Usutu Virus Africa 3 Lineage, Luxembourg, 2020

We detected Usutu virus in a dead Eurasian blackbird (Turdus merula) in Luxembourg in September 2020. The strain clustered within the Africa 3.1 lineage identified in Western Europe since 2016. Our results suggest maintenance of the virus in Europe despite little reporting during 2019–2020, rather than a new introduction.

Upon sample reception, nucleic acids were extracted from 140 µL of virus transport medium in which cloacal and tracheal swabs were stored immediately after sample collection using the QIAmp viral RNA mini kit (Qiagen, Venlo, the Netherlands). Tissues (30 mg; brain, liver, kidney, heart) (Table, https://wwwnc.cdc.gov/EID/article/28/5/21-2012-T1.htm) were lysed in 600 µL of RLT lysis buffer supplemented with 14.3 M β-mercaptoethanol in the TissueLyser II (Qiagen) with 5 mm steal beads for 2x 30 sec at 30 Hz. After centrifugation for 3 min at 16 000 g to remove tissue debris, nucleic acids were purified from the supernatant using RNeasy mini kit (Qiagen) according to the manufacturer's instructions. The PCR reaction and cycling conditions were similar to the initial USUV screening assay, except for 0.4 µM final concentration of primers and no additional MgCl2.

Page 2 of 4 USUV Genome Sequencing
RNA from the positive brain sample was first reverse transcribed into cDNA using Superscript III Reverse transcription (Invitrogen, Merelbeke, Belgium) and random hexamers (Invitrogen). The complete polyprotein coding sequence of USUV was then sequenced by amplifying overlapping amplicons using previously published (4) or newly designed primers (Appendix Table). Purified amplicons were sequenced on an ABI 3130 Avant capillary sequencer (Applied Biosystems) using PCR primers as sequencing primers. Contigs were assembled with Geneious Prime v2019.1.1 (Biomatters, Auckland, New Zealand) by aligning individual sequences to a similar reference strain (GenBank KY294723) identified by blasting amplicon sequences covering NS5. The consensus sequence of the complete polyprotein coding region was then used for phylogenetic analyses (GenBank accession no. OU674388).

Phylogenetic Analyses
All USUV sequences publicly available on 03.08.2021 were downloaded from GenBank into Geneious software. The curated sequence set (after removing sequences not belonging to USUV and clones) contained 843 partial and complete polyprotein coding sequences. All strains were renamed as follows: GenBank accession number_host/country/isolate/year. Given the nonuniform sequencing coverage obtained in various studies, phylogenetic analyses were performed on all complete polyprotein coding sequences (10,305 bp; n = 296) as well as partial sequences to obtain the best representation of the phylogenetic relationship of the strain from Luxembourg.
For this purpose, all sequences longer than 500 bp were aligned using MAFFT as implemented in Geneious software. The most commonly sequenced regions were included and comprised of partial NS5 (nucleotides 9088-9597, 510 bp; n = 376) and partial envelope sequences (nucleotides 1003-2067,1065 bp; n = 377). All sequences with unresolved regions (Ns) were removed from the alignments. Identical partial NS5 and envelope sequences were summarized by including only one unique sequence identified thanks to the DNACollapser tool available in FaBox (5). The best substitution model fitting each alignment was identified using MEGA v6.06 (6). The model with the lowest Bayesian Information Criterion, considered to best describe the substitution pattern observed in the alignment, was then implemented to calculate phylogenetic trees with the Maximum Likelihood method and 500 bootstrap replicates in MEGA v6.06.
Preliminary phylogenetic analyses showed that trees based on partial sequences lacked the power to define some lineages and the Africa 3 sublineages with confidence (bootstrap values at main Page 3 of 4 lineage defining nodes <75% and polyphyletic lineages) and led to some classification discrepancies (data not shown). Therefore further analyses focused on phylogenetic trees calculated on complete polyprotein coding sequences (Appendix Figure).