1. MSC Viral Infection and Host Responses
Mesenchymal stem cells (MSCs) as aforementioned have mostly critical roles during injury, participating in immunomodulation and tissue repair. In addition, their perivascular location in multiple organs makes them potential targets for viral infection, as summarized in Table 1.
Table 1. Examples of viruses targeting MSCs and related pathologies.
The following contents will address the different viral infections that affect MSCs and the resulting innate immune response given that MSC are immunocompetent gatekeepers known to express pattern recognition receptors (PRRs, e.g., RLR, TLR) and downstream signaling molecules (e.g., NF-κB, IRF3/7)
[64][65]. In parallel to macrophage polarization, two distinct subsets of MSC may exist in tissues (MSC1, proinflammatory and MSC2 anti-inflammatory)
[66][67]. In response to acute tissue injury and the release of alarmins derived from cell debris (apoptotic cells), MSC have a type 1 phenotype and help to recruit lymphocytes to sites of inflammation using MIP-1a and MIP-1b, RANTES/CCL5, CXCL9, and CXCL10 and to promote the clearance of cell debris and tissue repair. If the injury is too important and associated to cell necrosis and high levels of Tumor Necrosis Factor (TNF)-α and Interferon (IFN)-β produced by monocytes and T cells, respectively, MSC adopt an immune-suppressive phenotype (MSC2) by secreting high levels of soluble immunoregulatory factors, including kynurenine (Indoleamine 2 3-dioxygenase (IDO) pathway), PGE2 (COX2 pathway), NO, TGF-beta 1 (TGF-β1), IL10, Hepatocyte Growth Factor (HGF), and hemoxygenase (HO), that suppress adverse T cell proliferation and possible autoimmune response. TGF and IL10 will further control adaptive immune responses by mobilizing FoxP3+ T regulatory cells. Interestingly, viruses (e.g., those stimulating the TLR3 pathway by Polyinosinic:polycytidylic acid (Poly I:C)) may promote MSC2 anti-inflammatory and immunosuppressive phenotype to their own advantages by limiting the adaptive immune responses
[68].
2. MSCs a Gatekeper and/or Reservoir of Viruses
2.1. Bone Marrow-Derived MSC (BM-MSC)
BM-MSCs are among the first and best described subsets of MSCs
[69]. Genetic line tracing experiments using either Nestin or GLI-1 promoters indicated that perivascular BM-MSC could be derived from neural crest (NC) embryonic tissues, expressed neuroglial markers, as well as the beta3 adrenergic receptors arguing for a plausible role of the sympathetic nervous system to control MSC functions
[70][71][72]. Physiologically, BM-MSCs constitute a stromal cell niche for HSCs, supporting their stemness and education
[70][71][73]. Due to their relative ease of access, they represent a powerful source of cells as much for the study of the properties of MSCs as for the investigation of new therapeutic avenues. Indeed, BM-MSCs are now widely used as treatment given their beneficial immunoregulatory properties. They are used as carriers for the delivery of miRs or protein factors with therapeutic activities, physiologically expressed by MSCs or induced following an adenoviral or lentiviral infection
[74]. Additionally, BM-MSCs-derived exosomes have shown a similar therapeutic potential
[75].
However, due to their proximity with the HSC and hematopoiesis processes, their impairment, in particular through a viral infection, could be critical for BM-related diseases. Moreover, given bone marrow engraftment is the unique suitable treatment for certain diseases (e.g., hematological malignancies), the persistence of virus-infected cells in grafts could represent a major risk, especially in patients that are usually immunocompromised.
BM-MSCs have been shown to be targeted by a variety of viruses during the natural history of a clinical infection.
Human Immunodeficiency Virus (HIV), that causes Acquired Immuno Deficiency Syndrome (AIDS), was shown to infect BM-MSC and integrate its DNA in BM-MSC’s genome
[3][76]. Also, both intra- and extracellular interactions of HIV proteins Tat and Nef with BM-MSCs were observed
[3][4]. This leads to an impaired osteoblastic/proadipogenic differentiation therefore promoting respectively decreased bone marrow density and fat toxicity described in HIV-infected patients
[3][4]. Furthermore, HIV is able to induce senescence of BM-MSCs through its proteins Tat and Nef, but also p55-gag
[1][4]. BM-MSC senescence has been associated to HIV infection-related cytopenia
[1]. Finally, a role of BM-MSCs in HIV-related disease is the reactivation of HIV in latently-infected cells
[2].
Herpesviridae is a large family that includes diverse viruses causing human diseases, such as Human Cytomegalovirus (HCMV), or Human Herpesviruses (HHV). Among these viruses, HCMV, HHV-1, HHV-3 (or Varicella-Zoster Virus, VZV) and HHV-8 (Kaposi’s Sarcoma-Associated Herpesvirus) exhibited the ability to infect BM-MSCs
[5][7][8][77][78]. HCMV is a leading cause of congenital birth defects, as well as the major cause of diseases in immunocompromised individuals, notably following organ or BM transplant.
Of note, BM is an important site involved in the pathogenesis of chronic HCMV infection. The virus establishes latency in hematopoietic progenitors and can be transmitted after reactivation to neighboring cells. HCMV has deleterious effects on BM-MSCs function, changing the repertoire of cell surface markers expression (CD29, CD44, CD73, CD105, CD90, MHC class I and ICAM-1), and modifying the physiological interaction between BM-MSCs and HSC
[6]. In addition, similarly to HIV, HCMV alters BM-MSCs biology and might contribute to the development of diseases (impairment of osteoblast regeneration, cartilage regeneration, hematopoiesis and properties/functions of immune progenitor cells), due to a deterioration of BM-MSC differentiation capabilities
[6]. More strikingly, BM-MSCs were evoked as a potential reservoir for HCMV
[77], which could be crucial for treatments involving BM engraftment.
HHV-8 also induced alterations of BM-MSCs, infected cells displaying lower proliferation rates and altered expression of Kaposi’s Sarcoma markers as well as altered phenotypes related to malignant transformations
[7].
Even if BM-MSCs are physically less exposed to respiratory viruses, the influence of these viruses on BM-MSCs has been explored in several studies
[9][10][11]. Avian Influenza A (H5N1) virus productively infects BM-MSCs provoking cell death and IL-6, CCL2 and CCL4 secretion by co-cultured monocytes
[10]. The link between these findings and abnormal hematologic clinical descriptions such as lymphopenia, thrombocytopenia, and pancytopenia observed during avian flu remains to be addressed. Still, as for HCMV, the infection of BM-MSCs by Influenza virus might represent a risk of transmission during BM transplantation
[9]. BM-MSCs contribute to Respiratory Syncytial virus-related lung disease. Indeed, this virus is able to infect and replicate in BM-MSCs, altering their immunoregulatory functions via an increase of IFN-β and IDO expression
[11]. Zika virus (ZIKV) is a recently reemerging flavivirus, responsible for dengue-like syndrome in most of the cases but also associated with Guillain-Barré syndrome (GBS) in severe cases
[45]. ZIKV, which is more known for its implication in neurological induced disorders (as described in
Brain pericytes section), has the ability to infect and replicate in BM-MSCs
[12]. BM-MSCs infection by ZIKV causes increased IL-6 expression and impaired osteoblast differentiation, pointed out by a decreased expression of alkaline phosphatase (ALP) and Runt-related transcription factor 2 (RUNX2)
[12]. These data demonstrate the potential involvement of ZIKV in the development of bone pathologies. Finally, BM-MSCs might serve as an extrahepatic reservoir for Hepatitis B virus (HBV). Indeed, BM-MSCs are infected in vitro by HBV
[13]. Moreover, BM-MSCs are able to transport HBV to injured tissues, as evidenced by transplantation of BM-MSCs in a mouse model of myocardial infarction, resulting in HBV infection of injured heart and other damaged tissues
[13]. Thus BM-MSCs might play a critical role in HBV-associated myocarditis and other HBV-related extrahepatic diseases.
BM-MSCs are best known for their immunoregulatory activities during injury as aforementioned. They possess antiviral effector functions (
Figure 1), even if the innate immune response of BM-MSCs against each virus previously cited has not been examined in great depth. BM-MSCs are basically expressing several cytosolic PRR albeit at low levels
[79].
Figure 1. Infection of MSC by viruses may be controlled by a canonical innate immune response. Viruses may target perivascular MSC naturally expressing receptors (e.g., MXRA8/alphavirus) to grant entry. Among the up-regulated cytosolic pattern recognition receptors (PRRs), Retinoic acid-inducible gene-I (RIG-I)-like receptors (RLR: RIG-I, Melanoma-Differentiation-Associated Gene-5 or MDA5) and Toll-Like Receptor 3 (TLR3) are important to detect viral RNA. After sensing, MSCs engage different cell signaling pathways according to the stimulated PRR. A TLR3-dependent sensing activates mitogen-activated protein kinase pathways (through p38 MAPK and p46 JNK). RLR-dependent sensing stimulates IFN signaling pathway through TBK1/IKK-ε and subsequent interferon regulatory factor (IRF) 7 phosphorylation. These signaling pathways both trigger the production of pro-inflammatory cytokines and peptides with antiviral activities. Hence, in viral context MSCs express increased levels of IL-1β, IL-6, IL-8, IL-11, IL-12p35, IL-23p19, IL-27p28, TNF-α and CCL5/RANTES to recruit and activate adaptive immune cells (T/B lymphocytes). Furthermore, MSCs produce IFN-β and IFN-λ1). Classically, type I IFNs (such as IFN-β) induce the expression of Interferon Stimulated Genes (ISGs, e.g., RNASEL) by the interaction with the IFN receptor (IFNAR).
Yet in the context of a viral infection, it has been demonstrated (using Poly I:C to mimic RNA viruses) that BM-MSCs can up-regulate their PRR expression
[80]. Among the up-regulated cytosolic PRR, Retinoic acid-inducible gene-I (RIG-I)-like receptors (RLR: RIG-I, Melanoma-Differentiation-Associated Gene-5 or MDA5), and Toll-Like Receptor 3 (TLR3) are important to detect viral RNA
[80][81]. After sensing, BM-MSCs engage different cell signaling pathways according to the stimulated PRR. A TLR3-dependent sensing activate mitogen-activated protein kinase pathways (through p38 MAPK and p46 JNK)
[81]. Whereas RLR-dependent sensing stimulates IFN signaling pathway through TBK1/IKK-ε and subsequent interferon regulatory factor (IRF) 7 phosphorylation
[80]. These signaling pathways both trigger the production of pro-inflammatory cytokines and peptides with antiviral activities. Hence, in viral context BM-MSCs express increased levels of IL-1β, IL-6, IL-8, IL-11, IL-12p35, IL-23p19, IL-27p28, TNF-α, and CCL5/RANTES
[80][81][82]. Furthermore, BM-MSCs produce IFN-β and IFN-λ1 in a RIG-I-dependent manner in contrast to other MSC subtypes producing it in both TLR3 and RIG-I dependent manner
[80]. Classically, type I IFNs (such as IFN-β) induce the expression of Interferon Stimulated Genes (ISG) by the interaction with the IFN receptor (IFNAR). To date, the expression of ISGs has not been explored in BM-MSCs and needs further investigations. Interestingly, an antiviral activity of IDO has been demonstrated by decreasing HCMV and HSV-1 replication in BM-MSCs
[83].
Of note, exosomes derived from allogenic BM-MSC (e.g., EXOFLO
TM) have been used clinically to limit successfully the cytokine storm and the associated tissue injuries in patient with COVID-19
[84]. This would argue that MSC in tissues may be able to control the infection mediated by SARS-CoV-2.
2.2. Lung Resident-MSC and Viruses
To date, lung resident-MSCs (LR-MSCs) have been mostly studied in the context of fibrosis and bronchiolitis obliterans syndrome, notably after lung transplantation
[85][86]. Several important and recent gene tracing studies argued for a critical role of perivascular MSC, derived notably from NC progenitors, in lung diseases
[87][88]. However, LR-MSCs have been poorly studied in the virology field even if their pivotal position makes them potentially susceptible to viruses and particularly respiratory viruses. Supporting this, possible interactions between SARS-CoV-2 and MSC have been envisaged in COVID-19 positive patients experiencing pericytes loss and apoptosis (observed by cleaved-caspase 3 immunostaining)
[14]. Because pericytes often include MSCs, as discussed above, these findings suggest LR-MSCs as potential targets of SARS-CoV-2 in the lung
[15]. Currently available data regarding the expression of ACE2 and TMPRSS2 in MSCs are discordant
[89].
LR-MSCs have shown the highest permissiveness, replication rate and release during HCMV infection in cultured cells
[77]. Moreover, LR-MSC is now recognized as a natural reservoir for HCMV after its detection in seven out of nine individual donors
[77].
Lung PDGFβ+ MSC were recently shown as targets for HIV. These cells express both the primary receptor of HIV (CD4) and its major co-receptors (CXCR4 and CCR5), permitting productive infection and replication of HIV in vitro
[16]. This finding may indicate a possible implication of LR-MSCs in the HIV-related pulmonary complications
[16].
Surprisingly, in comparison to advances made for other MSCs subtypes, very few data are available to date regarding the specific antiviral response mounted by LR-MSCs during infection. Efforts should be made to fill this gap since it has already been demonstrated that a different antiviral response may be mobilized by tissue-specific MSCs
[80].
2.3. Adipose Stem Cells and Viruses
Adipose Stem Cells (ASCs) are a subtype of MSCs found in higher amounts than BM-MSCs and their isolation is easier, safer, less painful and less time consuming using liposuction techniques
[90][91]. Consequently, they became an alternative to BM-MSCs for cell-based or exosome-based therapy using MSCs
[92][93]. ASC can differentiate into different lineages and including into a Schwann-cell like phenotype expressing S100, glial fibrillary acidic protein (GFAP) with neurite outgrowth activity
[94]. They are now studied to be used in a broad range of diseases
[92]. MSCs are highly sensitive to their environment, influencing their responses through notably TLR or RLR
[80]. ASCs have important immunomodulatory functions shown by their high cytokine response
[80]. Notably, viral stimuli induce an important activation of these cells
[80]. Furthermore, the importance of immunoregulatory function of ASC is highlighted by their capacity to produce various adipokines (adiponectin, leptin) that influence the inflammatory responses in most tissues
[95]. Most of the studies regarding adipose tissue and viral infection have focused on adipocytes which are the main cell type of adipose tissue and also a derivative of ASCs. In the context of HIV, anti-viral treatments were shown to accumulate in adipocytes and alter their adipokine production, explaining some long-term complications of the therapies
[96]. Adipocytes are mostly resistant to direct viral infection by Hepatitis C Virus (HCV) or HBV
[97][98]. Due to their differentiation capabilities and their accessibility, ASC represents a good candidate for in vitro cell infection model
[97][98], research on factors influencing viral replication like miR-27a in HCV infection
[98] or stem-cell based therapies during HBV chronic infection
[97]. Nevertheless, further studies are needed to explore the direct role of ASC in the immune response in the course of viral infection and addressing the regulation in adipokine and miRNA production.
2.4. MSC of the Liver: Hepatic Stellate Cell/Ito Cell and Viruses
Hepatic Stellate Cells (HepSC, also named Ito cells) have been firstly described by Von Kupffer in the 19th century
[99][100], for review,
[101][102][103]. Using gene tracing experiments, they can be derived from myelin P zero (MPZ)+ NC cells and they express GFAP (astroglial marker) particularly in response to injuries (e.g.,
[104]). In physiological conditions, HepSCs are mostly long-lived quiescent cells residing in the perisinusoidal space of Disse
[99][105]. They represent 15% of resident liver cells
[106] and are fat-storing, vitamin-A rich, with presence of long processes
[99][103][107]. These cells interact with neighboring cells like hepatocytes, resident macrophages (Kupffer cells), endothelial cells or nerves
[99][106]. Also, they express markers of mesenchymal origin like desmin, alpha SMA
[105][106]. Under pathological conditions, they show proliferative activity and differentiate in myofibroblast-like cells
[99]. Activated HepSCs produce large amounts of extracellular matrix proteins (ECM) and matrix metalloproteases (MMPs)
[99][106][108][109].
Quiescent HepSCs participate in physiological conditions to liver homeostasis, regeneration, development, retinoid metabolism, extracellular matrix homeostasis, and drug metabolism
[110]. Upon liver injury, various stimuli, depending in the liver disease, activate HepSCs. The chronic activation of HepSC leads to excessive ECM accumulation and liver fibrosis
[21][110].
In the context of viral infection, liver fibrosis represents a critical complication of end-stage liver disease progression. It mainly occurs through the persistent activation of HepSCs during chronic infection. This process has been studied during either HCV, HBV, or HIV pathologies.
HepSC activation occurs through direct or indirect viral effects
[22]. HCV can directly activate HepSCs via either E2 protein binding to CD81
[23][110], NS3-NS5 proteins
[24][110] or dsRNA
[110][111]. HepSC expression of CCR5 and CXCR4 co-receptors for HIV has been described
[17]. Moreover, HepSC infection by HIV has been established in vitro but not in vivo
[17][18]. HIV through its envelope protein gp120 has been shown to induce HSC activation and chemotaxis
[17].
The indirect HepSC activation also represents an important mechanism. In fact, the hepatic inflammatory environment is an important factor of HepSC activation thus leading to pro-fibrogenic factor release
[110]. Hepatocytes produce profibrogenic factors like TGF-β1 or TIMP-1
[19][110], apoptotic bodies
[110] or ubiquitin carboxy-terminal hydrolase L1 (UCHL1)
[20] during HCV infection. Also, activated HepSC will release cytokine (i.e IL-1α) which will further stimulate hepatocytes to produce pro-inflammatory cytokines (IL-6, IL-8)
[110]. In the course of HBV infection, lymphocytes are implicated in inflammatory tissue damages
[25]. As shown for the hepatocytes, activated HepSCs may participate in the activation of immune cells thus amplifying the local inflammatory environment. For example during HBV infection, HepSCs could promote Th17 cell activation through IL-17RA-dependent proinflammatory cytokine expression
[26]. Similarly, during HIV chronic infection, many factors participate in HepSC activation and fibrogenesis like natural killer (NK) cells dysfunction, intestinal microbial translocation, Kupffer cell inflammatory response, hepatocytes apoptosis, and liver damages
[17].
The implication of HepSCs in the context of liver fibrosis during chronic viral diseases has been highly studied. However, their role during acute hepatotropic viral infection notably the importance of their immune response for viral clearance has been less explored.
2.5. MSC of the Kidney: Mesangial Cell and Viruses
Mesangial cells (MCs) is the name used to designate the resident perivascular MSCs of the kidney and chiefly involved in fibrosis
[88][112][113]. They are derived from the NC and responsible for mesangium matrix formation and glomerular capillaries support
[104]. As for other MSCs, their location at the periphery of the vessels allows them to act as gatekeepers and to sense their environment in case of injury such as viral infection
[114]. Similar to what is observed in other organs, viral infection in the kidney may lead either to a direct deleterious viral effect on targeted cells or to an indirect effect, due to an inappropriate immune response. MC participates in both mechanisms. Indeed, MCs were demonstrated to be targeted by various viruses
[27][29][34][115][116][117][118][119]. First, HIV was shown to infect MCs
[27] with an orphan G protein-coupled receptor 1 (GPR1) as a co-receptor
[28]. Additionally, HIV affects MCs either directly or indirectly leading to proliferation and matrix synthesis, thus MCs play a role in HIV-associated glomerulosclerosis
[120]. HCMV is the most threatening pathogen after a kidney transplantation, due to immunosuppression
[121] and potential virus rebound in immunocompromised host patient. HCMV is associated with the development of glomerulosclerosis due to the matrix deposition caused by MCs
[29][30]. Of note, MCs were shown to be targeted by HCMV and to allow its efficient replication
[31][32], even if the link between this infection and the pathology associated is poorly understood
[29]. Among the viruses of critical interest, HCV is a major problem worldwide, despite recent therapeutic improvements
[122]. The pathology associated with HCV is mostly liver disease, but with frequent extrahepatic complications, such as glomerulonephritis. HCV triggers TLR3 activation of MC leading to the release of procoagulant factors that causes vascular thrombosis and finally glomerulonephritis
[115]. Furthermore, MCs may be infected by ZIKV. It has been demonstrated that these cells may serve as a reservoir for the virus, thereby this finding may explain the high level persistent viruria observed
[34]. However, to date, no link has been established between ZIKV infection of MCs and kidney disease.
Of note, the antiviral response of MCs was not studied in the context of the viral infection aforementioned. However, viral RNA and DNA trigger a common antiviral response in MCs notably through RIG-I, MDA5 and TLR3
[123][124][125]. This response includes notably Interferon-type I response and secretion of proinflammatory cytokines and chemokines.
On the other hand, MCs also participate in inappropriate immune responses in case of infection. Indeed, immune complexes associated with viral RNA or DNA triggers overwhelmed immune response leading to glomerulonephritis. This mechanism was notably shown in a mouse model of Immunoglobulin A nephropathy (IgAN) induced by Sendai virus
[116], but also in patients suffering from Lupus nephritis
[126].
2.6. Brain MSC: Brain Pericytes and Viruses
Brain pericytes (BPs) are the resident MSCs of the brain
[127]. BPs are essentially derived from NC
[128][129]. With endothelial cells, and astrocytes, they constitute the blood-brain-barrier (BBB), the specialized vascular unit of the brain, responsible for its protection from external factors. Due to this major role in brain defense, an infection of BPs could lead to BBB disruption and increased permeability. BPs are well-equipped to ensure their sentinel function. They were shown to express PRR, notably TLR9, allowing them to be responsive in case of microbial infection, particularly by non-canonical inflammasome activation
[130]. Moreover, BPs are immunocompetent cells with NO, IL-1β, IL-3, IL-9, IL-10, IL-12, IL-13, TNF-α, IFN-γ, G-CSF, GM-CSF, Eotaxin, CCL3, and CCL4 secretion following LPS stimulation
[131]. Of note, BPs may also participate in the canonical antiviral response. Indeed, in response to IFN-γ and TNF-α, they showed upregulation of pro-IL-1β and pro-Caspase 1 mRNA expression
[130]. Moreover, BPs express CCL2 after Poly:IC or LPS stimulation, relaying inflammatory signals from circulation to neurons, leading to elevated excitability
[132].
Additionally, due to their location and the different route of entry for foreign agents (notably receptor-mediated endocytosis, unspecific transport by pinocytotic vesicles), BPs represent possible targets for viral infection. HIV infection can cause BBB disruption and contributes to the development of neurological dysfunction (e.g., HIV-associated neurological disorders—HAND). HIV is able to infect and replicate at low levels in BPs, using its co-receptors CXCR4 and CCR5
[37][133]. These cells even constitute one of HIV reservoirs in the central nervous system (CNS), the infection reactivation being possible thanks to genome integration mechanism. An increase in integrated genome in BPs is leading to a latent stage of infection following the initial peak of HIV production in the CNS
[134][135]. Moreover, HIV infection of BPs has deleterious effects on BBB. Of note, it was shown that a decreased BPs coverage of BBB was a consequence of BPs dysfunction in HIV-infected patients
[38], BPs coverage being negatively correlated with BBB permeability. The loss of pericytes observed has been associated with early PDGF-BB expression, which promotes pericytes migration away from their perivascular location
[36]. Moreover, HIV infection leads to IL-6 secretion
[37]. This may be another consequence of PDGF-BB secretion and downstream receptor signaling events
[136]. IL-6 secretion by BPs also concurs in HIV-induced CNS damage and BBB disruption observed
[37], the proinflammatory-induced response of endothelial cells causing BBB impairment
[137].
HCMV infection may cause neurological pathologies either in children or immunocompromised patients. If, astrocytes and endothelial cells were initially reported as the main targets of HCMV at the BBB level
[39], BPs were recently shown to be more permissive and sensitive to HCMV-induced lysis
[40]. Thus, BPs are contributing to HCMV dissemination in CNS
[40][77], as well as neuroinflammation via CXCL8, CXCL11, CCL5, TNF-α, IL-1β, and IL-6 secretion
[40].
Japanese Encephalitis Virus (JEV) is a neurotropic mosquito-borne flavivirus that causes encephalitis, with neurological and psychological sequelae among a majority of survivors. Neuroinflammation is the most outstanding mechanism associated with JEV pathogenesis. BBB integrity is critical to regulate neuroinflammation, by limiting circulating immune cells entry and avoiding overwhelmed immune response. BPs are targeted by JEV as least in vitro. Following infection, they secrete IL-6 promoting proinflammatory responses and proteasomal degradation of Zonula Occludens-1 (ZO-1), thus participating in JEV-induced BBB impairment
[41]. In addition, JEV was shown to trigger NF-κB through TLR7/MyD88 dependent axis, leading to PGE2 and CCL5/RANTES secretion
[42]. The latter causes attraction and infiltration of leukocytes to the site of infection, provoking the aforementioned inappropriate immune response responsible for JEV neuropathology. The same mechanism is evoked for Herpes Simplex encephalitis but need to be further explored
[43]. ZIKV, which is known for its neurological complications (e.g., microcephaly, Guillain-Barré Syndrome, and encephalitis)
[45], can invade CNS through BPs infection at the level of the choroid plexus (via the receptor AXL)
[44]. Both murine and human BPs are susceptible to ZIKV
[44]. Furthermore, ZIKV-infected BPs are associated with BBB defect in vitro, illustrated by increased cytosolic ZO-1, decreased transepithelial electrical resistance and higher degree of FITC-dextran transport
[44].
3. Osteoblasts and Viral Infections
Osteoblasts (OBs) are stromal cells, responsible for bone formation, through osteocalcin, ALP and type I collagen expression. As the other cells presented in this section, OBs (e.g., craniofacial bones) are derived from MSCs of the NC
[138]. A great deal of pathogens are able to target bone tissues, leading to a variety of diseases ranging from caries to osteomyelitis. Among these pathogens, bacteria are the most cited. Yet the implication of viruses, and more specifically arthritogenic ones, in development of bone disorders should not be underestimated.
Ross River Virus (RRV) is a mosquito-borne alphavirus that generally causes flu-like illness and polyarthritis. As other alphaviruses, RRV might promote the development of bone diseases by targeting osteoblasts through Mxra8
[139]. Chen et al. addressed the effect of RRV infection on osteoblasts and showed that RRV-infected osteoblasts are producing high levels of IL-6 and CCL2
[48]. Additionally, this infection led to an imbalance in the Receptor Activator of Nuclear factor-KappaB Ligand (RANKL)/Osteoprotegerin (OPG) ratio in favor of osteoclastogenic activities and bone loss
[48]. Moreover, the same team demonstrated an increased susceptibility to RRV in osteoblasts from osteoarthritic patients, further promoting the adverse effects of infection previously mentioned, due to a delayed IFN-β induction and RIG-I expression
[47].
Similar outcomes are observed following OBs infection by Chikungunya virus (CHIKV), which is another arthritogenic alphavirus, best known for its recent epidemics between 2005 and 2006 in Reunion Island. Thus, an increased RANKL/OPG ratio in favor of osteoclastogenesis
[49][140] can lead to bone loss through monocytic osteoclast recruitment
[140].
HCV, besides its hepatic-related disease, might be able to trigger bone disorders. Indeed, HCV-infected patients have a higher risk of osteoporosis
[141] and fracture
[52]. Conversely, HCV infection was also associated with bone density hardening and osteosclerosis
[50]. Once again, RANKL/OPG ratio impairment during HCV infection could be the cause of these detrimental consequences
[50]. Kluger and colleagues addressed the permissiveness of OBs to HCV and provided evidence that OBs and osteoblast progenitors harbor HCV replication in the bone
[51]. Thus, the implication of OBs in the pathophysiology of bone diseases following HCV infection should be explored given their involvement during alphaviruses-related disorders. On the other hand, as previously evoked, flavivirus ZIKV infection of BM-MSCs interferes with their ability to differentiate in osteoblasts, with a significant increase of IL-6 expression and a decrease of key osteoblast marker (e.g., ALP and RUNX2) in infected MSCs
[12]. This susceptibility leads to impaired osteoblast function during ZIKV infection, triggering an imbalance in bone homeostasis and inducing bone-related disorders
[12].
Measles morbillivirus (MeV), which belongs to the
Paramyxoviridae family, is the cause of measles, a highly contagious disease. Symptoms of measles associate general symptoms (fever, cough) with a generalized maculopapular and erythematous rash. Moreover, MeV has been evoked in bone-related diseases such as otosclerosis
[53][54] and Paget’s disease
[55], both characterized by disturbed equilibrium of bone resorption and new bone formation. Several teams reported that productive OBs infection by MeV participates in the development of these disorders. Hence during MeV infection, OBs exhibit a higher expression of several osteogenic markers: bone morphogenic proteins (BMP-1, -4, -5, -6, and -7), ALP, bone sialo-protein, collagen 1α1 and OPG
[53]. These findings highlight the ability of MeV to stimulate osteogenic differentiation of OBs thus participating in imbalance in bone formation. Such findings were corroborated by Potocka-Bakłażec et al., by demonstrating increased levels of TNF-α, IL-1β and decreased levels of OPG (conversely to Ayala-Peña et al. study) in MeV positive patients, also testifying bone remodeling subsequent to MeV infection
[54].
HIV-infected patients have a higher incidence of osteopenia and osteoporosis due to bone demineralization and reduced bone mass. This is the result of OBs increased apoptosis following HIV infection in response both to expression of TNF-α and impaired Wnt/β-Catenin signaling in infected cells
[46][142].
Osteoclasts, derived from macrophages, were thought to be the principal sensor of infection in the bone, due to their wide expression of PRRs. Notwithstanding, a growing body of evidence is in favor of PRRs’ expression by OBs, arguing for a sensing and antimicrobial role shared between osteoclasts and OBs
[143]. Even if exacerbated inflammatory response in case of infection may lead to detrimental consequences, this has notably been exemplified earlier with IL-6 and development of arthritic pathologies, antiviral response remains essential to counteract most of virus-induced pathogenesis processes. Thus, the response of OBs in viral context has been addressed by Nakamura et al. on mouse osteoblastic cells (MC3T3-E1), using Poly:IC. They showed that in context of viral infection, OBs produce IFN-β as early as 1 h after stimulation due to Poly:IC recognition by TLR3, with a peak at 12 h
[144]. Of note, they reported production of TLR3 and RIG-I in response to IFN-β, making OBs fully capable of viral dsRNA detection
[144][145]. Of further note, IFN-β triggers CXCL10 production through a IFN-α/β receptor-STAT1 pathway
[144]. Taken together these data suggest the ability of OBs to mount a proper IFN-type I response. Additionally, IL-27, a cytokine regulating immune responses as well as hematopoiesis and bone remodeling, was found to be expressed by OBs in inflammatory conditions (e.g., presence of type I IFN, IL-1β and TNF-α)
[146]. So, OBs are thought to be part of a negative-feedback mechanism, limiting bone erosion and dampening T cell-mediated immune pathology during bone inflammation (therefore antiviral response).
4. Schwann Cell of Peripheral Nerves and Viral Infection
Schwann Cells (SCs) originate from NC. Moreover, MSCs are able to give rise to Schwann Cell-like cells after induction
[147][148], these cells being able to drive a proper myelination. SCs are support cells that have a pivotal role in myelination of neurons from the peripheral nervous system (PNS). Thus, their affection is linked to demyelinating disorders of PNS (e.g., Guillain-Barré Syndrome).
As previously seen, HIV can cause neurological disorders, called HIV-associated neurological disorders (HAND). Among these HAND, distal neuropathy could be notably cited. Even if SCs are not primarily infected by HIV (infection has not been described in vitro), they express chemokine receptor CXCR4, a receptor for HIV-1 gp120
[56]. The pathway driven by HIV-1 gp120 leads to RANTES and TNF-α secretion, stimulating axon and neuron to release TNF receptor-1 (TNFR1)
[56]. This promotes dorsal root ganglion neurotoxicity, including axon and myelin injury
[56].
CNS tropism of Herpes Simplex Virus (HSV) and Herpes Zoster Virus/Varicella Zoster Virus (VZV) is well-established. Furthermore, infection of SCs by HSV or VZV was experimentally demonstrated
[149][150]. However, it is difficult to say if this infection is of clinical relevance, since the principal mechanism evoked for HSV-induced GBS is a molecular mimicry of viral proteins, leading to cross-reacting antibodies
[57]. However, this statement is based on case reports and needs to be further explored experimentally, in order to exclude any direct involvement of SCs in HSV-induced GBS.
Similarly, HCMV inclusions can be observed in SCs
[58]. HCMV infection is the second most frequent infectious etiology of GBS
[59], with notable cases in immunocompromised patients
[60]. The pathogenesis of CMV-associated GBS have been linked to a probable molecular mimicry generating autoantibodies against, among others, moesin expressed by SCs
[61]. Of note, these findings are called into question and need to be experimentally confirmed with an animal model, since no other team has obtained similar results
[62].
As aforementioned, ZIKV is increasingly implicated in neurological disorders affecting CNS (i.e., microcephaly) as PNS (i.e., GBS) and of critical concern in the last years
[45]. Initially, it has been demonstrated that CNS cells and oligodendrocytes, responsible for myelination in CNS, were more susceptible to ZIKV than PNS cells, causing axon and myelin injuries
[63]. Yet more recent data indicate SCs are susceptible to ZIKV
[151][152]. Volpi et al. showed that, in myelinating dorsal root ganglion explants from
Ifnar1−/− mice, ZIKV infection of SCs leads to endoplasmic reticulum stress pathway activation and apoptosis in these cells, which finally cause demyelination and axon degeneration
[151]. Additionally, Dhiman et al. showed a sustained viral production and a significant cell death at 96 h post-infection in vitro in human SCs, the infection inducing expression of proinflammatory cytokines (IL-6, TNF-α, IFN-β and IL-29)
[152]. These works pave the way for a direct viral pathogenic effect or a cell-mediated inflammation in pathogenesis of ZIKV-associated GBS, even if a possible antibody dependent enhancement with Dengue Virus (DENV) sera or a demyelination induced cross-reactivity of anti-ZIKV antibodies are also discussed elsewhere
[153][154].
Primary antiviral response mounted by SCs has not been well studied for each virus previously presented. However, SCs possess an efficient immune system to detect pathogens, represented by several PRRs. Among them, TLR3 and TLR7 are PRR devoted to virus detection and expressed by SCs
[155]. TLR3 and TLR7 trigger a classical antiviral response after PAMP recognition. This response implies IFN response, driven by NF-κB and IRF, ISG release, and ultimately apoptosis, subsequently to inflammatory cytokine stimulation of extrinsic pathway.
This entry is adapted from the peer-reviewed paper 10.3390/ijms23148038