Treatment of highly virulent mammarenavirus infections—Status quo and future directions

Abstract

Introduction:

Mammarenaviruses are negative-sense bisegmented enveloped RNA viruses that are endemic in Africa, the Americas, and Europe. Several are highly virulent, causing acute human diseases associated with high case fatality rates, and are considered to be significant with respect to public health impact or bioterrorism threat.

Areas covered:

This review summarizes the status quo of treatment development, starting with drugs that are in advanced stages of evaluation in early clinical trials, followed by promising candidate medical countermeasures emerging from bench analyses and investigational animal research.

Expert opinion:

Specific therapeutic treatments for diseases caused by mammarenaviruses remain limited to the off-label use of ribavirin and transfusion of convalescent sera. Progress to identify novel candidate medical countermeasures against mammarenavirus infection has been slow in part because of the biosafety and biosecurity requirements. However, novel methodologies and tools have enabled increasingly efficient high-throughput molecular screens of regulatory-agency-approved small-molecule drugs and led to identification of several compounds that could be repurposed for treatment of infection with several mammarenaviruses. Unfortunately, most of them have not yet been evaluated in vivo. The most promising treatment under development is a monoclonal antibody cocktail that is protective against multiple lineages of Lassa virus in nonhuman primate disease models.

1. Introduction

Arenaviridae, a family of order Bunyavirales, comprises a wide variety of segmented ambisense animal RNA viruses that produce enveloped spherical to pleomorphic (monopartite) particles [1]. Arenavirids are classified into five genera: Antennavirus, Hartmanivirus, Innmovirus, Mammarenavirus, and Reptarenavirus [2]. Antennaviruses have been discovered via metagenomic sequencing in actinopterygian fish [3–5]; hartmaniviruses and reptarenaviruses were discovered and isolated from captive and wild healthy and diseased boid and pythonid snakes [6–11]; and inmoviruses were discovered in river sediment samples [12] and are likely of actinopterygian and chondrichthyan fish origin [5, 13, 14]. Mammarenaviruses subclinically infect mammals, in particular phyllostomid bats [15–17] and possibly some of their ixodid ectoparasites [18], eulipotyphlans [19], lagomorphs [20, 21], and dipodid and muroid rodents [22–24]. Additional unclassified mammal viruses populate a sixth genus-rank clade that awaits taxonomic placement [15].

Nine arenavirids are known to be pathogenic for humans (Table 1). All human-infecting arenavirids are muroid-borne mammarenaviruses that are maintained in cricetid rodents (in South America) or in murid rodents (in Africa) in areas of disease endemicity. Zoonotic transmission of these viruses primarily occurs through contact with contaminated rodent excreta or secreta (e.g., inhaled aerosolized dried urine and feces particles produced by sweeping or consumption of food contaminated with rodent excrements) and infected rodents or their tissues (e.g., emptying of traps, aerosolization of whole rodents in harvest machinery, and preparation of rodents as food) [25, 26]. Thus, the natural geographic distribution, migration patterns, and population size of mammarenavirus-specific host rodents determine the endemicity of mammarenavirus diseases [25–28]. Person-to-person transmission is rare but may occur through contact with bodily secretions of infected individual or fomites, in particular in health care settings (e.g., aerosolization of bodily secretions during cardiopulmonary resuscitation, intubation, or sample centrifugation) [29, 30]. Sexual transmission of mammarenaviruses has not yet been proven but is not unlikely at least as an occasional occurrence [29, 31–33].

Table 1.

Mammarenaviruses pathogenic for humans (updated and modified from [208])

Virus (abbreviation) [2]	Endemicity [25, 26, 30, 209]	Natural reservoir host(s) [25, 26, 30, 209]	Human disease name (abbreviation; ICD-11 code [210])	Significance [48, 211–215]
South American-clade mammarenaviruses (“Tacaribe serocomplex”)
Chapare virus (CHAPV)	Bolivia	Unidentified cricetids	Other specified arenavirus disease (1D61.Y), informally also “Chapare hemorrhagic fever (CHHF)”	Biosafety: U.S. Risk Group 4 (BSL-4) agent Biosecurity: Australia List agent U.S. CDC Category A Bioterrorism Agent U.S. Select Agent Research priority: NIAID Category A Priority Pathogen
Flexal virus (FLEV)	Brazil	Unidentified cricetids	Other specified arenavirus disease (1D61.Y)	Biosafety: U.S. Risk Group 3 (BSL-3) agent
Guanarito virus (GTOV)	Venezuela	Predominantly cricetid short-tailed zygodonts (Zygodontomys brevicauda J.A. Allen & Chapman, 1893) but also cricetid Alston’s cotton rats (Sigmodon alstoni Thomas, 1881)	Venezuelan hemorrhagic fever (VeHF; 1D61.3)	Biosafety: U.S. Risk Group 4 (BSL-4) agent Biosecurity: Australia List agent U.S. CDC Category A Bioterrorism Agent U.S. Select Agent Research priority: NIAID Category A Priority Pathogen
Junín virus (JUNV)	Argentina	Predominantly cricetid drylands lauchas (Calomys musculinus (Thomas, 1913)) but also cricetid Azara’s akodonts (Akodon azarae J. Fischer, 1829) little lauchas (Calomys laucha Fischer, 1814)	Argentinian hemorrhagic fever (AHF; 1D61.0)	Biosafety: U.S. Risk Group 4 (BSL-4) agent Biosecurity: Australia List agent U.S. CDC Category A Bioterrorism Agent U.S. Select Agent Research priority: NIAID Category A Priority Pathogen
Machupo virus (MACV)	Bolivia	Cricetid big lauchas (Calomys callosus Rengger, 1830)	Bolivian hemorrhagic fever (BHF; 1D61.1)	Biosafety: U.S. Risk Group 4 (BSL-4) agent Biosecurity: Australia List agent U.S. CDC Category A Bioterrorism Agent U.S. Select Agent Research priority: NIAID Category A Priority Pathogen
Sabiá virus (SBAV)	Brazil	Unidentified cricetids	Other specified arenavirus disease (1D61.Y), informally also “Brazilian hemorrhagic fever (BzHF)”	Biosafety: U.S. Risk Group 4 (BSL-4) agent Biosecurity: Australia List agent U.S. CDC Category A Bioterrorism Agent U.S. Select Agent
African-clade mammarenaviruses (“Lassa–lymphocytic choriomeningitis serocomplex”)
Lassa virus (LASV)	Western Africa (Benin, Burkina Faso, Côte d’Ivoire, Guinea, Ghana, Liberia, Mali, Nigeria, Sierra Leone, Togo)	Predominanly murid Natal mastomys (Mastomys natalensis Smith, 1834) but also reddish-white mastomys (Mastomys erythroleucus (Temminck, 1853)), African hylomuscus (Hylomyscus pamfi Nicolas, Olayemi, Wendelen & Colyn, 2010), and Baoule’s mice (Mus baoulei (Vermeiren & Verheyen, 1980))	Lassa fever (LF; 1D61.2)	Biosafety: U.S. Risk Group 4 (BSL-4) agent Biosecurity: Australia List agent U.S. CDC Category A Bioterrorism Agent U.S. Select Agent Research priority: NIAID Category A Priority Pathogen WHO R&D Blueprint priority pathogen
Lujo virus (LUJV)	Zambia	unknown	Other specified arenavirus disease (1D61.Y), informally also “Lujo hemorrhagic fever”	Biosafety: U.S. Risk Group 4 (BSL-4) agent Biosecurity: Australia List agent U.S. CDC Category A Bioterrorism Agent U.S. Select Agent Research priority: NIAID Category A Priority Pathogen
lymphocytic choriomeningitis virus (LCMV)	worldwide	Predominantly murid house mice (Mus musculus Linnaeus, 1758), but also murid long-tailed field mice (Apodemus sylvaticus (Linnaeus, 1758)), soft-furred mice (Praomys spp.), and golden hamsters (Mesocricetus auratus Waterhouse, 1839)	Lymphocytic choriomeningitis (LCM; 1C8F), informally also “aseptic meningitis”	Biosafety: U.S. Risk Group 2 (BSL-2) agent Biosecurity: Australia List agent

BSL-4, biosafety level 4; CDC, Centers for Disease Control and Prevention; ICD-11, International Classification of Diseases 11th Revision; NIAID, National Institute of Allergy and Infectious Diseases; R&D, research and development; WHO, World Health Organization

As Table 1 clarifies, two of the human-pathogenic mammarenaviruses, lymphocytic choriomeningitis virus (LCMV) and Flexal virus (FLEV), are neither considered Select Agents nor Priority Pathogens and hence will not be discussed here. Of the remaining seven viruses, some are relatively rarely encountered. Chapare virus (CHAPV) [29, 34], Guanarito virus (GTOV) [35, 36], Lujo virus (LUJV) [30], and Sabiá virus (SBAV) [37–39] have caused very limited outbreaks, with one individual to dozens of cases. However, Junín virus (JUNV) [40, 41] (several hundred thousand infections over the last 60 years) and Machupo virus (MACV) [42–45] (>1,000 cases) have caused more extensive focal outbreaks with extremely high case fatality rates (sometimes >50%). Based on the low number of recorded cases but high virulence, these six viruses are primarily of bioterrorism/biowarfare concern. However, more clearly transmissible Lassa virus (LASV) causes hundreds of thousands of often subclinical infections annually over broad geographic areas, including a significant burden of disease associated with high case fatality, and is therefore primarily of public health concern [46–48].

1.1. Mammarenaviruses other than Lassa virus

The clinical presentation of human CHAPV [29, 34, 49], GTOV [35, 50, 51], JUNV [52–59], LUJV [60], MACV [42, 44, 61], and SBAV [38, 62] infections is similar: After an incubation period of approximately 1–2 weeks, patients develop influenza-like symptoms/signs, such as fever, headaches, myalgia and general malaise, and anorexia. A few days later, increasingly severe clinical manifestations involving multiple organ systems appear; in particular, early symptoms/signs are associated with the gastrointestinal (nausea with vomiting, constipation, or mild diarrhea), central nervous (photophobia with retro-orbital pain, disorientation), or coagulation (conjunctival injection, petechiae, or flushing of the head and upper torso) systems. In the second week after system onset, 20‒30% of patients develop severe neurologic signs (convulsions, seizures, tremors, delirium, coma) and hemorrhagic signs (mucosal hemorrhages ecchymoses) that are associated with fatal outcomes. Survivors convalesce over several weeks, often reporting sequelae (such as, alopecia, dizziness, fatigue, and hearing loss) that are generally poorly characterized [63–65].

1.2. Lassa virus

In contrast to CHAPV, GTOV, JUNV, LUJV, MACV, and SBAV infections, LASV causes predominantly asymptomatic infection or mild febrile disease. However, approximately 20% of patients develop acute disease that is generally reminiscent of the diseases caused by the other pathogenic mammarenaviruses; after a 1‒3-week incubation period, patients develop sudden and progressive fever and general malaise and then, within a few days, a wide array of clinical manifestations, including arthralgia/myalgia, headache (sometimes with dizziness or tinnitus), and gastrointestinal involvement (abdominal pain, nausea, diarrhea, and/or constipation). However, the clinical presentation also variably includes cardiac (chest pain, tachycardia), hepatic (hepatomegaly with jaundice), lymphatic (lymphadenopathy, splenomegaly), ocular (conjunctivitis), respiratory (cough, pharyngitis, rales, rhonchi, wheezing), and other manifestations. Patients either recover or develop progressive multi-system dysfunction, including acute kidney injury, pulmonary (pneumonia, pleural effusion, acute lung injury), cardiac (pericardial effusion, myocarditis), or hematologic dysfunction, hemorrhagic signs (ecchymoses, epistaxis, hematemesis, hematuria, hemoptysis, melena, petechia), and/or neurologic manifestations (confusion, convulsions, encephalopathy, tremors, coma) that often precede shock and death. LASV survivors may recover, sometimes over long periods of time, without sequelae, but approximately 13–30% of them suffer from permanent unilateral or bilateral sensorineural deafness, alopecia, and/or pericarditis; in general, post-acute clinical sequelae are poorly characterized. [66–74].

2. Mammarenavirus biology

2.1. Virions and genomes

Mammarenaviruses produce enveloped pleomorphic (50–200) virions containing evolutionary negative-sense (functionally ambisense) bisegmented RNA genomes, totaling about 10.5 kb. The virion surface contains club-shaped peplomers that are 8–10 nm in length and consist of clearly distinguishable head (GP1) and stalk (GP2) domains. The two internal ribonucleoprotein (RNP) complexes consist of the small (S) and large (L) genomic RNA segments that are independently encapsidated by nucleoproteins (NPs; 60–68 kDa) [75–77].

Both segments (which may be present in virions in multiple copies) each encode two structural proteins from open reading frames (ORFs) arranged in ambisense orientation, separated by highly structured intergenic regions (IGRs), and flanked by termini that include inverted complementary sequences containing highly conserved transcription and replication initiation signals. The complementary sequences force the segments into not covalently closed panhandle-like structures [78–80].

2.2. Proteins

The S genomic fragment encodes NP and the glycoprotein complex (GP) precursor (GPC; 70–80 kDa, whereas the L segment encodes a bunyaviral-typical L protein (250–450 kDa), including an RNA-directed RNA polymerase (RdRp) domain and a zinc-binding protein (Z; 10–14 kDa). NP, the most abundant mammarenavirus protein in infected cells, oligomerizes and encapsidates both genomic and antigenomic segments, functions as an exoribonuclease, and is an interferon antagonist [75, 81–83]. GP is heterotrimeric and consists of a stable signal peptide (SSP), a receptor-binding subunit (GP1), and a class I fusion membrane fusion subunit (GP2)—all generated via proteolytic cleavage of GPC. GP mediates the adsorption of the virion to the host cell and subsequent virion entry and, via the selectivity of GP for particular cellular molecules, thereby determines virus cell and tissue tropism [84–87]. The L protein mediates both transcription and replication via its RdRp domain as well as cellular mRNA cap-snatching for virus subgenomic RNA capping [88, 89]. Like NP, Z is an interferon antagonist that functions as a matrix protein for polymerization at membranes and thereby facilitates virion assembly and budding [89–92].

2.3. Lifecycle

Figure 1 depicts a simplified lifecycle of mammarenaviruses. Mammarenaviruses enter cells via adsorption to cell-surface attachment factors, which vary among viruses [93]. At the host–cell surface, LASV attaches to matriglycan, a unique disaccharide modification of dystroglycan 1 (DAG1; previously often called α-dystroglycan and abbreviated α-DG) [94, 95]. LUJV attaches to neuropilin 2 (NRP2) [96], whereas CHAPV, GTOV, JUNV, MACV, and SBAV engage transferrin receptor 1 (TFRC; previously often abbreviated TfR1) [97–99]. Attachment is followed by endocytosis. The increasingly acidic environment in the endosome triggers pH-dependent fusion, resulting in the release of the S and L RNPs into the cytoplasm, where viral RNA genome replication and gene transcription take place. In the case of LASV and LUJV, engagement of intracellular/endosomal proteins, lysosomal-associated membrane protein 1 (LAMP1) and the CD63 molecule (CD63), respectively, is necessary to progress to fusion [96, 100]. Transcription of NP- and L-encoding subgenomic RNAs, which lack poly(A) tails, is initiated by the RNP complexes at promoters located in the 3’ termini of the genomic RNA segments and terminated by structural motifs within the IGRs. The RNPs switch from transcription to replication, generating full-length antigenome RNAs that then serve as templates for transcription of the GPC- and Z-encoding ORFs. In addition, progeny genomes are amplified from the antigenomic RNA segment templates [101, 102]. Virion assembly occurs at the cell surface, near membranes enriched with GP and mediated by Z, which contains canonical late-budding motifs to recruit components of the endosomal sorting complex required for transport (ESCRT) pathway [90, 103, 104].

Figure 1. Mammarenavirus lifecycle.

(1) Mammarenavirions adsorb to cell-surface factors, followed by endocytic uptake and engagement of intracellular receptors; (2, 3) membrane fusion, uncoating, and release of S and L genomic RNP complexes into the cytosol; (4) transcription of viral subgenomic RNAs, translations of viral proteins NP, GPC, L, and Z, and genome replication via antigenomic intermediates (antigenomes); (5, 6) Z-mediated virion morphogenesis and budding. GP, glycoprotein complex; GP1, glycoprotein 1 subunit; GP2, glycoprotein 2 subunit; GPC, glycoprotein precursor; IGR, intergenic region; L, large; NP, nucleoprotein; RNP, ribonucleoprotein complex; S, small; Z, zinc-binding protein. Modified and updated from [2] as well as from [148] with permission of Taylor & Francis.

3. Progress in treatment of mammarenavirus infections

3.1. Clinical development

Identification, development, and evaluation (safety and efficacy) of candidate medical countermeasures (MCMs) against human mammarenaviruses are a priority because no prophylactic or therapeutic treatments are approved by the U.S. Food and Drug Administration (FDA) or other major agencies, leaving solely limited strategies to prevent (via minimizing rodent exposure and isolating patients) and treat (to supportive care or the non-standard use of virus-specific MCMs). Only four small molecules (ribavirin, favipiravir, LHF-535, and ARN-75039), all targeting LASV (Table 2), have been or are being evaluated in clinical trials. Despite limited evidence and no regulatory approval, ribavirin has been standardly used off-label in endemic areas for many years. Other candidates are only in advanced preclinical development or in early phases of human clinical trials.

Table 2.

Clinical trials to evaluate mammarenavirus antivirals [216].

https://clinicaltrials.gov/ Identifier	Title	Type/Phase	Status
NCT02483260	Intravenous Ribavirin Protocol to Treat Individuals With Viral Hemorrhagic Fever	Interventional/NA	Recruiting
NCT03889106	Cardiovascular Function and Ribavirin Pharmacokinetics and Pharmacodynamics in Patients With Lassa Fever	Observational/NA	Terminated
NCT03993704	Multiple Ascending Oral Dose 14-Day Trial of LHF-535 in Healthy Participants	Interventional/Phase 1	Completed
NCT04285034	Cardiovascular Function and Ribavirin PK/​PD in Lassa Fever in Lassa Fever	Observational/NA	Completed
NCT04907682	Pharmacokinetics, Tolerability and Safety of Favipiravir Compared to Ribavirin for the Treatment of Lassa Fever (SAFARI)	Interventional/Phase 2	Completed
NCT05735249	A Study to Assess the Safety, Tolerability, and Pharmacokinetics of Oral ARN-75039 in Healthy Adult Subjects	Interventional/Phase 1	Recruiting

3.1.1. Ribavirin

Ribavirin is a guanosine prodrug that is phosphorylated after administration to an active form that interferes with diverse cellular pathways. Ribavirin’s antiviral activity is hypothesized to be related to decreased cellular availability of guanosine triphosphate (GTP) and deoxyguanosine GTP (dGTP) and subsequent prevention of viral subgenomic RNA capping, leading to viral RNA degradation; interference with RdRp-mediated transcription and replication, leading to lethal mutagenesis in newly synthesized viral RNAs; and/or inhibition of macrophage activation and lymphocyte proliferation [105–112]. Off-label intravenous administration of ribavirin is the most-used specific treatment option for mammarenavirus infections. Rationale for this use is derived from a clinical study of Lassa fever patients in Sierra Leone in the late 1970s and early 1980s [113]. In this non-blinded study, early initiation of ribavirin for ten days in patients with high viral load appeared to reduce the case fatality rate compared to a control group [113]. However, this and other studies have come under scrutiny for methodologic and analytic flaws to truly determine efficacy; furthermore, reanalysis of the data indicates that ribavirin may actually have been harmful to some patients [114–116]. Ribavirin has also been used clinically to treat patients infected with JUNV, MACV, SBAV, or LUJV [60, 62, 117, 118], but therapeutic benefit in all of these cases continues to be hotly debated. Indeed, the adverse side effects of ribavirin, which include infusion-related rigors, severe anemia, thrombocytosis, and congenital disorders [119, 120], pose a conundrum for physicians weighing the potentially limited antiviral effects versus risks posed by the drug [110, 114]. That said, ribavirin has repeatedly been shown to have preclinical prophylactic and therapeutic value against a range of mammarenaviruses both in vitro and in a range of animal models, including nonhuman primates [121–126]; in the absence of and until identification of other effective therapeutics, these data suggest that ribavirin should not yet be dismissed. Indeed, efforts are underway to create and evaluate ribavirin analogues that may be equally effective while causing fewer side effects [127].

3.1.2. Favipiravir

The pyrazine derivative favipiravir (T-705) is a broad-spectrum purine nucleoside mimic that interferes with mammarenavirus transcription and replication [128–130], possibly through multiple mechanisms (termination of transcription via binding to RdRp and/or incorporation into nascent viral RNAs leading to error catastrophe) [131–134]. In domesticated guinea pigs exposed to the LASV Josiah isolate, treatment with favipiravir (300 mg/kg/day dosed subcutaneously) beginning at 2, 5, or even 7 days after virus exposure resulted in uniform survival with a 2–3 log infectious titer reduction compared to the control group [135]. Similarly, in domesticated guinea pigs exposed to JUNV, initiation (1 or 2 days after exposure) of favipiravir (300 mg/kg/day dosed orally or intraperitoneally twice daily for two weeks) led to 20% and 33–40% survival in the oral and intraperitoneally dosed groups, respectively [136]. In crab-eating macaques (Macaca fascicularis Raffles, 1821) exposed to LASV, initiation (4 days after exposure) of favipiravir treatment (300 mg/kg/day dosed intravenously once and subcutaneously thereafter for 13 days) led to uniform survival although all animal developed clinical signs [137]. Notably, in all published animal study comparisons, favipiravir outperformed ribavirin. Favipiravir escape mutants have been discovered in treated JUNV-infected cells [134], although these in vitro experiments used an attenuated JUNV (Candid#1 vaccine) strain rather than wild-type virus; consideration of combination treatments in the clinic to avoid emergence of therapeutic-resistant variants may be warranted. Combination treatment with favipiravir and ribavirin for two Lassa fever patients reduced viremia in both [31], but whether the combination led to the patients’ survival is unclear from these anecdotal descriptions. In an immunocompromised laboratory mouse model of Lassa fever, combination treatment with favipiravir and ribavirin led to a decline in viremia and protected against lethality; however, cessation of treatment led to a reoccurrence of viremia in all mice [138]. Ongoing phase 2 clinical studies are comparing the pharmacokinetics, safety, and tolerability of favipiravir compared to ribavirin in patients with Lassa fever in Nigeria.

3.1.3. LHF-535

Benzimidazole derivative LHF-535, related to previously promising antiviral ST-193 [139, 140], is a small-molecule viral-entry inhibitor that targets the mammarenavirus GP. Through interaction with the SSP–GP2 prefusion structure, LHF-535 is thought to prevent the conformational changes necessary within the complex to mediate membrane fusion [141]. The molecule was first shown to be inhibit cell transduction by noninfectious lentiviral particles pseudotyped with GPs of LASV lineages II, III, and IV (50% inhibitory concentration [IC₅₀] = 0.1–0.3 nM) but was equally active against JUNV (IC₅₀ = 0.1 nM) and MACV (IC₅₀ = 0.1 nM) pseudotypes, whereas much less activity was measured against CHAPV, SBAV, and LUJV pseudotypes [141]. In the established lethal domesticated guinea pig (Cavia porcellus (Linnaeus, 1758)) model of Lassa fever [142], intraperitoneal injection of 50 mg/kg/day of LHF-535 starting at day 1 or day 3 after exposure to a typically lethal dose of guinea pig-adapted LASV (lineage IV Josiah isolate) protected all animals from fatal outcome and led to reduced viremia and less severe disease compared to control animals [143]. In two phase 1 clinical trials in healthy volunteers, the safety (well-tolerated) and pharmacokinetics (rapid absorption and long half-life) of oral LHF-535 were favorable for further development [144]. LHF-535 pharmacokinetics at the protective dose in domesticated guinea pigs revealed plasma concentrations within the range observed in the phase 1 trial volunteers [143, 144], supporting the continued development of LHF-535 as a LASV-specific therapeutic.

3.1.4. ARN-75039

ARN-75039 is a potentially orally bioavailable imidazopyridine fusion inhibitor that is highly active at low doses against noninfectious vesicular stomatitis Indiana virus particles pseudotyped with JUNV, MACV, or LASV GPs (half-maximal response [EC₅₀] ≤ 1 nM) and wild-type JUNV and LASV in vitro (EC₅₀ <3.3 nM) [145]. In the domesticated guinea pig model of Argentinian hemorrhagic fever [146], oral administration of ARN-75039 led to uniform survival of JUNV-exposed animals when administered on day 2 after virus exposure and to partial survival when administered on day 4 or even day 6 [147]. Production company news indicate that the compound may be equally active against LASV in domesticated guinea pigs. Phase I studies of the tolerability, safety, and pharmacokinetics of oral ARN-75039 in healthy volunteers are ongoing.

4. Preclinical development

4.1. Small-molecule inhibitors

The identification of potentially promising novel small molecules that could be developed as future candidate MCMs is increasingly performed using standardized high-throughput screens. These screens are often established around mammarenavirus surrogates (e.g., reporter protein-encoding non-infectious retrovirion- or vesciculovirion-like particles pseudotyped with mammarenavirus GPs for identification of virion entry inhibitors or non-infectious minigenomes and derived systems consisting of transcription- and replication-regulatory sequences surrounding a reporter protein-encoding gene to identify L protein inhibitors) to enable screening at biosafety level 2. Alternatively, these screens are adapted for use of replicative viruses at biosafety level 4. Variations of screening include the modification of cell lines to encode reporter proteins that are induced (or abrogated) upon pseudotype transduction of virus infection and the use of recombinant mammarenaviruses that encode reporter proteins. Alternatively, screening platforms are developed specifically for identification of mammarenavirus protein-specific inhibitors [148–157].

“Repurposing” of regulatory agency-approved drugs, i.e., their use for a different indication (in this case treatment of mammarenavirus infections) could shorten the time and reduce costs and risks in preclinical development of novel drugs. Consequently, libraries of approved drugs are increasingly evaluated against mammarenaviruses (or surrogate systems) using in vitro high-throughput platforms [158–160]. Such approaches have led to the identification of, for instance:

• afatinib, a pan-ErbB tyrosine kinase inhibitor that potently inhibits LASV RNP activity and authentic JUNV replication [161];

• AM-251, AM-281, and rimonabant (SR141716), known biarylpyrazole cannabinoid receptor 1 (CB₁) antagonists, that robustly inhibit fusion affecting LUJV pseudotype transduction entry in the high nanomolar range [162];

• isavuconazonium, a triazole antifungal that inhibits CHAPV, GTOV, LASV (lineages I–IV), JUNV, MACV, and SBAV pseudotype transduction via inhibition of SSP-GP2-mediated membrane fusion [163];

• lacidipine, a known calcium channel blocker, as GTOV and LASV and d-phenothrin (sumithrin), a pyrethroid, as LASV pseudotype transduction inhibitors that act at the low-pH-induced membrane fusion step [164];

• losmapimod (GW856553X), a selective p38α/β mitogen-activated protein kinase (MAPK) inhibitor, as a LASV pseudotype transduction inhibitor that interferes with SSP-GP2-mediated membrane fusion [165];

• teriflunomide as a potent inhibitor of RNA synthesis and thereby JUNV replication [166];

• manidipine, lercanidipine, and trametinib as LUJV pseudotype transduction inhibitors in the micromolar range. Manidipine and lercanidipine were identified as dihydropyridine calcium channel blockers, whereas trametinib, which also inhibited CHAPV, GTOV, JUNV, LASV, MACV, and SBAV pseudotype transduction and is a known MEK inhibitor, was identified as a mammarenavirion fusion inhibitor [167]; and

• umifenovir, niclosamide, and sertraline as disruptors of LASV replication; and combinatorial umifenovir plus aripiprazole or sertraline as synergistic inhibitors of JUNV and LASV pseudotypes [168, 169].

Although numerous drugs have been identified as in vitro inhibitors via this approach, most have not progressed to in vivo preclinical evaluation. An exception is stampidine, the nucleoside analog derived from stavudine (d4T) and a known retroviral reverse transcriptase inhibitor, which increased survival of laboratory mice exposed to LASV as compared to controls [170].

Alternatively, high-throughput screens are performed with large libraries of chemical compounds that are not yet approved for any treatment. Such approaches have led to the identification of, for instance:

• 16G8, 17C8 as inhibitors of LASV, JUNV, MACV GTOV replication via entry interference (IC₅₀ ≈200–350 nM) [171];

• bergamottin and casticin as entry inhibitors of CHAPV, GTOV, JUNV, LASV, LUJV, and MACV transduction in the micromolar range [172];

• BEZ-235, a dual phosphatidylinositol 3-kinase (PI3K)/mammalian target of rapamycin (Rpm) (mTOR) inhibitor and LY294002, a PI3K inhibitor, as inhibitors of LASV particle production [173];

• BIBX 1382 and OSU-03012, two kinase inhibitors, as inhibitors of LASV replication [174];

• cap-dependent endonuclease inhibitors that inhibit JUNV and LASV replication at micromolar concentrations [175];

• CP100356, a specific P-glycoprotein inhibitor, as an inhibitor of LASV replication (IC₅₀ = 0.062 μM) and JUNV, LASV, MACV, and SBAV pseudotype transduction [176];

• F1204 and F1781, benzotriazole derivates that target the LASV endonuclease activity in a minigenome system [157];

• F1920 and F1965, which inhibit LASV pseudotype transduction [152];

• PF-429242, an inhibitor of cellular site 1 protease (S1P)-mediated processing of LASV GPV, as a disruptor of LASV maturation in the low micromolar range [177];

• quercetin, a flavonol that interrupts the PI3K/Akt pathway, as an inhibitor of JUNV replication via entry interference [178];

• BEZ-235 and LY294002, which together target the PI3K/Akt pathway inhibit viral budding of LASV and replication [173];

• ST-193, ST-294, ST-336, and TRAM-34, which are clotrimazole-derivatives that inhibit GTOV, JUNV, and MACV replication by interference with GP2 [179, 180], and ST-193 as a survival-increasing inhibitor in a LASV domesticated guinea pig experiment [140]; and

• tangeretin, a flavone, as a fusion inhibitor that interrupts replicative LASV entry in the submicromolar range and CHAPV, GTOV, JUNV, LASV, LUJV, MACV, and SBAV pseudotype transduction [181].

4.2. Antibodies

4.2.1. Passive antibody therapy

Transfusion of convalescent sera (or plasma) from survivors of mammarenavirus infections is a plausible and seemingly straightforward therapeutic intervention. For instance, administration of such sera reduces the case fatality rate of Argentinian hemorrhagic fever to <1–2% when administered within the first eight days of disease, and hence is the current treatment of choice for patients with acute disease. However, receipt of convalescent plasma has been associated with late neurologic sequelae (headache, cerebellar tremor, acranial nerve palsies) in 10% of treated patients [182–186]. Data supporting convalescent serum treatment of patients with Lassa fever are less clear; though repeatedly administered to individuals during outbreaks and after accidental laboratory infections [187–189] , these descriptive and observational data, often not obtained under well-controlled conditions, do not enable clear determination of safety and efficacy. Investigational animal studies indicated the potential usefulness of convalescent sera for treatment of LASV infections but only if repeated treatment begins almost immediately after virus exposure and contains high titers of virus-neutralizing antibodies [190, 191]. Likewise, very limited investigational animal studies indicated that GTOV and MACV infections could be treated with convalescent sera as well [192, 193].

4.2.2. Monoclonal antibodies

Even if efficacious, convalescent sera (or plasma) is typically available in limited quantity (in particular, due to a low number of cases and/or in the absence of a program for standardized collection, processing, and administration), is difficult to standardize, and poses a potential risk of transfusion-associated transmission of pathogens, such as HIV-1, hepatitis viruses, or plasmodia. Thus, a major focus of MCM development for mammarenavirids has focused on the identification and characterization of protective monoclonal antibodies (mAbs) from human survivors (or from animals after experimental mammarenavirus infections) which could then be manufactured as a standardized biological product. Indeed, a proof-of-concept study demonstrated that identified neutralizing murine anti-JUNV mAbs expressed as mouse-human chimeric antibodies provided complete protection to JUNV-exposed domesticated guinea pigs when administered on day 2 post-exposure. The mAb J199 provided complete protection when administered on day 6 and offered 92% protection when administered on day 7 [194]. Similarly, human Lassa fever survivor anti-GP mAbs 8.9F, 12.1F, 37.2D, and 37.7H [195] completely protected LASV-exposed crab-eating macaques when administered intravenously at 15 mg/kg at days 0, 4, and 8 post-exposure [196, 197]. Administration of a thoroughly characterized cocktail containing 15 mg/kg each of mAbs 8.9F, 12.1F, and 37.2D (now known as arevirumab-3) protected all macaques when treatment was initiated as late as 8 days after exposure during an advanced stage of disease [196–198]. Importantly, arevirumab-3 completely protected macaques against disease caused by lineage II, III, and IV LASV on day 8 and against lineage VII LASV when treatment was initiated on day 7 [199, 200].

5. Conclusion

Increased understanding of the molecular biology of arenavirids in general, and that of mammarenaviruses in particular, combined with the use of ever more sophisticated high-throughput small molecule screening platforms, resulted in the identification of a plethora of small molecules with anti-mammarenaviral activity. However, most of these molecules were only evaluated using mammarenavirus surrogate systems (usually viral pseudotypes or in minigenome systems) and, even in those settings, required micromolar concentrations to exert effects. Few molecules have been tested in vivo against authentic mammarenaviruses and rarely have they been evaluated against multiple mammarenaviruses side-by-side to identify broad-spectrum options that could, ideally, also protect against newly emerging viruses. Only a handful of promising molecules have resulted in considerable therapeutic benefit in investigational animal models of mammarenavirus disease and only three candidates (arevirumab-3, favipiravir, and ribavirin) were active in nonhuman primate models of, primarily, Lassa fever.

6. Expert opinion

At least two mammarenaviruses, JUNV and LASV, are significant public health concerns because they cause tens to hundreds of thousands of human infections and numerous deaths annually in endemic areas and result in case exportations to other continents (Argentina and Western Africa, respectively) [1, 71]. It is unsurprising, therefore, that the majority of research and development (R&D), including the limited data for the most advanced treatment options, is focused on these two viruses; treatments include off-label use of ribavirin for humans infected with any mammarenaviruses [62, 113, 117, 118], transfusion of convalescent sera to treat human JUNV infections [182–186], and arevirumab-3 as the currently most promising therapeutic under development against LASV infections [196–200]. However, real risk-benefit concerns regarding the safety and efficacy of ribavirin treatment remain unanswered by limited data[110, 114–116, 119, 120], and widespread use of convalescent sera is impractical, possibly unsafe, and poorly evidenced; absent data from well-designed clinical trials of either therapeutic, field clinicians are heavily reliant on supportive care strategies to improve patient outcomes.

At least historically, R&D strategies around mammarenavirus MCMs have unfortunately (but perhaps necessarily) been driven by and targeted more urgent needs in response to outbreak disease burden (and the resultant virus- and disease-specific questions and opportunities inherent to those “felt” needs); however, a narrow focus in the short term has not translated well into broader longer-term vision and strategy. First, the overall focus on JUNV and LASV neglects the possibility that the other highly virulent mammarenaviruses (CHAPV, GTOV, LUJV, MACV, SBAV), currently considered exotic pathogens causing only a handful to a few hundred of cases over decades, are potential pandemic pathogens [1]. Because all mammarenaviruses are mammal-borne, disturbances in, for instance, rodent habitats and concomitant rodent migrations leading to increased human–rodent contact makes this scenario a distinct possibility. Pathogenic mammarenaviruses are also the minority of known members of genus Mammarenavirus, which currently includes 43 species for 50 classified viruses and is associated with a list of numerous unclassified viruses that grows steadily [2]. Consequently, R&D efforts need to anticipate the emergence of novel pathogenic mammarenaviruses. Second, most R&D efforts rely on the use of surrogate systems rather than authentic viruses due to the risk to the laboratory worker that is associated with these pathogens (and hence the need for maximum containment laboratories, of which there a few worldwide) and/or the lack of access to the viruses. This reliance skews research towards identification of entry inhibitors, because pseudotyped viruses are the only truly amenable, available, and easily established for high-throughput drug screens. However, entry dynamics, mechanics, and kinetics of pseudotypes differ considerable from those of authentic viruses and EC₅₀ values and other important drug indices obtained with surrogate systems are rarely extrapolatable to or predictive for authentic virus infections. Third, when human clinical trials are either not logistically feasible or ethically unjustifiable, investigational animal model data become crucial for drug development, possible under the FDA Animal Rule [201]. However, only a few animal models of mammarenavirus diseases have been rigorously established and, once again, are biased toward JUNV and LASV infection models that require maximum containment facilities [142, 146, 202]. Fourth, there is likely considerable intra-species diversity among all pathogenic mammarenavirus. LASV is the only virus for which ample genomic data are available, and these data clearly delineate at least seven genomically diverse lineages that differ in geographic distribution, likely host spectrum [71], and also in response to MCMs [200]. However, until recently, the vast majority of MCM discovery and evaluation efforts have only targeted LASV lineage IV isolates (or, in case of pseudotype surrogate systems, GPs thereof). Together, these observations indicate a need for a more concerted, community-wide effort to create and make available standardized reagents that include well-characterized viruses (and genes and proteins thereof) that represent all major clades of mammarenaviruses and also all major intra-specific lineages so that drug candidates can be tested systematically to predict their value as broad-spectrum antivirals. In addition, MCM R&D needs to consider different field realities for their applications. Ideally, rigorous clinical trials show an MCM to be safe and efficacious as well as practicable in relevant settings (i.e., effective in the real world). Because most pathogenic mammarenaviruses are endemic in underdeveloped and often rural areas, affordability, and availability to local populations, including farm workers or hospital staff, are barriers to improving clinical outcomes. Studies cited in this manuscript rarely mention real-world applicability in connection with a given “promising” antiviral. To the contrary, the most products under development are based on monoclonal antibodies [196–200], which are expensive to produce and purify and require technological and manufacturing capabilities rarely available in areas most in need. Private–public partnerships could possibly mitigate these challenges in the long term by building local capabilities that are not pathogen-specific [203]. In the meantime, “repurposing” of regulatory-agency-approved drugs may be pursued to help mitigate the health impacts of local mammarenavirus disease outbreaks; drugs already approved for other indications a) may already be available in quantities necessary for distribution, b) may not be expensive especially if intellectual property protections have already expired, and c) may be quickly approved due to their known human safety profile. Repurposing screens ought to focus on drugs that are administered orally, stable at room temperature, have a long shelf life, and can be combined with additional out-of-class MCMs to prevent the evolution of drug-resistance traits in the targeted viruses.

It should be noted that recent pandemic preparedness efforts have underscored the need to more broadly characterize high-risk viral families (rather than just focus only their notorious members); however, the selection of “prototype pathogen” viruses that might serve as proxy for mammarenaviral family members is particularly challenging given known (and unknown) within-virus and within-family genetic variations. The degree of “representativeness” in the face of this diversity needs to be carefully considered, especially for prioritization of preclinical development efforts [204].

Even in the diseases that have received the most attention, there is a need for robust clinical trials data to inform clinical decision-making, most obvious in the common non-standard use of ribavirin in Lassa fever patients based on inadequate data from many decades ago. That non-standard (but now routine) use of ribavirin complicates the design of current and future randomized clinical trials, namely whether ribavirin’s use over many years obligates its inclusion either alone or in combination in clinical trials. The recent establishment of well-characterized Lassa fever clinical cohorts in Western Africa, as well as preparation and pre-positioning for clinical trials of advanced MCMs in these clinical networks, shows promise for clearing up some long-held therapeutic ambiguity. For example, results from LASCOPE (NCT03655561), a prospective observational cohort study of Lassa fever aiming to more fully characterizing the natural history of disease, identify risk factors, and standardize case management, also defined a reference mortality rate to inform future clinical trial endpoints and suggested that the need for dialysis should also be considered an important clinical outcome in any evaluation of therapeutics [205]. Increasingly, clinical trial evaluation also must consider the specific needs of vulnerable populations (e.g., those with HIV/AIDS or diabetes, pregnant women) that are at risk of acquiring mammarenavirus infections. Furthermore, as with many human viral infections, it is likely that combination therapeutics that include highly effective mAb and small-molecule antivirals (with potentially advantages in tissue distribution) may be needed to optimize acute and convalescent outcomes after mammarenavirus infections. Central nervous system manifestations (encephalopathy, coma, seizures) have been associated with poor outcomes in these cohorts [74] and LASV has been detected in the cerebrospinal fluid [206]; the relative penetration of advanced therapeutic candidates into the central nervous system is unknown but is plausibly less likely for mAb-based strategies [207]. The pathogenesis of hearing loss during acute Lassa fever and in survivors is uncertain; notably, ribavirin use during acute Lassa fever does not appear to prevent the development of Lassa-fever-associated sensorineural deafness [68]. Although related transmission events have not been documented, recent descriptions of LASV persistence in the semen of male survivors at least argue that therapeutic penetration into immune-privileged tissues merits consideration for public health reasons. Indeed, future clinical trials might include secondary endpoints targeting the post-acute clinical sequelae or viral persistence endpoints. Finally, though outside the scope of this review, mammarenavirus-specific therapeutic approaches cannot be considered in isolation; safe and effective antiviral strategies must be coupled and bundled with appropriate supportive or critical care to improve patient outcomes in an optimal future.

Article highlights.

• Several mammarenaviruses can cause severe human diseases with high case fatality rates and hence are considered public health and bioterrorism threats.

• Therapeutic options for mammarenavirus infections and diseases remain highly limited and controversial.

• High-throughput screening of regulatory-agency-approved drugs for the treatment of a variety of diseases identified numerous molecules that could be “repurposed” as medical countermeasures (MCMs) against mammarenaviruses.

• However, due to their classification as Risk Group 4 agents, further development of identified candidate MCMs has been slow and antiviral activity evaluation of identified candidates in animal models has rarely occurred.

• The most promising treatment under development is a cocktail consisting of three monoclonal antibodies that protects macaques against multiple lineages of Lassa virus even in advanced stages of disease.

This box summarizes key points contained in the article.

List of abbreviations used in the manuscript

CB1

cannabinoid receptor 1

CD63

CD63 molecule

CHAPV

Chapare virus

DAG1

dystroglycan 1

dGTP

deoxyguanosine GTP

EC₅₀

half-maximal response

ESCRT

endosomal sorting complex required for transport

FDA

U.S. Food and Drug Administration

FLEV

Flexal virus

GP

glycoprotein complex

GP1

glycoprotein subunit 1

GP2

glycoprotein subunit 2

GPC

glycoprotein precursor

GTOV

Guanarito virus

GTP

guanosine triphosphate

IC₅₀

50% inhibitory concentration

IGR

intragenic region

JUNV

Junín virus

L

large

LAMP1

lysosomal-associated membrane protein 1

LASV

Lassa virus

LCMV

lymphocytic choriomeningitis virus

LUJV

Lujo virus

MACV

Machupo virus

MAPK

mitogen-activated protein kinase

MCM

medical countermeasure

NP

nucleoprotein

NRP2

neuropilin 2

ORF

open reading frame

R&D

research and development

RdRp

RNA-directed RNA polymerase

RNA

ribonucleic acid

RNP

ribonucleoprotein

S

small

S1P

site 1 protease

SBAV

Sabiá virus

SSP

stable signal peptide

TFRC

transferrin receptor 1

Z

zinc-binding protein