Molecular and antigenic characterization of group C orthobunyaviruses isolated in Peru

Group C orthobunyaviruses (GRCVs) are a complex of viruses in the genus Orthobunyavirus and are associated with human febrile disease in tropical and subtropical areas of South and Central America. While numerous GRCVs have been isolated from mosquitoes, animals, and humans, genetic analysis of these viruses is limited. In this study, we characterized 65 GRCV isolates from febrile patients identified through clinic-based surveillance in the northern and southern Peruvian Amazon. A 500 base pair region of the S segment and 750 base pair regions of the M and L segments were sequenced. Pairwise sequence analysis of the clinical isolates showed nucleotide identities ranging from 68% to 100% and deduced amino acid sequence identities ranging from 72% to 100%. Sequences were compared with reference strains of the following GRCVs: Caraparu virus (CARV), Murutucu virus (MURV), Oriboca virus (ORIV), Marituba virus (MTBV), Itaqui virus (ITQV), Apeu virus (APEUV), and Madrid virus (MADV). Sequence comparison of clinical isolates with the prototype strains based on the S and L segments identified two clades; clade I included isolates with high genetic association with CARV-MADV, and clade II included isolates with high genetic association with MURV, ORIV, APEUV, and MTBV. Genetic relationships based on the M segment were at time inconsistent with those based on the S and L segments. However, clade groupings based on the M segment were highly consistent with relationships based on microneutralization assays. These results advance our understanding of the genetic and serologic relationships of GRCVs circulating in the Peruvian Amazon.


Introduction
, the Servico Especial de Saude Publica of Brazil and the Rockefeller Foundation conducted studies to isolate arboviruses in the Brazilian Amazon region and isolated a a1111111111 a1111111111 a1111111111 a1111111111 a1111111111

Virus culture and RNA extraction
Serum samples from febrile patients were inoculated onto Vero cell cultures and identified as GRCVs using an immunofluorescent assay as previously described [13]. Six GRCV reference strains obtained from Dr. Robert Tesh and the World Reference Center for Emerging Viruses and Arboviruses at the University of Texas Medical Branch, Galveston, TX and 65 isolates ( Table 1 and S1 Table) were propagated in Vero cells and harvested when 70-100% of cytopathic effect was exhibited. Infected supernatant were centrifuged at 2000 g at 4˚C for 20 minutes, and the clarified supernatants were recovered and stored in cryovials at -70˚C. Viral RNA was extracted from viral culture clarified supernatant using the QIAamp Viral RNA Kit (Qiagen, Valencia, CA).

Reverse transcription and cDNA amplification
More than 500 base pair region of the S segment (coding sequence of partial N and NSs protein) and more than 750 base pair regions of the M (coding sequence of partial polyprotein, corresponding to GC site) and L (coding sequence of partial RdRp) segments were amplified using one step reverse transcription and PCR (RT-PCR). Primers (Table 2) were designed based on previous published sequences [15,16].
RT-PCR was performed in a total volume of 50 μl containing 1 μM of each primer (forward and reverse; Table 2 For the partial S segment, reverse transcription was performed for 60 minutes at 42˚C and  95˚C for 2 minutes, followed by 35 PCR cycles, consisting of 94˚C for 40 seconds, 50˚C for 40  seconds, and 72˚C for 50 seconds, and finally at 72˚C for 5 minutes. For the partial M and L segments, reverse transcription was performed for 60 minutes at 42˚C and 95˚C for 2 minutes, followed by 40 PCR cycles, consisting of 94˚C for 40 seconds, 45˚C for 40 seconds, and 72˚C for 1 minutes, and finally at 72˚C per 10 minutes. The amplified products were visualized under UV light after electrophoresis in a 2.0% agarose gel for the S segment and a 1.5% agarose gel for the M and L segments.

Nucleotide sequencing
The cDNA amplicons from the partial S segment were purified using centri-sep columns (Princeton Separations, Inc.) after a single band was apparent in the gel. The amplicons from the partial M and L segments were recovered from excised DNA bands from the agarose gel and purified using Wizard SV Gel and PCR Clean-Up System (Promega, Madison, WI, USA). Sequencing was performed using BigDye1 Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) and the same primers as used for RT-PCR. Sequencing reactions were purified using centri-sep columns (Princeton Separations, Inc.) or an ethanol precipitation method. Nucleotide sequences were determined by running purified sequencing reactions on the ABI 3130 XL Genetic Analyzer (Applied Biosystems).

Sequence analysis and phylogeny
Nucleotide sequences were assembled using Sequencher 4.8 software (Gene Codes Corporation, Ann Arbor, MI). In addition to the sequences that were generated for this study (65 isolates and 6 reference strains), the following sequences from GenBank were included in the genetic analysis: Madrid Virus-BT4075 (GenBank accession numbers KF254781, KF254780 and KF254779) [16] and IQT9646 (GenBank accession numbers KM092514, KM092513, KM092512) [17].
ClustalW multiple alignment, pairwise comparisons, multiple sequence analyses and phylogenetic tree construction were performed using BioEdit 7.2 [18] and MEGA 5.2 [19] software. Trees were constructed using the neighbor-joining method and distances were calculated using Kimura 2-parameter model with bootstrap analysis using 1000 replicates.

Neutralization test
Hyperimmune mouse ascitic fluid (HMAF) against APEUV BeAn848, CARV BeAn3994, ITQV (isolate PE-00036, collected from a mosquito pool in Peru [20]), MTBV BeAn15, MURV BeAn974 and ORIV BeAn17 was prepared by serial immunization of 6-8 week old BALB/c female mice as previously described [21,22]. The six HMAFs produced (HMAF-A-PEUV, HMAF-CARV, HMAF-ITQV (PE00036), HMAF-MTBV, HMAF-MURV and HMA-F-ORIV) were heat inactivated at 56˚C for 30 minutes, and using 96 well plates, two fold serial dilutions were made (from 1/20 to 1/2560) using cell culture media (50 μl of final volume for each dilution). Equal volumes of the heat inactivated serially diluted HMAF and each previously diluted virus isolate was mixed and incubated for 1 hour at 37˚C to allow neutralization. The virus dilution used was determined by a preliminary titration, which contained between 30 and 300 TCID 50 , where one TCID 50 is the dilution of the original stock which results in half the wells being infected and half the wells remaining uninfected, calculated using the method of Reed and Muench [23]. After incubation, virus and HMAF were mixed with 100 μl of Vero 76 cells at a concentration of 5 X 10 5 / ml in cell culture media with 4% Fetal Bovine serum (FBS). The plates were incubated at 37˚C in 5% CO 2 . After 3 days, supernatant media were  removed and the cells were fixed and stained with a solution of naphthol blue black and acetic acid [24]. The neutralization titer of test sample was reported as the last dilution which cells are protected from cytopathic effect (last well in which at least 50% of the monolayer was protected, as identified by visual inspection of the monolayer).

S segment sequence analysis
Based on partial S segment sequence, GRCV isolates grouped into two main clades (Fig 2). Pairwise comparisons between isolates from clade I and clade II ranged from 69% -75% nucleotide identity and 75% -77% amino acid identity (Table 3). Clade I comprised CARV, MADV, ITQV, and 49 clinical isolates (including one previously described isolate [17]). Clade I clinical isolates shared 86% -88% nucleotide identity and 96% -97% amino acid identity with CARV. Isolates from northeastern Peru and southeastern Peru segmented into distinct lineages within clade I. One lineage was comprised of 35 isolates from northeastern Peru (Iquitos and Yurimaguas), and the second lineage was comprised of one isolate from northeastern Peru and 13 isolates from southeastern Peru (Puerto Maldonado). Clade II comprised ORIV, MURV, APEU, and MTBV strains and 17 clinical isolates, all from northeastern Peru. Isolates shared 83% -84% nucleotide identity and 97% amino acid identity with MTBV.

M segment sequence analysis
M segment sequences segregated into three clades (Fig 3). Clade I comprised CARV, MADV, APEUV and 47 clinical isolates (including one previously described isolate [17]). Clade I further separated into geographically stratified groupings related to CARV with 79% -84% nucleotide identity and 90% -99% amino acid identity (Table 4). Clade I clinical isolates separated into three lineages. Lineage 1 included a cluster of 30 isolates from northeastern Peru. Lineage 2 included a cluster of 11 isolates from southeastern Peru. Lineage 3 included a cluster of 6 isolates from both (northeastern-southeastern Peru). Clade II included MURV, MTBV, and 17 clinical isolates from northeastern Peru, clinical isolates shared 81% -82% nucleotide identity replicates on the bases of partial coding sequence of N and NSs protein. The evolutionary distances were computed using the Kimura 2-parameter method. Members of the Simbu, California and Bunyamwera serogroups were used as outgroups to root the tree.
https://doi.org/10.1371/journal.pone.0200576.g002 Table 3. Nucleotide and amino acid sequence identity values for a partial sequence of the S segment from some isolates and reference group C orthobunyaviruses. and 95%-96% amino acid identity with MTBV. Clade III included ORIV, ITQV, and two clinical isolates (one each from northeastern and southeastern Peru). Clade III clinical isolates were closely related to ITQV (99% -100% aa identity) and divergent from ORIV (87% aa identity).

L segment sequence analysis
Clustering of isolates based on L segment analysis mirrored that of the S segment. Based on partial L segment analysis, L segment sequences separated into two clades (Fig 4), with 71% -76% pairwise nucleotide identity and 78% -82% pairwise amino acid identity between clade I and clade II isolates (

Antigenic characterization
Clinical isolates fell into four antigenically distinct groups (

Categorization of isolates
The genetic characteristics of reference strains CARV BeAn3994 and MADV BT4075 tracked together for S, M, and L in clade I, and MURV BeAn974 and MTBV BeAn15 tracked together for S, M, and L in clade II. ITQV PE00036 tracked with CARV BeAn3994 and MADV BT4075 for S and L as part of clade I, but with ORIV BeAn17 for M, as part of clade III. APEUV BeAn848 and ORIV BeAn17 tracked with MURV BeAn974 and MTBV BeAn15 for S and L segments as part of clade II, but APEUV BeAn848 tracked with CARV BeAn3994 for the M segment as part of the clade I, and ORIV BeAn17 tracked with ITQV PE00036 for M, as part of clade III. These results are based on neighbor-joining method; similar topologies were identified using maximum likelihood and maximum parsimony approaches (S1-S6 Figs). Based on genetic and antigenic characteristics, clinical isolates could be grouped with prototype strains. Specifically, 41 isolates were most closely related to CARV (amino acid identity  > = 96% S, > = 98% M, > = 97% L, and highest neutralization titers with CARV HMAF), 17 isolates were most closely related to MURV-MTBV serocomplex (MTBV amino acid identity > = 97% S, > = 95% M, > = 95% L, and highest neutralization titers with MURV or MTBV HMAF), and 2 with ITQV (amino acid identity > = 96% S, > = 99% M, > = 98% L). No ORIV isolates were detected. Additionally, six reassortants relative to CARV reference strains were identified (90% -91% M segment amino acid identity). These isolates grouped with Itaya virus, a novel reassortant orthobunyavirus isolated from febrile patients in the cities of Iquitos and Yurimaguas in Peru [16,17]. Itaya virus S and L segments are closely related to the respective segments of CARV; in contrast, the Itaya virus M segment is distinct from currently characterized orthobunyaviruses.

Discussion
In this study, we identified 41 isolates of CARV, 17 isolates of MURV, 6 isolates of Itaya virus, and 2 isolates of ITQV associated with human febrile illness in the northern and southern Amazon basin of Peru. The ITQV isolates represent the first association of ITQV with human illness in Peru. Overall, while the Peruvian isolates were most closely related genetically to the indicated reference viruses, they still displayed greater than 10% divergence at the nucleotide level in comparison to the reference viruses, consistent with the fact that the reference viruses were collected in neighboring Brazil several decades earlier.
Our data confirm that the M segment gene product is the key determinant of virus neutralization specificity. Similarly, early studies on GRCVs [1,3,4] described cross-reaction based on hemagglutination inhibition assays and neutralization testing between CARV-APEUV, MARV-MURV and ORIV-ITQV and cross reaction based on complement fixation assays between MURV-ORIV, ITQV-CARV and APEUV-MTBV [3,4]. Our phylogenetic results are consistent with the notion that antibodies detected by hemagglutination inhibition assay and neutralization testing are made in response to surface glycoproteins encoded on the M segment, and antibodies detected by complement fixation are made in response to the nucleocapsid protein encoded on the S segment [25].
For all isolates described here, S and L segments segregated together, both grouping into two major clades. For S and L, clade I was divided into two lineages, with lineages most closely aligned with geographic distribution rather than year of isolation. Lineages 1 and 3 of clade I contained only isolates from the northern Amazon region of Peru, while lineage 2 mainly contained isolates from the southern Amazon region of Peru. These data suggest that there has been little flow of orthobunyaviruses within Peru over the past 20 years.
While the majority of isolates grouped consistently across all three segments, there were clear inconsistencies between S and L versus M segment for a subset of viruses, likely representing reassortants. In this virus set, six isolates of Itaya virus were identified, including four new cases in addition to the originally identified isolate [16] and the isolate characterized by Hontz et al [17]. Notably, two reference viruses, APEUV and ORIV, also appear to be reassortants. APEUV grouped with MTBV and related isolates for the S and L segments. However, APEUV grouped with CARV and MADV for the M segment, and all isolates that were neutralized by APEUV HMAF were also neutralized by CARV. ORIV grouped with MURV for the S  and L segments but with ITQV for the M segment. There also appeared to be some serological cross-reaction between ORIV and ITQV. These data demonstrate the diversity of GRCVs that infect and cause febrile illness in humans in the Peruvian Amazon. Our data also suggest that reassortment among GRCVs is a relatively common phenomenon, as described for other bunyaviruses [26]. Further ecological studies are warranted to better understand the arthropod vectors and vertebrate hosts that maintain these viruses in the Peru and potential for broader emergence, particularly in light of the known impact of other orthobunyaviruses in human (e.g., Oropouche virus) and livestock (e.g.,Schmallenburg virus) disease.
Supporting information S1 Table. Group C orthobunyavirus (GRCV), Peruvian isolates used in the study.