Dendritic transport of tick-borne flavivirus RNA by neuronal granules affects development of neurological disease

Abstract

Neurological diseases caused by encephalitic flaviviruses are severe and associated with high levels of mortality. However, little is known about the detailed mechanisms of viral replication and pathogenicity in the brain. Previously, we reported that the genomic RNA of tick-borne encephalitis virus (TBEV), a member of the genus Flavivirus, is transported and replicated in the dendrites of neurons. In the present study, we analyzed the transport mechanism of the viral genome to dendrites. We identified specific sequences of the 5' untranslated region of TBEV genomic RNA that act as a cis-acting element for RNA transport. Mutated TBEV with impaired RNA transport in dendrites caused a reduction in neurological symptoms in infected mice. We show that neuronal granules, which regulate the transport and local translation of dendritic mRNAs, are involved in TBEV genomic RNA transport. TBEV genomic RNA bound an RNA-binding protein of neuronal granules and disturbed the transport of dendritic mRNAs. These results demonstrated a neuropathogenic virus hijacking the neuronal granule system for the transport of viral genomic RNA in dendrites, resulting in severe neurological disease.

Keywords: dendritic mRNA; flavivirus; neuronal granule; neuropathogenicity; tick-borne encephalitis virus.

Fig. S1.

Genomic RNA of tick-borne encephalitis virus was transported to the neurites of infected PC12 cells. Differentiated PC12 cells were infected with TBEV and fixed at 48 or 72 h postinfection. The cells were stained with specific antibodies against a neurite marker (A; green), viral proteins (A and B; magenta), components of neuronal granules (C; magenta), and fluorescent RNA probes against viral genomic RNA (B and C; green). White arrows and arrowheads indicate viral protein and genomic RNA in neurites, respectively. (Scale bars, 5 μm.)

Fig. 1.

The 5′ untranslated region of tick-borne encephalitis virus functions as a signal of RNA transport to the neurites of PC12 cells. Differentiated PC12 cells were transfected with plasmids expressing the RNA of luciferase with TBEV sequences (A–C) or TBEV/WNV UTRs (D–F). Following fixation, the cells were hybridized with a fluorescent RNA probe for the luciferase gene (green), and stained with DAPI (blue) and antibodies against microtubule-associated protein 2 (MAP2; magenta). Fluorescence in situ hybridization signal in the neurites was analyzed from Z-stack images from five independent microscopic fields. (A) A coding sequence for luciferase (gray rectangles) was cloned with or without the partial sequence for TBEV replicon RNA. (B, C, E, and F) Fluorescent images (B and E) and fluorescence intensity (C and F) in PC12 neurites. (D) A CDS for luciferase was cloned with or without the 5′ and 3′ UTRs of TBEV (black lines) and WNV (striped lines). (Scale bars, 5 μm.) White arrows indicate the FISH signal for luciferase RNA in the neurites. Error bars represent SEM; **P < 0.02 and *P < 0.05.

Fig. S2.

Luciferase activity of the constructs expressing luciferase mRNA with or without flavivirus untranslated regions. Human embryonic kidney 293T cells were transfected with plasmids expressing luciferase mRNA with or without flavivirus UTRs. Deletion of the 3′ UTR (A) or mutations or substitutions in the 5′ and 3′ UTRs (B) were introduced into the plasmids. Following 48 h of incubation, cells were collected and luciferase activity was measured (n = 3). **P < 0.02.

Fig. 2.

Analysis of the roles of the stem-loop structure of the TBEV 5′ UTR in genome transport. Differentiated PC12 cells were transfected with plasmids expressing the mRNA of luciferase with the UTRs of TBEV with deletion (A–C) or mutation (D–F) of the 5′ UTR. Following fixation, the cells were hybridized with a fluorescent RNA probe for the luciferase gene (green), and stained with DAPI (blue) and antibodies against MAP2 protein (magenta). FISH signal in the neurites was analyzed from Z-stack images from five independent microscopic fields. (A) Schematic diagram of the predicted RNA secondary structure (Upper) and the constructs expressing mRNA with a deletion (Lower) are shown. The 5′ UTR has a predicted branched stem-loop structure (SL-1) and single-SL structure (SL-2). The SL-1 or SL-2 regions were deleted in pCMV-Luc (5′ TBEV/3′ TBEV). (B, C, E, and F) Fluorescent images (B and E) and fluorescence intensity (C and F) in PC12 neurites. (D) Schematic diagrams of the sequence and RNA secondary structure of TBEV SL-2 and the constructs used to analyze the role of SL-2 in transport. G-to-U and C-to-U in the loop, or four mutations in the stem, were introduced into pCMV (5′ TBEV/3′ TBEV). (Scale bars, 5 μm.) White arrows indicate the FISH signal for luciferase RNA in neurites. Error bars represent SEM; **P < 0.02 and *P < 0.05.

Fig. S3.

Prediction of the RNA secondary structure of TBEV. RNA secondary structures of the TBEV Oshima 5-10 strain (nucleotides 1 to 240) were predicted by mfold. The initiation codon of the viral coding sequences are underlined and in bold.

Fig. 3.

Mutation impeding genome transport to dendrites attenuated the neurological symptoms caused by TBEV infection. (A and B) Primary mouse neurons were infected with TBEV wild type (black squares with continuous lines) or SL-2 loop C-U (white circles with broken lines) at a multiplicity of infection (MOI) of 0.1. (A) The cells were fixed at 48 h.p.i. and the viral proteins and viral genomic RNAs were stained by indirect immunofluorescence assay (Upper; magenta) and FISH (Lower; green), respectively. (Scale bars, 5 μm.) White arrows indicate viral antigen accumulation or the viral genome in dendrites. (B) Viral antigen accumulation was counted at 24, 48, or 72 h.p.i. in five independent microscopic fields. (C–E) Five-week-old male C57BL/6 mice were inoculated with 100 plaque-forming units (PFUs) of TBEV WT or SL-2 loop C-U intracerebrally. (C) The Kaplan–Meier survival estimate was calculated (n = 10). (D) The neurological score of the mice (n = 5) was examined until 8 d.p.i. (E) The mice were killed at 3 or 6 d.p.i. (n = 3), and the viral titer in the brain was analyzed. Continuous and broken lines indicate the average of viral titer in the brain infected with TBEV WT and SL-2 loop C-U, respectively. Error bars represent SEM; **P < 0.02 and *P < 0.05.

Fig. S4.

Viral growth of TBEV WT and SL-2 loop C-U in primary neurons. Primary mouse neurons were infected with TBEV wild type (black squares with continuous lines) or SL-2 loop C-U (white circles with broken lines) at a multiplicity of infection of 0.1. The culture supernatant was collected at 24, 48, or 72 h.p.i., and virus titer was measured with a plaque-forming assay. Error bars represent SEM; **P < 0.02 and *P < 0.05.

Fig. 4.

Localization of the RNA-binding proteins of a neuronal granule in a neuron infected with TBEV. Primary mouse neurons were uninfected or infected with TBEV WT or SL-2 loop C-U at an MOI of 0.1. The cells were fixed at 48 h.p.i. and stained with antibodies against fragile X mental retardation protein, RNA granule protein 105, or Staufen (green), and antibodies against viral proteins (magenta). (A) Fluorescent images of the neurons. (B) The signals of RBPs in the cell body or neurites were analyzed in Z-stack images of five microscopic fields. FI, fluorescence intensity. Error bars represent SEM; **P < 0.02 and *P < 0.05.

Fig. S5.

Detailed localization of RNA-binding proteins of neuronal granules in dendrites of neurons infected with TBEV. Primary mouse neurons were infected with TBEV WT or SL-2 loop C-U at a multiplicity of infection of 0.1 and fixed at 48 h postinfection. The cells were stained with antibodies against FMRP, RNG105, or Staufen (green) and antibodies against viral proteins (magenta). (A) A differential interference contrast microscope image (DIC; Left) and fluorescent image of RNG105 and viral proteins are shown (Right). White continuous and broken arrows show the paths of RNG105-positive and -negative neurites, respectively. (B) Antigen accumulations in RNG105-positive or -negative neurites were counted in five microscopic fields independently. Statistical differences were assessed with the Student’s t test. Error bars represent SEM; *P < 0.05. (C) Fluorescence intensity (FI) line profiles of neurites infected with TBEV WT are defined with gray arrows. (Scale bars, 5 μm.)

Fig. S6.

Colocalization of RBPs and viral RNAs in dendrites. Primary mouse neurons were infected with TBEV WT or SL-2 loop C-U at an MOI of 0.1. Following fixation at 72 h.p.i., the cells were stained with antibodies against FMRP, RNG105, or Staufen (green) and fluorescent probes against viral RNA (magenta). (A) Fluorescent images of the neurons. Colocalized areas of RBPs and viral genomic RNA are shown in white (Right). (Scale bars, 5 μm.) (B) Colocalized areas of RBPs and viral genomic RNA in dendrites were measured in 10 AOIs. Error bars represent SEM; **P < 0.02 and *P < 0.05.

Fig. 5.

Interaction between the RBP of a neuronal granule and the genomic RNA of TBEV. Full-length RNAs of TBEV WT or SL-2 loop C-U (A) was mixed with cell lysate expressing Flag-FMRP WT or I304N (B). The mixture was immunoprecipitated (IP) with beads with anti-Flag antibody (Flag) or beads only (Control), and precipitated protein and RNA were detected by Western blotting (WB) and RT-PCR, respectively. (B, Right) Expression of the FMRP WT or I304N in total cell lysate.

Fig. 6.

TBEV infection and transport of viral RNA disrupted the localization of dendritic mRNAs. Primary mouse neurons were infected with TBEV WT or SL-2 loop C-U at an MOI of 0.1. (A) The cells were fixed at 48 h.p.i., and mRNA for Arc, brain-derived neurotropic factor, or Ca²⁺/calmodulin-dependent protein kinase II α was stained by FISH (green). (Scale bars, 5 μm.) (B) Fluorescent signal of the mRNA for Arc, BDNF, or CaMKIIα in dendrites was measured in 10 areas of interest (AOIs). Error bars represent SEM; **P < 0.02 and *P < 0.05.

Fig. S7.

Dendritic mRNA expression in primary neurons infected with TBEV. Primary neurons were infected with TBEV WT or SL-2 loop C-U. At 72 h postinfection, total RNAs were extracted. mRNAs for Arc, brain-derived neurotrophic factor, Ca²⁺/calmodulin-dependent protein kinase II α, or beta-actin and genomic RNA of TBEV were detected by reverse transcription–PCR.

Fig. S8.

Comparison of the stem-loop 2 sequence among flaviviruses. Alignment of the SL-2 sequences of tick-borne and mosquito-borne flaviviruses. AHFV, Alkhurma hemorrhagic fever virus; DENV2, dengue virus serotype 2; KFDV, Kyasanur forest disease virus; LGTV, Langat virus; LIV, louping ill virus; OHFV, Omsk hemorrhagic fever virus; POWV, Powassan virus; ZIKV, Zika virus. The squared region indicates conserved sequences.

Fig. S9.

Model of TBEV genomic RNA transport and dendritic dysfunction. Dendritic mRNAs bound to RBPs of the neuronal granule are transported and locally translated in dendrites for maintenance of synaptic plasticity. In neurons infected with TBEV, viral genomic RNAs bind to RBPs (such as FMRP) via the SL-2 region and are transported to dendrites. This transport disturbs the transport of dendritic mRNAs. The transported viral genomic RNAs alter RBP distribution, and the local replication initiates the degeneration of dendrites, resulting in neuronal dysfunction.