In the context of cellular targets, one of the most explored to date is represented by cellular enzymes involved in the various metabolic pathways related to cell RNA synthesis and also engaged in virus RNA transcription and/or replication. Precisely RIB, the only antiviral in clinical use for HF caused by arenaviruses, is a prototypical example of this class of agents.

RIB is a guanosine analog which once inside the cell is phosphorylated first to RIB-5’-monophosphate and then to RIB-5’-triphosphate. The primary target for this compound is supposed to be the cellular enzyme inosine monophosphate dehydrogenase (IMPDH) that converts IMP to xanthosine monophosphate, a limiting step for de novo intracellular synthesis of guanosine nucleotides [34,35]. The blockade of IMPDH by RIB-5’-monophosphate decreased the intracellular GTP pool with the consequent reduction of viral RNA synthesis and virus yield, an effect reversed by exogenous addition of guanosine. However, multiple mechanisms of action have been proposed for RIB according to the virus/host system analyzed [36,37]. In the particular case of the arenaviruses, it is currently in discussion which is the real mode of action for RIB. There is evidence for JUNV and LASV that the mechanism of action of RIB is not mainly focused on IMPDH inhibition and additional proposed targets may involve the viral RNA polymerase or lethal mutagenesis [38,39,40,41].

To overcome the shortcomings reported by the treatment of humans with RIB [9,10], other known IMPDH inhibitors were evaluated against different arenaviruses decades ago with variable results [42]. More recently, 5-ethynyl-1-β-D-ribofuranosylimidazole-4-carboxamide (EICAR) and mycophenolic acid (MPA) were successfully tested against pathogenic arenaviruses showing greater antiviral potency and selectivity than RIB in vitro against JUNV and LASV with MPA as the most active inhibitor (IC₅₀ values around 0.3 µg/mL) [38,39]. EICAR is a competitive inhibitor of IMPDH, whereas MPA is an uncompetitive inhibitor, and both compounds possess antiproliferative, immunosuppressive and antiviral properties [43,44]. JUNV and LASV inhibition was highly reversed by exogenous guanosine addition, indicating that antiviral activity is effectively associated with GTP depletion through IMPDH blockade as well as the high susceptibility of both HF arenavirus replication to this cellular pathway of inhibition [38,39].

Acridones are another class of compounds apparently related to cellular targets that have attracted attention in recent years for their wide range of biological properties. As for the antiviral properties of this class of compounds, several studies have shown the inhibitory action of acridone derivatives against DNA viruses such as herpes simplex virus and cytomegalovirus [45,46], Epstein-Barr virus [47] and adenovirus [48], as well as RNA viruses, including human immunodeficiency virus [49,50], bovine viral diarrhea virus [51] and hepatitis C [52,53]. The precise target and the mode of action of antiviral acridones are not clearly determined at present, although its predominant action seems to be centered on the synthesis of nucleic acids. It has been identified specific inhibitory activity against cellular enzymes such as DNA topoisomerase II [45,54] and IMPDH [55], and viral enzymes such as RNA helicase and RNA polymerase [53] as well as the ability to intercalate into the nucleic acid molecules, blocking the recognition and association of the enzyme to the modified molecule [56].

With respect to arenaviruses, a screening of antiviral activity of diverse novel N-substituted acridone derivatives identified a group of 10-allyl-9(10H)-acridones as effective and very selective inhibitors of arenaviruses [57]. In particular, the 10-allyl-6-chloro-4-methoxy-9(10H)-acridone, designated 3f, was the most active compound that blocked replication of the pathogenic arenaviruses JUNV and LCMV as well as the four serotypes of dengue virus, another HF-causing RNA virus belonging to the family Flaviviridae. Mechanistic studies demonstrated that JUNV RNA synthesis was strongly inhibited and the addition of exogenous guanosine rescued the infectivity and viral RNA levels in a dose-dependent manner [39]. However, the guanosine reversal was partial, suggesting that GTP pool reduction may contribute to the inhibitory action but it is not the only operative mechanism: a possible cellular target is represented by IMPDH whereas another still unidentified target, presumably viral, given the high selectivity of the compound, is also involved in the antiviral activity of the acridone against JUNV. Then, the possibility of a multiple effect on different targets, as reported for other acridones [53], turns promising perspectives for these agents in treatment of pathogenic arenavirus infections allowing a potent antiviral efficiency and less feasibility of generating resistance.

Inhibitors of the biosynthetic pathway of pyrimidine nucleotides were also studied for anti-arenavirus activity. Compounds evaluated include cytidine analogs targeted to cytosine triphosphate synthetase (CTPS); carbanucleoside analogs which are inhibitors of the orotidylic acid decarboxylase (OMPD) which converts OMP to uridine monophosphate (UMP); and analogs of adenosine, most of them acting as inhibitors of S-adenosylhomocysteine hydrolase (SAHH), a key enzyme in the transmethylation to obtain the 5 ’cap of the mRNA [42,58,59]. However, the selectivity and effectiveness of these series of analogs were not very promising without showing appreciable advantages over RIB.

The finding and characterization of molecules with broad-spectrum inhibitory action against viral RNA synthesis and minor cellular toxicity, as MPA and acridones above commented, represent novel promising alternatives for chemotherapy, although further in vivo evaluation is necessary to renew the consideration of this class of metabolic inhibitors against arenaviruses.

Heterogeneous nuclear ribonucleoproteins (hnRNPs) constitute a family of cellular RNA binding proteins which participate in RNA metabolic processes including mRNA splicing, nucleo-cytoplasmic trafficking, translation and turnover [60,61,62]. More than 20 hnRNPs have been characterized, called from hnRNP A1 to hnRNP U, and those belonging to the A/B group (hnRNP A1, A2/B1, A3) are among the smallest but most abundant hnRNPs [62]. In virus-infected cells, several hnRNPs are able to interact with viral components, RNA and proteins, modulating the multiplication of both DNA and RNA viruses. Several studies have proved the involvement of hnRNPs in different stages of viral replicative cycle, mainly in RNA synthesis, processing and translation [63,64,65]. In addition, several viruses induce changes in the level of expression and intracellular localization of these predominantly nuclear proteins and these alterations may facilitate viral replication and/or help the virus to counteract the host cell antiviral response [66,67].

There is some evidence about changes in hnRNP expression in arenavirus-infected cells that suggest a possible role of these cell factors on both viral infectivity and/or pathogenesis. Pichinde virus (PICV) infection of guinea pigs produces a pathology similar to Lassa fever in humans [68] and two viral variants, P2 (an attenuated virus that causes a mild infection) and P18 (a virulent virus that produces a severe HF in the guinea pig model), have been used to analyze cell signaling pathways and responses following infection of a murine macrophage-like cell line [28,29,69]. These studies identified several cellular signaling events which may be involved in PICV pathogenesis [28,29]. Virus-host interactions were further characterized by investigating the differential nuclear proteomes induced by the infection of these two PICV variants [69]. This analysis revealed that several hnRNPs (hnRNP A2/B1, hnRNP A3, hnRNP D0 and hnRNP L) exhibited substantial differential expression after infection with the attenuated or the virulent isolates. The subcellular localization of hnRNPs during infection was also investigated: at early times after infection only the attenuated variant P2 induced the cytoplasmic localization of hnRNP A2/B1, whereas hnRNP A1 remained predominantly nuclear for both P2 and P18. The results obtained with hnRNP A2/B1 are in accordance with previous data showing that at the initial stages of the infection cells infected with the virulent P18 variant resemble uninfected cells. Bowick et al., proposed either a failure in the activation of cell response or the suppression of this response after infection with the virulent virus, suggesting that virus-mediated alterations in hnRNP intracellular localization may be implicated in PICV pathogenesis [69].

Infection of macaques with a lethal dose of LCMV also constitutes a model for LASV infection of humans [70] and the analysis of the expression of cell genes in a blood sample by using DNA microarray technology made possible the identification of genes that may serve as markers for disease progression and to gain insights into the molecular mechanisms triggering viral HF. Moreover, the comparison of the blood profiles after infection with the virulent LCMV strain WE or the non-virulent strain Armstrong [71] enabled the finding of genes that would be associated with viral pathogenesis [72]. This comparative study revealed that hnRNP C is downregulated in peripheral blood mononuclear cells isolated at early times after infection from rhesus macaques that were infected with WE [72]. It is interesting to note that virus-mediated changes in hnRNP C expression or localization would result in the inhibition of interferon (IFN) signaling since this hnRNP has been proved to be involved in phospho-STAT1 nuclear import, so a reduced level of hnRNP C in WE-infected cells may contribute to the severity of the infection [63,67].

The first report providing evidence about the importance of hnRNPs on JUNV replication was a comparative study of acute and persistent JUNV infections performed in Vero cell cultures [73]. Silencing of hnRNP A1 or hnRNP A2, by using small interfering RNAs (siRNAs), caused a decrease in JUNV protein synthesis and progeny virus production during acute infection. It was also demonstrated that acute JUNV infection does not affect hnRNP A/B levels of expression in comparison with uninfected cells and promotes the cytoplasmic accumulation of hnRNPA1. Cytoplasmic re-distribution of hnRNP A1 could also be detected in cells expressing the viral nucleoprotein and co-immunoprecipitation studies revealed the interaction between both proteins. Unlike the acute infection, persistently infected cultures do not produce infectious virus and exhibit a continuous synthesis of nucleoprotein, associated with blockage of the expression of the glycoprotein G1 and absence of cytopathic effect [74,75]. A marked diminishment in the hnRNP A/B expression was observed in persistently JUNV-infected cultures. Furthermore, overexpression of hnRNP A1 allowed determining that cells persistently infected with JUNV do not induce hnRNP A1 cytoplasmic re-localization. These results indicate that hnRNP A1 would favor JUNV productive infection through the interaction with the nucleoprotein. On the other hand, JUNV non-productive persistent infection might involve downregulation of hnRNPs A/B, as well as changes in nucleo-cytoplasmic trafficking [73].

Despite much remains to be investigated about the relation between hnRNPs and arenavirus multiplication, persistence and pathogenesis, the data available to date indicate that these molecules can be considered as possible targets for antiviral therapy. However, it should be noted that hnRNPs do not always act as promoters of viral replication [63] and there are situations, such as hnRNP C downregulation by LCMV-WE infection mentioned above [72], in which a decrease in the level of expression of these cell factors could be associated with an increased viral pathogenicity.

The phosphorylation of proteins, either viral or cellular, is an event of key importance that determines the fate of several features during the infection process related to viral multiplication and cell survival as well. Phosphorylation is a reversible post-translational modification that proteins may suffer in order to regulate stability, biological activity and interactions with other proteins. This process is chemically characterized by the addition of a negatively charged phosphate group to specific serine or threonine residues of the protein and may be accomplished by cellular or viral kinases. On the contrary, removal of phosphate groups from protein backbone is achieved through the action of phosphatases which, together with kinases, control most of the cellular signaling pathways involved in metabolism, transcription and translation, cell cycle, apoptosis, differentiation and immune response [76,77].

Solid evidence shows that viruses require the modulation of cell signaling pathways in order to ensure their successful replication. In this context, modulation of cell survival pathways is a key target for both DNA and RNA viruses as a strategy to temporarily overcome cell death and prolonging viral replication. The phosphatidylinositol-3-kinase/protein kinase B (PI3K/Akt) is one of the major signaling pathways involved in cell survival, as well as proliferation, migration and apoptosis and, as such, has been shown to be widely involved in critical viral events like virus uptake, replication, translation and budding [78,79,80]. Several signaling pathways are also modulated by PI3K/Akt, among which are Bad, caspase-9, IKK, the Forkhead transcription factors, Mdm2, YAP, mTOR and the Forkhead box O (FoxO) transcription factors [81,82] and hence may be potentially regulated by viruses through PI3K modulation.

It has been reported that JUNV is able to activate Akt-phosphorylation mediated by PI3K at an early stage of infection. This activation triggered during virus uptake would be necessary for an efficient recycling of transferrin receptor, main receptor involved in JUNV binding to the cell surface [83,84]. The inhibition of Akt phosphorylation by the PI3K inhibitor Ly294002 impaired virus production in different cell types indicating the essential role of PI3K/Akt signaling for JUNV infection. In view that PI3K/Akt activation is also involved in the stabilization of the mRNA encoding hnRNPA1 [85] and taking into account, as previously mentioned, the participation of this factor during JUNV replication [73], it is tempting to speculate that activation of this pathway may contribute in other stages of virus multiplication besides virus uptake.

The participation of PI3K during the early stages of LASV and LCMV infection has also been described. For these Old World arenaviruses, which use α-dystroglycan as cellular receptor [86], the biosynthesis of lipid membrane phosphatidylinositol 3-phosphate (PI3P), mediated by PI3K, would be necessary to support the formation of multivesicular bodies, mechanism employed by these viruses to invade host cells [87]. In contrast, in a recent work it has been reported that inhibition of PI3K/Akt, but not of the downstream effector mTOR/raptor, during LCMV infection impaired budding and, at a lesser extent, RNA synthesis but not virus entry. Moreover, inhibition of PI3K/Akt reduced virus budding mediated by the matrix protein Z in a dose- dependent manner, suggesting the involvement of this cellular pathway in virus budding [88]. This apparent discrepancy about the involvement of PI3K in LCMV entry may be ascribed to differences in LCMV strains employed for each report as well as to the different mechanisms of action of PI3K inhibitors used by both groups. Noticeably, Urata et al., demonstrated that BEZ-235, a PI3K inhibitor currently in cancer clinical trials, inhibited LCMV multiplication in cultured cells, reinforcing the good perspectives of cell-targeted agents in general and PI3K blockade in particular for arenavirus chemotherapy [88].

Interestingly, modulation of PI3K/Akt pathway was observed in CD8+ T cells during acute LCMV infection. In vivo phosphorylation levels of Akt, mTOR and FoxO1/O3, a critical regulator of T cell homeostasis activated by PI3K/Akt pathway, showed dynamic alterations during the course of infection: while Akt activation correlated with mTOR activity, no correlation was observed in FoxO1/O3 phosphorylation, inhibiting thus apoptosis of infected lymphocytes [89].

As mentioned above, the comparison of signaling proteins regulated during macrophage infection with attenuated or virulent strains of PICV also allowed detect a differential response to each strain in three of the central signaling networks involving p53, c-Myc, and Akt, respectively [28]. Further studies revealed that macrophages isolated from guinea pigs infected with the P2 or P18 strain of PICV showed differential phosphorylation status of ERK [29], a component of the RAF-MEK-ERK1/2 signaling pathway, other central route that regulates many cellular functions related to virus multiplication [90]. Although this pathway controls cell differentiation and proliferation leading to changes in gene expression, cell metabolism and apoptosis, little is known about its modulation by arenaviruses. In this regard, increase in ERK phosphorylation has been reported during PICV multiplication [91]. On the contrary, LASV infection impaired MEK/ERK activation after virus binding to its receptor [92].

The trafficking mechanism of arenavirus is also an important point to be considered as a possible cellular target [93]. Upon binding receptor, endosomal acidification is a mandatory step for internalization of either New or Old world arenaviruses [94,95,96,97], although cellular receptor and hence the trafficking mechanisms are different. In case of LCMV and LASV, virus internalization is cholesterol-dependent but is independent of clathrin, caveolin, and dynamin. On the other hand the proteins Rab-5 and Rab-7 are partially required while ARF6, flotillin and actin would not to be necessary for LCMV and LASV entry [98,99,100,101]. On the contrary for the New World arenaviruses JUNV and PICV, virus internalization is clathrin- and dynamin-dependent but is partially dependent of cholesterol and independent of caveola. The proteins Eps15, Rab-5 and Rab-7 have also been shown necessary for both viruses [102,103,104]. In this regard, although novel approaches have been evaluated for antiviral drug development [105], the possibility to consider inhibition of kinases participating in the internalization process remains in the speculative field [106].

In conclusion, the modulation of survival signaling pathways is a critical event in the replication cycle of arenavirus. Given that these pathways are highly involved in the cell deregulation that occurs during cancer, much progress has been obtained in the development of potential chemotherapy agents against different components of these pathways that might be used as anti-arenaviral drugs.

The host immune response has been widely accepted to be one of the essential factors to determine the outcome of LCMV infection in mice [107]. From many of these studies it can be concluded that viral modulation of phosphorylation of proteins involved in the IFN response is a key event to explain pathogenicity.

Type I interferon (IFN-I) has been pointed out as an important factor that may have beneficial or detrimental consequences during LCMV infection of mouse [108]. On this basis, it has been demonstrated that inhibition of the innate immune response, mainly supported by the production of IFN, is necessary for the establishment of viral persistence in the mouse [109]. In vitro infection of several cell types by different arenaviruses showed the induction of IFN-I [110,111]. It has been reported that LCMV nucleoprotein negatively modulates IFN production in persistently infected A549 cells by inhibiting nuclear translocation of the interferon regulatory factor 3 (IRF-3) [112]. This property has been also demonstrated for nucleoproteins of several representatives of New and Old World arenaviruses except for Tacaribe virus nucleoprotein [113]. LCMV nucleoprotein is able to interact with the IκB kinase (IKK)-related kinase IKKε leading to a loss of its phosphorylation activity towards IRF-3 [114]. At the same time, LCMV nucleoprotein is able to interfere with the nuclear translocation of the nuclear factor kappa B (NF-κ B) suggesting a multi-factorial role of this protein in the modulation of innate immune response and inflammation [115]. Also, Z protein from New World arenaviruses, but not from Old World members, antagonize IFN response by binding to the retinoic acid-inducible gene I (RIG-I) and inhibiting downstream activation of IFN-I expression signaling pathway [116].

One of the main mechanisms mediated by IFN and employed by cells to restrict viral infection is the up-regulation of the protein kinase R (PKR), which is activated by double-stranded RNA, a species that can be readily detected during the replication of many viruses. PKR expression is augmented during the IFN response and, in the case of infection, it phosphorylates the α subunit of eukaryotic translation initiation factor 2 (eIF2α), resulting in the blockage of cap dependent translation and impairment of the bulk of cellular protein synthesis and in many cases, viral protein synthesis as well [117]. At the same time, phosphorylation of eIF2α leads to the formation of stress granules, cytoplasmic aggregates that contain pre-initiation complexes that accumulate as a consequence of translation blockage. Many viruses can modulate eIF2α phosphorylation in order to counteract cell restriction [118]. On this line, it has been reported that JUNV is able to impair eIF2α phosphorylation upon stressing of infected Vero cells with sodium arsenite. Both, the nucleoprotein and the glycoprotein precursor were able to maintain low eIF2α phosphorylation levels, similar to those observed for control cells [119]. Although the mechanism through which arsenite induces eIF2α phosphorylation is under discussion, PKR may be one of the kinase targets for this drug [120].

Another point that might be considered as a prospective target for anti-arenaviral therapy is the Toll-like receptor 2 (TLR2) pathway that activates host defense mechanisms in order to prevent and eradicate infections and participates in the generation of the adaptive immune response [121]. Several years ago, Zhou et al., studied the role of TLR2 and myeloid differentiation factor 88 (MyD88) during an LCMV infection and demonstrated that TLR-mediated responses are necessary for an efficient innate immune response against this virus and that MyD88 would be of crucial importance for the maturation and activation of virus-specific CD8 (+) T cells [122]. Further studies revealed that LCMV infection of glial cells leads to the production of inflammatory chemokines, such as MCP-1, RANTES and TNF-alpha in a TLR2-MyD88 dependent manner [123]. On this perspective, a comprehensive study of prospective inhibitors of TLR2 signaling described the characterization of one of the active compounds that also exhibited antiviral properties against LCMV, suggesting that this approach may be a promising tool for anti-arenaviral therapy [124]. It should be emphasized that the studies mentioned above were performed with the Armstrong strain of LCMV. Recent findings demonstrate that the WE strain of LCMV and LASV, both able to induce a fatal disease in rhesus macaques, suppressed pro-inflammatory cytokine responses that depend on TLR2. By contrast, nonpathogenic Mopeia arenavirus is able to activate TLR2 pathway, as described for the Armstrong strain of LCMV [125]. Similar results were reported for JUNV infection of murine macrophages, where expression of the TNF-alpha and beta IFN through a TLR2 pathway are activated by the viral glycoproteins [110].

In view of the results obtained to the present, impairment/disruption of innate immune response would be a key parameter that controls arenavirus pathogenicity. This approach deserves further investigation in order to determine if kinases involved in the modulation of IFN production/response would be attractive targets for antiviral therapy.