Understanding host responses to equine encephalitis virus infection: implications for therapeutic development

ABSTRACT

Introduction

Venezuelan, eastern, and western equine encephalitis viruses (VEEV, EEEV, and WEEV) are mosquito-borne New World alphaviruses that cause encephalitis in equids and humans. These viruses can cause severe disease and death, as well as long-term severe neurological symptoms in survivors. Despite the pathogenesis and weaponization of these viruses, there are no approved therapeutics for treating infection.

Areas covered

In this review, we describe the molecular pathogenesis of these viruses, discuss host-pathogen interactions needed for viral replication, and highlight new avenues for drug development with a focus on host-targeted approaches.

Expert opinion

Current approaches have yielded some promising therapeutics, but additional emphasis should be placed on advanced development of existing small molecules and pursuit of pan-encephalitic alphavirus drugs. More research should be conducted on EEEV and WEEV, given their high lethality rates.

KEYWORDS:

• Alphavirus
• encephalitis
• equine
• therapeutics
• host-pathogen interaction
• pathogenesis

1. Overview of VEEV, EEEV, and WEEV biology

1.1. Equine encephalitic alphaviruses

Equine encephalitic alphaviruses are New World zoonotic pathogens belonging to the Togaviridae family. They are enveloped, positive sense single-stranded RNA viruses. There are three of these viruses that cause encephalitic disease in both equines and humans: Venezuelan, eastern, and western equine encephalitis viruses (VEEV, EEEV, and WEEV, respectively). These viruses are classified as Category B pathogens by the National Institute of Allergy and Infectious Diseases (NIAID) and are transmitted naturally by mosquitos, but can also become infectious upon aerosolization, making them a concern for bioweaponization. Non-South American genotypes of EEEV and strains of VEEV excluding enzootic IE and ID subtypes are considered as selected agents and require additional clearances and regulations for laboratory study. Natural transmission occurs between the reservoir hosts (typically rodents and/or birds) through a mosquito vector during the enzootic cycle. Bites from infected mosquitos to non-host species initiate the epizootic cycle, resulting in infection in more susceptible species, including horses and humans. Different mosquito species are involved in epizootic versus enzootic cycles and can vary between the viruses. As opposed to other pathogenic alphaviruses, which cause arthritic syndromes and are largely found in the Eastern hemisphere, VEEV, EEEV, and WEEV can cause severe encephalitis in humans (Table 1). Long-term costs for health-care support for individuals with neurological defects due to alphavirus-induced encephalitis have been estimated at over $400,000 per year per person [1]. In an alphavirus outbreak in Texas in 1971, Earnest et al. [2] estimated that the costs for caring for infected patients was $320,000 per individual. In the event of a large outbreak or attack on military personnel, the cost in mortality, morbidity, and short- and long-term care are likely to be extremely high. Significantly, there are no approved therapeutics for VEEV, EEEV, or WEEV-infected individuals, and infections are treated with only supportive care. Therefore, these viruses are dangerous, cause severe disease with significant mortality and long-term morbidity, have a history of bioweaponization, and have no protective countermeasures available.

Table 1. Comparison of VEEV, EEEV, and WEEV.

	Neurological symptoms	mortality	# cases	Location
VEEV	1–4%	<1%	>250,000	Central America, South America, southern US
EEEV	50–90%	30–70%	>380	Eastern US, eastern Canada, Central America, South America (Madariaga virus)
WEEV	3–15%	3–7%	>4,000	Western US, western Canada, South America

1.1.1. VEEV

VEEV was first discovered in equines (horses, donkeys, and mules) in 1936 after an investigation into sick horses in Venezuela. Human VEEV infections were first reported in the 1960s, where a connection with mosquitos as the transmitting vector were described, as were cases of laboratory-acquired infections [3]. Since that time there have been hundreds of thousands of human infections by VEEV in the Americas. Large individual outbreaks in humans with VEEV have been reported, such as in 1995 in Columbia, which resulted in ~75,000 infections, 3,000 cases of neurological sequelae, and 300 deaths. A 1971 outbreak in Texas involved 86 infected individuals, and at least 12 had long-term neurological complications [4]. In the 1960s in Columbia [5], approximately 200,000 human cases of VEEV occurred, along with lethal infection in 100,000 equines [6]. There are several subtypes of VEEV. Subtypes I-A, I-B, and I-C are associated with human disease and are also highly virulent in horses. Importantly, mosquitos can carry the virus from horses and infect other horses and humans. However, subtypes I-D, I-E, I-F, II, III, IV, V, and VI are not typically pathogenic. There is an experimental VEEV vaccine (TC-83) given to select personnel, but its efficacy is unknown, and it has substantial side effects [7,8] . Although an inactivated vaccine exists, it is given to equids but not typically administered to humans.

1.1.2. EEEV

Disease caused by EEEV was first discovered in 1831 in Massachusetts, USA when horses started becoming ill [9]. The virus itself was not isolated until 1933 from an infected horse’s brain [9]. As its name implies, EEEV is found in the eastern part of North America and is largely a seasonal virus, with peaks in summer months [10]. EEEV is most frequently found in the mosquito Culiseta melanura and is thought to spread amongst several bird species, but can also feed on other mammals and reptiles. EEEV-induced disease in humans is rare, with 380 cases reported in the United States between 1964–2020. In recent years, human EEEV infections have risen, with 38 cases reported in 2019, which is more than was cumulatively reported in the previous decade [11]. Infection is extremely pathogenic, with the case fatality rate being approximately 30% in humans (up to 70% in some outbreaks) and as high as 90% in equids. Those that survive can have permanent brain damage. EEEV is divided into two main groups – EEEV, the more pathogenic form found in North America, and Madariaga virus, the less pathogenic virus found in South America. In 2010, investigation of an outbreak of encephalitis in Panama revealed both EEEV and VEEV cases, including one patient co-infected with both viruses [12]. There are no vaccines available for human use, but there is an inactivated vaccine that is used in equids.

1.1.3. WEEV

WEEV is primarily found in the western parts of North America. WEEV is severely pathogenic in younger patients, with 90% mortality in children less than 1 year old and an overall mortality rate ranging from 3 to 15% [13,14]. Aerosol infections in laboratory accidents indicate that this route of infection may result in a higher mortality rate [15], highlighting the concerns for bioweaponization of this virus. Fatality rates in horses are approximately 30% [14]. There have been over 600 human cases reported in humans in the United States from 1964 to 2010. Although the frequency of WEEV has steadily decreased over the last several decades [16,17], horses in Mexico were reported to be infected in 2019 [18]. The primary vector for WEEV is thought to be Culex tarsalis, which circulates the virus amongst birds and small mammals. Similar to EEEV, no human vaccine is available, although an inactivated vaccine is approved for use in equids.

1.1.4. Veterinary Prospective

Encephalitic alphaviruses are important pathogens from a veterinary perspective, as these viruses are much more pathogenic in equines than in humans. Although equine vaccines are available, there is still widespread infection in horses and related animals. From 2003 to 2021, VEEV seroprevalence rates in unvaccinated horses from 23 municipalities in three different Mexican states was a total of 52%, ranging from 10 to 100% in each municipality [19]. EEEV has primarily been found in horses in the Southeastern US, but in the last few decades it has spread towards Canada, with one study finding more than 8% of horses having antibodies against the virus [20]. Madariaga virus is endemic in Central and South America and causes significant morbidity and mortality. Some studies have found seroprevalence to be over 26% in Central and South America (reviewed in [21]).

WEEV caused large outbreaks in Canada and the US in the 1930s and 1940s, with hundreds of thousands of cases in horses. The last reported case of WEEV in humans in North America was 1994 and the most recent identification of WEEV in mosquitos in North America was 2008, but dozens of horses were infected in Mexico in 2019 [18,22] and a child died of the virus in Uruguay in 2009 [23].

Recent seroprevalence in horses in Uruguay was 4.6–13% for VEEV and 3–4% for WEEV and Madariaga virus[24]. Neutralizing antibody frequency was reported in serum from equids and sheep in Brazil in 2009–2011 [25], where only 0.2% of unvaccinated equids had neutralizing serum antibody titers against VEEV and 0.8% against WEEV, while 9.9% of unvaccinated equids had neutralizing antibodies against EEEV. Therefore, the encephalitic alphaviruses are still circulating, widespread pathogens that have both human and agricultural impacts.

1.2. Genomes of equine encephalitis viruses

VEEV, EEEV, and WEEV have similar genetic organization. All are positive strand enveloped RNA viruses. Their single-stranded, positive sense RNA genomes are approximately 12 kB and consist of two open reading frames [26]. The first encodes four nonstructural genes and the second encodes the six structural genes. The four non-structural proteins (aptly named nsP1-4) are involved in viral transcription, translation, and genome replication. The six structural proteins have standard roles such as formation of virion structure, encapsidation of the viral genome, cellular attachment and entry, budding, and immune evasion and antagonism of cellular antiviral activities [27] (Table 2).

Table 2. Function of encephalitic alphavirus genes.

Protein	Viral function
nsP1	RNA capping
nsP2	Helicase, protease
nsP3	Replication
nsP4	RNA dependent RNA polymerase (RdRp)
Capsid	Virion structure, genome encapsidation, budding, antagonism of cellular antiviral response
E1, E2, E3	Cellular attachment/entry
TF	virulence factor
6K	Budding, glycoprotein spacer, viroporin

2. Pathogenesis

The mosquito-borne encephalitic alphaviruses (EEEV, VEEV, and WEEV) cause outbreaks of encephalomyelitis and are a significant threat to the civilian population and the warfighter. Neurological sequelae have been documented in up to 75%, 14%, and 90% of EEEV, VEEV, and WEEV survivors, respectively [28]. Sequelae observed following VEEV infection includes convulsions, somnolence, confusion, photophobia, and coma in 4–14% of survivors [28]. EEEV infections can result in convulsions, seizures, and paralysis in 50–90% of survivors, whereas 15–30% of WEEV survivors have confusion, visual disturbances, photophobia, seizures, somnolence, coma, and spastic paresis (28). In addition, intellectual disability and behavioral changes are sequalae common amongst VEEV, EEEV, and WEEV infection survivors (28).

Neuropathologic data for VEEV in humans are limited. However, cerebral edema and meningeal infiltrates containing neutrophils, lymphocytes, and monocytes have been observed in lethal encephalitic cases [14,29]. Similarly, in horses, VEEV infection results in meningoencephalitis and brain lesions characterized by gliosis, satellitosis, hemorrhage, and vasculitis (reviewed in [14]). Damage within the brain is widespread, but the cerebral cortex is most severely impacted [14]. Animal models for VEEV include nonhuman primates, as well as guinea pigs and hamsters. In addition, different strains of mice (including BALB/c, C57BL6, CD-1, and C3H/HeN) are commonly used for VEEV studies. VEEV infection of mice results in overt encephalitis. Neuronal damage is observed and can be caused by apoptosis or necrosis, with apoptosis found in the granule-type neurons in the hippocampus and the cerebellum [14]. In the CNS, VEEV infection results in the upregulation of numerous genes in the inflammatory response and apoptotic pathway [6,30]. Specifically, pro-inflammatory cytokines including interleukin-1β (IL-1β), IL-6, IL-12, and tumor necrosis factor-α (TNF-α) play a role in VEEV pathogenesis [26]. Sharma et al. analyzed gene expression changes in the brain tissue of VEEV infected CD-1 mice and discovered alterations in immune pathways involved in antigen presentation, inflammation, apoptosis and the traditional antiviral response (Cxcl10, Cxcl11, Ccl5, Ifr7, Ifi27, Oas1b, Fcerg1, Mif, Clusterin and MHC class II) [31]. C57BL/6 mice intranasally infected with the vaccine strain VEEV TC-83 survive infection and display neurological sequalae, including hippocampus involvement [32]. In this model, symptomatic C57BL/6 mice (ruffled fur, hunched back, and reduced activity) displayed deficits in prepulse inhibition (the inability to filter out extraneous sensory stimuli). Histopathological analysis of brain sections, indicated that symptomatic C57BL/6 mice had thalamus damage (glial nodules and calcification) and increased glial fibrillary acidic protein (GFAP) expression in the thalamus posterior complex and the hippocampus dentate gyrus. Overall, these studies suggest that VEEV-induced inflammation contributes substantially to neurological damage.

In humans, EEEV infections have been more thoroughly documented [29]. Brain lesions preliminary occur within the basal ganglia, thalamus, and brainstem. Fatal infections display neuronal injury with vasculitis and thrombosis, demyelination, necrosis, and caspase 3 activation. Parenchymal, meningeal, and perivascular infiltrates containing monocytes/macrophages, neutrophils, and lymphocytes have been observed. Animal models of acute EEEV infection to date include mice (including CD-1, BALB/c, C57BL/6), hamsters, guinea pigs, and non-human primates (NHPs) [14]. NHPs develop many features of EEEV disease observed in humans such as neuronal tropism and necrosis, meningoencephalitis, and vascular damage. Mice are susceptible to EEEV infection when exposed subcutaneously (SQ), intranasally, and through aerosol, and they die fairly rapidly, within 4–9 days post infection (depending on the mouse strain and route of infection) [33–35]. Infection of young mice results in a biphasic disease course, with virus detected in the periphery followed by central nervous system (CNS) invasion [14]. Neuroinvasion likely occurs through the hematogenous route, rather than through the olfactory or peripheral nerves [36]. A recent study suggests that following peripheral infection, EEEV is able to gain access to the brain through circumventricular organs, which naturally lack a typical blood-brain barrier [37]. Studies with adult CD1 mice found that EEEV does not replicate well in lymphoid tissues (specifically in macrophages and dendritic cells) [38], resulting in quick clearance of the virus from the periphery. EEEV replicates primarily in neurons with some detection of antigen found in glia cells [36,39]. Outside the brain, EEEV has been found in multiple cells types including fibroblasts, cardiac myocytes, skeletal muscle, and osteoblasts [39]. Mice display encephalitis, paralysis, neuronal necrosis, and inflammation as indicated by infiltration of eosinophils and neutrophils [14,39,40]. The development of seizures or epileptogenesis is a commonly observed neurological sequalae in EEEV patients as well as mice [28,34].

Humans infected with WEEV typically have a mild clinical course, but significant neurological pathology has been documented in lethal cases and 15–30% of infected individuals develop debilitating neurological sequalae. Pathological alterations include perivascular cuffing, multifocal necrosis, and gliosis as well as lesions on the spinal cord [14]. Animal model development is ongoing for WEEV, but NHP and hamster models have been established. Different strains of mice, including BALB/c, CD-1, and Swiss, are well-established and commonly used. Mice infected with WEEV develop encephalitis with most models having lethal outcomes, but the severity of disease depends on the age and strain of the mice [14]. Non-lethal infection of CD1 mice with WEEV induces a persistent neuroinflammatory response, including activation of microglia and astrocytes and loss of dopaminergic neurons [41]. Protein aggregation was also noted and gene expression changes observed which were similar to alterations observed in Parkinson disease. This model provides an intriguing system to begin to understand the molecular underpinnings of WEEV neurological sequalae, especially given that WEEV infection in humans have resulted in ‘parkinsonian syndrome’ [28,42].

3. Host immune responses to infection

3.1. Cells targeted in alphavirus infection

VEEV replication occurs in two phases [43]. In the first phase, the virus targets lymphoid and myeloid cells in the absence of infection in the neurological system. VEEV targets dendritic cells and macrophages early in infection, but EEEV does not, preferring non-immune cells such as osteoblasts, fibroblasts, and muscle cells [38,39]. This restriction of EEEV in dendritic cells is not due to type I interferon (IFN) activity [38]. Similarly, a second study found that neither EEEV nor WEEV productively infects monocytes or lymphocytes, while VEEV replicates well in monocytes [44]. The first stage of encephalitic alphavirus replication leads to high viremia, which is then cleared and followed by the second phase, which is infection of neural tissues, including astrocytes and neurons. This infection can cause neurological complications through either replication-induced cell death or, potentially, induction of harmful inflammatory responses. The question of whether the immune response to the second phase of viral replication is beneficial or harmful is as yet unresolved, but pathogenesis in humans is correlated with an increase of neutrophils followed by lymphocytes in the CSF and increased neutrophils, lymphocytes, and mononuclear cells in the CNS (reviewed in [11]).

3.2. T and B cells

Adaptive immune cells may play a role in VEEV pathogenesis, as one study showed the time to death of VEEV-infected wild-type CB17 mice was 6.8 days, while SCID mice lacking T and B cells died at an average of 8.9 days [45]. An earlier study showed increased inflammatory demyelination in wild-type BALB/c mice but no alterations in white matter of mice lacking B and T cells [46]. Other studies showed that treatment of C3H HeJ mice with anti-thymocyte serum that depleted T cells actually delayed time-to-death of VEEV-infected mice by two days and delayed the peak replication of VEEV in spleen, blood, and brain [47].

Subcutaneous infection of mice lacking B cells (uMT mice) develop a rapid, severe disease with paralyzing encephalomyelitis after infection with an attenuated, non-pathogenic strain of VEEV (V3533) [48]. However, 93% of mice survived infection and recovered from severe disease. Recovery was associated with an increase of T cells and inflammatory monocytes in the CNS, although peak cellular influx occurred days after peak viremia. Depletion of T cells with anti-CD3 antibody in uMT mice exacerbated weight loss through day 6, but these animals had increased weight compared to uMT mice between days 8–15, and then had more severe weight loss between days 20–24. Depletion of either CD8 + T cells or CD4 + T cells in uMT mice mitigated early disease symptoms but resulted in more severe long-term pathology. Mice lacking both B and T cells (Rag -/-) showed severe disease with 93% mortality and a mean time to death of 30 days.

Therefore, it appears that both B and T cells are required for control of VEEV infection, although CD4+ and CD8 + T cells may also contribute to disease in the acute phase of infection.

3.3. Cytokines and chemokines

Delivery of pegylated IFN-alpha, but not IL-12, protected BALB/c mice from VEEV infection and was associated with increased macrophage activation, decreased TNF-alpha production in macrophages, and decreased CD4 + T cell, CD8 + T cell, and B cell activation (based on CD69 expression). VEEV titers were reduced in blood, spleen, lungs, and brain throughout infection as opposed to untreated animals [49]. Mice lacking IFNAR or IRF-2 were more susceptible to VEEV pathogenesis [50], showing the importance of the type I IFN antiviral pathway in protection.

Infection of astrocytes with VEEV results in increased production of TNF-alpha and low inducible nitric oxide synthase [51]; iNOS may play a role in VEEV pathogenesis in vivo [50]. One study found no difference in cytokines produced during pathogenic versus non-pathogenic VEEV infection in C57BL/6 mice, but importantly the timing of cytokine expression varied [52]. In non-pathogenic infection, there was a 24-hour delay in IFN-gamma, IL-6, IL-12, TNF-alpha, and IL-10 gene expression in draining lymph nodes compared to pathogenic infection; there was not as significant a trend in non-draining lymph nodes or spleens. However, neither IL-12 injection nor depletion affected VEEV pathogenesis, whereas blocking type I IFN resulted in increased virulence. In contrast, in EEEV infection, BALB/c mice infected via the aerosol or intranasal route, developed rapid onset of clinical manifestations and neuroinvasion [34] and had elevated serum levels of the inflammatory cytokines IFN-γ, MIP-1β, MIG, and G-CS. However, BALB/c mice infected via the subcutaneous route showed a spike in IFN-γ, RANTES, MIP-1β, and MIG serum levels 1 day earlier than aerosol or intranasal infected animals, and a late increase in G-CSF. Whether these differences in EEEV and VEEV-induced cytokine expression delay is due to route of infection, strain of mice used, or pathogenic versus non-pathogenic strains requires further study.

Multiple studies have confirmed that infection with virulent VEEV induces inflammatory gene pathways, including TLRs, inflammatory chemokines and cytokines, signaling pathways, antigen presentation, and apoptosis in the brain of infected CD-1 mice [31,53]. Infection of CD-1 mice with V3000 (recombinant TrD) and the attenuated but partially virulent V3034 clone was used to assess gene expression differences [54]. Overall, pathogenesis correlated with increased gene signatures for inflammatory pathways in the brain of infected animals and increased induction of innate immune cells in the spleen.

As noted above, WEEV infection in CD-1 mice also results in rapid death, and pre-treatment with innate immune cell-stimulating cationic lipid-DNA complexes causes partial protection of animals and correlated with decreased MCP-1, IFN-gamma, TNF-alpha, IL-12, and IL-10 protein levels in brain [55].

3.4. Other immune components

Complement may play a role in protective responses, as mice lacking C3 demonstrate severe encephalitis after infection with an attenuated VEEV (V3533) that causes a transient and mild disease course in wild-type mice [56]. Disease in C3-deficient mice was associated with a more rapid viral dissemination into the CNS and delayed viral clearance, even though the antibody responses were similar to wild-type mice. Interestingly, C5 deficiency had no effect on pathogenesis.

Adhesion molecules on microvascular endothelial cells that comprise the BBB can have an effect on inflammatory cell entry and resultant pathogenesis. Several of these components are differentially regulated in VEEV infection. ICAM-1 KO mice had decreased brain inflammation and a slightly delayed time to death [57]. Treatment with the anti-inflammatory NSAID drug naproxen did not decrease viral replication in the brain, but showed decreased brain inflammation.

3.5. Summary of host immune responses in encephalitic alphavirus infection

It appears that the immune response to encephalitic alphavirus infection is complex, with many understudied areas of control of alphaviral pathogenesis. Antibody and T cell responses are likely both necessary to control infection, but can also play a role in promoting pathogenesis. Likewise, innate immune cells can serve as reservoirs for initial viral replication but are important in promoting adaptive immune responses. It is also probable that VEEV, EEEV, and WEEV may require different types of immune responses for control of pathogenesis, as their replication targets and disease severity are different. A balanced innate and adaptive response, correctly tuned for timing and magnitude, is likely required for protection against these viruses (Figure 1).

Figure 1. Innate and adaptive immune responses can contribute to control of encephalitic alphavirus infection or can augment pathogenesis depending on a number of factors. Created with Biorender.com.

4. Host pathways impacting alphavirus replication

Viruses are intracellular obligate parasites and as such, require a large number of host proteins and processes to facilitate their replication. Conversely, innate immune response factors function to suppress viral replication. Insights into potential therapeutic targets can be gained by examining host factors that are both proviral and antiviral in nature.

4.1. Viral entry, fusion, and uncoating

The alphavirus glycoproteins E2 and E1 mediate entry and fusion, respectively. Alphaviruses enter host cells through receptor-mediated endocytosis within clathrin coated vesicles [58]. Once inside the vesicles, pH dependent fusion occurs, which is mediated by E1, a type II membrane fusion protein. E2 and E1 proteins form trimers consisting of three heterodimers in the virion membrane [59–62]. Previous work with the prototypic alphavirus Sindbis virus (SINV) indicated that E2 utilizes Dendritic Cell-Specific Intercellular adhesion molecule-3-Grabbing Non-integrin (DC-SIGN) [63], liver/lymph node-specific intracellular adhesion molecules-3 grabbing non-integrin (L-SIGN) [63], and Heparin sulfate [64,65] as receptors or attachment factors. However, the receptors for the encephalitis alphaviruses remained elusive until recently. The use of a CRISPR–Cas9 screening approach identified the very low-density lipoprotein receptor (VLDLR) as an entry receptor for multiple alphaviruses, including EEEV [66]. The closely related apolipoprotein E receptor 2 (ApoER2) also facilitated EEEV entry [66]. Interestingly, the invertebrate receptor orthologues of these receptors were also capable of mediating EEEV entry, suggesting that this family of receptors may enable alphavirus infection of multiple hosts [66]. VEEV was not able to use VLDLR or ApoER2 for entry [66]. However, another member of the low-density lipoprotein receptor (LDLR) family of proteins, the low-density lipoprotein receptor class A domain-containing 3 (LDLRAD3), was identified as an entry receptor for VEEV [67]. LDLRAD3 was not capable of supporting entry of EEEV, WEEV, SINV or Mayaro virus (MAYV), an alphavirus known to cause arthralgia. Ldlrad3-D1-Fc protein was able to block infection of VEEV in vitro as well as prevent mortality of C57BL/6 J and CD-1 mice infected with virulent strains of VEEV, TrD, and ZPC738, respectively [67]. Mice with Ldlrad3 gene deletions were also resistant to VEEV induced lethality, further supporting the role of LDLRAD3 as an entry receptor for VEEV.

Interferon-induced transmembrane protein 3 (IFITM3) is an endosomal protein that restricts multiple viruses including VEEV [68,69]. Alphavirus pH-dependent fusion was inhibited by IFITM3, indicating that IFITM3 is an antiviral restriction factor. To further support this notion, Ifitm3−/− C57BL/6 mice showed greater mortality that WT C57BL/6 mice infected with VEEV TC-83 A3G [68]. In contrast, a number of other endosomal proteins, including epidermal growth factor receptor substrate 15 (Eps15), Rab5, and Rab7, have been shown to facilitate VEEV endocytosis [70,71].

Fusion of the virion membrane with the endosomal membrane allows release of the nucleocapsid into the cytoplasm. However, there is limited information regarding host factors needed for viral uncoating. The 60S ribosomal RNA interacts with the alphavirus capsid protein to enable the release of viral RNA and the immediate translation of the non-structural proteins 1–4 (nsP1-4) [58]. VEEV capsid is K48 ubiquitinated and a model was suggested wherein ubiquitination and proteasomal degradation of VEEV capsid would enable viral uncoating [72]; however, this model has yet to be experimentally verified. Host factors involved in alphavirus entry, fusion, or uncoating are summarized in Table 3.

Table 3. Known host factors involved in alphavirus entry, fusion, or uncoating.

General Function	Host Protein	Inhibitor	Virus	Cell Type	In Vivo Model	Reference
Receptors	VLDLR	N/A	EEEV, SFV, SINV	HEK 293 T	N/A	[66]
	ApoER2	N/A	EEEV	HEK 293 T	N/A	[66]
	LDLRAD3	Ldlrad3-D1-Fc	VEEV	Neuro-2a	CD1 or C57BL/6 J mice infected s.c. with VEEV TrD or ZPC738	[67]
Endocytosis	Eps15	N/A	VEEV	N/A	N/A	[70]
	Rab5	N/A	VEEV, SFV	N/A	N/A	[70,71]
	Rab7	N/A	VEEV	N/A	N/A	[70,71]
	IFITM3	N/A	VEEV	WT or Ifitm3−/− MEFs	WT C57BL/6 or Ifitm3−/− C57BL/6 mice	[68]
Ubiquitin proteasome system	unknown	Bortezomib	VEEV, EEEV, and WEEV	U87MG, Vero	ND	[72]

*N/A = not applicable, **ND = not determine, s.c. = subcutaneous

4.2. Viral translation and genomic replication

Stress granules, which contain translation initiation factors, non-translated RNAs, and ribosomal components, are formed in eukaryotic cells in response to stress, such as viral infection [73]. Their formation coincides with translational inhibition. During alphavirus infections, stress granule formation is suppressed by nsP3; nsP3 interacts with stress granule proteins to repurpose them to assist with viral replication [74,75]. nsP3 from the Old World alphaviruses, SINV and chikungunya virus (CHIKV), interact with GTPase-activating protein (SH3 domain)-binding protein (G3BP) family members, whereas VEEV nsP3 was found in complex with Fragile X syndrome (FXR) family members [76,77]. nsP3 interaction with G3BP and FXR proteins is mediated through the nsP3 hypervariable domains (HVDs) and these complexes assist in the assembly of viral replication complexes (vRCs). Interestingly, EEEV nsP3 interacts with both G3BP and FXR family members and utilizes both for efficient viral replication [78]. Out of 20 alphaviruses tested (including WEEV and EEEV), all nsP3 proteins except salmonid alphavirus (SAV), Tai Forest virus (TAFV) and VEEV colocalize with G3BP [79]. This indicates that among the encephalitic equine alphaviruses, VEEV is unique in its lack of G3BP binding.

VEEV nsP3 interacts with DEAD-Box Helicase 1 (DDX1) and DDX3 [80], which are RNA helicases involved in translation and associated with stress granules [81,82]. Loss of DDX1 and DDX3 through siRNA transfection significantly reduced VEEV TC-83 and VEEV TrD replication in U87MG astrocytes [80]. Likewise, treatment of U87MG cells with the DDX3 small molecule inhibitor, RK-33, decreased VEEV infectious titers. VEEV nsP3 was also found in complex with components of the cellular translation machinery including eIF4G, eIF4A, and poly A binding protein (PABP), which are known interacting partners of DDX3, and treatment with the DDX3 inhibitor RK-33 disrupted the interaction of nsP3 with eukaryotic initiation factor 4G (eIF4G), eukaryotic initiation factor 4A (eIF4A) and PABP [80]. These data suggest that the interaction of nsP3 with DDX3 could be facilitating recruitment of translational initiation factors to assist in alphavirus translation.

A recent VEEV nsP3 interactome study identified 160 putative host interacting proteins, including eukaryotic initiation factor 2 subunit 2 (eIF2S2) and transcription factor AP-2 alpha (TFAP2A) which were validated for their importance in VEEV production through siRNA studies [83]. eIF2S2 was found to facilitate VEEV genomic RNA translation, but not translation of the subgenomic RNA [83]. Citalopram HBr and Z-VEID-FMK, inhibitors of TFAP2A, and Tomatidine, a small molecule inhibitor of eIF2S2, decreased VEEV production by >10 fold. Citalopram HBr, Z-VEID-FMK, and Tomatidine also suppressed EEEV replication.

Protein kinase R-like endoplasmic reticulum kinase (PERK) was found to be important for translation of alphavirus nsPs from the genomic RNA [84]. PERK is activated via the unfolded protein response (UPR) and phosphorylates eIF2α, which subsequently suppresses host translation [85]. Loss of PERK through siRNA significantly impacted VEEV and EEEV production in primary human astrocytes, pericytes and human umbilical vein endothelial cells, but had no impact on VEEV production in transformed U87MG astrocytes or 293 T cells [84].The mechanism by which PERK facilitates viral nsP translation is not yet known, but these results indicate a noncanonical role of PERK in alphavirus production via supporting viral translation. It is important to note that alphavirus subgenomic translation is resistant to eIF2α translational inhibition [86,87], but translation from the genomic RNA is not, being regulated similar to host translation [86,88].

There is little known about host factors needed for subgenomic RNA translation. However, a recent study found that Src kinase facilitates subgenomic RNA translation [89]. Inhibitors of Src family kinases, namely Torin 1 and Dasatinib, decreased CHIKV infectious titers and reduced CHIKV and VEEV structural protein synthesis, with no impact on viral RNA levels. Torin 1 and Dasatinib treatment had a broad impact on alphaviruses, reducing production of CHIKV, VEEV, o’nyong’nyong virus (ONNV), Ross River virus (RRV), and MAYV. Sorafenib, traditionally defined as a Raf kinase inhibitor, was also found to inhibit VEEV production through inhibition of subgenomic protein translation [90]. However, sorafenib’s activity was not dependent on c-Raf and b-Raf, but rather on its suppression of phosphorylation of multiple translational proteins, including eukaryotic initiation factor 4E (eIF4E), 70-kDa ribosomal protein S6 kinase (p70S6K), and ribosomal S6. siRNA studies confirmed the importance of the cap-binding protein eIF4E for VEEV replication. Sorafenib had broad alphavirus activity inhibiting replication of VEEV, EEEV, SINV, and CHIKV.

The Nuclear Factor Kappa Beta (NF-κB) pathway, and more specifically the Inhibitor of Nuclear Factor Kappa B Kinase Subunit Beta (IKKβ), plays an important role in regulating alphavirus genomic replication. NF-κB is a transcription factor that is stimulated following many viral infections and can be exploited by viruses to promote viral replication and/or alter the immune response to infection [91]. Phosphorylation of p65 at serine 536, degradation and phosphorylation of IκBα, and p65 nuclear localization, all of which are markers of NF-kB activation, were observed during early time points after VEEV infection [92]. The IKK complex is responsible for phosphorylation of IκBα and p65 (Serine 536) [93] and of the three IKK subunits: IKKα, IKKβ and IKKγ, only IKKβ underwent molecular reorganization in VEEV infected cells [92]. IKKβ is a VEEV and WEEV nsP3 interacting protein [92] and IKKβ phosphorylates VEEV nsP3 at amino acids 204/5, 142, and 134/5 [94]. VEEV nsP3 mutants with alanine substitutions at positions 204/5, 142, and 134/5 are replication deficient, due to a lack of negative strand RNA synthesis. Conversely, phospho-mimetics at these positions rescued VEEV production and negative strand RNA synthesis [94]. These results demonstrated that IKKβ phosphorylation of nsP3 is critical for negative strand viral RNA production. Small molecule inhibitors of IKKβ, BAY-11-7082, BAY-11-7085 and IKK2 compound IV, reduced VEEV infectious titers in U87-MG astrocytoma cells and BAY-11-7082 was also capable of reducing EEEV and WEEV titers [92]. Treatment of VEEV infected cells with the IKKβ inhibitor BAY-11-7082 inhibited nsP3 phosphorylation [94], indicating that the impact of BAY-11-7082 of VEEV replication is at least partially due to its impact on nsP3 phosphorylation. Treatment of VEEV infected C3H/HeN mice with BAY-11-7082 increased animal survival, reduced viremia at Day 3, and decreased viral titers in the brain at Day 7 [92], further reinforcing the importance of IKKβ for VEEV replication.

While nsP3 interacts with multiple host factors to facilitate viral translation and replication, Old World alphavirus nsP2 has a well-defined role in suppressing the antiviral response through inhibiting host translation (reviewed in [27]). VEEV nsP2, but not EEEV nsP2, is able to suppress host translation and VEEV nsP2 does so even in IFN-primed cells [95]. Table 4 summarizes host factors that affect alphavirus replication and translation.

Table 4. Known host factors involved in alphavirus translation and/or genomic replication.

General Function	Host Protein	Inhibitor	Virus	Cell Type	In Vivo Model	Reference
Translation	eIF2S2	Tomatidine	VEEV, EEEV	HMC3, U87MG, SVGp12	ND	[83]
Transcription	TFAP2A	citalopram HBr	VEEV, EEEV	HMC3, U87MG, SVGp12	ND	[83]
		Z-VEID-FMK	VEEV, EEEV	HMC3, U87MG, SVGp12	ND	[83]
Translation, Stress Granule Formation	DDX1	N/A	VEEV	U87MG	ND	[80]
Translation, Stress Granule Formation	DDX3	RK-33	VEEV	U87MG	ND	[80]
Stress Granule Formation	G3BP family members	N/A	EEEV, WEEV	BHK-21	N/A	[78,79]
Stress Granule Formation	FXR family members	N/A	EEEV, VEEV	BHK-21	N/A	[76,77]
Translation	eIF4E	Sorafenib	VEEV, EEEV, SINV, CHIKV	Vero	ND	[90]
Innate Immune Response	IKKβ	BAY-11-7082	VEEV, EEEV, WEEV	U87MG	VEEV TC-83 i.n. infection	[92,94]
		BAY-11-7085	VEEV	U87MG	ND	[92]
		IKK2 compound IV	VEEV	U87MG	ND	[92]
Cell Signaling	Src kinase	Torin 1 and Dasatinib	CHIKV, VEEV, ONNV, RRV, and MAYV	Normal human dermal fibroblasts	ND	[89]
UPR and translation	PERK	siRNA	VEEV, EEEV	Primary human astrocytes and pericytes, HUVECs	ND	[84]

*N/A = not applicable, **ND = not determined, i.n. = intranasal

4.3. Viral assembly and release

Host proteins that play a role in encephalitic alphavirus assembly and release are summarized in Table 5. Multiple host protein-interacting partners of VEEV E2 have been identified that facilitate late viral events. The ER chaperone, GRP78, was identified as an VEEV E2 protein partner through a proteomics approach. HA15, a GRP78 inhibitor, suppressed infectious viral titers without reducing viral RNA production, suggesting that GRP78 was important for viral assembly or viral protein trafficking [96].

Table 5. Known host factors involved in alphavirus viral assembly or exit.

General Function	Host Protein	Inhibitor	Virus	Cell Type	In Vivo Model	Reference
Protein folding, chaperone	GRP78	HA15	VEEV, EEEV, CHIKV, SINV	Vero	ND	[96]
Phosphatase	PP1α	1E7-03	VEEV, EEEV, WEEV, CHIKV, SINV	Vero	ND	[99]
Kinase	PKCδ	N/A*	VEEV	Vero	VEEV TC-83 or VEEV CPD infection of C3H/HeN mice	[100]
G Protein Signaling	Rac1	EHT1864	VEEV	Hela	ND	[97]
	Rac1	NSC23766	VEEV	Hela	ND	[97]
Actin Organization	Arp3	CK548	VEEV	Hela	ND	[97]
	Arp3	CK869	VEEV	Hela	ND	[97]
Cytoskeleton	Actin	Cytochalasin D	VEEV	Hela	ND	[97]
	Actin	Latrunculin A	VEEV	Hela	ND	[97]

*N/A = not applicable, **ND = not determine

The actin cytoskeleton facilitates E2 trafficking in cytopathic vacuole type II (CPV-II) vesicles from the trans-Golgi network to the plasma membrane [97]. Actin filament rearrangements are observed during late stages of VEEV infection, VEEV E2 protein interacts with actin and colocalizes with actin filaments, and inhibition of actin polymerization with the small molecule inhibitors cytochalasin D or latrunculin A suppresses infectious viral titers [97]. Furthermore, siRNA knockdown of the trafficking proteins Ras-related C3 botulinum toxin substrate 1 (Rac1), phosphatidylinositol-4-phosphate 5-kinase type-1 (PIP5K1-α), and actin related protein 3 (Arp3) reduces VEEV production. Treatment of VEEV infected cells with small molecule inhibitors of Rac1 (EHT1864 and NSC23766) and Arp3 (CK548 and CK869) decreased E2 trafficking from the Golgi to the cell surface and reduced viral infection [97]. Based on these data, a model emerged wherein Rac-1 and Arp3 dependent actin remodeling mediates E2 trafficking to the plasma membrane.

Capsid interaction with both viral RNA and the E2 protein are critical for viral assembly [98]. Phosphorylation of VEEV capsid has been shown to regulate capsid interaction with viral RNA [99,100]. VEEV capsid is phosphorylated on Thr 93, Thr 108, Thr 127, and Ser 124 [99]. Protein kinase C delta (PKCδ) and protein phosphatase 1α (PP1α) both interact with VEEV capsid and regulate its phosphorylation [99,100]. Loss of PP1α and PKCδ through siRNA suppressed VEEV replication. PP1α appears to be important for multiple alphaviruses, as treatment with the small molecule inhibitor, 1E7-03, decreased VEEV, EEEV, WEEV, CHIKV, and SINV infectious titers [101]. In contrast, loss of PKCδ through siRNA impacted VEEV production, but not EEEV and WEEV [100]. Mutation of the 4 phosphorylated residues to alanine (VEEV CPD mutant) resulted in increased viral RNA binding and attenuation of VEEV TC-83 pathogenesis in C3H/HeN mice [100], further highlighting the importance of VEEV capsid phophorylation for viral replication.

4.4. Viral inhibition of nucleocytoplasmic trafficking

A number of RNA viruses such as dengue and respiratory syncytial virus replicate almost exclusively within the cytoplasm, yet infection remains dependent on the efficient nuclear import of specific viral protein(s) in order to inhibit the host cell anti-viral response [102–104]. New World alphaviruses follow this pattern with a portion of the capsid protein population being found in the nucleus and the nuclear role of VEEV and EEEV capsid proteins being critical for pathogenesis [105]. VEEV and EEEV capsid proteins contribute to viral induced cytopathogenicity, at least partially due to their ability to shut down host cell transcription [105–107]. The transcriptional inhibition ability has been mapped to the N-terminal region of capsid, which contains a nuclear localization sequence (NLS) and a nuclear export sequence (NES) [107]. VEEV capsid binds to importin-α (Impα), Impβ1, and CRM1 to form a tetrameric complex which results in obstructing the nuclear pore, thus blocking host cell nuclear export and import [108]. siRNA mediated knockdown of Impα, Impβ1, and CRM1 altered capsid localization, confirming their critical role in modulating capsid trafficking [109]. Mutation of the NLS of VEEV capsid results in its inability to suppress host transcription and the antiviral response and prevention of VEEV virulence in NIH Swiss mice [107,110]. The NLS and NES in VEEV is highly conserved amongst VEEV, EEEV, and WEEV [108], but there is no evidence to date indicating that EEEV and WEEV inhibit nucleocytoplasmic trafficking. However, mutation of the EEEV NLS sequence resulted in delayed replication kinetics in mammalian cells, increased interferon sensitivity, and loss of virulence in NIH Swiss mice, confirming the importance of this sequence in EEEV [106].

Given the importance of VEEV and EEEV capsid for viral pathogenesis, there have been a number of studies aimed at developing therapeutics to target capsid’s nucleocytoplasmic trafficking ability (Table 6). The FDA-approved drugs mifepristone and ivermectin were found to inhibit Imp α/β-mediated import [109,111] and were able to reduce nuclear-associated VEEV capsid. Mifepristone was able to reduce virus titers of both VEEV TC-83 and VEEV TrD, whereas ivermectin was a less potent inhibitor of VEEV replication [109,112]. Mifepristone also reduced cell death induced by VEEV infection [109], which suggested that nuclear import inhibitors may be able to protect cells from apoptosis in addition to disrupting the function of essential viral proteins. Novel compound scaffolds that inhibit the VEEV capsid-Impα interaction were identified through an in silico structure-based-drug-design (SBDD) approach [113]. From an initial screen of 1.5 million compounds, followed by in silico refinement and screening for biological activity in vitro, 21 hit compounds which inhibited capsid:Impα/β1 binding with IC₅₀ values as low as 5 μM [113]. Four compounds were found to inhibit nuclear import of capsid in transfected cells. Of the four, compound 1111684 was able to reduce VEEV replication at μM concentrations, concomitant with reduced capsid nuclear accumulation in infected cells. Further, this compound was inactive against a mutant VEEV that lacks high affinity capsid:Impα/β1 interaction (TC-83-CM), supporting the mode of its antiviral action is through inhibiting capsid nuclear localization. A second compound, referred to as G281-1485, was identified through a high-throughput screening (HTS) approach using AlphaScreen technology which identifies compounds able to inhibit protein-protein interactions (e.g. capsid binding to Impα/β1) [114]. Compound G281-1485 in particular inhibited VEEV replication at low μM concentrations, while showing minimal toxicity [114]. Closely related compounds G281-1564 and G281-1481 also inhibited VEEV replication, but both were less potent inhibitors of capsid: Impα/β1 and VEEV replication. Targeting CRM1 has also been successful for inhibition of alphavirus replication in vitro. Treatment of VEEV infected cells with CRM1 inhibitors, termed selective inhibitors of nuclear export (SINE) compounds, resulted in retention of capsid in the nucleus, decreasing viral assembly and reducing infectious titers [115]. SINE compounds were also capable of inhibiting EEEV and WEEV infectious titers. It is important to note, that to date no capsid inhibitors have been tested in vivo.

Table 6. Known host factors involved in capsid nuclear import and export.

General Function	Host Protein	Inhibitor	Virus	EC50	Cell Type	In Vivo Model	Reference
Nuclear import receptor	Impα/ Impβ	Mifepristone	VEEV	19.9 µM	Vero, U87MG	ND*	[109,112]
		Ivermectin	VEEV		Vero, U87MG	ND	[109]
		1111684	VEEV	9.9 µM	Vero	ND	[113]
		G281-1564	VEEV	10.8 µM	Vero	ND	[114]
Nuclear export receptor	CRM1	KPT-185	VEEV, EEEV, WEEV	0.09 µM^#	Vero	ND	[115]
		KPT-335	VEEV, EEEV, WEEV	0.17 µM^#	Vero	ND	[115]
		KPT-350	VEEV, EEEV, WEEV	0.62 µM^#	Vero	ND	[115]

*ND = not determined, ^#EC50 was only determined for VEEV

5. Host factors and pathways identified through transcriptomic analysis

Global analysis of the transcriptome following alphavirus infection has identified multiple cellular pathways that are altered upon infection and also shed light on specific cellular factors that are utilized for more efficient viral replication. As mentioned above, there have been multiple transcriptomic studies that identified upregulation of proinflammatory cytokines and chemokines, as well as immune response factors, such as TLRs in the brains of mice infected with virulent VEEV [26,31,53,54]. Likewise, a transcriptomics study on brain samples from C3H/Ne mice infected intranasally with VEEV TC-83 found significant alterations in immune and inflammatory response pathways, including gene expression changes of factors involved in signaling between natural killer (NK) cells and antigen presenting cells [116]. These changes led the authors to deplete NK cells, which protected the C3H/Ne mice from VEEV-induced mortality. NHPs exposed to aerosolized VEEV also displayed upregulation of innate immune response genes in their brain including IFITM1, IFITM2, Mx1 and STAT1 as well as MHC class I molecules [117]. The induction of immune response factors following VEEV exposure has also been observed in patient samples. Specifically, blood samples from individuals vaccinated with live attenuated VEEV TC-83 revealed transcriptomic alterations in multiple immune pathways at Days 3 and 7 post-vaccination, including interferon response, interferon-response factors, activation of pattern recognition receptors, and engagement of the inflammasome [118]. However, it is important to note that this study examined the response to VEEV vaccination and transcriptomic studies from human VEEV pathogenic cases have not been performed.

Microarray analysis has been used to identify multiple interferon stimulated genes (ISGs) induced in cells infected with VEEV, including PARP12, Ifi27, Trim30, and Bst2 [119]. These studies were performed with a VEEV capsid mutant, which is capable of persistently infecting cell deficient in interferon signaling [119,120]. PARP12 was capable of suppressing alphavirus replication, including VEEV, SINV, and CHIKV, as well as replication of a broad range of viruses such as Rift Valley fever virus (RVFV), vesicular stomatitis virus (VSV), and encephalomyocarditis virus (EMCV).

RNA sequencing (RNAseq) of VEEV TrD infected U87MG astrocytoma cells identified alterations in the type 1 interferon response pathway and the unfolded protein response (UPR) [121], with the UPR being induced at late time points after infection. Early growth response 1 (EGR1), a stress induced transcription factor, was identified as a link between these pathways and is induced in VEEV infected cells via the PERK and ERK pathways [121,122]. Loss of EGR1 had minimal impacts on VEEV replication, but reduced cell death of VEEV infected cells [121,122]. Single-cell RNAseq (sc-RNAseq) analysis identified subpopulations of U87MG cells that produced exceptionally high amounts of VEEV viral RNA and corresponding host factors that were correlated with high vs low levels of viral RNA production [123]. siRNA depletion of CXCL3, ATF3, TNFAIP3, and CXCL2, all factors that were positively correlated with viral RNA levels, reduced VEEV replication. Interestingly ATF3 is involved in the UPR and previously found in transcriptomic analysis of VEEV infected U87MG cells [121]. Conversely, loss of TAF7, SURF4, and RAB1A, which were negatively correlated with viral RNA levels, resulted in enhancement of VEEV replication.

Finally, there has only been one study examining changes in miRNA gene expression following encephalitic New World alphavirus infection. Analysis of brain samples from VEEV V3000 infected CD-1 mice revealed 32 differentially expressed miRNAs at 2 days post-infection and 36 at 3 days post-infection. Eleven of these miRNAs, Mir-136, 203, 216b, 297b-5p, 320, 331–3p, 339–5p, 449a, 501–3p, 690 and 720, were correlated with nervous system development and function [124].

6. Antiviral candidates showing efficacy in animal models

Development of small molecule compounds for encephalitic alphavirus infections has garnered recent attention [reviewed in [27]], but only a few candidates to date have shown efficacy in animal models (Table 7). CID15997213, a quinazolinone compound that likely targets the N-terminal domain VEEV nsP2, has antiviral effects for VEEV with EC₅₀ values ranging 0.36 to 1.3 μM, depending on the strain (Table 7) [125]. It also had moderate in vitro activity against WEEV (EC₅₀ of 10 μM), but no activity against EEEV. In vivo studies demonstrated no toxicity and complete protection of C3H/HeN mice upon VEEV TC-83 infection [126]. Medicinal chemistry efforts to improve CID15997213 led to the development of ML-336, a benzamidine compound, which has shown promising in vitro and in vivo activity against VEEV [126–129]. ML-336 inhibits VEEV in Vero cells with an EC₅₀ of 32 nM. ML-336 inhibits viral RNA synthesis with resistant strains developing mutations in nsP2 and nsP4 [127,129]. Concerns about ML-336ʹs solubility and limited stability led researchers to investigate the use of lipid-coated mesoporous silica nanoparticles (LC-MSN) as delivery vehicles for ML-336 in vivo [130]. Neither free ML-336 or LC-MSN loaded with ML-336 (at 1 mg/kg delivered via intraperitoneal (i.p.) injection twice a day) were capable of protecting C3H/HeN mice from VEEV TC-83 infection; however, LC-MSN loaded with ML-336, but not free ML-336, reduced viral titers in the brain by 10 fold. It should be noted that ML-336 was previously shown to be capable of protecting VEEV TC-83-infected C3H/HeN mice when given at a dose of 5 mg/kg via intraperitoneal injection twice a day [126]. ML-336 was further refined to create three derivatives, BDGR-4, BDGR-69 and BDGR-70, which displayed anti-VEEV activity in the nanomolar range [128]. BDGR-4 was considered an especially promising candidate, as it had comparable in vitro activity, but displayed increased solubility and microsomal stability as compared to ML-336. ML-336 and BDGR-4 dosed at 12.5 mg/kg twice a day via i.p. provided 90–100% protection from a lethal VEEV TrD subcutaneous challenge in BALB/c mice [128]. BDGR-69 and BDGR-70 protected 50 and 100% of BALB/C mice, respectively, from VEEV infection. BDGR-4 was also effective when given up to 48 hours after VEEV infection and protected 90% of C57BL/6 mice from subcutaneous challenge with EEEV.

Table 7. Antivirals with in vivo activity against alphaviruses.

Antiviral Compound	Target/MOA	EC50	Cell Type tested in vitro	In vivo Efficacy	References
CID15997213	nsP2	1.3 µM (VEEV TC-83) 1.01 µM (VEEV V3526) 0.36 µM (VEEV TrD) 10 µM (WEEV)	Vero 76	100% protection of C3H/HeN mice from i.n. VEEV TC-83 infection	[126]
ML-336	inhibits viral RNA synthesis, nsP2 and/or nsP4	32 nM (VEEV TC-83)	Vero 76	80% protection from i.n. VEEV TC-83 infection 90–100% protection from s.c. VEEV TrD infection	[126–129]
BDGR-4	ND*	47 nM (VEEV TC-83)	Vero 76	100% protection of C3H/HeN mice from i.n. VEEV TC-83 infection 100% protection of BALB/c mice from s.c. VEEV TrD infection 90% protection of C57BL/6 mice from s.c. EEEV infection	[128]
BDGR-69	ND	38 nM (VEEV TC-83)	Vero 76	50% protection of C3H/HeN mice from i.n. VEEV TC-83 infection	[128]
BDGR-70	ND	25 nM (VEEV TC-83)	Vero 76	100% protection of C3H/HeN mice from i.n. VEEV TC-83 infection	[128]
Favipiravir	nucleoside analog	312 µM (WEEV)	Vero	50% protection of C57BL/6 mice from WEEV infection	[132]
EIDD-1931	nucleoside analog, induces mutations in the viral genome	0.426 µM (VEEV TC-83)	Vero	90–100% protection of CD-1 mice from i.n. VEEV TrD infection	[133,134]
BAY-11-7082	IKKβ	ND	NA^#	Reduced viremia and viral load in the brain of VEEV TC-83 infected C3H/HeN mice	[92]
Compound 4210	MyD88	10.23 µM (VEEV TrD)	A549	90–100% protection of C3H/HeN mice from i.n. VEEV TC-83 infection	[142]
Acriflavine	Ago2	0.20 µM (VEEV TC-83)	U87MG	60% protection of C3H/HeN mice from i.n. VEEV TC-83 infection	[141]
Bioder	GSK-3β	~0.5 µM (VEEV TC-83)	U87MG	50% protection of C3H/HeN mice from i.n. VEEV TC-83 infection	[140]

*ND = not determine, ^#NA = not applicable

Favipiravir (originally known as T-705) is a nucleoside analog with activity against viral RNA-dependent RNA polymerases that inhibits replication of viruses from multiple families [131]. Despite having a relatively high EC₅₀ of 312 μM against WEEV in vitro, WEEV infected C57BL/6 mice treated with 400 mg/kg/day favipiravir by oral gavage had a significant improvement in mean time to death and survival [132]. Perhaps the most advanced antiviral candidate is the ribonucleoside analog EIDD-1931, which impacts VEEV through inducing extensive mutations in the VEEV genome [133]. It is orally available and protects 90–100% of CD-1 mice from an intranasal challenge of VEEV TrD even when administered 24 h post-exposure [134]. It is interesting to note that EIDD-1931 has broad spectrum activity against multiple RNA viruses including Ebola virus, SARS-CoV-2, and influenza virus [135–138]. In addition, molnupiravir (EIDD-2801), an orally available prodrug of EIDD-1931, has FDA emergency use authorization for the treatment of COVID-19 [139].

Finally, there are a few host-based small molecule inhibitors that have demonstrated in vivo activity against VEEV, including BAY-11-7082 [92] (see the ‘Viral translation and genomic replication’ section above), BIOder [140], acriflavine [141], and compound 4210 [142]. BIOder, a GSK-3β inhibitor, was effective at suppressing VEEV replication in vitro, protecting neuronal cells from viral induced cell death, and partially protecting C3H/HeN mice from VEEV TC-83 i.n. infection [140]. GSK-3β is a master regulator of inflammation, inducing proinflammatory factors and cell migration and also impacting apoptosis [143]. The ability of BIOder to impact VEEV pathogenesis could be partially due to its regulation of apoptotic factors, as BIOder treatment altered BID (pro-apoptotic) and survivin (anti-apoptotic) gene expression.

Acriflavine suppressed VEEV TC-83, VEEV TrD, EEEV, WEEV, WNV, but not VSV replication. Acriflavine is typically used as an antiseptic, but more recently was shown to inhibit Ago2, a protein involved in miRNA processing [144]. The importance of Ago2 for alphavirus replication was further shown through MEFs lacking Ago2 having decreased VEEV TC-83 viral RNA levels, structural protein production, and infectious titers [141]. In addition, multiple inhibitors of the miRNA machinery, including aurintricarboxylic acid (ATA), oxidopamine hydrochloride (OXD) and suramin (SUR), acriflavine, and poly L lysine, suppressed VEEV TC-83 replication, with acriflavine being the most potent. Acriflavine partially protected C3H/HeN mice from VEEV TC-83 infection, but did not have a protective effect on BALB/c mice infected with VEEV TrD, indicating that inhibition of Ago2 alone is not sufficient to protect mice from a fully virulent VEEV infection.

Compound 4210, a MyD88 inhibitor, was shown to suppress replication of multiple viruses in vitro, including VEEV, EEEV, ebolavirus (EBOV), RVFV, Lassa virus and dengue virus [142]. MyD88 is an adapter protein in the Toll like receptor (TLR) signaling cascades and can act as both a positive and negative regulator of the innate immune response [145]. Loss of MyD88 increased IFNβ production and inhibition through compound 4210 treatment increased IRF3 phosphorylation [142]. Mice (C3H/HeN or BALB/C) treated with compound 4210 displayed increased survival from both VEEV and EBOV infection, respectively.

7. Conclusion

The encephalitic alphaviruses VEEV, EEEV, and WEEV are important human and agricultural pathogens for which there are no therapeutics to alleviate disease. Recent advances in understanding mechanisms of pathogenesis offers potential cellular and viral targets for directed small molecule development. There appears to be conserved but also divergent mechanisms of pathogenesis between VEEV, EEEV, WEEV, which offers opportunities for the generation of broadly-acting or specific drugs.

8. Expert opinion

The advances in identification of host responses to VEEV, EEEV, and WEEV infection has unveiled multiple pathways that could be targeted using existing or novel drugs to reduce viral pathogenesis. These pathways include innate immune responses, translational regulation, and control of viral entry and egress. Several small molecules that have shown protective responses in small animal studies and should be further assessed in higher animals for preclinical development. Additionally, the use of repurposed drugs with in vivo activity could simplify the transition into clinical trials.

More studies have been conducted for VEEV compared to EEEV and WEEV, in part due to the larger outbreaks associated with VEEV. However, EEEV and WEEV cause higher rates of morbidity and mortality, and all three viruses are devastating in equids. Additional work evaluating similarities and differences in pathogenesis of these three viruses will enable development of broadly-protective drugs or will reveal whether this approach is feasible. Furthermore, given the recent expansion of small molecules used against multiple viruses (such as remdesivir), more research should be done testing existing broadly active small molecules against alphaviruses.

Research with VEEV has been facilitated by being able to work safely at BSL-2 with the live attenuated strain of VEEV, TC-83. However, no such attenuated strains are available for EEEV and WEEV, restricting EEEV and WEEV studies to BSL-3 and limiting the labs that can safely work with these pathogens. Work with EEEV is further restricted due to it being a select agent pathogen with corresponding biosecurity regulations. Therefore, the generation of live attenuated strains of EEEV and WEEV that could serve as BSL-2 models for these viruses would be a valuable resource to the scientific community, although it should be noted that validation of BSL-2 viruses with wild-type virus studies is necessary to assess the scientific validity of attenuated strains.

These viruses are Category B pathogens, can be spread rapidly by mosquitos, and have no approved therapeutics. Therefore, future research is critical to a) advance existing therapeutics, b) increase the pipeline for new and more broadly-acting drugs, and c) better understand the molecular pathogenesis of these viruses.

Finally, while host-directed antivirals are the focus of the current review, the impact of direct acting antivirals cannot be discounted. This is evidenced by multiple direct acting alphavirus antivirals showing efficacy in animal models and by the majority of FDA approved antivirals being direct acting [146,147]. Rather, the alphavirus community would benefit from continued development of both host-directed and direct acting antivirals with the goal of future combination therapies which can target multiple aspects of viral replication and prevent the development of viral resistance.

Article highlights

• Encephalitic alphaviruses are important livestock and human pathogens

• No approved therapeutics are available for use

• Therapeutics with in vivo activity have been identified

• Multiple viral and host targets are candidates for therapeutic development

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

KKH and SBB have (1) equally contributed to the conception and design of the review article and interpreting the relevant literature, and (2) been equally involved in writing the review article and revising it for intellectual content.

Additional information

Funding

This manuscript was funded by Defense Threat Reduction Agency, Department of Defense HDTRA1-21-1-0008 (KKH) and HDTRA1-20-1-0015 (SBB).