Fitness adaptations of Japanese encephalitis virus in pigs following vector-free serial passaging

Japanese encephalitis virus (JEV) is a zoonotic mosquito-transmitted Flavivirus circulating in birds and pigs. In humans, JEV can cause severe viral encephalitis with high mortality. Considering that vector-free direct virus transmission was observed in experimentally infected pigs, JEV introduction into an immunologically naïve pig population could result in a series of direct transmissions disrupting the alternating host cycling between vertebrates and mosquitoes. To assess the potential consequences of such a realistic scenario, we passaged JEV ten times in pigs. This resulted in higher in vivo viral replication, increased shedding, and stronger innate immune responses in pigs. Nevertheless, the viral tissue tropism remained similar, and frequency of direct transmission was not enhanced. Next generation sequencing showed single nucleotide deviations in 10% of the genome during passaging. In total, 25 point mutations were selected to reach a frequency of at least 35% in one of the passages. From these, six mutations resulted in amino acid changes located in the precursor of membrane, the envelope, the non-structural 3 and the non-structural 5 proteins. In a competition experiment with two lines of passaging, the mutation M374L in the envelope protein and N275D in the non-structural protein 5 showed a fitness advantage in pigs. Altogether, the interruption of the alternating host cycle of JEV caused a prominent selection of viral quasispecies as well as selection of de novo mutations associated with fitness gains in pigs, albeit without enhancing direct transmission frequency.


S4 Methods: Construction of JEV RNA for quantitative RT-PCR
For PCR amplification from JEV cDNA, Phusion Hot Start II DNA Polymerase (ThermoFisher Scientific), dNTP mix (ThermoFisher Scientific), the forward primer 5' -CTC ATC GGA ACG CGA TCC AG -3' and the reverse primer 5'-TAG ACG AGG CTT GGG CCG TGG G -3' were used.The PCR followed 32 cycles, with an annealing temperature of 57°C.For the adenylation of the PCR fragments, the reaction was incubated for 10 minutes at 72°C with a Taq polymerase (Thermo Fisher) prior to pCR4-TOPO cloning.The TOPO cloning was performed using the TOPO TA cloning Kit for Sequencing (Thermo Fisher).Briefly, after overnight incubation of the purified PCR product (4µl) with 1µl of pCR™4-TOPO TA vector, 2µl of the reaction was used to transform electrocompetent TOP10 E. coli cells.After kanamycin selection on LB-agar plates, single colonies were used to inoculate liquid LB-kanamycin cultures for the amplification and subsequent isolation of plasmid DNA.To select positive clones, the plasmids were digested with the restriction enzyme EcoRI (ThermoFisher), and the resulting DNA fragments were separated by gel electrophoresis.To screen for the correct orientation of the insert within the plasmid, miniprep DNA preparations were digested with the restriction enzymes Tth111l and NotI (Thermo Fisher) and Sfil/NotI (Thermo Fisher).The final construct was verified by Sanger sequencing using pCR4-TOPOspecific M13_F and M13_R oligonucleotides.The RNA was synthesized using the T7 RNA Polymerase (MEGAscript T7 Transcription Kit, Ambion) with a Bcul linearized, phenol/chloroform-extracted plasmid template.The RNA was further purified by two rounds of RNA extraction using the NucleoSpin RNA Kit (Macherey Nagel).The purity and integrity of the RNA was analyzed by gel electrophoresis.The concentration of the RNA was determined using a Nanodrop 2000C Spectrometer (Thermo Fisher).

S5 Methods: Transcriptomics
To isolate blood leukocytes, EDTA blood was treated with a red blood cell lysis buffer (1.5M NH4Cl, 100mM NaHCO3, 10mM EDTA in H2O) at 37°C for 2.5 minutes to obtain complete erythrocyte lysis.The remaining cells were washed three times with cold Dulbecco′s Phosphate Buffered Saline (PBS, Thermo Fisher) before lysing the pellet in Trizol (Thermo Fisher).After the addition of a 49:1 mixture of chloroform (Sigma) and isoamyl alcohol (Sigma), the aqueous phase was transferred, and an equal volume of 75% ethanol (Merck) was added.Then, RNA was finally extracted by loading the mixture on the column of the Nucleospin RNA extraction Kit following the manufacturer's instructions.The RNA was then sent for Bulk RNA barcoding and sequencing (BRB-seq) at the sequencing center of the University of Bern.A quality control by fragment analysis (5200 Fragment Analyzer CE instrument, Agilent) was performed prior to the sequencing with the Illumina NovaSeq6000 sequencer (Illumina).Reads were mapped to the pig genome (Sus scrofa 11.1, Ensembl release) using Tophat v.2.0.11[4][5][6][7].The number of reads overlapping with each gene was evaluated with Htseq-count v.0.6.1 [8,9].The Bioconductor package DESeq2 v1.38.3 [10] was used to test for differential gene expression between the experimental groups.Gene set enrichment analysis (GSEA) was performed following ranking of genes based on differential gene expression using the "stat" value [11,12].Calculations of normalized enrichment scores and false discovery rates (FDR) was performed using online tools available on https://www.gsea-msigdb.org[13].As gene sets we employed the blood transcriptional modules (BTM) defined by Li et al. [14] for humans and the modified version for pigs as previously described [15].Figures were created in R 4.3.0 using ggplot2 Bioconductor package.

S6 Methods: Histopathology
Half of the brain (including one complete cerebral and cerebellar hemisphere) and cervical spinal cord, tonsil and mandibular lymph node were fixed in 4% buffered formalin for four weeks.For histopathological examination cross sections of collected tissues were performed, including representative areas of the central nervous system (CNS): olfactory bulb, frontal lobe, basal nuclei, hippocampus, cerebral cortex, thalamus, midbrain, pons, cerebellum and cervical spinal cord.The tissue sections were embedded in paraffin, cut at 4 µm of thickness and stained with hematoxylin and eosin (HE).Lesions in the CNS were semi-quantitatively scored from 0 to 3, as previously described [16], with minor modifications: 0 = no lesions, 1 = mild lesions, 2 = moderate lesions, and 3 = severe lesions.
Grade 1 corresponds to few and thin perivascular cuffs and/or few glial nodules.Grade 2 is characterized by prominent perivascular cuffs and multiple dense multifocal glial nodules with neuronal degeneration and neuronophagia.Grade 3 corresponds to extended lesions with prominent perivascular cuffs and numerous glial nodules that may coalesce, and frequent evidence of neuronophagia.All scoring was done as blind trial.A sum of the scores per pig was calculated by adding all the scored areas as previously described.[16] S7 Methods: Serum neutralization assay Firstly, focus forming units of JEV Laos stocks were determined by tenfold serially diluting the virus stock in MEM starting at 1:100 dilution in quadruplicates.200µl of this viral dilution was added to confluent Vero cells, seeded in a 24-well plate.The inoculum was incubated for 1.5 hours at 37°C and 5% CO2.Thereafter, the cells were washed with pre-warmed D-PBS and incubated with 1ml of medium supplemented with 1% methylcellulose (Sigma Aldrich) at 37°C and 5% CO2 for 72 hours.Then, the overlay medium was aspirated, cells were washed once with pre-warmed PBS and fixed with 4% buffered formalin solution (Polyscience Inc).The virus infected cells were stained by an immuno-peroxidase staining, described in the S9 Method.The foci were counted and multiplied by the dilution to calculate the Focus forming units per milliliter (FFU/ml).
For the neutralization test, all sera samples were treated at 56°C for 20 minutes to inactivate the complement.Then, sera were serially twofold diluted in media, starting with a 1/10 dilution.Thereafter, 200 FFU/ well of JEV Laos were added and the plates were incubated for 1h at 37°C with 5%CO2.Each mix containing 100 FFU of JEV Laos was then pipetted to wells with confluent Vero cells and incubated for 48h at 37°C with 5% CO2.After removal of the supernatant, the cells were washed once with PBS, before fixation with 4% buffered formalin solution.This was followed by E protein immune-peroxidase staining, as described in S9 Methods.Finally, the plates were photographed using the Immunospot machine (Cellular Technology LTD), and the area of E protein staining was quantified for each well using Image J software.Wells with naïve sera were used as controls as negative controls, and wells were 100 FFU of JEV were added in absence of immune serum were included as positive controls.The serum dilution that allowed a reduction of at least 50% E protein staining was used as cut off.The 50% neutralizing dose (ND50) was calculated using the Spearman-Kaerber formula [17].

S8 Methods: Viral titration
For the titration of JEV from cell culture supernatants, the cells were seeded in a 96-well plate.Once cell confluency was reached, the medium was replaced.Then, virus samples were ten-fold serially diluted, starting at 1:10 dilution in quadruplicates and incubated on the cells for three days.The cells were washed with PBS and fixed with 4% formalin.Finally, virus-infected cells were labelled by an immunoperoxidase-staining as described in S9 Methods.Wells containing stained JEV-infected cells were counted as positive, while wells with unstained cells were counted as negative.The number of positive and negative wells were used to determine the TCID50/ml using the Reed-Muench formula [18,19]

S9 Methods: Immunoperoxidase staining for JEV infected cells
The virus-infected cells were visualized by immunoperoxidase staining using the mouse anti-Flavivirus E protein monoclonal antibody (ATCC HB 112, D1-4G2-4-15) diluted in 0.3% saponin (PanReac AppliChem) solution in PBS for 30 min at 37°C.After 3 washes with wash buffer (0.14M NaCl, 1.47mM KH2PO4, 7.8mM Na2HP4, 2.68mM KCl and 0.05% v/v Tween 20), the rabbit anti-mouse HRP (Agilent Technology) diluted in 0.3% saponin solution in PBS was added on the cells and incubated for another 30 min at 37°C, then washed 3 times, as before.Then infected cells were visualized after 30 min of incubation in AEC solution in the dark.The AEC solution contains 50mM Na-Acetate with 5% v/v 3-Amino-9-ethylcarbazol dissolved in N,N-dimethylformamide (all Sigma) and 5% v/v H2O2.

S10 Methods: Virus isolation from sera
To enhance efficiency of virus isolation from sera, the samples were diluted 1/20 in DMEM and mixed with immune sera (0.1 % v/v) from vaccinated pigs that was previously demonstrated to strongly enhance JEV infection.[20] After 1 hour at 37°C the sera mix was added to monocyte derived macrophages (MDM) for another 1.5 h.Then, cells were washed twice with pre-warmed PBS and incubated with infection media for 72h at 37°C, 5% CO2.This resulted in an effective JEV replication in MDM, allowing us to rescue the virus from all serum of viremic pigs.The rescued viruses were titrated like described in S8 Methods for the determination of the TCID50/ml.

S11 Methods: Virus sequencing
For the viral genome sequencing, RNA was extracted from serum or swab samples using the TRIzol™ (Thermo Fisher).After the addition of a 49:1 mixture of Chloroform (Sigma) and isoamyl alcohol (Sigma), the aqueous phase was transferred, and an equal volume of 75% ethanol (Merck) was added.The mixture was then loaded onto the RNA column of the NucleoSpin RNA extraction Kit (Macherey Nagel).The RNA was extracted following the guidelines of the kit.
The quantity and quality of the purified total RNA was assessed using the Qubit 4.0 fluorometer with the Qubit RNA BR & HS Assay Kits (Thermo Fisher), and an Advanced Analytical Fragment Analyzer System using a Fragment Analyzer RNA Kit (Agilent).Input RNA samples varying between 0.5ng-1000ng were first depleted of ribosomal RNA and globin mRNA using RiboCop for HMR + Globin Depletion Kit following the Lexogen user guide (144UG288V0101).Thereafter, cDNA libraries were generated using a CORALL Total RNA-Seq V2 library Prep.kit with UDIs 12nt set A1-A4 (Lexogen) according to the protocol for short insert sizes and with 11-19 PCR cycles (Lexogen user guide 171UG394V0100).The resulting cDNA libraries were evaluated using a Qubit 4.0 fluorometer with the Qubit dsDNA HS Assay Kit (Thermo Fisher) and an Agilent Fragment Analyzer (Agilent) with a HS NGS Fragment Kit (Agilent), respectively.In a pilot experiment pooled libraries were sequenced 50 bp paired end using an Illumina NovaSeq 6000 SP Reagent Kit v1.5 (100 cycles; Illumina, 200228401) on an Illumina NovaSeq 6000 instrument.For the main experiment and majority of samples, the cDNA libraries were equimolar-pooled and sequenced 100 bp paired-end using a shared Illumina NovaSeq 6000 S4 Reagent Kit v1.5 (200 cycles; Illumina) on an Illumina NovaSeq 6000 instrument.The quality of the sequencing run was assessed using Illumina Sequencing Analysis Viewer (Illumina version 2.4.7) and all base call files were demultiplexed and converted into FASTQ files using Illumina bcl2fastq conversion software v2.20.