Usutu Virus

Introduction and History

USUV was first identified in South Africa in Culex naevei mosquitoes in 1959 and subsequently spread to regions thousands of kilometers apart, such as Senegal, Central African Republic, Nigeria, Uganda, Burkina Faso, Cote d’Ivoire, and Morocco. At that time, it was not associated with clinical disease in birds or mammals. Recently, these and other flaviviruses have spread and are now found in Tunisian oases, where 10% of the tested anthropophilic mosquitoes as well as the abundant population of loving doves were seropositive for WNV and 4% were seropositive for USUV. Occurrence of such antiflavivirus antibodies increases with dove age and proximity to the coast, rather than vegetation type. Although surrounded by dry, barren desert, oases are wet and densely vegetated areas that support rich local bird populations as well as providing climatic conditions appropriate for mosquitoes. Oases also have concentrated human populations, many of whom are farmers with a high rate of exposure to mosquitoes.

Outside of Africa, USUV was first documented in Central European birds in 2001 and the following years. Retrospective studies, however, suggest that this virus was present in 6% of those tested in Italy in 1996 and may have arrived in Europe in the early 1990s. In 2001, USUV was first detected near or in Vienna and elsewhere in Austria and later in Switzerland, where it caused avian deaths, especially among wild blackbirds, captive great gray owls, and barn swallows. In the following years, this viral strain continued to kill birds in Austria, suggesting that the virus is able to overwinter and establish a local bird–mosquito transmission cycle in that region.

Between 2006 and 2009, USUV caused outbreaks in free and captive wild birds in Budapest, Hungary and Italy (2008–2009), in which over 1000 birds died. In a Zurich zoo, USUV was particularly deadly for wild members of Passeriformes (particularly blackbirds and house sparrows), and, to a lesser extent, Strigiformes species. Based upon the 99.9% identity of the nearly complete genomes of the Austrian and Hungarian viruses, it is most likely that the USUV strains from wild birds in Austria spread to Hungarian wild birds rather than from a separate introduction from Africa. Despite winter temperatures as low as −20°C in Central Europe, USUV has adapted to local mosquito species and is endemic in several European countries, with different viral genetic lineages emerging and undergoing multiple dispersal events. In Italy alone from 2008 to 2009, at least 11 viral strains were found in sentinel horses, dead blackbirds, and a very small number of magpies. It should be noted that almost half of the blackbirds examined were coinfected by hemosporidia (blood protozoans), perhaps weakening the hosts and contributing to disease severity of Plasmodium species. This is similar to the situation occurring in a large disease outbreak in the Netherlands in 2016, in which many blackbirds and great gray owls were coinfected with USUV and Plasmodium , the hemosporidian causative agent of malaria. It was suggested that the coinfections may be partially explained by the fact that mosquitoes are vectors for both pathogens; however, Culex species mosquitoes are the primary vectors for USUV in Europe, while Anopheles mosquito species primarily transmit Plasmodium . Alternatively, coinfection may lead to increased severity of USUV-related disease. Notably, USUV-infected Aedes albopictus mosquitoes have been found in Italy and are also vectors for DENV and Zika virus (ZIKV).

USUV has been found in a variety of mosquitoes, birds, and bats in Europe, including Hungary, Switzerland, Spain, Italy, the Czech Republic, Germany, Belgium, and France. Moreover, antibodies to the virus in birds are present in Poland and Greece and in a horse in Serbia. USUV infection is also recurrent in Austria (2001–2006), Hungary (2005–2006), Italy (2009–2016), Spain (2006, 2009, 2012), and Germany (2010–2015). A large European outbreak in humans occurred in 2016 in Belgium, Germany, France, and the Netherlands. The complete genome of the responsible USUV strain is closely related to isolates from Germany (Europe 3 lineage, described later). The North Eastern Italian strain nucleosides are identical to an USUV strain later detected in an immunocompromised human patient’s brain and are highly similar to viruses isolated from birds in Vienna and Budapest. Interestingly, the Italian isolates from birds as well as those from Spanish Cx. pipiens mosquitoes were 97%–99% identical to the 1959 South African reference strain. In the 2009 USUV outbreak in Spain, the virus was detected in Cx. perexiguus mosquitoes and, later, in two ill song thrushes in 2012. Comparisons of these Spanish strains, however, found differences between the mosquito and bird isolates.

The first report of human infection was in the early 1980s in Africa. In 2009, the first two human cases of USUV infection were reported in Europe after isolating the virus from patients’ blood and detecting viral RNA in the cerebrospinal fluid (CSF) in immunocompromised patients in Italy. One of the patients who developed severely impaired cerebral functions had very recently undergone orthotropic liver transplantation as a treatment for thrombotic thrombocytopenic purpura.

At least 21 human cases of USUV infection have been reported as of 2017 with moderate (rash, fever, and headache) to severe (neurological disorders) illness. Two human cases from the Central African Republic (1980s) and Burkina Faso (2004) presented with moderate symptoms, such as fever, rash, and jaundice. The first cases of meningoencephalitis due to USUV were reported in 2009 in two immunocompromised subjects in Italy. USUV was also present in the CSF from three patients with acute meningoencephalitis in 2008–2009. A retrospective study in Italy detected USUV infection in patients with suspected viral encephalitis or meningoencephalitis and comorbidities during the same time period. Intriguingly, in a retrospective study of CSF and serum, not only was USUV infection first detected in Italy in 2008, but also the first case of WNV neuroinvasive disease in humans was found in Italy in that year. The number of human USUV-related neurological disease cases in 2009 was fivefold higher than the corresponding number of WNV cases in that area. Additionally, 6.6% of well, ill, or hospital inpatients and outpatients tested ( n = 609) had anti-USUV antibodies, while 3.0% had anti-WNV antibodies. From 2008 to 2009, in the tested region of Italy, therefore, incidence of USUV infection and neurological disease in humans was much higher than those of WNV. Prevalence of USUV-specific antibodies in healthy blood donors, however, ranged from 0.02% to 1.1% in Italy and Germany from 2010 to 2011. The prevalence of USUV may be underestimated due to its high serological cross-reactivity and cocirculation with the closely related WNV and tickborne encephalitis virus (TBEV). In 2013, neuroinvasive infection was found in three patients from Croatia as well.

The Disease

Symptomatic infection of wild birds may be either mild or lead to serious infection or death. Multiorgan failure, especially brain lesions, seems to be the most likely cause of death. Symptoms involve depression, ruffled plumage, half-closed eyes, anorexia, inability to fly, jerky movements, torticollis, nystagmus, incoordination, and seizures, as well as loss of mistrust of humans. These symptoms are associated with brainstem and cortical neuron necrosis. Upon necropsy, most birds have marked splenomegaly, mild hepatomegaly, and pulmonary hyperemia. Lesions are very discrete and are composed primarily of neuronal necrosis, leukocytolysis in and around brain blood vessels, myocardial degeneration, coagulative necrosis of the liver and spleen, hyperemia and edema of the lungs, perivascular infiltration of the kidneys, and intestinal inflammation. USUV infection of domestic chickens and domestic geese, however, leads to minimal pathology.

In humans, infection is typically mild. USUV was found in an African patient with fever and rash, and viral RNA was present in an Austrian patient with rash. In a small number of cases, however, infection leads to neurological disease or death. In a retrospective survey of over 650 CSF samples collected in 2016 from patients with infectious or neurologic syndromes in France, only 1 CSF sample contained USUV RNA.

In addition to meningitis or meningoencephalitis, symptoms in severe human cases include persistent high fever, headache, hepatitis, or idiopathic facial paralysis (one patient each for the latter two symptoms). Only one case of neurological disease is known to have occurred in an infected immunocompetent person: a 29-year-old female from Croatia in 2013. The two other cases reported in this outbreak both had comorbidities and were middle-aged (56 and 61 years of age), indicating that disease is not confined to the young and elderly. The immunocompetent woman displayed disorientation and somnolence preceding an elongated period of memory and speech dysfunction. She and the two other patients in this outbreak had nuchal rigidity, hand tremor, and hyperreflexia. Their electroencephalography showed diffusely slow activity. Since USUV neurological signs may be very similar to those of WNV neuroinvasive disease, it is important to consider USUV infection in areas in which both viruses are present. Unfortunately, no USUV-specific treatment or vaccines are currently available.

Both African and European strains of USUV upregulate the cellular autophagic pathway by inducing the unfolded protein response. During this pathway, cytoplasmic components are sequestered into double-membrane vesicles where they are degraded and which, under normal circumstances, help to maintain cellular homeostasis as well as limiting the extent of apoptosis. Rapamycin, an inductor of autophagy, increased the virus yield of both USUV strains, while two inhibitors of autophagy decreased virus yield. Rapamycin activates type I IFN production that protects normal adult mice, but not in those animals lacking the IFN receptor or in suckling mice, the latter of which develop depression, disorientation, paraplegia, paralysis, and coma. Moderate neuron cell death was also seen in the spinal cords of very young mice, especially in the ventral horns, in addition to multifocal demyelination, while apoptosis was spread throughout the brain, particularly the brain stem, but also in white matter of the cerebellum and medulla, most likely in oligodendrites. This was often in conjunction with primary demyelination. Skeletal muscle appears to be the only location of peripheral USUV replication.

USUV-induced mortality in suckling mice is dose-dependent, with 60% dying after infection with 10 ⁴ plaque-forming units of virus, while all tested adult mice survived at all doses tested. Accordingly, while the brains of suckling mice were positive for USUV RNA on day 7 postinfection, no such RNA was detected in the brains of adults. In contrast, suckling and adult mice had higher mortality rates when infected by WNV. USUV infection does not lead to cross-protection against WNV infection in adult mice; however, almost all USUV-infected mice survive infection with a neuroinvasive strain of WNV. Sera from WNV-infected mice also cross-react with USUV, neutralizing both flaviviruses and, upon challenge, protect animals from disease and death. Additionally, immunization with WNV lineage 1 recombinant subviral particles, which share antigenic features with infectious virions, stimulates the production of low levels of USUV-specific antibodies, which are active during subsequent USUV infection. Comparison of multiple USUV strains to WNV defines residues of the E protein that are important or critical to the observed serological cross-reactivity.

The Virus

Most USUV characteristics are typical of other members of the JEV complex of flaviviruses, especially WNV, with which it may interact at the population level. USUV and WNV are antigenically cross-reactive and infect human and animal populations throughout Europe, where they cocirculate in 10 countries. USUV and WNV both infect 34 bird species from 11 orders, including carrion crows, Eurasian blackbirds, Eurasian blackcaps, European robins, and magpies. At least four mosquito species are potential vectors for both flaviviruses, most importantly, Cx. pipiens with a feeding preference for blackbirds and Eurasian magpies. USUV-infected Ae. albopictus has an expanding geographical range and is also another important vector in urban transmission cycles of DENV. Such cocirculation may alter the epidemiology of these viruses, such as occurred with the displacement of St. Louis encephalitis virus by WNV in the United States and may also increase disease severity, such as the increase in DENV pathology in individuals previously infected by JEV in Thailand.

USUV does not display tropism to any specific cell, tissue, or organ. One study of dead birds in Austria detected viral nucleic acid in brain neurons (100%; n = 40 birds), proventricular glands (86%; n = 22), myocardial fibers (73%; n = 33), renal tubular cells (53%; n = 36), intestinal tunica muscularis (41%; n = 32), splenic macrophages (35%; n = 34), hepatic Kupffer cells (32%; n = 38), and lungs (21%, n = 34). USUV is able to grow in cultured cells from a wide range of vertebrates, including humans, monkeys, horses, pigs, rabbits, cattle, dogs, cats, hamsters, rats, and turtles, as well as primary goose embryo fibroblasts. USUV causes cytopathic effect in Vero (monkey), PK-15 (pig), and goose embryo fibroblast cells in vitro. USUV also infects astrocytes, microglia, and neuronal stem cells. The infection rate of human astrocytes and the amount of viral production is greater than that caused by Zika virus, while astrocyte proliferation is decreased. USUV also induces massive amounts of caspase-dependent cell death in neural stem cells ex vivo and a strong anti-USUV immune response that may lead to inflammation.

The USUV reference strain (SAAR-1776; from South African mosquitoes in 1958) is not associated with avian deaths, while European strains have caused massive bird deaths. Austrian and South African isolates, nevertheless, have 97% nucleoside and 99% amino acid identities. Among tested members of the JEV serogroup, USUV is most closely related to Murray Valley encephalitis virus (73% nucleoside and 82% amino acid identities), followed by JEV (71% and 81%), and is furthest from West Nile virus (68% and 75%).

Sequence analysis indicates the existence of at least seven USUV strains with nucleoside and amino acid identities of 96%–99% and 99%, respectively. USUV also shares nucleoside and amino acid identities of 81% and 94%, with an outlier USUV subtype strain, respectively. The first USUV strain isolated from an immunocompromised patient with neuroinvasive disease has overall genome identities of 99% and 96%, respectively, with those of isolates from Europe and Africa. Comparison of the complete human USUV polyprotein sequence to those of bird-derived strains revealed two unique amino acid substitutions, one that is believed to be in the receptor-binding domain of the E protein. In WNV, substitutions within this domain alter virus infectivity, virulence, antigenicity, and escape from neutralizing antibodies. The other substitution is in a residue in the viral RNA polymerase domain of the NS5 protein. This domain is highly conserved among USUV isolates that are not associated with human infection.