I'd like to block a percentage of PGC-1β or Sod2 expression. According to the following paper's figure beta-blockers inhibit some expression of PGC-1α. Are there any medicines/chemicals which safely and temporarily block 50%-60% expression of PGC-1β or Sod2 in fully developed mice or humans? Thank you
Fig. 1. Interactions of β-blockers and mitochondrial biogenesis. Aerobic exercise activates β2-adrenergic receptors (β-AR) on skeletal muscle and induces peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α) transcription, a regulator of mitochondrial biogenesis. Selective and nonselective β-blockers can blunt β2-AR signaling, which restricts the expected PGC-1α response following exercise and can impair adaptations to mitochondria and aerobic capacity. V̇O2max, maximum oxygen consumption.
Dexamethasone can be used to regulate PGC-1β http://www.ncbi.nlm.nih.gov/pubmed/20647557
Nonlinear transcriptomic response to dietary fat intake in the small intestine of C57BL/6J mice
A high caloric diet, in conjunction with low levels of physical activity, promotes obesity. Many studies are available regarding the relation between dietary saturated fats and the etiology of obesity, but most focus on liver, muscle and white adipose tissue. Furthermore, the majority of transcriptomic studies seek to identify linear effects of an external stimulus on gene expression, although such an assumption does not necessarily hold. Our work assesses the dose-dependent effects of dietary fat intake on differential gene expression in the proximal, middle and distal sections of the small intestine in C57BL/6J mice. Gene expression is analyzed in terms of either linear or nonlinear responses to fat intake.
The highest number of differentially expressed genes was observed in the middle section. In all intestine sections, most of the identified processes exhibited a linear response to increasing fat intake. The relative importance of logarithmic and exponential responses was higher in the proximal and distal sections, respectively. Functional enrichment analysis highlighted a constantly linear regulation of acute-phase response along the whole small intestine, with up-regulation of Serpina1b. The study of gene expression showed that exponential down-regulation of cholesterol transport in the middle section is coupled with logarithmic up-regulation of cholesterol homeostasis. A shift from linear to exponential response was observed in genes involved in the negative regulation of caspase activity, from middle to distal section (e.g., Birc5, up-regulated).
The transcriptomic signature associated with inflammatory processes preserved a linear response in the whole small intestine (e.g., up-regulation of Serpina1b). Processes related to cholesterol homeostasis were particularly active in the middle small intestine and only the highest fat intake down-regulated cholesterol transport and efflux (with a key role played by the down-regulation of ATP binding cassette transporters). Characterization of nonlinear patterns of gene expression triggered by different levels of dietary fat is an absolute novelty in intestinal studies. This approach helps identifying which processes are overloaded (i.e., positive, logarithmic regulation) or arrested (i.e., negative, exponential regulation) in response to excessive fat intake, and can shed light on the relationships linking lipid intake to obesity and its associated molecular disturbances.
Get full journal access for 1 year
All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.
Get time limited or full article access on ReadCube.
All prices are NET prices.
Fission Versus Fusion in Cancer
Re-modeling of the mitochondrial network in cells is mechanically regulated by key dynamin-related fission and fusion gene products and takes place in response to hypoxia, cell cycle cues, changing energy demands, and other cellular stresses (1, 2, 39, 40, 44).
The Mechanics of Fission and Fusion
Mitochondrial fusion is promoted by homotypic/heterotypic interactions of the Mitofusin 1 and Mitofusin 2 dynamin-related GTPases at the outer mitochondrial membrane (OMM) of adjacent mitochondria and by Opa-1, also a dynamin-related GTPase, at the inner mitochondrial membrane (IMM) (1, 39, 40, 45). Inhibition or loss of any one of these proteins impedes mitochondrial fusion leading to increased mitochondrial fragmentation and is associated with clinical neuropathy in Charcot–Marie–Tooth disease and Autosomal Dominant Optic Atrophy, highlighting the critical role played by mitochondrial fusion in cellular homeostasis, particularly in the nervous system (39, 42).
Mitochondrial fission requires the recruitment of a different dynamin-related GTPase, Drp1 to the OMM where it forms ring-like oligomers that pinch off mitochondria into smaller fragmented mitochondria (39). Fission is important ahead of mitosis to ensure even distribution of mitochondria to daughter cells (47) but also occurs when cells undergo mitophagy or apoptosis (2, 40, 44, 50, 51). Recruitment of cytosolic Drp1 to the mitochondria during fission is a regulated process involving post-translational modification of Drp1 (49, 52) and its interaction with putative receptors at the OMM such as Mitochondrial Fission Factor (Mff), Fis1, MiD49, MiD51, and possibly other proteins with which Drp1 interacts (45, 55). Recent work has also highlighted a role for the endoplasmic reticulum (ER) that is intimately associated with mitochondria, in determining the sites at which fission will occur (60). The constriction of mitochondria at points of contact with the ER are set up prior to recruitment of Drp1 to mitochondria and independent of Mff. Intriguingly, the mitochondrial fusion protein Mfn2 also plays a role in tethering mitochondria to the ER, in a manner required for proper calcium uptake by the mitochondria from the ER (61). Screens in yeast have identified additional putative molecular regulators of mitochondrial tethering to the cell cortex and the ER in ways that regulate mitochondrial positioning in cells and inheritance by daughter cells (62) but whether similar mechanisms operate in mammalian cells is unclear.
Stress-Induced Changes in Rates of Fission and Fusion
Beyond the actual mechanics of mitochondrial fission and fusion, the signaling pathways that regulate these processes are only just coming to the fore as mechanisms that may be disrupted in cancer. These signaling pathways respond to specific stresses and serve to coordinate mitochondrial dynamics with other aspects of cellular physiology. Drugs that inhibit protein synthesis, including mTOR inhibitors, as well as other stresses such as ultra-violet (UV) light have been shown to promote so-called “stress-induced mitochondrial hyperfusion” that relies upon canonical fusion proteins (Opa-1, Mfn1) (46), although, how these stress signaling pathways activate the fusion machinery is not clear. Stress-induced hyperfusion of this kind promotes ATP production through more efficient oxidative phosphorylation (OXPHOS) (63), and also inhibits mitophagy and prevents apoptosis (46, 64, 65).
The functional consequences of altered rates of fission or fusion for cellular metabolism, cell cycle kinetics and cell viability are a “work in progress” relying on numerous systems and technical approaches. In Drosophila, Yorkie-mediated up-regulation of Opa-1 and mitochondrial fusion was required for Yorkie/YAP-dependent proliferation and tumorigenesis (66). In mammalian systems, glucose deprivation and use of galactose as an alternative source of carbon, resulted in cells switching from glycolysis to OXPHOS, as expected, but significantly caused a marked increase in mitochondrial fusion and an accompanying increase in respiratory chain protein expression and cristae density, without any increase in mitochondrial mass (67).
Consistent with a critical role for mitochondrial fusion in regulating metabolism, inactivation of Mitofusin 1, Mitofusin 2, or Opa-1 inhibits oxidative metabolism and cell growth (68, 69). If fusion promotes mitochondrial ATP production (63), it is reasonable to postulate that this is achieved by increasing the efficiency of the electron transport chain (ETC) that is key to OXPHOS. This may be achieved by increasing the numbers/density of ETC complexes through increased expression of key components, as has been reported (67), or by altering the composition of specific respiratory chain components into different supercomplexes to maximize utilization of specific substrates, such as NADH in the presence of glucose versus FAD in the presence of fatty acids (70). Recent data has shown that assembly of respiratory chain complexes (RCC) into supercomplexes and increased respiratory efficiency is promoted by tighter cristae formation that is dependent on the fusion protein, Opa-1 (71). Conversely, disruption of cristae formation by knockout of Opa-1 in mice promoted mitochondrial fragmentation, cristae dissolution, and reduced RCC formation and respiration (71). Thus fusion may promote respiration through Opa-1 dependent effects on cristae density and formation of respiratory chain supercomplexes. Additionally, fusion may induce mitochondrial membrane potential changes that promote uptake of pyruvate or other substrates that fuel OXPHOS. Clearly, a more continuous mitochondrial lumen achieved by increased fusion is likely to promote more rapid diffusion of ADP, NADH, FADH2, and other matrix metabolites required to drive more efficient OXPHOS. In this way, fusion would also likely promote increased carbon flux through the Krebs’ cycle, more efficient rates of fatty acid oxidation and possibly increased activity of other metabolic pathways located at the mitochondria.
It has been suggested that fusion may promote respiratory efficiency by promoting complementation of mtDNA mutations (51). While this may be the case in some instances, the failure to detect homoplasmic mutations in mtDNA encoded subunits of CO, ND, ATP synthase, or cytochrome b in primary cells, or indeed more widely in tumor cells suggests that complementation of mtDNA mutations is not the key role of mitochondrial fusion in metabolism. Indeed, increased fusion has been shown to promote OXPHOS in very short time frames (67), indicating that the effects of fusion on oxidative metabolism are post-translational and not primarily dependent on gene complementation between mitochondrial genomes.
The effects of mitochondrial fusion in promoting oxidative metabolism at the mitochondria imply a contrasting role for mitochondrial fission in inhibiting oxidative metabolism, perhaps by decreasing substrate uptake, disrupting cristae, and respiratory complex formation and/or limiting diffusion of reducing equivalents. Oxygen is the major electron acceptor from complex IV of the respiratory chain, and thus it benefits the cell to decrease OXPHOS under limiting oxygen conditions (hypoxia) both to maximize efficient use of the smaller amounts of oxygen available but also to prevent generation of damaging ROS. Hypoxia limits OXPHOS in a number of ways but promoting mitochondrial fission may be one of the key mechanisms. Hypoxia promotes mitochondrial fission by modulating Drp1 activity and interaction with Fis1 (72). Hypoxia-induced expression of Siah2, an E3 ubiquitin ligase (73) targets the mitochondrial scaffold protein, anchoring protein 121 (AKAP121) for degradation (72) thereby preventing protein kinase A (PKA) dependent inhibition of Drp1. These observations, amongst others, are consistent with an inhibitory role for mitochondrial fission in OXPHOS. Increased fission linked to deregulated expression of Drp1 (increased) and Mfn2 (decreased) has been observed in tumor cells (74) but to what extent this contributes to the Warburg effect or other aspects of tumor growth remains to be determined.
Coordination of Rates of Fusion/Fission with Cell Cycle
Several reports indicate that increased mitochondrial fusion is required not only for efficient oxidative metabolism (68, 69) but is necessary for proliferation and entry into S-phase (63). In this latter study, mitochondrial membrane polarization and hyperfusion of mitochondria occurring at the G1/S transition was required for cyclin E (CCNE) expression and S-phase entry (63). Additional studies have also shown that mitochondrial bioenergetics are linked to cell cycle progression (75). However, artificially inducing hyperfusion, either through treatment with mdivi (a drug that inhibits Drp1) (63) or expression of dominant negative Drp1 (76), resulted in untimely induction of hyperfusion and sustained cyclin E over-expression at inappropriate phases of cell cycle, such as G2/M (76). This in turn was accompanied by replication stress, DNA damage, centrosomal amplification, and chromosomal instability (63, 76), all known features of cells over-expressing cyclin E (77). Interestingly, while cyclin E over-expression was required for proliferation and genome instability arising from hyperfusion or dysfunctional fission (63, 76), the specific factors produced by hyperfusion that resulted in cyclin E up-regulation have not been identified. Increased ROS has been previously reported to modulate levels of key cell cycle proteins, such as Emi-1 (78) but neither increased mitochondrial ROS nor altered ATP production (76) explain increased proliferation induced by mitochondrial hyperfusion (although this arguably requires further validation), leaving us with the unanswered question of what drives cyclin E expression in these circumstances.
Glycolysis is also cell cycle regulated and increases at the G1/S-phase boundary due in part to stabilization of 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase, isoform 3 (Pfkfb3) that is normally turned over by APC/Cdh1 in early stages of G1 phase of cell cycle (79). Interestingly, Drp1 is also turned over by APC/Cdh1 in G1 phase of cell cycle in a CDK-dependent manner (47, 49) suggesting a possible mechanism by which up-regulation of glycolysis may be coordinated with mitochondrial dynamics. Evidence that glutaminolysis is cell cycle regulated is sparse but since c-Myc and RB/E2F both modulate glutamine uptake (80), it would not be surprising if glutaminolysis, like glycolysis, was up-regulated as cells enter S-phase.
Dual Role of Bcl-2 Family Members in Mitochondrial Dynamics and Apoptosis
In addition to effects on cellular metabolism and cell cycle, mitochondrial fusion can delay cytochrome c release and apoptosis (46, 84), with Opa-1 oligomerization inhibiting pro-apoptotic cristae remodeling independent of its role in fusion (85) (Figure 2). Disruption of Opa-1 oligomers in a Bax/Bak-dependent manner by the pro-apoptotic BID protein was independent of mitochondrial membrane permeabilization (86) but was associated with pro-apoptotic cristae remodeling (a requirement for cytochrome c release) and Opa-1 deficient cells are more susceptible to apoptosis (87, 88). Interestingly, prohibitins promote cellular proliferation and resistance to apoptosis by inhibiting OPA-1 cleavage at the IMM thereby promoting fusion and normal cristae morphology (89).
Figure 2. Dual and apparently opposing roles of Bcl-2 family members and fission/fusion proteins in apoptosis and mitochondrial dynamics. Mitochondrial fission and fusion proteins appear to modulate apoptosis through activities that are distinct from their roles in mitochondrial dynamics but which involve members of the Bcl-2 family. Conversely, Bcl-2 family members modulate mitochondrial fission and fusion in a manner that appears to be independent of their functions in apoptosis.
In contrast to mitochondrial fusion, mitochondrial fission promotes mitochondrial membrane depolarization, cytochrome c release and apoptosis (90), with Drp1 promoting Bax oligomerization through mechanisms independent of its GTPase activity (91), possibly explaining how fragmented mitochondria are more amenable to Bax/Bak channel formation. Thus both mitochondrial fission and fusion proteins appear to modulate apoptosis through activities that are distinct from their roles in mitochondrial dynamics but which involve members of the Bcl-2 family (Figure 2).
Members of the Bcl-2 superfamily of cell death regulators are extensively characterized for their key role in regulating apoptosis (2, 44) but they also have an emerging role in mitochondrial dynamics (44, 92, 93). Bak and Bax are essential for apoptosis, such that Bax/Bak double knockout cells are resistant to apoptosis (94, 95). When Bak and Bax are activated by pro-apoptotic signals, they undergo oligomerization to form a channel in the OMM through which cytochrome c is released resulting in the formation of the apoptosome and activation of caspases (96). Apoptosis-resistant Bax/Bak null cells exhibit extensive mitochondrial fragmentation that is rescued by over-expression of Bak in the absence of apoptotic signaling suggesting that Bak and Bax can promote mitochondrial fusion (92). Indeed, the soluble form of Bax interacts directly with Mfn2 to promote its GTPase activity while both Bak and Bax interact with Mfn1 (90, 92, 97). Conversely, the anti-apoptotic Bcl-XL has been shown to promote mitochondrial fission in neurons through interactions with Drp1 that promote its GTPase activity (98). More recently, Mcl-1 has been implicated in modulating mitochondrial dynamics through an amino-terminal truncated form that localizes to the mitochondrial matrix, in contrast to full-length Mcl-1 at the OMM (99). Truncated Mcl-1 in the matrix is required for mitochondrial fusion and assembly of the F0F1-ATP synthase and for efficient respiration (99). This novel function for Mcl-1, distinct from its anti-apoptotic function may explain the observed heart failure in Mcl-1 null mice, in which cardiomyocytes exhibited aberrant mitochondrial structure and defects in respiration that were not rescued by Bax/Bak deletion (100), although the defects were partially rescued by deletion of cyclophilin D, a key regulator of the mitochondrial permeability transition pore (101). Thus there is regulation of the apoptotic activity of Bcl-2 related proteins by fusion/fission proteins and conversely regulation of fission/fusion protein activity by Bcl-2 related proteins. What is not clear is whether the activities of Bcl-2 proteins, and Bax/Bak in particular, in apoptosis and fission/fusion are exclusive. The increase in mitochondrial fragmentation taking place during apoptosis suggests that the pro-fusion activity of Bak/Bax is suppressed by their pro-apoptotic functions but formal experimental evidence supporting this is lacking. Similarly, it is not known whether reduced OXPHOS resulting from increased fission may contribute to cellular susceptibility to apoptosis. Finally, it is not clear whether altered mitochondrial dynamics contributes to the oncogenic activity of key Bcl-2 family members, such as Bcl-2 or Bcl-XL that are both over-expressed in certain human malignancies (102). For example, increased Bcl-XL expression in tumors may promote mitochondrial fission, as seen in neurons (98), that in turn would be predicted to limit mitochondrial OXPHOS thereby promoting the Warburg effect.
DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS
Anthracyclines, such as doxorubicin, are can be used as highly effective anticancer therapies. However, these compounds are known to cause damage to cardiomyocytes in a dose-dependent manner. Because of this patient's receiving anthracycline therapy, such as treatment with doxorubicin, have to be carefully monitored to detect signs of heart damage and/or administered secondary agents to help limit cardiac damage. This dangerous side effect of anthracycline-therapy has been the major limitation of its clinical usage and no method was previously known for predicting whether a patient would display cardiotoxicity in response to therapy or how serious cardiotoxicity would be.
Studies detailed herein demonstrate that cardiomyocyte-specific deletion of Top2b expression was able to protect cardiomyocytes from doxorubicin-induced DNA double strand breaks. Cardiomyocytes that lacked Top2b expression also showed changes in their transcriptome that are responsible for defective mitochondrial biogenesis and ROS formation. Furthermore, deletion of Top2b protected mice from development of doxorubicin-induced progressive heart failure, suggesting that Top2b is the molecular basis of doxorubicin-induced cardiotoxicity. In view of the results studies were undertaken to determine if Top2b expression would be a useful biomarker for predicting cardiotoxicity. Since detection of Top2b in patient's heart would be difficult and invasive, ELISA was used to determine the peripheral blood expression level of Top2b as a surrogate for Top2b expression in the heart. Studies shown in FIG. 8 demonstrate that patients with low Top2b level in the peripheral blood are more resistant to doxorubicin-induced cardiotoxicity, whereas patients with high Top2b level are much more sensitive to doxorubicin. Thus, Top2b expression level (such the level of Top2b protein measured in the peripheral blood) can be used as a genetic marker to predict susceptibility to doxorubicin-induced cardiotoxicity.
The new methods provided here could vastly improve therapeutic use of anthracyclines compounds such as doxorubicin. Before treatment, patients can be tested to determine their Top2b expression level. If Top2b level is low, then patients can safely receive a high dose of therapeutic, such as up to 450 mg/m 2 of doxorubicin even without regular monitoring for cardiac damage. On the other hand, if Top2b expression level is high, then patients should receive a lower initial dose of drug, intensive monitoring for cardiac damage and cardiac protective medication as they receive increased amounts of anthracycline-therapy.
Posttranslational Modification of PGC-1α
Numerous posttranslational modifications impact PGC-1α. 9, 15-25 Among them, phosphorylation, acetylation, methylation and ubiquitination are predominantly responsible for the transcriptional activity and stability of PGC-1α (Fig. 2).
Phosphorylation of PGC-1α
The adenosine monophosphate (AMP)-activated protein kinase (AMPK), p38 mitogen-activated protein kinase (MAPK), Akt, S6 kinase and glycogen synthase kinase 3β (GSK3β) are the best-characterized protein kinases targeting PGC-1α for phosphorylation. 16, 26
Phosphorylation can promote the activation of PGC-1α. For example, AMPK binds to and activates PGC-1α in skeletal muscle by direct phosphorylation on Thr177 and Ser538. This increases the transcriptional activity of PGC-1α and is required for AMPK-induced gene expression of glucose transporter 4 (GLUT4), mitochondrial genes, as well as PGC-1α itself. 20 The p38 MAPK phosphorylates PGC-1α on Thr 262, Ser265 and Thr298 resulting in increased protein stability. 16
On the other hand, like a two-edged sword, phosphorylation can also decrease PGC-1α activity as well. The C-terminus of PGC-1α contains multiple potential sites of Akt phosphorylation (RXRXXS/T, where X is any amino acid). Akt suppresses both gluconeogenesis and fatty acid oxidation (FAO) through its phosphorylation and inhibition of PGC-1α. 19 Similarly, when induced by nutrient signals, the gluconeogenic genes program is also inhibited through the phosphorylation of PGC-1α at Ser568 and Ser572 by serine/threonine kinase S6K1. However, these phosphorylations do not affect the stability of PGC-1α or other gene programs mediated by PGC-1α, such as mitochondrial electron transport and FAO. 15 Because all the sites phosphorylated by Akt or S6K1 are located within the RS region (amino acids 551–635) speculating that phosphorylation of the RS domain might attenuate the transcriptional activity of PGC-1α is reasonable.
Unlike the full-length-PGC-1α, NT-PGC-1α is likely to be exported to the cytoplasm by interacting with the nuclear protein, chromosomal maintenance 1 (also known as exportin 1, CRM1), through its LxxLL domain (nuclear export sequence). This process can be blocked by protein kinase A-dependent phosphorylation of NT-PGC-1α on Ser194, Ser241 and Thr256. The phosphorylation of NT-PGC-1α by PKA not only inhibits CRM1-mediated export from the nucleus but also affects its transcriptional potential. 11
Acetylation modification of PGC-1α
PGC-1α is controlled by acetylation to modulate its ability to recruit chromatin-remodeling complex and initiate gene transcription. PGC-1α is dynamically acetylated and inhibited by GCN5 (general control nonrepressed protein 5) when deacetylated and activated by SIRT1 (silent information regulator 2 homolog 1).
Although PGC-1α can interact with several acetyltransferases including GCN5, p300, SRC-1 and SRC-3, only GCN5 is characterized in its acetylation and inhibition of PGC-1α. 17 GCN5 is the first discovered histone lysine acetyltransferase, linking histone acetylation to transcriptional activation. 27 Besides histone proteins, GCN5 can also acetylate nonhistone proteins, such as PGC-1α, which harbors multiple acetylation sites spanning the whole sequence of the protein. 16 The acetylation status of PGC-1α could be modulated by GCN5 depending on the amounts of nuclear acetyl-CoA. In mammals, acetyl-CoA can be generated from TCA (tricarboxylic acid cycle)-derived citrate by ACL (ATP citrate lyase). Considering that GCN5, like all other acetyltransferases, requires acetyl-CoA as a substrate for the acetylation reaction, ACL might connect energy balance to GCN5 and affect the acetylation status of PGC-1α by controlling the nuclear production of acetyl-CoA. 28
SRC-3 facilitates PGC-1α acetylation in muscle and BAT by enhancing the expression of GCN5. 18 Under conditions of caloric restriction, decreased expression of SRC-3 and GCN5 leads to reduced acetylation levels, but enhanced PGC-1α activity, which promotes the use of fatty acids stored in adipose tissues as a source of energy. 18 Although SRC-1 can also interact with PGC-1α, whether it affects GCN5 activity in a manner similar to SRC-3 remains to be further investigated.
To date, SIRT1 and histone deacetylase 3 (HDAC3) are the only identified deacetylases for PGC-1α. SIRT1 is a member of the Sir2 family of nicotinamide adenine dinucleotide (NAD + )-dependent histone deacetylases that targets histone H4 at Lys16 and histone H3 at Lys9. As the principal NAD + -dependent deacetylase in mammalian cells, SIRT1 also deacetylates a wide set of nonhistone proteins, such as p53, p300, FOXO-1, 3a,4, Ku70 and PGC-1α. 29 Because SIRT1 requires the coenzyme NAD + as a substrate for its function, when in high energy need, PGC-1α is activated by SIRT1-mediated deacetylation presumably in response to changes in the amounts of NAD + or NADH or the NAD + /NADH ratio in cells. In turn, PGC-1α promotes the induction of fatty acid utilization and mitochondrial metabolism. 23 In addition, AMPK can perceive the energy stress and modulate SIRT1 activity by altering the intracellular level of NAD +. 30 Therefore, AMPK increases PGC-1α activity by direct phosphorylation and also by inducing SIRT1-mediated PGC-1α deacetylation.
Most recently, Emmett et al. reported that HDAC3 can also deacetylate PGC-1α and reverse GCN5-mediated repression of PGC-1α coactivator activity. In BAT, HDAC3 acts as a coactivator of ERRα. Deacetylation of PGC-1α mediates the coactivation of ERRα induced by HDAC3 and is required for the transcription of oxidative phosphorylation (OXPHOS) genes, which prepares BAT for acute thermogenic challenge. 21
All these findings suggest that reversible acetylation of PGC-1α connects cell energy status to transcriptional regulatory circuits mediating mitochondrial function and guarantees metabolic flexibility.
Methylation of PGC-1α
Another posttranslational modification contributing to metabolic flexibility is reversible methylation of PGC-1α. Arginine methylation has been implicated in the regulation of multiple cellular processes, including subcellular localization of proteins, protein–protein interactions and transcriptional activation. PRMT1 is the major protein arginine methyltransferase (PRMT) family member in mammalian cells. It serves as a coactivator for NRs and other DNA-binding transcription factors, such as p53 and initiator element binding factor (YY1) and mediates chromatin remodeling and initiation of transcription. 31, 32 PGC-1α has been confirmed to be methylated by PRMT1 at arginines 665, 667 and 669. These sites reside in a RERQR sequence within the C-terminal Glu-rich E region of PGC-1α. The methylation of PGC-1α by PRMT1 strongly enhances the transcription of target genes important for the biogenesis and function of mitochondria. 24 Theoretically, the E region of PGC-1α contains a nuclear localization signal and arginine methylation of PGC-1α in this region might regulate its nuclear export function.
SET7/9 is a mono-methyltransferase that methylates histone 3 Lys4 (H3K4), which is predominately found in active chromatin. Although the major role of SET7/9 is believed to regulate gene activation through histone methylation and alteration of chromatin structure, many nonhistone protein substrates of SET7/9 have been discovered, including PGC-1α, p53, nuclear hormone estrogen receptor alpha (ERα) and DNMT1. 9, 33, 34 Lysine-specific demethylase 1 (LSD1) is the first enzyme identified to be capable of removing the methyl group from methylated lysines in histone proteins. LSD1 reportedly demethylates histone H3 on Lys4 (H3K4) and on Lys9 (H3K9) 35, 36 and PGC-1α is methylated by SET7/9 and demethylated by LSD1 at Lys779.
Most recently, a new mechanism reinforcing the coactivator function of PGC-1α by methylation was revealed. 9 SET7/9 mediates the methylated status of PGC-1α. In particular, in response to metabolic stress of fasting in liver, the interaction of methylated PGC-1α (K779Me) with RNA methytransferase 7 (NSUN7) could increase the enrichment of m 5 C-modified eRNAs (enhancer RNAs) at enhancers of PGC-1α-targeted genes, heightening the transcriptional program of these genes. Accordingly, ablation of SET7/9 and NSUN7 resulted in depletion of the PGC-1α-targeted genes. 9 Collectively, methylation may link epigenetic circuit with transcriptional initiation of PGC-1α-targeted gene programs and promote mitochondrial proliferation under conditions of energy crisis.
Ubiquitination of PGC-1α
PGC-1α is a short-lived protein and its cellular levels are tightly controlled by a dynamic balance of synthesis and degradation. Ubiquitination and subsequent proteasome-mediated degradation greatly affect the expression level of PGC-1α. The N-terminal AD (amino acids 1–185) of PGC-1α has previously been considered to have no effect on protein stability and subcellular distribution. Ubiquitination of PGC-1α relies on the integrity of the C-terminal region, which might interact with PGC-1α-specific E3 ligases. Deletion of the C-terminal fragment of PGC-1α (amino acids 565–798) was demonstrated to completely block ubiquitination and stabilized the protein. Interestingly, another study revealed that instead of entering into the proteasome for degradation, the polyubiquitinated PGC-1α was prone to form intranuclear aggregation and the N-terminal of PGC-1α was required for targeting the polyubiquitinated PGC-1α to the proteasome. 22 Thereby, intramolecular interactions between the N- and C-terminal regions might cooperatively tune PGC-1α coactivator function by regulating protein stability and sublocalization.
In fact, multiple posttranslational modifications always act in concert to finely control the activity and stability of PGC-1α, which is exemplified by the regulation of ubiquitination of PGC-1α in a phosphorylation-dependent manner. SCF Cdc4 is identified as an E3 ubiquitin ligase that regulates PGC-1α stability through ubiquitin-mediated proteolysis. PGC-1α harbors two Cdc4 phosphodegrons that bind Cdc4 when phosphorylated by GSK3β on T295 and by p38 MAPK on Thr262, Ser265, and Thr298, resulting in SCF Cdc4 -dependent ubiquitination and degradation of PGC-1α in the nucleus. 37
Overall, various posttranslational modifications provide an effective and flexible system for the regulation of activity and organelle localization of PGC-1α, which contributes substantially to its nodal roles in mitochondrial energy metabolism.
Impairment of the IRS/PI3-K/Akt-mediated signal transduction module is at the very core of insulin resistance phenotypes. As a result, these molecules have been under intense scrutiny for a number of years. Morris Birnbaum discussed the regulation of glucose and lipid metabolism by Akt/PKB. He summarized the phenotypes of knockout mice for the individual three Akt isoforms. While redundant in some aspects, the three isoforms play unique functions in other areas. An example is Akt2 in the context of insulin signaling in the liver. Although there is partial compensation by Akt1 for the loss of Akt2, 80% of insulin-mediated FoxO1 phosphorylation is mediated by Akt2, while only 20% is due to Akt1 activity. Despite this difference in activity, Akt1 is able to reduce the expression of the FoxO1 target genes, PEPCK and G6Pase, in Akt2 (−/−) mice. Nevertheless, insulin is still unable to suppress glucose production in Akt2(−/−) mice, and this may be due to decreased Akt-mediated Ser570 phosphorylation of PGC-1α.
Morris White stressed the finding of largely overlapping and compensatory roles of IRS1 and IRS2 in the liver on the regulation of gluconeogenic and lipogenic gene expression. Only the loss of function of both IRS1 and IRS2 in the liver caused an increase in PEPCK, GLUT2, and PGC-1α as well as major changes in glucose metabolism, a mild increase in triglyceride synthesis, and changes in whole-body insulin resistance. IRS proteins are rather big and offer a large number of regulatory phosphorylation sites. Sarah Dunn identified >30 Ser/ Thr phosphorylation sites on murine Irs2, but focused on Ser965, which is found phosphorylated at high levels in the basal state of serum-starved cells and slightly increased upon insulin treatment. Both basal and insulin-stimulated phosphorylation events depend on the ERK/ MAP kinase cascade, and the basal phosphorylation at this site is essential for IRS2 function. Other MAPKs, such as JNK, play differential roles and are required for growth factor-induced phosphorylation. Ronald Kahn also focused on the insulin receptor signal transduction cascade. He emphasized “the critical nodes” IRS, PI3K, and AKT because all three are essential for the action of insulin, have multiple isoforms, play both positive and negative roles, and cross-talk to each other. PI3K is composed of a p110 catalytic subunit and a p85 regulatory subunit. Interestingly, Pik3r1 (p85α) or Pik3r2 (p85β) knockout mice show improved insulin sensitivity. In addition, Pik3r1-null mice display hypoglycemia, hypoinsulinemia, and decreased hepatic glucose output. Pik3r1 knockout mice decrease JNK activity and have a decreased phosphorylation state of Ser307 on IRS1 in response to insulin. The viral reconstitution of p85 rescues JNK activity and phosphorylation of Ser307 on IRS1.
Insulin signal transduction in the periphery has long been the only area of insulin action. However, recent exciting data on central effects of insulin highlight the important role of insulin action in the brain. Luciano Rossetti elaborated on insulin signaling in the medial basal hypothalamus (MBH) and its impact on hepatic gluconeogenesis. Using a PI3K activator or infusing PtdIns(3,4,5)P3 in the hypothalamus mimicked the effects of insulin on suppression of gluconeogenic gene expression and glucose production in the liver. These effects were inhibited in Sur1-KATP channel (−/−)
mice and also in Akt2 (−/−) , but not in Akt1 (−/−) , mice. Additionally, overfeeding rats for 3 d induces liver insulin resistance, which is partially alleviated with PtdIns(3,4,5)P3 infusion in the hypothalamus and completely restored with diazoxide administration. Therefore, defective brain-to-liver signaling appears to be an early event in the development of insulin resistance. Other downstream targets of insulin action include the Foxo transcription factors. Domenico Accili discussed the involvement of forkhead transcription factors in diabetes and obesity, focusing on the role of the forkhead protein FoxO1 in protecting pancreatic β cells against oxidative stress. FoxO1 forms a complex with two other proteins, Pml and Sirt1, in order to activate factors involved in transcription of the insulin gene, such as MafA. Glucose promotes FoxO1 acetylation, which targets it to PML bodies in the nucleus and protects it from ubiquitin-mediated degradation. Beyond the insulin signal transduction pathway, FoxO1 also plays a role the leptin pathway. Adenoviral overexpression of a constitutively active FoxO1 resulted in a loss of leptin action in AGRP neurons, through a mechanism that may involve competition with Stat3 for binding to AGRP and POMC promoters. FoxO1 overexpression also results in hepatic lipid accumulation, increased VLDL secretion, increased triglyceride synthesis, and decreased FFA oxidation.
AMP-activated protein kinase (AMPK) is an enzyme that serves as a fuel gauge, activated in states of low energy. AMPK can modify diverse metabolic pathways. In liver, activation of AMPK results in enhanced fatty acid oxidation and represses gluconeogenesis and production of cholesterol and triglycerides. Reuben Shaw described the regulation of AMPK and glucose homeostasis in the liver by the serine/threonine kinase, LKB1, and its requirement for the beneficial effects of metformin. LKB1 is required for the phosphorylation and activation of AMPK in the liver, but not in the muscle. Mice with a liver specific knockout of LKB1 have decreased P-AMPK and increased PGC-1α and G6Pase expression, and they are hyperglycemic and glucose intolerant. Additionally, in the livers of these mice, the CREB coactivator TORC2 is hypophosphorylated on Ser171, leading to increased TORC2 and CREB activity and the subsequent PGC-1α and gluconeogenic gene expression. TORC2 RNAi reduced PGC-1α expression in the liver and blood glucose levels in the knockout mice. LKB1 therefore leads to AMPK and Sik2 activation, which then leads to the phosphorylation and nuclear exclusion of TORC2 and CREB and the reduced expression of PGC-1α and gluconeogenic genes. Markus Stoffel focused on insulin-regulated transcription factor FoxA2, which is phosphorylated by Akt at the critical and conserved residue Thr156. In wild-type mice, insulin phosphorylates FoxA2 and excludes it from the nucleus, resulting in an inhibition of the transcription of its target lipolytic and ketogenic genes in the liver. When a T156A FoxA2 mutant that cannot be phosphorylated is expressed, there is an increase in lipolytic, mitochondrial β-oxidation, and ketogenic genes. Adenoviral expression of this mutant in livers of ob/ob mice decreased plasma glucose and insulin levels and increased oxygen consumption and heat production. Furthermore, FoxA2 appears to be required for ApoM expression and subsequent preβ-HDL formation. Therefore, decreased FoxA2 activity leads to reverse cholesterol efflux. Additional results were shown suggesting that FoxA2 phosphorylation depends on IRS1 or IRS2, while FoxO1 phosphorylation strictly depends on IRS2.
Takashi Kadowaki discussed the differences in globular and full-length adiponectin signaling: Full-length adiponectin affects the liver and muscle, signals mostly through the adiponectin receptor-2 (adipoR2), and leads to increased AMPK and PPARα activity and increased fatty acid oxidation in the liver globular adiponectin affects only muscle, signals mostly through adiponectin receptor-1 (adipoR1), and leads to p38 MAPK and AMPK activation and increased fatty acid oxidation. He discussed knockout models of adipoR1 and adipoR2 and their metabolic phenotypes, as well as presented data on adopoR1/2 double-knockout mice. These mice are viable, a finding different from data presented by Phil Scherer, who showed that double knockout mice are not viable in his hands. The differences in viability may be explained by the different knockout strategies employed, since the Kadowaki group has identified low levels of alternatively spliced messages in their double-null mice, which may explain why these mice manage to survive. Many questions remain to be answered in terms of how these receptors convey an adiponectin signal into the cytoplasm, what their specificity is, and whether or not additional receptors exist for adiponectin. Many groups have, however, described that adiponectin potently activates AMPK in a number of different cell types, thereby positively affecting the cellular energy metabolism. Juleen Zierath discussed a mechanism of AMPK activation that is independent of adiponectin AMPK can be activated pharmacologically by 5-aminoimidazole-4-carbox-amide-1-β-4-ribofuranoside (AICAR), which is phosphorylated in vivo to the AMP analog AICAR-monophosphate (ZMP). Treatment with AICAR leads to a striking normalization of blood glucose levels and glucose tolerance in ob/ob mice after in vivo treatment, providing evidence that targeting the AMPK pathway effectively improves metabolic defects in diabetes.
Circulating Mitochondrial DAMPs: Much More Than Waste-Derived Molecules
In agreement with the original “danger theory” of inflammation, 46 several syndromes characterized by systemic inflammatory response, including trauma, HIV and cancer, have been found to be associated with increasing levels of circulating mitochondrial DAMPs. 15,57,58
Interestingly, together with proinflammatory cytokines (e.g., IL6, tumor necrosis factor alpha [TNF-α], Regulated on Activation Normal T Cell Expressed and Secreted [RANTES], and IL1 receptor antagonist), circulating levels of mtDNA molecules have been found to increase past the age of 50. 59 In this context, the finding of increased TNF-α production following exposure of monocytes to mtDNA concentrations similar to those detected in vivo supports the idea of circulating mtDNA as factor contributing to systemic inflammation during aging (“inflamm-aging”). 59
TFAM has recently been proposed to act as a mitochondrial DAMP both in rodents and humans. 60 Mouse embryonic fibroblasts expressing only one TFAM allele show a 50% decrease in mtDNA content associated with constitutive activation of the cGAS-STING-IRF3 pathway. 61 Moreover, TFAM acts as a specific DAMP-inducing proinflammatory and cytotoxic response in in vitro models of human brain microglia. 60 In addition, TFAM appears to be relevant during inflammation through the modulation of its binding to mtDNA. Indeed, TFAM has been involved in rerouting oxidized mtDNA to lysosomes for degradation in neutrophils. 16 The extrusion of oxidized nucleoids by neutrophils in SLE represents a powerful immune system activator. 16
Systemic inflammation is a recognized feature of several musculoskeletal disorders. Recently, a fracture-initiated systemic inflammatory response syndrome (SIRS), characterized by increased circulating levels of cell-free mtDNA, has been documented in patients with hip fracture. 62 In such a context, circulating mtDNA appears to promote the development of inflammation by recruiting leukocytes. 62 It is noteworthy that specific alterations of the MQC axis have been identified in muscle biopsies obtained from old hip-fractured patients. 63 Taken together, these findings suggest the existence of a functional link between muscular mitochondrial dysfunction and systemic inflammation, possibly mediated by the release of mtDNA into the circulation. Conversely, moderate aerobic exercise, a well-known anti-inflammatory intervention, 64 decreases systemic cell-free mtDNA levels in healthy adults. 65
Despite the evidence supporting cell-free mtDNA and TFAM-bound mtDNA as DAMPs, the exact mechanism of mtDNA delivery into the cytosol and then into the circulation is currently unknown. As elegantly reviewed by Safdar et al., 66 one of the mechanisms through which eukaryotic cells communicate with each other is a signaling system relying on the exocytosis of proteins containing secretion-targeting sequences. 67 Being too labile within the extracellular environment, proteins and other macromolecules (including mtDNA), may be secreted within small membranous extracellular vesicles. 68–70 A system of vesicles called exosomes is thought to use such a pathway to release a set of molecules (exerkines) in muscle under endurance exercise. 66 Exerkines contribute to mediating the beneficial effect of exercise by allowing systemic adaptations through autocrine, paracrine, and/or endocrine signaling. 66 Interestingly, cell-free mtDNA has been identified among the molecules released via exosomes. 66 Regardless of the actual mechanisms generating and releasing DAMPs, which goes beyond the purpose of this review, their accumulation has been shown to activate tissue resident macrophages and also favor tissue leukocyte infiltration. 71
The involvement of mitochondrial dysfunction and inflammation in the pathogenesis of muscle wasting disorders is well established. However, if and how mitochondrial DAMPs are involved in the inflammatory response associated with those conditions is presently unknown. The existence of crosstalk between mitochondria and the inflammasome is a highly promising area of investigation, especially in the context of muscle wasting disorders.
Nrf2 AND RADIATION EXPOSURES
In two articles published in 2010, it was reported that exposing mammalian cells and tissues to radiation activated the Nrf2-mediated transcription of various antioxidant proteins. Thus, Tsukimoto et al. (46 ) showed that exposure of murine Raw 264.7 macrophage cells to low-dose γ rays in the range 0.1–2.5 Gy caused an early (1–2 h) dose-dependent increase in cytoplasmic Nrf2 levels as well as its nuclear accumulation by 4 h and corresponding elevation of HO-1 levels by 24 h. McDonald et al. (5 ) similarly observed Nrf2/ARE-dependent gene induction after γ irradiation, albeit with markedly delayed kinetics, in several cell types. Although exposure of the stably expressing ARE-luciferase reporter cell line MCF7-AREc32 (derived from the MCF7 human breast cancer cell line) to either single dose of radiation in the range 0.05–10 Gy or to three daily fractions of 0.5, 2 or 4 Gy did not significantly increase luciferase expression at 24 h after completion of the irradiation, a dose-dependent activation was observed in cells receiving five daily fractions of 0.5, 1, 2, 3 or 4 Gy and evaluated at 3 h after the final dose (5 ). The delayed nature of this activation was confirmed by the observation that single doses of 2–8 Gy did indeed enhance the ARE-reporter signal at days 5–15 postirradiation before subsiding, suggesting that it represents a second-tier antioxidant response. Wild-type mouse embryo fibroblasts (MEFs), NIH-3T3 murine fibroblasts and DC2.4 murine dendritic cells (but not Nrf2-knockout MEFs) showed a similar pattern of delayed induction of mRNAs for two Nrf2-regulated genes, HO-1 and GSTA2, after single or fractionated doses. Corresponding increases in HO-1 protein levels were seen in wild-type MEFs, NIH-3T3 and primary murine bone marrow cells, but again not in Nrf2-knockout MEFs. Functional GSH activity at day 5 after a single 8 Gy dose increased by ∼50% in wild-type (but not in Nrf2-knockout) MEFs. Similar responses were observed in vivo in C57BL/6 mice after five daily 2 Gy fractions, where splenic HO-1 (but not GSTA2) gene expression increased significantly. Nrf2-knockout MEFs and C57BL/6 mice showed increased radiosensitivity compared to their wild-type counterparts (5 ). Nrf2 activation in MEFs after 8 Gy irradiation was temporally correlated with delayed ROS production, peaking at ∼5 days, with ROS induction greatly increased in Nrf2-knockout MEFs (5 ).
Subsequent reported studies have described phenotypic and mechanistic features of the Nrf2 response to radiation in various model systems in relationship to enhanced cell survival. Many of these studies were reviewed by Sekhar and Freeman (37 ). For example, activation of Nrf2/ARE signaling has been shown to lower intracellular ROS and confer radioresistance in fibroblasts, bronchial and breast epithelial cells, DU145 prostate cells, glioblastoma and squamous cell lung cancer cells. Knockdown or inhibition of Nrf2 in human cancer cell lines typically results in elevated ROS levels and radiosensitization, as it did in Nrf2-knockout MEFs. Collectively, these findings suggest that Nrf2 does indeed promote a pro-survival response in irradiated cells. Similar observations have been made in in vivo models, including the increased radiosensitivity of Nrf2-knockout mice (37 ). In addition to antioxidant responses, some Nrf2/ARE-regulated enzymes are involved in the repair of radiation/ROS-induced DNA damage, such as the BER protein 8-oxoguanine DNA glycosylase (OGG1) and the homologous recombination repair (HRR) protein RAD51 (47 ) furthermore, interactions between Nrf2 and factors such as p53-binding protein 1 (53BP1) and breast cancer 1 (BRCA1) protein may influence the cellular choice of double-strand break (DSB) repair pathway, i.e., non-homologous end joining (NHEJ) versus HRR (37 ). Sekhar and Freeman (37 ) also reiterate the importance of inflammatory cytokines in the radiation response of normal tissue and of Nrf2 in regulating cytokine expression, and they suggest that the pro-survival role of Nrf2 is related to its ability to modulate the pro-inflammatory cytokine response that can generate ROS as an effector against microbial challenge. Several published studies using various cell types [e.g., (46, 48, 49 )] also indicate that radiation-induced Nrf2 activation and downstream effects can be suppressed by mitogen-activated protein kinase (MAPK)1/3 inhibitors such as U0126 or by shRNA MAPK knockdown, suggesting that this pathway might play a key role in this response.
The Nrf2-mediated induction of antioxidants has also been implicated in the above-mentioned adaptive response to radiation in which priming cells with a low dose of radiation can invoke resistance to a higher challenge dose delivered several hours later. However, defining the mechanism of this effect is complicated not only because it is not universally observed (20 ) but also because many contributing processes are likely to modify adaptive responses, including DNA repair, cell cycle checkpoints, stress chaperone proteins and intercellular signaling pathways involving, e.g., p53 and MAPK1/3 (50 ). Bravard et al. (51 ) examined the effect of γ rays on the activity and levels of various antioxidant proteins in AHH-1 human lymphoblast cells using an adaptive paradigm of a 0.02 Gy priming dose followed 6 h later by a 3 Gy challenge dose. Although the priming dose itself had little effect, adapted cells did exhibit slightly elevated activity/levels of SOD2, GST, GPx and catalase versus control cells at 3 h after the 3 Gy challenge dose, suggesting that such events may contribute to adaptation. Similarly, delivery of a priming dose of 5 cGy to AG1522 normal human skin fibroblasts caused Nrf2 translocation from the cytoplasm to the nucleus and induction of the HO-1 gene and protein, which presumably contributed to the observed adaptation to a subsequent 2 Gy challenge dose of X rays given 12 h later (52 ). In contrast, Miura (50 ) reported no changes in the activity of GST, GSR and catalase in rat glial cells receiving a 0.1 Gy priming dose 3 h prior to a 2 Gy challenge dose. Although in the latter two cases the phenotypic adaptive response was not demonstrated, there is evidence that Nrf2 responds to low-dose irradiation. For example, in human hematopoietic stem cells (HSCs), hypersensitivity to low doses of radiation was dependent on immediate increased levels of ROS that activated the Keap1-Nrf2 antioxidant pathway leading to autophagy (52 ).
The adaptive response to radiation is clearly multifactorial and extensive mechanistic discussion is beyond the scope of this review. Here, we consider two recently published studies that shed some light on the role of Nrf2 in such responses. First, the studies described so far typically involved low-LET beams (X rays or γ rays). In their published study, Chen et al. (53 ) examined whether high-LET α particles evoked a similar adaptive response in A549 human lung adenocarcinoma cells and, if so, whether the Nrf2-mediated induction of antioxidants might play a role therein. A clear adaptive response to α particles (increase in cell survival) was apparent in cells receiving a priming dose of 5 cGy delivered 6 h prior to a challenge dose of 75 cGy. Nrf2 elevation and accumulation in the nucleus as well as transcriptional activation of its target gene HO-1 were seen at 6 h after 5 cGy irradiation. Also, DSB levels (γ-H2AX foci) at 3 h after the 75 cGy challenge dose were decreased in cells that received the 5 cGy priming dose compared to non-primed cells, presumably reflecting enhanced DNA repair. Knockdown of Nrf2 using shRNA suppressed the adaptive response, as did the MAPK1/3 inhibitor U0126. The autophagy inhibitor 3-methyladenine and the ROS scavenger n -acetyl cysteine also blocked the increase in Nrf2 and HO-1 levels and the adaptive response, with the latter also blocking the autophagy response. Collectively, these observations suggest that the adaptive response to α particles in A549 cells is mediated by ROS elevation, autophagy and activation of the Nrf2 antioxidant pathway, which is very similar to findings in human HSCs (52 ). Second, there appears to be a more general aspect of Nrf2-mediated adaptation insofar as a whole-body priming X-ray dose of 7.5 cGy given to C57BL/6J diabetic mice in 1 or 3 daily fractions (but not 1 × 2.5 cGy) was able to protect against manifestations of diabetes injury that are related to excessive ROS generation in the kidney (54 ). The priming dose upregulated Nrf2 expression in the kidney at 3–6 h postirradiation as well as Nrf2 function, as reflected by levels of its downstream antioxidants (NQO1 at 3–6 h and HO-1 at 3–9 h) it also attenuated various manifestations of diabetes-induced oxidative damage to the kidney (inflammation, dysfunction).
The cellular regulation of Nrf2 is actually more complex than outlined above. Indeed, there is evidence of crosstalk between the Ref1 and Nrf2 oxidative stress-sensing proteins. For example, Ref1 has been reported to negatively regulate Nrf2 in a variety of cell types via its redox function (55 ). The sestrin proteins described earlier have been suggested to stimulate the Nrf2-mediated antioxidant response by enhancing the disruption of the Keap1-Nrf2 interaction via the p62/SQSTM1 (sequestosome 1) protein, thereby promoting the autophagic degradation of Keap1 (56 ). Both the p62 and sestrin2 genes are themselves transcriptional targets of Nrf2, resulting in a positive feedback loop. Also worth noting is that the cellular homeostatic response to radiation/ROS-mediated stress involves a critical element of crosstalk between the Nrf2 and p53-p21 WAF1 pathways. Such interactions will be discussed later.
A final but important point relates to the kinetics of Nrf2 induction by radiation in various model systems. Although the two initial reports of this effect (5, 46 ) both highlighted the importance of Nrf2 induction in the cellular response to radiation, they also indicated very different activation kinetics. Whereas a rapid (≤24 h) induction of Nrf2/ARE-dependent events by exposure was observed by Tsukimoto et al. (46 ) in mouse Raw 264.7 macrophage cells and in subsequent studies with a variety of human cell lines, McDonald et al. (5 ) found that induction was minimal at less than 48 h in several cell types and required a delay of ∼5 days to fully manifest. However, Rodrigues-Moreira et al. (52 ) suggest a recurring state of persisting oxidative stress that might drive delayed waves of Nrf2 induction. This may be driven by waves of further cell death and micronuclei formation and be dependent on cell line differences related to cell fate e.g., the late activation seen by McDonald et al. (5 ) appears to reflect an enzyme-mediated delayed ROS production related to radiation-induced senescence (a form of cell growth arrest that will be discussed below), which Nrf2 generally inhibits. In keratinocytes, for example, radiation-induced senescence is associated with increased ROS levels at four days postirradiation due to activation of enzymes such as the NOX proteins, events that are blocked by the B lymphoma Mo-MLV insertion region 1 homolog (Bmi-1) polycomb complex protein that, among other activities, regulates mitochondrial oxidative stress levels and increases radioresistance (57 ). Senescence can also trigger cell fate decisions associated with reprogramming in which Nrf2 appears to play a critical role. In particular, Nrf2, perhaps induced by late ROS production, orchestrates the metabolic shift from oxidative to glycolytic energy production that is associated with reprogramming of induced pluripotent stem cells (58 ) and drives anabolic pathways essential for metabolic reprogramming through enhanced glucose utilization and the pentose phosphate cycle (59–61 ), which is also associated with increased radioresistance. Although the focus here has been on Nrf2's role in regulating redox homeostasis, these additional roles in regulating a variety of metabolic enzymes that contribute to the rapid synthesis of macromolecules (e.g., via activation of enzymes involved in nucleotide synthesis) and to cell proliferation are clearly of relevance in the cellular response to oxidative stress. Also relevant is the role of Nrf2 in the early replacement of proteasome subunits after the rapid radiation-induced disassembly of the 26S proteasome, which is cell line dependent, presumably being regulated by redox and metabolic status (62 ). Indeed the early Nrf2 response could be for this very purpose.
Mitochondrial fidelity and metabolic agility control immune cell fate and function
Laboratory of Mitochondrial Biology and Metabolism, Cardiovascular Branch, National Heart, Lung, and Blood Institute, NIH, Bethesda, Maryland, USA.
Address correspondence to: Michael N. Sack, Laboratory of Mitochondrial Biology and Metabolism, Cardiovascular Branch, National Heart, Lung, and Blood Institute, NIH, Building 10CRC, Room 5-3150, 10 Center Drive, Bethesda, Maryland 20892, USA. Phone: 301.402.9259 Email: [email protected]
Remodeling of mitochondrial metabolism plays an important role in regulating immune cell fate, proliferation, and activity. Furthermore, given their bacterial ancestry, disruption in mitochondrial fidelity leading to extravasation of their content initiates and amplifies innate immune surveillance with a myriad of physiologic and pathologic consequences. Investigations into the role of mitochondria in the immune system have come to the fore, and appreciation of mitochondrial function and quality control in immune regulation has enhanced our understanding of disease pathogenesis and identified new targets for immune modulation. This mitochondria-centered Review focuses on the role of mitochondrial metabolism and fidelity, as well as the role of the mitochondria as a structural platform, for the control of immune cell polarity, activation, and signaling. Mitochondria-linked disease and mitochondrially targeted therapeutic strategies to manage these conditions are also discussed.
Given the preponderance of mitochondria within cells, the late physician-scientist and essayist Lewis Thomas posited that he “could be taken for a very large, motile colony of respiring bacteria” ( 1 ). At that time, despite their bacterial origins, there was little appreciation of the integral role of mitochondria in the regulation of immunity. Today, our recognition of the influence of mitochondrial fidelity on immune function extends beyond classical immunology to include mitochondria-linked immune perturbations that contribute to inflammation of degenerative diseases and to autoimmunity ( 2 ). In contrast, recognition of the metabolic underpinnings of immune cell function has been evident since the 16th century as manifested by the persistence of the maxim that “starving a fever and feeding a cold” hastens recovery from infectious illnesses ( 3 ). Components of this adage have been experimentally validated ( 4 ), and intermittent fasting and caloric restriction show evidence of blunting inflammation-linked diseases ( 5 – 10 ).
The field of study exploring the interdependence between metabolism and immune cell fate and function is termed immunometabolism ( 11 – 13 ). Additionally, given the persistence of bacterial signatures within mitochondria, mitochondrial components play diverse roles in triggering immune surveillance programs ( 2 , 14 ). These aspects of mitochondrial biology and metabolism, in addition to their role in disease pathophysiology and their potential for therapeutic targeting, are the focus of this Review.
Metabolic plasticity and immune cell fate. The bidirectional control between cellular metabolism and innate and adaptive immune cell fate and function has been most extensively explored in the modulation of macrophage polarization and of CD4 + T cell differentiation/activation, respectively ( 11 – 13 ). The metabolic pathways that contribute to immune cell activation and differentiation are schematized in Figure 1A, and examples of how they support immune cell fate are summarized here. On the one hand, inflammatory macrophages (M1) employ aerobic glycolysis for energy production and divert glycolytic intermediates into the pentose phosphate pathway for NADPH synthesis. NADPH in turn is catabolized by NADPH oxidase to generate reactive oxygen species (ROS) to facilitate antimicrobial effects ( 15 ). The corresponding preference for glutamine oxidation additionally promotes mitochondrial electron transfer chain (ETC) ROS production ( 16 ). On the other hand, alternatively polarized macrophages (M2) rely more on glucose and fat oxidative metabolism to orchestrate reparative functions ( 17 , 18 ). However, the requirement for fatty acid oxidation in reparative macrophage function has recently been disputed by observations that genetic disruption of macrophage mitochondrial fat uptake did not prevent M2 polarization ( 19 ). It should be noted that the binary classification of macrophages is partially an experimental construct, given that macrophage polarity is a spectrum with multiple distinct signatures and functional phenotypes between these “poles” ( 20 ).
Overview of metabolic remodeling with immune cell differentiation and activation. Overview of metabolic remodeling with immune cell differentiation and activation. (A) Major pathways linked to immunometabolism include the generation of ATP from cytosolic glycolysis or from oxidative phosphorylation of pyruvate, fatty acids, and glutamine. Cytosolic NADPH oxidase or mitochondria produce ROS, which can signal, oxidize proteins, or exert antimicrobial effects. The major sites of mitochondrial ROS production linked to immunometabolism are generated at complexes I and III of the ETC. (B) In monocytes, metabolic remodeling is most extensively characterized in the differentiation into the M1 and M2 cell fates. In CD4 + T cells, immunometabolism has been explored at multiple levels of differentiation, proliferation, activation, and migration. B cell immunometabolic control is less well explored however, metabolic remodeling is important for different B cell fates. FAO, fatty acid oxidation Glut Ox, glutamine oxidation Gly, glycolysis GO, glucose oxidation PPP, pentose phosphate pathway RET, reverse electron transport in the electron transfer chain for ROS generation.
The transition from quiescent to activated effector CD4 + T cells is a growth- and energy-demanding event that is accomplished, in part, by a switch from preferential oxidative phosphorylation to an increased reliance on aerobic glycolysis, the pentose phosphate pathway, and glutamine oxidation ( 21 – 24 ). Conversely, regulatory T cells (Tregs) retain reliance on oxidative phosphorylation ( 25 ), although they depend on glycolysis for migration ( 26 ).
The genetic requirement of these metabolic regulatory effects on immune cell fate is evident where the experimental disruption of glucose uptake impairs effector T cell proliferation and activation ( 21 ) and its promotion augments effector T cell function ( 27 ) and inflammatory macrophage polarity ( 28 ). This reliance on defined metabolic preference is further validated by the promotion of M2 polarization by experimental induction of glucose oxidation ( 18 ), and by the facilitation of M1 polarization by genetic disruption of oxidative metabolism and promotion of glycolysis ( 29 ).
The characterization of immunometabolism is extending into other immune cells, and metabolic preferences support dendritic cell immunogenic function ( 30 ) and B cell quiescence, proliferation, antibody-generating capacity, and survival ( 31 – 35 ). As is discussed in the final section of this Review, therapeutics to modulate substrate preference are being tested in T and B cell–linked disease such as systemic lupus erythematosus. A schematic summarizing metabolic remodeling links to immune cell fate is shown in Figure 1B.
TCA cycle “fragmentation” in immune activation. The tricarboxylic acid (TCA) cycle generates reducing equivalents for ETC function through oxidation of acetyl-CoA derived from glucose, fat, and amino acids. Immune Toll-like receptor (TLR) and cytokine signaling ( 36 ) disrupts coordinated flux of intermediates through the TCA-inciting accumulation of distinct metabolites. Lipopolysaccharide (LPS) signals through TLR4 to upregulate immunoresponsive gene 1 (IRG1), which encodes an enzyme that decarboxylates the TCA intermediate cis-aconitase to produce itaconic acid ( 37 ). Itaconic acid exhibits antimicrobial activity ( 37 ) and macrophage antiinflammatory effects ( 38 ) and functions as an endogenous succinate dehydrogenase (SDH) inhibitor ( 39 ). Inhibition of SDH, in turn, promotes succinate accumulation in the TCA cycle in addition to inhibition of complex II of the ETC ( 38 ). The accumulation of succinate additionally facilitates transactivation of the proinflammatory cytokine IL-1β ( 40 ).
Interestingly, modulation in metabolic intermediate flux through the TCA cycle also controls macrophage polarity ( 41 ). Here, levels of specific metabolites confer distinct macrophage functions, as evidenced by succinate and extramitochondrial acetyl-CoA functioning as key intermediates for optimal M1 function. These macrophages exhibit reduced isocitrate dehydrogenase (IDH) transcript levels with diminished IDH enzyme activity ( 41 ). This results in reduced citrate conversion to α-ketoglutarate, resulting in citrate accumulation. Excess mitochondrial citrate translocates into the cytosol through the mitochondrial citrate carrier (CIC). Cytosolic citrate conversion to acetyl-CoA promotes production of proinflammatory prostaglandin, nitric oxide, and ROS ( 42 ). Interestingly, increased mitochondrial citrate also undergoes decarboxylation to generate itaconic acid, which functions as described above ( 37 – 39 ), and genetic knockdown of CIC impairs LPS-induced M1 activation ( 42 ). Alternatively, increased TCA glutamine utilization fuels M2 polarization by upregulating TCA-encoding genes and chemokine production ( 41 ). Recent evidence also shows that itaconate directly alkylates cysteine residues on a transcriptional repressor, KEAP1. This in turn enables the transcriptional induction of antioxidant and antiinflammatory encoding genes to limit inflammation and modulate type I IFNs ( 43 ). A dimethyl derivative of another TCA intermediate, fumarate, also modulates metabolism via a posttranslational mechanism. Here, dimethyl fumarate covalently modifies catalytic cysteine residues on the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in a process termed “succination.” This blunts glycolysis in myeloid and lymphoid cells, conferring antiinflammatory effects ( 44 ). Together these data illustrate the integration of metabolism and immune cell function where inflammatory signaling induces alterations in intermediary metabolism to generate metabolites that serve as signal transducers and posttranslational modifiers to regulate subsequent immune cell function.
ETC perturbations in the control of immune signaling. A role of LPS signaling in impairing the ETC has been recognized for decades ( 45 ), and recently was linked to LPS-induced mitochondrial ROS production ( 46 ). A mechanism underpinning the LPS-ROS nexus is TLR4-driven succinate oxidation that orchestrates mitochondrial hyperpolarization with concomitant excess macrophage ROS production. This ROS signaling–induced inflammatory gene programing upregulates IL-1β, and inhibition of succinate oxidation during TLR4 activation promotes antiinflammatory gene expression signatures with induction of the canonical antiinflammatory cytokine IL-10 ( 16 ). The observation of blunting of this ROS generation by rotenone-mediated ETC complex I inhibition ( 16 ) implicates reverse electron transport (RET) as a mechanism propagating complex I ROS generation ( 47 ).
Another identified mechanism is operational following macrophage exposure to live E. coli or bacterial RNA. Here, bacterial components promote disassembly of mitochondrial supercomplexes with disruption of complex I respiration ( 48 ). This mechanism depends on phagosomal NADPH oxidase ROS generation to oxidize specific complex I proteins to facilitate complex disassembly ( 49 ). Complex I disruption in turn promotes increased respiration through complex II in parallel with IL-1β-induction, and the inhibition of complex II under these conditions reduces IL-1β and increases IL-10 levels ( 48 ). Interestingly, activation of TNF-α by E. coli is independent of complex II activity, a finding consistent with prior studies showing that TNF-α production was less sensitive to macrophage metabolic preference ( 40 ). Notably, complex I disassembly and its link to complex II activation are independent of LPS and TLR4 signaling ( 48 ). Although the TLR4/RET signaling ( 16 ) and microbial RNA/NADPH oxidase ( 48 ) mechanisms have not been reconciled, they collectively support that numerous mechanisms of ETC disruption regulate IL-1β–driven inflammation. ETC-derived ROS similarly plays a critical role in antigen-specific T cell activation ( 50 ).
The role of mitochondrial constituents in immune activation stems in part from the incorporation of a prokaryotic endosymbiont into more complex (archaeal) bacteria ( 51 ). During subsequent evolution into eukaryotic cells, the majority of the engulfed bacterial genes were transferred to the nuclear genome ( 52 ). The residual bacterial genome persisted as the mitochondrial genome, and the hypomethylated CpG motifs of mitochondrial DNA (mtDNA) retained immunogenic properties of bacterial CpG DNA motifs ( 53 ). Additional mitochondrial features that phenocopy bacteria include protein synthesis, where mtDNA translation is initiated at N-formyl-methionine residues ( 54 ), and retention of the phospholipid cardiolipin within the inner mitochondrial membrane (IMM) ( 55 ). A consequence of these bacterial remnants is that mitochondrial cargo is viewed as foreign by mammalian intracellular and extracellular immune surveillance programs. These mitochondrial features are in a sense “immune-privileged” and isolated from the cytosol by encasement by the outer mitochondrial membrane (OMM). Loss of integrity of the mitochondria exposes these components to immune pattern recognition receptors (PRRs), functioning as mitochondrial alarmins to trigger immune activation. The canonical immune surveillance programs that recognize signatures/components of mitochondrial stress are discussed below. It should be noted that although these programs are reviewed separately here, the potential for them to be activated contemporaneously is beginning to be recognized ( 56 – 58 ). Additionally, the OMM functions as a structural platform for the assembly of immune regulatory complexes ( 59 – 61 ), and mitochondria–endoplasmic reticulum (ER) contact sites operate as a signaling hub enabling metabolic remodeling to reactivate memory CD8 + T cells ( 62 ).
NLRP3 inflammasome. The inflammasome is an intracellular immune surveillance program that recognizes either pathogen-associated molecular patterns (PAMPs) or host cell–derived damage-associated molecular patterns (DAMPs). The inflammasome is a multiprotein complex that assembles and self-oligomerizes to promote cleavage and activation of canonical cytokines, namely IL-1β and IL-18, to amplify immune responses ( 63 , 64 ). The nucleotide oligomerization domain–like (NOD-like) receptor family pyrin domain 3 (NLRP3) inflammasome is activated by sterile inflammation associated with crystal-linked diseases like gout, and by metabolic diseases including obesity, diabetes, hyperlipidemia, and cardiovascular disease ( 65 ), and its activation further exacerbates these diseases ( 66 – 69 ). For example, in obesity, the inflammasome is initiated through adipose tissue hypertrophy with macrophage infiltration and cytokine secretion elevated circulating saturated fatty acids, glucose, and lipids and/or obesity-linked endotoxemia mediated by upregulation of NLRP3 and its canonical cytokines (pro–IL-1β and pro–IL-18) via NF-κB ( 70 – 73 ). Transcriptional induction of NLRP3 inflammasome components is termed “priming,” and mitochondrial ROS signaling contributes to this regulation ( 74 , 75 ). Priming-induced mitochondrial ROS also facilitates externalization of cardiolipin and promotes the association of NLRP3 and caspase-1 with mitochondria ( 76 ).
Following priming, subsequent activation is instigated by assembly and oligomerization of canonical inflammasome constituents ( 65 ). NLRP3 complex oligomerization initiates caspase-1 activation and cleavage of pro–IL-1β and pro–IL-18 into bioactive cytokines that function to amplify inflammation ( 63 , 64 ). Multiple mitochondrial components function as NLRP3 DAMPs following disruption of mitochondrial integrity, including mtDNA ( 53 ), cardiolipin ( 55 , 76 ), and mitochondrial ROS ( 56 , 77 ). The role of mitochondria in orchestrating this program is further supported by (a) the close association between mitochondria and NLRP3 complex formation, with coordinated perinuclear localization during inflammasome induction ( 59 , 76 ) (b) the role of the mitochondrial antiviral signaling protein (MAVS) in NLRP3 recruitment to mitochondria following noncrystalline inflammasome activation ( 60 , 78 ) and (c) evidence that depletion of mtDNA in ρ 0 (mtDNA-deficient) cells blunts inflammasome activation ( 77 ). Interestingly, mtDNA binds to both the NLRP3 inflammasome and the distinct double-stranded DNA–sensing inflammasome AIM2 ( 79 ), whereas oxidized mtDNA interacts exclusively with NLRP3 ( 79 ). Finally, mitochondria also function as a structural platform upon which multiple proteins that constitute the NLRP3 inflammasome nucleate and assemble ( 59 , 60 , 76 , 79 , 80 ). Interestingly, numerous degenerative conditions, including type 2 diabetes, atherosclerosis, chronic kidney disease, and aging, manifest with sterile inflammation and mitochondrial dysfunction ( 66 , 69 , 81 , 82 ), although whether these features are mechanistically integrated has not always been experimentally confirmed.
Whether mitochondrial metabolic dysfunction rather than disruption per se activates the NLRP3 inflammasome is less well characterized ( 75 ). However, inhibition of mitochondrial respiration, promotion of aerobic glycolysis ( 75 , 83 ), and/or the selective induction of fatty acid oxidation ( 84 ) facilitate inflammasome activation. Additionally, NLRP3 activators diminish mitochondrial membrane potential and reduce intracellular NAD + levels, thereby blunting SIRT2 deacetylase activity. The consequent acetylation of the SIRT2 target α-tubulin facilitates trafficking of NLRP3 inflammasome components to spatially align and bind to mitochondria and ER to facilitate inflammasome activation ( 80 ).
Investigation into whether the mitochondrial permeability transition pore (mPTP) regulates extrusion of mitochondrial content into the cytoplasm has also been explored. Although the structure of the mPTP remains contentious ( 85 , 86 ), its ability to facilitate release of high–molecular weight molecules (approximately 1.5 kDa) through the IMM is established, as is the role of increased intracellular calcium (Ca 2+ ) in opening the pore ( 87 ), and the inhibition of opening by cyclosporin A (CsA) ( 85 , 86 ). The contribution of the mPTP to NLRP3 activation is supported by inflammasome activation as a result of increasing intracellular Ca 2+ ( 88 , 89 ) and by CsA blunting of mtDNA extrusion ( 77 ). The investigation of the mechanisms underpinning mtDNA extrusion through the mPTP has not been comprehensively defined. Nevertheless, given its proximity to ROS and its lack of protective histones, mtDNA is susceptible to oxidative stress with formation of 8-hydroxydeoxyguanosine modifications. These base modifications preferentially reside on mtDNA fragments rather than on intact circular mtDNA ( 90 ), and mPTP activation is linked to mtDNA fragment extrusion in a CsA-dependent manner ( 91 ). Additionally, the extent of this leak appears to be dependent on mtDNA fragment size ( 92 ). CsA similarly blunts NLRP3 activation in response to numerous ROS-dependent and ROS-independent inflammasome triggers ( 55 ). Finally, the interdependence of different immune activation triggers is further illustrated by the demonstration that E. coli–induced complex II–driven respiration is dependent on NLRP3 ( 48 ), although the mechanisms underpinning this are unclear.
mtDNA TLR9 signaling. Intracellular TLR9 is a distinct immune surveillance PRR that recognizes unmethylated DNA derived from bacteria, viruses, and extruded mtDNA ( 93 , 94 ) to initiate a MyD88 signaling– and NF-κB transactivation–dependent inflammatory cascade ( 95 ). Interestingly, impaired mitochondrial turnover ( 94 ), traumatic injury ( 54 ), and nonalcoholic steatohepatitis ( 96 ) promote mtDNA extravasation, triggering TLR9 signaling that results in organ-specific ( 94 , 96 ) or systemic inflammation ( 54 ). Although the mechanisms underpinning recognition of nucleotides by TLR9 and their subsequent activation are better characterized for bacterial and viral nucleotides, the mtDNA-induced programs probably mirror these mechanisms. One potential mechanism whereby mtDNA engages TLR9 is via trafficking within mitochondrial-derived vesicles to the endosomes ( 97 ). This mechanism, or other mechanisms that traffic mtDNA, are postulated to facilitate antigen presentation given that TLR9, which resides on the ER, is recruited to the endosome-lysosome compartments ( 98 ) to undergo proteolytic cleavage and receptor activation ( 99 ). In general, the CpG motifs that function as ligands for TLR9 activation sequester in the endosome-lysosome compartment before triggering this signaling pathway. In contrast, extracellular free or microparticle-encased mtDNA activates this pathway via canonical immune cell surface PRRs or via the immunoglobulin superfamily member receptor for advanced glycation end products (RAGE), which undergoes endocytosis to initiate TLR9-dependent signaling ( 100 , 101 ). Interestingly, this RAGE-mediated inflammatory pathway regulates mitochondrial bioenergetics ( 102 ), although whether this potential mitochondrial feedback loop is triggered by mtDNA release does not appear to have been investigated.
cGAS-STING signaling. The cytosolic DNA-sensing cGAS/STING (cyclic GMP–AMP synthase linked to stimulator of interferon genes) pathway is activated in response to exogenous viral infections ( 103 ) and in response to endogenous DNA leakage associated with cancer and aging ( 104 ). The detection of cytoplasmic DNA by cGAS generates cyclic dinucleotide cGAMP as a second messenger to activate STING. The downstream phosphorylation and activation of IFN regulatory factor 3 (IRF3) transactivate expression of IFN-β and canonical IFN-stimulated genes. In parallel with NLRP3 activation and TLR9 signaling, mtDNA activates cGAS/STING and type I IFN signaling ( 2 ). An intriguing route for the escape of mtDNA during cGAS/STING activation has been identified through the OMM in the context of apoptosis, in which Bak and Bax apoptotic channels release mtDNA ( 105 , 106 ). Normally this leakage is inert because mtDNA is lysed by concurrent activation of apoptotic caspases ( 105 , 106 ). However, when apoptotic caspases are inhibited or genetically disrupted, type I IFN signaling is initiated and manifests as enhanced resistance to viral infections ( 106 ). The enigma remains of how the IMM is breached given that this membrane is not necessarily disrupted during apoptosis ( 107 ). A recent finding shows that the OMM proteins Bak and Bax generate macropores that facilitate IMM herniation with the efflux of mtDNA ( 108 ). The possibility that this mechanism is operational in initiating cGAS/STING signaling is intriguing, but as yet unproven.
Nevertheless, the molecular machinery orchestrating mtDNA-initiated cGAS/STING activation has been further characterized following genetic disruption of transcription factor A, mitochondrial (TFAM), a regulatory protein that controls mitochondria transcription and translational and mitochondrial nucleoid integrity ( 109 ). Haploinsufficient TFAM cells exhibit disrupted mtDNA repair capacity and distorted mtDNA distribution in parallel with cGAS/STING activation ( 110 ). The role of mtDNA in TFAM deficiency-induced inflammation was validated by evidence of blunted antiviral signaling following depletion of mtDNA through dideoxycytosine-mediated inhibition of mtDNA replication. In contrast, disruption of mitochondrial quality control programs (expanded upon below) in TFAM-heterozygous cells exacerbated type I IFN signaling ( 110 ). Interestingly, and possibly owing to the role of TFAM in controlling nucleoid packaging, distinct mtDNA fragments from the D-loop regulatory region were implicated in cGAS/STING activation ( 111 ). Whether the D-loop mtDNA region confers specificity for IFN signaling requires exploration. In parallel with caspase-defective apoptotic signaling, reduction in TFAM levels also primes antiviral signaling in response to pathogen invasion and viral challenge.
Mitochondrial antiviral signaling. An additional PRR family linked to mitochondria are retinoic acid–inducible gene I–like (RIG-I–like) receptors (RLRs) ( 14 ). An OMM-anchored adaptor molecule, termed the mitochondrial antiviral signaling protein (MAVS), is an essential platform for antiviral RLR signal transduction ( 61 ). The MAVS C-terminal transmembrane domain resembles other OMM protein transmembrane domains, and its depletion negated OMM MAVS localization ( 61 ). This transmembrane domain was shown to facilitate MAVS dimerization, providing an interface for direct binding and activation of downstream immune-effector molecules ( 112 ). Although mechanisms orchestrating recruitment of RLR signaling molecules to mitochondria remain uncertain, other OMM proteins function as MAVS cofactors. The TOM70 protein, which functions as an OMM pre-protein receptor for mitochondrial protein import, binds to MAVS, and the subsequent interaction between TOM70 and the chaperone HSP90 orchestrates binding of RLR effector molecules to coordinate antiviral signaling ( 113 ). Interestingly, this requirement for mitochondrial localization to activate the MAVS signalosome is exploited by viruses, as epitomized by hepatitis C virus, which encodes a serine protease that cleaves MAVS from the OMM to prevent antiviral signaling ( 114 ).
An additional mtDNA immune triggering pathway has recently been described. Here, exposure of T cells, B cells, NK cells, monocytes, and neutrophils to distinct oligonucleotides promulgates ejection of mtDNA out of the cell, which creates filamentous web-like structures that activate type I IFN signaling in adjacent cells ( 115 ). This process is insensitive to inhibition by TLR9, cGAS/STING, AIM2 inflammasome, ROS, or the permeability transition pore inhibitors ( 115 ). Although this process requires further characterization, it adds a previously unrecognized trigger for rapid immune activation. The contribution of mitochondrial components to immune regulation is shown in Figure 2.
Mitochondria-linked activation of immune pathways. Mitochondrial content and structure can play integral roles in the activation of inflammatory signaling in response to stress effects or viral infections. Extracellular mtDNA can activate NF-κB–driven inflammation via the intracellular TLR9 receptor. Alternatively, released mtDNA can initiate type I IFN signaling in adjacent immune cells. In response to mitochondrial stressors, ROS-damaged mtDNA, mitochondrial ROS, and the release of cardiolipin (CL) from the IMM can activate the NLRP3 inflammasome to promote IL-1β and IL-18 signaling. The loss of mitochondrial integrity with the extrusion of mtDNA can also activate the cGAS/STING pathway and type I IFN signaling. Finally, the mitochondrion functions as a platform for the dimerization of MAVS to activate a combination of NF-κB and IFN regulator signaling in response to viral infections.
Taken together, there is evidence of considerable overlap between mitochondrial cargo–triggered immune surveillance programs, and it is conceivable that they may function independently and/or may be coordinately activated ( 56 , 57 ). Additional studies are required to assess whether shared or distinct mechanisms of mtDNA extrusion and mitochondrial ROS signaling ( 65 , 116 ), as well as different mtDNA species, dictate engagement of these various intracellular immune surveillance sensors.
It stands to reason that if disruption of mitochondrial integrity propagates innate immune sensing and activation, enhancing mitochondrial quality control should confer immune resistance. Although an expansive discussion of these mitochondrial quality control programs is beyond the scope of this Review, emerging evidence supporting this proposition is briefly reviewed. The quality of mitochondria is controlled by numerous regulatory programs, including control of mitochondrial biogenesis ( 117 ) and dynamics ( 118 ), autophagy ( 119 ), and proteostasis ( 120 , 121 ), and the role of “nutrient-sensing” programs to sustain mitochondrial fidelity ( 122 ).
Mitochondrial dynamics. The control of mitochondrial fission and fusion plays an important role in controlling mitochondrial integrity, metabolism, and levels of mtDNA ( 118 ). Furthermore, mitochondrial dynamics facilitate turnover of damaged mitochondrial components in coordination with other housekeeping programs including autophagy and proteostasis. The molecular machinery controlling mitochondrial dynamics is well characterized ( 123 ), and their genetic disruption is linked to immune modulation. This is epitomized by observations that knockdown of the OMM fusion protein mitofusin 2 (Mfn2) abrogated viral infection–mediated NLRP3 inflammasome activation ( 124 ) and that modulation of Mfn2 levels controlled CD4 + T cell activation ( 125 ). A non-canonical function of Mfn2 similarly inhibits antiviral signaling through interaction with MAVS ( 126 ), and genetic induction of mitochondrial fission to remove damaged mtDNA abrogates type I IFN signaling in TFAM-depleted cells ( 110 ). The regulation of mitochondrial dynamics is also tightly coupled to metabolic remodeling in T cells in that genetically driving mitochondrial fusion in glycolytic effector T cells forces reprogramming toward oxidative metabolism and memory T cell function ( 127 ). Interestingly, the disruption of mitochondrial fusion in nonimmune cells can also evoke inflammation through TLR9 signaling in response to mtDNA leakage ( 128 ).
Mitochondrial autophagy. The recycling of complete or damaged components of mitochondria occurs in isolation (mitophagy) or as a component of broader recycling of cellular content (macroautophagy/autophagy) to maintain mitochondrial quality control. The molecular machinery governing both mitophagy and autophagy is well defined, and the molecular or pharmacologic disruption of these programs demonstrates their link to innate immunity ( 129 ). The genetic depletion of autophagy mediators disrupts autophagy and mitophagy with subsequent extrusion of mtDNA into the cytosol. The consequences of reduced mitochondrial clearance activate NLRP3 inflammasome and RLR type I IFN production ( 59 , 77 , 116 ). Conversely, activation of a bacterially induced inflammasome family member, NLRC4, by Pseudomonas aeruginosa that exhibited mitochondrial damage is abrogated by autophagy induction ( 130 ). The control of mitochondrial content by autophagy is also operational in sustaining the regenerative potential of hematopoietic stem cells (HSCs). Here, aging-associated reductions in autophagy increase mitochondrial content and metabolic activity, resulting in accelerated myeloid differentiation and reduced HSC renewal ( 131 ). Mitophagy is also regulated, in part, by the E3 ligase parkin ( 132 ), and mutations in the parkin-encoding gene PARK2 lead to early-onset Parkinson disease (PD) ( 133 ). Neuroinflammation is increasingly recognized in PD ( 134 ), and parkin regulates inflammatory pathways ( 135 – 137 ). However, whether the parkin-linked neuroinflammation is linked to mitophagy does not appear to have been investigated.
Sirtuins. The sirtuin family of deacylase enzymes, including three major NAD + -dependent deacetylases, play important roles in mitochondrial homeostasis ( 122 ). Briefly, the mitochondrial regulatory functions of these deacetylases can be exemplified by (a) SIRT1 contribution to nuclear regulatory control of mitochondrial biogenesis and to nutrient-sensing signal transduction ( 122 ) (b) SIRT2-mediated modulation of mitochondrial subcellular localization and control of nutrient-sensing signal transduction ( 138 , 139 ) and (c) SIRT3 deacetylation and regulation of mitochondrial metabolic and homeostatic proteins ( 140 ). Studies linking these sirtuin effects to immune modulation are limited. The plurality of SIRT1 immunoregulatory effects is not directly linked to mitochondrial function per se, but rather is linked via deacetylation and inactivation of the RelA/p65 subunit of the canonical inflammatory NF-κB transactivator and via activation of the nutrient-sensing kinase AMPK, which modulates bioenergetics and activates autophagy ( 141 ). Nevertheless, SIRT1 activation skews macrophages to their reparative polarity via increased fatty acid oxidation ( 142 ). SIRT2 deacetylates and destabilizes microtubules, resulting in disrupted intracellular organelle trafficking. As previously described, inflammasome-mediated SIRT2 inhibition optimized microtubule-directed spatial alignment of mitochondria to amplify the NLRP3 inflammasome ( 80 ). Activation of SIRT3 has been shown to deacetylate and activate mitochondrial superoxide dismutase (SOD2), which inhibits the NLRP3 inflammasome by blunting mitochondrial ROS generation and by reducing mtDNA leakage ( 6 , 56 ). The absence of SIRT3 similarly augments fatty acid–induced renal tubular inflammation via impaired mitochondrial oxidative metabolism and increased mitochondrial ROS ( 143 ).
Additional mitochondrial regulatory control programs linked to inflammation include regulation of mitochondrial biogenesis and modulation of uncoupling protein 2 (UCP2). Interestingly, IFN-γ drives a transcriptional regulatory program governing mitochondrial biogenesis ( 144 ). Induction of biogenesis augments mitochondrial respiratory function and ROS to control bacterial infections, and genetic disruption of mitochondrial biogenesis reduces ROS production with diminished murine bacterial-clearance capacity ( 144 ). At a distinct regulatory level, the inner mitochondrial membrane UCP2 modulates myeloid cell mitochondrial ROS. Here, genetic ablation of UCP2 heightens TLR4-initiated ROS-sensitive inflammatory signaling ( 145 ). Moreover, UCP2-KO mice display increased myeloid mitochondrial ROS and increased resistance against intracellular microbial infections ( 146 ).
In addition to the readily apparent disruption of mitochondrial integrity in response to intracellular viral and bacterial infections as well as traumatic cellular injury ( 54 ), the pathophysiology of autoimmune and inflammatory diseases is similarly linked to immunometabolism and to mitochondrial immunogenicity.
Systemic lupus erythematosus. The autoimmune disease systemic lupus erythematosus (SLE) is characterized by multiorgan damage from dysregulated innate and adaptive immunity. Immune cell metabolic remodeling, mitochondrial ROS, and mtDNA perturbations contribute to this pathophysiology ( 147 ). Immunometabolism signatures include evidence that SLE-linked cytokine B cell–activating factor ( 148 ) promotes B cell metabolic remodeling with enhanced glycolysis and a greater antibody generation capacity ( 33 ), and that CD4 + T cells from humans with lupus and mouse models exhibit increased oxidative phosphorylation and glycolysis compared with controls ( 149 ).
Mitochondrial ROS and mtDNA concurrently contribute to neutrophil extracellular trap (NET) formation, in which ROS-induced oxidized mtDNA incorporates into, and amplifies, NETosis. Oxidized mtDNA also activates STING-dependent type I IFN signaling to exacerbate lupus ( 150 ). The roles of immunometabolism and mitochondrial-linked immune activation in SLE is schematized in Figure 3.
Metabolic remodeling and mitochondrial mediators of SLE. SLE pathology encompasses a broad array of immunometabolism- and mitochondria-initiated events. Immunometabolic remodeling is evident in both effector T cells, driving cytokine production, and in B cells, promoting antibody generation. Activation of neutrophils by immune complexes increases mitochondrial ROS, which promotes the oxidation of mtDNA. Subsequently, the concurrent extrusion of ROS-damaged mtDNA and nuclear/genomic DNA from NETs triggers mtDNA-dependent, STING-mediated type I IFN signaling. In parallel, the ROS-damaged mtDNA can also trigger cGAS/STING signaling in immune cells in a mechanism distinct from NETosis.
Rheumatoid arthritis and other autoimmune diseases. Rheumatoid arthritis (RA) is predominantly characterized by inflammation of the joint synovium resulting in a polyarthritis. The immunogenic role of mitochondria in this pathophysiology is supported by markedly increased mitochondrial ROS production in circulating monocytes ( 151 ) and evidence of excessive mtDNA levels in the synovial fluid of RA patients ( 152 , 153 ). Furthermore, the intra-articular injection of mtDNA but not nuclear DNA evokes arthritis in animal models ( 153 ), and in humans, levels of joint mtDNA correlate with inflammatory biomarkers of disease severity ( 151 , 152 ). It has yet to be determined whether these mitochondrial triggers initiate or amplify RA.
Additional immune diseases linked to mitochondria include fibromyalgia, a disease in which cytochrome B gene mutations correlate with NLRP3 inflammasome activity ( 154 ). Additionally, antibodies directed against nuclear-encoded mitochondrial enzymes, termed anti-mitochondrial antibodies, are pathognomonic biomarkers of primary biliary cirrhosis ( 155 ).
Mitochondrial haplotypes have also been linked to inflammatory diseases such as age-related macular degeneration ( 156 ), and mtDNA-dependent inflammation is associated with cardiovascular and liver disease and linked to immunosenescence and other aging-associated degenerative conditions ( 2 ).
Given the importance of mitochondrial metabolic remodeling in immune cell polarity and activation and the impact of mitochondrial integrity on inflammation, the modulation of these mitochondrial effects is an intriguing target to regulate immune activity.
Metabolic modulators of immune function. Numerous compounds that remodel metabolic pathways could be potential targets to control immunometabolism. The use of these compounds to temper human diseases or for evaluation in murine disease models is beginning to be explored, and examples are briefly reviewed below.
Similar to findings that succinate oxidation promotes M1 polarity, dimethyl malonate (DMM) and 3-nitroproprionic acid blunted SDH activity and ameliorated E. coli infection in mice, and DMM similarly blunted LPS-induced murine sepsis ( 16 , 48 ). As described in the TCA cycle section, the TCA intermediates dimethyl fumarate and itaconate directly modulate metabolic and immune regulatory proteins via posttranslational modifications ( 43 , 44 ). In an extension of these findings, heptelidic acid, which covalently modifies the same catalytic cysteine residues on GAPDH to inhibit its activity, blunts autoimmune encephalopathy in mice ( 44 ), and the cell-permeable itaconate derivative 4-octyl itaconate counteracts LPS-induced inflammation in macrophages and in mice ( 43 ).
Furthermore, given the parallel induction of mitochondrial oxidative phosphorylation and glycolysis in SLE CD4 + T cells, a combination of metformin to block mitochondrial respiration and 2-deoxyglucose (2-DG) to reduce glucose uptake and glycolysis was tested in primary human cells and in murine lupus models ( 149 , 157 ). Given separately, metformin or 2-DG prevented murine T cell activation, and metformin alone blunted human lupus CD4 + T cell IFN-γ secretion ( 149 , 157 ), whereas the combined agents were required to reverse murine lupus ( 157 ).
Interventions to enhance mitochondrial integrity. Given that NAD + functions as a cofactor in sirtuin activation, numerous NAD + metabolic intermediate precursors have been employed to test their effects on improving mitochondrial integrity in disease ( 158 – 160 ). Nicotinamide riboside (NR), a precursor in the NAD + salvage pathway, activated SIRT3 targets and blunted the NLRP3 inflammasome in primary human peripheral blood mononuclear cells ( 6 ) and in a murine model of hepatic steatosis ( 161 ). Additionally, administration of the NAD + precursor nicotinamide mononucleotide similarly blunted pancreatic islet IL-1β production in diabetogenic mice ( 162 ). Finally, the administration of a neutralizing antibody to disrupt the CD38 NADase increased T cell NAD + levels, conferred metabolic remodeling, and augmented antitumor functioning of these effector cells ( 163 ).
TLR9 antagonists. TLR9 antagonists are being developed for clinical use, and their efficacy following mtDNA extrusion–mediated TLR9 activation was evident in disease models including protection against murine mitophagy-defective myocarditis and nonalcoholic steatohepatitis ( 94 , 96 ). TLR9 antagonists similarly blunt mtDNA-induced NETosis in primary human neutrophils ( 164 ) and reduce autoimmune symptoms in a chemically induced murine lupus ( 165 ).
Mitochondrial ROS inhibitors. The role of mitochondrial ROS in immune modulation has been described throughout this Review, and the effect of blunting mitochondrial ROS on immune modulation has been explored. Examples of this include administering the mitochondrial-targeted SOD mimetic mitoTEMPO to blunt the inflammasome in human macrophages ( 6 ) and to ameliorate type I IFN signaling in murine lupus ( 150 ).
Caloric-restricted diets. Although intermittent fasting has been found to blunt inflammatory diseases and markers of innate immune activation, the mechanisms underpinning this regulation are incompletely characterized ( 5 , 7 – 10 ). To explore the mitochondrial contribution to this biology, prolonged fasting was interrogated in human subjects and in a mouse model and showed that the NLRP3 inflammasome was blunted, in part, via SIRT3 activation with improvement in mitochondrial respiration and enhanced mitochondrial fidelity ( 6 , 56 ).
In light of the recognition of the role of mitochondria in immune modulation, numerous ongoing studies will exploit this link to explore the role of mitochondrial modulation to control immune responses.
The field of immunometabolism is in its infancy, and our understanding of how diverse immune surveillance programs integrate with mitochondrial fidelity is identifying novel targets for the modulation of innate immune activity. Further characterization of these pathways should cement the centrality of the mitochondrion in directing health and disease. An important caveat in this field is the significant differences between murine and human immune systems ( 11 , 57 , 63 , 166 , 167 ). This caveat reinforces our need to sustain the exploration of human subjects to understand the role of mitochondria in orchestrating immune function and pathology ( 168 ). Conversely, the relative accessibility of human immune cells affords the opportunity for rapid advancement in the exploration and interrogation of mitochondria to understand their role in immune function. Finally, given our expanding understanding of the role of mitochondria in immune activation, the commentary by Lewis Thomas that “mitochondria are stable and responsible lodgers, and I choose to trust them” ( 169 ) should include the proviso, “if they retain their fidelity and appropriate location within cells.”
This Review is mitochondria-centric, focusing on the central role of mitochondria the regulation of immune cell polarity, activation, and immune signaling. Immune pathway signaling has not been extensively reviewed. Additionally, space constraints have limited inclusion of all original articles supporting the concepts explored. MNS is supported by the NIH, National Heart, Lung, and Blood Institute Division of Intramural Research (ZIA-HL006047-07 and ZIA-HL005102-12)
Conflict of interest: The author has declared that no conflict of interest exists.