Viruses are small infectious agents that cause disease in all forms of life (Koonin et al., 2006). Based on their genomic material, they are classified as double-stranded (ds) DNA viruses, single-stranded (ss) DNA viruses, dsRNA viruses, (+)ssRNA viruses, (−)ssRNA viruses, ssRNA- reverse transcriptase (RT) viruses, and dsDNA-RT viruses (Baltimore, 1971).

To protect themselves from infection by viruses, hosts evolved immune systems (Janeway et al., 2004) consisting of many barriers and biological processes (Delves and Roitt, 2000). The immune system can be subdivided into innate immunity and adaptive immunity. The innate immunity is trigged when pathogens are identified by their pathogen associated molecular patterns (PAMPs) or when cells signal in response to damage, injury, or stress (Takeuchi and Akira, 2009). Although the innate system is non-specific and does not confer long-lasting protection (Mackay et al., 2000) it is the major defense mechanism against pathogens in most organisms (Litman et al., 2005). It consists of physical and chemical barriers including phagocytes and dendritic cells (DC), inflammation, the complement system, and natural killer cells. In contrast, the adaptive immunity relies on antigens and is highly specific to pathogens or pathogen-infected cells (Dörner and Radbruch, 2007). Lymphocytes are key role players in adaptive immunity and include both T cells and B cells (Janeway et al., 2004). Of the various components of the immune system the following are critical in clearance of viral infections; natural killer cells, interferons, dendritic cells, B cells, and T cells (Aoshi et al., 2011).

The immune system can be modulated by various factors including prostaglandins (PGs) (Harris et al., 2002). Prostaglandins are lipid molecules, derived from arachidonic acid (AA) and are produced by cyclooxygenase (COX), and PG synthases (Phipps et al., 1991). One of the most studied PGs is prostaglandin E₂ (PGE₂) which is produced by many cells including fibroblasts, macrophages and some malignant cells (Harris et al., 2002). PGE₂ regulates various processes in the body via PGE₂ receptors (EP1–EP4) (Sugimoto et al., 2000). Both the innate and adaptive immunity can also be regulated by the levels of PGE₂, which can either have adverse or beneficial effects on the immune system's ability to fend off pathogens (Kalinski, 2012). This review focuses on the regulatory role of PGE₂ on the immune system in the course of some well- viral infections caused by herpes simplex virus, Epstein-Barr virus, rotavirus, influenza A viruses, human immunodeficiency virus and hepatitis B virus, and the development of related potential therapies for the treatment of these infections.

Prostaglandins are eicosanoids that are produced by nearly all mammalian cells (Park et al., 2006). They are not stored within cells but rather produced in response to specific trauma, signaling molecules or stimuli such as infections (Smith, 1989; Funk, 2001). PGE₂ is the most abundant prostanoid (Serhan and Levy, 2003) in the mammalian body and under normal physiological conditions and plays a role in regulation of immune responses, blood pressure, gastrointestinal integrity, fertility (Ricciotti and FitzGerald, 2011), and inflammation (Davies et al., 1984).

The innate immune system is critical for pathogen recognition (Mackay et al., 2000) and functions as the first level of defense. In general the immune system relies upon leukocytes, antibodies, the complement system and cytokines to remove pathogens or toxins from the host (Janeway et al., 2004). Leukocytes are subdivided into different types, namely neutrophils, eosinophils, basophils, lymphocytes and monocytes which are distinguished by their physical and functional characteristics. Of particular importance to this review are the lymphocytes, which can be divided into B cells, T cells and natural killer cells, as well as monocytes which differentiate into macrophages in resident tissues. Antibodies are produced by plasma cells (differentiated B-cells) in response to specific antigens and bind to antigens. They bind to antigens, before the complexes are phagocytized. The complement system enhances the ability of antibodies and phagocytic cells to remove pathogens and damaged cells from an organism and promotes inflammation (Figure 3I). Although the complement system is considered to be part of the innate immune systems, it can be brought into action by the adaptive immune system. The complement system is known to consist of three components. These include, the classical pathway (relies on the activation of C1-complex by antibodies), the alternative pathway (relies on direct interaction between the pathogen and C3b) and the lectin pathway (relies on binding of certain sugars to mannose-binding lectin). Lastly, cytokines are a group of small molecules that are important in cell signaling and some as immunomodulatory agents. Cytokines play an important role in the balance between humoral and cell-based immunity.

Immune defenses against viruses are initiated when viral infection occurs or when antibodies from a previous infection recognize the virus and neutralize it (Mackay et al., 2000) (Figure 3A). Natural killer (NK) cells (Brandstadter and Yang, 2011) and interferons (IFN) are the most effective mechanism of the innate immunity against viral infections (Figures 3B,C). Natural killers cells are particularly important for the defense against the herpesvirus, poxvirus, papillomavirus, cytomegalovirus, influenza A virus, and human immunodeficiency virus (HIV) (Biron et al., 1989; Orange, 2002; Jost and Altfeld, 2013). These cells function mainly by killing infected cells by releasing cytotoxins early in infection, preventing spread to neighboring cells (Vivier et al., 2008). They also play a critical immunoregulatory role during the development of adaptive immunity. Thus, NK cell-mediated immunity plays a role in the control and clearance of viruses in the early stages of infection, but the continual stimulation of NK cells can contribute to viral pathology (Jost and Altfeld, 2013). However, there is still limited knowledge about the mechanisms of NK recognition and antiviral function. Another important response of the innate immunity toward viral infection is IFN (Samuel, 2001). Interferons are produced when a virus infects a cell (Sonnenfeld and Merigan, 1979). They are secreted as cell surface receptors causing the transcription of multitudes of IFN-stimulated genes which encode for protein products with antiviral, antimicrobial, antitumor and immunomodulatory effects (Fensterl and Sen, 2009). Type I (IFN-α, -β, -ω, -ϵ, -κ, -τ, -δ, -ν, and -ζ) and type III (IFN-λ) IFNs are induced when any cell recognizes PAMPs whereas type II (IFN-γ) is induced by other cytokines like IL-12 and expression is restricted to T cells and NK cells (Samuel, 2001; Fensterl and Sen, 2009). Viruses have also acquired various mechanisms to circumvent these actions of the innate immunity (Cerwenka and Lanier, 2001; Katze et al., 2002).

Another critical role player in innate immunity is a class of proteins called Toll-like receptors (TLRs) (Figure 3H; Xagorari and Chlichlia, 2008). Toll-like receptors are single, membrane spanning, non-catalytic receptors that are mainly expressed by macrophages and DCs. They are responsible for recognizing and responding to the PAMPs, leading to the activation of intracellular signaling pathways and altered gene expression. In turn, this allows the host immune system to detect pathogens and respond to their stimuli (Kawai and Akira, 2005). The signaling pathways that are activated by TLRs are responsible for the production of type I IFN, inflammatory cytokines and chemokines as well as the induction of immune responses responsible for eliminating pathogens (Kawai and Akira, 2006). Toll-like receptors signal via two distinct pathways, the MyD888-dependent and the TRIF-dependent pathways (Kawai and Akira, 2010). The MyD888-dependent pathway's primary effect is to activate nuclear factor kappa B (NFκB) and mitogen-activated proteins kinases, while activation of the TRIF-dependent pathway leads to the production of type I IFN and the transcription of NFκB. Sensing via TLRs is also responsible for the induction of DC maturation, which in turn initiates adaptive immune responses (Pasare and Medzhitov, 2004).

Dendritic cells and their system of antigen-presenting cells (APC) (Figures 3D,E) are another important component of immunity as they bridge the innate and adaptive immunity (Banchereau and Steinman, 1998; Chan et al., 2006). The main function of DCs are to activate naïve T cells by presenting processed antigen material on their surface (Banchereau and Steinman, 1998). However, before DCs can present the antigenic material, they have to complete a maturation cycle which is induced by either direct contact with pathogens or by interactions with other innate immune cells. Invading pathogens are sensed via extensive PAMP receptors which control the secretion of cytokines, migration, proliferation, and expression of major histocompatibility complex (MHC) II and co-stimulatory molecules in DCs (Steinman, 2003). The activation of DCs can be enhanced by activated NK cells (Chan et al., 2006).

The adaptive immunity's B and T cells are stimulated by DCs (Hess et al., 2016) (Figure 3G). B cells, which are the precursors to antibody secreting cells, directly recognize antigens through their B-cell receptors, while T cells are dependent on APCs for antigen recognition (Banchereau and Steinman, 1998). In viral infections, B cells (part of humoral immunity) begin their maturation process when they come in contact with viral antigens within the lymphatic system (Dörner and Radbruch, 2007). Naïve B cells have membrane-bound antibodies that effectively bind the viral antigen. Immediately following its contact with a specific viral antigen the B cell divides to become B memory cells and/or plasma cells. B memory cells express the same membrane-bound antibody as original naïve B cells and this is essential for a faster immune response to the particular viral antigen in future infections. Plasma cells also produce the same antibody as the original B cell, but are secreted into the bloodstream and neutralize the viral pathogen. T cells (part of cell-mediated immunity) express T cell receptors (TCR) and either CD4⁺ or CD8⁺ receptors (Sant and McMichael, 2012). As stated earlier, T cells cannot recognize antigens without receptor molecules (either MHC I or MHC II) which are membrane-bound on APCs. T cells mature into T helper cells (T_H), cytotoxic T cell (T_c), and T regulatory cells (Tregs). T helper cells express CD4⁺ receptors and are responsible for activation of T_c, B cells and other immune cells (Alberts et al., 2002) (Figure 3F). As mentioned earlier, T_H function by producing either Th1 or Th2 responses (Figure 3F). T helper cells 1 are produced when DCs secrete IL-2 and IFN-γ cytokines. These Th1 cells then secrete their own cytokines (IFN-γ and TNF-β) which stimulate recruitment of other lymphocytes to inflammation areas, induce B cell antibody switching and activate T_c (Romagnani, 1992). The Th1 responses are thus essential for the removal of intracellular pathogens. T helper cells 2 are produced when APC present antigens to TCR along with costimulatory molecule B7, IL-4, and IL-2. T helper cells 2 then secrete their own cytokines (IL-4, IL-5, and IL-13) which promotes IgE production, blocks IFN-γ receptors, recruits and activates basophil and eosinophil leukocytes (Romagnani, 1992). The Th2 response is thus responsible for the control of extracellular pathogens. Cytotoxic T cells remove pathogens and infected host cells and express CD8⁺ receptors, while Tregs express CD4⁺ and CD25⁺ and help in distinguishing of self from non-self-molecules (Alberts et al., 2002).

All the classic signs of inflammation (swelling, redness, heat, and pain) can be attributed to PGE₂ (Funk, 2001). PGE₂ causes redness and edema (of skin) by augmenting arterial dilation and microvascular permeability, increasing the blood flow into inflamed tissues. The pain caused by PGE₂ results from the action on peripheral sensory neurons and on central sites within the spinal cord and brain. In addition, the binding of PGE₂ to one of its various receptors can regulate the functions of macrophages, dendritic cells (DCs) and T and B lymphocytes (Ricciotti and FitzGerald, 2011). These regulating functions of cell types can in turn lead to both pro- and anti-inflammatory effects.

PGE₂ also plays a role in the regulation of cytokine expression in DCs and has shown bias in T cell differentiation toward either Th1 or Th2 responses (Kirkpatrick, 2005) (Figure 4). Disruption in the early stages of DC differentiation is noted as one of the effects of PGE₂ on DCs (Kalinski et al., 1997). In addition to the suppressive function of PGE₂ on differentiation of functionally competent Th1-inducing DCs, they also suppress responses of T_c (Obermajer et al., 2011). The induction of the DC migratory phenotype permitting their homing to drain lymph nodes, is enhanced by PGE₂ (Kabashima et al., 2003; Legler et al., 2006). During early maturation, PGE₂ can stimulate DCs to express co-stimulatory molecules which enhance T-cell activation (Krause et al., 2009). PGE₂ can also enhance DC production of suppressive factors, but the net effect on DCs is to enhance promotion of T cell expansion (Kalinski, 2012). Dendritic cells that have matured in the presence of PGE₂ have an impaired ability to induce Th1 while enhancing Th2 responses. PGE₂ also has suppressive effects on naïve T cell activation and expansion as well as direct inhibitory effects on interleukin 12 (IL-12) production and the expression of IL-12 receptors (Kalinski, 2012). Furthermore, PGE₂ also balances the Th cell responses by inhibiting interferon (IFN)-γ, a Th1 response. It does, however, not inhibit IL-4 and IL-5, Th2 responses, in CD4⁺ T cells (Snijdewint et al., 1993) (Figure 4). PGE₂ is responsible for the suppression of IL-2 production and IL-2 responsiveness in T cells, leading to the suppression of T cell activation and proliferation at high doses (Kalinski, 2012). At lower doses PGE₂ already shows a great modulatory effect on the shifting patterns of CD4⁺ T cell responses form the aggressive Th1 cells toward Th2 and Th17 cells that cause less tissue destruction. The Th1 suppressive effect of PGE₂ also relies on the suppression of IL-12 in macrophages and DCs (van der Pouw Kraan, 1995; Kalinski, 2012). Thus, PGE₂ shifts the immune response from Th1 to Th2, which leads to a reduced protective ability against intracellular pathogens (viruses and bacteria). In addition to the direct effect of PGE₂ on Th1 immune cells, recent studies have showed the indirect effect of PGE₂ in enhancing the development and activity of suppressive types of immune cells (Kalinski, 2012). PGE₂ has been shown to promote the development of Tregs in both mice and humans (Baratelli et al., 2005). The EP2- and EP4-dependent induction of Tregs in murine cancer (Sharma et al., 2005) and skin UV irradiation (Soontrapa et al., 2011), have been shown to rely on COX-2 and PGE₂. The Tregs have been shown to have with an role in human tumor tissues (Bergmann et al., 2007), The interaction between DCs and Tregs are also promoted by PGE₂, suggesting a role in the promotion of the expansion of pre-existing Tregs (Muthuswamy et al., 2008). It has also been shown the PGE₂ is involved in mediating the suppressive effect of Tregs (Mahic et al., 2006).

T helper cells 17 are important in the maintenance of mucosal barriers and the subsequent clearance of pathogens from these areas and have been implicated in autoimmune disorders and infectious diseases (Zambrano-Zaragoza et al., 2014). Cytokines such as IL-6, TGFβ, IL-21, and IL-23 are involved in the development of Th17 cells (Holzer et al., 2013). In addition, it has been shown that PGE₂ plays a role in the differentiation and function of Th17 cells via cAMP and EP2/EP4 signaling (Boniface et al., 2009). Napolitani et al. (2009) concluded that PGE₂ may dramatically influence the balance between the highly inflammatory IL-17 and the negative feedback loop exerted by IFN-γ. PGE₂ accomplishes this by triggering the expression of EP2 and EP4 receptors on T cells, leading to a rapid increase in retinoic-acid-related orphan receptor-γt (controls T-cell precursors) and a decrease of T-cell specific T-box transcription factor 21 mRNA (encodes for a Th1 cell-specific transcription factor). The role Th17 plays in viral infection will be discussed in detail in a later section. Although no direct link has been found between PGE₂ and autoimmune diseases, the role of Th17 cells have been reviewed in Zambrano-Zaragoza et al. (2014), implicating causative roles in Glioma, Multiple Sclerosis, Type 1 Diabetes and others. Since PGE₂ can play a regulatory role in Th17 cell differentiation and functioning, it may be speculated to have some role in autoimmune diseases.

Toll like receptor 4 is well-known for its ability to recognize LPS and other ligands, including viral proteins, polysaccharides, and endogenous proteins (Brubaker et al., 2015). PGE₂ has strong suppressive effects on LPS-induced IFN-β production at mRNA and protein levels in murine J774A.1 macrophages (Xu et al., 2008). This inhibitory effect is mediated through EP2 and EP4 which in turn activates the Epac/PI3K/Akt pathway. In contrast, PGE₂ inhibits LPS-induced TNF-α by a PKA pathway independent of the Epac/PI3K/Akt pathway. Importantly, it was found that blocking COX activity in vivo resulted in higher concentrations of IFN-β post-LPS. Degraaf et al. (2014) showed that PGE₂ can decrease the expression of TLR4. The authors showed that the effect was not by regulation of TLR4 mRNA but rather by the inhibition of its translation. They concluded that this reduction was mediated by EP2-dependent cAMP activated by PKA. The reduction in TLR4 expression was enough to decrease the transcription and section of TNF-α alveolar macrophages in responses to LPS. This thus shows that lipid mediators can modulate TLR4-mediated immune responses.

PGE₂ can also interfere with early B cell activation and play a cAMP-mediated role in the regulation of immunoglobulin (Ig) class switching in B cells (Simkin et al., 1987; Phipps et al., 1991). Antibody class switching is a process where B cells change the production of antibodies from one type to another by a mechanism called class switch recombination (reviewed in Stavnezer and Amemiya, 2004). One of the most important consequences of these effects is the promotion of IgE production by PGE₂ contributing to atopic disease (Carini et al., 1981).

Furthermore, PGE₂ can also exert anti-inflammatory effects on neutrophils, macrophages, mast cells and NK cells which all form part of the innate immune system (Kalinski, 2012) (Figure 4). In NK cells, PGE₂ functions by suppressing the responsiveness of IL-12 (Walker and Rotondo, 2004) and IL-15 (Joshi et al., 2001) which suppresses the cytolytic effects of NK cells (Bankhurst, 1982). PGE₂ abrogates NK cell “helper” function by inhibiting the ability of NK cells to produce IFN-γ (Mailliard et al., 2005). The function of macrophages are modulated by PGE₂ in an EP2-dependent (Aronoff et al., 2004) and phosphatase and tensin homolog-dependent manner (Canetti et al., 2007) which limits phagocytosis by alveolar macrophages and their pathogen-killing function (Serezani et al., 2007). This inhibition is in part mediated by induction of IL-1R–associated kinase-M, which blocks the scavenger receptor-mediated phagocytosis and the TLR-dependent activation of tumor necrosis factor-α (Hubbard et al., 2010). Furthermore, it has been shown that PGE₂ in combination with LPS can promote anti-inflammatory phenotypes in macrophages by high expression of IL-10 and regulatory markers, SPHK1 and LIGHT, via a protein kinase A-dependent pathway (MacKenzie et al., 2013). The local attraction and degranulation of mast cells are induced by PGE₂ in a mechanism involving EP1 and EP2 (Hu et al., 1995; Gomi et al., 2000).

Various steps of inflammation can thus be modulated by PGE₂ in either a pro-inflammatory or anti-inflammatory manner. PGE₂ can also modulate the immune system by shifting Th responses and thus affect the interplay between innate and adaptive immunity.

As reviewed above, PGE₂ has varying effects on the immune system. In some instances viruses can interact with PGE₂ and possibly benefit from the effects of PGE₂ (Table 1). A few of the potential effects of PGE₂ on various viral infections are described:

The current review highlights the potential of the biosynthetic pathway of PGE₂ (Figure 1) as a therapeutic target in viral infections. This is possible as PGE₂ has been shown to play a role in various viral infection, by either having a stimulatory/inhibitory effect on the viral life cycle or host's immune system.

One of the first potential therapies is limiting the amount of AA (or FA that can be converted to AA) that is taken in up in the diet of an individual (Calder, 2005, 2010). This can be done by limiting the amount of n-6 polyunsaturated fatty acids (PUFAs) (vegetable oils, animal sources) and rather ingesting n-3 PUFAs (fish oils, marine sources) which are preferentially converted into either docosahexaenoic acid (DHA) or eicosapentaenoic acid (EPA). They act by replacing AA as an eicosanoid substrate and inhibiting AA metabolism (directly) or altering the effects of inflammatory genes through effects on transcriptional activation (indirectly) (Calder, 2005). Although EPA also gives rise to eicosanoids, these have anti-inflammatory effects in contrast to AA-derived eicosanoids (Calder, 2006).

Both of the COX isoforms can also serve as a potential target as they are directly responsible for the downstream production of PGE₂. Of the two isozymes, COX-2 is a more attractive target as COX-1 is constitutively expressed and particularly important in gastrointestinal protection (Hawkey, 2001). This means that COX-2 selective drugs are required to regulate the production of PGE₂. These drugs include coxibs (celecoxib and rofecoxib), etodalac, meloxicam, and nimesulide. Coxibs function by binding to side pockets close to the COX-2 active site, while etodalac, meloxicam and nimesulide bind in the active site (Jackson and Hawkey, 2000). All of these selective inhibitors have some side effects and can only partially inhibit PGE₂ production.

The next possible therapeutic targets include the PGES (Murakami and Kudo, 2006). Microsomal prostaglandin E synthase-1 is the most promising target as the deletion of mPGES-1 in animal and cells models leads to no severe adverse effects (Chen Y. et al., 2015). There are no known mPGES-1 inhibitors due to the fact that compounds are not stable or effective in cellular assays or animal models. The search for an appropriate inhibitor is further complicated by specific differences in rodent and human models as they differ in key amino acid residues of mpGES-1 active site. Inhibitors also need to overcome the challenge of the complex binding mechanisms of both cofactors and substrates. Although all of these enzymes are potential therapeutic targets, care should be taken in the way these targets are inhibited as several of the downstream products of the of the PGE₂ biosynthetic pathway are required for normal physiological functioning. The effect of the target should also take into account the effect that lowering/increasing PGE₂ level may have on preexisting autoimmune conditions or in causing such conditions.

In response to viral infections the host elicits a defense by activating inflammation and immunity. The activation of inflammation by viral infections leads to the production of PGE₂ which has a modulatory role in immunity and other roles in normal cell physiology. In this review, viruses from specific classes including, double-stranded DNA viruses (HSV, CMV, EBV, MAV-1), double-stranded RNA viruses (RV), (+) single-stranded RNA viruses (CVB3, EV71, PSaV), (−) single-stranded RNA viruses (VSV, IVA, PIV3, LCMV, RSV), single-stranded RNA-RT viruses (HTLV-III and HIV-l) and double-stranded DNA-RT viruses (HBV) and their interaction with PGE₂ were reviewed. PGE₂ had either an inhibitory, stimulatory or in some cases a dual role in there viral replication cycles. The stimulation of viral pathogenicity (HSV, CMV, EBV, RV, CVB3, EV71, PSaV, VSV, LCM, RSV, and HTLV-III) occurred mainly by affecting the host immunity, viral transcription/translation, and/or viral replication, while the inhibition (PIV3 and HBV) affected viral replication itself. In the case of both IAV and HIV-1, PGE₂ was shown to have both a stimulatory and inhibitory effect. Both these viral stimulations were dependent on immune suppression and gene expression, while the inhibitory effects were dependent on IL-27 expression (IAV) and inhibition of spreading (HIV-1). In many of the viral infections the effect of PGE₂ was negated when inhibitors of PGE₂ synthesis was added, but the effects were overcome by the addition of exogenous PGE₂. Interestingly, it appears that Th17 cells have a regulatory role in retrovirus infections. In turn Th17 cells are regulated by PGE₂ which might play an indirect role in these viral infections. There are a number of possible therapeutic targets in the PGE₂ biosynthetic pathway, although more research into their effects and modes of actions are required. This paper suggest that it may be possible to design therapeutic strategies to target selective pathways in an effort to attenuate inflammation associated with virus infection.

When looking into the majority of studies included in this review, the effect that PGE₂ has in viral infections seem to have been a coincidental discovery and not the primary objective of these studies. In some cases the discovery was noted, but not further investigated. There is, therefore, a major gap in our knowledge of the exactly role of PGE₂ in viral infections. These gaps include, determining the exact signaling pathways that viruses might use to induce PGE₂, the role that PGE₂ can play in the induction of autoimmunity and clarifying the dependence of viral infections of PGE₂. The underlying mechanisms should be further investigated to determine if viruses require PGE₂ to enhance their pathogenicity or require PGE₂ for optimal viral replication. Furthermore, light should be shed on the possible effect that the interplay between TLRs, PGE₂ and its receptors may have on viral infections. Taken together these mechanisms could also shed light on other host-pathogen interactions and may facilitate selection or development of the more optimal therapies against viral infections.

WS compiled the information, co-wrote the manuscript, and approved the final version submitted. HO and CP provided scholarly input in placing the literature into context, edited the manuscript, and approved the final version submitted.

The study was funded through the South African National Research Foundation (Grant No: 103395 to HO).