Mitophagy-mediated immune evasion: Shared strategies of pathogens Authors: Jun Li, Weijian Li, Dan Wang, Yulan Liu Published: 2026 ABSTRACT Mitophagy selectively eliminates dysfunctional mitochondria, playing a pivotal role in mitochondrial quality control and cellular homeostasis. Emerging evidence reveals that certain pathogens exploit mitophagy to evade host immune defenses. Here, we provide novel insights into the regulatory mechanisms of mitophagy by integrating it with mitochondrial dynamics, and systematically review the mechanisms by which intracellular bacteria, viruses, and parasites utilize mitophagy to subvert host innate immunity. Notably, some pathogens dynamically regulate mitophagy at different infection stages to facilitate their survival, and the mitophagy show a positive correlation with mitochondrial fission/fragmentation. This review further summarizes four therapeutic strategies to counteract pathogen-induced immune evasion via mitophagy: 1) pharmacological modulation of mitophagy pathways; 2) mitochondria-targeted nanomaterials delivery systems; 3) mitochondria transplantation; 4) nanoengineered mitochondria. Moreover, two core mechanistic questions that remain to be addressed: (1) The mechanisms of time-dependent mitophagy-mediated immune evasion during infection, and (2) the mechanistic connection between mitochondrial dynamics and mitophagy. Future studies could employ label-free holographic tomography microscopy combined with artificial intelligence to visualize and quantify pathogen-induced subcellular alterations, enhancing our understanding of how mitophagy is manipulated, particularly through stage-specific regulation. These insights may open new avenues for treating infections resistant to conventional therapies. Introduction Mitochondria are dynamic and multifunctional organelles, primarily recognized for their central role in oxidative phosphorylation and adenosine triphosphate (ATP) production in eukaryotic cells [1]. Beyond energy metabolism, mitochondria participate in a wide range of cellular processes, including membrane potential generation, calcium homeostasis, lipid oxidation and synthesis, redox signaling, iron-sulfur cluster biosynthesis, innate immune signaling, and the regulation of cell death [2,3]. However, the inevitably generated byproduct reactive oxygen species (ROS) during oxidative phosphorylation can compromise the mitochondrial DNA, protein, and lipids. Environmental stressors further exacerbate this vulnerability, underscoring the need for robust quality control mechanisms [4]. Eukaryotic cells have therefore evolved mitochondrial quality control (MQC) systems to continuously monitor and maintain mitochondrial homeostasis. Among these, mitophagy is a selective autophagic process that degrades and recycles damaged mitochondria within the cell, thereby preserving both mitochondria and cellular homeostasis [3]. Mitochondrial dyshomeostasis, characterized by structural and functional defects, contributes to the pathogenesis of numerous diseases [5]. Given its central role within MQC, dysregulation of mitophagy has emerged as a critical factor in neurodegenerative, metabolic, and infectious diseases [[6], [7], [8]]. Among various disease contexts, pathogen infection exhibits particularly striking functional heterogeneity in mitophagy. Its divergence is far greater than the relatively consistent roles of mitophagy observed in neurodegenerative and cancer disorders [[9], [10], [11]]. This renders the mechanisms of mitophagy in pathogen infection a particularly complex endeavor. Although mitophagy is often regarded as cytoprotective, growing evidence highlights its detrimental effects during pathogen infection [7,12]. The role of mitophagy appears highly context-dependent, especially in the setting of pathogen infection. For instance, Mycobacterium bovis (Multiplicity of infection, MOI = 10 for 24 h) facilitates intracellular survival in J774A.1 and BMDM macrophages by inducing mitophagy [13], whereas spike protein of Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus inhibits mitophagy in both mice (5 μg/mouse/d for 10 days via tracheal intubation) and H9C2 cells (2.5 μg/mL for 3 h), suggesting induction of mitophagy may serve as a novel potential therapeutic target for COVID-19-associated cardiopulmonary manifestations [14]. These contrasting findings suggest that mitophagy operates as a context-dependent double-edged sword, with its specific function shaped by pathogen-specific tactics and host cellular conditions. Nevertheless, our systematic understanding of mitophagy in pathogen immune evasion and disease pathogenesis remains fragmented and insufficient, highlighting the urgent need for a comprehensive review to integrate current knowledge. Specifically, this gap manifests in three key understudied dimensions. First, existing findings are largely siloed across individual pathogen studies, with limited integrative analysis of regulatory mechanisms across bacteria, viruses, and parasites. Moreover, mitophagy may exert opposite roles at different stages of infection (early vs. persistence infection). For example, herpes simplex virus type 1 (HSV-1, MOI = 1) rapidly induces PINK1-Parkin-mediated mitophagy at early intracellular infection stages (3 h post-infection) in mouse microglial BV2 cells, which was inhibited in the late stages of HSV-1 infection (24 h post-infection) [15], suggesting that stage-dependent dynamic regulation of mitophagy may be a more general pattern that remains to be verified across more pathogen infection models. Furthermore, the roles of mitophagy in many common infection models remain underexplored. In this review, we systematically summarize the involvement of mitophagy in the immune evasion across a range of pathogens, and highlight the key signaling pathways and regulatory mechanisms underlying these processes. We further summarize four promising intervention strategies targeting mitophagy for anti-infective therapy: Pharmacological modulation of mitophagy pathways, mitochondria-targeted nanomaterials delivery systems, mitochondria transplantation, and nanoengineered mitochondria. We also synthesize the limited available data from these understudied models to identify priority directions for future investigation. Based on this integrative analysis of existing literature, we outline two core mitophagy-related mechanistic questions that urgently need to be addressed. 1) the mechanisms underlying time-dependent mitophagy regulation during pathogen-mediated immune evasion, and 2) the crosstalk between mitochondrial dynamics (fission and fusion) and mitophagy in shaping infection outcomes. Here, we integrate existing evidence to examine the association between mitochondrial fission/fragmentation and mitophagy across infectious contexts, which has not been systematically evaluated across the full spectrum of bacteria, viruses, and parasites in infectious contexts to date. Additionally, to address the technical bottleneck of studying stage-specific mitophagy dynamics, we also highlight an emerging imaging technology that enables real-time, quantitative analysis of subcellular mitochondrial alterations during infection. Specifically, this approach integrates label-free holographic tomography microscopy (HTM) with artificial intelligence to visualize and quantitatively assess pathogen-induced subcellular alterations. This technology offers enhanced capability to investigate the manipulation of mitophagy, particularly through stage-specific regulatory mechanisms, providing deeper insights into its dynamic roles during infection. Elucidating the context-specific functions of mitophagy may not only enhance our understanding of pathogen pathogenesis but also pave the way for novel anti-infective strategies. An introduction to mitophagy Currently, mitophagy is commonly classified based on its dependence on ubiquitin into the well-characterized PINK1-Parkin-dependent (ubiquitin-mediated) pathway and alternative PINK1-Parkin-independent (non-ubiquitin-based) mechanisms [16,17]. Notably, growing evidence indicates that mitochondrial dynamics, such as fission and fusion, critically influence mitophagy and contribute to various disease mechanisms [18]. In this review, an integrated classification system for mitochondrial removal that incorporates these dynamics is proposed (Fig. 1). This system comprises three major categories: 1) The ubiquitin-dependent pathway, represented by the canonical PINK1-Parkin axis; 2) receptor-mediated pathways, encompassing ubiquitin-independent receptors such as BCL2-interacting protein 3 (BNIP3) and FUNDC1; and 3) non-classical mitochondrial clearance pathways, including microvesicles and mitochondrial-derived vesicles (MDVs). This framework provides novel insights into the regulatory mechanisms of mitophagy within the context of mitochondrial fission and fusion dynamics.Fig. 1Schematic overview of mitophagy pathways.A. The ubiquitin-dependent pathway is exemplified by PINK1-PRKN-mediated mitophagy mechanism. Upon stress, PINK1 stabilizes and accumulates on the outer mitochondrial membrane, where it recruits PRKN and catalyzes the formation of phosphorylated ubiquitin (p-ubiquitin) chains on mitochondrial proteins. This leads to the ubiquitination and degradation of mitofusin 1 (MFN1), a key mediator of mitochondrial fusion. Additionally, PINK1 promotes mitochondrial fission and mitophagy initiation through phosphorylation of DNM1L. Finally, the adaptor proteins, such as optineurin, NDP52, and p62, bind to p-ubiquitin chains via their structural domains, thereby facilitating recognition by LC3 on the autophagosomal membrane.B. FUNDC1 receptor-mediated mitophagy. Under basal conditions, continuous phosphorylation of FUNDC1 at Ser13 by CSNK2 promotes its binding to mitochondrial fusion protein optic atrophy 1 (OPA1), while phosphorylation at Tyr18 by SRC kinase sterically hinders the LC3B-binding domain, collectively suppressing mitophagy initiation. Upon stress, ULK1-mediated phosphorylation of FUNDC1 at Ser17, couple with dephosphorylation at Ser13 and Tyr18 (regulated by PGAM5 and reduced SRC activity, respectively), coordinately modulates mitochondrial fission (via DNM1L) and fusion (through OPA1) to promote mitophagy.C. BNIP3 and BNIP3L receptor-mediated mitophagy. BNIP3 and BNIP3L promote mitophagy initiation through three primary mechanisms: First, they enhance their binding to LC3B and GABARAPL2 upon activation. Moreover, both proteins competitively dissociate BECN1 from BCL2 at the BH3 domain, resulting in BECN1 release and subsequent promotion of mitophagy. Additionally, BNIP3 and BNIP3L facilitate DNM1L translocation to mitochondria, driving mitochondrial fragmentation. Collectively, these actions coordinate to initiate mitophagy.D. Noncanonical mitophagy via microvesicles and MDVs. Microvesicles facilitate the extracellular release of mitochondrial content through budding from the plasma membrane, which is subsequently subjected to phagocytosis by microphages. In contrast, mitochondrial-derived vesicles (MDVs) form via the budding of mitochondrial fragments, a process dependent on PINK1–Parkin signaling. MDVs subsequently traffic to lysosomes for degradation and material recycling. PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase. Schematic overview of mitophagy pathways. A. The ubiquitin-dependent pathway is exemplified by PINK1-PRKN-mediated mitophagy mechanism. Upon stress, PINK1 stabilizes and accumulates on the outer mitochondrial membrane, where it recruits PRKN and catalyzes the formation of phosphorylated ubiquitin (p-ubiquitin) chains on mitochondrial proteins. This leads to the ubiquitination and degradation of mitofusin 1 (MFN1), a key mediator of mitochondrial fusion. Additionally, PINK1 promotes mitochondrial fission and mitophagy initiation through phosphorylation of DNM1L. Finally, the adaptor proteins, such as optineurin, NDP52, and p62, bind to p-ubiquitin chains via their structural domains, thereby facilitating recognition by LC3 on the autophagosomal membrane. B. FUNDC1 receptor-mediated mitophagy. Under basal conditions, continuous phosphorylation of FUNDC1 at Ser13 by CSNK2 promotes its binding to mitochondrial fusion protein optic atrophy 1 (OPA1), while phosphorylation at Tyr18 by SRC kinase sterically hinders the LC3B-binding domain, collectively suppressing mitophagy initiation. Upon stress, ULK1-mediated phosphorylation of FUNDC1 at Ser17, couple with dephosphorylation at Ser13 and Tyr18 (regulated by PGAM5 and reduced SRC activity, respectively), coordinately modulates mitochondrial fission (via DNM1L) and fusion (through OPA1) to promote mitophagy. C. BNIP3 and BNIP3L receptor-mediated mitophagy. BNIP3 and BNIP3L promote mitophagy initiation through three primary mechanisms: First, they enhance their binding to LC3B and GABARAPL2 upon activation. Moreover, both proteins competitively dissociate BECN1 from BCL2 at the BH3 domain, resulting in BECN1 release and subsequent promotion of mitophagy. Additionally, BNIP3 and BNIP3L facilitate DNM1L translocation to mitochondria, driving mitochondrial fragmentation. Collectively, these actions coordinate to initiate mitophagy. D. Noncanonical mitophagy via microvesicles and MDVs. Microvesicles facilitate the extracellular release of mitochondrial content through budding from the plasma membrane, which is subsequently subjected to phagocytosis by microphages. In contrast, mitochondrial-derived vesicles (MDVs) form via the budding of mitochondrial fragments, a process dependent on PINK1–Parkin signaling. MDVs subsequently traffic to lysosomes for degradation and material recycling. PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase. The ubiquitin-dependent pathway The PINK1-Parkin pathway is the best characterized ubiquitin-dependent mitophagy mechanism (Fig. 1A). PINK1 acts as a key sensor of mitochondrial damage and helps maintain mitochondrial functions under stress [19]. Under basal conditions, PINK1 is continuously imported into the inner mitochondrial membrane (IMM) through the translocase of the outer membrane (TOM) complex, and cleaved by the IMM protease PARL, leading to its degradation [19,20]. Loss of mitochondrial membrane potential (MMP, ΔΨm) stabilizes PINK1 on the outer membrane of mitochondria (OMM), where it phosphorylates ubiquitin at Ser 65, recruiting and activating the E3 ubiquitin ligase Parkin [[21], [22], [23]]. Parkin subsequently ubiquitinates numerous OMM proteins, including the fusion GTPases MFN1 and MFN2, targeting them for proteasomal degradation and thereby inhibiting mitochondrial fusion [24,25]. This process prevents the dissemination of damaged components and facilitates the elimination of impaired mitochondria. PINK1 also promotes mitochondrial fission through phosphorylation of DNM1L [26]. Parkin-mediated ubiquitination generates phosphorylated ubiquitin chains, which serve as docking sites for autophagy adaptors such as calcium-binding and coiled-coil domain 2 (CALCOCO2, also known as NDP52), optineurin (OPTN), sequestosome-1 (p62/SQSTM1), neighbor of BRCA1 gene 1 (NBR1), and Tax1 binding protein 1 (TAX1BP1) [[27], [28], [29]]. These mitochondrial proteins/p-ubiquitin/adaptor protein complexes (also known as ubiquitinated cargos) then interact with microtubule-associated protein light chain 3 (MAP1LC3/LC3)- decorated autophagosomal membrane via LC3-interacting regions (LIR) motifs of adaptors proteins, ultimately leading to lysosomal degradation. Notably, PINK1−/− or Parkin−/−mice exhibit mild or no phenotype abnormalities, and mitochondrial fusion remains inhibited, suggesting the existence of alternative pathways for maintaining mitochondrial homeostasis [25]. Receptor-mediated pathways Apart from ubiquitin-dependent pathways, several OMM-localized receptors mediate mitophagy through direct LC3 interaction via LIR domains, independent of ubiquitin. Well-characterized examples include FUN14 domain-containing protein 1 (FUNDC1), BNIP3, and BCL2/adenovirus E1B interacting protein 3-like (BNIP3L/Nix) [30,31]. FUNDC1 is a ubiquitously expressed mitochondrial protein with three transmembrane domains anchored in the OMM (Fig. 1B). Its cytosolic N-terminal domain bears an LIR motif that binds LC3B, while its C-terminus extends into the intermembrane space. FUNDC1 primarily mediates mitophagy under hypoxia, selenite, and carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) treatment [9,32]. Its activity is regulated by phosphorylation and is closely linked to mitochondrial dynamics [9]. A phosphorylation-based “on-off” switch mechanism operates via three key sites: Ser13, Tyr18, and Ser17 (Table 1) [35,36]. Under basal conditions, casein kinase 2 (CSNK2)-mediated phosphorylation at Ser13 promotes binding to optic atrophy 1 (OPA1), supporting mitochondrial fusion, while SRC kinase-mediated phosphorylation at Tyr18 impedes LC3 interaction [9,32,33]. During hypoxia, PGAM5 dephosphorylates Ser13, dissociating OPA1 and promoting DNM1L-mediated fission [9,33]. Concurrent reduced SRC kinase activity allows spontaneous Tyr18 dephosphorylation, exposing the LC3B binding interface and activating mitophagy [32,33], and UNC-51-like autophagy-activating kinase 1 (ULK1)-mediated phosphorylation at Ser17 enhances LC3 affinity, driving mitophagy [34]. Taken together, ULK1 and PGAM5/SRC regulate FUNDC1 to coordinate mitochondrial dynamics and removal.Table 1Phosphorylation-dependent regulation of mitophagy by FUNDC1.Table 1Regulatory siteBasal conditionStress conditionMechanismOutcomeRefsSer13CSNK2 induced-phosphorylation-Promoting FUNDC1 binding to mitochondrial fusion protein OPA1↓Mitophagy[9]-PGAM5 induced-dephosphorylationPromoting mitochondrial fission by disrupting OPA1 binding while facilitating FUNDC1-DNM1L interaction↑Mitophagy[9,33]Tyr18SRC induced-phosphorylation-Hindering the LC3B-binding domain of FUNDC1↓Mitophagy[32,33]-Dephosphorylation induced by reduced SRC kinase activityExposing the LC3B binding interface↑Mitophagy[32,33]Ser17-ULK1 induced-phosphorylationEnhancing FUNDC1′ affinity for LC3B↑Mitophagy[34]CSNK2, a constitutive serine/threonine kinase; DNM1L/DRP1, Dynamin 1 Like; FUNDC1, FUN-14 domain-containing protein 1; MAP1LC3B/LC3B, microtubule associated protein 1 light chain 3 β; OPA1, Optic atrophy 1; PGAM5, phosphoglycerate mutase family member 5; SRC, SRC proto-oncogene; ULK1, UNC-51-like autophagy-activating kinase 1. Table 1Phosphorylation-dependent regulation of mitophagy by FUNDC1.Table 1Regulatory siteBasal conditionStress conditionMechanismOutcomeRefsSer13CSNK2 induced-phosphorylation-Promoting FUNDC1 binding to mitochondrial fusion protein OPA1↓Mitophagy[9]-PGAM5 induced-dephosphorylationPromoting mitochondrial fission by disrupting OPA1 binding while facilitating FUNDC1-DNM1L interaction↑Mitophagy[9,33]Tyr18SRC induced-phosphorylation-Hindering the LC3B-binding domain of FUNDC1↓Mitophagy[32,33]-Dephosphorylation induced by reduced SRC kinase activityExposing the LC3B binding interface↑Mitophagy[32,33]Ser17-ULK1 induced-phosphorylationEnhancing FUNDC1′ affinity for LC3B↑Mitophagy[34]CSNK2, a constitutive serine/threonine kinase; DNM1L/DRP1, Dynamin 1 Like; FUNDC1, FUN-14 domain-containing protein 1; MAP1LC3B/LC3B, microtubule associated protein 1 light chain 3 β; OPA1, Optic atrophy 1; PGAM5, phosphoglycerate mutase family member 5; SRC, SRC proto-oncogene; ULK1, UNC-51-like autophagy-activating kinase 1. Phosphorylation-dependent regulation of mitophagy by FUNDC1. CSNK2, a constitutive serine/threonine kinase; DNM1L/DRP1, Dynamin 1 Like; FUNDC1, FUN-14 domain-containing protein 1; MAP1LC3B/LC3B, microtubule associated protein 1 light chain 3 β; OPA1, Optic atrophy 1; PGAM5, phosphoglycerate mutase family member 5; SRC, SRC proto-oncogene; ULK1, UNC-51-like autophagy-activating kinase 1. BNIP3 and BNIP3L are hypoxia-induced OMM proteins belonging to the BH3-only subfamily of BCL-2 proteins and serve as pro-apoptotic factors [37,38]. Each contains a single C-terminal transmembrane domain that anchors it to the OMM, with the N-terminal region projecting into the cytosol where it harbors both a BH3 domain for BCL-2 family interaction and an LIR motif for binding LC3B and GABARAPL2 [38]. This unique topology enables dual regulation of apoptosis and mitophagy [39]. Their expression is upregulated by hypoxia-inducible factor 1-alpha (HIF-1α) under hypoxia, positioning BNIP3 and BNIP3L as key mediators coupling oxygen sensing to MQC [40]. The HIF-1α-BNIP3/BNIP3L signaling induces mitophagy through diverse pathways: 1) Phosphorylation of BNIP3 at Ser17 and Ser24 or BNIP3L at Ser34, Ser35, and Ser81 promote LIR binding to LC3/GABARAPL2 [38,41,42]; 2) disruption of the BECN1–BCL2 complex, releasing BECN1 to induce autophagy [38,43]; 3) facilitation of DNM1L recruitment and mitochondrial fragmentation (Fig. 1C) [[44], [45], [46]]. Furthermore, both BNIP3 and BNIP3L are transcriptionally regulated by forkhead box O3 (FOXO3), which binds their promoters to enhance expression [47,48]. Although BNIP3 and BNIP3L share structural and functional similarities, they exhibit distinct roles. BNIP3 is widely expressed across multiple tissues and participates in diverse pathophysiological processes. It is regulated through AMP-activated protein kinase (AMPK)-mediated phosphorylation of ULK1/2 at Ser555, thereby connecting cellular energy sensing with mitophagy [49]. In contrast, BNIP3L plays an essential role in erythrocyte maturation by mediating the programmed removal of mitochondria, a critical process in terminal differentiation. Deficiency in BNIP3L leads to impaired erythropoiesis, delayed erythroid maturation, anemia, and splenomegaly [[50], [51], [52]]. Pathologically, BNIP3 often acts as a tumor suppressor [53], whereas BNIP3L may facilitate oncogenesis in specific malignancies [54]. These differences underscore their complementary roles in MQC. However, their interplay with mitochondrial dynamics proteins such as OPA1 and DNM1L remains poorly characterized. Non-classical mitochondrial clearance pathways Non-classical pathways also contribute to MQC. Among these, MDVs represent a distinct and selective mechanism (Fig. 1D). Unlike plasma membrane-derived microvesicles [55], MDVs bud directly from mitochondrial membranes, forming 70–150 nm vesicles that deliver damaged components to endo-lysosomal compartments for degradation [56,57]. MDV biogenesis requires PINK1 and Parkin [58], and often precedes PINK1-Parkin-mediated mitophagy under mild stress [58,59]. This suggests that MDVs may represent an early, selective response to mitochondrial damage, preferentially removing toxic components without triggering full organellar autophagy [57]. Notably, MDVs formation is independent of core autophagy-related (ATG) proteins but depends on DNM1L [58,60]. Although sharing PINK1-Parkin regulation with mitophagy, MDVs operate via vesicular budding rather than autophagosome encapsulation, and do not involve conventional autophagic machinery. Therefore, this review supposes classifying MDVs as a selective mitochondrial component clearance pathway. The precise role of ubiquitination signaling in MDVs biogenesis, however, warrants further investigation. Mitophagy is intricately linked to mitochondrial fission, fusion, and biogenesis, collectively maintaining functional homeostasis. Mitochondrial fission is essential for efficient mitophagy, as it segregates damaged mitochondrial components into smaller fragments that are more readily encapsulated and degraded by autophagosomes [20], through DNM1L recruitment of DNM1L and action of mitophagy mediators such as Parkin, FUNDC1, BNIP3, and BNIP3L [33,61,62]. Conversely, suppression of mitochondrial fusion helps confine dysfunctional components and supports the fragmentation required for efficient stress responses, via PINK1-Parkin-mediated degradation of MFN1 and MFN2 or OMA1-mediated cleavage of OPA1 upon mitochondrial depolarization [25,63,64]. While mitophagy is increasingly implicated in immune evasion, the roles of mitochondrial dynamics in this process remain underexplored. Roles of mitophagy in immune evasion of pathogens Although originally recognized as a guardian of mitochondrial quality, mitophagy has been co-opted by numerous pathogens to undermine host immunity. This subversion arises from the ability of pathogens to reprogram mitochondrial homeostasis, either by hyperactivating or suppressing mitophagy, to establish a permissive environment for their survival. Intracellular pathogens, in particular, have evolved sophisticated mechanisms to exploit mitophagy to achieve immune evasion. Immune evasion is, by definition, the process by which pathogens or tumors evade recognition and elimination by the host immune system under immunological pressure through diverse mechanisms. Based on mechanistic differences, immune evasion can be categorized into three types: (1) Signaling-based evasion, such as modulating signaling pathways; (2) metabolic evasion, such as mitophagy-mediated metabolic reprogramming; and (3) inflammatory evasion, such as regulating inflammasome activation and reshaping cytokine profiles. Rather than following a universal “one-size-fits-all" regulatory pattern, pathogens modulate host mitophagy in a direction-specific manner (induction or inhibition) tailored to their immune evasion demands. For example, the canonical strategy for most viruses is to induce mitophagy to degrade core immune signaling molecules such as mitochondrial antiviral signaling (MAVS) protein, while intracellular bacteria employ bidirectional regulation of mitophagy (either induction or inhibition) to secure metabolic support or disrupt inflammatory signaling, depending on their intracellular niche and infection stage. Importantly, the regulatory direction of mitophagy is tightly coupled to the three mechanistic modes of immune evasion defined above. In this section, we systematically reviewed how intracellular bacteria (e.g., Listeria monocytogenes, Mycobacterium, Salmonella), viruses, and parasites manipulate mitophagy to subvert host innate defenses, organized according to the three mechanistic modes of immune evasion defined above. Bacteria Bacteria are prokaryotes and do not have mitochondria (Table 2). Although they have a complete prokaryotic metabolic system and can carry out core metabolic reactions autonomously, they still highly rely on the host for specific nutrients (e.g., certain amino acids, nucleotides, or cofactors) or metabolic intermediates. Mitochondria serve as the central hub of host cellular metabolism. Importantly, many facultative intracellular bacteria have evolved sophisticated strategies to manipulate host mitophagy, thereby subverting host immunity through the three mechanistically distinct modes of immune evasion defined above. In the following sections, we systematically summarize how different bacterial pathogens exploit mitophagy to achieve immune evasion via these three modes (Table 3).Table 2Characteristics of bacteria, viruses, and parasites.Table 2Core CharacteristicsBacteriaVirusesParasitesMechanism of PersistencePersisters, BiofilmGenome integration, Latent infectionAntigen variation, Complex life historyPresence of mitochondriaNo (Prokaryote)No (Cell-free structure)Yes (Eukaryote)Metabolic systemPossess a prokaryotic metabolic system, can metabolize independently but is highly rely on the hostWithout an independent metabolic system, completely relies on the host.Possess independent simplified and unique metabolic pathways, can partially rely on its own metabolismCore metabolic strategyHighly flexible, can utilize host nutrientsCompletely hijack the host's metabolic machinery1)Hijacks the host's nutrients;2)Relies on its own mitochondriaTypical infection outcomeChronic infection, antibiotic resistancePersistent/latent infectionChronic/subclinical/inapparent infection, host malnutritionTable 3Mechanisms of mitophagy-mediated immune evasion in bacterial infection.Table 3PathogenBacterial factor(s)Infected subjectInfection timeMOIMitochondrial dynamicsCellular response and target (s) Immune nodes affectedMitophagy directionOutcomeRefsListeria monocytogenes-BMDM, MEF0-4 h10-↑RNF144A, ↑Ubiquitinated BECN1 (K117/K427), ↓BECN1, ↓LC3-contained autophagosome, ↓Autophagy-Promoting pathogen survival[65]InlKTissue cultured cells-N/A (treatment with InlK protein)-↑MVP/InlK interaction, ↑Recruitment MVP to bacterial surface, ↓Autophagic marker (Ubiquitin, p62, LC3B), ↓Autophagy-Promoting pathogen survival[66]LLOPMs, THP-13 h5- (mitochondrial fission inhibitor Mdivi-1 alleviates infection)↓MMP, ↑Ca2+ influx, ↑Dissociation LRR domain of NLRX1 with NACHT, ↑NLRX1 oligomerizationUp (mitophagy inhibitor Mdivi-1 alleviates infection)Promoting pathogen survival[67]LLOHeLa, Jeg3, Vero1-4 h and overnight50Biphasic: ↑Fragmentation (1-4 h), Normalization (overnight)↑Ca2+ influx, ↓MMP, ↓ATP-Promoting pathogen survival[68]LLOHeLa1 h50↑Fragmentation↓DNM1L mitochondrial localization, ↓MFF-Promoting pathogen survival[69]LLOHCT116, U2OS2 h20 for HCT116,50 for U2OS↑Fragmentation↑Mic10-Promoting infection[70]LLOHeLa2 h50↑Fragmentation↑Mic60 binding to LLO-Promoting pathogen survival[71]Mycobacterium bovis-J774A.1, BMDM24 h10-↑PINK1, ↑Parkin, ↑Mitochondrial LC3B-II, ↓XenophagyUpPromoting pathogen survival[13]Mycobacterium tuberculosis-BMDM, RAW 264.724, 48 h1-↑HIF1α, ↑BNIP3, ↑Mitochondrial LC3B-IIUpPromoting pathogen survival[72]PE_PGRS47 (Rv2741)BMDCs3 h20-↓LC3 puncta, ↓LC3-II/LC3-I, ↑p62, ↓Xenophagy-Promoting pathogen survival[73]PE6 (Rv0335c)RAW264.724 hN/A (treatment with PE6 protein)-↓Autophagy flux, ↓P-ULK1, ↓LC3B-II, ↑MTOR, ↑p62, ↓Autophagy-Promoting pathogen survival[74]RipA (Rv1477)RAW264.724 hN/A (treatment with RipA protein)-↑PI3K-AKT1-mTOR signaling axis, ↓LC3B-II, ↑MTOR, ↑p62, ↓Autophagy-Promoting pathogen survival[75]S. Tm-RAW264.72, 4, 8 h100- (mitochondrial fission inhibitor Mdivi-1 aggravates infection)↑MMP, ↓Autophagy, ↓LC3, ↓PINK1Down (mitophagy inhibitor Mdivi-1 aggravates infection)Promoting pathogen survival[76,77]SipABMDM, RAW 264.72, 8 h10↓Fragmentation, ↓DNM1L (fission protein, mitochondrial fission inhibitor Mdivi-1 aggravates infection)↑MMP, ↓mtROS, ↓mtDNA, ↓Cytochrome c- (mitophagy inhibitor Mdivi-1 aggravates infection)Promoting pathogen survival[78]SopFHeLa20 min - 3 h100-Biphasic: After 40 min (mainly 2 h): ↓LC3B, ↓LC3A, ↓LC3C, and ↓GABARAP, ↓V-ATPase-ATG16L1, ↓Xenophagy-Promoting pathogen survival[79]SseLHeLa, RAW 264.74-14 hN/A (treatment with SseL protein)-12 h: ↓LC3, ↓p62, ↓ubiquitination near the SCV, ↓Autophagy flux-Promoting pathogen survival[80]spvCJ774A.1,HeLa2, 8, 16, 24 h10-2 h: ↓LC3-II, ↓Beclin-1, ↓NLRP3, ↓Autophagy-Promoting pathogen survival[81,82]SopFHEK293A, HCT116, HeLa4 h180-↓FAM134B, ↓ER-phagy-Promoting pathogen survival[83]SseJHeLa, RAW 264.72, 4 h100↑Fragmentation (mitochondrial fission inhibitor Mdivi-1 alleviates infection)↑LC3, ↑PHB2, ↑PGAM5, ↑PINK1-Parkin pathway, ↓COX4, ↓MFN2Up (mitophagy inhibitor Mdivi-1 alleviates infection)Promoting pathogen survival[84]Salmonella EnteritidisRfbDRAW264.70.5, 2 h100-↓LC3 puncta, ↓LC3-II, ↓Xenophagy-Promoting pathogen survival[81]S. Tm,Shigella flexneri-HeLa1 h50Unchanged--Promoting pathogen survival[68]Brucella abortus-HeLa48, 72 h2000↑Fragmentation↑Brucella-containing vacuoles, ↑LC3, ↑HIF1α-BNIP3L-MitophagyUpPromoting pathogen survival[85,86]Legionella pneumophilaLpg1137HeLa4 h1, 5, 50-↓Stx17, ↓ER-mitochondria communication, ↓Bax, ↓Apoptosis, ↓LC3, ↓Autophagy-Promoting pathogen survival[87]MitFRAW 264.7, hMDMs6 h10↑Fragmentation, ↑DNM1L (fission protein, mitochondrial fission inhibitor Mdivi-1 further aggravates infection)↑Ran-RanBP2-WASP/Arp2/3-DNM1L functional axis, ↑Warburg-like metabolism- (mitophagy inhibitor Mdivi-1 further aggravates infection)Promoting pathogen survival[88,89]Burkholderia pseudomalleiBipDRAW264.7, PMs4 h10↑Fragmentation, ↑DNM1L (fission protein)↑KLHL9-KLHL13-CUL3, ↑LC3, ↓MMP, ↓mtDNA, ↑IMMT ubiquitinationUpPromoting pathogen survival[90]KT/PKB, AKT serine/threonine kinase; ATG16L1, Autophagy-related 16-like 1; ATP, Adenosine triphosphate; BAX, BCL2 associated X; BECN1, Beclin 1; BMDM, Bone marrow-derived macrophages; BNIP3, BCL2-interacting protein 3; BNIP3L/Nix, BCL2/adenovirus E1B interacting protein 3-like; CUL3, Cullin 3; DNM1L/DRP1, Dynamin 1 Like; ER-phagy, Reticulophagy; FAM134B, Family with sequence similarity 134 member B; GABARAP, GABA type A receptor-associated protein; HIF-1α, Hypoxia-inducible factor 1-alpha; IL-10, Interleukin-10; IMMT, Inner membrane mitochondrial protein; KLHL9, Kelch-like 9; KLHL13, Kelch-like 13; LIR, LC3-interacting region; LLO, Listeriolysin O; LRR, Leucine-rich repeat; MAP1LC3/LC3, Microtubule associated protein 1 light chain 3; MCU, Mitochondrial calcium uniporter; MEFs, Mouse embryonic fibroblasts; MFF, Mitochondria fission factor; MMP, Mitochondrial membrane potential; MOI, Multiplicity of infection; mtDNA, Mitochondrial DNA; mTOR, Mammalian target of rapamycin; MVP, Major vault protein; NLRX1, Nod-like receptor family member with a mitochondrial targeting sequence; PHB2, Prohibitin 2; PI3K, Phosphatidylinositol-3-kinase; PINK1, PTEN-induced putative kinase 1; PMs, Peritoneal macrophages; PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase; RNF144A, Ring finger protein 144A; ROS, Reactive oxygen species; RPS6KB, Ribosomal protein S6 kinase; RanBP2, Ran binding protein 2; STAT3, Signal transducer and activator of transcription 3; S. Tm, Salmonella typhimurium; STX17, Syntaxin 17; ULK1, UNC-51-like autophagy-activating kinase 1; WASP, Wiskott-Aldrich syndrome protein; “-” means not mentioned; “N/A” means not applicable. Table 2Characteristics of bacteria, viruses, and parasites.Table 2Core CharacteristicsBacteriaVirusesParasitesMechanism of PersistencePersisters, BiofilmGenome integration, Latent infectionAntigen variation, Complex life historyPresence of mitochondriaNo (Prokaryote)No (Cell-free structure)Yes (Eukaryote)Metabolic systemPossess a prokaryotic metabolic system, can metabolize independently but is highly rely on the hostWithout an independent metabolic system, completely relies on the host.Possess independent simplified and unique metabolic pathways, can partially rely on its own metabolismCore metabolic strategyHighly flexible, can utilize host nutrientsCompletely hijack the host's metabolic machinery1)Hijacks the host's nutrients;2)Relies on its own mitochondriaTypical infection outcomeChronic infection, antibiotic resistancePersistent/latent infectionChronic/subclinical/inapparent infection, host malnutrition Characteristics of bacteria, viruses, and parasites. 1)Hijacks the host's nutrients; Hijacks the host's nutrients; 2)Relies on its own mitochondria Relies on its own mitochondria Table 3Mechanisms of mitophagy-mediated immune evasion in bacterial infection.Table 3PathogenBacterial factor(s)Infected subjectInfection timeMOIMitochondrial dynamicsCellular response and target (s) Immune nodes affectedMitophagy directionOutcomeRefsListeria monocytogenes-BMDM, MEF0-4 h10-↑RNF144A, ↑Ubiquitinated BECN1 (K117/K427), ↓BECN1, ↓LC3-contained autophagosome, ↓Autophagy-Promoting pathogen survival[65]InlKTissue cultured cells-N/A (treatment with InlK protein)-↑MVP/InlK interaction, ↑Recruitment MVP to bacterial surface, ↓Autophagic marker (Ubiquitin, p62, LC3B), ↓Autophagy-Promoting pathogen survival[66]LLOPMs, THP-13 h5- (mitochondrial fission inhibitor Mdivi-1 alleviates infection)↓MMP, ↑Ca2+ influx, ↑Dissociation LRR domain of NLRX1 with NACHT, ↑NLRX1 oligomerizationUp (mitophagy inhibitor Mdivi-1 alleviates infection)Promoting pathogen survival[67]LLOHeLa, Jeg3, Vero1-4 h and overnight50Biphasic: ↑Fragmentation (1-4 h), Normalization (overnight)↑Ca2+ influx, ↓MMP, ↓ATP-Promoting pathogen survival[68]LLOHeLa1 h50↑Fragmentation↓DNM1L mitochondrial localization, ↓MFF-Promoting pathogen survival[69]LLOHCT116, U2OS2 h20 for HCT116,50 for U2OS↑Fragmentation↑Mic10-Promoting infection[70]LLOHeLa2 h50↑Fragmentation↑Mic60 binding to LLO-Promoting pathogen survival[71]Mycobacterium bovis-J774A.1, BMDM24 h10-↑PINK1, ↑Parkin, ↑Mitochondrial LC3B-II, ↓XenophagyUpPromoting pathogen survival[13]Mycobacterium tuberculosis-BMDM, RAW 264.724, 48 h1-↑HIF1α, ↑BNIP3, ↑Mitochondrial LC3B-IIUpPromoting pathogen survival[72]PE_PGRS47 (Rv2741)BMDCs3 h20-↓LC3 puncta, ↓LC3-II/LC3-I, ↑p62, ↓Xenophagy-Promoting pathogen survival[73]PE6 (Rv0335c)RAW264.724 hN/A (treatment with PE6 protein)-↓Autophagy flux, ↓P-ULK1, ↓LC3B-II, ↑MTOR, ↑p62, ↓Autophagy-Promoting pathogen survival[74]RipA (Rv1477)RAW264.724 hN/A (treatment with RipA protein)-↑PI3K-AKT1-mTOR signaling axis, ↓LC3B-II, ↑MTOR, ↑p62, ↓Autophagy-Promoting pathogen survival[75]S. Tm-RAW264.72, 4, 8 h100- (mitochondrial fission inhibitor Mdivi-1 aggravates infection)↑MMP, ↓Autophagy, ↓LC3, ↓PINK1Down (mitophagy inhibitor Mdivi-1 aggravates infection)Promoting pathogen survival[76,77]SipABMDM, RAW 264.72, 8 h10↓Fragmentation, ↓DNM1L (fission protein, mitochondrial fission inhibitor Mdivi-1 aggravates infection)↑MMP, ↓mtROS, ↓mtDNA, ↓Cytochrome c- (mitophagy inhibitor Mdivi-1 aggravates infection)Promoting pathogen survival[78]SopFHeLa20 min - 3 h100-Biphasic: After 40 min (mainly 2 h): ↓LC3B, ↓LC3A, ↓LC3C, and ↓GABARAP, ↓V-ATPase-ATG16L1, ↓Xenophagy-Promoting pathogen survival[79]SseLHeLa, RAW 264.74-14 hN/A (treatment with SseL protein)-12 h: ↓LC3, ↓p62, ↓ubiquitination near the SCV, ↓Autophagy flux-Promoting pathogen survival[80]spvCJ774A.1,HeLa2, 8, 16, 24 h10-2 h: ↓LC3-II, ↓Beclin-1, ↓NLRP3, ↓Autophagy-Promoting pathogen survival[81,82]SopFHEK293A, HCT116, HeLa4 h180-↓FAM134B, ↓ER-phagy-Promoting pathogen survival[83]SseJHeLa, RAW 264.72, 4 h100↑Fragmentation (mitochondrial fission inhibitor Mdivi-1 alleviates infection)↑LC3, ↑PHB2, ↑PGAM5, ↑PINK1-Parkin pathway, ↓COX4, ↓MFN2Up (mitophagy inhibitor Mdivi-1 alleviates infection)Promoting pathogen survival[84]Salmonella EnteritidisRfbDRAW264.70.5, 2 h100-↓LC3 puncta, ↓LC3-II, ↓Xenophagy-Promoting pathogen survival[81]S. Tm,Shigella flexneri-HeLa1 h50Unchanged--Promoting pathogen survival[68]Brucella abortus-HeLa48, 72 h2000↑Fragmentation↑Brucella-containing vacuoles, ↑LC3, ↑HIF1α-BNIP3L-MitophagyUpPromoting pathogen survival[85,86]Legionella pneumophilaLpg1137HeLa4 h1, 5, 50-↓Stx17, ↓ER-mitochondria communication, ↓Bax, ↓Apoptosis, ↓LC3, ↓Autophagy-Promoting pathogen survival[87]MitFRAW 264.7, hMDMs6 h10↑Fragmentation, ↑DNM1L (fission protein, mitochondrial fission inhibitor Mdivi-1 further aggravates infection)↑Ran-RanBP2-WASP/Arp2/3-DNM1L functional axis, ↑Warburg-like metabolism- (mitophagy inhibitor Mdivi-1 further aggravates infection)Promoting pathogen survival[88,89]Burkholderia pseudomalleiBipDRAW264.7, PMs4 h10↑Fragmentation, ↑DNM1L (fission protein)↑KLHL9-KLHL13-CUL3, ↑LC3, ↓MMP, ↓mtDNA, ↑IMMT ubiquitinationUpPromoting pathogen survival[90]KT/PKB, AKT serine/threonine kinase; ATG16L1, Autophagy-related 16-like 1; ATP, Adenosine triphosphate; BAX, BCL2 associated X; BECN1, Beclin 1; BMDM, Bone marrow-derived macrophages; BNIP3, BCL2-interacting protein 3; BNIP3L/Nix, BCL2/adenovirus E1B interacting protein 3-like; CUL3, Cullin 3; DNM1L/DRP1, Dynamin 1 Like; ER-phagy, Reticulophagy; FAM134B, Family with sequence similarity 134 member B; GABARAP, GABA type A receptor-associated protein; HIF-1α, Hypoxia-inducible factor 1-alpha; IL-10, Interleukin-10; IMMT, Inner membrane mitochondrial protein; KLHL9, Kelch-like 9; KLHL13, Kelch-like 13; LIR, LC3-interacting region; LLO, Listeriolysin O; LRR, Leucine-rich repeat; MAP1LC3/LC3, Microtubule associated protein 1 light chain 3; MCU, Mitochondrial calcium uniporter; MEFs, Mouse embryonic fibroblasts; MFF, Mitochondria fission factor; MMP, Mitochondrial membrane potential; MOI, Multiplicity of infection; mtDNA, Mitochondrial DNA; mTOR, Mammalian target of rapamycin; MVP, Major vault protein; NLRX1, Nod-like receptor family member with a mitochondrial targeting sequence; PHB2, Prohibitin 2; PI3K, Phosphatidylinositol-3-kinase; PINK1, PTEN-induced putative kinase 1; PMs, Peritoneal macrophages; PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase; RNF144A, Ring finger protein 144A; ROS, Reactive oxygen species; RPS6KB, Ribosomal protein S6 kinase; RanBP2, Ran binding protein 2; STAT3, Signal transducer and activator of transcription 3; S. Tm, Salmonella typhimurium; STX17, Syntaxin 17; ULK1, UNC-51-like autophagy-activating kinase 1; WASP, Wiskott-Aldrich syndrome protein; “-” means not mentioned; “N/A” means not applicable. Mechanisms of mitophagy-mediated immune evasion in bacterial infection. KT/PKB, AKT serine/threonine kinase; ATG16L1, Autophagy-related 16-like 1; ATP, Adenosine triphosphate; BAX, BCL2 associated X; BECN1, Beclin 1; BMDM, Bone marrow-derived macrophages; BNIP3, BCL2-interacting protein 3; BNIP3L/Nix, BCL2/adenovirus E1B interacting protein 3-like; CUL3, Cullin 3; DNM1L/DRP1, Dynamin 1 Like; ER-phagy, Reticulophagy; FAM134B, Family with sequence similarity 134 member B; GABARAP, GABA type A receptor-associated protein; HIF-1α, Hypoxia-inducible factor 1-alpha; IL-10, Interleukin-10; IMMT, Inner membrane mitochondrial protein; KLHL9, Kelch-like 9; KLHL13, Kelch-like 13; LIR, LC3-interacting region; LLO, Listeriolysin O; LRR, Leucine-rich repeat; MAP1LC3/LC3, Microtubule associated protein 1 light chain 3; MCU, Mitochondrial calcium uniporter; MEFs, Mouse embryonic fibroblasts; MFF, Mitochondria fission factor; MMP, Mitochondrial membrane potential; MOI, Multiplicity of infection; mtDNA, Mitochondrial DNA; mTOR, Mammalian target of rapamycin; MVP, Major vault protein; NLRX1, Nod-like receptor family member with a mitochondrial targeting sequence; PHB2, Prohibitin 2; PI3K, Phosphatidylinositol-3-kinase; PINK1, PTEN-induced putative kinase 1; PMs, Peritoneal macrophages; PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase; RNF144A, Ring finger protein 144A; ROS, Reactive oxygen species; RPS6KB, Ribosomal protein S6 kinase; RanBP2, Ran binding protein 2; STAT3, Signal transducer and activator of transcription 3; S. Tm, Salmonella typhimurium; STX17, Syntaxin 17; ULK1, UNC-51-like autophagy-activating kinase 1; WASP, Wiskott-Aldrich syndrome protein; “-” means not mentioned; “N/A” means not applicable. Signaling-based evasion The most common strategy employed by bacteria to exploit mitophagy for immune evasion is to hijack host signaling pathways, including both mitophagy regulatory cascades and innate immune signaling axes, thereby blocking host's immune recognition and clearance. Listeria monocytogenes, a facultative intracellular pathogen, dynamically interacts with host autophagic pathways. Although approximately 35% of internalized bacteria are enveloped by LC3-positive autophagosomes within 2 h post-infection, this percentage drops to about 1% by 6 h post-infection, indicating effective evasion mechanisms in later stages (MOI = 500 for human H4 cells) [91]. Mechanistically, the E3 ubiquitin ligase ring finger protein 144A (RNF144A) induced by Listeria monocytogenes infection has been shown to target BECN1 at residues K117/K427 for degradation, ultimately promoting Listeria monocytogenes infection [65]. Given BECN1's central role in both canonical autophagy and mitophagy through autophagosome formation, its depletion may impair host mitochondrial clearance [92]. Apart from Listeria, Mycobacterium tuberculosis (M. tb) and Mycobacterium bovis (M. bovis), major causative agents of tuberculosis, have evolved subtle and multifaceted strategies to hijack the mitophagy signaling axes for immune evasion, aiming to adapt intracellular niche to prolong survival time rather than simply destroy the host [[73], [74], [75],93,94]. Growing evidence indicates that mitochondria serve as a key battlefield where Mycobacteria hijack mitophagy to evade immunity [95]. Specifically, M. bovis exploits PINK1-Parkin-mediated mitophagy pathway to competitively deplete phosphorylated-TBK1 (MOI = 10 for 24 h in J774A.1 macrophages), thereby suppressing host xenophagy (also termed antibacterial autophagy, a selective autophagic process that targets invading pathogens) and promoting its intracellular survival and pathogenicity [13]. Apart from the ubiquitin-dependent mitophagy pathway, M. tb also upregulates BNIP3 via the ROS-HIF1α axis in BMDM and RAW 264.7 cells (MOI = 1 for 24 and 48 h), triggering mitophagy to remove damaged mitochondria and maintain cellular homeostasis conducive to bacterial persistence [72]. Salmonella also manipulates multiple signaling axes to regulate mitophagy for immune evasion. Notably, current studies have reported seemingly conflicting phenotypes regarding how mitophagy modulates Salmonella enterica serovar Typhimurium (S. Tm) survival. S. Tm infection suppresses PINK1-mediated mitophagy in RAW 264.7 cells (MOI = 100 for 4 h), and the mitophagy inhibitor Mdivi-1 enhanced bacterial survival by impairing phosphorylated TANK-binding kinase 1 (p-TBK1) mitochondrial translocation, while the mitophagy inducer fisetin restored mitophagy and limited intracellular proliferation [76]. Intriguingly, a recent study demonstrated that the T3SS-2 effector SseJ activates the PINK1-Parkin pathway via OMM protein PHB2 in RAW 264.7 cells (MOI = 100 for 2 and 4 h), recruiting LC3B to encapsulate damaged mitochondria, and inhibition of mitophagy with Mdivi-1 reduces S. Tm survival [84]. Due to the limited direct literature, the role of mitophagy in S. Tm infection remains poorly understood. We, therefore, have supplemented indirect evidence concerning mitophagy or even autophagy to infer mitophagy's function during S. Tm infection. Pharmacological inhibition of mitochondrial fission via Mdivi-1 that can also inhibit mitophagy increases the survival of S. Tm while FCCP (a mitochondrial uncoupling agent that can activate mitophagy)-mediated fission induction impairs its survival in BMDM cells (MOI = 10 for 8 h) [78], suggesting that S. Tm represses mitophagy for intracellular replication and immune evasion. Parallel evidence from analogous selective autophagy pathways further strengthens this inference. S. Tm also targets the ER-phagy (also a selective autophagic process that similar to mitophagy but targets the endoplasmic reticulum) receptor FAM134B via its effector protein SopF, inhibiting its oligomerization and ER membrane scission (similar to mitochondrial fission), thereby blocking ER-phagy and remodeling ER morphology to create a favorable replication niche for bacterial survival (MOI = 180 for 1, 2, 3, and 4 h in HEK293A and BMDM cells) [83]. Consistent with this organelle-specific autophagy suppression strategy, a large body of literature has confirmed that S. Tm broadly inhibits autophagic pathways to support survival [81,82,96]. For example. effectors such as SopF inhibit the recruitment of LC3 paralogs by targeting the V-ATPase-ATG16L1 axis [79,97], and the deubiquitinase SseL reduces autophagic flux, impairing aggregate autophagy, and promoting intracellular replication [80]. These findings on ER-phagy and autophagy inhibition provide strong indirect evidence that Salmonella likely employs a conserved strategy to suppress mitophagy, thereby evading host innate immunity and supporting intracellular survival. Additionally, the direct inconsistent results regarding the role of mitophagy in S. Tm infection aforementioned may also be attributed to the dynamic mitophagic flux. Consistent with this dynamic nature, S. Tm has been shown to rapidly induce xenophagy in HeLa cells at an MOI of 100, approximately 20% of intracellular S. Tm were labeled with GFP-LC3 within 40 min post-infection (early infection), before this proportion returned to basal levels (5%-10%) by 180 min post-infection (persistence infection stage) [79]. This stage-dependent regulation of xenophagy strongly implies that mitophagy is similarly dynamically modulated across infection, with suppression occurring at later persistence stages to support long-term survival. Discrepancies among studies, particularly regarding mitophagy inhibitors, merit further investigation, and the potential confounding effect of cell culture supernatant (containing abundant dead cells) on experimental outcomes warrants careful consideration. Metabolic evasion Apart from signaling pathway manipulation, bacteria also exploit mitophagy to reprogram host cellular metabolism, creating a nutrient-permissive metabolic niche that supports their intracellular replication and persistence. Emerging evidence suggests that this mitophagy hijacking strategy may be a key immune evasion tactic for intracellular pathogens. Listeria monocytogenes can compromise host defenses via virulence factor listeriolysin O (LLO)-induced mitochondrial dysfunction. LLO triggers mitochondrial calcium uniporter (MCU)-dependent Ca2+ influx, resulting in MMP loss, mitochondrial fragmentation, and mitophagy, while pharmacological inhibition of mitophagy by Mdivi-1 (a mitochondrial fission inhibitor targeting DNM1L) has been reported to alleviate Listeria infection [[67], [68], [69], [70]]. The importance of mitochondrial dynamics relating to mitophagy is further highlighted by studies showing that MFN2 (mitochondrial fusion protein) knockout exacerbates infections by Listeria monocytogenes and M. tb infections, suggesting that activation of mitophagy contributes to immune evasion [98]. Mechanistically, LLO-induced Ca2+ overload triggers NLRX1 oligomerization, exposing its LIR domain and recruiting LC3 to damaged mitochondria to initiate mitophagy [67]. Moreover, Listeria monocytogenes infection promotes mitochondrial fragmentation [68,70,71], which may facilitate mitophagic engulfment [13,99]. Based on the documented link between mitochondrial fission and mitophagy in the field of cancer and neurological disorders [[9], [10], [11]], these observations support that mitophagy manipulation may act as a potential immune evasion strategy employed by Listeria. However, more direct genetic evidence is needed to clarify the role of mitophagy during Listeria monocytogenes infection, particularly given the highly time-dependent changes in xenophagy and mitochondrial dynamics [91,100]. For S. Tm, it manipulates mitochondrial dynamics through the T3SS effector SipA, which delays LPS-induced mitochondrial fission by impeding DNM1L recruitment. Consistently, inhibition of mitochondrial fission by Mdivi-1 further enhances S. Tm survival, whereas inducing fission with FCCP restricts its survival in BMDM cells (MOI = 10 for 8 h), indicating that S. Tm suppresses mitophagy to establish a replicative niche [78]. Beyond this unique inhibitory strategy, other intracellular pathogens also exploit mitophagy to evade immune clearance, though their mechanisms are less characterized. Brucella abortus, the causative agent of brucellosis, induces mitochondrial fragmentation and iron-HIF1α-BNIP3L-dependent mitophagy in HeLa cells (MOI = 2000 for 48 h), facilitating bacterial egress and cell-to-cell spread [85,86]. Legionella pneumophila employs a dual-pronged strategy: The effector Lpg1137 cleaves syntaxin 17 (Stx17) to disrupt ER-mitochondria contact sites, suppressing autophagy and apoptosis [87]; while the T4SS-secreted MitF triggers mitochondrial fission via the Ran-RanBP2-WASP-Arp2/3-DNM1L axis, thereby promoting bacterial survival [88,89]. Burkholderia pseudomallei, through its T3SS effector BipD, recruits the KLHL9-KLHL13-CUL3 E3 ubiquitin ligase complex to ubiquitinate IMM (Mitofilin) at K211 in RAW264.7 cells (MOI = 10 for 4 h), targeting it for autophagic degradation (mitophagy) [90]. Collectively, these findings indicate that Brucella, Legionella, and Burkholderia pseudomallei all induce mitochondrial fragmentation and utilize mitophagy to enhance intracellular survival. Inflammatory evasion In addition to signaling and metabolic strategies, some bacteria modulate mitophagy to reshape the inflammatory milieu. Specifically, mitophagy can modulate inflammatory responses by clearing damaged mitochondria that would otherwise release pro-inflammatory damage-associated molecular patterns (DAMPs) such as mtROS and mtDNA [101]. By inducing or suppressing mitophagy, bacteria can regulate the release of these DAMPs, thereby controlling inflammatory response to avoid host immune clearance. Consistent with this mechanism, the mitophagy induced by Listeria monocytogenes may serve to limit excessive inflammatory responses by removing pro-inflammatory damaged mitochondria, allowing these pathogens to persist without triggering overwhelming host inflammation [67,102,103]. A typical example is S. Tm, which has evolved a finely tuned strategy to balance inflammatory responses. Unlike many pathogens, S. Tm relies on intestinal inflammation for replication, employing T3SS effectors to induce inflammation independently of innate immune receptors [77]. It also employs multiple effectors to temper inflammation and maintain host homeostasis, reflecting a delicate host-pathogen equilibrium. Specifically, S. Tm infection induces robust STAT3-dependent inflammation while concomitantly suppressing autophagy, and STAT3 inhibition has been proven to restore autophagic function both in vitro and in vivo, further suggesting that S. Tm inhibits autophagy [77]. Notably, current research on S. Tm has predominantly focused on Type III Secretion System (T3SS) effectors, with limited attention to mitophagy. Although direct links between mitophagy and inflammatory regulation in bacterial infection remain to be fully established, the intertwining of mitophagy with inflammatory signaling represents an important mode of immune evasion that warrants further investigation. Viruses Viruses are acellular and therefore do not possess mitochondria. Owing to the absence of an autonomous metabolic system, their metabolism completely relies on the host. Host cellular metabolism is fundamentally centered on mitochondria. Notably, diverse viruses have developed strategies to manipulate host mitophagy to evade immune responses. Consistent with the mechanistic framework we defined, viruses exploit mitophagy to achieve immune evasion through the same three distinct modes: signaling-based, metabolic, and inflammatory regulation. In the following sections, we systematically summarize how different viruses co-opt host mitophagy to support their infection via these three modes (Table 4).Table 4Mechanisms of mitophagy-mediated immune evasion in virus infection.Table 4PathogenVirus gene(s)/protein (s)Infected subjectInfection timeMOIMitochondrial dynamicsCellular response and target (s) Immune nodes affectedMitophagy directionOutcomeRefsSARS-CoV-2ORF10HeLa, HEK293T24 h0.1- (mitochondrial fission inhibitor Mdivi-1 alleviates infection)↑BNIP3L, ↑LC3B, ↓IFN-I, ↓MAVS, ↓TOMM20Up (mitophagy inhibitor Mdivi-1 alleviates infection)Promoting pathogen survival[104]ORF9bCalu-348 h0.05, 0.1, 0.2, 0.3- (inhibiting mitophagy leads to fusion)↓RIG-I, ↓IRF3, ↓MAVS, ↓P-TBK1, ↓IFNsUpPromoting pathogen survival[105]Spike proteinH9C2, HPAEC24 hN/A (treatment with Spike protein)-↓PINK1, ↓LC3B-II, ↑IL-18, ↑NLRP3DownHost immune response[14]SARS-CoV-2-HEK293T, Calu-324, 48 h1↑Mitochondrial elongation, ↓DNM1L (fission protein), ↓MFF (fission protein)↑MMP, ↑OXPHOS, ↑ATP, ↑EGFR/AKT, EGFR inhibitor vandetanib exhibits antiviral effect-Promoting pathogen survival[106]SARS-CoV-2-Calu-372 h0.01↑Mitochondrial fission, ↑DNM1L (fission protein)↓MMP, ↓ATP, ↑BCL2L11, ↓NDUFAB1, ↓NDUFS7, ↓NDUFS8-Host immune response[107]SARS-CoV-2-U2OS2-48-Biphasic: ↓Early (15 h) mitochondrial length, ↑later (15 h) mitochondrial lengthNucleolus enlargement, ↑number and size of lipid droplets, ↓mitochondrial density-Promoting pathogen survival[108]SARS-CoV-2Spike S proteinBEAS-2B, A54924 hN/A (Spike protein)↑Fragmentation, ↑DNM1L (fission protein), ↓MFN2 (fusion protein)↑MtROS, ↓MMP, ↓OXPHOS, ↑NLRP3, ↑GSDMD, ↑IL-1β, ↑Glycolysis, ↑HK2, ↑PKM2, ↑Lactic acidUpHost immune response[109]Influenza ANucleoproteinU2OS, HEK293T20 h1-↓MAVs, ↓COX4, ↓TOMM20, ↓IFN-β, ↓TOLLIPUpPromoting pathogen survival[12]Influenza APB1–F2A549, HEK293T24 h0.1↑Fragmentation↓TOMM20, ↓MT-CO2, ↓MAVs, ↑LC3B-II, ↑TUFMUpPromoting pathogen survival[110]Influenza APB1–F2HeLa, J774A.1, HEK29322 h10↑Fission (Depend on DNM1L and OMA-1)↓MMP, ↓RIG-I, ↓NLRP3, ↓IFN-β, ↓NF-κB, ↓IL-1β-Promoting pathogen survival[111]Influenza A-BMDCs, BMDMs18 h10-↑NOD2-RIPK2-ULK1 signalingUpHost immune response[112]Influenza AM2 proteinHEK293T, A54924 hN/A (M2 protein)↑Fusion protein: OPA1, MFN1, MFN2; ↓Fission protein: DNM1L, MFF and FIS1↑Mitochondrial Ca2+ influx, ↑ATG5, ↓PI3K-AKT-mTOR signaling, ↓Autophagosome-lysosome fusion-Promoting pathogen survival[113]Classical swine fever virusNSP3PK-15, 3D4/224 h0.1↑Fragmentation, ↑Ubiquitination of MFN2↓LDHB, ↓Pyruvic acid, ↓Apoptosis, ↓VDAC1, ↓TOMM20, ↑Lactic acid, ↑BECN1, ↑LC3B, ↑Autophagosome-lysosome fusionUpPromoting pathogen survival[114]Human herpesvirus 8vIRF-1iBCBL-148 h- (iBCBL-1 infected with the complete genome of Human herpesvirus 8 latently)↑Fragmentation (mitochondrial fission inhibitor Mdivi-1 inhibits infection)↑Mitochondrial vIRF-1, ↑Mitochondrial LC3B-II,Up (mitophagy inhibitor Mdivi-1 inhibits infection)Promoting pathogen survival[115]↑BNIP3L (NIX), ↓ApoptosisAfrican swine fever virusProtein p17PAM, HEK293T6, 12, 24 h1 (mainly p17 protein)↑Fission↓MAVs, ↓MDA5, ↓RIG-I, ↓TOMM70, ↓TOMM20, ↑LC3B-II, ↑autophagyUpPromoting pathogen survival[116]HCV-Huh712 h5↑Fragmentation, ↑P-DNM1L (fission protein), ↑MFF (fission protein)↑Glycolysis, ↓MAVs-ISRE signaling, ↓Cytochrome C, ↓apoptosisUpPromoting pathogen survival[117]HCV-Huh724 h1↑Fission, ↑P-DNM1L (fission protein), ↓MFN2 (fusion protein)↓MMP, ↓P21, ↓VDAC1, ↑CDK1UpPromoting pathogen survival[118]HCV-Huh772 h1-↑PINK1-Parkin signaling, ↑Mitochondria-lysosome fusionUpPromoting pathogen survival[119]Herpes simplex virus 1ICP34.5, US11BV2, N2a1.5, 3, 6, 12, 24, 48, 72 h0.1, 0.5, 112 h: ↑Fission protein: DNM1L, and FIS1; ↓Fusion protein: OPA1, MFN2↓MMP, ↑mtDNA,Biphasic: Up (3 h), Down (24 h, mitophagy activator CCCP inhibits infection)Promoting pathogen survival[15,120]↓EIF2S1-ATF4- PINK1-PRKN signaling (24 h)PrionPrP106-126N2a6, 12, 24, 36 hN/A↑Fragmentation (24 h)6h: ↑LC3-II, ↑P-TBK1 ↑OPTN;Biphasic: Up (6 h); Down (12-36 h)Host immune response[121]12-24 h: ↓LC3-II, ↓P-TBK1 ↓OPTN;12-36 h: ↓PINK1-PRKN signalingAKT/PKB, AKT serine/threonine kinase; ATG5, Autophagy related 5; ATP, Adenosine triphosphate; BCL2L11, BCL2-like 11; BECN1, Beclin 1; BNIP3L/Nix, BCL2/adenovirus E1B interacting protein 3-like; CDK1, Cyclin-dependent kinase 1; CDKN1A/p21, Cyclin-dependent kinase inhibitor 1A; COX4, Cytochrome c oxidase 4; DNM1L/DRP1, Dynamin 1 Like; EGFR, Epidermal growth factor receptors; FIS1, Mitochondrial fission 1 protein; GSDMD, Gasdermin D; HCV, Hepatitis C virus; HK2, Hexokinase 2; IFN-β, Interferon-beta; IFN-I, Type I interferon; IFNs, Interferons; IL-18, Interleukin-18; IL-1β, Interleukin-1β; ISRE, IFN-stimulated response elements; LDHB, Lactate dehydrogenase B; MAP1LC3/LC3, Microtubule associated protein 1 light chain 3; MAVs, Mitochondrial antiviral signaling; MDA5, Melanoma differentiation-associated protein 5; MFF, Mitochondria fission factor; MFN1, Mitofusin-1; MFN2, Mitofusin-2; MMP, Mitochondrial membrane potential; MOI, Multiplicity of infection; MT-CO2, Mitochondrially encoded cytochrome c oxidase II; mtDNA, Mitochondrial DNA; mTOR, Mammalian target of rapamycin; mtROS, Mitochondrial reactive oxygen species; NDUFAB1, NADH ubiquinone oxidoreductase subunit AB1; NDUFS7, NADH ubiquinone oxidoreductase core subunit S7; NDUFS8, NADH ubiquinone oxidoreductase core subunit S8; NF-κB, Nuclear factor κB; NLRP3, NLR family pyrin domain containing 3; NOD2, Nucleotide-binding oligomerization domain-containing protein 2; NSP3: nonstructural protein 3; OMA1, Overlapping activity with m-AAA protease 1; OPA1, Optic atrophy 1; OPTN, Optineurin; ORF9b, Open reading frame 9 b; ORF10, Open reading frame 10; OXPHOS, Oxidative phosphorylation; PINK1, PTEN-induced putative kinase 1; PKM2, Pyruvate kinase M2; PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase; P-TBK1, Phosphorylated TANK-binding kinase 1; RIG-I, Retinoic acid-inducible gene I; RIPK2, Receptor interacting protein kinase 2; ROS, Reactive oxygen species; SARS-CoV-2, Severe acute respiratory syndrome coronavirus 2; TNF-α, Tumor necrosis factor-alpha; TOLLIP, Toll interacting protein; TOM40, Translocase of outer mitochondrial membrane 40; TOMM20, Translocase of outer mitochondrial membrane 20; TOMM70, Translocase of outer mitochondrial membrane 70; TUFM, Tu translation elongation factor, mitochondrial; ULK1, UNC-51-like autophagy-activating kinase 1; VDAC1, Voltage-dependent anion channel 1; vIRF-1, Viral interferon regulatory factor 1; “-” means not mentioned; “N/A” means not applicable. Table 4Mechanisms of mitophagy-mediated immune evasion in virus infection.Table 4PathogenVirus gene(s)/protein (s)Infected subjectInfection timeMOIMitochondrial dynamicsCellular response and target (s) Immune nodes affectedMitophagy directionOutcomeRefsSARS-CoV-2ORF10HeLa, HEK293T24 h0.1- (mitochondrial fission inhibitor Mdivi-1 alleviates infection)↑BNIP3L, ↑LC3B, ↓IFN-I, ↓MAVS, ↓TOMM20Up (mitophagy inhibitor Mdivi-1 alleviates infection)Promoting pathogen survival[104]ORF9bCalu-348 h0.05, 0.1, 0.2, 0.3- (inhibiting mitophagy leads to fusion)↓RIG-I, ↓IRF3, ↓MAVS, ↓P-TBK1, ↓IFNsUpPromoting pathogen survival[105]Spike proteinH9C2, HPAEC24 hN/A (treatment with Spike protein)-↓PINK1, ↓LC3B-II, ↑IL-18, ↑NLRP3DownHost immune response[14]SARS-CoV-2-HEK293T, Calu-324, 48 h1↑Mitochondrial elongation, ↓DNM1L (fission protein), ↓MFF (fission protein)↑MMP, ↑OXPHOS, ↑ATP, ↑EGFR/AKT, EGFR inhibitor vandetanib exhibits antiviral effect-Promoting pathogen survival[106]SARS-CoV-2-Calu-372 h0.01↑Mitochondrial fission, ↑DNM1L (fission protein)↓MMP, ↓ATP, ↑BCL2L11, ↓NDUFAB1, ↓NDUFS7, ↓NDUFS8-Host immune response[107]SARS-CoV-2-U2OS2-48-Biphasic: ↓Early (15 h) mitochondrial length, ↑later (15 h) mitochondrial lengthNucleolus enlargement, ↑number and size of lipid droplets, ↓mitochondrial density-Promoting pathogen survival[108]SARS-CoV-2Spike S proteinBEAS-2B, A54924 hN/A (Spike protein)↑Fragmentation, ↑DNM1L (fission protein), ↓MFN2 (fusion protein)↑MtROS, ↓MMP, ↓OXPHOS, ↑NLRP3, ↑GSDMD, ↑IL-1β, ↑Glycolysis, ↑HK2, ↑PKM2, ↑Lactic acidUpHost immune response[109]Influenza ANucleoproteinU2OS, HEK293T20 h1-↓MAVs, ↓COX4, ↓TOMM20, ↓IFN-β, ↓TOLLIPUpPromoting pathogen survival[12]Influenza APB1–F2A549, HEK293T24 h0.1↑Fragmentation↓TOMM20, ↓MT-CO2, ↓MAVs, ↑LC3B-II, ↑TUFMUpPromoting pathogen survival[110]Influenza APB1–F2HeLa, J774A.1, HEK29322 h10↑Fission (Depend on DNM1L and OMA-1)↓MMP, ↓RIG-I, ↓NLRP3, ↓IFN-β, ↓NF-κB, ↓IL-1β-Promoting pathogen survival[111]Influenza A-BMDCs, BMDMs18 h10-↑NOD2-RIPK2-ULK1 signalingUpHost immune response[112]Influenza AM2 proteinHEK293T, A54924 hN/A (M2 protein)↑Fusion protein: OPA1, MFN1, MFN2; ↓Fission protein: DNM1L, MFF and FIS1↑Mitochondrial Ca2+ influx, ↑ATG5, ↓PI3K-AKT-mTOR signaling, ↓Autophagosome-lysosome fusion-Promoting pathogen survival[113]Classical swine fever virusNSP3PK-15, 3D4/224 h0.1↑Fragmentation, ↑Ubiquitination of MFN2↓LDHB, ↓Pyruvic acid, ↓Apoptosis, ↓VDAC1, ↓TOMM20, ↑Lactic acid, ↑BECN1, ↑LC3B, ↑Autophagosome-lysosome fusionUpPromoting pathogen survival[114]Human herpesvirus 8vIRF-1iBCBL-148 h- (iBCBL-1 infected with the complete genome of Human herpesvirus 8 latently)↑Fragmentation (mitochondrial fission inhibitor Mdivi-1 inhibits infection)↑Mitochondrial vIRF-1, ↑Mitochondrial LC3B-II,Up (mitophagy inhibitor Mdivi-1 inhibits infection)Promoting pathogen survival[115]↑BNIP3L (NIX), ↓ApoptosisAfrican swine fever virusProtein p17PAM, HEK293T6, 12, 24 h1 (mainly p17 protein)↑Fission↓MAVs, ↓MDA5, ↓RIG-I, ↓TOMM70, ↓TOMM20, ↑LC3B-II, ↑autophagyUpPromoting pathogen survival[116]HCV-Huh712 h5↑Fragmentation, ↑P-DNM1L (fission protein), ↑MFF (fission protein)↑Glycolysis, ↓MAVs-ISRE signaling, ↓Cytochrome C, ↓apoptosisUpPromoting pathogen survival[117]HCV-Huh724 h1↑Fission, ↑P-DNM1L (fission protein), ↓MFN2 (fusion protein)↓MMP, ↓P21, ↓VDAC1, ↑CDK1UpPromoting pathogen survival[118]HCV-Huh772 h1-↑PINK1-Parkin signaling, ↑Mitochondria-lysosome fusionUpPromoting pathogen survival[119]Herpes simplex virus 1ICP34.5, US11BV2, N2a1.5, 3, 6, 12, 24, 48, 72 h0.1, 0.5, 112 h: ↑Fission protein: DNM1L, and FIS1; ↓Fusion protein: OPA1, MFN2↓MMP, ↑mtDNA,Biphasic: Up (3 h), Down (24 h, mitophagy activator CCCP inhibits infection)Promoting pathogen survival[15,120]↓EIF2S1-ATF4- PINK1-PRKN signaling (24 h)PrionPrP106-126N2a6, 12, 24, 36 hN/A↑Fragmentation (24 h)6h: ↑LC3-II, ↑P-TBK1 ↑OPTN;Biphasic: Up (6 h); Down (12-36 h)Host immune response[121]12-24 h: ↓LC3-II, ↓P-TBK1 ↓OPTN;12-36 h: ↓PINK1-PRKN signalingAKT/PKB, AKT serine/threonine kinase; ATG5, Autophagy related 5; ATP, Adenosine triphosphate; BCL2L11, BCL2-like 11; BECN1, Beclin 1; BNIP3L/Nix, BCL2/adenovirus E1B interacting protein 3-like; CDK1, Cyclin-dependent kinase 1; CDKN1A/p21, Cyclin-dependent kinase inhibitor 1A; COX4, Cytochrome c oxidase 4; DNM1L/DRP1, Dynamin 1 Like; EGFR, Epidermal growth factor receptors; FIS1, Mitochondrial fission 1 protein; GSDMD, Gasdermin D; HCV, Hepatitis C virus; HK2, Hexokinase 2; IFN-β, Interferon-beta; IFN-I, Type I interferon; IFNs, Interferons; IL-18, Interleukin-18; IL-1β, Interleukin-1β; ISRE, IFN-stimulated response elements; LDHB, Lactate dehydrogenase B; MAP1LC3/LC3, Microtubule associated protein 1 light chain 3; MAVs, Mitochondrial antiviral signaling; MDA5, Melanoma differentiation-associated protein 5; MFF, Mitochondria fission factor; MFN1, Mitofusin-1; MFN2, Mitofusin-2; MMP, Mitochondrial membrane potential; MOI, Multiplicity of infection; MT-CO2, Mitochondrially encoded cytochrome c oxidase II; mtDNA, Mitochondrial DNA; mTOR, Mammalian target of rapamycin; mtROS, Mitochondrial reactive oxygen species; NDUFAB1, NADH ubiquinone oxidoreductase subunit AB1; NDUFS7, NADH ubiquinone oxidoreductase core subunit S7; NDUFS8, NADH ubiquinone oxidoreductase core subunit S8; NF-κB, Nuclear factor κB; NLRP3, NLR family pyrin domain containing 3; NOD2, Nucleotide-binding oligomerization domain-containing protein 2; NSP3: nonstructural protein 3; OMA1, Overlapping activity with m-AAA protease 1; OPA1, Optic atrophy 1; OPTN, Optineurin; ORF9b, Open reading frame 9 b; ORF10, Open reading frame 10; OXPHOS, Oxidative phosphorylation; PINK1, PTEN-induced putative kinase 1; PKM2, Pyruvate kinase M2; PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase; P-TBK1, Phosphorylated TANK-binding kinase 1; RIG-I, Retinoic acid-inducible gene I; RIPK2, Receptor interacting protein kinase 2; ROS, Reactive oxygen species; SARS-CoV-2, Severe acute respiratory syndrome coronavirus 2; TNF-α, Tumor necrosis factor-alpha; TOLLIP, Toll interacting protein; TOM40, Translocase of outer mitochondrial membrane 40; TOMM20, Translocase of outer mitochondrial membrane 20; TOMM70, Translocase of outer mitochondrial membrane 70; TUFM, Tu translation elongation factor, mitochondrial; ULK1, UNC-51-like autophagy-activating kinase 1; VDAC1, Voltage-dependent anion channel 1; vIRF-1, Viral interferon regulatory factor 1; “-” means not mentioned; “N/A” means not applicable. Mechanisms of mitophagy-mediated immune evasion in virus infection. AKT/PKB, AKT serine/threonine kinase; ATG5, Autophagy related 5; ATP, Adenosine triphosphate; BCL2L11, BCL2-like 11; BECN1, Beclin 1; BNIP3L/Nix, BCL2/adenovirus E1B interacting protein 3-like; CDK1, Cyclin-dependent kinase 1; CDKN1A/p21, Cyclin-dependent kinase inhibitor 1A; COX4, Cytochrome c oxidase 4; DNM1L/DRP1, Dynamin 1 Like; EGFR, Epidermal growth factor receptors; FIS1, Mitochondrial fission 1 protein; GSDMD, Gasdermin D; HCV, Hepatitis C virus; HK2, Hexokinase 2; IFN-β, Interferon-beta; IFN-I, Type I interferon; IFNs, Interferons; IL-18, Interleukin-18; IL-1β, Interleukin-1β; ISRE, IFN-stimulated response elements; LDHB, Lactate dehydrogenase B; MAP1LC3/LC3, Microtubule associated protein 1 light chain 3; MAVs, Mitochondrial antiviral signaling; MDA5, Melanoma differentiation-associated protein 5; MFF, Mitochondria fission factor; MFN1, Mitofusin-1; MFN2, Mitofusin-2; MMP, Mitochondrial membrane potential; MOI, Multiplicity of infection; MT-CO2, Mitochondrially encoded cytochrome c oxidase II; mtDNA, Mitochondrial DNA; mTOR, Mammalian target of rapamycin; mtROS, Mitochondrial reactive oxygen species; NDUFAB1, NADH ubiquinone oxidoreductase subunit AB1; NDUFS7, NADH ubiquinone oxidoreductase core subunit S7; NDUFS8, NADH ubiquinone oxidoreductase core subunit S8; NF-κB, Nuclear factor κB; NLRP3, NLR family pyrin domain containing 3; NOD2, Nucleotide-binding oligomerization domain-containing protein 2; NSP3: nonstructural protein 3; OMA1, Overlapping activity with m-AAA protease 1; OPA1, Optic atrophy 1; OPTN, Optineurin; ORF9b, Open reading frame 9 b; ORF10, Open reading frame 10; OXPHOS, Oxidative phosphorylation; PINK1, PTEN-induced putative kinase 1; PKM2, Pyruvate kinase M2; PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase; P-TBK1, Phosphorylated TANK-binding kinase 1; RIG-I, Retinoic acid-inducible gene I; RIPK2, Receptor interacting protein kinase 2; ROS, Reactive oxygen species; SARS-CoV-2, Severe acute respiratory syndrome coronavirus 2; TNF-α, Tumor necrosis factor-alpha; TOLLIP, Toll interacting protein; TOM40, Translocase of outer mitochondrial membrane 40; TOMM20, Translocase of outer mitochondrial membrane 20; TOMM70, Translocase of outer mitochondrial membrane 70; TUFM, Tu translation elongation factor, mitochondrial; ULK1, UNC-51-like autophagy-activating kinase 1; VDAC1, Voltage-dependent anion channel 1; vIRF-1, Viral interferon regulatory factor 1; “-” means not mentioned; “N/A” means not applicable. Signaling-based evasion The most common strategy employed by viruses to exploit mitophagy for immune evasion is to hijack host innate immune signaling pathways, particularly the MAVS-mediated interferon signaling axis, by targeting key antiviral signaling molecules for mitophagic degradation, thereby blocking host immune recognition and clearance. Influenza A virus (IAV), a segmented RNA virus with high zoonotic potential, poses persistent pandemic threats due to its antigenic drift and shift capabilities [122]. This virus employs multiple strategies to manipulate mitochondrial function and mitophagy, including degrading MAVS, suppressing interferon responses, and modulating inflammasome activation, to facilitate its immune evasion and replication [12,110]. Specifically, IAV's nucleoprotein binds MAVS and recruits the mitophagy receptor toll interacting protein (TOLLIP) via Tyr 313 residue, inducing mitophagy to degrade MAVS and suppress IFN-I signaling in HEK293T cells (MOI = 1 for 24 h) [12]. In addition, the viral protein PB1–F2 disrupts mitochondrial immunity by translocating into the mitochondrial intermembrane space via Tom40, inducing MMP loss and suppressing RIG-I-NLRP3 signaling in HEK293 cells (MOI = 10 for 22 h) [111]. This protein also acts as a mitophagy receptor through its C-terminal LIR motif interaction with TUFM, triggering mitochondrial clearance and MAVS degradation [110]. SARS-CoV-2, the causative agent of COVID-19, adopts a comparable MAVS-targeting strategy. Its encoded viral protein open reading frame 10 (ORF10) recruits BNIP3L to induce mitophagy in HeLa cell (MOI = 0.1 for 24 h), which in turn degrades MAVS, thereby blunting type I interferon (IFN–I) responses and facilitating viral immune evasion [104]. Similarly, African swine fever virus utilizes the P17 protein to bridge SQSTM1 and TOMM70, thereby inducing MAVS degradation via mitophagy and suppressing the production of IFN-α, IL-6, and TNF-α to evade host immunity in PAM cells (MOI = 1 for 24 h) [116]. Notably, the mitochondrial peptide MOTS-c has been shown to inhibit hepatitis B virus (HBV) infection through promoting MAVS aggregation and inducing ISG15 expression [123]. Notably, aside from targeting the MAVS signaling axis, some viruses exhibit biphasic regulation of PINK1-Parkin pathway mediated mitophagy. For instance, HSV-1 induces mitochondrial damage and exerts biphasic regulation over PINK1-Parkin signaling. It activates the pathway at the early stage of infection (3 h post-infection), then suppresses it from 12 to 24 h post-infection via the viral proteins ICP34.5 and US11. Mechanistically. this dynamic modulation, mediated through the EIF2S1-ATF4 axis, is critical for viral neurovirulence, as evidenced by the enhanced HSV-1 resistance observed in Parkin-overexpressing mice [15]. Similarly, prion diseases share this biphasic regulatory pattern of mitophagy. PrP106-126 peptides induce transient mitophagy activation at 6 h post-infection, followed by inhibition of the PINK1-Parkin pathway during 12-36 h post-infection. This dysregulation leads to the accumulation of dysfunctional mitochondria and ultimately triggers neuronal apoptosis, driving the neurodegenerative pathology of prion diseases [121]. Collectively, these examples illustrate that temporal regulation of mitophagy often determines pro-versus anti-viral outcomes. Consequently, stage-specific interventions may offer promising therapeutic strategies. Successfully leveraging these approaches requires precise timing to either enhance protective mitophagy in contexts such as in prion disorders or inhibit viral-hijacked pathways like HSV-1 encephalitis. These insights highlight the context-dependent therapeutic potential of targeting mitophagy in viral infections. Metabolic evasion Apart from signaling manipulation, viruses also exploit mitophagy to reprogram host cellular metabolism, preserving functional mitochondria to ensure a continuous supply of energy and metabolic intermediates that support viral replication. For IAV, the M2 protein, typically involved in viral entry and assembly, localizes to mitochondria to enhance mitochondrial fusion and suppress mitophagy. Paradoxically, this increases ROS production and MAVS signaling, potentially amplifying innate immune activation, but it also preserves intact mitochondrial metabolic function to support viral replication [113]. Likewise, SARS-CoV-2 has been shown to hyperactivate mitochondrial bioenergetics via the EGFR-Akt pathway, increasing ATP production to support viral replication [106]. We suppose that SARS-CoV-2 suppresses mitophagy to preserve functional mitochondria necessary for energy and biomolecule production, as their elimination would impede viral proliferation (e.g., by disrupting nucleotide synthesis and virion assembly). Supporting this dynamic viewpoint, Saunders et al. [108] employed HTM, an innovative label-free approach enabling high-resolution refractive index imaging for quantitative analysis of cellular dynamics in living systems, to unveil biphasic mitochondrial alterations post SARS-CoV-2 infection: Initial mitochondrial fragmentation (15 h) followed by elongation and density loss, suggesting impaired mitochondrial translation. These observations imply that mitophagy is dynamically regulated to meet evolving viral metabolic and immune evasion needs. SARS-CoV-2 orchestrates a delicate balance between mitophagy suppression (sustaining energy supply for viral replication) and targeted mitophagy activation (degrading antiviral signaling proteins like MAVS to evade host immunity), making mitochondrial homeostasis a pivotal battlefield in host-virus conflicts. Corresponding therapeutic strategies, such as EGFR inhibitors mitophagy modulators, may require precise timing to effectively counteract viral evasion. Apart from SARS-CoV-2 and IAV, studies exploring the interplay between mitophagy and other viral infections remain relatively limited. Other viruses also rely on mitophagy regulation to maintain metabolic homeostasis for their survival. For instance, classical swine fever virus (CSFV) employs its NS3 protein to suppress LDHB expression, creating a lactate-rich microenvironment that induces fission and mitophagy, while concurrently inhibiting apoptosis via NF-κB to enhance replication [114]. Similarly, hepatitis C virus (HCV) activates DNM1L via phosphorylation (Ser616)/dephosphorylation (Ser637) to induce fission, followed by PINK1-Parkin upregulation to clear damaged mitochondria, promoting hepatocyte survival and sustaining metabolic support for viral replication [[117], [118], [119]]. In addition, human herpesvirus 8 (HHV-8), on the other hand, exploits BNIP3L-dependent mitophagy through its viral IRF-1 protein, sustaining infected cell viability and suppressing apoptosis in iBCBL-1 cells [115]. Inflammatory evasion The third mode of mitophagy-mediated immune evasion in viruses is the regulation of inflammatory responses, allowing pathogens to reshape cytokine profiles and avoid excessive inflammatory clearance while maintaining a favorable inflammatory niche for replication. Inconsistently with previous report that SARS-CoV-2 activates mitophagy in Hela cell (MOI = 0.1 for 24 h) in signaling-based evasion of virus [104], the spike protein (2.5 μg/mL can be deemed as the later stage of infection in this research) encoded by SARS-CoV-2 suppresses mitophagy, leading to mitochondrial reactive oxygen species (mtROS) accumulation and activation of the NLRP3-IL-18 inflammasome axis, contributing to cardiopulmonary injury. This effect is reversible with mitophagy inducers (e.g., urolithin A) or mitochondrial antioxidants, highlighting therapeutic potential by activating mitophagy [14]. This apparent discrepancy is consistent with the stage-specific dynamic regulation of mitophagy. ORF10-mediated mitophagy activation occurs at early infection stages to degrade MAVS for immune evasion, while spike protein-mediated mitophagy inhibition occurs at late stages to trigger inflammatory injury, reflecting the virus's evolving needs across infection phases. For IAV, the virus has evolved sophisticated mechanisms to balance inflammatory responses to avoid excessive immunopathology. Apart from viral protein-mediated regulation [111], IAV RNA also activates mitophagy through the NOD2-RIPK2-ULK1 axis, dampening NLRP3 inflammasome overactivation and IL-18/IFN-γ-mediated immunopathology in BMDCs cells (MOI = 10 for 18 h) [112]. While this mechanism may reflect host self-regulation, we suppose that it concurrently serves viral immune evasion by mitigating excessive antiviral responses. Intracellular parasites Parasites are eukaryotes and thus possess their own mitochondria. Unlike viruses and bacteria, which have relatively simple intra- or extracellular life cycles, protozoan parasites typically undergo multi-host, pleomorphic differentiation stages. Their living environments and metabolic demands vary dramatically across different stages, which can be simply divided into the following four core stages: (1) Environmental adaptation stage, (2) invasion and early intracellular establishment stage, (3) intracellular replication and proliferation stage, and (4) differentiation and transmission preparation stage. Consequently, the function and regulatory pattern of mitophagy are stage specific. Although parasites have independent and unique metabolic pathways and can partially carry out energy metabolism through their own mitochondria, they still rely on the host for nutrients by hijacking host metabolic intermediates. Host cellular metabolism is largely orchestrated by mitochondria, and importantly, parasites can manipulate host mitophagy to support infection [124]. Notably, the host mitophagy and the parasite mitophagy differ fundamentally in definition and biological function. Host mitophagy refers to the selective autophagy removal of damaged or dysfunctional mitochondria by host cell, maintaining cellular homeostasis and contributing to innate immune defense [3]. In contrast, parasite mitophagy involves the selective degradation of the parasite's own mitochondria, which is related to quality control and metabolic adaptation within the parasite. Under normal replicative conditions, it maintains mitochondrial integrity to ensure energy supply; during differentiation, it adapts to metabolic transitions; under extreme nutrient deprivation, it can trigger programmed cell death through complete mitochondrial fragmentation and autophagic degradation, thereby regulating population density or coping with stress [125,126]. Most parasites lack homologs of the classic mitophagy regulatory elements such as PINK1/Parkin, and instead rely solely on simplified autophagy core components like Atg3, Atg7, and Atg8 to carry out the regulation [125,127,128]. However, how parasites manipulate host mitophagy to facilitate immune evasion remains an understudied yet critical direction. Notably, consistent with the three-dimensional immune evasion framework (signaling evasion, metabolic evasion, and inflammatory evasion) established for bacteria and viruses in the preceding sections, parasites can theoretically modulate host mitophagy to achieve immune escape across these three dimensions. However, unlike bacteria and viruses that often coordinate all three dimensions to evade host immunity, the complex multi-stage life cycle of parasites makes metabolic adaptation their most urgent survival demand. This biological priority, paired with the fact that research on parasite-host mitophagy interactions remains in its infancy, accounts for the current landscape where evidence is predominantly restricted to the metabolic evasion dimension. Given the available literature, here we mainly focus on the metabolic evasion strategies via host mitophagy during Toxoplasma gondii and Plasmodium infection (Table 5).Table 5Mechanisms of mitophagy-mediated immune evasion in parasites infection.Table 5Intracellular parasitesParasites' gene(s)/protein (s)Infected subjectInfection timeMOIParasites' mitophagyHost mitochondrial dynamicsImmune nodes affectedHost mitophagy directionOutcomeRefsT. gondii-MEFs, Caco-26, 12, 24 h2-↑Mitochondrial elongation (6 h),↑Lipophagy, ↓lipid droplets-Promoting pathogen survival129↑Fragmentation (24 h)T. gondiiTgMAF1----↓Fusion, ↓MFN1 (Fusion protein), ↓MFN2 (Fusion protein)↓TOM70-SAM50 signaling, ↑SPOTs-Promoting pathogen survival130Plasmodium spp-MSQ43, Anopheles stephensi, mice24 hCell or Anopheles stephensi: -;-Host of Anopheles stephensi: ↑mitochondrial sizeHost of Anopheles stephensi: ↓5-HT, ↓ROS↑Host of Anopheles stephensi Mitophagy (mitophagy activator CCCP or Spermidine aggravates infection)Host immune response124Mice: injected intraperitoneally with 106 infected RBCs with P. bergheiHost of vertebrates: No detectedHost of mice: 5-HT↓5-HT, 5-hydroxytryptamine; MFN1, Mitofusin-1; MFN2, Mitofusin-2; MOI, Multiplicity of infection; ROS, Reactive oxygen species; SPOTs, structures positive for the outer mitochondrial membrane; T. gondii, Toxoplasma gondii; TgMAF1, T. gondii mitochondrial association factor 1; “-” means not mentioned. Table 5Mechanisms of mitophagy-mediated immune evasion in parasites infection.Table 5Intracellular parasitesParasites' gene(s)/protein (s)Infected subjectInfection timeMOIParasites' mitophagyHost mitochondrial dynamicsImmune nodes affectedHost mitophagy directionOutcomeRefsT. gondii-MEFs, Caco-26, 12, 24 h2-↑Mitochondrial elongation (6 h),↑Lipophagy, ↓lipid droplets-Promoting pathogen survival129↑Fragmentation (24 h)T. gondiiTgMAF1----↓Fusion, ↓MFN1 (Fusion protein), ↓MFN2 (Fusion protein)↓TOM70-SAM50 signaling, ↑SPOTs-Promoting pathogen survival130Plasmodium spp-MSQ43, Anopheles stephensi, mice24 hCell or Anopheles stephensi: -;-Host of Anopheles stephensi: ↑mitochondrial sizeHost of Anopheles stephensi: ↓5-HT, ↓ROS↑Host of Anopheles stephensi Mitophagy (mitophagy activator CCCP or Spermidine aggravates infection)Host immune response124Mice: injected intraperitoneally with 106 infected RBCs with P. bergheiHost of vertebrates: No detectedHost of mice: 5-HT↓5-HT, 5-hydroxytryptamine; MFN1, Mitofusin-1; MFN2, Mitofusin-2; MOI, Multiplicity of infection; ROS, Reactive oxygen species; SPOTs, structures positive for the outer mitochondrial membrane; T. gondii, Toxoplasma gondii; TgMAF1, T. gondii mitochondrial association factor 1; “-” means not mentioned. Mechanisms of mitophagy-mediated immune evasion in parasites infection. 5-HT, 5-hydroxytryptamine; MFN1, Mitofusin-1; MFN2, Mitofusin-2; MOI, Multiplicity of infection; ROS, Reactive oxygen species; SPOTs, structures positive for the outer mitochondrial membrane; T. gondii, Toxoplasma gondii; TgMAF1, T. gondii mitochondrial association factor 1; “-” means not mentioned. Metabolic evasion The major strategy employed by parasites to exploit mitophagy for immune evasion is to hijack host cellular metabolism. For example, parasites can exploit the host PINK1/Parkin pathway to induce mitophagy, clearing dysfunctional organelles to recycle metabolic intermediates, while removing mitochondrial DNA and MAVS that would otherwise disrupt host cellular metabolic homeostasis, thereby securing a continuous, stable supply of energy and biosynthetic precursors to support their intracellular replication [131]. Here, the parasite is the regulator, and the targets are host mitochondria and autophagic machinery. A well-characterized example of this strategy in the vertebrate intracellular replication stage is observed in Toxoplasma gondii infection. Toxoplasma gondii exploits host lipophagy to acquire essential fatty acids for proliferation. The host counteracts this through mitochondrial fusion and enhanced fatty acid oxidation at 6 h post-infection (corresponding to stage 2: Invasion and early intracellular establishment stage) to limit parasite access to these critical nutrients. Notably, host mitochondrial morphology exhibits distinct stage-specific changes during infection. Elongation around parasitophorous vacuoles occurs early (6 h post-infection), shifting to fragmentation later (24 h post-infection, corresponding to stage 3: Intracellular replication and proliferation stage). This mitochondrial fragmentation is an upstream trigger for host PINK1/Parkin-mediated mitophagy [[9], [10], [11]]. The fragmented, dysfunctional mitochondria are subsequently targeted for mitophagic degradation, releasing large amounts of free fatty acids and tricarboxylic acid cycle intermediates that are hijacked by Toxoplasma gondii for membrane biosynthesis and energy supply, enabling Toxoplasma gondii to secure fatty acids and ensure its survival [129]. The parasite effector TgMAF1 binds the host TOM70-SAM50 mitochondrial import complex, forming SPOTs structures that degrade host MFN1/2, disrupt mitochondrial fusion, and compromise host metabolic defenses [130]. Aside from the vertebrate intracellular stage, parasites also employ host mitophagy for metabolic evasion during their transmission stage in invertebrate hosts, as exemplified by Plasmodium infection of Anopheles mosquitoes. Plasmodium's complex life cycle involves alternating infections of vertebrate and Anopheles mosquitoes (the mosquito stage corresponds to stage 4: differentiation and transmission preparation stage). Plasmodium-mediated depletion of host 5-hydroxytryptamine (5-HT) reduces mosquito 5-HT uptake during blood feeding, leading to the inhibition of mosquito mitophagy and the subsequent accumulation of dysfunctional mitochondria. This mitochondrial dysfunction not only reduces ROS production in mosquitoes to avoid immune clearance, but also triggers mosquito metabolic reprogramming, shifting energy production from oxidative phosphorylation to aerobic glycolysis. The accumulated glucose and glycolytic intermediates in the mosquito midgut are then hijacked by Plasmodium to support its ookinete development and oocyst formation, ultimately facilitating Plasmodium infection [124]. Interventions to prevent mitophagy-mediated immune evasion Mitophagy serves as a critical host mechanism for eliminating damaged mitochondria, yet numerous pathogens, including bacteria, viruses, and parasites, exploit or inhibit this pathway to evade immune responses. Given that mitophagy often undergoes biphasic regulation (activation followed by inhibition, and vice versa) at different stages of infection in certain pathogen infections [15,121,129], pathogen-specific treatment strategies should be time-dependent rather than adopting a uniform approach. In this section, we outline several mitophagy-targeted interventions that may help to block pathogen immune evasion, including small-molecules modulators, mitochondria-targeted nanomaterials delivery system, mitochondria transplantation, and nanoengineered mitochondria. Small molecules pharmacological regulation of mitophagy Recent advances have identified small-molecule modulators of mitophagy as promising anti-infective agents, primarily through their targeted regulation on core mitophagy pathways such as PINK1-Parkin pathway. These compounds serve as valuable pharmacological tools for dissecting complex biological systems and identifying new drug targets, owing to their distinct advantages: 1) High membrane permeability. Their low molecular weight facilitates efficient absorption, oral bioavailability, and traversal of biological barriers such as the intestinal epithelium and blood-brain barrier. 2) Precise subcellular targeting. Conjugation with mitochondrial-localizing moieties enhances organelle-specific delivery [132]; 3) Predictable drug-target interactions. Molecular docking tools like AutoDock Vina and AlphaFold3 enable rational design and mechanistic validation; 4) Versatile delivery compatibility. Formulation with liposomes or polymeric nanoparticles improves bioavailability, particularly for hydrophilic compounds [133]; 5) Well-established synthesis. Chemical production methods are well-established, with 64% of FDA-approved small-molecule drugs in 2024 being synthetically derived [134]. Here we reviewed pharmacologically active small molecules that disrupt pathogen immune evasion by modulating mitophagy (Table 6). For a comprehensive summary of mitophagy modulators in other diseases (e.g., neurodegeneration, cancer, metabolic disorders), please see the report by Lu et al., [17] and many of these compounds may also hold promise for infectious contexts.Table 6Small molecules pharmacological regulation of mitophagy.Table 6NameTargetDisease modelMechanismOther applicationsRefsInhibitorsMdivi-1, ChloroquineNLRX1Listeria infectionListeria hijacks host mitophagy through a novel mitophagy receptor NLRX1 to evade killing, which is abrogated by Mdivi-1 or chloroquineAnti-Listeria infection[67]Mdivi-1PINK1-ParkinMycobacterium bovis infectionMycobacterium bovis induces mitophagy to suppress xenophagy by competitive utilizing p-TBK1, which is alleviated by mitophagy inhibitor Mdivi-1Anti-Mycobacterium bovis infection[13]PevonedistatCUL3Burkholderia pseudomallei infectionBurkholderia pseudomallei BipD initiates mitophagy to evade killing by recruiting host KLHL9-KLHL13-CUL3 E3 ligase to ubiquitinate IMMT, while CUL3 inhibitor Pevonedistat can increase the survival rate of infected miceAnti-Burkholderia pseudomallei infection[90]BafA1, Mdivi-1BNIP3LSARS-CoV-2 infectionORF10 degrades MAVS through BNIP3L-mediated mitophagy, thereby inhibiting the IFN-I signaling pathway and promoting viral replication, which can be reversed by autophagy inhibitors (Baf A1, Chloroquine)Anti-SARS-CoV-2 infection[104]Liensinine, Mdivi-1BNIP3LHHV-8 infectionHHV-8 activates vIRF-1-BNIP3L mediated mitophagy to promote viral replication, which is repressed by mitophagy inhibitors Liensinine and Mdivi-1Anti-HHV-8 infection[115]Ginsenoside Rg3Parkin, MFN2HCV infectionGinsenoside Rg3 suppresses HCV replication by upregulating p21, which inhibits CDK1 activity, leading to reduced DNM1L phosphorylation and restoration of HCV-induced mitophagyAnti-HCV infection[118]InducersCCCP, Rotenone, and TaurineParkinHSV-1 infectionHSV-1 inhibits mitophagy via the EIF2S1-ATF4-PRKN pathway, while mitophagy inducer CCCP, rotenone, and taurine inhibit HSV-1 infectionAnti-HSV-1 infection[15]Nicotinamide mononucleotidePINK1-ParkinPrion diseasePrP106-126 inhibits PINK1-Parkin-mediated mitophagy, while nicotinamide mononucleotide activates mitophagy and increase the survival rate of neuronsTherapeutic methods for Prion disease[121]Nicotinamide n-oxideULK1HSV-1 infectionHSV-1 induces mitochondrial damage and inhibits PINK1-Parkin-mediated mitophagy, while Nicotinamide n-oxide activates ULK1-dependent mitophagyAnti-HSV-1 infection[120,135]AcacetinTFEB (shown to modulate mitophagy)Salmonella infectionAcacetin restores the function of TFEB by dephosphorylation, thereby enhancing autophagy and lysosomal biogenesisAnti-Salmonella infection[[136], [137], [138]]FisetinPINK1-ParkinSalmonella infectionFisetin restricts the intracellular survival by promoting NAP1L1 to interact with p-TBK1, and subsequently recruiting p-TBK1 complex to mitochondria for mitophagyAnti-Salmonella infection, Anti-Escherichia coli infection, Anti-Leishmania infection[76,[139], [140], [141]]ATF4, Activating transcription factor 4; Baf A1, Bafilomycin A1; BNIP3L/Nix, BCL2/adenovirus E1B interacting protein 3-like; CCCP, Carbonyl cyanide m-chlorophenylhydrazone; CDK1, Cyclin-dependent kinase 1; CDKN1A/p21, Cyclin-dependent kinase inhibitor 1A; CUL3, Cullin 3; DNM1L/DRP1, Dynamin 1 Like; EIF2S1/eIF2α, Eukaryotic initiation factor 2α; HCV, Hepatitis C virus; HHV-8, Human herpesvirus 8; HSV-1, Herpes simplex virus; IFN-I, Type I interferon; KLHL9, Kelch-like 9; KLHL13, Kelch-like 13; MAVs, Mitochondrial antiviral signaling; Mdivi-1, Mitochondrial division inhibitor 1; MFN2, Mitofusin-2; NAP1L1, Nucleosome assembly protein 1-like 1; NLRX1, Nod-like receptor family member with a mitochondrial targeting sequence; ORF10, Open reading frame 10; PINK1, PTEN-induced putative kinase 1; PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase; PrP, prion protein; P-TBK1, Phosphorylated TANK-binding kinase 1; SARS-CoV-2, Severe acute respiratory syndrome coronavirus 2; TFEB, Transcription factor EB; ULK1, UNC-51-like autophagy-activating kinase 1; vIRF-1, viral interferon regulatory factor 1. Table 6Small molecules pharmacological regulation of mitophagy.Table 6NameTargetDisease modelMechanismOther applicationsRefsInhibitorsMdivi-1, ChloroquineNLRX1Listeria infectionListeria hijacks host mitophagy through a novel mitophagy receptor NLRX1 to evade killing, which is abrogated by Mdivi-1 or chloroquineAnti-Listeria infection[67]Mdivi-1PINK1-ParkinMycobacterium bovis infectionMycobacterium bovis induces mitophagy to suppress xenophagy by competitive utilizing p-TBK1, which is alleviated by mitophagy inhibitor Mdivi-1Anti-Mycobacterium bovis infection[13]PevonedistatCUL3Burkholderia pseudomallei infectionBurkholderia pseudomallei BipD initiates mitophagy to evade killing by recruiting host KLHL9-KLHL13-CUL3 E3 ligase to ubiquitinate IMMT, while CUL3 inhibitor Pevonedistat can increase the survival rate of infected miceAnti-Burkholderia pseudomallei infection[90]BafA1, Mdivi-1BNIP3LSARS-CoV-2 infectionORF10 degrades MAVS through BNIP3L-mediated mitophagy, thereby inhibiting the IFN-I signaling pathway and promoting viral replication, which can be reversed by autophagy inhibitors (Baf A1, Chloroquine)Anti-SARS-CoV-2 infection[104]Liensinine, Mdivi-1BNIP3LHHV-8 infectionHHV-8 activates vIRF-1-BNIP3L mediated mitophagy to promote viral replication, which is repressed by mitophagy inhibitors Liensinine and Mdivi-1Anti-HHV-8 infection[115]Ginsenoside Rg3Parkin, MFN2HCV infectionGinsenoside Rg3 suppresses HCV replication by upregulating p21, which inhibits CDK1 activity, leading to reduced DNM1L phosphorylation and restoration of HCV-induced mitophagyAnti-HCV infection[118]InducersCCCP, Rotenone, and TaurineParkinHSV-1 infectionHSV-1 inhibits mitophagy via the EIF2S1-ATF4-PRKN pathway, while mitophagy inducer CCCP, rotenone, and taurine inhibit HSV-1 infectionAnti-HSV-1 infection[15]Nicotinamide mononucleotidePINK1-ParkinPrion diseasePrP106-126 inhibits PINK1-Parkin-mediated mitophagy, while nicotinamide mononucleotide activates mitophagy and increase the survival rate of neuronsTherapeutic methods for Prion disease[121]Nicotinamide n-oxideULK1HSV-1 infectionHSV-1 induces mitochondrial damage and inhibits PINK1-Parkin-mediated mitophagy, while Nicotinamide n-oxide activates ULK1-dependent mitophagyAnti-HSV-1 infection[120,135]AcacetinTFEB (shown to modulate mitophagy)Salmonella infectionAcacetin restores the function of TFEB by dephosphorylation, thereby enhancing autophagy and lysosomal biogenesisAnti-Salmonella infection[[136], [137], [138]]FisetinPINK1-ParkinSalmonella infectionFisetin restricts the intracellular survival by promoting NAP1L1 to interact with p-TBK1, and subsequently recruiting p-TBK1 complex to mitochondria for mitophagyAnti-Salmonella infection, Anti-Escherichia coli infection, Anti-Leishmania infection[76,[139], [140], [141]]ATF4, Activating transcription factor 4; Baf A1, Bafilomycin A1; BNIP3L/Nix, BCL2/adenovirus E1B interacting protein 3-like; CCCP, Carbonyl cyanide m-chlorophenylhydrazone; CDK1, Cyclin-dependent kinase 1; CDKN1A/p21, Cyclin-dependent kinase inhibitor 1A; CUL3, Cullin 3; DNM1L/DRP1, Dynamin 1 Like; EIF2S1/eIF2α, Eukaryotic initiation factor 2α; HCV, Hepatitis C virus; HHV-8, Human herpesvirus 8; HSV-1, Herpes simplex virus; IFN-I, Type I interferon; KLHL9, Kelch-like 9; KLHL13, Kelch-like 13; MAVs, Mitochondrial antiviral signaling; Mdivi-1, Mitochondrial division inhibitor 1; MFN2, Mitofusin-2; NAP1L1, Nucleosome assembly protein 1-like 1; NLRX1, Nod-like receptor family member with a mitochondrial targeting sequence; ORF10, Open reading frame 10; PINK1, PTEN-induced putative kinase 1; PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase; PrP, prion protein; P-TBK1, Phosphorylated TANK-binding kinase 1; SARS-CoV-2, Severe acute respiratory syndrome coronavirus 2; TFEB, Transcription factor EB; ULK1, UNC-51-like autophagy-activating kinase 1; vIRF-1, viral interferon regulatory factor 1. Small molecules pharmacological regulation of mitophagy. ATF4, Activating transcription factor 4; Baf A1, Bafilomycin A1; BNIP3L/Nix, BCL2/adenovirus E1B interacting protein 3-like; CCCP, Carbonyl cyanide m-chlorophenylhydrazone; CDK1, Cyclin-dependent kinase 1; CDKN1A/p21, Cyclin-dependent kinase inhibitor 1A; CUL3, Cullin 3; DNM1L/DRP1, Dynamin 1 Like; EIF2S1/eIF2α, Eukaryotic initiation factor 2α; HCV, Hepatitis C virus; HHV-8, Human herpesvirus 8; HSV-1, Herpes simplex virus; IFN-I, Type I interferon; KLHL9, Kelch-like 9; KLHL13, Kelch-like 13; MAVs, Mitochondrial antiviral signaling; Mdivi-1, Mitochondrial division inhibitor 1; MFN2, Mitofusin-2; NAP1L1, Nucleosome assembly protein 1-like 1; NLRX1, Nod-like receptor family member with a mitochondrial targeting sequence; ORF10, Open reading frame 10; PINK1, PTEN-induced putative kinase 1; PRKN/Parkin, Parkin RBR E3 ubiquitin protein ligase; PrP, prion protein; P-TBK1, Phosphorylated TANK-binding kinase 1; SARS-CoV-2, Severe acute respiratory syndrome coronavirus 2; TFEB, Transcription factor EB; ULK1, UNC-51-like autophagy-activating kinase 1; vIRF-1, viral interferon regulatory factor 1. Most pathogens exploit mitophagy to subvert host immunity, rendering pharmacological inhibition a promising therapeutic strategy (Table 6). For instance, HHV-8 enhances viral replication through vIRF-1-BNIP3L-mediated mitophagy, which can be suppressed by liensinine (an inhibitor of autophagosome-lysosome fusion) or Mdivi-1 (an inhibitor of mitochondrial fission that is the prerequisite for mitophagy) [115]. Similarly, Ginsenoside Rg3 curbs HCV replication by upregulating p21, thereby inhibiting CDK1 activity, reducing DNM1L phosphorylation (p21-CDK1-DNM1L pathway), and counteracting HCV-induced PINK1-Parkin-mediated mitophagy [118]. Mycobacterium bovis hijacks p-TBK1 to induce PINK1-Parkin mitophagy and suppress xenophagy, while CCCP facilitate and Mdivi-1 or siPINK1 inhibits the survival of Mycobacterium bovis by modulating mitophagy [13]. SARS-CoV-2 ORF10 degrades MAVS via BNIP3L-mediated mitophagy to inhibit IFN-I signaling, which can be reversed by autophagy inhibitors (Bafilomycin a1, chloroquine) [104]. Additionally, Burkholderia pseudomallei initiates mitophagy to evade killing by recruiting host KLHL9-KLHL13-CUL3 E3 ligase to ubiquitinate IMMT, whereas CUL3 inhibitor Pevonedistat improves survival of infected mice [90]. Listeria monocytogenes further exemplifies this strategy by exploiting the mitophagy receptor NLRX1 for immune evasion, which is effectively blocked by Mdivi-1 or chloroquine [67]. Converse, certain pathogens inhibit mitophagy or modulate it dynamically, making mitophagy induction being therapeutic strategy. For example, HSV-1 suppresses mitophagy via the EIF2S1-ATF4-Parkin axis, but overexpressing Parkin or using mitophagy inducers such as CCCP (via ATF4-PINK1-Parkin axis), rotenone, and taurine (via PINK1-Parkin signaling) restore mitochondrial clearance and limit infection [15]. In an alternative mechanism, HSV-1 impairs mitochondria while blocking PINK1-Parkin signaling, yet nicotinamide N-oxide reactivates ULK1-dependent mitophagy to combat infection [120,135]. Fisetin exhibits anti-Salmonella activity by promoting NAP1L1-p-TBK1 complex assembly on mitochondria, thereby inducing PINK1-Parkin mediated mitophagy and restricting Salmonella survival [76]. Likewise, the prion fragment PrP106-126 inhibits PINK1-Parkin mitophagy, whereas nicotinamide mononucleotide (NMN) reactivates this pathway and improves neuronal survival [121]. Additionally, acacetin enhances lysosomal biogenesis and autophagic flux by dephosphorylating and activating TFEBs, providing another strategy to combat pathogens [[136], [137], [138]]. While these small molecules have shown promising preliminary effects in modulating mitophagy and blocking pathogen immune evasion in current preclinical studies, the non-specificity of these small molecules means that additional validation is still required to definitively confirm that the observed therapeutic effects are indeed mitophagy-dependent, rather than off-target metabolic toxicity. These findings demonstrate the feasibility of small molecule to modulate mitophagy and block immune evasion. Candidate compounds can be screened via molecular docking (e.g., AutoDock Vina, AlphaFold3) and validated through multi-level experiments including surface plasmon resonance (SPR), cellular assays, and in vivo models. Mitochondria-targeted nanomaterials delivery system Conventional therapeutics often suffer from poor mitochondrial specificity and limited access to intracellular pathogen niches. Nanomaterials delivery systems (NDS) offer superior targeting, stability, and reduced toxicity [142], making them promising for combating persistent infections. Mitochondria-targeted NDS is a class of subcellular precision delivery platforms that integrate nanoscale drug carriers with mitochondria-specific targeting ligands, aiming to overcome the biological barriers of the mitochondrial double-membrane structure and precisely deliver therapeutic or diagnostic agents to the mitochondrial compartment [[143], [144], [145], [146]]. In these systems, a diverse array of nanocarriers have been developed to accommodate different types of cargoes, ranging from polymers, lipids, hybrid polymers, dendrimers, inorganic nanoparticles, and so on [147,148]. To confer specific mitochondrial targeting capability, these nanocarriers are typically functionalized with various targeting ligands that leverage the unique properties of mitochondria. These targeting ligands mainly include delocalized lipophilic cations, mitochondrial targeting sequences (MTS) or mitochondria-penetrating peptides (MPPs) [147,149,150]. Among various mitochondria-targeting strategies, lipophilic cations, especially triphenylphosphonium (TPP), stand out due to their optimal balance of lipophilicity/hydrophilicity and efficient mitochondrial accumulation (Fig. 2) [155,156]. The cationic phosphonium group in TPP facilitates rapid mitochondrial membrane penetration, making TPP-conjugated nanoparticles highly effective for organelle-specific drug delivery [154,157,158]. These properties position TPP-based systems as lead candidates for clinical trials aimed at modulating mitochondrial function [159]. While mitochondria-targeted NDS have been explored in cancer [152,160], neurological disorders [161], and cardiovascular diseases [162], their application in infectious contexts remains limited.Fig. 2Schematic illustration of mitochondria-targeted nanomaterials delivery system.A, Schematic illustration of targeting the mitochondria; Lipophilic cations and other positively charged molecules can passively deliver cargo into mitochondria by leveraging the high mitochondrial membrane potential. Alternatively, N-terminal mitochondrial targeting sequences (MTS) are recognized by receptors on the outer mitochondrial membrane and subsequently processed by the translocation machinery. Nanoparticles functionalized with MTS can thus be internalized into mitochondria through this active transport pathway. Reproduced and modified with permission [151]. Copyright © 2025 Advanced Science.B, Schematic illustration of mitochondria-targeted nanomaterials via 5-(4-carboxyphenyl)-10,15,20-triphenylporphyrin (TCPP) conjugation in repressing tumor progression. KCKT is an in situ transformable nanoparticle via TCPP conjugation with pH-responsive self-assembly and sonodynamic therapy (SDT) functions. Its workflow enhances lysosomal escape, damages mitochondria, and inhibits autophagic flux by sequentially damaging lysosomes and mitochondria via nanofiber transformation and SDT, followed by suppression of lysosomal function to block mitochondrial autophagy flux. Reproduced and modified with permission [152]. Copyright © 2025 Advanced Science.C, Schematic illustration of mitochondria-targeted nanomaterials via triphenylporphyrin (TPP) conjugation in treating Leber's hereditary optic neuropathy (LHON). TISUH is a mitochondria-targeted nanomaterial that contains Idebenone (Ide)-containing fluorinated polymer via TPP conjugation and Ide-terminated poly (Ide-SUH). TISUH can delivery NADH ubiquinone oxidoreductase subunit 4 (ND4) gene to treating eyes of LHON. After cellular uptake, the positively charged TPP drives TISUH to cross the mitochondrial double membrane and reach the matrix. The fluorinated group further enhances membrane translocation, overcoming the poor uptake characteristic of diseased mitochondria. Notably, these complexes remain functional under high reactive oxygen species (ROS) conditions and can release the ND4 gene and Ide in response to elevated ROS levels. Once released, ND4 restores complex I activity while Ide facilitates electron transfer, thereby collectively improving mitochondrial function. Reproduced and modified with permission [153]. Copyright © 2021 Advanced Materials.D, Schematic illustration of mitochondria-targeted nanomaterials via dimethylmaleic anhydride moiety-modified PEG conjugation for breast cancer treatment. This study describes the synthetic scheme, mitochondria-targeting capability, and delivery mechanism of α-CD-based nanoparticles (denoted α-CD-DOX-DA). These nanoparticles carry NO via a GSH-sensitive linker and DOX via a hydrazine linker, and are functionalized with a dimethylmaleic anhydride moiety-modified PEG conjugated to the MPP peptide (KLAKLAK)2CGKRK. DOX, Doxorubicin; GSH, Glutathione; MPP, Mitochondria-penetrating peptide. Reproduced and modified with permission [149]. Copyright © 2024 Advanced Drug Delivery Reviews.E, Schematic illustration of mitochondria-targeted nanomaterials via Atv/PTP-TCeria in ameliorating acute kidney injury. Atv/PTP-TCeria is a ceria-based nanoparticle. Specifically, ceria nanoparticles were first modified with triphenylphosphine (TPP) to obtain TCeria nanoparticle, then coated with the reactive oxygen species (ROS)-responsive organic polymer mPEG-TK-PLGA and loaded with atorvastatin. Upon passive accumulation in the kidney, the Atv/PTP-TCeria nanoparticle respond to high levels of ROS by releasing the drug, while the TCeria component specifically targets mitochondria to scavenge excessive ROS, thereby ameliorating acute kidney injury. Reproduced and modified with permission [154]. Copyright © 2020 Theranostics. Schematic illustration of mitochondria-targeted nanomaterials delivery system. A, Schematic illustration of targeting the mitochondria; Lipophilic cations and other positively charged molecules can passively deliver cargo into mitochondria by leveraging the high mitochondrial membrane potential. Alternatively, N-terminal mitochondrial targeting sequences (MTS) are recognized by receptors on the outer mitochondrial membrane and subsequently processed by the translocation machinery. Nanoparticles functionalized with MTS can thus be internalized into mitochondria through this active transport pathway. Reproduced and modified with permission [151]. Copyright © 2025 Advanced Science. B, Schematic illustration of mitochondria-targeted nanomaterials via 5-(4-carboxyphenyl)-10,15,20-triphenylporphyrin (TCPP) conjugation in repressing tumor progression. KCKT is an in situ transformable nanoparticle via TCPP conjugation with pH-responsive self-assembly and sonodynamic therapy (SDT) functions. Its workflow enhances lysosomal escape, damages mitochondria, and inhibits autophagic flux by sequentially damaging lysosomes and mitochondria via nanofiber transformation and SDT, followed by suppression of lysosomal function to block mitochondrial autophagy flux. Reproduced and modified with permission [152]. Copyright © 2025 Advanced Science. C, Schematic illustration of mitochondria-targeted nanomaterials via triphenylporphyrin (TPP) conjugation in treating Leber's hereditary optic neuropathy (LHON). TISUH is a mitochondria-targeted nanomaterial that contains Idebenone (Ide)-containing fluorinated polymer via TPP conjugation and Ide-terminated poly (Ide-SUH). TISUH can delivery NADH ubiquinone oxidoreductase subunit 4 (ND4) gene to treating eyes of LHON. After cellular uptake, the positively charged TPP drives TISUH to cross the mitochondrial double membrane and reach the matrix. The fluorinated group further enhances membrane translocation, overcoming the poor uptake characteristic of diseased mitochondria. Notably, these complexes remain functional under high reactive oxygen species (ROS) conditions and can release the ND4 gene and Ide in response to elevated ROS levels. Once released, ND4 restores complex I activity while Ide facilitates electron transfer, thereby collectively improving mitochondrial function. Reproduced and modified with permission [153]. Copyright © 2021 Advanced Materials. D, Schematic illustration of mitochondria-targeted nanomaterials via dimethylmaleic anhydride moiety-modified PEG conjugation for breast cancer treatment. This study describes the synthetic scheme, mitochondria-targeting capability, and delivery mechanism of α-CD-based nanoparticles (denoted α-CD-DOX-DA). These nanoparticles carry NO via a GSH-sensitive linker and DOX via a hydrazine linker, and are functionalized with a dimethylmaleic anhydride moiety-modified PEG conjugated to the MPP peptide (KLAKLAK)2CGKRK. DOX, Doxorubicin; GSH, Glutathione; MPP, Mitochondria-penetrating peptide. Reproduced and modified with permission [149]. Copyright © 2024 Advanced Drug Delivery Reviews. E, Schematic illustration of mitochondria-targeted nanomaterials via Atv/PTP-TCeria in ameliorating acute kidney injury. Atv/PTP-TCeria is a ceria-based nanoparticle. Specifically, ceria nanoparticles were first modified with triphenylphosphine (TPP) to obtain TCeria nanoparticle, then coated with the reactive oxygen species (ROS)-responsive organic polymer mPEG-TK-PLGA and loaded with atorvastatin. Upon passive accumulation in the kidney, the Atv/PTP-TCeria nanoparticle respond to high levels of ROS by releasing the drug, while the TCeria component specifically targets mitochondria to scavenge excessive ROS, thereby ameliorating acute kidney injury. Reproduced and modified with permission [154]. Copyright © 2020 Theranostics. Although direct applications in infection models are scarce, untargeted nanocarriers delivering immune agonists (e.g., TLR7/RIG-I dual agonists) have enhanced antiviral responses against influenza and SARS-CoV-2 [[163], [164], [165]]. Meanwhile, TPP-conjugated nanotherapeutics (e.g., TPP-triptolide, TPP-paclitaxel) show efficacy in cancer and neurodegeneration through mitochondrial modulation [166,167]. Notably, nanoparticles targeting mitophagy flux suppress bladder cancer progression [152], and TPP-derived MitoQ induces autophagy in HepG2 cells [168]. These successes suggest that mitophagy-regulating NDS could be repurposed to counter pathogen evasion, though their effects (mitophagy inhibition vs. induction) depend on the conjugated drug. Apart from TPP, alternative mitochondrial-targeting ligands, such as rhodamine derivatives, and small-molecule mitophagy modulators [151], warrant exploration. Coupling these agents with TPP or other nanocarriers may unlock novel anti-infective mechanisms. The proven utility of mitochondria-targeted NDS in other diseases, combined with emerging insights into pathogen-induced mitophagy manipulation, highlights their potential for infection control. Future studies should focus on adapting these strategies to disrupt pathogen-host interactions, offering an innovative shift in antimicrobial therapy. Mitochondria transplantation Mitochondria transplantation is an emerging therapeutic strategy that involves delivering functional exogenous mitochondria to recipient cells to restore cellular homeostasis [169], providing a novel strategy to block pathogens hijacking of mitophagy. Unlike pharmacological or nanomaterial approaches, this technique directly introduces intact mitochondrial function, bypassing pathogen subversion of mitophagy. This section will systematically discuss its mechanisms, applications and challenges. Mitochondria can be transferred via both contact-dependent and contact-independent pathways (Fig. 3A) [170,172]. Tunneling nanotubes (TNT) and dendrites are two forms of membranous protrusions that facilitate intercellular mitochondrial transfer. TNT extend from the plasma membrane, varying in diameter (50–1500 nm) and length (tens to hundreds of microns), and represent the most common route. Dendrites, such as those in osteocytes, connect cells into an intercellular network that serves as a highway for mitochondrial transfer. Contact-independent pathways intercellular mitochondrial transfer include microvesicle shed from the plasma membrane, as well as direct extrusion or internalization of free mitochondria without carriers. Therapeutic mitochondria transplantation currently emphasizes contact-independent mechanisms due to their clinical feasibility, using methods such as direct tissue injection or intravascular infusion [172]. Mitochondrial transfer involves two main mechanisms [170]. 1) In stressed recipient cells, ROS activate p53 and its downstream Akt/PI3K/mTOR pathway, upregulating TNFαip2 to promote actin polymerization and TNT formation. Meanwhile, p53-activated caspase-3 cleaves S100A4, creating a gradient that directs TNT growth from low to high S100A4 levels. 2) Excess ROS in recipients trigger mitochondrial fragmentation for mitophagy, while damaged mitochondria and damage-associated molecular patterns (DAMPs) are released and taken up by donors for transmitophagy. Donor lysosomal degradation releases heme, activating HO-1 to enhance mitochondrial biogenesis and fusion; functional mitochondria are then returned. Intra-TNT mitochondrial movement along microtubules relies on the Miro1/Milton complex and kinesin, akin to axonal transport.Fig. 3Mechanisms, functions, and challenges of mitochondrial transplantation.A, Routes and mechanisms of mitochondria transfer. Mitochondria can be transferred via both contact-dependent and contact-independent pathways. Tunneling nanotubes (TNT) and dendrites are two forms of membranous protrusions that facilitate intercellular mitochondrial transfer. Contact-independent pathways intercellular mitochondrial transfer include microvesicle shed from the plasma membrane, as well as direct extrusion or internalization of free mitochondria without carriers. Mitochondrial transfer involves two main mechanisms. 1) In stressed recipient cells, ROS activate p53 and its downstream Akt/PI3K/mTOR pathway, upregulating TNFαip2 to promote actin polymerization and TNT formation. Meanwhile, p53-activated caspase-3 cleaves S100A4, creating a gradient that directs TNT growth from low to high S100A4 levels. 2) Excess ROS in recipients trigger mitochondrial fragmentation for mitophagy, while damaged mitochondria and DAMPs are released and taken up by donors for mitophagy. Donor lysosomal degradation releases heme, activating HO-1 to enhance mitochondrial biogenesis and fusion; functional mitochondria are then returned. Intra-TNT mitochondrial movement along microtubules relies on the Miro1/Milton complex and kinesin, akin to axonal transport.Reproduced and modified with permission [170]. Copyright © 2021 Signal Transduction and Targeted Therapy.B, Functional roles of mitochondria transfer. Mitochondrial transfer occurs through tissue- and cell-type-specific mechanisms and is implicated in a range of physiological and pathological processes across multiple tissues. Reproduced and modified with permission [171]. Copyright © 2020 MedComm.C, Challenges of mitochondrial transplantation include isolation and purification, storage, rapid validation of integrity and functionality, targeted delivery, potential immune responses, and long-term efficacy and safety. Reproduced and modified with permission [171]. Copyright © 2020 MedComm. Mechanisms, functions, and challenges of mitochondrial transplantation. A, Routes and mechanisms of mitochondria transfer. Mitochondria can be transferred via both contact-dependent and contact-independent pathways. Tunneling nanotubes (TNT) and dendrites are two forms of membranous protrusions that facilitate intercellular mitochondrial transfer. Contact-independent pathways intercellular mitochondrial transfer include microvesicle shed from the plasma membrane, as well as direct extrusion or internalization of free mitochondria without carriers. Mitochondrial transfer involves two main mechanisms. 1) In stressed recipient cells, ROS activate p53 and its downstream Akt/PI3K/mTOR pathway, upregulating TNFαip2 to promote actin polymerization and TNT formation. Meanwhile, p53-activated caspase-3 cleaves S100A4, creating a gradient that directs TNT growth from low to high S100A4 levels. 2) Excess ROS in recipients trigger mitochondrial fragmentation for mitophagy, while damaged mitochondria and DAMPs are released and taken up by donors for mitophagy. Donor lysosomal degradation releases heme, activating HO-1 to enhance mitochondrial biogenesis and fusion; functional mitochondria are then returned. Intra-TNT mitochondrial movement along microtubules relies on the Miro1/Milton complex and kinesin, akin to axonal transport. Reproduced and modified with permission [170]. Copyright © 2021 Signal Transduction and Targeted Therapy. B, Functional roles of mitochondria transfer. Mitochondrial transfer occurs through tissue- and cell-type-specific mechanisms and is implicated in a range of physiological and pathological processes across multiple tissues. Reproduced and modified with permission [171]. Copyright © 2020 MedComm. C, Challenges of mitochondrial transplantation include isolation and purification, storage, rapid validation of integrity and functionality, targeted delivery, potential immune responses, and long-term efficacy and safety. Reproduced and modified with permission [171]. Copyright © 2020 MedComm. Notably, mitochondria transplantation directly modulates mitophagy to rescue cellular function. Specifically, TNT facilitate mitochondrial transfer between mesenchymal stromal cells (MSCs) and endothelial cells (ECs), with approximately 25% of ECs acquiring MSC-derived mitochondria within 24 h [173]. Direct co-incubation of isolated MSCs mitochondria with ECs resulted in rapid uptake (approximately 90% within 4 h) via micropinocytosis. Although transplanted mitochondria depolarize and undergo mitophagy due to isolation stress, they still enhance ECs function without fusing with endogenous mitochondria [173]. In a cerebellar neurodegeneration model (conditionally DNM1L-knockout mice), mitochondria transplantation (using liver-derived mitochondria) transiently rescues motor dysfunction, improves mitochondrial function, and reduces mitophagy [174]. These studies confirm that mitochondria transplantation can remodel mitophagy to restore cellular homeostasis, laying the biological foundation for the therapeutic feasibility of this strategy. Mitochondria transplantation has thus emerged as a promising therapeutic strategy to restore cellular bioenergetics in damaged tissues, with substantial evidence and early clinical data supporting its preliminary feasibility, exerting broad therapeutic effects, including neuroprotection [175], regulation of cancer metabolism and proliferation [176], MQC [169], promotion of wound healing and bone remodeling [[177], [178], [179]], metabolic modulation [180], and support of hematopoiesis and attenuation of ischemia-reperfusion injury (Fig. 3B) [181]. Notably, mitochondria transplantation promotes protective CD4+ T cell responses during Mycobacterium tuberculosis infection and diminishes exhaustion and senescence in aged cells [182], suggesting great potential in infection contexts. Despite the promising prospects, several critical biosafety constraints remain to be addressed before widespread clinical translation. While isolated mitochondria have been successfully transplanted into diverse recipient cells in vivo and in vitro via multiple strategies (direct injection, co-culture, magnetomitotransfer, and cell-penetrating peptide) [183], only direct injection is currently feasible for in vivo applications. Specific unresolved obstacles span the entire therapeutic workflow, covering the source of mitochondrial donor, mitochondrial isolation/purification, storage and preservation, rapid validation of mitochondrial integrity and functionality, targeted delivery, extracellular viability constraints, potential immune responses, and long-term safety/efficacy (Fig. 3C) [171]. 1) Among these, the source of donor mitochondria represents a critical biosafety bottleneck. Autologous mitochondria are preferred due to their low immunogenicity but pose substantial logistical challenges for rapid application in acute settings [184]. Moreover, autologous sources are often unsuitable for patients with congenital mitochondrial disorders or elderly individuals, as systemic mitochondrial defects are commonly present across tissues. Allogeneic and xenogenic transplantation circumvents these issues but entails risks of immune rejection and mtDNA–nDNA incompatibility [185]. 2) Mitochondrial isolation/purification carries inherent functional and biosafety risks, and this procedure can induce mitochondrial stress, OMM damage, and release of immunogenic DAMPs that include mtDNA, RNA, N-formyl peptides, and cytochrome C, triggering unintended immune responses [186,187]. Maintaining mitochondrial integrity is thus a critical prerequisite for safe transplantation. 3) Mitochondrial storage and preservation. Mitochondrial viability declines rapidly ex vivo, with isolated mitochondria remaining functional for only approximately 1 h on ice, severely restricting the therapeutic time window for clinical use [188]. Although cryopreservation represents a potential solution, conventional protocols (e.g., −65°C or −80°C with DMSO) compromise respiratory activity and long-term stability [[189], [190], [191]]. Additionally, Damaged or poorly preserved exogenous mitochondria can release mitochondrial DAMPs, which can activate innate immune response and may limit the therapeutic efficacy of mitochondria transplantation and even aggravate injury [184]. Developing optimized preservation strategies to maintain mitochondrial stability and activity is therefore critical for translating mitochondria transplantation into clinical practice. 4) Rapid validation of mitochondrial integrity and function is a critical prerequisite for safe clinical translation [192]. Current assessment methods, including fluorescent staining, respirometry, ATP/function assays, and purity checks (Western blot, electron microscopy), are all time-consuming and complex, incompatible with acute clinical demands [183,193]. 5) Precise targeted delivery of viable mitochondria to diseased tissues is a major translational bottleneck, particularly for the central nervous system (CNS), where the blood-brain barrier severely limits access. Current preclinical and clinical approaches rely on direct local injection, which is infeasible for most clinical scenarios, especially CNS disorders with limited surgical access [183,194]. 6) Extracellular viability constraints. Of note, isolated skeletal muscle mitochondria cannot survive in the ionic milieu of blood or extracellular space, since they undergo immediate, irreversible membrane depolarization and damage, driven by rapid Ca2+ overload mediated by the mitochondrial calcium uniporter [195]. 7) Potential immune responses. Mitochondria transplantation faces unresolved immunological safety concerns, with conflicting preclinical evidence. While multiple studies have demonstrated that autologous or allogeneic mitochondria transplantation is well-tolerated, with no significant elevation of pro-inflammatory cytokines or alloimmune responses across murine, porcine, and early clinical models [[196], [197], [198]], contradictory findings indicate that allogeneic [199], and exogenous mitochondria can act as DAMPs [200,201]. These can activate neutrophils and endothelial cells, triggering pro-inflammatory cascades and alloreactive rejection that may eliminate transplanted mitochondria. 8) long-term safety/efficacy. Exogenous mtDNA introduction raises concerns since it may trigger mitochondrial genetic drift, heteroplasmy, and elevated inflammatory/oxidative stress risks [202,203], though the magnitude of this genetic risk remains debated in the field [[204], [205], [206]], alongside associated ethical considerations for genetic material manipulation. Although preclinical investigations have not detected obvious short-term abnormalities, the long-term functional and genetic consequences remain largely unknown. Furthermore, xenogenic transplantation raises additional ethical considerations. Future progress requires advanced tools (e.g., single-cell sequencing for fate tracking) to accelerate precision applications in infection diseases. Within the complex pathogen-mitochondria interplay, multimodal combination strategies, such as nanoengineered mitochondria, may yield superior therapeutic outcomes.Risk/limitation box for mitochondria transplantationOff-target effects: Systemic administration of exogenous mitochondria may distribute to non-target tissues and be internalized by healthy cells [207]. This not only dilutes the local therapeutic efficacy but may also disrupt normal mitochondrial homeostasis in off-target cells. Specifically, these excess exogenous mitochondria can trigger aberrant ROS production, or release of DAMPs from damaged mitochondria, inducing unintended local inflammatory responses, abnormal cell activation, and even undesired cell proliferation in some contexts.Systemic consequences of altering mitophagy: The impact of mitochondria transplantation on mitophagy is highly context-dependent. In pathological contexts with inherent mitophagy dysregulation, such as the conditional DNM1L-knockout cerebellar degeneration model, where exogenous liver-derived mitochondria were found to normalize aberrantly elevated mitophagy, rescue motor function, and improve mitochondrial function [174], this intervention can restore disrupted mitophagy balance to exert therapeutic effects. However, it remains unclear how this intervention affects mitophagy in other contexts. Whether exogenous mitochondria perturb the normal basal mitophagy equilibrium in healthy bystander tissues remains unknown.Heterogeneity across cell types: The uptake, processing, and integration of exogenous mitochondria vary drastically across cell types. Cardiomyocytes rely on actin-mediated endocytosis [208,209], while hepatocytes and fibroblasts utilize macropinocytosis [210,211], leading to divergent uptake efficiencies [184]. Additionally, cells with distinct energy demands exhibit differential responses: High-energy cells (e.g., cardiomyocytes, neurons) show robust bioenergetic improvements from donor mitochondria, while low-energy cells display limited benefits [207,212]. Furthermore, donor mitochondrial sources also introduce heterogeneity. Mitochondria from high-energy tissues (muscle) confer stronger protective effects in inflammatory/infection models compared to those from stem cells, complicating the development of standardized therapeutic protocols [213,214].Challenge of translating cell-culture mitophagy phenotypes into in vivo infection outcomes: Most observations of mitophagy modulation by mitochondria transplantation are derived from isolated cell culture systems, which fail to recapitulate the complex in vivo infection microenvironment. In vitro studies often show that mitochondria transplantation can normalize mitophagy and suppress inflammatory responses in LPS-stimulated macrophages [213,215,216], but these phenotypes do not always translate to in vivo infection settings. For example, in polymicrobial sepsis models, only muscle-derived mitochondria improved bacterial clearance and survival, while the beneficial anti-inflammatory effects observed in vitro with other donor sources (liver cells and MSCs) were lost in vivo [213,214]. Furthermore, in vitro systems cannot model how pathogen-mediated hijacking of autophagy pathways may interact with transplant-altered mitophagy.CRediT authorship contribution statementJun Li: Conceptualization, Funding acquisition, Writing – original draft. Weijian Li: Data curation, Validation. Dan Wang: Conceptualization, Data curation. Yulan Liu: Funding acquisition, Supervision, Writing – review & editing. Risk/limitation box for mitochondria transplantation Off-target effects: Systemic administration of exogenous mitochondria may distribute to non-target tissues and be internalized by healthy cells [207]. This not only dilutes the local therapeutic efficacy but may also disrupt normal mitochondrial homeostasis in off-target cells. Specifically, these excess exogenous mitochondria can trigger aberrant ROS production, or release of DAMPs from damaged mitochondria, inducing unintended local inflammatory responses, abnormal cell activation, and even undesired cell proliferation in some contexts. Systemic consequences of altering mitophagy: The impact of mitochondria transplantation on mitophagy is highly context-dependent. In pathological contexts with inherent mitophagy dysregulation, such as the conditional DNM1L-knockout cerebellar degeneration model, where exogenous liver-derived mitochondria were found to normalize aberrantly elevated mitophagy, rescue motor function, and improve mitochondrial function [174], this intervention can restore disrupted mitophagy balance to exert therapeutic effects. However, it remains unclear how this intervention affects mitophagy in other contexts. Whether exogenous mitochondria perturb the normal basal mitophagy equilibrium in healthy bystander tissues remains unknown. Heterogeneity across cell types: The uptake, processing, and integration of exogenous mitochondria vary drastically across cell types. Cardiomyocytes rely on actin-mediated endocytosis [208,209], while hepatocytes and fibroblasts utilize macropinocytosis [210,211], leading to divergent uptake efficiencies [184]. Additionally, cells with distinct energy demands exhibit differential responses: High-energy cells (e.g., cardiomyocytes, neurons) show robust bioenergetic improvements from donor mitochondria, while low-energy cells display limited benefits [207,212]. Furthermore, donor mitochondrial sources also introduce heterogeneity. Mitochondria from high-energy tissues (muscle) confer stronger protective effects in inflammatory/infection models compared to those from stem cells, complicating the development of standardized therapeutic protocols [213,214]. Challenge of translating cell-culture mitophagy phenotypes into in vivo infection outcomes: Most observations of mitophagy modulation by mitochondria transplantation are derived from isolated cell culture systems, which fail to recapitulate the complex in vivo infection microenvironment. In vitro studies often show that mitochondria transplantation can normalize mitophagy and suppress inflammatory responses in LPS-stimulated macrophages [213,215,216], but these phenotypes do not always translate to in vivo infection settings. For example, in polymicrobial sepsis models, only muscle-derived mitochondria improved bacterial clearance and survival, while the beneficial anti-inflammatory effects observed in vitro with other donor sources (liver cells and MSCs) were lost in vivo [213,214]. Furthermore, in vitro systems cannot model how pathogen-mediated hijacking of autophagy pathways may interact with transplant-altered mitophagy. CRediT authorship contribution statement Jun Li: Conceptualization, Funding acquisition, Writing – original draft. Weijian Li: Data curation, Validation. Dan Wang: Conceptualization, Data curation. Yulan Liu: Funding acquisition, Supervision, Writing – review & editing. Nanoengineered mitochondria Apart from the three aforementioned strategies, we introduce nanoengineered mitochondria as a fourth distinct and independent therapeutic approach. This strategy is fundamentally distinct from the two mitochondria-related technologies previously discussed. In contrast to conventional mitochondria transplantation, which directly transplanting the unmodified healthy exogenous mitochondria, nanoengineered mitochondria represent biohybrid systems formed by integrating isolated mitochondria with synthetic nanomaterials or biomolecules to confer mitochondria new functionalities [[217], [218], [219], [220]], aiming to overcome the limitations of conventional mitochondria transplantation [221,222]. Unlike mitochondria-targeted nanomaterial delivery systems, which function primarily as drug carriers to deliver therapeutic agents to endogenous mitochondria within host cells, nanoengineered mitochondria involve direct engineering of the isolated mitochondria themselves, transforming the organelle into a standalone, functional therapeutic platform [217]. This key difference firmly establishes its status as an independent therapeutic modality. Pathogen infection generates a harsh local microenvironment featured by excessive ROS, hypoxia, and acidity [223]. Under such stressful conditions, conventionally transplanted mitochondria are prone to rapid damage and MMP collapse. Instead of exerting therapeutic functions, these damaged mitochondria release DAMPs that exacerbate inflammation and tissue injury [183,201,224]. In comparison, nanomaterial modification may act as a protective shield for isolated mitochondria. For instance, antioxidant coatings effectively scavenge excess local ROS and stabilize mitochondrial membrane structures [225], enabling nanoengineered mitochondria to sustain respiratory function for an extended period in hostile microenvironment, continuously supply energy to host cells, and restore metabolic homeostasis. This may disrupt pathogen hijacking of mitophagy and block immune evasion. Previous studies demonstrate that these tailored nanobiohybrid systems enable the isolated mitochondria to acquire stronger targeting ability, boost movement capability, enhance the metabolic activity, and reduce ROS [[226], [227], [228]]. Beyond structural protection, the nanomaterials or biomolecules for mitochondria engineering also enhance host immunity, alleviate oxidative stress, while suppressing excessive inflammatory response [229,230]. Currently, a variety of nanomaterials have been employed for mitochondrial engineering. On one hand, the nanoengineered mitochondria enable the achievement of systemic mitochondrial pool homeostasis via synchronized supplementation of healthy mitochondria and elimination of damaged ones. For instance, Wang et al. [231] developed mitophagy-enhanced nanoparticles (MENs) encapsulating Parkin mRNA, which anchored intact mitochondria via electrostatic interactions to form Mito-MEN. This system simultaneously boWosted the delivery of functional exogenous mitochondria and promoted mitophagy of damaged endogenous mitochondria. In a bleomycin-induced pulmonary fibrosis mouse model, Mito-MEN restored mitochondrial function and alleviated fibrotic phenotypes, providing a promising strategy for maintaining mitochondrial homeostasis and treating mitochondria-related diseases [231]. Moreover, the engineered mitochondria exhibit efficient antitumor effects [[232], [233], [234]], targeting potential for endothelial repair [235], potential in ocular mitochondrial disease therapy [236], and treating radiation-induced skin injury [237]. On the other hand, the nanomaterials also provide remotely controllable emerging strategies to regulate mitochondrial-associated metabolic networks for biotechnological applications. Guo et al. [238] constructed a Saccharomyces cerevisiae-indium phosphide nanoparticle biohybrid system, where photo-generated electrons from the light-activated nanomaterials achieve efficient intracellular NADPH regeneration via the mitochondria-associated redox network. This design bypasses the conventional carbon-consuming cofactor biosynthesis pathway, demonstrating the great potential of nanomaterials in breaking cellular biosynthetic bottlenecks through modulating mitochondrial metabolism. These studies suggest that nanoengineered mitochondria can achieve systemic mitochondrial pool homeostasis to treat a broad spectrum of mitochondria-related diseases, while nanomaterials can also remotely regulate mitochondrial-associated metabolic networks to break cellular biosynthetic bottlenecks for biotechnological applications. However, direct applications of engineered mitochondria in infection models are scarce to date. To enable such functional engineering for infection therapy, the development of biocompatible and tunable modification materials is essential. As an emerging supramolecular assembly material, metal-polyphenol networks (MPN) serve as an ideal platform for mitochondrial engineering. Through dynamic coordination between natural polyphenols and metal ions, MPN rapidly form ultrathin functional coatings on diverse biological surfaces under mild conditions without compromising intrinsic bioactivity, thus attracting extensive interest in biointerface engineering [219,239]. Of note, the MPN-derived nanoarmor was proven to shield probiotics from antibiotic damage, and effectively ameliorated antibiotic-associated diarrhea in gut infection models [240]. Recent advance further demonstrated that polyphenol-based supramolecular coatings could reshape the gut immuno-microbiome axis, boosting the therapeutic efficacy against infection-related pathological microenvironments [220]. Nevertheless, this strategy still faces several limitations, including challenges in large-scale standardized preparation, the requirement for long-term biosafety evaluation, and substantial barriers to clinical translation. Future efforts should prioritize the design of functionalized coatings that better adapt to the infectious microenvironment, thereby promoting the clinical translation of MPN-based mitochondrial engineering for infectious diseases. Outlook and conclusion In this review, we systematically summarized the mechanisms of mitophagy in the context of dynamic mitochondrial remodeling and the multifaceted roles of mitophagy in pathogen-mediated immune evasion, filling the gap between the well-documented tumor immune evasion and the under-explored infectious disease context. Apart from cataloging how diverse pathogens (intracellular bacteria, viruses, parasites) hijack mitophagy to promote their survival, we highlighted the dynamic nature of this regulation. For instance, HSV-1 induce mitophagy at early infection stages but suppress it at later phases. Notably, given the limited direct evidence regarding the role of mitophagy during pathogen infections, this review also reveals a positive correlation between fission and mitophagy based on the most included studies in Table 3, Table 4, Table 5, which is consistent with the conclusion that mitochondrial fission is a prerequisite for mitophagy in the field of cancer and neurological disorders [[9], [10], [11]]. We further proposed that integrating mitophagy analysis with dynamic mitochondrial behavioral profiling (e.g., fission/fusion) is critical to accurately interpret this highly regulated process, rather than assessing it as a static event. Despite these advances, two core mechanistic questions remain to be addressed: (1) The mechanisms of time-dependent mitophagy-mediated immune evasion during infection, and (2) the mechanistic connection between mitochondrial dynamics and mitophagy. Addressing these questions not only fills a critical gap in understanding how pathogens exploit mitophagy for immune evasion but also provides a theoretical foundation for developing targeted therapies. We also propose three priority directions to advance this field and tackle these open questions: 1) High-throughput single-cell organelle dynamics profiling. Future studies could leverage label-free high-throughput microscopy combined with machine learning to quantify pathogen-induced subcellular alterations (organelle morphology, positioning, dry mass) across hundreds of single cells, covering the full infection timeline. Follow-up Bayesian network modeling could reconstruct inter-organelle regulatory networks disrupted by infection, which would not only clarify mitophagy exploitation by pathogens but also accelerate drug screening and therapeutic target identification [108]. 2) Precise small-molecule mitophagy modulator development. To address the off-target effects of current compounds, future efforts could integrate molecular docking and surface plasmon resonance (SPR) techniques to discover novel, specific mitophagy-regulating compounds (Fig. 4), which could be further combined with mitochondrial-targeted nanotherapeutics to boost therapeutic efficacy. 3) Translational exploration of emerging mitochondrial therapies. Notably, mitochondria-targeted nanodelivery systems, mitochondria transplantation, and the newly emerged engineered mitochondrial therapy remain largely underexplored in infectious disease settings. These strategies hold great potential to overcome pathogen immune evasion, warranting in-depth mechanistic and preclinical investigation in the future.Fig. 4Artificial intelligence-enhanced label-free microscopic methods.A. This innovative approach combines dynamic label-free holotomographic microscopy (HTM) with artificial intelligence-powered analytical capabilities. Briefly, cell infected by pathogen are dynamically imaged by HTM, followed by machine learning to determine the cells and organelles in time-lapse recordings.B. The pixel-scale mitochondria are subsequently identified using a two-class pixel categorization, utilizing an extra-tree classifier trained to assign pixel categories based on their position in a derived feature space. This method enables reliable detection of these sparse organelles within the complex texture of HTM images.C. Larger objects like nuclei and whole cells are optimally segmented with an adapted U-NET34 fully convolutional network, and a sharpening of the outlines was performed by object propagation within the refractive index signal [241]D. The refractive index (RI) U2OS mitotracker is obtained by HTM based on the RI signal, with their respective fluorescent label signal made partially transparent by analyzing RI images through machine learning algorithms such that underlying cellular structures are visible. Conversely, mitotracker pred is visualized with epifluorescence microscopy generated, and the structures are thicker when compared with HTM.E. This approach is proved to be accessible as evidenced by the immunofluorescence signal, analysis by standard F1, intersection over union (IoU) scores, and the structural similarity index measure (SSIM).F. Bayesian networks can be employed to construct organelle cross-regulation, to reveal the dynamic regulatory relationship among subcellular organelles during the pathogens' infection process. LD, Lipid droplets. Reproduced and modified with permission [108]. Copyright © 2024 Nature Communications. Artificial intelligence-enhanced label-free microscopic methods. A. This innovative approach combines dynamic label-free holotomographic microscopy (HTM) with artificial intelligence-powered analytical capabilities. Briefly, cell infected by pathogen are dynamically imaged by HTM, followed by machine learning to determine the cells and organelles in time-lapse recordings. B. The pixel-scale mitochondria are subsequently identified using a two-class pixel categorization, utilizing an extra-tree classifier trained to assign pixel categories based on their position in a derived feature space. This method enables reliable detection of these sparse organelles within the complex texture of HTM images. C. Larger objects like nuclei and whole cells are optimally segmented with an adapted U-NET34 fully convolutional network, and a sharpening of the outlines was performed by object propagation within the refractive index signal [241] D. The refractive index (RI) U2OS mitotracker is obtained by HTM based on the RI signal, with their respective fluorescent label signal made partially transparent by analyzing RI images through machine learning algorithms such that underlying cellular structures are visible. Conversely, mitotracker pred is visualized with epifluorescence microscopy generated, and the structures are thicker when compared with HTM. E. This approach is proved to be accessible as evidenced by the immunofluorescence signal, analysis by standard F1, intersection over union (IoU) scores, and the structural similarity index measure (SSIM). F. Bayesian networks can be employed to construct organelle cross-regulation, to reveal the dynamic regulatory relationship among subcellular organelles during the pathogens' infection process. LD, Lipid droplets. Reproduced and modified with permission [108]. Copyright © 2024 Nature Communications. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.