ATG9B regulates bacterial internalization via actin rearrangement

Summary Invasive bacterial pathogens are internalized by host cells through endocytosis, which is regulated by a cascade of actin rearrangement signals triggered by host cell receptors or bacterial proteins delivered into host cells. However, the molecular mechanisms that mediate actin rearrangement to promote bacterial invasion are not fully understood. Here, we show that the autophagy-related (ATG) protein ATG9B regulates the internalization of various bacteria by controlling actin rearrangement. ATG knockout screening and knockdown experiments in HeLa cells identified ATG9B as a critical factor for bacterial internalization. In particular, cells with ATG9B knockdown exhibited an accumulation of actin filaments and phosphorylated LIM kinase and cofilin, suggesting that ATG9B is involved in actin depolymerization. Furthermore, the kinase activity of Unc-51-like autophagy-activating kinase 1 was found to regulate ATG9B localization and actin remodeling. These findings revealed a newly discovered function of ATG proteins in bacterial infection rather than autophagy-mediated immunity.

(c-f) ATG9A, ATG9B KO cells or siRNA treated ATG9A, ATG9B KD cells were infected with GAS at an MOI of 10, and CFUs were quantified at 1 hpi (c, e) and 2 hpi (d, f).
(g-j) A549 cells (g, h) and HBEpCs (i, j) were treated with ATG9B-targeted siRNA and were infected with GAS at an MOI of 10 and CFUs were quantified at 1 hpi (g, i) and 2 hpi (h, j). NS.

Figure S1 .
Figure S1.Confirmation of ATG knockout and ATG9A or ATG9B knockdown cells, related to Figure 1.(a) Immunoblotting analysis of ATG KO Hela cell lysates using the indicated antibodies.(b) Genome sequence of the indicated region in ATG9B KO HeLa cells.Genomic DNA purified from ATG9B KO HeLa cells, generated by knocking out ATG9A using ATG9B #2, was subjected to Sanger sequencing.(c) Immunoblotting analysis of cell lysates of HeLa cells undergoing a 48-h transfection with the indicated siRNA oligonucleotides (48 h) using an ATG9A-specific antibody.(d-f) qPCR analysis of ATG9B mRNA expression in HeLa cells (d), A549 cells (e), or HBEpCs (f) at 48 h post-transfection with ATG9B siRNA.Data shown represent individual values and mean ± SEM.P values were calculated by one-way ANOVA, followed by Tukey's multiple comparison test; *P < 0.05, **P < 0.01, ***P < 0.001.
Figure S2 Figure S2.Measurement of CFUs by gentamicin protection assay, related to Figure 1.(a,b) Hela cells with ATG KO were infected with GAS at an MOI of 10 and CFUs were quantified at 1 hpi (a) and 2 hpi (b).(c-f)ATG9A, ATG9B KO cells or siRNA treated ATG9A, ATG9B KD cells were infected with GAS at an MOI of 10, and CFUs were quantified at 1 hpi (c, e) and 2 hpi (d, f).

Figure S3 .
Figure S3.Immunoblot quantification of autophagy receptor and LC3 variation in ATG9A and ATG9B KO cells during GAS infection, related to Figure2.(a-d) HeLa cells with ATG KO were infected with GAS for 0 h (non-infection; NI), 2 h, or 4 h.NDP52 (a), OPTN (b), p62 (c), and LC3 (d) protein levels were quantified.GAPDH was used as the loading control.ATG7 and FIP200 KO cells were used as a control that cannot be induced into autophagy.Data shown represent individual values and means ± SEM.P values were calculated by one-way ANOVA, followed by Tukey's multiple comparison test; *P < 0.05, **P < 0.01, ***P < 0.001.

Figure S5 .
Figure S5.Immunoblot quantification of RhoGTPase activity, p-LIMK and p-cofilin levels, and measurement of bacterial CFUs by gentamicin protection assay, related to Figure 4. (a-c) Pull-down assay of RhoGTPase activity in ATG9B KD cells.Active RhoA (a), Rac1 (b), Cdc42 (c) levels were calculated from the respective RhoGTPase levels in whole cell lysates (WCL).(d, e) Quantification of protein levels of p-LIMK (d) and p-cofilin (e) using LIMK and cofilin levels, respectively, in ATG9B KD cells.(f-k)HeLa cells treated with ATG9B-targeted siRNA and infected with S. aureus (f, g), L. monocytogenes (h, i) or S. typhimurium (j, k) at an MOI of 10.CFUs were quantified at 1 hpi (f, h, j) and 2 hpi (g, i, k).Data shown represent individual values and means ± SEM.P values were calculated by one-way ANOVA, followed by Tukey's multiple comparison test; *P < 0.05, **P < 0.01, NS: not significant.
Figure S6 a b