PPE31^Mm was required for the resistance of M. marinum to acid medium in vitro. (A) CRISPR interference has been performed for knockdown of ppe31^Mm. Expression of ppe31^Mm determined by RT-qPCR with or without induction of aTc. (B)M. marinum contain dCas9 and sgRNA targeting ppe31^Mm were exposed to 7H9 medium of pH4.5 with or without (control) induction of aTc. The bacterial survival was monitored by CFU counting at the indicated time. (C) WT, Δppe31^Mm, or comp-Δppe31^Mm were exposed to 7H9 medium of pH4.5 at 30°C. The bacterial survival was monitored by CFU counting for 9h. Data are shown as mean ± S.E.M. of three independent experiments. *p < 0.05, **p < 0.01.

Δppe31^Mm infection fails to partly induce the expression of proinflammatory cytokines and ROS generation in BMDMs. BMDMs were infected with WT, Δppe31^Mm, or comp-Δppe31^Mm at different MOIs (0.1, 1 and 10) for 6 h. The level expressions of inflammatory cytokines, TNF (A) and IL-6 (B) were assessed by RT-qPCR. (C) BMDMs were stimulated with WT, Δppe31^Mm or comp-Δppe31^Mm for 30 min. Cells were then labeled with DCFH-DA that were used detecting cytosolic ROS, and were analyzed for cytosolic ROS levels using flow cytometry. Data are shown as mean ± S.E.M. of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001. WT, wild-type M. marinum; UI, uninfected.

Macrophages infected with Δppe31^Mm show reduced caspase- independent cell death. (A) BMDMs were infected with WT, Δppe31^Mm, or comp-Δppe31^Mm (MOI = 10) for 24h for analysis of morphological changes. Scale bar in 100 μm. (B) BMDMs were infected with WT, Δppe31^Mm, or comp-Δppe31^Mm (MOI = 10). The cytotoxicity was assessed by LDH release assay at indicated times. (C) BMDMs were infected with either WT, Δppe31^Mm, or comp-Δppe31^Mm (MOI = 10 or 5) for 24 h, and cell death was detected by PI staining and then examined by flow cytometry. (D) BMDMs were infected with WT, Δppe31^Mm or comp-Δppe31^Mm (MOI = 10) in the presence or absence of Z-VAD-FMK (20 μM), a caspase inhibitor, then cells were stained with PI and then examined by flow cytometry. Data are shown as mean ± S.E.M. of three independent experiments. *p < 0.05, **p < 0.01. WT, wild-type M. marinum; UI, uninfected; SC, solvent control (0.1% DMSO); ns, no significant.

Infection with Δppe31^Mm reduces cell death through JNK-dependent signaling. (A) BMDMs were infected with WT, Δppe31^Mm or comp-Δppe31^Mm (MOI = 10) for the indicated periods of time, and then subjected to Western blot analysis using antibodies raised to p-ERK1/2, p-p38, p-JNK, and β-actin. (B) BMDMs were pretreated with the following MAPK signaling pathways inhibitors U0126 (20 μM), SB203580 (SB; 10 μM), or SP600125 (SP; 20 μM) for 1h, and then infected with WT, Δppe31^Mm or comp-Δppe31^Mm for 30 min, Cells were then incubated with DCFH-DA and analyzed immediately for ROS generation by flow cytometry. (C) BMDMs were pretreated with U0126 (20 μM), SB (10 μM), or SP (20 μM) for 1h, and then infected with WT, Δppe31^Mm or comp-Δppe31^Mm for 24h, cells were stained with PI and then examined by flow cytometry. (D) BMDMs were infected with WT, Δppe31^Mm or comp-Δppe31^Mm (MOI = 10) in the presence or absence of DPI (10 μM). After 24h, cells were then stained with PI and analyzed immediately for cell death using flow cytometry. Data are shown as mean ± S.E.M. of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001. WT, wild-type; M, marinum; UI, uninfected; SC, solvent control (0.1% DMSO); ns, no significant.

Mutants for ppe31^Mtb decreases the inflammatory cytokine expressions and reduces host cell death. RAW264.7 cells were infected with H37Rv or Δppe31^Mtb (MOIs=10) for 24 h. The supernatants were collected, sterile-filtered, and assayed for TNF (A) and IL-6 (B). (C) RAW264.7 were infected with H37Rv or Δppe31^Mtb (MOI = 10) for 24h, and then subjected to Western blot analysis using antibodies raised to cleaved PARP and β-actin. (D) RAW264.7 were infected with H37Rv or Δppe31^Mtb (MOI = 10) for 24h, and cell death was detected by TUNEL staining and then examined by confocal microscope. Data are shown as mean ± S.E.M. of three independent experiments. *p<0.05, ***p<0.001, ****p<0.00001. UI, uninfected; M, marker.

PPE31 promotes mycobacteria survival in macrophage and in zebrafish. (A) WT, Δppe31^Mm or comp-Δppe31^Mm-infected BMDMs were analyzed for number of surviving internalized bacilli by CFU counting at the indicated time. (B) H37Rv or Δppe31^Mtb-infected RAW264.7 cells were analyzed for number of surviving internalized bacilli by CFU counting at the indicated time. (C) Bacterial survival advantage in the infected zebrafish. Three hundred CFU of M. marinum strains were injected to reach a comparable infection level at 4 dpi. Representative burden pictures and representative burden analysis derive from larvae collected at 4 dpi. Each point in represents 1 infected larva from a representative pool. Scale bar in 200 μm. Data are shown as mean ± S.E.M. of three independent experiments. **p < 0.01. WT, wild-type M. marinum; UI, uninfected.