Role of MicroRNAs in Bone Pathology during Chikungunya Virus Infection

Chikungunya virus (CHIKV) is an alphavirus, transmitted by mosquitoes, which causes Chikungunya fever with symptoms of fever, rash, headache, and joint pain. In about 30%–40% of cases, the infection leads to polyarthritis and polyarthralgia. Presently, there are no treatment strategies or vaccine for Chikungunya fever. Moreover, the mechanism of CHIKV induced bone pathology is not fully understood. The modulation of host machinery is known to be essential in establishing viral pathogenesis. MicroRNAs (miRNAs) are small non-coding RNAs that regulate major cellular functions by modulating gene expression. Fascinatingly, recent reports have indicated the role of miRNAs in regulating bone homeostasis and altered expression of miRNAs in bone-related pathological diseases. In this review, we summarize the altered expression of miRNAs during CHIKV pathogenesis and the possible role of miRNAs during bone homeostasis in the context of CHIKV infection. A holistic understanding of the different signaling pathways targeted by miRNAs during bone remodeling and during CHIKV-induced bone pathology may lead to identification of useful biomarkers or therapeutics.


Antiviral Role of miRNAs
The generation of antiviral miRNAs has been observed in many viral infections [21,22]. Computational analysis showed a number of significantly modulated miRNAs in early CHIKV infection are involved in apoptosis and JAK-STAT signaling pathways [32]. Interestingly, the JAK/STAT pathway is known to be one of the key signaling pathways in the interferon (IFN) response against viral infection [33]. Reverse genetic approaches and functional studies in Ae. aegypti mosquitoes revealed that increased resistance to Dengue virus (DENV) and Zika virus (ZIKV) infections is mediated by the JAK/STAT pathway [34]. Moreover, CHIKV non-structural protein 2 (nsP2), has been associated with the JAK/STAT pathway [35]. Thus, evaluating the interaction among viral proteins, miRNAs and their involvement in the JAK-STAT pathway holds potential for exploratory studies in CHIKV pathogenesis. miRNA profiling in CHIKV-infected human skin fibroblasts showed differential expression of a number of miRNAs in the early stage of CHIKV infection [26]. The miRNAs were predicted to target immune-related signaling pathways including JAK/STAT, MAPK, WNT, and retinoic acid inducible gene I (RIG-I)-like receptor pathways [26]. Interestingly, both JAK/STAT and MAPK pathways have been associated with CHIKV infection [35,36]. Additionally, the WNT signaling pathway can regulate IFN response in flaviviruses [37]. The expressions of hsa-miR-15 and hsa-miR-16 were altered during CHIKV infection [26]. In normal physiology, a number of cellular processes are regulated by hsa-miR-15 and hsa-miR-16 and altered expression of these miRNAs is observed in many other viral infections and diseases [38][39][40][41][42]. Interestingly, hsa-miR-15 and hsa-miR-16 play important roles in inducing apoptosis by targeting the anti-apoptotic protein BCL2 [43]. Additionally, downregulated expression of hsa-miR-15 was found in arthritic synovial tissue, whereas hsa-miR-16 level was high in sera of RA patient [38,44]. rno-miR-32-5 p is a negative regulator of phosphatase and tensin homolog (PTEN) [45]. Thus, understanding the functional relevance of these miRNAs during CHIKV infection would be helpful for the development of novel drug targets.

Pro-Viral Role of miRNAs
The ability of miRNAs to regulate gene expression makes them particularly useful for viruses. Often viruses employ cellular miRNAs to target specific genes and downregulate their expression to establish infection [21,22]. The expression of hsa-miR-146a was found to be upregulated in CHIKV-infected human synovial fibroblasts where TRAF6 and IRAK1 were predicted as targets (Table 1) [24]. The expression of these targets was restored in cells transfected with hsa-anti-miR-146a. In addition, overexpression of hsa-miR-146a leads to decreased phosphorylation of NF-кB during infection [24]. The conclusions were similar to another finding which demonstrated that increased expression of hsa-miR-146a enhanced DENV replication by targeting the TRAF6-mediated NF-кB pathway [49]. These results suggested a role of hsa-miR-146a-mediated targeting of the NF-кB pathway during CHIKV pathogenesis.
In human synovial fibroblasts, an miRNA microarray identified a subset of 26 differentially expressed miRNAs (DEMs) during CHIKV infection (Table 1) [25]. Among the DEMS, expression of hsa-miR-4717-3p, hsa-miR-4299, hsa-miR-1264, and hsa-miR-21-5p were significantly upregulated. AKT3 was predicted as a target for hsa-miR-4717-3p (Table 1). The AKT3 protein is a key regulator of the PI3K/AKT/mTOR signaling pathway which influences various cellular processes including metabolism, growth, proliferation, survival, transcription, and protein synthesis [50]. Moreover, dysregulation of the PI3K/AKT/mTOR pathway has been implicated in many diseases [51,52]. Interestingly, this pathway is moderately activated during CHIKV infection [53]. Thus, hsa-miR-4717-3p may mediate suppression of the robust inflammatory response during CHIKV infection, by targeting AKT3 [25]. In response to an infection, the host cellular system elicits a cytokine mediated immune response, but often due to inefficient pathogen clearance, the immune response results in inflammation [54]. The expression of hsa-miR-4299 was upregulated during CHIKV infection and the suppressor of cytokine signaling 7 protein (SOCS7) was predicted as the target (Table 1) [25]. SOCS7 is known to negatively regulate the STAT3 protein which can either induce IL-6-mediated inflammation or IL-10-mediated suppression of inflammation during CHIKV infection [55]. STAT3 can also promote viral replication and persistence [56]. Thus, hsa-miR-4299 could mediate suppression of SOCS7 which may result in increased STAT3 expression contributing to suppressed immune response during CHIKV infection. As earlier mentioned, viruses employ different strategies to exploit cellular pathways for optimizing chances of survival [57,58]. Agrawal et al. showed that expression of hsa-miR-1264 increased during CHIKV infection and TRIM26 was predicted to be its target (Table 1) [25]. TRIM proteins function as E3 ubiquitin ligase playing an important role in antiviral responses through ubiquitination and proteasomal degradation of IRF3 genes during viral infections [59]. Thus, this indicates that CHIKV infection may lead to hsa-miR-1264-mediated suppression of TRIM26 resulting in depleted antiviral response and enhanced viral replication and persistence [25]. Another E3 ubiquitin ligase, PELI1 was targeted by hsa-miR-21-5p whose expression was upregulated during infection (Table 1) [25]. PELI1 can suppress the NF-кB pathway by ubiquitination and degradation of an NF-кB-inducing kinase (NIK) [60]. As the NF-кB pathway plays a critical role in antiviral response, thus, increased expression of miR-21 during CHIKV infection may contribute to suppression of cytokine signaling by modulating the NF-кB pathway.
A genome-wide miRNA screen using high throughput RNA sequencing in Huh-7.5.1 cells revealed that alphaviruses have a binding site for hsa-miR-124 [27]. A significant increase in CHIKV production was observed on overexpressing hsa-miR-124, whereas inhibiting hsa-miR-124 led to reduced CHIKV infection. In rare cases, CHIKV infection can result in encephalitic symptoms [9,10]. hsa-miR-124 is predominantly found in neurons and act as a key negative regulator of neuroinflammation [61]. An altered expression of hsa-miR-124 has been associated with brain disease [61]. Thus, it would be interesting to evaluate whether hsa-miR-124 is associated with encephalitic pathology during CHIKV infection. In a study by Nakamachi et al., decreased expression of hsa-miR-124 was observed in fibroblast like synoviocytes (FLS) of patients with RA where hsa-miR-124a contributed to the inflammatory processes in RA pathogenesis by targeting the monocyte chemoattractant protein-1 (MCP-1) and cyclin-dependent kinase-2 (CDK-2) [62]. Thus, hsa-miR-124 may have a role in contributing to inflammation observed during CHIKV infection.

Aberrant Expression of miRNAs in Mosquito Cells during CHIKV Infection
To establish infection and increase virus survival in a mosquito vector, viruses modify the transcriptional profile of the vector [63]. In Aag-2 cells, aae-miR-2944b-5p and aae-miR-2b were observed to have binding sites for the 3 UTR of CHIKV [29]. When mosquitoes were treated with antagomiR-2944b-5p, they showed more susceptibility to CHIKV infection compared to untreated control which suggested the role of antagomiR-2944b-5p in viral replication. The host vacuolar protein sorting-associated protein 13 (VPS-13) was predicted as a target of aae-miR-2944b-5p. In Ae. aegypti, VPS-13 functions in maintaining the mitochondrial membrane potential (MtMP) [29]. Interestingly, studies report that host mitochondria are involved in combating the oxidative stress induced during viral infections [64]. Silencing aae-miR-2944b-5p in Aag-2 cells and infecting with CHIKV increased cellular MtMP, which indicated that aae-miR-2944b-5p interacts with VPS-13 to maintain MtMP [29]. In humans, VPS-13 is involved in post Golgi apparatus sorting and trafficking. Thus, studying the effect of hsa-miR-2944b-5p on VPS-13 expression in human cell lines during CHIKV infection can lead to identification of novel drug target.
Using next generation RNA sequencing, the expressions of a set of eight miRNAs were found to be altered during CHIKV infection [30]. Among them, the expressions of aae-miR-100, aae-miR-283, aae-miR-305-3p, and aae-miR-927 were significantly upregulated and the expressions of aae-miR-1000, aae-miR-2b, aae-miR-2c-3p, and aae-miR-190-5p were downregulated. Target prediction revealed that aae-miR-100, aae-miR-283, and aae-miR-305-3p commonly affected NK cell-mediated cytotoxicity and protein processing in ER pathways. The analysis also revealed that the metabolic pathways such as the TCA cycle, dorso-ventral axis formation, and valine, leucine, and isoleucine degradation pathways were affected by aae-miR-100 and aae-miR-305-3p. aae-amiR-927 and aae-miR-305-3p were predicted to target SNARE interactions in vesicular transport. Among these, aae-miR-305-3p was predicted to target pathways essential for viral entry such as ECM receptor interaction, endocytosis, and SNARE interactions in vesicular transport. The downregulated aae-amiR-1000, aae-miR-2b, and aae-miR-2c targeted the ribosomal pathway. The upregulated miRNAs targeted genes which encodes for protein tyrosine phosphatase SHP2, ERK1/2, and ubiquitin fusion degradation protein, respectively, whereas the downregulated miRNAs targeted the gene that encodes for the 40S ribosomal protein S16. In another study, next-generation sequencing identified the altered expression of 13 miRNAs during CHIKV infection in Aag-2 cells [31]. Target prediction analysis showed aae-miR-2b targets URM and ubiquitin whereas aae-miR-100 targets CDC42 and sumo-ligase. When cells were treated with aae-antagomiR-2b, increased CHIKV replication was observed. The expression of URM was also significantly high in CHIKV infected cells. Furthermore, CHIKV replication was reduced to 50% in URM knock down cells, indicating that aae-miR-2b-mediated regulation of URM plays a significant role in chikungunya replication.
Usually in mosquito vectors, viruses establish infection in the salivary gland during a blood meal [63]. For establishing a successful infection, viruses often modulate the gene expression of several proteins in the salivary gland [63]. Next generation sequencing showed that aae-miR-bantam, aae-miR-263a, aae-miR-125, and aae-miR-285 were significantly upregulated in CHIKV-infected Ae. aegypti saliva [65]. In Ae. albopictus saliva, aal-miR-43b, aal-miR-43a, aal-miR-413a, aal-miR-5, and aal-miR-249 were upregulated [65]. In addition, Aag-2 cells and BHK-21 cells showed decreased CHIKV titers when treated with inhibitors against selected miRNAs indicating the role of salivary gland miRNAs in modulating CHIKV replication. Another study predicted a set of miRNAs that commonly targeted the different genotypes of CHIKV where aae-miR-282-5p, aae-miR-34-3p, and aae-miR-11-5p had binding sites for CHIKV [66]. Moreover, aae-miR-11-5p was conserved among the different lineages of CHIKV and was predicted to target the end of subgenomic untranslated RNA region, thus, indicating that the CHIKV structural proteins may regulate the formation of a miRNA-viral RNA (vRNA) complex at the end of subgenomic RNA untranslated regions, thereby preventing the binding of host translational factors on vRNA.

Possible Role of miRNAs in Bone Homeostasis in the Context of CHIKV Infection
CHIKV infection is associated with bone pathology and it was first indicated by the presence of bony lesions in CHIKV-infected IRF 3/7 -/mice [67]. In CHIKV-infected patients, MRI results showed the presence of erosive arthritis [68]. Bone is one of the most dynamic organs in the body that continuously undergoes remodeling. Bone homeostasis is a highly regulated and complex process involving a fine balance between osteoblastogenesis and osteoclastogenesis [69]. Osteoblastogenesis is the process of bone formation which results from differentiation of mesenchymal stem cells (MSCs) into osteoblastic cell lineage forming the bone cells or osteoblasts (OBs) and later into osteocytes, the mature OBs [70]. Conversely, osteoclastogenesis is the process of bone resorption where the formation of multinucleated osteoclasts (OCs) occurs from the fusion of myeloid precursors which arise by differentiation of hematopoietic stem cells (HSCs) [71]. Many complex processes, signaling pathways, and transcription factors govern osteoblastogenesis and osteoclastogenesis in maintaining normal bone homeostasis (Figures 1 and 2).   During osteoblastogenesis, the key signaling pathways activated are canonical WNT, NOTCH, Hedgehog, BMP, SMAD, MAPK, and the receptor activator of nuclear factor kβ (RANK), osteoprotegerin (OPG)-and RANK ligand (RANKL) [72,73]. These pathways result in expression of the key transcription factors identified during osteoblastogenesis which are the runt-related transcription factor 2 (RUNX2), and Osterix (OSX) [74,75]. However, there are other transcription factors that also function in bone homeostasis [76,77]. These transcription factors subsequently induce the expression of other osteogenic genes including alkaline phosphatase (ALP), type I collagen (COl-I), osteocalcin (OCN), osteonectin (ON), and bone sialoprotein (BSP) [76][77][78]. Similarly, processes such as OC differentiation from myeloid precursors, maturation, and survival of the OC are regulated by a variety of environmental factors including cytokines, growth factors, and hormones which influence the RANK-RANKL, MAPK, PI3K/AKT, and NF-kβ pathways [72,73]. These signaling pathways in turn regulate the expression of various transcription factors among which nuclear factor of activated T-cells, cytoplasmic 1 (NFATC1) is critical [79,80]. NFATC1 is the major regulator of the early phase of osteoclastogenesis, which induces the expression of other osteoclastic genes in the late phase such as tartrate-resistant acid phosphatase (TRAP), cathepsin k (CTSK), and dendrocyte expressed seven transmembrane proteins (DCSTAMP) [79,[81][82][83][84][85]. Conversely, few transcription factors can also negatively regulate osteoclastogenesis [86][87][88][89][90][91].
In osteogenically differentiated MSCs derived from dental and craniofacial tissues, the expression of hsa-miR-21 was down-regulated, and its overexpression suppressed osteoblastogenesis [94]. SMAD5, the upstream regulator of RUNX2 during osteogenesis, was the target of hsa-miR-21 [94]. However, in mouse osteoblast MC3T3-E1 cells, mmu-miR-21 induced osteogenic differentiation by targeting SMAD7 [95]. Sun et al., also showed that mmu-miR-21 induced osteogenesis as overexpression of mmu-miR-21 resulted in increased mineralization and bone healing properties in a femur fracture model in rats [143]. During CHIKV infection, an upregulated expression of hsa-miR-21-5p was observed, and PELI1, a E3 ubiquitin protein ligase, was predicted as the target [25]. PELI1 has been shown to inhibit the NF-κB signaling pathway, which is an important pathway during osteoclastogenesis [60]. Thus, further studies may be conducted to investigate whether miR-21 can impair osteoclastogenesis during CHIKV infection. FAK signaling pathway acts as a critical signaling pathway in the early stages of osteoblastogenesis [144]. During CHIKV infection, hsa-miR-138-2-3p is upregulated and a number of genes are predicted as targets including MAPK13 [145]. MAPK13 encodes p38 MAPK which plays an important role in bone homeostasis [146]. hsa-miR-138 is also downregulated during osteoblastogenesis as it can target FAK and inhibit the FAK-mediated signaling pathway [145]. Additionally, suppression of hsa-miR-138 expression with antagomiR-138 increased ectopic bone formation in vivo and overexpression of hsa-miR-138 reversed the effects, thus indicating that hsa-miR-138 impairs osteogenic differentiation by targeting FAK and its downstream signaling pathways. However, whether miR-138 regulates any FAK-mediated MAPK-signaling pathway during CHIKV infection is not yet known. Many reports have suggested that joint inflammation is associated with arthritic-like symptoms during CHIKV infection [147]. hsa-miR-146 has been associated with many viral and microbial infections and also with inflammatory conditions such as RA [49,148,149]. An upregulated expression of hsa-miR-146a was observed in synovial fibroblasts during CHIKV infection [24]. Furthermore, TRAF6 and IRAK1 were predicted as targets of hsa-miR-146a [25]. It is known that during viral infections, TRAF6 and IRAK1 activate the NF-кB signaling pathway to produce pro-inflammatory cytokines for combating infection [49]. Additionally, the NF-кB signaling pathway has been shown to play an important role during osteoclastogenesis. However, the effect of hsa-miR-146a on the NF-кB signaling pathway during CHIKV infection remains unknown.

Conclusions
Understanding the involvement of miRNAs during bone homeostasis in the context of CHIKV infection is of much interest as identification of novel biomarkers and/or development of miRNA-based therapeutics against viral infections is a promising area of research. Several miRNAs already serve as biomarkers and have been associated with pathologies, stages, and/or progression of different diseases. However, the use of miRNAs as biomarkers to diagnose viral diseases is still uncommon. Recently, hsa-miR-181c-5p and hsa-miR-1254 were identified as biomarkers for detection of H1N1 virus influenza [150]. In miRNA-based therapeutics, the developed miRNA either targets the pathogen or host factor during infection. During CHIKV pathogenesis, the two broad areas that can be targeted for drug development are (1) to directly impact virus replication or (2) to modulate host factors to mitigate arthritic-like symptoms caused due to infection. At present, a number of bioinformatic databases and high throughput screens are available to predict miRNA targets during preclinical therapeutic investigations. Additionally, a variety of in vitro cell culture models and in vivo mouse and non-human primate models are available to investigate the efficacy, toxicity, and safety of miRNA therapeutics. A phase 2 clinical trial with miravirsen (locked nucleic acid-modified DNA phosphorothioate antisense oligonucleotide that sequesters the mature hsa-miR-122 in a stable heteroduplex, thereby suppressing its function) in chronic hepatitis C virus (HCV)-infected patients showed reduced HCV RNA levels that persisted beyond the end of active therapy [151]. Another product, RG-101 (an N-acetyl-D-galactosamine-conjugated RNA antagomiR that targets hsa-miR-122 in HCV infected hepatocytes), was used in a clinical trial, which also resulted in undetectable HCV RNA in patients; however, it produced adverse effects due to which the trial was put on hold [152]. Thus, the transition of laboratory findings to clinical applications of miRNA-based diagnostics and therapeutics still remains a challenge and warrants further research. Funding: This work was supported by NIH/NIAID 1R21AI140026-01.