Induction of Cell Cycle and NK Cell Responses by Live-Attenuated Oral Vaccines against Typhoid Fever

Abstract

The mechanisms by which oral, live-attenuated vaccines protect against typhoid fever are poorly understood. Here, we analyze transcriptional responses after vaccination with Ty21a or vaccine candidate, M01ZH09. Alterations in response profiles were related to vaccine-induced immune responses and subsequent outcome after wild-type Salmonella Typhi challenge. Despite broad genetic similarity, we detected differences in transcriptional responses to each vaccine. Seven days after M01ZH09 vaccination, marked cell cycle activation was identified and associated with humoral immunogenicity. By contrast, vaccination with Ty21a was associated with NK cell activity and validated in peripheral blood mononuclear cell stimulation assays confirming superior induction of an NK cell response. Moreover, transcriptional signatures of amino acid metabolism in Ty21a recipients were associated with protection against infection, including increased incubation time and decreased severity. Our data provide detailed insight into molecular immune responses to typhoid vaccines, which could aid the rational design of improved oral, live-attenuated vaccines against enteric pathogens.

Keywords: NK cell; Ty21a; cell cycle regulation; functional genomics; typhoid; typhoid vaccines; vaccine immunity.

Figure 1

Gene expression (GEX) following oral, live-attenuated typhoid vaccination. (A,B) Study design and sample overview for Ty21a (orange; three doses administered at D-32, D-30, and D-28) and M01ZH09 (green; one dose administered at D-28). Numbers in circles represent N of which samples were taken at each time point/assay. Gray shaded time points indicated where no samples were taken. Samples were collected for GEX, antibody serology (Ab), and in vitro infection [peripheral blood mononuclear cell (PBMC)] analysis. (C,D) Number of differentially expressed (DE) genes over pre-vaccination baseline following Ty21a (C) and M01ZH09 (D) vaccination at time points after vaccination corresponding to panel (A,B). (E) Circos plot indicating the overlap of DE genes between each vaccine arm and time point. Blue links: overlap between M01ZH09 time points. Purple links: overlap between M01ZH09 and Ty21a. Yellow and red links: overlap between Ty21a time points. Red and blue dots indicate up- or downregulation of each gene, respectively. (F) Significant pathways overrepresented by DE genes following Ty21a and M01ZH09 vaccination (increasing bubble size depicts increasing significance level).

Figure 2

Gene Set Enrichment Analysis was performed at each time point following Ty21a and M01ZH09 vaccination. (A) Tile graph with significant blood transcriptional modules (BTMs) (adjusted p < 0.05) following vaccination (blue: negative enrichment; red: positive enrichment). Non-significant BTMs were set to normalized enrichment score (NES) = 0 (gray: not significant). Bar graph to the right indicates the number of significantly enriched BTMs at each time point/group. (B) NESs of BTMs at the D-21 time point following vaccination. Middle: scatter plot displaying NES of significantly enriched BTMs (adjusted p < 0.05) following either Ty21a (x-axis) or M01ZH09 (y-axis) vaccination. If a BTM was not significantly enriched in one of the two groups, the respective NES was set to 0. Colors represent different BTM categories. Dot plots depict enrichment scores of cell cycle (left) and NK cell (right) related modules. Solid dots: p < 0.05.

Figure 3

Single sample GSEA (ssGSEA) following vaccination at time point D-21. (A) Normalized enrichment score (NES) (mean + SEM) of blood transcriptional modules (BTMs) related to NK cells (blue bar), the cell cycle (purple), B cells (pink bar), and T cells (cyan bar) across all participants in the Ty21a (orange) and M01ZH09 (green) arm at time point D-21 after vaccination is plotted. (B) Tile graph representing the NES for all NK cell-related BTMs for each participant (red: positive NES; blue: negative NES). Color bar on the right represents the vaccine group membership. Bar graph along the tile graph signifies the heterogeneity within each participant (depicted by inter-decile range). (C) NK cell modules were combined into one network and superimposed with mean gene expression values at time point D-21 following Ty21a (top) and M01ZH09 (bottom). (D) Mean expression of genes within the NK cell-related BTMs in the Ty21a and M01ZH09 group at time point D-21 following vaccination. (E) Association networks. NESs derived from the ssGSEA at time point D-21 were correlated with antibody responses to vaccination and parameters reflecting outcome following challenge [temperature (Temp); time to diagnosis (ttDx)] in Ty21a (left) and M01ZH09 (right) recipients. BTM colors: purple: cell cycle. Blue: NK cells. Cyan: T cells. Red: inflammation, monocytes, and dendritic cells. Pink: antibodies, B cells, and antigen presentation. Green: platelet activation. Colors of edges represent significance level and Spearman’s rank correlation coefficients (orange: p < 0.05, rho > 0; blue: p < 0.05, rho < 0; gray: n.s.). Gray shaded nodes: BTMs associated with post-challenge outcomes. (F) Correlation plot of maximum Temp following challenge and ttDx. Modules indicated on the plot represent modules inversely associated with Temp and ttDx. Top left: positive and negative correlation with Temp and ttDx, respectively. Bottom right: positive and negative correlation with ttDx and Temp, respectively. Orange: Ty21a recipients; dark green: M01ZH09 recipients.

Figure 4

In vitro infection of peripheral blood mononuclear cells (PBMCs) from study participants. (A) Experimental design. PBMCs harvested from study participants at pre-vaccination (D-32: Ty21a; D-28: M01ZH09), 14 days (D-14), and 28 days after vaccination (D0, day of challenge) were infected in vitro with Ty21a (orange) or M01ZH09 (dark green). (B) Differential induction of activation markers on NK, T, and NKT cells following in vitro stimulation of PBMCs from vaccine-naïve volunteers with Ty21a (orange) and M01ZH09 (green). Multiplicity of infection = 0.1:1. Statistics: paired t-test: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 5

Autologous recall responses in peripheral blood mononuclear cells (PBMCs) from individuals post-vaccination. PBMCs harvested from study participants at 14 days (D-14) and 28 days after vaccination (D0, day of challenge) were infected in vitro with the autologous vaccine strain and NK cell responses measured. Fold-change increases of baseline responses are presented. Statistics: two-sided t-test.