Volume 64, Issue 5 p. 397-401
Critical Review
Free Access

Hijacking Mitochondria: Bacterial Toxins that Modulate Mitochondrial Function

Jhih-Hang Jiang

Jhih-Hang Jiang

Department of Biochemistry & Molecular Biology, Monash University, Clayton, Australia

Department of Microbiology and Immunology, The University of Melbourne, Parkville, Australia

Search for more papers by this author
Janette Tong

Janette Tong

Department of Biochemistry & Molecular Biology, Monash University, Clayton, Australia

Search for more papers by this author
Kipros Gabriel

Corresponding Author

Kipros Gabriel

Department of Biochemistry & Molecular Biology, Monash University, Clayton, Australia

Kipros Gabriel, Department of Biochemistry & Molecular Biology, Monash University, Clayton Campus, Melbourne 3800, Australia. Tel: +61 3 9902 9213. Fax: 61 3 9905 3726Search for more papers by this author
First published: 11 April 2012
Citations: 21


Bacterial infection has enormous global social and economic impacts stemming from effects on human health and agriculture. Although there are still many unanswered questions, decades of research has uncovered many of the pathogenic mechanisms at play. It is now clear that bacterial pathogens produce a plethora of proteins known as “toxins” and “effectors” that target a variety of physiological host processes during the course of infection. One of the targets of host targeted bacterial toxins and effectors are the mitochondria. The mitochondrial organelles are major players in many biological functions, including energy conversion to ATP and cell death pathways, which inherently makes them targets for bacterial proteins. We present a summary of the toxins targeted to mitochondria and for those that have been studied in finer detail, we also summarize what we know about the mechanisms of targeting and finally their action at the organelle. © 2012 IUBMB IUBMB Life, 2012


mitochondrial membrane potential (ΔΨ); EPEC, Enteropathogenic Escherichia coli; OM, Outer membrane; IM, Inner membrane; TOM, Translocase of the outer mitochondrial membrane; TIM, Translocase of the inner mitochondrial membrane.


As more information about bacterial mechanisms of pathogenesis is uncovered, one can only marvel at the diversity, complexity, and intricate nature of the mechanisms of pathogenesis that have evolved. Bacterial toxins and effector proteins offer bacterial pathogens an advantage over their host by modulating host cell function, killing host cells, or even preventing programmed cell death. This in turn might give a pathogen access to nutrients, create a more stable niche, and allow evasion of the immune system. For bacterial proteins to reach eukaryotic host cells, bacteria must first release them. Most bacterial pathogens usually engage large multisubunit complex secretion systems to do this. To date, seven different secretion systems have been discovered (types I-VII) (1–3) each being responsible for the secretion of different classes of bacterial proteins and or toxins. Once the toxins/effector proteins reach the host surface and gain access to the inside of the host cells, they must also traffic to the correct compartment to mediate their effects. This means that the bacterial protein must not only have the information to be efficiently secreted but also the information to be correctly trafficked within the host cell.

One critical organelle targeted by pathogens is the mitochondria. Over the last 20 years or so, a significant research effort has been spent on exposing the molecular mechanisms behind the mitochondrial targeting of pathogenic proteins or toxins, and understanding how the toxins subvert host physiological processes. Our review summarizes the known mitochondria targeted toxins and effectors and their effects.


Mitochondria play an important role in many aspects of biological function, such as energy conversion for ATP synthesis, ion homeostasis, fatty acid biosynthesis, iron sulfur cluster biogenesis, calcium storage, and importantly the cell death pathways.

To convert energy to ATP via oxidative phosphorylation, mitochondria oxidize carbohydrates, fats, and amino acids. This finally converges on the creation of a proton gradient across the inner membrane known as the mitochondrial electrochemical membrane potential (Δψ), which is the driving force of ATP generation by the F0F1 ATPase (4, 5). Importantly, mitochondria are also deeply entrenched in events of the programmed cell death pathways, apoptosis. As more toxins are being discovered, there is a trend toward proteins that affect this cell death aspect of mitochondrial biology. As our understanding of mitophagy and autophagy, processing that result in mitochondrial turnover, improves there is also likely to be discoveries of multiple toxins affecting these pathways.

When human cells experience a stimulus such as a severe cellular stress that usually leads to death, such as DNA damage and in some cases infection, the BH3-only proteins will translocate from the cytosol to the mitochondrial outer membrane to inhibit Bcl-2 from binding to Bax. Then, this would trigger a change in Bax conformation which forms a channel in outer mitochondrial membrane, leading to the release of cytochrome c from the intermembrane space (IMS) into the cytosol, which in turn results in the activation of the apoptosome by cytochrome c (6). Then, the apoptosome activates Caspase 3 to cleave substrates, which results in apoptosis reaching an irreversible phase (6, 7). Furthermore, the release of Smac/Diablo from mitochondria removes the inhibition of pro-caspase 8 by XIAP (8, 9). The matured Caspase-8 further activates more BH3-only proteins leading to a vicious cycle that accelerates DNA degradation, cell lysis, and death (10). Modulating one or more steps in the pathway, particularly before activation of the apoptosome and Caspase 3, can drastically change the fate of a host cell. Mitochondrial targeted toxins that can sensitize cells to cell death or inhibit cell death have been discovered (Table 1), however, the action of most toxins remains controversial. For example, PorB has been reported to act in both proapoptotic and antiapoptotic fashion by various groups (11–14). Differences in these observations could be due to protein dose, the strain of bacteria used, mammalian cells used for infection, and the mode of PorB delivery. This controversy is likely to continue until the exact mechanism of PorB function is elucidated. For the FimA toxin, an important virulence factor for Escherichia coli, Shigella species, and Salmonella typhimurium, it has been suggested that interaction of the toxin with a hexokinase-voltage dependent anion channel complex in the mitochondrial outer membrane is likely to be the underlying cause of cell death inhibition (15). However, for many other toxins, mostly those that are predicted to form channels, the exact mode of action remains mysterious and of intensive research focus.

Table 1. Bacterial toxins targeted to mitochondria
Pathogens Potentially pore forming Bacterial factor References
Cell death induction
Acinetobacter baumannii Y Omp38(AbOmpA) (16, 17)
Bacillus cereus Y Cereulide (18–21)
Clostridium difficile N Toxin A, B (22–24)
Clostridium sordellii N Lethal toxin (25, 26)
EPEC N/A EspF, Orf19(Map), Tir (27–31)
EPEC, EHEC N/A EspZ (31, 32)
ETEC N/A STb (33)
Helicobacter pylori Y VacA (34–39)
Mannheimia haemolytica Y Leukotoxin A (25, 40, 41)
Neisseria gonorrhoeae Y PorB (42–44)
Pseudomonas syringae N HopAA1-1 (45)
Salmonella N SipB (46)
Staphylococcus aureus Y Panton–Valentine leukocidin (47, 48)
Streptococcus pneumonia Y Pneumolysin (49)
Cell death inhibition
Anaplasma phagocytophilum N Ats-1 (50)
Neisseria meningitidis Y PorB (13, 43, 44, 51, 52)
Porphyromonas gingivalis N Gingipain adhesin peptide A44 (53, 54)
Escherichia coli K1 Salmonella Shigella N FimA (15)
Plant immunity
Pseudomonas syringae N HopG1 (55)
Coiella Burnetii N/A AnkJ (56)
EPEC N/A EspJ (57)
Legionella pneumophila N/A LegS2, LncP (58, 59)
Salmonella N SopA (60, 61)
  • EPEC: Enteropathogenic E. coli, EHEC: Enterohaemorrhagic E. coli, ETEC: Enterotoxigenic E. coli


Efficient delivery of bacterial effector proteins is central to bacterial pathogenesis. This not only requires transport across the bacterial membranes but also across one or more host lipid bilayers. After synthesis in the bacterial cytoplasm, usually as protoxin proteins with N-terminal signals that direct their movement out of the bacterial cell (62), the proteins are then transported across the inner membrane and additionally the outer membrane in the case of gram-negative bacteria. The mechanisms for achieving this are diverse with seven secretion systems characterized to date in gram-negative bacteria. In general, we can group the bacterial secretion systems into two groups, those that “inject” proteins into the host cytoplasm and those that secrete proteins into the extracellular milieu or deliver toxins to the host surface. However, it must be noted that both types usually function in tandem in most bacterial pathogens. The type III and type IV systems involve the assembly of large modular secretion apparatus spanning both bacterial membranes that can also cross host plasma membrane (63). This allows for bacterial proteins to be “injected” into the host cytoplasm and from there the toxins can target to mitochondria. The mechanism for delivery to mitochondria for proteins that are not “injected into the host cytosol is much more complex as it anticipated that most proteins in this class must be endocytosed, are trafficked within vesicles in the host and must at some point be liberated from trafficking vesicles or the endosome/lysosome system into the cytosol before relocating to the mitochondria (Fig. 1A). The mechanisms behind such trafficking pathways have so far remained largely elusive but examples of release from intracellular vesicles do exist, Toxin A and B from Clostridium difficile which glycosylate proteins in the cytoplasm after “self release” from the lysosome by pore formation after pH activation are such examples (64, 65). Toxin A and B have been reported to localize to mitochondria and increase the sensitivity or indeed induce host cell apoptosis (23, 24, 66). Other toxins, which do not target to mitochondria, have been reported to navigate through the secretory system after endocytosis, culminating in retrograde transport through the secretory (SEC) translocon into the cytosol at the endoplasmic reticulum membrane. This includes the Shiga and Cholera toxins (67, 68). Whether toxins that are targeted to mitochondria use such a system remains to be elucidated.

Details are in the caption following the image

Trafficking of bacterial toxins to mitochondria. A. Bacterial proteins are usually either released and taken up by host cells (1) or “injected” into host cells (2). Injected proteins can then directly target mitochondria. Endocytosed toxins need to be liberated from membranous vesicles before being targeted to mitochondria. This process is not well understood. B. Mitochondria have two membranes (OM and IM) which enclose the IMS and matrix respectively. Protein import machineries exist in all compartments and internally targeted toxins and some toxins targeted to the outer membrane would need to “hijack” these machineries for entry. Arrows depict potential targeting pathways of individual mitochondrial translocases. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

The structure of mitochondria also presents another set of hurdles for a bacterial toxin destined for the organelle. To carry out their diverse functions mitochondria have multiple compartments that are formed by the inner and outer membranes (IM and OM) that separate the matrix and IMS, respectively. This allows critical cellular reactions to be separated and allows for the generation and maintenance of the electrochemical potential across the inner mitochondrial membrane. At the same time, this obviously brings another level of complexity to pathogenesis, as a bacterial protein might not only need to be targeted to mitochondria but also to a specific compartment within the organelle to target a specific function. This means that a toxin must, in addition to signals for export from bacteria and entry into the host, also contain all of the information for targeting to a specific mitochondrial compartment. Bacterial toxins have been shown to be targeted to the matrix, inner membrane, and outer membrane. The major class of toxins and effectors targeted to mitochondria are predicted to form pores, and these are the best studied to date (Table 1) with various effects on apoptosis reported (summarized in Table 1).


To carry out their many functions, mitochondria, import nearly all their resident proteins after they are synthesized in the cytosol. Mitochondrial protein translocation is a highly selective and very efficient process (69). Specific translocases exist in every compartment to courier proteins to their place of function where they are then folded. Likewise, bacterial proteins that must enter mitochondria must in some way “trick” the protein translocation system to think that they are native mitochondrial proteins to gain entry in the organelle and be transported and folded in the correct mitochondrial compartment. All proteins destined for mitochondria are first delivered to the translocase of the outer mitochondrial membrane (TOM), also known as general import pore, before they are transferred to subsequent machinery depending on their final destination (69). Proteins destined for the mitochondrial matrix are passed onto the presequence Translocase of the inner mitochondrial membrane (TIM23 complex) and with the assistance of the Presequence translocase-Associated Motor (PAM) complex, of which mitochondrial Hsp70 is a key component, proteins are translocated across the inner membrane (70). The recently characterized EspZ from Enteropathogenic E. coli (EPEC) and Enterohaemorrhagic E. coli (EHEC) was shown to interact with the TIM23 complex, specifically the Tim17 subunit causing toxicity (31). Polytopic membrane proteins are transferred to the inner or outer membrane by the small TIM proteins of the intermembrane space. At the outer membrane polytopic β-barrel proteins, such as the PorB toxin, and some single α-helical transmembrane domain proteins are integrated and folded by the sorting and assembly machinery (SAM) (71–73). PorB follows the normal import pathway for β-barrel proteins, using the TOM complex for entry into the intermembrane space and the SAM for membrane integration. Interestingly, it requires only the core subunit of the SAM, Sam50, which is conserved from bacteria to man for its import and assembly. At the inner membrane, the polytopic α-helical “carrier” class of proteins are inserted and folded by the TIM22 complex (74) (Fig. 1B). It has very recently been shown that Legionella pneuomophila secrete a “carrier-like” protein, which is targeted to mitochondria using the TOM and TIM22 complexes, which is thought to not only hijack the import pathway of a native mitochondrial carrier protein but also mimic the function of a carrier protein at the inner mitochondrial membrane (59).

Hijacking these machineries in part has been made easier for modern day bacteria by the ancestral bacterial origin of mitochondria. It has been noted that some signals at the N-termini of bacterial proteins, which usually target proteins out of bacteria share characteristics with mitochondrial-targeting signals (29). Other signals for insertion of proteins into the bacterial outer membrane, such as those of β-barrel proteins, might also play a role when these proteins are targeted to mitochondria. It has already been shown that at least the N-terminal regions of many bacterial effectors are responsible for their mitochondrial entry. The amino-terminal region of bacterial effectors, such as Orf19 and EspJ (Enteropathogenic E. coli), Ats-1 (Anaplasma phagocytophilum), and leukotoxin A (Mannheimia haemolytica) have been identified as putative mitochondrial targeting signals (29, 50, 57). Signals that are not at the N-terminus are likely to exist on proteins such as PorB from Neisseria gonorrhoeae and VacA from Helicobacter pylori (11, 36, 43). It has been shown experimentally that deletion of the N and C termini of both of the VacA subunits (p33 and p55) does not impact on their mitochondrial importability (36) and furthermore green fluorescent protein (GFP) tagging at the N-terminus of VacA p33 does not inhibit p33 mitochondrial localization suggesting that an N-terminal targeting signal may not be responsible for its translocation (35). It has also been shown experimentally that the general import pore, the TOM complex, is required for the import of several bacterial effectors into mitochondria. This includes Orf19 and PorB (29, 76). Orf19 is further transported to the matrix with the help of Hsp70 which is the central component of the PAM (27).

Further studies are required to characterize the molecular mechanisms behind the import of bacterial effectors into mitochondria as it is essential for understanding the interactions between hosts and pathogens. Furthermore, as mitochondria are derived from a α-proteobacterial ancestors, understanding the molecular mechanisms will provide the link between evolutionary conserved process and pathogenesis. For example, the core of SAM complex in mitochondria, Sam50, is the homolog to BamA that is essential for bacterial protein assembly in outer membrane in gram-negative bacteria. Both the BAM and the SAM can import and fold the Neisserial PorB toxin.


Although we have not yet elucidated the function of most bacterial effectors targeted to mitochondria, there has been some work done on elucidating their suborganellar localization, which hopefully will provide insights into role of these effectors in pathogenesis. Studies so far have indicated that the mechanisms at play are very complex and diverse. It will be difficult to dissect the events that control cell fate for many toxins, particularly those that on initial examination seem to have multiple effects on cell biology. In addition, despite the considerable interest in studying these bacterial effector proteins, there is few high-resolution structures published over the years. Determination of structures or motifs in bacterial effectors will also provide clues to their biochemical functions in host targets. Advances in sequencing technologies have also greatly improved our understanding of bacterial pathogenic proteins, as bioinformatics approaches have become key in identifying differences between nonpathogenic and pathogenic strains and the proteins that they produce.