Autophagosome-independent essential function for the autophagy protein Atg5 in cellular immunity to intracellular pathogens

Abstract

The physiologic importance of autophagy proteins for control of mammalian bacterial and parasitic infection in vivo is unknown. Using mice with granulocyte- and macrophage-specific deletion of the essential autophagy protein Atg5, we show that Atg5 is required for in vivo resistance to the intracellular pathogens Listeria monocytogenes and Toxoplasma gondii. In primary macrophages, Atg5 was required for interferongamma (IFN-gamma)/LPS-induced damage to the T. gondii parasitophorous vacuole membrane and parasite clearance. While we did not detect classical hallmarks of autophagy, such as autophagosomes enveloping T. gondii, Atg5 was required for recruitment of IFN-gamma-inducible p47 GTPase IIGP1 (Irga6) to the vacuole membrane, an event that mediates IFN-gamma-mediated clearance of T. gondii. This work shows that Atg5 expression in phagocytic cells is essential for cellular immunity to intracellular pathogens in vivo, and that an autophagy protein can participate in immunity and intracellular killing of pathogens via autophagosome-independent processes such as GTPase trafficking.

Figure 1

Atg5 Is Required for Cellular Immunity to Toxoplasma gondii and L. monocytogenes In Vivo (A) Survival of BCG in Atg5-deficient and control macrophages after starvation. Results are pooled from three independent experiments. Data are presented as mean ± SEM. (B) Survival of female mice after i.p. infection with 100 T. gondii parasites expressing luciferase. These data were pooled from four independent experiments. (C) Weight of mice in (B) over the course of T. gondii infection. Data are presented as mean ± SEM. (D) Light emission from mice in (B) after injection of luciferin. Data are presented as mean ± SEM. (E) Quantification of parasites in the indicated tissues using the methods and standard curve in Figure S3. These data were pooled from two independent experiments and presented as mean ± SEM; ^∗∗P < 0.001, ^∗∗∗P < 0.0001. (F) Representative images of mice 8 days after infection with T. gondii. The full data set of which these are representatives is provided in Figure S2. (G) Survival of male mice after i.p. infection with 200 T. gondii parasites. This dose is higher than that used in female mice in (B). (H) Survival of mice after inoculation with 2 × 10⁵ CFUs of L. monocytogenes. (I) L. monocytogenes colony forming units in spleen or liver 3 days after infection. These data were pooled from at least three independent experiments (20 control mice and 19 ATG5^flox/flox-Lyz-Cre mice).

Figure 2

Atg5 Is Required for IFN-γ-Induced Clearance of Toxoplasma gondii from Macrophages (A) The proportion of macrophages containing at least one T. gondii parasite; ^∗p < 0.05, ^∗∗∗p < 0.0001. (B) The number of T. gondii parasites per vacuole in infected cells; ^∗∗∗p < 0.0001. For (A) and (B), data were pooled from at least three independent experiments in which at least 210 cells were counted per condition and presented as mean ± SEM. (C) Representative immunofluorescence images of T. gondii-infected macrophages 20 hr after infection, stained with anti-mouse F4/80 (red), anti-T. gondii (green), and DAPI (blue). Scale bar = 10 μm.

Figure 3

Atg5 Is Not Globally Required for IFN-γ-Induction of Transcription or NO Production (A) Shown is the induction of expression of the indicated genes as measured by qRT-PCR at 16 hr after treatment of control, or Atg5-deficient macrophages with LPS, IFN-γ, or the combination of IFN-γ/LPS. Data were pooled from two independent experiments and presented as mean ± SEM. There were no statistically significant differences between control and Atg5-deficient macrophages. (B) Shown are the levels of NO produced by macrophages stimulated by IFN-γ/LPS for 18 hr. DPI is an inhibitor of NO production. Data were pooled from two independent experiments and presented as mean ± SEM.

Figure 4

Electron Microscopic Analysis of the Clearance of T. gondii by IFN-γ-Activated Macrophages (A) Macrophages treated as indicated were infected for 5 hr and then analyzed by EM. All images shown are from IFN-γ/LPS-activated macrophages. In (A), parasites (Tg) within Atg5-deficient cells reside within conventional parasitophorous vacuoles (PV) bounded by host ER (arrow heads). (B) The parasitophorous vacuole membrane is a single unit membrane (arrow), surrounded in places by host ER (arrow head). (C) Parasites within activated control macrophages are found within vacuoles that are undergoing membrane blebbing and vesiculation (arrows). (D) Enlarged view from (C) shows membrane blebs protruding from the parasitophorous vacuole membrane (arrows). Scale bars = 0.5 μm.

Figure 5

Electron Microscopic Analysis of Damage to the Parasitophorous Vacuole in IFN-γ/LPS-Activated Control Macrophages (A) Clearance of T. gondii infection by control macrophages treated with IFN-γ/LPS is associated with a process of vacuolar membrane damage including vesiculation and blebbing 5 hr after infection. Findings here were specific for control macrophages and were not observed in Atg5-deficient macrophages. (A) shows a heavily damaged parasite within the cytosol following dissolution of the parasitophorous vacuole membrane. The parasite plasma membrane shows evidence of damage. Scale bar = 0.5 μm. (B–E) Examples of double-membrane-bound compartments forming in the vicinity of the degraded parasite. Scale bars = 0.1 μm. (F) Parasite residing within a parasitophorous vacuole that is undergoing extensive membrane blebbing and vesiculation. A prominent cluster of membrane vesicles and flattened cisternae are found at the posterior end (arrows). Scale bar = 0.5 μm. (G–I) Enlarged views of the membrane vesicles showing flattened cisternae (arrows). Scale bars = 0.1 μm.

Figure 6

Atg5 Is Required for IFN-γ/LPS-Induced Targeting of IIGP1 to the T. gondii Parasitophorous Vacuole (A and B) Macrophages were stimulated with IFN-γ/LPS and infected with T. gondii as in Figure 2, and then stained for the indicated markers 2 hr after infection. Scale bar = 10 μm for (A)–(F). In (A) and (B), localization of IIGP1 (red) with GFP-expressing T. gondii (green) after infection of IFN-γ/LPS-activated control or Atg5-deficient macrophages is shown. DAPI (blue) staining of nuclei. In control cells, IIGP1 decorates the parasitophorous vacuole in a cap, while in Atg5-deficient cells, IIGP1 (red) was shunted to large intracellular inclusions and was not recruited to the parasite-containing vacuole. (C) Intracellular fate of GFP-expressing parasites in IFN-γ/LPS-activated control cells. LAMP1 (green) was recruited selectively to IIGP1 (red) positive vacuoles and often formed a partial cap, associated with material sloughing from the vacuole surface. Nuclei were visualized with DAPI (blue) staining. (D) Similar to (C), except wild-type parasites were detected with rabbit anti-Toxoplasma and secondary antibodies conjugated to Alexa355 (blue). (E) Parasites within Atg5-deficient cells did not recruit IIGP1 or become LAMP1-positive. Asterisks indicate GFP-expressing parasites. However, IIGP1-positive inclusions (red in Figure 6B) were strongly associated with LAMP1. Insert shows enlarged view of intracellular inclusion. IIGP1 (red) and LAMP1 (green) were closely associated, but not strictly colocalized (scale bar = 1 μm). (F) Similar to (E), except wild-type parasites were detected with rabbit anti-Toxoplasma and secondary antibodies conjugated to Alexa355 (blue). (G) Quantitation of IIGP1 colocalization with T. gondii parasites in IFN-γ/LPS-activated macrophages 1, 2, and 5 hr postinfection. Data were collected from two independent experiments counting at least 600 vacuoles and presented as mean ± SEM.