Many arboviruses, including viruses of the Flavivirus genus, are known to cause severe neurological disease in humans, often with long-lasting, debilitating sequalae in surviving patients. These emerging pathogens impact millions of people worldwide, yet still relatively little is known about the exact mechanisms by which they gain access to the human central nervous system. This review focusses on potential haematogenous and transneural routes of neuroinvasion employed by flaviviruses and identifies numerous gaps in knowledge, especially regarding lesser-studied interfaces of possible invasion such as the blood–cerebrospinal fluid barrier, and novel routes such as the gut–brain axis. The complex balance of pro-inflammatory and antiviral immune responses to viral neuroinvasion and pathology is also discussed, especially in the context of the hypothesised Trojan horse mechanism of neuroinvasion. A greater understanding of the routes and mechanisms of arboviral neuroinvasion, and how they differ between viruses, will aid in predictive assessments of the neuroinvasive potential of new and emerging arboviruses, and may provide opportunity for attenuation, development of novel intervention strategies and rational vaccine design for highly neurovirulent arboviruses.
11 September 2025 in Clinical Microbiology Reviews
SUMMARY: Mosquito-borne orthoflaviviruses pose an ever-increasing risk to public health. A prime example is West Nile virus (WNV), which continues to expand its geographical distribution, leading to new outbreaks of severe disease in humans and animals. In contrast, the closely related Usutu virus (USUV) has rarely been associated with neurological disease despite its wide distribution in Europe where human cases of WNV-associated disease largely outnumber those of USUV. Thus, the risk to human health posed by USUV appears to be lower compared with WNV. However, considering their close relation and similarities in ecology, it is still not known why the clinical impact differs for these viruses. In this review, we explore the potential pathogenesis of USUV in human infection by comparing the clinical presentations reported in the few cases of USUV-associated disease with the extensive clinical and pathological data available for WNV. With such a comparison, we aim to determine the clinical significance of USUV and identify key differences between WNV and USUV, which could explain the disparity in the incidence of neurological disease. We identify that while the clinical reports of USUV infection do indicate involvement of similar brain regions as WNV, clinical presentations are not as severe and not as frequent and do not seem to progress to the same extent as many cases of WNV-related neuroinvasive disease. Additionally, we highlight the urgent need to standardize reporting of (arbo)viral meningoencephalitis cases to allow for a better comparison and understanding of viral pathogenesis and the risks posed by new and emerging arboviruses.
Vertebrate animals and many small DNA and single-stranded RNA viruses that infect vertebrates have evolved to suppress genomic CpG dinucleotides. All organisms and most viruses additionally suppress UpA dinucleotides in protein-coding RNA. Synonymously recoding viral genomes to introduce CpG or UpA dinucleotides has emerged as an approach for viral attenuation and vaccine development. However, studies that investigate the effects of this recoding strategy on viral replication and pathogenesis in vivo are still limited. Flaviviruses, including West Nile virus (WNV), are transmitted between vertebrate hosts by invertebrate vectors. In humans, WNV infection can cause flu-like symptoms and neuroinvasive disease. We investigated how alterations in WNV dinucleotide frequencies impact virus replication, transmission by vector mosquitoes, as well as pathogenesis and neuroinvasiveness in vertebrates. In Culex pipiens vector mosquitoes and Culex cell lines, only WNV with elevated UpA frequencies displayed attenuated replication. In vertebrate cell lines and primary human neuro-astrocyte co-cultures, both UpA and CpG enrichment reduced viral replication. In mice, the CpG-high WNV mutant demonstrated partial attenuation with delayed weight loss compared with wild-type WNV, although infection still resulted in 100% mortality. In contrast, 75% of animals survived inoculation with the UpA-high WNV mutant and were protected against wild-type WNV challenge. Notably, all animals that succumbed to infection had similar levels of virus in the brain, irrespective of the WNV mutant. Our results underscore the complex interplay between viral genome composition and host immune responses, highlighting potential safety concerns for dinucleotide manipulation as a strategy for live-attenuated vaccine development in flaviviruses. IMPORTANCE: Flaviviruses such as West Nile virus (WNV) pose significant public health concerns due to their potential to cause severe neurological disease. Synonymously recoding flavivirus genomes to introduce CpG or UpA dinucleotides has emerged as an approach for viral attenuation and vaccine development. However, the in vivo effects of manipulating these frequencies across the complete transmission cycle remained unexplored. Our study provides comprehensive insights into how CpG and UpA recoding affects WNV replication in both the mosquito vector and vertebrate hosts. We demonstrate that elevated UpA content attenuates virus replication throughout the transmission cycle, whereas CpG enrichment only impacts replication in the vertebrate host. Although UpA-high WNV shows significant attenuation and provides protection against wild-type infection, animals that succumb exhibit similar brain viral loads as wild-type infections. These findings have critical implications for live-attenuated vaccine development based on dinucleotide manipulation, specifically highlighting the importance of carefully evaluating the risk of neuroinvasion.
16 April 2026 in Journal of Investigative Dermatology
Rabies is a fatal zoonotic infection caused by rabies virus (RABV), which is transmitted through the saliva of infected animals. Although most cases are caused by dog bites, superficial exposures such as bat bites or scratches can also lead to infection, although the underlying mechanisms remain poorly understood. We recently detected RABV-positive keratinocytes in skin samples of experimentally infected mice and naturally infected dogs, prompting us to investigate their role in rabies pathogenesis and the antiviral immune response. We confirmed that both the keratinocyte cell line HaCaT and primary keratinocytes are susceptible to RABV infection in vitro, with higher infection rates observed for the attenuated SAD P5/88 Potsdam and silver-haired bat RABV strains than for a dog-related strain. Keratinocytes showed a significant immune activation in response to these 2 strains, as measured by expression of relevant surface markers and cytokine release. To explore whether RABV can be transmitted from infected keratinocytes to neurons, we developed a coculture model using HaCaT and a neuronal cell line separated by a microporous membrane. Virus transfer to the neurons was observed while the barrier integrity remained intact, suggesting that keratinocytes may contribute to RABV transmission and neuroinvasion after superficial exposure.
West Nile virus (WNV) leads to thousands of cases of severe neurological disease in humans each year. Usutu virus (USUV) is closely related to WNV, but rarely induces disease in humans. We hypothesised that USUV is less able to cross the blood-brain barrier, and is therefore less likely to infect the brain. Therefore, we developed an in vitro BBB model consisting of primary human brain microvascular endothelial cells (BMECs), pericytes and astrocytes. Both USUV and WNV invaded across the in vitro BBB via a transcellular mechanism in absence of barrier disruption. USUV replicated to lower titres than WNV but induced a comparable cytokine and chemokine response, with modulation of key factors associated with barrier function and immune-cell migration. In conclusion, USUV appears attenuated in its ability to replicate at this interface compared with WNV, but further work must to done to identify key determinants underlying the differing clinical presentations.
Vertebrate animals and many small DNA and single-stranded RNA viruses that infect vertebrates have evolved to suppress genomic CpG dinucleotides. All organisms and most viruses additionally suppress UpA dinucleotides in protein coding RNA. Synonymously recoding viral genomes to introduce CpG or UpA dinucleotides has emerged as an approach for viral attenuation and vaccine development. However, studies that investigate the effects of this recoding strategy on viral replication and pathogenesis in vivo are still limited. Flaviviruses including West Nile virus (WNV) are transmitted between vertebrate hosts by invertebrate vectors. In humans, WNV infection can cause flu-like symptoms and neuroinvasive disease. We investigated how alterations in WNV dinucleotide frequencies impact virus replication, transmission by vector mosquitoes, as well as pathogenesis and neuroinvasiveness in vertebrates. In Culex pipiens vector mosquitoes and Culex cell lines only WNV with elevated UpA frequencies displayed attenuated replication. In vertebrate cell lines and primary human neuro-astrocyte co-cultures both UpA and CpG enrichment reduced viral replication. In mice, the CpG-high WNV mutant demonstrated partial attenuation with delayed weight loss compared to wild-type WNV, though infection still resulted in 100% mortality. In contrast, 75% of animals survived inoculation with the UpA-high WNV mutant and were protected against wild-type WNV challenge. Notably, all animals that succumbed to infection had similar levels of virus in the brain, irrespective of the WNV mutant. Our results underscore the complex interplay between viral genome composition and host immune responses, highlighting potential safety concerns for dinucleotide manipulation as a strategy for live-attenuated vaccine development in flaviviruses. Importance: Flaviviruses such as West Nile virus (WNV) pose significant public health concerns due to their potential to cause severe neurological disease. Synonymously recoding flavivirus genomes to introduce CpG or UpA dinucleotides has emerged as an approach for viral attenuation and vaccine development. However, the in vivo effects of manipulating these frequencies across the complete transmission cycle remained unexplored. Our study provides comprehensive insights of how CpG and UpA recoding affects WNV replication in both the mosquito vector and vertebrate hosts. We demonstrate that elevated UpA content attenuates virus replication throughout the transmission cycle, while CpG enrichment only impacts replication in the vertebrate host. Although UpA-high WNV shows significant attenuation and provides protection against wild-type infection, animals that succumb exhibit similar brain viral loads as wild-type infections. These findings have critical implications for live-attenuated vaccine development based on dinucleotide manipulation, specifically highlighting the importance of carefully evaluating the risk of neuroinvasion.