
ADORA3 |
15 |
| ITGA1 |
24 |
CCL19 |
16 |
ADRA1D |
15 |
| SHC1 |
23 |
CXCL11 |
16 |
HCRTR1 |
15 |
| AGTR1 |
23 |
TBXA2R |
16 |
ADRA1B |
15 |
| KRAS |
21 |
PPP2R1A |
16 |
HLA-DQA1 |
15 |
| ADCY6 |
21 |
CD3D |
16 |
F2RL1 |
15 |
| ADCY9 |
21 |
CXCL9 |
15 |
CCL21 |
15 |
| GCGR |
20 |
HLA-DPB1 |
15 |
GHSR |
15 |
| EDNRA |
19 |
HLA-DRB5 |
15 |
|
|
PPI = protein-protein interaction.
Figure 3.: PPI network of DEGs. Circular nodes represent DEGs. Increasing degrees is indicated by larger nodes, labels and darker color. Red represents that the nodes degree > 20, yellow represents that the nodes degree >10 and <20, blue represents that the nodes degree >2 and <10, green represents that the nodes degree = 1. DEGs = differentially expressed genes, PPI = protein-protein interaction.
3.4. TF-DEGs regulation networks
Firstly, we selected 40 DEGs related to OP from 352 genes based on existing literature in PubMed then Enrichr software was applied to predict TFs of the 40 genes related to OP. A total of 12 common TFs were obtained by comparing these TFs and 352 DEGs. At last, 12 TFs were predicted to regulate 35 of the 352 DEGs, and a TF-DEGs regulation network was constructed. As shown in Figure 4, CREB1, RUNX1, STAT3, CREBBP, GLI1, PML, COL11A2, THRA, MYH11, RELA, TBL1X, and SPL1 were involved in regulation of 35 DEGs (e.g.: TNF, CBS, EGF, WNT10B). And some of 12 TFs (CREB1, RUNX1, STAT3, CREBBP, GLI1) were also DEGs. Therefore, the above 5 TFs are considered as key TFs.
Figure 4.: Diagram of TF and TF-DEGs regulation networks. (A) Diagram of TF. List 1 represents all TFs, list 2 represents 352 DEGs. (B) TF-DEGs regulation networks. The red diamond represents TF, and the green circle represents DEGs. DEGs = differentially expressed genes, TFs = transcription factors.
3.5. Validation of DEGs
In order to verify the microarray result, the gene expression levels of RNF215, SULT1B1, ALCAM, and KRAS were detected by RT-PCR in pairs of samples between OP and NC from PBMs of patients. The results indicated that the expression of RNF215, SULT1B1, ALCAM, and KRAS in PBMs was consistent with the results of microarray analysis in PBMs (Fig. 5). Therefore, the RT-PCR data validated the credibility of the microarray result.
Figure 5.: Confirmation of microarray results by quantitative PCR. Based on the overlapped genes in PBMs and bone tissue, we selected 4 common DEGs (RNF215, SULT1B1, ALCAM, KRAS) randomly for further qPCR validation. The PBMs were collected from OP patients (n = 3) and NC patients (n = 3). The expression of RNF215, SULT1B1, ALCAM, KRAS in PBMs was consistent with the results of microarray analysis in PBMs. DEGs = differentially expressed genes, NC = normal control, OP = osteoporosis, PBM = peripheral mononuclear blood cell, qPCR = quantitative polymerase chain reaction.
4. Discussion
In this study, we obtained 13 common DEGs from PBMs and bone tissue. Except for ALCAM, the other 12 DEGs have not been reported related to OP or osteogenesis. ALCAM, namely, CD166, encodes activated leukocyte cell adhesion molecule, which is involved in the processes of cell adhesion and migration. Most articles focus on ALCAM regulation of hematopoietic function.[16,17] It has been reported that global ALCAM deficiency leads to a marked reduction in hemopoietic stem cells, as well as promoting bone formation, indicating an increase in trabecular bone fraction, OB number, and bone formation rate. Additionally, CD166−/− OB shows increased alkaline phosphatase activity and mineralization.[18] CD166 expression in multiple myeloma cells promoted osteoclastogenesis by activating TRAF6-dependent signaling pathways.[19] In our study, ALCAM expression is upregulated in PMBs and bone tissue from OP groups. Thus, these findings suggest that ALCAM was associated with low BMD, which could be potential diagnostic biomarkers for OP in the future.
The GO analysis indicated that DEGs in PBMs were more involved in immune response, combining the literature reviews, we speculated monocytes can regulate bone metabolism by several cytokines related to immunity, such as interleukin 1, IL-6, and TNF. That is the reason why PBMs were more involved in immune response. While DEGs in bone were more involved in renal response and urea transmembrane transport in this study. Modern histoembryology has confirmed that both kidney and bone are derived from the mesoderm, indicating the kidney is related to the growth and development of bone. The kidney can change 25-(OH)-VitD3 into the more active 1,25-(OH)-2-VitD3, and then regulate calcium and phosphorus metabolism to maintain blood calcium level, promoting bone calcification and bone salt deposition through a series of regulatory effects. Our results also confirmed the kidney is involved in the molecular mechanism of OP. Through KEGG pathway analysis, 90 common pathways were identified in the 2 tissues. We can see that almost all of the pathways in PBMs were overlapped with those in the bone tissue. Perhaps monocytes are just precursors of osteoclasts, while bone tissue includes OBs, osteoclasts, and osteocytes and so on. Therefore, we conclude that studies of OP from PBMs are limited, while bone tissue involves more metabolic pathways. It may be the best material for studying the molecular mechanism of OP.
Through PPI network construction, we obtained 6 hub proteins in this study: PI3K1, APP, GNB5, FPR2, GNG13, and PLCG1. According to the literature survey, APP has been found to be associated with OP, whereas PI3K1, FPR2, and PLCG1 could be implicated in associated with OP. Additionally, there is still no evidence to support the association between GNB5, GNG13, and OP.
APP encodes a cell surface receptor and transmembrane precursor protein, and its physiological role remains unclear. However, its contribution to the pathogenesis of OP has been extensively studied. A previous study noted that the mRNA and protein expression levels of APP were significantly increased in the osteoporotic bone tissues compared with control groups both from human and ovariectomized rats.[20] Additionally, Cui demonstrated a role for APP in regulating osteoclast differentiation in vitro and in vivo.[21] And mutant APP has been shown to suppress OB differentiation and bone formation in culture and in mouse.[22] Furthermore, mutant APP mice, which exhibited a low BMD phenotype by 9 months, are susceptible to OP.[23,24] These studies have indicated APP is related with low BMD.
A recent study has reported that increased PI3K1 expression can promote MC3T3-E1 cell proliferation and inhibit MC3T3-E1 cell apoptosis during fracture healing. The study suggests that circRNA AFF4 promotes fracture healing by targeting the miR-7223-5p/PIK3R1 axis.[25] Choudhary found that FPR2 mediated the inhibitory actions of Saa3 on OBs differentiation.[26] PLCG1 plays an important role in the transduction of receptor-mediated tyrosine kinase activators using calcium as a cofactor. A study demonstrated PLCG1 were activated by RANKL during osteoclastogenesis, produces inositol-1,4,5-trisphosphate, which prompts calcium release from endoplasmic reticulum. Then the reaction enhances intracellular calcium, thereby induces NFATc1 dephosphorylation, and finally induces many osteoclast specific genes expression.[27] Another study indicated by activation PLCG1 phosphorylation, RGS10 regulates osteoclast differentiation an in vivo model.[28] In our study, PI3K1, FPR2, PLCG1 were indented as hub proteins, which indicated these proteins may play important roles in OP. However, there is no direct evidence of a connection between these genes and OP. Perhaps they can be used as objects to study the molecular mechanism of OP in the future.
In the present study, we identified 5 key TFs by TF-DEGs regulation networks analysis: CREB1, RUNX1, STAT3, CREBBP, GLI1, and CREB1, as a homodimer, binds to the cAMP-responsive element to be an octameric palindrome, then induces transcription of genes. Several studies indicated cAMP-PKA-CREB signaling pathway promotes the osteogenic response of BMSCs and OBs in vitro and in vivo.[29–32] RUNX1, as the master regulator of hematopoiesis, is expressed in preosteoclasts and negatively regulates osteoclast formation and activity.[33] Genome-wide association study meta-analysis has indicated SNP of RUNX1 are involved in endochondral bone formation.[34] And several studies have reported STAT3 can promote osteogenic differentiation of OB and inhibit osteoclast differentiation.[35,36] CREB-binding protein (CBP), functions mainly as a coactivator of many TFs, or acetylates both histone and non-histone proteins. Mutations in this gene cause Rubinstein–Taybi syndrome, including a marked decrease in trabecular bone that was predominantly caused by increased osteoclastogenesis.[37] Another analysis provides evidence that CBP can promote OB differentiation and osteoblastic cells growth by mediating other gene acetylation.[38,39] Greenblatt’s study revealed that the TAK1–MKK3/6–p38 MAPK axis phosphorylated Runx2, promoting its association with the coactivator CBP, which was required to regulate OB genetic programs.[40] GLI1, encodes a member of the Kruppel family of zinc finger proteins. The encoded TF is activated by the sonic hedgehog signal transduction cascade. Hedgehog-Gli1 signaling plays an essential role in OB proliferation and differentiation as well as bone formation.[41,42] Moreover, Hedgehog-Gli1 signaling is reported to accelerate bone fracture healing.[43,44] The above studies and our results suggest that these key TFs may be the important regulator of OP.
In conclusion, we have successfully inferred several potential diagnostic biomarkers for OP based on the combination of multi grouping data. Remarkably, we found that PBMs were more involved in immune response, while bone was more involved in renal response and urea transmembrane transport. And many overlapped pathways were identified in the 2 tissues. Compared with PBMs, bone is a more suitable material to study the pathogenesis of OP; however, it is not easy to obtain certainly. In addition, there are some limitations in this study that should be acknowledged. First of all, PBMs and bone tissue did not originate from the same batch of people, and the samples are not particularly sufficient. Similarly, when using RT-PCR to verify DEGs, we should also pay attention to the sample size. Secondly, there are regional limitations in this study, but it is hard to avoid. Despite these shortcomings, we are still satisfied because the study enhances our understanding of the potential molecular mechanisms of OP and provides targets for further research.
Acknowledgments
We thank all the participants for their willingness to participate in this study.
Author contributions
Conceptualization: Lihua Xie, Jirong Ge.
Data curation: Shengqiang Li, Juan Chen.
Formal analysis: Lihua Xie.Resources: Eryou Feng.Writing – original draft: Lihua Xie.
Writing – review & editing: Lihua Xie, Hao Chai, Li Li, Jirong Ge.
References
[1]. Nuti R, Brandi ML, Checchia G, et al. Guidelines for the management of
osteoporosis and fragility fractures. Intern Emerg Med. 2019;14:85–102.
[2]. Xie Y, Zhang L, Xiong Q, et al. Bench-to-bedside strategies for osteoporotic fracture: from osteoimmunology to mechanosensation. Bone Res. 2019;7:25.
[3]. Yaacobi E, Sanchez D, Maniar H, et al. Surgical treatment of osteoporotic fractures: an update on the principles of management. Injury. 2017;48:S34–40.
[4]. Hansson S, Rolfson O, Åkesson K, et al. Complications and patient-reported outcome after hip fracture. A consecutive annual cohort study of 664 patients. Injury. 2015;46:2206–11.
[5]. Yang C, Ren J, Li B, et al. Identification of gene biomarkers in patients with postmenopausal
osteoporosis. Mol Med Rep. 2019;19:1065–73.
[6]. Liu YZ, Zhou Y, Zhang L, et al. Attenuated monocyte apoptosis, a new mechanism for
osteoporosis suggested by a transcriptome-wide expression study of monocytes. PLoS One. 2015;10:e0116792.
[7]. Xiang-Ding C, Peng X, Shu-Feng L, et al. Gene expression profiling in monocytes and SNP association suggest the importance of the STAT1 gene for
osteoporosis in both Chinese and Caucasians. J Bone Miner Res. 2010;25:339–55.
[8]. Yao-Zhong L, Volodymyr D, Yan L, et al. A novel pathophysiological mechanism for
osteoporosis suggested by an in vivo gene expression study of circulating monocytes. J Biol Chem. 2005;280:29011–6.
[9]. Li JJ, Wang BQ, Fei Q, et al. Identification of candidate genes in
osteoporosis by integrated microarray analysis. Bone Joint Res. 2016;5:594–601.
[10]. Fischer V, Haffner-Luntzer M. Interaction between bone and immune cells: Implications for postmenopausal
osteoporosis. Semin Cell Dev Biol. 2022;123:14–21.
[11]. Sprangers S, Schoenmaker T, Cao Y, et al. Different blood-borne human osteoclast precursors respond in distinct ways to IL-17A. J Cell Physiol. 2016;231:1249–60.
[12]. Yukiko K, Toshihiro N, Kenji H, et al. Identification of a human peripheral blood monocyte subset that differentiates into osteoclasts. Arthritis Res Ther. 2006;8:R152.
[13]. Liu Y, Wang Y, Yang N, et al. In silico analysis of the molecular mechanism of postmenopausal
osteoporosis. Mol Med Rep. 2015;12:6584–90.
[14]. Lundberg GD. The international code of medical ethics of the World Medical Association. MedGenMed: Medscape General Medicine. 2004;6:37.
[15]. Liu ZH, Zhu HM, Wang HF, et al. Chinese people recommended diagnostic criteria for
osteoporosis (second draft). Chin J Osteoporos. 2019;6:1–3.
[16]. Chitteti BR, Bethel M, Kacena MA, et al. CD166 and regulation of hematopoiesis. Curr Opin Hematol. 2013;20:273–80.
[17]. Chitteti BR, Cheng YH, Kacena MA, et al. Hierarchical organization of osteoblasts reveals the significant role of CD166 in hematopoietic stem cell maintenance and function. Bone. 2013;54:58–67.
[18]. Hooker RA, Chitteti BR, Egan PH, et al. Activated leukocyte cell adhesion molecule (ALCAM or CD166) modulates bone phenotype and hematopoiesis. J Musculoskelet Neuronal Interact. 2015;15:83–94.
[19]. Xu L, Mohammad KS, Wu H, et al. Cell adhesion molecule CD166 drives malignant progression and osteolytic disease in multiple myeloma. Cancer Res. 2016;76:6901–10.
[20]. Li S, Liu B, Zhang L, et al. Amyloid beta peptide is elevated in osteoporotic bone tissues and enhances osteoclast function. Bone. 2014;61:164–75.
[21]. Cui S, Xiong F, Hong Y, et al. APPswe/Aβ regulation of osteoclast activation and RAGE expression in an age-dependent manner. J Bone Miner Res. 2011;26:1084–98.
[22]. Xia WF, Jung JU, Shun C, et al. Swedish mutant APP suppresses osteoblast differentiation and causes osteoporotic deficit, which are ameliorated by N-acetyl-L-cysteine. J Bone Miner Res. 2013;28:2122–35.
[23]. Yang MW, Wang TH, Yan PP, et al. Curcumin improves bone microarchitecture and enhances mineral density in APP/PS1 transgenic mice. Phytomedicine. 2011;18:205–13.
[24]. Zhao L, Liu S, Wang Y, et al. Effects of curculigoside on memory impairment and bone loss via anti-oxidative character in APP/PS1 mutated transgenic mice. PLoS One. 2015;10:e0133289.
[25]. Mi B, Xiong Y, Chen L, et al. CircRNA AFF4 promotes osteoblast cells proliferation and inhibits apoptosis via the Mir-7223-5p/PIK3R1 axis. Aging (Milano). 2019;11:11988–2001.
[26]. Choudhary S, Goetjen A, Estus T, et al. Serum amyloid A3 secreted by preosteoclasts inhibits parathyroid hormone-stimulated cAMP signaling in murine osteoblasts. J Biol Chem. 2016;291:3882–94.
[27]. Zeng X, Zhang Y, Wang S, et al. Artesunate suppresses RANKL-induced osteoclastogenesis through inhibition of PLCγ1-Ca-NFATc1 signaling pathway and prevents ovariectomy-induced bone loss. Biochem Pharmacol. 2017;124:57–68.
[28]. Yang S, Li YP. RGS10-null mutation impairs osteoclast differentiation resulting from the loss of [Ca2+].i oscillation regulation. Genes Dev. 2007;21:1803–16.
[29]. Wang YY, Pu XY, Shi WG, et al. Pulsed electromagnetic fields promote bone formation by activating the sAC-cAMP-PKA-CREB signaling pathway. J Cell Physiol. 2019;234:2807–21.
[30]. Park KH, Choi Y, Yoon DS, et al. Zinc promotes osteoblast differentiation in human mesenchymal stem cells via activation of the cAMP-PKA-CREB signaling pathway. Stem Cells Dev. 2018;27:1125–35.
[31]. Chen M, Cui Y, Li H, et al. Icariin promotes the osteogenic action of BMP2 by activating the cAMP signaling pathway. Molecules (Basel, Switzerland). 2019;24:3875.
[32]. Kim JM, Choi JS, Kim YH, et al. An activator of the cAMP/PKA/CREB pathway promotes osteogenesis from human mesenchymal stem cells. J Cell Physiol. 2013;228:617–26.
[33]. Paglia DN, Yang X, Kalinowski J, et al. Runx1 regulates myeloid precursor differentiation into osteoclasts without affecting differentiation into antigen presenting or phagocytic cells in both males and females. Endocrinology. 2016;157:3058–69.
[34]. Baird DA, Evans DS, Kamanu FK, et al. Identification of Novel Loci Associated With Hip Shape: A Meta-Analysis of Genomewide Association Studies. J Bone Miner Res. 2019;34:241–51.
[35]. Yang Y, Chung MR, Zhou S, et al. STAT3 controls osteoclast differentiation and bone homeostasis by regulating NFATc1 transcription. J Biol Chem. 2019;294:15395–407.
[36]. Jo S, Wang SE, Lee YL, et al. IL-17A induces osteoblast differentiation by activating JAK2/STAT3 in ankylosing spondylitis. Arthritis Res Ther. 2018;20:115.
[37]. Zimmer SN, Zhou Q, Zhou T, et al. Crebbp haploinsufficiency in mice alters the bone marrow microenvironment, leading to loss of stem cells and excessive myelopoiesis. Blood. 2011;118:69–79.
[38]. Lu J, Qu S, Yao B, et al. Osterix acetylation at K307 and K312 enhances its transcriptional activity and is required for osteoblast differentiation. Oncotarget. 2016;7:37471–86.
[39]. You W, Song L, Wang K. Acetylation of GATA4 on lysine residue K313 promotes osteoblastic cells growth. Cell Physiol Biochem. 2018;46:269–78.
[40]. Greenblatt MB, Shim JH, Zou W, et al. The p38 MAPK pathway is essential for skeletogenesis and bone homeostasis in mice. J Clin Invest. 2010;120:2457–73.
[41]. Lin Y, Huang Y, He J, et al. Role of Hedgehog-Gli1 signaling in the enhanced proliferation and differentiation of MG63 cells enabled by hierarchical micro-/nanotextured topography. Int J Nanomedicine. 2017;12:3267–80.
[42]. James AW, Leucht P, Levi B, et al. Sonic Hedgehog influences the balance of osteogenesis and adipogenesis in mouse adipose-derived stromal cells. Tissue Eng Part A. 2010;16:2605–16.
[43]. Shi Y, He G, Lee WC, et al. Gli1 identifies osteogenic progenitors for bone formation and fracture repair. Nat Commun. 2017;8:2043.
[44]. Takebe H, Shalehin N, Hosoya A, et al. Sonic hedgehog regulates bone fracture healing. Int J Mol Sci. 2020;21:677.
