Macrophages are required for Shigella training and contribute to zebrafish survival (A, D, and G) Schematic of the zebrafish immune training model in the HBV in the absence of macrophages using IRF8 zebrafish mutants (A) and MTZ treatment during the 1st infection (D) or 2nd infection (G). (B) Survival rates of IRF8+/+ and IRF8−/− zebrafish larvae (N = 3 with >8 larvae per experiment). Left graph represents collated data and right graph represents only IRF8−/− larvae, a subset of the data on the left. (C) Bacterial burden of naive and Shigella-trained (IRF8+/+ 1.1 × 103 ± 8.3 × 102 colony-forming units [CFUs] and IRF8−/− 1.3 × 103 ± 7.6 × 102 CFUs) larvae upon secondary infection (IRF8+/+ PBS 2.6 × 104 ± 7 × 103 CFUs, IRF8+/+ Shigella 2.7 × 104 ± 7.5 × 103 CFUs, IRF8−/− PBS 2.7 × 104 ± 7.5 × 103 CFUs, and IRF8−/− Shigella 2.3 × 104 ± 6.8 × 103 CFUs). (E) Survival rates of DMSO- and MTZ-treated zebrafish larvae (N = 3 with >18 larvae per experiment). Left graph represents collated data and right graph represents only MTZ-treated larvae, a subset of the data on the left. (F) Bacterial burden of naive and Shigella-trained (DMSO 8 × 102 ± 2.4 × 102 CFUs and MTZ 1.2 × 103 ± 4 × 102 CFUs) larvae upon secondary infection (DMSO PBS 2.6 × 104 ± 8 × 103 CFUs, DMSO Shigella 2.7 × 104 ± 1.1 × 104 CFUs, MTZ PBS 2.9 × 104 ± 1 × 104 CFUs, and MTZ Shigella 3.3 × 104 ± 1.2 × 104 CFUs). (H) Survival rates of DMSO- and MTZ-treated zebrafish larvae (N = 3 with >14 larvae per experiment). Left graph represents collated data and right graph represents only MTZ-treated larvae, a subset of the data on the left. (I) Bacterial burden of naive and Shigella-trained (1.3 × 103 ± 6.8 × 102 CFUs) larvae upon secondary infection (PBS DMSO 2.1 × 104 ± 1.5 × 104 CFUs, Shigella DMSO 2.5 × 104 ± 4.7 × 103 CFUs, PBS MTZ 2.2 × 104 ± 8.8 × 103 CFUs, and Shigella MTZ 2.7 × 104 ± 5.1 × 103 CFUs). ns, non-significant; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.01, and ∗∗∗∗p < 0.0001. Error bars represent standard deviation (SD). (B, E, and H): log rank (Mantel-Cox) test. (C, F, and I) Two-way ANOVA with Tukey’s multiple comparisons test.

Shigella-trained macrophages shift to a pro-inflammatory state (A) Representative images from the CHT region (top, white box indicates area imaged) of macrophages (magenta) expressing TNF-α (green) in naive (PBS, left) and Shigella-trained (right) larvae at 48 hp1i. Scale bar: 100 μm. Scale bars in insets: 10 μm. (B) Quantification of tnfa:GFP expression. (C) Quantification of H3K4me1 (left) and H3K4me3 (right) deposition in macrophages from naive (black) and Shigella-trained (green) larvae at 48 hp1i. Data were pooled from 3 independent experiments, with >20 larvae per experiment. ∗∗p < 0.01 and ∗∗∗∗p < 0.0001. (B and C) Student’s unpaired t test. Percentages indicate population with normalized values ≥ 1.

Shigella-trained macrophages produce more ROS (A) Schematic of the zebrafish immune training model in the HBV and re-infection in the duct of Cuvier. (B and C) Survival rates (B; N = 3 with >14 larvae per experiment) and bacterial burden (C) of naive and Shigella-trained (1.5 × 103 ± 2.5 × 102 CFUs) larvae upon systemic secondary infection (PBS 2.4 × 104 ± 8.2 × 103 CFUs and Shigella 1.7 × 104 ± 9.9 × 103 CFUs). (D) Quantification of tnfa:GFP expression in naive and Shigella-trained macrophages at 3 hp2i (N = 3 with >20 larvae per experiment). Single dots indicate Shigella-infected macrophages. (E) Quantification of LIVE/DEAD staining internalized by macrophages in naive and Shigella-trained larvae at 3 hp2i (N = 4 with >20 larvae per experiment). (F) Quantification of CellROX in naive and Shigella-trained macrophages at 3 hp2i (N = 3 with >20 larvae per experiment). Single dots indicate Shigella-infected macrophages. ns, non-significant; ∗∗∗∗p < 0.0001. Error bars represent standard deviation (SD). (B) Log rank (Mantel-Cox) test. (C) Two-way ANOVA with Sidak’s multiple comparisons test. (D–F) One-way ANOVA with Tukey’s multiple comparisons test). Percentages indicate population with normalized values ≥ 1.