Pathological features of West Nile and Usutu virus natural infections in wild and domestic animals and in humans: A comparative review

Mosquito-borne flaviviruses are emerging pathogens with zoonotic potential. Due to the recent climate and environmental changes, they are spreading across Europe, becoming a major threat for public and veterinary health. West Nile virus (WNV) and Usutu virus (USUV) are arboviruses that are responsible for multiple disease outbreaks in different species of birds, reptiles, and mammals, including humans. This review reports and compares the clinical signs as well as the gross and microscopic pathological features during natural infection with WNV and USUV in wild and domestic animals, as well as in humans. The main objective of this comparative review is to delineate the common features and the specific differences that characterize WNV- and USUV-induced diseases in each group of species and to highlight the main gaps in knowledge that could provide insight for further investigation on the pathogenesis and neurovirulence of these viruses.


Introduction
Emerging mosquito-borne flaviviruses are a major threat to human and animal health worldwide, in part due to the continued geographical expansion of their vector species [1]. West Nile Virus (WNV) and Usutu Virus (USUV) belong to the family Flaviviridae, genus Flavivirus, and are responsible for multiple outbreaks of disease in animals and humans spreading across Europe and North America alongside with climate changes [2][3][4][5]. USUV isolates are currently classified into eight lineages, clustered into the African (AF1, -2, -3) and the European group (EU1, -2, -3, -4, -5) [6]. WNV isolates can be classified in at least eight lineages. Pathogenic strains belong to lineage 1, that include also Kunjin virus previously considered as a separate virus, and lineage 2 [7].
In mammals, tropism for the central nervous system (CNS) is a main feature of West Nile virus infection, characterized by the occurrence of encephalitis or encephalomyelitis, with or without meningeal involvement [3]. In contrast to what is commonly observed in mammals, in avian species, the involvement of the CNS is part of a severe systemic disease that often involves also other organs. Lesions associated with USUV infection are only reported in birds and, like WNV, affect multiple organs. Characteristics of WNV and USUV associated disease have been reported in the last decades in numerous studies, but a collection and comparison of common features in the same and different species is yet to be carried out. The aim of this paper is to give a general overview of the pathology during WNV and USUV infections in both wildlife, domestic animals, and in humans, focusing and delineating the main traits that characterize the disease for each virus, highlighting possible pathogenetic insights and main gaps in knowledge, that should represent topics for further investigations.

Transmission cycle and host range of WNV and USUV
WNV and USUV are both characterized by a similar enzootic transmission cycle that includes several species of mosquitoes as vectors and birds as amplifying hosts. Humans and other mammals act as incidental dead-end hosts as they do not develop high levels of viremia to maintain the transmission cycle.
WNV has been isolated from several species of mosquitoes worldwide, but the most relevant of these as vectors in the transmission cycle belong all to the genus Culex [8]. USUV has been isolated from Culex pipiens, which is considered the main vector, and its competence as vector has also been demonstrated under laboratory conditions together with Cx. neavei and Cx. quinquefasciatus [6]. WNV and USUV infections have been identified in a wide range of bird species, both clinically healthy animals in the context of active surveillance campaigns, as well as in diseased or dead bird species (listed in Tables 1 and 2). Among them, American crows (Corvus brachyrhynchos) [9] are considered highly susceptible to WNV infection while Eurasian blackbirds (Turdus merula) and great grey owls (Strix nebulosa) are highly susceptible to USUV [6]. In addition to infection in birds, WNV infects incidental hosts, mainly humans and horses, and sporadically other species listed in Table 3. USUV infection has been detected in various mammals including horses [10], dogs [11], bats [12], squirrels [13], wild boars, wild ruminants [14], lizards [15] and humans [16]; however, lesions are only reported in birds.

Pathogenesis of WNV-and USUV-associated disease in animals and humans
Disease associated with WNV and USUV infection has been reported in several animal species. Birds infected with USUV or WNV usually show a multisystemic disease involving the central nervous system, liver, spleen, heart and kidney [9,59]. In addition, in raptors the eye is reported as one of the main target organs during WNV natural infection [26]. However, whilst incidental hosts may show pathology in numerous organ systems, clinical disease primarily manifests with neurological symptoms. The pathogenesis and dissemination of WNV has been widely studied in mice models and avian natural hosts, although many of its aspects are still to be elucidated. Virus is transmitted by mosquito bite, although, based on identification of virus antigen in the neurons of enteric ganglia, a possible oral transmission has also been proposed [59]. In the skin, WNV is thought to replicate in keratinocytes and Langerhans dendritic cells, the latter migrate to draining lymph nodes, from which a primary viremia starts with subsequent infection of peripheral organs such as the spleen [60]. Neuroinvasion is one of the most relevant and investigated aspects of WNV infection. Regardless of the transmission mechanism of the virus, two overarching routes of WNV neuroinvasion have been proposed: transneural and hematogenous. Transneural pathways include transport along neurons of the periphery or olfactory tract directly into the brain. Hematogenous invasion could occur transcellularly via transcytosis of virus across the blood-brain barrier (BBB), or paracellularly as a result of inflammatory disruption of the BBB resulting in invasion of virus, either free in the blood or trafficked within infected immune cells acting as a so called 'Trojan horse' [60]. Knowledge on pathogenesis of USUV is still lacking, but due to the close relatedness with WNV, similar pathways of neuroinvasion are hypothesized.
In general, the main pathologic features of WNV and USUV infection include necrosis and inflammation. The pathogenesis behind the cell death during WNV and USUV infection is still unknown. For WNV two Table 1 Avian species reported as susceptible to WNV disease development.
overarching mechanisms are thought to be involved: (1) a direct cytopathic effect of the virus on the cells and (2) the host antiviral immune response [61]. It has been shown, using both in vivo and in vitro models, that WNV can induce cell death in neurons through different mechanisms such as apoptosis [62] and pyroptosis [63]. The inflammation induced by WNV is mainly represented by infiltration of lymphocytes, plasma cells and histiocytes and is suggested to be triggered by the production of cytokines such as IL-1β, -6, -8, and tumor necrosis factor (TNF)-α in the tissue [64]. Innate immunity during WNV infection is mediated by several pathogen recognition receptors (PRRs), RNA binding produces downstream activation of pathways that ultimately induce expression of IFN I and III molecules with direct or indirect antiviral functions. The adaptive immune response against flaviviruses is either mediated by antibodies produced by B-lymphocytes or T-cell mediated cytotoxicity. Regarding T-cell immunity CD4+ cells provide help for antibody response and sustain CD8+ mediated viral clearance. CD8+ cells mediated immunopathology is described in mammals and is mainly related to aberrant cytokine production and or cytolytic activity [65]. In this review we report the clinical features, lesions and viral antigen distribution associated with WNV and USUV infection in the main target organs of different host species and summarize these findings in Tables 4 to 8.

Clinical features of WNV-and USUV-associated disease in animals
Clinical disease in birds infected with WNV and USUV presents as non-specific (e.g., immobility, apathy, ruffled feather) and neurological signs (ataxia, paresis, tremors, torticollis, inability to fly and seizures) [24,59], furthermore vision loss is reported for WNV natural infection in diurnal raptors [26]. WNV infection of incidental host species is most often asymptomatic but presents mainly as neurological disease in the small percentage of symptomatic cases. Specifically, in patients around 80% of the cases are asymptomatic. Most symptomatic patients present with a mild, self-limiting febrile disease called West Nile fever (WNF), often accompanied by headache, myalgia, lethargy and gastrointestinal complaints. Approximately 1% of human infections results in severe neurological involvement, called West Nile neuroinvasive disease (WNND), with a large range of manifestations such as generalized weakness [16,66], hyperreflexia [67], transient paralysis of limbs [68], facial paralysis [69], dysgeusia, aphasia [66], ataxia and dysmetria [16]. In the few reported human clinical cases described so far, USUV infections present with mild symptoms such as rash and fever, however, evidence of USUV neuroinvasive disease has been reported in several countries [16,66,67,70], mainly concerning patients with contributing risk factors. Nevertheless, pathology data are lacking.

Pathological features of WNV-associated disease in animals
The description of WNV-associated disease lesions will focus on the most commonly affected organs.

Table 4
Comparison of histologic lesions in the nervous system of birds infected with WNV or USUV.
In the present table, the term "Game birds" is referred to the species belonging to the order Galliformes as listed in Table 1  Table 5 Comparison of histologic lesions in extra neurologic organs in bird infected with WNV or USUV.
In the present table, the term "Game birds" is referred to the species belonging to the order Galliformes as listed in Table 1  Table 6 Comparison of presence of virus antigen by immunohistochemistry in birds infected with WNV or USUV.
Gastrointestinal system Mucosal epithelial cells Endocrine system Thyroid follicular epithelium -

Eye
Ocular lesions have only been described in raptors and remain under investigated in other relevant avian groups such as Passerine birds. Gross lesions include intraocular hemorrhage [24,26], discoloration of the retina, vitreous humor opacity [26], corneal erosion [24], and areas of opacity in the fundus [27,28]. Microscopically, lymphoplasmacytic pectenitis is most frequently reported, followed by iridocyclitis and keratoconjunctivitis [72]. Chorioretinal lesions have been classified based on their severity and include degeneration and disarray of the retinal architecture and lymphoplasmacytic infiltration of the choroid, optic disc and optic nerve [24,27,28]. WNV antigen is seen in choroid and pecten in passerine birds without lesions, while in raptors it is commonly observed in retinal neurons, pigmented epithelial cells, macrophages in the iris and melanocytes in the pecten [26][27][28].

Other organs
Lesions and viral antigen distribution associated with WNV infection are reported in detail in Tables 5 and 6 for birds and Tables 7 and 8 for mammals and reptiles.

Pathological features of USUV-associated disease in birds
The description of USUV-associated disease lesions will focus on the most commonly affected organs.

Spleen
Splenomegaly is one of the major macroscopic findings associated with USUV-infection [38,80,84]. Microscopically, zonal necrosis centered in the sheathed arteries in association with histiocytic hyperplasia is seen in raptors and passerines [37][38][39]41,84]. The same lesions are also hallmarks of avian malaria, caused by highly prevalent haemoprotozoan infection in passerine birds [86] possibly also in coinfection with USUV. Regarding virus antigen distribution is seen in capsule spindle cells, endothelial cells and mononuclear cells [37,41,84]. "+" IHC positive cells; "-" IHC negative cells; "NI" infection not reported in the animal group; "NE" not evaluated.
In the present table, the term "Game birds" is referred to the species belonging to the order Galliformes as listed in Table 1 6
A detailed list of the species included within the animal groups reported in the present table is provided on Table 3 6

Other organs
Lesions associated with USUV infection have been observed in several organs of the respiratory and gastrointestinal system. Furthermore, site-specific hyperkeratosis of peri cloacal skin is observed in blackbirds [38,59]. Associated histological lesions and viral antigen distribution are reported in detail in Tables 5 and 6.

Pathological features of WNV-associated disease in humans
There are no fatal human cases of USUV associated disease described in literature, therefore only pathology resulting from WNV-associated disease is described in humans.

Nervous system
Histological findings of WNND vary depending on the severity and presentation of disease. In general, non-supportive encephalitis, characterized by lymphohistiocytic perivascular cuffs, and neuronal necrosis with neuronophagia and microglial nodules are observed in the midbrain, pons, medulla, substantia nigra and cerebellum [69,[87][88][89][90][91][92][93]. Gliosis is also observed in the lumbar region [88,90] of the spinal cord. Presence of viral antigen is seen in neurons and Purkinje cells often associated with foci of inflammation, except in severely immunocompromised patients who also show extensive WNV staining independent of inflammatory lesions [69,88,91,93].

Eye
Ocular manifestations such as uveitis, optic neuritis, occlusive vasculitis and chorioretinitis are reported, as well as isolated cases of chorioretinitis [96][97][98][99], neuritis [96,97,100] and vitritis [96,99], most often associated with meningitis and encephalitis. However, cases involving only the eye suggest that ocular symptoms can also occur in absence of descending infection from the brain to the ocular nerves. Neuronal atrophy was observed in the inner retinal layer, however no retinal tissue was available for IHC to confirm or discount presence of virus.

Kidney
In severely immunocompromised patients, virus antigen and inflammation has been reported in the kidney [69,91], but how such findings may contribute to disease and correlate with clinical presentations during acute infection and convalescence is yet to be fully determined.

Other organs
Histopathological changes have been observed also in the skin, including hemorrhages and thrombi within the dermis, as well as perivascular inflammatory infiltrates. WNV antigen was identified within the cytoplasm of perivascular infiltrates, as well as the vascular endothelium of the small dermal vessels. Lesions and viral antigen distribution associated with WNV infection in humans are reported in detail in Tables 7 and 8. "+" IHC positive cells; "-" IHC negative cells; "NI" infection not reported in the animal group; "NE" not evaluated. Investigation of the presence of the viral antigen through immunohistochemistry in non-avian species is rarely reported. Positivity is present in only few organs listed in the table. A more extensive overview is provided in the supplementary materials.

Discussion
Lesions associated with WNV infection are reported in a wide range of animals and in humans. In USUV infection, despite evidence of infection in mammals and disease reported in humans, pathological data are available only for birds. In order to extrapolate the known mechanisms and tropism involved in WNV pathogenesis to the lesser-known mechanisms of USUV pathogenesis, WNV and USUV-associated disease can be compared according to the type of lesions, distribution of viral antigen by IHC and clinical presentation.

WNV and USUV neurologic disease
For WNV, in all hosts, the main pattern of lesions in the CNS is characterized by lymphohistiocytic meningoencephalomyelitis and neuronal loss [101,102]. The same pattern is observed in birds naturally infected with USUV, though lesions are not reported in the spinal cord, probably due to lack of investigation in this organ. The pathogenetic mechanisms behind the cellular damage and subsequent functional impairment associated with WNV and USUV infection are still to be completely elucidated [61]. WNV can induce cell damage either directly or indirectly through the immune reaction of the host. In vivo and in vitro experimental studies have shown that WNV can induce programmed cell death, with a more severe effect if it occurs in non-renewing cell populations, such as neurons [62,63,102]. Additionally, WNV infection can trigger an immune reaction by inducing the production of several inflammatory mediators; this promotes neuroinflammation and appears to be a main factor driving WNV-induced neuronal damage [64]. At the same time, virus-induced immune response is suggested to have a negative impact on neurotransmission, this has been highlighted by the upregulation of pleiotropic genes functioning at several level of the immune-neural-synaptic axis in a Non-Human Primate (NHP) model of WNV-ND [103]. Nevertheless, how much cell injury can be attributed to viral cytopathology and how much to the inflammatory response is still not known.
Distribution of viral antigen in the CNS is similar in all hosts but does not always reflect the localization of the lesions. The uniformity of virus distribution between hosts suggests the involvement of receptor molecules highly preserved among animal species; these are thought to belong to the C-type lectins and glycans families, although specific molecules still need to be identified [104]. The discrepancy between viral antigen distribution and the location of lesions supports the theory of an immunologic component in the pathogenesis of WNV disease. In horses has been observed either presence of viral antigen in absence of microscopic changes or low amount of antigen compared to lesion severity, both findings suggest that the pathological changes may follow the virus distribution in time and may therefore be seen even without the presence of virus antigen [35,44]. Lesions and viral tropism in the CNS associated with USUV infection are reported only in birds and are comparable to what is observed in birds infected with WNV. The similarity between the two viruses in the histopathological patterns of lesions and viral antigen distribution in the CNS provides insightful indications to drive future investigations on the pathogenesis of USUVassociated disease and, at the same time, can represent a diagnostic challenge raising the need to develop specific techniques for the differentiation of the two agents. This is difficult due to the high chance of cross-reaction between closely related flaviviruses belonging to the same serocomplex, as in this case.
Neuronal dysfunction or loss in specific areas of the brain can be linked to the clinical signs observed in both humans and animals. Involvement of cerebellar and vestibular nuclei of the brainstem that are responsible for the control of motor activity, may cause incoordination of movements, such as ataxia and dysmetria. Facial paresis, observed in humans, could be linked to the affection of the facial nucleus in the brainstem and damage to the dopaminergic neurons of the substantia nigra could cause tremors due to loss of modulation on motor function. Impairment of the motor neurons of the ventral horns of the spinal cord can be associated with clinical evidence of paraparesis [105]. Seizures are commonly associated with damage to cerebral cortex, that for WNV infection has been observed only in psittacine. Nevertheless, seizures could be additionally explained by the more widespread effect of IL-6 and TNF-α produced by macrophages in response to viral infection in the CNS. These cytokines increase neural hyperexcitability by disrupting the neuronal excitation/inhibition balance [106]. The same symptoms have been observed during USUV infection in humans, for which neuropathology is still to be investigated. Since the close relatedness of the two viruses, the comparison of similar neurological symptoms observed in USUV neuroinvasive disease and in WNND, may suggest a similar neurolocalization and pattern of the lesions for USUV in humans.

WNV and USUV extra neurologic disease
In addition to the changes in the CNS, necrosis and lymphoplasmacytic inflammation mainly affecting heart, liver, spleen, kidney and eye, represent a typical pattern of lesions during WNV and USUV infection in birds. In the accidental hosts, extra neurological lesions are less commonly reported, even though they can sometimes be a prominent feature of the infection, such as myocarditis in canids and marsupials or renal necrosis in reptiles. Similar pathologic changes are observed in multiple organs in course of avian malaria, posing this disease in the differential diagnosis of flaviviral infection [86]. Furthermore, due to the high prevalence of haemoprtozoa in birds, co-infection with either WNV or USUV is common and needs further investigation to understand its implications in the pathology and pathogenesis of these infections. Viral antigen in both definitive and accidental animal host is widely distributed in several extra neurological tissue and cell types also in albescence of microscopic changes sustaining the previously discussed immune-mediated tissue damage. The differences in lesions and viral antigen distribution between avian and mammalian species could be due to (1) differences in the host immune reaction: in humans a widespread distribution of WNV antigen in several non-CNS organs has been observed in immunocompromised patients showing WNND [91]; or (2) the levels of viremia: it is known that birds develop higher levels of viremia compared to other hosts such as humans in which WNV is rarely isolated from patients with WNND [91].
Although vascular changes such as vasculitis and fibrinoid necrosis are common additional findings in WNV infection, a histopathological pattern, in which vascular lesions are predominant, is observed in house sparrow nestlings and is also reported as main finding in a killer whale that died acutely without premonitory signs. In both the birds and the mammal, the pattern was associated with minimal inflammation that can indicate a rapidly fatal course of the infection in these animals [17,53]. In a fatal human case of fulminant hemorrhagic fever associated with WNV infection there was widespread endothelial positivity for WNV antigen in several organs including the CNS [107]. This evidence poses the hemorrhagic disease in the spectrum of possible clinicpathological manifestations of WNV infection in both animals and humans.
Both WNV and USUV antigen has been highlighted by IHC in the cytoplasm of neurons of the myenteric plexuses in raptors and passerine birds with evidence of ganglion neuritis, that are never reported in mammals and humans, probably due to lack of systematic investigation of the GI tract in these hosts. Contribution of flaviviral infection to the development of intestinal dysmotility has been suggested in a mouse model, posing these viruses among the possible causes of GI dysmotility syndrome also in humans, highlighting the necessity of extending pathological and immunohistochemical investigation to other organs such as the GI tract in the accidental hosts [108].

Reservoir potential and surveillance implications
Clinical presentation of WNV disease in birds has been experimentally studied in several species, while data on USUVassociated disease mainly comes from passive surveillance on wild birds. Therefore, insight into the clinical progression of USUV-associated disease is lacking. Blackbirds and great grey owls are considered particularly susceptible species to USUV infection mainly due to the high mortality rate observed during the outbreaks of the disease. However, highly susceptible species might not be the best reservoir of the infection in nature. To define a species as reservoir, several indicators are required; among them there is the susceptibility to natural infection [109] and the ability to maintain a sufficient level of viremia long enough to infect a new mosquito. Animals that die acutely are less likely to spread the infection and animals that may chronically carry the infection with a high viremic level are difficult to identify through surveillance on wildlife due to the ethical concerns and practical difficulties posed by repeated sampling.
Currently, tissues (found dead birds), blood and swabs (live caught birds) are collected as samples for monitoring and surveillance. However, these samples are often difficult to obtain and invasive for live birds. Determining the viral tropism in tissues through IHC provides insights in the locations of the lesions and the infected cell types in order to improve surveillance in dead and live birds. As was shown in previous studies, USUV and WNV antigen have been demonstrated in mature feather follicles of naturally infected birds [23,59], suggesting feathers to be an easy, effective and non-invasive method for testing in monitoring and surveillance.

Future perspective
The effect of climate change and increasing urbanization is changing the dynamics of mosquito-borne diseases, making them more likely to spread across countries where they were not commonly observed before. With the outbreaks of WNV and USUV in birds and the presence of WNV and USUV in mosquitos, mammals and humans, it is important to be prepared for increased incidence of human cases. Therefore, a One Health approach to improve preparedness strategies to deal with future outbreaks is needed. The identification of the host range of these viruses can give important information to improve monitoring and surveillance, including an early warning system in case of an outbreak with reliable diagnostic methods and relevant therapeutics and perhaps more preventive measures. For early warning systems the use of sentinel birds is crucial in detecting and monitoring mosquito-borne flaviviruses. To determine the appropriate sentinel species but also to develop therapeutic and preventive measures, knowledge on the clinical and pathological features as well as the pathogenesis in amplifying and incidental hosts for WNV and USUV is essential.

Funding
This publication is part of the project 'Preparing for Vector-Borne Virus Outbreaks in a Changing World: a One Health Approach' (NWA.1160.1S.210), which is (partly) financed by the Dutch Research Council (NWO).

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
No data was used for the research described in the article.

Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi. org/10.1016/j.onehlt.2023.100525.