Fig 1.
Summary of virus-positive tissues, saliva, and urine samples from Oligoryzomys spp.
In (A), we denote the samples identified to be positive (dark blue) or negative (light blue) for antibodies (whole blood) from prior work [33,42] and for viral RNA (lung, saliva, urine) herein using RNASeq and in (B) the overall percent positive or negative is presented. All samples are from Oligoryzomys nigripes except for one O. mattogrossae, TK66745. In (C), we present data from the screening of mice, with the majority of sample types available, for those that were NGS-positive. The unique rodent TK identification is presented on the y-axis, and the sample type is on the x-axis in A and C. We set a cutoff of >1,000 reads mapped across a region of>500 bp of the S- and M-segment reference sequences to define a vRNA-positive sample. In light blue, we illustrate those samples that are negative. Grey indicates that no sample was available for screening. Two samples were previously reported as positive by PCR [32,41] and are highlighted in white.
Fig 2.
Phylogenetic relationships of JUQV S and M-segment coding regions and representative South American hantaviruses.
(A) Maximum likelihood tree of the S-segment mRNA coding sequence (CDS; 1,065 bp after alignment and trimming corresponding to nucleotides 265-1329 of reference OR184959). Orthohantavirus sinnombreense (Sin Nombre virus; KF537003.1) was used as the outgroup, and phylogeny was inferred using the GTR + F + G4 model. (B) Maximum likelihood tree of the M-segment mRNA coding sequence (CDS; 3,420 bp after alignment and trimming corresponding to nucleotides 52-3468 of reference OR184986). Orthohantavirus sinnombreense (Sin Nombre virus; OQ999167.1) was used as the outgroup, and phylogeny was inferred using the TIM2 + F + I + G4 model. For visualization, branch lengths were proportionally rescaled in FigTree [46]; scale bar is not shown. Ultrafast bootstrap support values are shown at nodes. Sequences from this study are labeled by TK number, grid, year, and sample type. Two JUQV subclades are indicated in light blue (Clade I) and dark blue (Clade II). GenBank accession numbers and virus names are provided in the tree and the S4 Table.
Fig 3.
Heatmap of Shannon entropy for JUQV S- and M-segments and clustered by Shannon score for nucleotide.
Entropy values reflect nucleotide diversity at each genomic position, with white indicating low diversity (conserved sites) and blue indicating high diversity (mutational hotspots). Hierarchical clustering of samples based on entropy profiles highlights similarities in genetic diversity. We used 19 S-segment NGS datasets in the analysis: TK132709, TK133245, TK141528, TK141638, TK141660, TK141672, TK141765, TK170224, TK184781, TK184858, TK184889, TK184992, TK186283, TK186318, TK186352, TK186353, TK246023, TK246099, TK66745. The M-segment included the same TK specimens, except for TK184781 and TK141528, which were not included because they did not meet the criteria (≥80% genome coverage with coverage depths ≥ 500x).
Fig 4.
Number of consensus and minority polymorphisms in vRNA genomes from lung, saliva, and urine samples.
Consensus polymorphisms (≥50% frequency, x-axis) are shown in light blue, and minority polymorphisms (<50% frequency) in dark blue. The y-axis provides the TK number for the sample type. TK141528, TK184781, and TK170224 data sets were not included as they did not meet the inclusion criteria (≥80% genome coverage with coverage depths ≥ 500x).
Fig 5.
Frequency of JUQV S- and M-segment mutations.
The frequencies of (A) SNPs per 1000 nt and (B) amino acid mutations per 100 amino acids (aa) are illustrated for lung, saliva, and urine sequences. The height of the bar represents the mean of the data set, and the error bars represent the standard deviation.
Fig 6.
Distributions of SNPs in NCR and nonsynonymous mutations in coding regions of S- and M-segments.
(A) Proportion of lung, saliva, and urine samples that share specific SNPs in the cRNA of S- and M-segments. The x-axis shows the nucleotide (nt) within the cRNA position, and the y-axis indicates the proportion of samples with a specific mutation. Each SNP is represented by a colored bar: synonymous (light blue), non-synonymous mutation (dark blue), or noncoding regions (NCR, grey). A grey line in the background of each plot represents the depth of coverage, with values indicated on the right y-axis. (B) Heat maps of nonsynonymous mutations in the JUQV N and GP display within-sample frequency of amino-acid substitutions across lung, saliva, and urine samples. Each row represents an individual genome and each column an amino-acid position; color intensity increases with mutation frequency (scale 0 – 100%).
Fig 7.
Within-host network analysis of JUQV S- and M-segment sequences from eight rodent specimens.
For those samples with sufficient sequence integrity (≥80% genome coverage and ≥500x depth of coverage), we used the program PopArt to map within-host relationships. The samples chosen were from the set of eight specimens in Table 1C, which had the most significant number of available sequences of all those listed in Table 1A. Analyses were restricted to samples with long, contiguous nucleotide regions to maximize the number of shared informative sites for haplotype network reconstruction. The hash marks in the networks denote nucleotide differences between samples. Three areas (grids) were associated with these eight samples. As the samples were collected in areas with different levels of degradation this information was included from our prior published efforts [42]. The nucleotide regions analyzed are shown below each network and correspond to cRNA consensus sequences.
Table 1.
Codon sites under purifying selection detected by Fixed Effects Likelihood across JUQV consensus sequences.
Fig 8.
Mutation frequency distributions in acute and persistently-infected rodents.
We compare the nucleotide (Nt) and amino acid (AA) mutation frequencies of orthohantaviral genomes in rodents with acute Ab-/RNA+) or persistent (Ab + /RNA+) infections. Statistical comparisons were performed using a one-sided permutation test (Persistent > Acute). Permutation tests were implemented in Python (version 3.10) using SciPy (version 1.11) with 10,000 random permutations.
Fig 9.
Assessment of JUQV GP entry using high content screening.
(A) Representative confocal images are shown from the second dilution (3-fold) of JUQV WT GP, JUQV Q292H GP, and JUQV V504I GP, which were pseudotyped onto VSV-GFP. The three pseudotyped viruses were diluted to several concentrations to assess entry. Multiple images were automatically acquired for each well using the Yokogawa CQ1 at 20 × magnification. Shown are TO-PRO-3 nuclei (left), GFP signal (middle), and merged channels (right). Identical acquisition settings were applied across treatments. (B) The GFP signals were quantitated and used to construct dose–response curves for JUQV GP WT, Q292H, and V504I pseudotyped viruses. Data were normalized to the positive control (VSV at MOI 5) and background-subtracted. Data represent mean ± SEM of triplicates, fitted with a four-parameter logistic regression. Estimated EC₅₀ values of virus MOI (i.e., the effective MOI concentration for entry) were 3.17 (WT), 2.95 (Q292H), and 1.42 (V504I). The parameter estimates for (B) are given in S8 Table.
Fig 10.
Structural comparison of JUQV and ANDV GPs using AlphaFold.
The AlphaFold-predicted structure of the JUQV Gn (green) and Gc (purple) based on the ANDV GP prefusion complex (PDB ID: 6Y5F). The Q292H mutation is located near the fusion loop, a critical region for membrane fusion. In contrast, the V504I mutation is predicted to be within the transmembrane. This region was predicted with low confidence by AlphaFold and is absent in the ANDV crystal structure.