Identication of potent novel biomarkers related to Osteoporosis through bioinformatics approach

Objective Calcium is fundamental component of bone tissue. It is present in extracellular uid in ionized form, some are bounded with albumin while a few are of the complex anionic form. Calcium regulates many biochemical processes. Loss of calcium causes hypocalcemia further lead to osteopenia and osteoporosis. Proton pump inhibitor causes mal-absorption of calcium that leads to poor bone metabolism might causes hip fracture. Some studies via microRNA gene regulatory networks have been analyzed in the present study. bone signaling or osteoporosis. For the further cross validation wet lab experiments are required to validate gene roles. Uniform manifold approximation and projection plot (umap) shows the clustering of dataset with neighborhood scoring 4. Heat-map was constructed for DEG’s, which shows a data matrix where coloring gives an overview of the numeric differences for gene expressions with upregulation in orange and downregulation in blue color code. While gradation from blue to orange shows changing of gene expressions from small to large. mesenchymal and endothelial; organization of extracellular matrix and actin cytoskeleton. While novel DEGs are a part of cellular components like lysosomal, endosome and plasma membrane; synapse and postsynaptic density; cell-cell junction and focal adhesion; cytosol, actin cytoskeleton, cytoplasm and nucleoplasm. Novel DEGs are involved in molecular functions like binding of metal ion, histone acetyltransferase, ephrin receptor, enzyme, laminin, chromatin, sequence-specic and transcription regulatory DNA, identical protein, ion-channel, transcription factor, proteins; ATPase coupled transmembrane transporter and transcription factor Further KEGG pathway reveals that novel DEGs involved in major signicant pathways like NF-kappa B, PI3K-Akt, Wnt and Epithelial cell signaling; regulation of actin cytoskeleton;


Introduction
Calcium is a basic elemental constituent of human body. It is a fundamental constituent of bone structural design and is necessary for deposition of bone mineral throughout the life [1]. About 99% of the body calcium constitutes bone tissue, and remaining are found in the extracellular uid (ECF) [2]. About 50% of calcium that's present in ECF is active in ionized form, 40% is bounded in albumin, while 10% is in complex anionic form ie., phosphate, citrate, sulphate and lactate [3]. Calcium regulates many biochemical processes like blood coagulation, intracellular signal transduction, neural transmission, muscle function, cellular membrane integrity and cellular enzymatic activities, cell differentiation, and bone mineralization [4].
Osteoporotic fractures now causing an enormous public health issue globally and it has been estimated that, around 10 million Americans, over the age of 50 years, have been suffering with it [5]. Accordingly a clinical data analysis estimated that about 1.5 million people per year undergoes an osteoporotic fracture and this can lead to poor health quality and increased risk of death rates [6]. Serum calcium levels<8mg/dL or ionized calcium <4.4mg/dL causes hypocalcemia, while in infants of weight~1500g at birth time, <7mg/dL of the total serum calcium level isleads to hypocalcemia [7]. It's well known that hypocalcemia leads to osteopenia and further osteoporosis (loss of bone density) [8]. In most cases at its early stages neonatal hypocalcemia are asymptomatic, but in late/advanced stages aponea, cyanosis, poor feeding, vomiting, trachicardia, heart failure, prolonged QT interval, tremor, laryngospasm, jerking and twitching episodes, tetanyand seizures are the main generalized clinical features [9]. Primarily three major hormones regulate the calcium homeostasis in our body which are; parathyroid hormone (PTH), 1, 25-dihydroxyvitamin D-3 (Vitamin D3), and calcitonin, and these three-control calcium transport in our gut, bone and kidney [10].
Gastric acid suppression via Proton pump inhibitors (PPIs) causes malabsorption of calcium and hypergastrinemia, and may results in poor bone metabolism through path of induction of hyperparathyroidism [11]. What's more, the diagnosis of OP is intricate until the episode of bone fractures.
So that further study is necessary in order to trace out biomarkers and therapeutic targets that will say something about pathogenesis and molecular mechanisms of osteoporosis. Mutations have been detected in various genes associated with osteoporosis, like CSTA gene interacts with various genes, involved in epidermal development and maintenance, and is associated with the immune regulation of osteoclast [12]. Additionally, FGF21 is inversely coupled with regional bone mass density, affecting the bone development, while its main function is to regulate glucose and lipid metabolism [13]. Even though previous researches have detected regulatory genes and protein linked with osteoporosis [14], still there is need to decipher the candidate biomarkers related to osteoporosis. To generate narrative promises for osteoporosis research, some studies via microRNA gene regulatory networks have been explored here [15]. Identifying DEG's and pre-processing of datasets Benjomini and Hochberg algorithm was used for processing the raw data and was performed next to the retrieval of raw gene expression. Furthermore, log transformation was implemented to the data set by avoiding limma prescision weights and force normalization was done. Probe dataset values related to the particular gene were averaged and then with the help of BiGGEsTs software up-regulated and down regulated genes were selected.
Applying GEO2R(https://www.ncbi.nlm.nih.gov/geo/geo2r/) tool, probe level symbols were exchanged to the gene level symbols. For the assortment of DEG's, parameters were adjusted as, p value<0.05 and threshold logFC values for up regulated genes>0.1 and <-0.1for down regulated genes. Through volcano plot screened DEG's were shown that emphasizes the upregulated and down regulated genes. A mean difference graph were plotted with log2 fold variant versus average log2 expressed gene, with the help of LIMMA package of GEO2R [17].

Generation of heatmap plots and principal component analysis plot
Heatmap plot and principal component analysis (PCA) plot were constructed by applying an online tool ClustVis[18] for DEGs. It was unable to generate whole dataset pro le of PCA, since this tool can support le size only up to 2MB hence only DEGs were analysed for its principal component.

Construction of PPI network and sub network
For the ndings of functional interactions amongst protein, STRING v 10.5 [19](https://www.stringdb.org/), an online tool was used here. For the current study, preferred DEG's with included parameter of combined sore >0.9 were put forwarded for analysis. For the various networks and co-networking creation bioinformatics software Cytoscape V 3.2.1(http://www.cytoscape.org) [20] was applied, and while constructing networks degree and edge betweenenss criteria were employed.
A large set of functionally annotated genes records were integrated by using an online tool DAVID (Database for Annotation, Visualisation and Integrated Discovery) software [21] (https://david.abcc.ncifcrf.gov).DAVID v 6.8 tools were applied for performing gene Ontology (GO) enrichment analysis that includes analysis of molecular function(MF), cellular component (CC) and biological process(BP). Based upon closely associated function DAVID uses an entire set of genes that is based on hypergeometric distribution.  Heat map and Venn diagram of the differentially expressed genes (DEGs). Blue to orange gradation is for small to large changes in gene expression values.   Gene Ontology (GO) enrichment analysis for up-regulated and down-regulated DEG's, and KEGG pathway enrichment for DEG's (5C). Table 1 GENE ONTOLOGY (GO) analysis for novel DEGs related to hypocalcemia and osteoporosis.

Differentially expressed genes (DEGs)
A total of 3390 DEG's were identi ed initially out of which … were found to be up-regulated while …. were found to be down-regulated and these were selected on the basis of their average gene expression values.
Out of 3390 DEGs, only 2473 DEGs were found to be signi cant with p-values (<0.05). In total 2473 DEGs, here 1929 DEG's were found to be novel in which DEGs related to only Hypocalcemia, Bone Signaling and Osteogenesis were 4, 1147 and 1 respectively in their number. While DEGs related to Hypocalcemia and Bone Signaling were 53; Osteogenesis and Bone Signaling Both were 86; and DEGs related to Hypocalcemia, Osteogenesis and Bone Signaling were 8 in number.

Principal component and hierarchical clustering analysis of DEGs
Uniform manifold approximation and projection plot (umap) shows the clustering of dataset with neighborhood scoring 4. Heat-map was constructed for DEG's, which shows a data matrix where coloring gives an overview of the numeric differences for gene expressions with upregulation in orange and downregulation in blue color code. While gradation from blue to orange shows changing of gene expressions from small to large.
The protein-protein interaction and Sub-network construction Based on the combined score calculated by STRING, a total of gene pairs (combined score>0.9) was found to interact together, forming a unique network having nodes and edges respectively. From main network,sub-networks were extracted separately( Figure 4B), the clustering coe cient and edge betweenness were taken as a basic criterion for the selection of hub nodes.

Discussion
Osteoporosis is a systemic bone disease, which leads to the deterioration of microstructure of bone tissues causing lowering of the bone mass and so consequent fracture. The imbalance of bone remodeling process leading to bone resorption and its main pathophysiological process of osteoporosis [22]. Various types of non-skeletal factors add to fracture risk and thus, diagnostic tool for osteoporosis is the estimation of risk factors and Bone Mass Density measurement [22]. [22]. Bone is an active tissue thatis continuously remodeled with the help of speci c bone forming cells, osteoblast and bone resorbing cells, osteoclasts. The imbalance of bone metabolism like decreasing bone resorption causes lower blood calcium level, called as hypocalcemia. Hypocalcemia is de ned as blood calcium level below 8.5mg/dL or ionized blood calcium level lower than 4.6mg/dL in blood plasma [23]. It is most common in advance stage prostate cancer patients and about approximately in 30% of cases [24].
Osteogenesis and bone formation involve many genes, some have positive whilst some have negative impacts [25]. Like WNTB increases osteoblast activity[26] and RUNX2, a transcription factor involved in osteoblast differentiation and shortage of sclerostin (SOST) enhances bone formation [27]. Signaling pathways plays a decisive role in the regulation of osteoblasts and osteoclasts that regulate the bone turn over. Osteoclast activation escort to the loss of bone, and at present therapeutic agents for osteoporosis mainly works on inhibition of bone resorption [28]. PTHrP1-36 stimulates bone formation by reducing bone resorption. Member of TGF that's Activin A stimulates osteoclastogenesis and also involved in the process of inhibition of bone mineralization [29].
Here, using different bioinformatic tools a high throughput gene expression datasets of osteoporosis vs control were studied and as the output it has revealed about 1929 were novel DEGs out of 3390. From total of 1929 novel DEGs, about 1147 DEGs were related to bone signaling only; 86 were related to osteogenesis and bone signaling both; 53 in hypocalcemia and bone signaling both; 8 DEGs were together involved in hypocalcemia, osteogenesis and bone signaling while only 4 were related to hypocalcemia and 1 in osteogenesis.
The hypocalcemia related novel DEGs forming hub nodes are mainly ITCH, CKAP4, FBXW11 and RAB37 from these the rst two are upregulated and rest two are downregulated respectively. ITCH (Itchy E3ubiquitin protein ligase) gene, a regulator of ubiquitination of T-cell receptors, and any mutation in ITCH gene can cause syndromic Multisystem Autoimmune Disease with acute liver failure [30].
CKAP4 protein is upregulated in hepatocellular carcinoma tissues and has been identi ed as speci cally in it [31]. FBXW11, a F-box family members contributing to tumorigenesis and tumour development [32]; [33]. A recent work revealed the role of Fbxw11 in the proliferation of lymphocytic leukemia cells and implies that it can be served as a potential molecular target for the disease treatment [34].
RAB37 is a small GTPase, playing important roles in several cellular processes through intracellular membrane tra c. It has been identi ed as a tumour suppressor and regulates exocytosis of several proteins including TIMP metallopeptidase inhibitor 1 (TIMP1) [35].
Similarly, the bone signaling novel DEGs forming hub nodes mainly include CHML, ATP11A, TMEM30A, YWHAE, AP1M1 and FYN from these the rst three are upregulated and rest three are downregulated respectively.
CHML (Choroideremia-like) protein is essential for the prenylation modi cation of various Rab proteins and it exertsbiological effects on vesicle tra cking and signal transduction. CHML gene is now considered as an independent factor to evaluate the prognosis of Multiple Myeloma, and is associated with poor survival of myeloma cells [36].
However,ATP11A, a ubiquitously expressed gene in various tissues and deleterious effect are lethal for an organism. It plays an important role in myotube formation, while detailed cellular function of ATP11A is still remain exclusive. Mutation in this gene affects localization of Golgi and plasma membrane and Phosphatidylserine ippase activity [37].
TMEM30A(Transmembrane protein 30A), is a ubiquitously expressed terminally-glycosylated membrane protein [38]. The TMEM30A phospholipid ippase complex plays role in cell migration via the formation of membrane ru es as a result of phospholipid translocation [39]. YWHAE (Tyrosine 3-Monooxygenase/Tryptophan 5-Monooxygenase Activation Protein Epsilon), is a protein coding gene and its association has been found with Endometrial Stromal Sarcoma and Kidney Clear Cell Sarcoma. Mutation in this gene has now been considered a risk for Major Depressive Disorder in the Han Chinese Population [40]. AP1M1 (Adaptor protein-1 µ subunit-1), mediates late secretory and vascular tra c and is required for growth [41]. Adaptor proteins (AP) are complexes predominantly as coat proteins of membrane vesicles in post-Golgi tra cking of mammalian cells. AP-1 is crucial for cell division and plant growth.
FYN, is a Src family non-receptor tyrosine kinase, which interacts with tau via SH3 domain [42], where tau is enriched in axons and regulates the microtubule assembly. Fly is found to be critical for neuro brillary tangle formation and tau hyperphosphorylation and if depletion in Fly occurs it causes tau induced neuropathy [43].
GO cluster analysis identi ed several other major up regulated novel DEGs like; CTNNB1, UBE2D1, RAP1A and few major down regulated DEGs like EGFR, MAPK1 and AKT1 which were forming hub nodes.
Among these CTNNB1 (β-catenine-1), a fundamental component of the canonical Wnt signaling pathway that controls cell growth and celladhesion [44]. Nonsense and missense mutations in CTNNB1 were identi ed in patients with ASD [45] and intellectual disability (ID) [46]. UBE2D1 has a crucial role in hepatocellular carcinoma progression, itis one of the family members of E2 ubiquitin conjugating enzyme, mediating the ubiquitination and degradation of tumor suppressor protein p53 [47].
RAP1A, a small G protein similar to Ras oncogene and has role in different cellular processes [48]. Other studies show that RAP1A mediates Glioblastoma cell proliferation [49] and oral cavity squamous cell carcinoma [50].
EGFR (Epidermal Growth Factor Receptor (EGFR), a member of the ErbB family of receptor tyrosine kinase (RTK) proteins, is aberrantly expressed in tumors [51]. The ZNF216 (zinc nger 216) in human carcinoma cells has been proved to be a potential regulator of EGFR activity [52]. MAPK1 (Mitogen-activated protein kinase 1) is a serine/threonine kinase that plays critical roles in several cellular processes like cell proliferation, survival, adhesion, migration via phosphorylation of hundreds of nuclear and cytosolic substrates in the cell. It is a master regulator of stem cell differentiation and is responsible for stem cell fate [53].
A recent study tells MAPK-RAP1A signalling plays an important function as clinical diagnosis and prognostic value in Hepatocellular carcinoma (HCC), and is related with immune in ltration and clinical prognosis [54]. AKT1 (Protein kinase B) is a member of AGC family of serine-threonine kinases and transduces signals through the phosphoinositide 3-kinase (PI3K)/AKT cell-signalling cascade. It is involved widely in signal transduction, metabolism, cell-cycle regulation, transcription, cell-proliferation and angiogenesis processes [55].

Conclusion
Further we can now conclude, that these genes ie; ITCH, CKAP4, FBXW11, RAB37, CHML, ATP11A, TMEM30A, YWHAE, AP1M1, FYN, CTNNB1, UBE2D1, RAP1A, EGFR, MAPK1 and AKT1, whether are upregulated or downregulated in the diseased tissue samples, but must be playing a crucial in the disease progression. These hub genes are involved in different biological, molecular and cellular functions which can directly or indirectly may be related to osteoporosis or hypocalcaemia or bone signalling. Most of these hub genes have their roles in cancer progression while their role in occurrence of osteoporosis has to be depicted. But a clear picture of these genes' role needs wet lab experiments, observations, analysis and further cross validations.

Declarations
Ethics approval and consent to participate Not applicable Consent for publication Not applicable

Availability of data and materials
Request for additional materials can be addressed to the Ravi Bhushan. All the relevant data are enclosed in the manuscript and provided as supplementary le. The RNA-seq. raw data were retrieved from NCBI's Gene Expression Omnibus and are accessible through GEO accession number i.e. GSE35958 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE35958).      Protein-protein interaction (PPI) network of DEG's. Red-circle: up-regulated genes; blue-circle: downregulated genes. Blue diamond for similar related genes and lines shows the correlation between genes, where thickness of lines (edges), is proportional to the combined scores.

Figure 6
Protein-protein interaction (PPI) network of novel and reported DEG's related to bone signaling and hypocalcemia. Red-circle: up-regulated genes; blue-circle: down-regulated genes. Blue diamond for similar related genes and lines shows the correlation between genes, where thickness of lines (edges), is proportional to the combined scores. Figure 7