The viral isolates used in this study are summarized in Table 1. The first strain of Itaya virus (IQT9646) was isolated in 1999 from samples from a 25-year-old man in Iquitos, Peru. The second strain (FSL2923) was isolated in 2006 from a 59-year-old febrile man in Yurimaguas, Peru (11). The origin of the Caraparu strain BeAn3994 was described by Causey et al. (16). The source and isolation of group C virus prototype strains have been described elsewhere (9,17,18).

The locations of the study sites where IQT9646 and FSL2923 were isolated are depicted in Figure 1. Iquitos is a city of ≈380,000 inhabitants located 120 meters above sea level in the Amazon Basin of northeastern Peru. Yurimaguas is a city of ≈63,000 inhabitants located ≈184 meters above sea level and ≈388 km southwest of Iquitos.

The human use study protocols were approved by the Peruvian Ministry of Health and by the NMRC Institutional Review Board (protocol NMRCD.2000.0006). Study subjects enrolled were >5 years of age and sought treatment for an acute, febrile, undifferentiated illness at military or civilian outpatient clinics at predetermined study sites. The criteria for inclusion of patients were fever >38°C of no more than 5 days in duration and nonspecific symptoms, such as headache, fatigue, or myalgia. Demographic and clinical information were obtained from each patient at the time of voluntary enrollment, and patients >18 years of age signed individual consent forms. Paired blood samples were collected, the first during the acute phase of illness and the second 2–4 weeks after symptom onset.

Serum samples collected during the acute phase of illness were used to isolate viruses by cell culture techniques, and samples collected during both acute and convalescent phases of illness were assayed by using IgM ELISA for evidence of arboviral infections (9). The procedure used for virus isolation was described by Aguilar et al. (8). In brief, serum samples were inoculated into flasks containing either confluent monolayers of African green monkey kidney cells of the Vero lineage or Aedes albopictus mosquito (C6/36) cells and maintained at 37°C and 28°C, respectively. The cell cultures were examined daily for 10 days for evidence of viral cytopathic effects (CPE); on the appearance of CPE, spot-slides were prepared and an immunofluorescence assay was done by using a polyclonal antibody against specific arboviruses that are known to circulate in Peru (9).

Prototype group C viruses were inoculated into flasks containing confluent monolayers of African green monkey Vero kidney cells and maintained at 37°C, as described previously. On the appearance of CPE, cell culture supernatants were harvested and clarified by centrifugation, and viral RNA was extracted.

For the extraction of RNA, cell culture supernatants were harvested and clarified by low-speed centrifugation (2,000 × g, 10 min at 4°C), filtered through a 0.45-μM pore size filter (EMD Millipore, Billerica, MA, USA), and treated with a combination of DNases: 14 U Turbo DNase (Ambion, Austin, TX, USA); 20 U Benzonase (EMD Millipore); and 20 U RNase One (Promega, Madison, WI, USA) for 1 h at 37°C. Next, 24 mL of supernatant was loaded on top of 8 mL 30% sucrose (in TEN, pH 7.4), and centrifuged at 15,000 × g for 4 h at 4°C. Finally, the pellet was resuspended in 250 μL RNase/DNase and protease-free water (Ambion), and viral RNA was extracted using the Trizol reagent (Invitrogen, Carlsbad, CA, USA) by using the manufacturer’s protocols.

Viral RNA (≈0.9 μg) was fragmented by incubation at 94°C for 8 min in 19.5 μL of Illumina fragmentation buffer 15016648 (Illumina, San Diego, CA, USA). A sequencing library was prepared from the sample RNA by using an Illumina TruSeq RNA Sample Preparation Kit v2 using the manufacturer’s protocol (Illumina). The sample was sequenced on a HiSeq 1000 by using the 2 × 50 paired-end protocol. Reads in fastq format were quality filtered, and any adaptor sequences were removed by using Trimmomatic software (19). The de novo assembly program ABySS (20) was used to assemble the reads into contigs, using several different sets of reads and k values from 20 to 40. In all samples, host reads were filtered out before de novo assembly. The longest contigs were selected, and reads were mapped back to the contigs by using bowtie2 (21) and visualized with the Integrated Genomics Viewer (22) to verify that the assembled contigs were correct. Total reads ranged from 1.5 to 12 million; the percentage of reads mapping to the virus genome in each sample ranged from 12% to 33%. (Details are available upon request from the authors.)

We deposited the complete genome sequences of Itaya virus and other group C viruses obtained for this study in GenBank under accession numbers KM092512-KM092514 and KM280924-KM280938. We used the neighbor-joining method available in MEGA5 (23) for phylogenetic analysis. The support for each node was determined by using 1,000 bootstrap replicates.

We used established methods to obtain hyperimmune ascitic fluid for classic cross-neutralization tests (24). Mice received 4 weekly intraperitoneal injections of 10% virus-infected newborn mouse brain suspension with Freud’s adjuvant. Sarcoma 180 cells were also given intraperitoneally with a final injection to induce ascites formation. Antigenic differences among the viruses were then investigated by using cross-neutralization assays (25).

In 1999, the virus strain IQT9646 was isolated from a 25-year-old male febrile patient who resided in Belen, Iquitos, Peru. The patient had an illness with symptoms of fever, headache, retro-orbital pain, arthralgia, chills, cough, and nasal congestion. These clinical symptoms are also characteristic of dengue, malaria, and other tropical infectious diseases common in the region (26,27). The strain was initially classified as a Maguari isolate based on the results of serologic reactivity in an indirect immunofluorescence test. Maguari virus has been previously isolated from mosquitoes of the Aedes, Mansonia, and Psorophora spp. in Brazil; a variety of other mosquito species in Ecuador, Brazil, Trinidad, Colombia, Argentina, and French Guiana; and from horses in Guyana and Colombia (28,29). Nevertheless, evidence of human infection with Maguari virus is lacking. Therefore, we attempted to further identify and genetically characterize this strain using primers specific for different orthobunyaviruses, including Maguari; however, our attempts to amplify partial or complete genomic segments were unsuccessful. We therefore sought to obtain the complete sequences of the S, M, and L segments using an unbiased sequencing approach, then phylogenetic analyses to determine the relationship of the IQT9646 isolate to other viruses within the Orthobunyavirus genus.

Phylogenetic trees based on the S and L gene segments placed the IQT9646 virus among isolates of Caraparu virus, a member of the group C virus serogroup (Figures 2, 3), a pathogen known to cause febrile illness in humans and animals in the Amazon region of Peru (9,11). However, the M segment phylogenetic tree indicated that the IQT9646 virus had an M segment sequence divergent from Caraparu. In an attempt to determine the source of the M segment, we used a comprehensive full-genome sequence approach of other group C viruses and determined that the M segment of the IQT9646 strain was not closely related to other group C viruses but instead was most similar to the Caraparu and Apeu viruses (Figure 4). The S and L segment fragment sequences of the IQT9646 strain exhibited 97.6% and 96.6% nucleotide identity, respectively, to the prototype Caraparu strain BeAn3994 and to the Peruvian Caraparu strain IQD5973, whereas the M segment sequence displayed ≈75% nucleotide identity to the prototype and Peruvian Caraparu strains. In summary, these results suggest that IQT9646 is possibly a group C reassortant virus.

To determine if the IQT9646 strain is antigenically distinct from Caraparu virus, we investigated the serologic relationships of IQT9646 using cross-neutralization tests. Mouse antisera were prepared against the IQT9646 strain and the prototype Caraparu strain BeAn 3994. These antisera displayed a minimum 4-fold difference in neutralization titers between these viruses, indicating that IQT9646 is serologically distinct from the prototype Caraparu virus. Because some genetic variation exists between the prototype Caraparu and recent Caraparu isolates from Peru (11) that may translate into minor antigenic differences, we included in our analyses a recent Caraparu strain isolated in Peru. The results were consistent with those obtained with the prototype Caraparu strain (Table 2). In summary, we found that the IQT9646 strain is antigenically distinct from Caraparu virus.

Genomic characterization of some group C prototype viruses and other isolates found in South America were recently reported (11). The strain FSL2923, isolated from a febrile patient residing in Yurimaguas, Peru, in 2006 was found to possess L and S RNA segments closely related to Caraparu virus; the M segment was only ≈75% identical to the M segment of Caraparu virus (11). However, this strain was not characterized antigenically, causing uncertainty of whether this virus strain was distinct from Caraparu virus.

Considering our findings with the IQT9646 strain, which, like FSL2923, possesses the S and L RNA segment sequences closely related to Caraparu virus and the M RNA segment sequence derived from an unidentified group C virus, we made genetic comparison and observed that the S, M, and L RNA segments of IQT9646 and FSL2923 shared >98% nucleotide and >99% amino acid sequence homology for all 3 viral RNA segments. These findings indicate that the FSL2923 isolate is the same virus as strain of IQT9646. The data also suggest that this reassortant virus is widely distributed throughout the Amazon basin in northeastern Peru, because it was isolated in Iquitos and Yurimaguas, in the Department of Loreto, within a 7-year time period (Figure 1).