Table 1.
Subjects with confirmed or suspected CHIKV infection were enrolled in either an endemic cohort (Ponce, Puerto Rico; color coded in orange), a non-endemic cohort (Portland, Oregon; color coded in blue), or an alphavirus naïve cohort (Portland, Oregon; color coded in black). Subjects are assigned an ID with age, country of birth, country of infection, range of time-post infection for serum collection, CHIKV PRNT50 at time of primary blood draw, and reported symptoms displayed.
Fig 1.
Longitudinal serology for endemic and non-endemic patients.
(A) Phylogenetic tree produced using the E1, 6k, and E2 amino acid sequences for the six alphaviruses under investigation; viruses are color coded to match serology graphs. (B, C) Sera samples from each subject were tested for neutralization activity against CHIKV, ONNV, RRV, MAYV, Una, and VEEV by plaque reduction neutralization titer assays (PRNT) performed on confluent monolayers of Vero cells. Shown are the average 50% reduction titer values (PRNT50) calculated by variable slope non-linear regression using Prism software. Longitudinal serology is shown for 9/12 human subjects. Additional samples for the other human subjects were unavailable. Endemic subject serologic profiles are shown in (B). Serology for non-endemic subjects is shown in (C). (D) Summarizes the breadth of cross-neutralization data for both endemic and non-endemic subjects at all time points presented in (B) and (C). The statistical analysis to compare grouped cross-neutralizing PRNT50 values to CHIKV PRNT50 was completed using an ANOVA and Friedman’s test ***p = 0.0001, **** p = <0.0001. Limit of detection (LOD) is 40, samples below the LOD were assigned an arbitrary value of 39.
Fig 2.
Antigenic cartography to map human subject alphavirus cross-neutralization by human sera.
Antigenic map shows the relative antigenic relatedness between CHIKV, ONNV, RRV, MAYV, UNAV, and VEEV. Each unit of antigenic distance (AU), the length of one side of a grid square, is equivalent to a two-fold dilution in a neutralization assay. Sera are shown as open ellipses and labeled by subject number. Each virus is shown as a color filled ellipses and is colored according to virus strain (Fig 1A). The size and shape of each ellipse is the confidence area of its position. In making the map, each sera is initially plotted on top of the virus it most potently neutralizes and then pairwise distances between each sera:virus combination are calculated as a fold-difference in titer between the most potently neutralized virus and each other virus. The map is then optimized to place each virus relative to the serum samples in a manner that minimizes error between pairwise fold-differences. The closer a virus is to another virus, the more antigenically related the two are. Sera are initially plotted nearest to the virus they most potently neutralize with subsequently increasing distance to other viruses in descending neutralization potency against each virus. The antigenic map in (A) reflects each human subject at the primary blood draw, and (B) is representative of longitudinal sampling.
Fig 3.
Comparison of Alphavirus E2 B domains.
(A) Amino acid sequence alignment was performed using Geneious software for the E2 B domains of the alphaviruses examined in this study. Regions of 100% homology are highlighted in black, 80–100% similarity is dark grey, 60–80% similarity is light grey, and less than 60% similarity is in white. (B) Matrix depicts the amino acid sequence identity as a percentage. (C) Top-down view of the organization of the Mayaro Virus E1:E2 monomer (Teal:Brown) shown with the E2 B domain annotated in purple. (D) E1:E2 trimer spike organization depicted with the E2 B domain annotated in purple, E1 in shades of teal, and E2 in shades of brown.
Fig 4.
Impact of depletion of E2B-binding antibodies on CHIKV and MAYV neutralization.
His-tagged CHIKV or MAYV E2 B domain bound to magnetic beads (or control beads alone) was adsorbed by diluted human serum for 4 hours and the beads were pulled off with a magnet. Following depletion, the sera was used in both CHIKV and MAYV neutralization assays. Human sera samples from the first blood draw were diluted 1:2 from 1:100 to 1:102,400. "No beads" is diluted serum only in black, CHIKV E2B absorbed human sera is in purple, MAYV E2B absorbed human sera is in teal, and control beads bound to diluted human sera is in pink. The data are representative of 3 biological experiments completed with duplicate samples.
Fig 5.
Analysis of changes in CHIKV and MAYV neutralizing antibody titers following E2 B domain depletion.
Fold change in neutralizing antibody titers (nAb) of subject serum samples following adsorption against E2B domain coated Ni-NTA or control beads was calculated against non-bead-treated serum samples. Depletion of MAYV E2 B domain-specific antibodies and impact on (A) MAYV or (B) CHIKV neutralizing antibody titer fold change compared to serum with control beads. A paired t-test for comparison of fold change heterotypic MAYV neutralization following MAYV E2B depletion yielded a p value *** = 0.0003 and 0.3276 (ns) for homotypic CHIKV neutralization. Depletion of CHIKV E2 B domain-specific antibodies and impact on (C) MAYV or (D) CHIKV neutralizing antibody titer fold change compared to serum with control beads. A paired t-test for comparison of fold change in heterotypic MAYV neutralization following CHIKV E2B depletion yielded a p value *** = 0.0006 and ** 0.0013 for homotypic CHIKV neutralization. Comparison of changes in (E) CHIKV PRNT50 or (F) MAYV PRNT50 following E2B depletion relative to no beads or control samples. LOD = 100 with values below the limit of detection graphed as 99. Data were analyzed using a one-way ANOVA with the significant comparison in (F) being ** p = 0.0045. Note Subject 17 was excluded from this statistical analysis as the MAYV neutralization in this subject was low, therefore, the impact on E2B depletion was not detectable.
Fig 6.
Antigen-specific MBC frequency per 106 PBMC over time in non-endemic cohort (blue n = 6), endemic (orange n = 5), and naïve subjects (black n = 3).
(A) CHIKV-specific MBC frequency as determined by whole CHIKV-ELISA. (B) MAYV-specific MBC frequency determined by whole MAYV-ELISA. (C) E2B-specific MBC frequency determined by MAYV-E2B ELISA. Negative samples and those below the limit of detection were assigned an arbitrary value between 0.05 and 0.09 (LOD = 0.1). (D) Summary of antigen-specific MBC frequency for CHIKV, MAYV, and MAYV E2B with subjects grouped together. Geometric mean frequencies are reported on the graph. P values are the result of a one-way ANOVA ** p = 0.0044, *** p = 0.006.
Table 2.
Antigen-specific MBC frequency for non-endemic and endemic cohorts.
Table summarizes subject sampling time post-infection, MBC frequencies for the three antigens tested, and % MAYV MBC attributable to E2B, determined by E2B MBC frequency divided by total MAYV-MBC frequency. ND = not detected, N/A = not applicable.