PMC

Three groups of vaccinia-naïve individuals from protocol #4 [–] were given vaccinations as follows. A: MVA on d0 and d28, followed by Dryvax on d112 (group 2). B: MVA on d0 and d28, followed by placebo on d112 (Group 3). C. Dryvax on d112 (group 4). Proteome-wide antibody profiles were determined at regular time points by microarrays; only responses to WR148 (present in Dryvax only), WR118/D13 and WR113/D8 are shown.

Peak MVA and Dryvax antibody profiles from protocol #1 (groups 1a+1b and group 2, respectively) were defined by microarray and compared by t test. Histograms are plotted from raw signal intensities subtracted of d0 signals, and ranked by the Dryvax signals. The overlaid p-values were determined from log transformed and normalized signals at the peak subtracted of d0. Antigens are segregated into differentially and non-differentially reactive according to Benjamini-Hochberg corrected p-value (p_BH) for each reactive antigen (< or ≥ 0.05, respectively).

A: Protocol #3 [], vaccinia naïve individuals (n=47) before and 28d after Dryvax. B: Protocol #5, previously vaccinated individuals (n=25) before and 28d after Dryvax revaccination. C: Protocol #1, vaccinia naïve individuals (n=48) before and 42d after two MVA inoculations in d0 and d28. Horizontal line represents cutoff, above which a serum was considered seropositive. Figures in the top left of each panel indicate the percentages of seropositive samples before/after vaccination. The cutoff was defined the average+2SD of the pre-vaccination (d0) signals using log transformed data. To help evaluate pre-existing immunity in the previously vaccinated individuals before revaccination (d0 in panel B), the cutoff was defined from the vaccinia naïve population.

ELISA data for vaccinia naïve individuals given Dryvax or MVA (described in ) were subjected to receiver operator characteristic (ROC) analysis to find antigens that discriminate between profiles on d0 and at the peak response (d28 for Dryvax and d42 for MVA). The best discriminators are antigens with an area under the curve (AUC) approaching 1.0. An antigen with no discriminatory power would have an AUC of 0.5 (represented by the diagonal). A: Dryvax, single antigens; B: Dryvax, multi-antigen analysis; C: MVA, single antigens; D: MVA, multi-antigen analysis; E: AUC values for single and multi-antigen analyses, ranked by AUC of the Dyvax response.

Antigens from vaccinia virus strain Western Reserve (VACV-WR) were printed on microarrays and probed with sera from groups of MVA and Dryvax vaccinees. Nomenclature for antigens is WR number/ Copenhagen strain ortholog. The plots show the average raw signal intensity ±SEM of each antigen for each vaccination group. Day 0 signals were subtracted from all time points. The reactive antigens are ranked in descending order independently for each group. A: MVA given on d0 and d28 (group 1a, n=25); B: Dryvax given on d0 (group 2, n=8); A and B are longitudinal sera from a single representative protocol (protocol #1). C: Dryvax given on d0 from protocol #2 (n=10) [].

Individuals known to have been vaccinated previously against smallpox were given boosts of MVA or Dryvax, and sera probed against VACV-WR proteome microarrays (A, B) or were vaccinia naïve (C, D). The ratio of post/pre signal intensities for each antigen are shown to the right of each graph. The values shown in each panel are raw signal intensities without d0 subtraction, and antigens are ranked separately by the profile prior to boosting. The position of the WR148 response is highlighted by the asterisk (*). A: MVA arm from protocol #6 (n=25). Individuals were inoculated twice (d0/d28) and profiles on d0 (`pre') and d56 (`post') plotted. B: Dryvax arm from protocol #5 (n=25). Individuals were inoculated on d0, and profiles on d0 (`pre') and d28 (`post') plotted. Protocol #5 examined revaccination of individuals with 3 different dilutions of Dryvax (1/10, 1/5 and undiluted). All individuals showed a take to the inoculation. A three-way T test analysis revealed no significant different differences between the profiles engendered by the three different doses (data not shown), thus the data could be pooled for this figure. C: Corresponding MVA profile in vaccinia naïve individuals (group 1a from protocol #1, n=25). D: Corresponding Dryvax profile in vaccinia naïve individuals (group 2 from protocol #1, n=8).