Unraveling the Intricacies of Xenophagy: Lessons from Legionella pneumophila

The prevalence of drug-resistant bacterial strains has made it imperative to seek novel antibacterial approaches to fight infectious diseases. Autophagy, a fundamental eukaryotic process by which cells develop phagocytic vesicles containing cellular proteins or organelles, plays an essential role in regulating cellular homeostasis and various metabolic processes¹. As knowledge of autophagy deepens, researchers have discovered that autophagic vesicles can selectively target specific organelles, proteins, or pathogens². Xenophagy refers to the process by which cells selectively recognize intracellular pathogens through autophagy and deliver them to lysosomes for clearance³. This process is critical for the innate immune system's defense mechanism in eukaryotic organisms⁴. Consequently, elucidating its mechanisms and regulation is an essential area of research, with significant implications for the development of novel anti-infective strategies.

Legionella pneumophila is a Gram-negative intracellular pathogen that usually infects aquatic unicellular protists. However, humans can develop opportunistic infections from contact with contaminated water, leading to Legionnaires' disease, a severe form of pneumonia⁵. Upon entering the host cell, Legionella employs the type IV secretion system (T4SS) to transport roughly 330 effector proteins, aiding in the evasion of the host immune response and ensuring bacterial survival and proliferation within cells. Legionella, like Salmonella, hinders host xenophagy using effector proteins, although the precise molecular mechanism is yet to be established⁶. For example, Salmonella Typhimurium delivers type III secretion effectors (e.g., SopF) that ADP-ribosylate the vacuolar V-ATPase and block autophagy initiation⁷, whereas L. pneumophila uses the T4SS and effectors like RavZ and SidE to interfere with later stages of xenophagy^6,8. Despite these different tactics, both pathogens ultimately evade xenophagic clearance, highlighting convergent strategies in autophagy evasion. Hence, investigating these virulence proteins not only contributes to the development of specific anti-infective therapies but also has the potential to shed light on the molecular intricacies of xenophagy.

Mechanisms of xenophagy in the clearance of pathogens

In 1984, Rikihisa et al. made a significant discovery that Rickettsia-infected neutrophils produce autophagosomes, marking the first instance of involvement of autophagy in bacterial infections⁹. Further study found that the induction of autophagy in macrophages through rapamycin results in the co-localization of autophagy-related proteins LC-3 and Beclin-1 with Mycobacterium tuberculosis containing vesicles, preventing bacterial proliferation within cells¹⁰. Additionally, it is known that intracellular pathogens such as Salmonella typhi, Listeria, Shigella flexneri, and Burkholderia mallei induce autophagy, which subsequently controls their growth within host cells¹¹.

Xenophagy, a form of selective autophagy, comprises three distinct processes: cargo recognition, autophagy receptor recruitment, and lysosome-mediated degradation (as depicted in Figure 1). However, the mechanism by which cells recognize invading pathogens (cargo recognition) and initiate autophagy remains unclear. Studies suggest that when bacteria infiltrate the cytoplasm, ubiquitin ligase attaches ubiquitin to pathogen-infected vacuoles or proteins present on the surface of the pathogens^12,13,14. These labeled proteins then assemble into a ubiquitin coat, encasing the pathogen and acting as a signaling platform for the recruitment of autophagy adaptors, mainly including p62/SQSTM1, NDP52, NBR1, and optineurin. Autophagy adaptors facilitate xenophagy by interacting with proteins through the LC3-interacting domain (LIR), apart from the ubiquitin-binding domain (UBD) that mediates the binding of ubiquitinated proteins¹⁵. For instance, host E3 ligases like LRSAM1 and Parkin attach ubiquitin to invading bacteria, thereby generating the ubiquitin coat signal that flags them for xenophagic clearance^16,17. Furthermore, xenophagy receptors can also initiate xenophagy by binding to galectin on the bacterial phagosome surface or by directly interacting with bacterial surface proteins^18,19. Another perspective proposes that bacteria within the phagosome may alter the ion gradients of the vesicle through damaging the membrane, leading to conformational changes in V-ATPases on the phagosomal membrane, which recruit the Atg5-Atg12-Atg16L1 complex through binding to the WD40 domain of ATG16L1^7,20. ATG16L1 can also bind to complement C3 that coats the engulfed bacteria through its WD40 domain, recruiting the Atg5-Atg12-Atg16L1 complex to the bacterial phagosome²¹. It remains unclear whether these different recognition mechanisms target distinct bacterial species or whether they occur at different stages of infection. Further research and exploration are necessary to fully understand these mechanisms.

Strategies by pathogens to evade xenophagy

During the co-evolution with their host, intracellular pathogens have developed various mechanisms to inhibit xenophagy^22,23,24. This results in a weak xenophagic response during infections, making it challenging to observe²⁵. For instance, beta-hemolytic Streptococcus secretes a cysteine protease called SpeB that targets p62, NDP52, and NBR1 for degradation. Mutants with SpeB deletion were unable to resist autophagy, and their growth within cells was suppressed. These findings confirm the essential role of these autophagy adaptors in xenophagy. The virulence protein VirG of Shigella flexneri has been found to induce xenophagy by recruiting Atg5 to the bacterial surface¹⁹. However, the bacterium also employs the type III secretion system to transport the effector protein IcsB, which inhibits the binding of ATG5 to VirG¹⁹. Consequently, autophagy is prevented from occurring¹⁹. This competition between bacterial proteins and autophagy machinery suggests that Atg proteins can directly recognize bacterial surface proteins to initiate xenophagy. Similarly, virulence protein CpbC of Streptococcus pneumoniae induces autophagic degradation of Atg14 by forming a p62-CpbC-Atg14 complex, which in turn decreases the level of Atg14. This affects the Atg14-STX17 complex, which mediates lysosome interaction, thereby inhibiting xenophagy²⁶.

In a study utilizing Salmonella as a model bacterium, Xu et al. discovered that the membrane vesicle damage caused by the infection is detected by V-ATPase, which leads to activation of xenophagy through binding to the WD40 domain of ATG16L1⁷. This activation is not part of the typical autophagy pathway. In addition, the virulence protein SopF has the ability to mediate the ADP-ribosylation (ADPRylation) of V-ATPases and inhibit their ability to initiate xenophagy⁷. This illustrates how bacteria and their secreted virulence proteins can be used as valuable tools for investigating and understanding the complex phenomenon and mechanisms of autophagy, which are often difficult to explore under normal physiological conditions.

Legionella and xenophagy

L. pneumophila establishes a replicative niche called Legionella-containing vacuole (LCV) by manipulating essential biological activities such as host ubiquitination²⁷, vesicle transport²⁸, energy metabolism²⁹, and cell motility³⁰. Despite the presence of numerous ubiquitinated proteins within the intracellular bacteria³¹Legionella has developed various mechanisms to counteract autophagy, preventing the fusion of LCV with lysosomes^6,8.

RavZ inhibits autophagy by cleaving LC3-II

The first Legionella effector protein found to regulate autophagy is RavZ (Lpg1683). As a cysteine protease, RavZ cleaves the amide bond between the carbon-terminal glycine and the previous amino acid residue of LC3, leading to an irreversible impairment of LC3 lipidation⁸. Compared to wild-type Legionella, the mutant strain with RavZ deletion (ΔravZ) showed a significant increase in LC3-II levels and the number of LC3 puncta in the host cells after infection⁸. A genetically modified Listeria strain (ΔhlyΔprfAcLLO) can trigger strong xenophagic response upon entering cells. When co-infecting cells, wild-type Legionella can suppress the accumulation of LC3 around the ΔhlyΔprfAcLLO strain, but ΔravZ loses its ability to inhibit the strain's induction of xenophagy⁶. Furthermore, transient expression of RavZ in host cells inhibits starvation-induced autophagy⁸. These results suggest that RavZ exerts a broad trans-acting effect (i.e., acting diffusely in the cytosol) to antagonize autophagy. Surprisingly, the replication of the ΔravZ mutant was unaffected in host cells, and there was no apparent recruitment of LC3 on the LCV membrane containing the ΔravZ mutant^6,8. These findings suggest that Legionella may utilize other effector proteins, in addition to RavZ, to evade host xenophagy.

SidE family and xenophagy

The SidE family consists of four homologous proteins over 170 kDa in size, named SdeA, SdeB, SdeC, and SidE³². These virulence proteins are structurally similar with functional redundancy, all of which contain a mono ADP-ribosyltransferase (mART) domain and a phosphodiesterase (PDE) domain. Studies have found that in the absence of ubiquitin-activating enzyme (E1) and ubiquitin-conjugating enzyme (E2), the SidE family functions as an independent ubiquitin ligase (E3), catalyzing unique phosphoribosyl ubiquitination (PR-Ub) modification of substrates^33,34,35. This process occurs in two steps. Taking SdeA as an example, first, SdeA modifies the Arg42 residue of ubiquitin (Ub) with the adenosine diphosphate ribose (ADPR) group derived from NAD+ using its mART domain, generating an intermediate product of ADPR-Ub^33,36. Then, the phosphodiester bond of ADPR-Ub is cleaved by the functional PDE domain of SdeA, resulting in the release of AMP and the connection of PR-Ub with the serine residue of the substrate protein³⁶.

The SidE family has numerous eukaryotic protein substrates. In addition to mediating phosphoribosyl ubiquitination of GTPases such as Rab1A, Rab6A, and Rab33B to interfere with vesicular transport^33,37, recent studies have found that the SidE family also has antagonistic functions against xenophagy^6,38,39,40. Compared to wild-type Legionella, mutant strains lacking the SidE family coding genes (ΔsidEs) show a significant increase in the recruitment of p62 on the LCV⁶. Unlike RavZ, which acts in trans, the SidE family works in cis (acting locally at the LCV) to protect Legionella's own vacuole from xenophagy. Legionella transporting SidE family proteins cannot suppress the recruitment of p62 around co-infected ΔhlyΔprfAcLLO Listeria strains⁶. Similar to the ΔravZ strain, the ΔsidEs mutant can also evade the host xenophagy. The mechanism by which SidE antagonizes xenophagy is not clear and may be related to its interference with the host ubiquitination system mediated by phosphoribosyl ubiquitination. For example, phosphoribosylation of ubiquitin by SidE could prevent recognition of the LCV by ubiquitin-binding receptors, thereby blocking recruitment of adaptors like NDP52 or optineurin to the vacuole.

Other effectors engaged in xenophagy

In addition to RavZ and SidE, L. pneumophila encodes other effectors - including LpSpl, Lpg1137, and Lpg2936 - that target various stages of the autophagy pathway. LpSpl is a secreted enzyme that mimics sphingosine-1-phosphate lyase, cleaving sphingosine-1-phosphate (S1P) into metabolites⁴¹. This depletion of S1P by LpSpl prevents sphingosine accumulation and results in mTORC1 activation, thereby inhibiting the initiation of autophagy⁴¹. Lpg1137 is a protease that cleaves the SNARE protein Syntaxin 17 (Stx17), which is required for autophagosome-lysosome fusion⁴². By degrading Stx17, Lpg1137 interferes with autophagosome maturation and blocks the final fusion step of xenophagy⁴². Lpg2936 is a nuclear effector that causes epigenetic modifications of autophagy-related genes (such as ATG7 and LC3B), leading to their reduced expression⁴³. This down-regulation of core autophagy components inhibits the biogenesis of autophagosomes at the transcriptional level⁴³. However, little is known about the roles of LpSpl, Lpg1137, and Lpg2936 in Legionella infection; their contributions to xenophagy evasion in vivo remain to be fully elucidated.

Table 1 summarizes key L. pneumophila effectors involved in xenophagy suppression, their targets and functions, and the stages of xenophagy they affect.

Future perspective

Key uncertainties remain and suggest concrete next steps. First, the temporal hierarchy of Legionella effectors acting on xenophagy is unresolved (early LpSpl/mTORC1 versus mid-/late RavZ, SidE); second, several autophagy-modulating effectors (e.g., LpSpl, Lpg2936) lack in vivo validation; third, the cell-type heterogeneity of xenophagic responses across macrophage subsets, epithelium, and neutrophils is poorly mapped. We propose time-resolved infection atlases (pulse-chase, LC3 lipidomics, PR-ubiquitin proteomics, live reporters) to order effector actions; paired effector-deletion strains (ΔravZ/ΔsidE/ΔlpSpl/Δlpg2936) in cell-type-specific mouse models to establish causality; and single-cell/spatial multi-omics plus CRISPR/perturb-seq focusing on the V-ATPase-ATG16L1 axis, LRSAM1/Parkin tagging, and adaptor modules (OPTN/NDP52/p62). Translationally, we envision effector-directed inhibitors that block RavZ (cysteine protease active site) or SidE (mART/PDE PR-ubiquitination) to restore LC3-II pools and adaptor recruitment on the LCV, alongside host-directed boosters that enhance cargo tagging (activate LRSAM1/Parkin), stabilize adaptor-LC3 engagement, and transiently tune flux (TFEB activation, moderated mTORC1). Lung-targeted delivery (e.g., inhaled nanoparticles combining an anti-RavZ/SidE agent with a TFEB agonist) should be benchmarked by time-stamped, cell-type-resolved readouts (LC3 flux, PR-Ub occupancy, adaptor recruitment, fusion metrics) and in vivo efficacy (bacterial load, pathology, survival).

L. pneumophila has evolved an impressive array of strategies to evade xenophagy, primarily through its diverse effector proteins that interfere with cargo recognition, ubiquitination, autophagosome maturation, and lysosomal fusion. These mechanisms not only enable the pathogen to survive and replicate within host cells but also reveal novel aspects of autophagy regulation. Understanding these evasion tactics provides valuable insights into the complex interplay between host defense and microbial adaptation. Future research should focus on identifying additional effectors, elucidating their molecular targets, and clarifying their temporal regulation during infection. Moreover, leveraging these mechanistic insights could guide the development of host-directed therapies or novel antimicrobial agents that restore xenophagy function. Continued integration of advanced imaging, structural biology, and omics approaches will be crucial for dissecting the dynamic interactions between L. pneumophila and the autophagy machinery, potentially leading to innovative strategies for combating intracellular pathogens.

Finally, many Legionella effectors are multifunctional, affecting multiple host processes beyond autophagy. The same effectors that help L. pneumophila evade xenophagy can also modulate host cell death (apoptosis) and immune signaling pathways, further promoting bacterial survival23,30,35,44,45. This crosstalk between autophagy evasion and other host-defense mechanisms underscores the need to study Legionella's virulence tactics holistically. Exploring how autophagy-targeting effectors simultaneously influence apoptosis and inflammation will provide a more comprehensive understanding of host-pathogen interactions and may reveal new targets for therapeutic intervention.

Figure 1: Schematic representation of the mechanism of xenophagy. When bacteria invade the cytoplasm, ubiquitin is conjugated by ubiquitin ligase to proteins present on the pathogen's surface or pathogen-infected vacuoles, followed by the formation of a ubiquitin coat that encases the pathogen. This ubiquitin coat acts as a signaling platform for the recruitment of autophagy adaptors. Additionally, the autophagy receptors can initiate xenophagy by binding to the bacterial phagosome surface galectin or directly interacting with bacterial surface proteins. Please click here to view a larger version of this figure.

Table 1: Xenophagy-evasion strategies of L. pneumophila effectors. This table lists selected L. pneumophila effector proteins, their molecular targets or activities, and the stage of the xenophagy pathway they disrupt.