Redox Regulation of Homeostasis and Proteostasis in Peroxisomes

Abstract

Peroxisomes are highly dynamic intracellular organelles involved in a variety of metabolic functions essential for the metabolism of long-chain fatty acids, d-amino acids, and many polyamines. A byproduct of peroxisomal metabolism is the generation, and subsequent detoxification, of reactive oxygen and nitrogen species, particularly hydrogen peroxide (H₂O₂). Because of its relatively low reactivity (as a mild oxidant), H₂O₂ has a comparatively long intracellular half-life and a high diffusion rate, all of which makes H₂O₂ an efficient signaling molecule. Peroxisomes also have intricate connections to mitochondria, and both organelles appear to play important roles in regulating redox signaling pathways. Peroxisomal proteins are also subject to oxidative modification and inactivation by the reactive oxygen and nitrogen species they generate, but the peroxisomal LonP2 protease can selectively remove such oxidatively damaged proteins, thus prolonging the useful lifespan of the organelle. Peroxisomal homeostasis must adapt to the metabolic state of the cell, by a combination of peroxisome proliferation, the removal of excess or badly damaged organelles by autophagy (pexophagy), as well as by processes of peroxisome inheritance and motility. More recently the tumor suppressors ataxia telangiectasia mutate (ATM) and tuberous sclerosis complex (TSC), which regulate mTORC1 signaling, have been found to regulate pexophagy in response to variable levels of certain reactive oxygen and nitrogen species. It is now clear that any significant loss of peroxisome homeostasis can have devastating physiological consequences. Peroxisome dysregulation has been implicated in several metabolic diseases, and increasing evidence highlights the important role of diminished peroxisomal functions in aging processes.

I. GENERAL OVERVIEW OF PEROXISOMES

Peroxisomes were first identified using electron microscopy in 1954 by Rhodin and termed microbodies (198), which were thought at the time to be the “garbage pail” of the cell or a “fossil organelle.” Later, in 1966, de Duve and Baudhuin (44) biochemically characterized these single membrane structures as organelles with a matrix containing a high number of hydrogen peroxide (H₂O₂)-producing oxidases, and the H₂O₂ degrading enzyme, catalase. It was this high level of peroxide-consuming (peroxidase) enzymatic activity that led to the renaming of this organelle as the ‟peroxisome.ˮ

Peroxisomes are found in both prokaryotes and eukaryotes (for a review of the phylogenetic tree of organisms with peroxisomes, see Ref. 73). In mammals, for example, peroxisomes are found in all cell types except sperm cells and red blood cells. Morphologically and metabolically, peroxisomes exhibit a high degree of plasticity. The enzyme composition and structure of peroxisomes vary greatly from organism to organism and reflect the adaptability of these organelles to different environmental conditions. Their size (0.1–1.5 µM), shape (generally spherical but can be elongated, or can even form a reticulum in some cells), and even number vary depending on the cell type and environment (209, 213, 269). In mammals, the liver is the tissue with the highest density of peroxisomes, accounting for around 2% of total hepatic protein content. The functional role(s) of peroxisomes can be highly unique to specific organisms. For example, plant peroxisomes, when first discovered, were mistakenly classified as glycosomes, due to these organelle’s high content of enzymes involved in the glyoxylate cycle (91). Similarly, peroxisomes, known as the Wornin body in filamentous fungi, are crucial for cellular integrity (267).

Peroxisomes are involved in numerous catabolic and anabolic functions essential for development and overall health. Oxidation of fatty acids (and metabolism of the resultant H₂O₂) is a central function of peroxisomes in all organisms, where this organelle is present (FIGURE 1). Peroxisomes also serve essential functions in lipid metabolism and detoxification, synthesis of ether phospholipids, bile acids, and cholesterol, but the specific pathways may vary depending on the organism. Functionally, peroxisomes are integrated into the complex network of communication involving other subcellular organelles. Peroxisomes have very close physical and metabolic connections with the endoplasmic reticulum (ER) and with mitochondria (27, 211, 214), although the exact nature of any physical contact between peroxisomes and other organelles, or how information is exchanged between them, is not fully understood.

FIGURE 1.

Metabolic functions of peroxisomes in different organisms. Oxidation of fatty acids and H₂O₂ degradation by catalase is a common function that is performed by all peroxisomes irrespective of the organism type. Other functions are detailed in plants, protozoa, fungi, and animals.

Similar to mitochondria, peroxisomes are capable of autonomous replication via fission. However, in contrast to mitochondria, peroxisomes do not contain their own DNA. Over 100 nuclear genes in Homo sapiens and over 60 genes in Saccharomyces cerevisiae have been identified that encode peroxisomal proteins (207), and peroxisome biogenesis relies on the cytosolic translation of nuclear-encoded proteins that are transported to these organelles. Defects in peroxisome biogenesis and/or function lead to diseases classified as Peroxisomal Biogenesis Disorders (PBD), consisting largely of various severe developmental brain disorders that often lead to death during childhood (257). Functional defects in peroxisomes are also associated with aging and neurological disorders, such as Parkinson’s and Alzheimer’s diseases (204, 268). Peroxisome homeostasis in general needs to remain adaptable to the metabolic state of the cell, which is ensured by a combination of peroxisome proliferation, the removal of excess organelles by autophagy (pexophagy), as well as by processes of peroxisome inheritance and motility.

II. CELLULAR FUNCTIONS

Peroxisomes contain more than 50 different enzymes that perform a vast array of metabolic, homeostatic, and proteostatic functions inside the cell, which vary by cell type and under different environmental conditions (FIGURE 1). Some of their central functions in mammalian cells include the β-oxidation of very-long-chain and branched-chain fatty acids, production and decomposition of H₂O₂, and the synthesis of plasmalogens. These key catabolic and anabolic roles render peroxisomes essential for cellular homeostasis. In addition, excessive peroxisome numbers, or production of high levels of peroxisomal reactive oxygen species, may increase oxidative stress in the cell and contribute to aging and cancer development.

In mammalian cells, β-oxidation of lipids occurs in both peroxisomes and mitochondria (183). Multiple enzymes have been identified in the peroxisomal matrix that catalyze various steps in the β-oxidation of very-long-chain fatty acids, branched-chain fatty acids, prostaglandins, and eicosanoids, as well as the oxidation of d-amino acids and branched-chain fatty acids (89, 207). Peroxisomes metabolize very-long-chain and branched-chain fatty acids to smaller, eight-carbon (or less) units that are subsequently conjugated to carnitine and shuttled into mitochondria for further oxidation and energy liberation via the mitochondrial electron transport chain. Alternatively, some of the eight carbon (or less) fatty acids can serve as substrates for the generation of complex lipids within peroxisomes. However, in lower organisms, such as yeast and plants, β-oxidation of fatty acids occurs only in the peroxisome. In plants, peroxisomes perform two additional functions. 1) In germinating seeds, peroxisomes provide energy and raw materials for growth by conversion of stored fatty acids to carbohydrates (164). The special peroxisomes in which this takes place are known as glyoxysomes, and the process is called the glyoxylate cycle. 2) In leaves, peroxisomes are involved in photorespiration, which serves to metabolize a side product formed during photosynthesis (129, 146).

Recently, peroxisomes have become recognized as signaling organelles (243). Dixit and colleagues first demonstrated that the mitochondrial antiviral signaling protein (MAVS, also known as IPS-1, Cardif, or VISA) localizes to peroxisomes and plays a role in antiviral innate immunity (47) by increasing expression of type III interferon in response to diverse pathogenic stimuli (167). Peroxisomes have also been identified as important sites for cross-talk between several cell signaling pathways including mechanistic target of rapamycin (mTOR) signaling in which the tuberous sclerosis complex-2 (TSC-2) participates, damage response signaling in which ataxia telangiectasia mutated (ATM) participates, and energy responsiveness in which AMP-activated kinase (AMPK) participates (279, 280). As described later, the nexus of these signaling pathways at the peroxisome plays an important role in directed autophagy of peroxisomes (pexophagy).

III. REDOX ACTIVE ENZYMES AND REACTIVE OXYGEN AND NITROGEN SPECIES

As mentioned above, the canonical role of peroxisomes is metabolism of fatty acids to form the critical tricarboxylic acid (TCA) cycle intermediate acetyl-coenzyme A (acetyl-CoA). Even in mammals, where mitochondria are also capable of β-oxidation, peroxisomes are still essential for β-oxidation of very-long-chain and branched fatty acids that cannot be broken down in the mitochondria. As part of their normal catalytic activity, several peroxisomal enzymes produce reactive oxygen or nitrogen species as byproducts of metabolism (TABLE 1). These various oxidases include acyl-CoA oxidase, l-pipecolic acid oxidase, l-α-hydroxy acid oxidase, xanthine oxidase, urate oxidase, d-amino acid oxidase, and d-aspartate oxidase, all of which generate H₂O₂ in oxidizing their substrates (TABLE 1) (6).

Table 1.

Human peroxisomal enzymes that produce reactive oxygen or nitrogen species as byproducts of their normal catalytic activity, and the antioxidant enzymes present in cells that help to maintain redox homeostasis

Name of ROS Producing Enzyme	Protein Symbol	ROS/RNS Produced	Localization of ROS Producing Enzyme	Name of Antioxidant Enzymes Responsible for Removal of Peroxisomal ROS/RNS	Protein Symbol	Localization of Antioxidant Enzyme in the Cell
Acyl-CoA oxidase-1	ACOX1	H2O2	PO	Catalase	CAT	PO/C?
				Peroxiredoxin 5	PDX5	C/MT/PO/N
Acyl-CoA oxidase-2	ACOX2	H2O2	PO	Catalase	CAT	PO/C?
				Peroxiredoxin 5	PDX5	C/MT/PO/N
Acyl-CoA oxidase-3	ACOX3	H2O2	PO	Catalase	CAT	PO/C?
				Peroxiredoxin 5	PDX5	C/MT/PO/N
d-Amino acid oxidase	DAO	H2O2	PO	Catalase	CAT	PO/C?
				Peroxiredoxin 5	PDX5	C/MT/PO/N
d-Aspartate oxidase	DDO	H2O2	PO	Catalase	CAT	PO/C?
				Peroxiredoxin 5	PDX5	C/MT/PO/N
l-Pipecolic oxidase	PIPOX	H2O2	PO	Catalase	CAT	PO/C?
				Peroxiredoxin 5	PDX5	C/MT/PO/N
l-α-Hydroxyacid oxidase 1	HAO1	H2O2	PO	Catalase	CAT	PO/C?
				Peroxiredoxin 5	PDX5	C/MT/PO/N
l-α-Hydroxyacid oxidase 2	HAO2	H2O2	PO	Catalase	CAT	PO/C?
				Peroxiredoxin 5	PDX5	C/MT/PO/N
Polyamine oxidase	PAOX	H2O2	PO	Catalase	CAT	PO/C?
				Peroxiredoxin 5	PDX5	C/MT/PO/N
Xanthine oxidase	XDH	H2O2	PO/C/MT	Catalase	CAT	PO/C?
		NO·		Peroxiredoxin 5	PDX5	C/MT/PO/N
		O2.−		Superoxide dismutase 1	SOD1	C/MT/PO/N
Inducible nitric oxide synthase	NOS2	NO·	C/PO	Peroxiredoxin 5	PDX5	C/MT/PO/N
		ONOO–		Peroxiredoxin 5	PDX5	C/MT/PO/N

ROS, reactive oxygen species; RNS, reactive nitrogen species.

Importantly, peroxisomes contain several scavenger enzymes such as catalase, which utilize the H₂O₂ synthesized by these processes to oxidize and detoxify a variety of other substrates, including alcohols, formaldehyde, polyphenols, and formic acid. Both the superoxide anion radical (O₂^·−) and H₂O₂ are generated within peroxisomes, as well as the even more reactive hydroxyl radical (^·OH), which can be formed through metal-catalyzed Fenton or Haber-Weiss relations (33, 41, 42, 70, 187), leading to formation of various alkyl hydroperoxides (ROOH). Alternatively, O₂^·− can react with nitric oxide (NO·) to form the strongly oxidizing radical peroxynitrite (ONOO⁻), and both ^·OH and ONOO⁻ are highly damaging, redox-active species (171, 226, 273). Thus generation of a variety of reactive oxygen and nitrogen species is an inevitable consequence of the normal catalytic role of peroxisomes, and indeed, the very name peroxisome highlights the fact that this organelle not only generates, but also degrades H₂O₂, as a consequence of normal metabolism (43).

Two of the most significant sources of peroxisome-derived H₂O₂ are xanthine oxidase and acyl-CoA oxidase. Xanthine oxidase (which generates O₂^·− directly) was originally identified in the core of peroxisomes in rat livers, but was later found in the peroxisomal matrix (72). It exists in two distinct configurations, as xanthine dehydrogenase and as xanthine oxidase. Xanthine dehydrogenase couples xanthine oxidation to urate, with concomitant NAD⁺ reduction to NADH. In contrast, xanthine oxidase oxidizes hypoxanthine to xanthine and reduces oxygen to O₂^·−. The enzyme then oxidizes xanthine to uric acid, again reducing oxygen to O₂^·−. Xanthine oxidase may also generate some H₂O₂ directly, and certainly increases H₂O₂ levels via superoxide dismutase-dependent O₂^·− dismutation. Xanthine oxidase is the dominant form of the enzyme found in the peroxisomes, whereas the dehydrogenase form is primarily found in the cell cytoplasm (259). This is especially important during periods of high cellular stress, such as after ischemia-reperfusion, when excess conversion of the dehydrogenase form to the oxidase form in the cytoplasm can further increase the intracellular generation of O₂^·− and H₂O₂ (57).

Another critical oxidase within the peroxisome is acyl-CoA oxidase. During the first and rate-limiting step of peroxisomal β-oxidation of lipids, peroxisomes are dependent on two forms of the H₂O₂-generating acyl-CoA oxidases. These are palmitoyl-CoA oxidase, which oxidizes the CoA esters of long-chain fatty acids, and the branched-chain acyl-CoA called pristanoyl-CoA oxidase which breaks down branched-chain fatty acids (194, 249). Both enzymes work by desaturating acyl-CoAs of fatty acids to form 2-trans-enoyl-CoA, with the byproduct of H₂O₂ formation. Humans and rodents have been found to contain three forms of the acyl-CoA oxidase: palmitoyl-CoA oxidase (ACOX1), trihydroxycoprostanoyl (THC)-CoA oxidase (ACOX2), and pristanoyl-CoA oxidase (ACOX3) (10). Mice lacking acyl-CoA oxidase show growth retardation and increased intracellular H₂O₂ generation in early life, and go on to develop hepatocellular carcinomas, due the chronic upregulation of PPARα from lipid accumulation attributed to the loss of acyl-CoA (60). Similarly, humans lacking an enzymatically active acyl-CoA oxidase are diagnosed with the rare hereditary disease, pseudoneonatal adrenoleukodystrophy, which results in the accumulation of long-chain fatty acids and the complete absence of peroxisomes. So severe is this deficiency that many individuals diagnosed with the disease die within 4–5 years of life (258).

Although the peroxisome is correctly considered to be a major producer of O₂^·− and H₂O₂, it also contributes to the production of reactive nitrogen species. Specifically nitric oxide (NO^·), which is predominantly a consequence of nitric oxide synthase activity in plants (NOS and its inducible form, NOS2) (171), resulting from the oxidation of l-arginine to NO^· and citrulline (119). In mammalian systems, NOS has only been identified in mitochondria from rat liver (78). Secondary sources include xanthine oxidase activity and polyamines. Additionally, in the absence of its substrate, NOS has been found to produce large amounts of superoxide (231). In turn, peroxynitrite (ONOO⁻) and nitrosoglutathione (GSNO) form as a consequence of the reaction between NO^· and O₂^·− coupled with reduced glutathione (GSH) (36). Moreover, the high instability of certain nitrogenous species may trigger direct oxidative and nitrosative damage, often manifested as lipid peroxidation or protein oxidation. To counteract this problem, peroxisomes rely heavily on superoxide dismutase to prevent the accumulation of O₂^·− and minimize its reaction with NO·, thus limiting the generation of ONOO⁻ (203). In addition, catalase and glutathione peroxidase, traditional antioxidant enzymes in the peroxisome, are downregulated in the presence of NO^·. Conversely, β-oxidation is upregulated in the presence of NO^·, which may provide evidence for NO^·-mediated H₂O₂ accumulation within peroxisomes (49). At low concentrations, however, NO^· can act as a sensor of oxidative stress, specifically during high salinity and cadmium stress (35), and in some instances (specifically in plants) can help to restore catalase activity (126).

IV. HYDROGEN PEROXIDE AS A SIGNALING MOLECULE

At high concentrations, H₂O₂ can damage cellular proteins and membranes (either directly, or through iron- or copper-catalyzed reactions), and in some instances can even trigger cell death. Hence, H₂O₂ was originally solely regarded as an unwanted byproduct of cellular metabolism (197). This resulted in many studies to determine methods that would rapidly eliminate all H₂O₂ from the cell, with the paradoxical result of only increasing cellular sensitivity to oxidative stress (252). Subsequent studies revealed that cells rely on the continual generation of nanomolar to low micromolar concentrations of H₂O₂ (e.g., 10⁻⁸ to 10⁻⁷) for intracellular signaling (29). It is now accepted that, at low levels, H₂O₂ participates in cellular differentiation, migration, and proliferation and can dramatically alter gene expression profiles, hence the recent prominence of H₂O₂ as an important cell signaling molecule (66, 70, 197, 277, 278). It is also now clear that H₂O₂ plays a major role in redox-dependent mechanisms of adaptive homeostasis, the process by which signaling molecules or events cause reversible extensions of the homeostatic range to cope with transient increases in internal or external stressors (39).

Since H₂O₂ can exert very different concentration-dependent effects, for example, cytotoxicity at high concentrations and physiological signal transduction at low concentrations, its production and removal requires tight regulation within the cell. Excess accumulation of H₂O₂ alters the cellular redox state, which if not corrected will promote cellular dysfunction (202). Mitochondria are often considered to be the primary culprit for excess H₂O₂ production via leakage from the electron transport chains of these organelles into the cytosol (25, 130). However, due to the high abundance of peroxisomal oxidases that participate in lipid metabolism, peroxisomes are increasingly also viewed as prominent endogenous sources of H₂O₂ (22). This is especially apparent in the liver, with peroxisomes consuming ~20% of the available oxygen (44), resulting in an H₂O₂ production rate of 0.1–0.4 nmol·min^-1·mg protein^-1, which equates to ~35% of the total cellular H₂O₂ generation (20).

The relative stability, long lifetime and mobility, and high diffusion length of H₂O₂ all add to the its effectiveness as a signaling molecule as well as to its potential for toxicity. For example, H₂O₂ can directly modify some sulfur containing amino acids within proteins through the formation of sulfenic acids, or can form disulfide bonds via the oxidation of thiol groups (62, 69, 166) which in turn impacts the functionality of a wide array of proteins. The importance of peroxisomal H₂O₂ as a signaling molecule, was demonstrated following knockdown of catalase (271), or the overexpression of acyl-CoA oxidase (56), both of which promoted increased accumulation/generation of H₂O₂. Additionally, knockdown of catalase triggered upregulation of proteins involved in free fatty acid oxidation and mitochondrial biogenesis, such as fibronectin (97). Conversely, studies that triggered the overexpression of catalase actually increased susceptibility to oxidative stress (31).

V. DETOXIFICATION ENZYMES OF THE PEROXISOME

Peroxisomal metabolism is associated with high levels of production of reactive oxygen and nitrogen species. As noted above, if allowed to accumulate, these reactive species can lead to peroxisome enzyme inactivation and membrane damage. To minimize such potentially destructive reactions, peroxisomes appear to use a relatively simple strategy: convert all O₂^·− to H₂O₂ as efficiently as possible, and remove the H₂O₂ (almost) as rapidly as it forms. To accomplish this, peroxisomes contain both manganese superoxide dismutase (225) and a copper-zinc superoxide dismutase (108) (FIGURE 1 and TABLE 1). The high activities of these two superoxide dismutases achieve the extremely efficient dismutation of O₂^·− to H₂O₂ and a combination of peroxiredoxin 5, catalase, and glutathione peroxidase serve to eliminate (most) H₂O₂ safely. However, in cells with especially active peroxisomes, excess escape of even a small percentage of these reactive species can cause immediate damage to peroxisomal proteins, peroxidation of the organelle’s lipid membrane, and longer range effects if released into the cytosol. Once in the cytosol, reactive oxygen and nitrogen species may cause protein oxidation or damage ribosomes or RNAs, and can also disrupt the cellular oxidation-reduction (redox) state (51), and may even represent an underlying cause of chronic disease, including atherosclerosis, cancer, and type II diabetes (202).

A. Peroxiredoxins

Peroxiredoxins are a superfamily of thiol-dependent peroxidases that break down H₂O₂, various hydroperoxides (118), and peroxynitrites (67) into water, alcohols, and nitrites, respectively (TABLE 1). Peroxiredoxins rely on cysteine-dependent peroxidase activity for substrate breakdown, which is not only a shared trait among the six mammalian peroxiredoxins, but across phyla (265). Peroxiredoxin 5 is found within mitochondria, the cytosol, peroxisomes, and nucleus. It exists as both a short and a long isoform, with the short isoform being found in peroxisomes (253). The amount of peroxiredoxin 5 within peroxisomes was found to be similar to that of other peroxiredoxins, ranging from 0.2 to 1.5 µg/mg in rat tissues (218).

Similar to other peroxiredoxins, peroxiredoxin 5 relies on the interaction of two cysteines to break down H₂O₂ and various other hydroperoxides (118). Typically, most peroxiredoxins function by the combined activity of two monomers. In this reaction, one monomer’s cysteine is oxidized during the peroxidase reaction, and the second monomer then binds to form a stable intermolecular disulfide bond, which can subsequently be reduced by an external reductant. Unlike other mammalian peroxiredoxins, however, peroxiredoxin 5 has an atypical 2-cysteine peroxiredoxin structure (218). The 2-cysteine structure utilizes two internal cysteines in the same protein, which allows for an intramolecular disulfide bond to be formed (218). As a result, peroxiredoxin 5 shows a lower reaction rate with H₂O₂ (~10⁵ M⁻¹·s⁻¹) in comparison to the much faster breakdown of ONOO⁻ (~10⁷ M⁻¹·s⁻¹) and ROOH (~10⁶ M⁻¹·s⁻¹) (245). Thus the atypical 2-cysteine structure found in peroxiredoxin 5 allows the enzyme to be more resistant to inactivation by H₂O₂ compared with other peroxiredoxins. This is crucial in the peroxisome where excess H₂O₂ must be quickly neutralized, to limit organelle and cellular damage. Indeed, various yeast studies have identified peroxisomal peroxiredoxin 5 as a vital cytoprotective enzyme (3, 94).

B. Catalase

Another critical peroxisome protein to combat excess H₂O₂ accumulation is catalase ( FIGURE 1 and TABLE 1) . Catalase enters peroxisomes as a monomer, but it subsequently self-associates to become the catalytically active homotetrameric enzyme, which contains four binding positions for NADPH. Active peroxisomal catalase relies on a heme molecule in its active site to perform two critical reactions: the conversion of H₂O₂ into oxygen and water (catalytic reaction) and the oxidation of hydrogen donors, such as methanol and formic acid through the decomposition of peroxide (peroxidatic reaction) (1). The concentration of hydrogen donors present and the rate of H₂O₂ generation dictate which reaction will take precedence. The catalytic reaction is a two-part process that occurs at low concentrations of H₂O₂ (ranging from 0.01 to 0.05 M). It relies on the heme Fe³⁺ to generate a water molecule from the reduction of H₂O₂ to form a covalent Fe⁴⁺ = O oxyferryl species (compound I). The subsequent second step is a very rapid, first-order catalytic reaction (K = 10⁷ l·mol⁻¹·s⁻¹) that utilizes a second H₂O₂ molecule to oxidize compound I, thus forming oxygen and releasing the oxyferryl species as water. In contrast, the peroxidatic reaction is much slower in its formation of complex I between catalase and organic peroxides (K = 10²–10³ l·mol⁻¹·s⁻¹) (1, 16).

In vitro studies have shown that catalase does not follow Michaelis Menten kinetics (82). Thus even small amounts of H₂O₂ should be efficiently removed by catalase, and the enzyme should not become saturated with substrate, even at extremely high concentrations of the oxidant (up to 5 M H₂O₂), theoretically allowing for the continual breakdown of H₂O₂. Thus one might expect that catalase would be the primary mechanism for H₂O₂ breakdown, especially under homeostatic conditions involving rather low H₂O₂ levels. It has been suggested, however, that even low amounts of endogenous H₂O₂ may be poorly degraded by catalase, despite its rapid reaction rate. This is, in part, because catalase’s active site requires the interaction of two H₂O₂ molecules and, although this prevents saturation, is difficult to achieve at low H₂O₂ concentrations. As a result, catalase actually may not play the major role in eliminating low levels of H₂O₂ from the peroxisome (88). Additionally, catalase is highly sensitive to oxidation and can be inactivated with increasing amounts of H₂O₂, which can occur at H₂O₂ levels above 0.1 M (28). These factors all contribute to the possibility of H₂O₂ escaping into the cytosol, acting as a peroxisome redox signal at low concentrations, or causing oxidative damage in higher amounts. Indeed, early studies in rat liver peroxisome estimated that ~20–60% of H₂O₂ synthesized, even under homeostatic conditions, escapes into the surrounding cytosolic environment (15).

C. Glutathione Peroxidase

Glutathione peroxidase was initially discovered in rat liver peroxisomes as a complementary enzyme for the breakdown of peroxides and the removal of H₂O₂ during periods of oxidative stress (224). Glutathione peroxidase converts H₂O₂ into water via the oxidation of two reduced glutathione molecules (GSH) to form a disulfide-bridged dimeric GSSG form. The GSSG produced is then reduced back to two GSH molecules by glutathione reductase, completing the cycle and ‟rechargingˮ the system. Despite its dependence on reduced GSH, glutathione peroxidase is very efficient at removing H₂O₂, and peroxisomal GSH levels do not appear to be rate-limiting. Some indirect evidence hints that GSH diffuses into the peroxisome via the nonselective pore PXMP2 (199), but further studies will be needed to confirm this finding.

The limitations noted above for catalase’s efficiency would point to glutathione peroxidase as the main mechanism for H₂O₂ breakdown in peroxisomes, but absolute comparative quantification is difficult under physiological conditions. During periods of acute elevated endogenous H₂O₂ generation, however, such as ischemia-reperfusion injury (281), hypolipidemic drug treatment (46), or endotoxemia (105), catalase clearly undergoes inactivation. Under such conditions, it seems likely that glutathione peroxidase becomes the primary mechanism for H₂O₂ breakdown.

VI. PEROXISOMAL PROTEIN QUALITY CONTROL

A unique safety measure for protein quality control within peroxisomes is the import of fully folded proteins (254). This helps to eliminate misfolding of native peptides in the highly oxidizing environment of the peroxisome matrix. Peroxisome-destined proteins are synthesized in the cytosol and folded by cytosolic chaperones, such as Hsp70, whose interaction is also believed to enhance peroxisome protein import (256). Correct cytosolic folding is also important for recognition by the peroxisome import machinery, as defects in protein folding will prevent target protein recognition by import receptors (68). However, even with this additional precaution, peroxisomal proteins are still susceptible to damage. Unique to the peroxisome, enzymes that function solely as chaperones have not been identified within the organelle (122). Rather, the peroxisome appears to rely on enzymes that fulfill the dual role of chaperone and protease (FIGURE 2).

FIGURE 2.

Quality control checks for peroxisomal proteins. Peroxisome-destined proteins have various checks, before and after import, to ensure proper protein folding. In mammalian cells, the first check relies on newly synthesized proteins from ribosomes containing the correct sequence to ensure that proper configuration is achieved. This is further assisted during the second quality control check, with correct folding assisted by the heat shock protein 70 (HSP70). If the newly synthesized protein is incorrectly folded, it is targeted for degradation by the 26S Proteasome while still in the cytoplasm. The third check relies on the presence of the correct peroxisome targeting signal (PTS). Without the correct PTS signal, PEX5 is unable to recognize the peroxisome-targeted protein, resulting in degradation. Upon binding to PEX5, fully folded peroxisome proteins are imported into the peroxisome. The fourth check relies on the LonP2 enzyme, which can act as either a chaperone or a protease. Due to the highly oxidizing environment, peroxisomal proteins are prone to either lose their configuration or become oxidized (or both). If the protein loses proper folding, LonP2 can assist in helping to refold the protein. However, if the protein becomes oxidized, LonP2 will degrade the damaged protein.

It is well established that intracellular proteins undergo oxidation in vivo and that the ability to remove them by proteolytic digestion plays a vital role in preventing aggregation and cross-linking, and maintaining cellular homeostasis (38, 85, 228). In the cytoplasm, ER, and nucleus, this vital proteolytic function is carried out by the 20S Proteasome and the Immunoproteasome, both often activated by the 11S (or Pa28) Proteasome regulator; the nuclear form of the 20S Proteasome is also activated by poly-ADP ribose polymerase (PARP) (40, 178, 179, 223, 228 , 246). Mitochondria do not have an endogenous proteasome, however, and the selective degradation of oxidatively damaged proteins (to maintain mitochondrial protein quality control) is instead conducted by the ATP-stimulated LonP1 protease (17–19, 162, 163). Like mitochondria, peroxisomes also do not contain any form of proteasome, but they do contain a form of the ATP-stimulated Lon protease called LonP2 (111, 184). LonP2 is localized in the peroxisomal matrix, where it acts as both a protease and a chaperone. Also like LonP1, which prevents the aggregation of oxidized mitochondrial matrix proteins such as the critical Krebs cycle enzyme aconitase (17), LonP2 preferentially degrades misfolded catalase (9).

The ability of the peroxisome to quickly remove damaged proteins is crucial, especially in the highly oxidizing environment (due to fatty acid oxidation byproducts) of the peroxisome matrix. The close proximity of peroxisome matrix proteins to the sites of oxidant generation makes them highly susceptible to oxidative damage; hence, LonP2 is thought to play a vital role in maintaining proteostasis by acting to either refold destabilized proteins, or to proteolytically degrade proteins that have become too badly damaged to repair. Both functions help to prevent protein aggregation and covalent cross-linking (9). LonP2 also plays a regulatory role in the process of fatty acid β-oxidation within peroxisomes. Upon removal of peroxisome proliferation signals, the peroxisome relies on LonP2 to quickly degrade unnecessary fatty acid oxidases to prevent excessive fatty acid breakdown and decrease reactive oxygen species generation (275). Additionally, LonP2 was recently identified as the enzyme responsible for degrading the trypsin-domain containing 1 protease (Tysnd1) (169, 184). Tysnd1 processes some of the amino-terminal presequences from PTS2-containing proteins, and in rare cases, those containing PTS1 (124). Plants also contain an enzyme equivalent to Tsynd1, called DEG15 (215). Upon downregulation of Tysnd1, it undergoes self-cleavage, and the byproducts are then degraded by LonP2 (168). Conversely, deletion of LonP2 has been found to be detrimental to peroxisome homeostasis and cell survival (4). Taken together, these findings indicate the importance of LonP2 in peroxisome proteostasis and homeostasis.

VII. PEROXISOME BIOGENESIS

Peroxisome biogenesis is a process that consists of three components: 1) the formation of the peroxisome membrane, 2) import of fully folded peroxisomal proteins, and 3) organelle proliferation (54). To accomplish these goals, peroxisomes rely on peroxins for protein import: a unidirectional process starting with recognition of the peroxisome-targeted protein, migration to the peroxisomal membrane, followed by pore formation, and translocation of the cargo into the peroxisome matrix. There are currently over 30 known peroxins in yeast (a number that is increasing as research progresses) and 13 orthologs that have been identified in humans (157), which are involved in various aspects of peroxisome biogenesis and division, as well as matrix and membrane protein import from the cytosol (TABLE 2). Peroxisomal proteins are translated on free ribosomes in the cytoplasm before their import. The vast majority of proteins destined for the peroxisome contain specific peroxisomal targeting signals (PTS) located either at their carboxy (PTS1) or amino (PTS2) terminus (83, 233). In S. cerevisiae, Pex5 recognizes proteins containing the PTS1 signal and directly targets them to the peroxisomal membrane, where it interacts with its docking partner, Pex14, for successful import. Proteins containing the PTS2 signal are recognized by Pex7 and the cofactors Pex18 and Pex21. Once at the peroxisomal membrane, Pex7, along with the docking complex proteins Pex13 and Pex14, ensures protein import (55, 157) (FIGURE 3). In the mammalian system too, PEX5 and PEX7 are known to function as import receptors for delivery of proteins to the peroxisome. PEX5 recognizes PTS1-containing proteins, whereas PTS2-containing proteins require both PEX7 and PEX5 for import in mammalian systems (21, 50).

Table 2.

Functions, distribution, and disease associations of the peroxin proteins

					Yeast
Peroxins	Function	Disease (PBD)	Human	Mouse	Sc	Yl	Pp
PEX1	Receptor export (recycling), AAA-type ATPase, mediate fusion of preperoxisomal vesicle in the de novo formation of peroxisomes	IRD, NALD, ZSS	Yes	Yes	Yes	Yes	Yes
PEX2	Receptor export (ubiquitylation), form the RING finger complex with PEX10 and PEX12	ZS, mild ZSS	Yes	Yes	Yes	Yes	Yes
PEX3	Receptor docking, form a complex required for de novo generation of peroxisome	ZS, mild IRD	Yes	Yes	Yes	Yes	Yes
Pex4	Receptor export (ubiquitylation), ubiquitin conjugating enzyme		No	No	Yes	Yes	Yes
PEX5	PTS1 and PTS2 cargo shuttling receptor, cargo translocating channel	NALD, ZSS	Yes	Yes	Yes	Yes	Yes
PEX6	Receptor export (recycling), AAA-type ATPase, mediate fusion of preperoxisomal vesicle in the de novo formation of peroxisome	NALD, ZSS	Yes	Yes	Yes	Yes	Yes
PEX7	PTS2 cargo, shuttling receptor	RCDP	Yes	Yes	Yes	Yes	Yes
Pex8	Docking and export complex conjugation, importomer assembly		No	No	Yes	Yes	Yes
PEX10	Receptor export (ubiquitylation), form the RING finger complex with PEX2 and PEX12	NALD, ZSS	Yes	Yes	Yes	Yes	Yes
PEX11	Membrane elongation, recruits the fission machinery	Mild ZSS	Yes	Yes	Yes	Yes	Yes
PEX12	Receptor export (ubiquitylation), form the RING finger complex with PEX2 and PEX10	IRD, NALD, ZSS	Yes	Yes	Yes	Yes	Yes
PEX13	Receptor docking complex	NALD, ZSS	Yes	Yes	Yes	Yes	Yes
PEX14	Receptor docking complex	ZS	Yes	Yes	Yes	Yes	Yes
Pex15	Receptor export (recycling), membrane receptor for Pex6		No	No	Yes	No	No
PEX16	Recruits PMPs in the ER	ZSS	Yes	Yes	No	Yes	No
Pex17	Receptor docking complex		No	No	Yes	No	Yes
PEX18	PTS2 cargo, co-receptor		No	No	Yes	No	No
PEX19	Direct targeting of PMPs, soluble chaperones and receptor, form a complex required for de novo generation of peroxisomes	ZS	Yes	Yes	Yes	Yes	Yes
Pex20	PTS2 cargo, co-receptor		No	No	No	Yes	Yes
Pex21	PTS2 cargo, co-receptor		No	No	Yes	No	No
Pex22	Receptor export (ubiquitylation), Pex4 anchor		No	No	Yes	Yes	Yes
PEX23	Form a complex with reticulon homology domain containing proteins and establish peroxisome contact sites at ER subdomains, contain dysferlin domains, regulate the de novo generation of peroxisome		Yes	No	Yes	Yes	Yes
Pex24	Form a complex with reticulon homology domain containing proteins and establish peroxisome contact sites at ER subdomains, contain dysferlin domains		No	No	Yes	Yes	Yes
Pex25	Membrane elongation and remodeling		No	No	Yes	Yes	Yes
PEX26	Receptor export (recycling), membrane receptor for Pex1 and Pex6	ZSS, NALD, IRD	Yes	Yes	No	Yes	No
Pex27	Negatively affect fission		No	No	Yes	No	No
Pex28	Form a complex with reticulon homology domain containing proteins and establish peroxisome contact sites at ER subdomains, contain dysferlin domains		No	No	Yes	Yes	Yes
Pex29	Form a complex with reticulon homology domain containing proteins and establish peroxisome contact sites at ER subdomains, contain a dysferlin domain		No	No	Yes	Yes	Yes
Pex30	Form a complex with reticulon homology domain containing proteins and establish peroxisome contact sites at ER subdomains, contain a dysferlin domain, regulate the de novo generation of peroxisomes		No	No	Yes	Yes	Yes
Pex31	Contain a dysferlin domain		No	No	Yes	No	No
Pex32	Form a complex with reticulon homology domain containing proteins and establish peroxisome contact sites at ER subdomains, also contain a dysferlin domain		No	No	Yes	No	No
Pex33	Receptor docking complex, found in fungus Neurospora crassa		No	No	No	No	No
Pex34	Positive regulator of fission		No	No	Yes	No	No

PEX, peroxin; IRD, infantile refsum disease; NALD, neonatal adrenoleukodystrophy; PBD, peroxisomal biogenesis disorder; Pp, Pichia pastoris; RCDP, rhizomelic chondrodysplasia punctata; Sc, Saccharomyces cerevisiae; Yl, Yarrowia lipolytica; ZS, Zellweger syndrome; ZSS, ZS spectrum.

FIGURE 3.

Import of peroxisomal proteins. Newly synthesized peroxisome-targeted proteins are directed to the peroxisome predominantly due to recognition of the peroxisome targeting signal 1 (C’-PTS1), a tripeptide, located on the carboxy terminal of peroxisome-targeted proteins. In yeast, to ensure correct protein folding, various heat shock proteins, such as the heat shock 70 (HSP70) chaperone protein, bind to the newly synthesized peptide and assist in folding. The newly folded protein is recognized in the cytosol by the soluble receptor Pex5. In turn, Pex5 binds to the protein and brings the protein into the peroxisome, where it docks with Pex14, before being recycled back to the cytosol. A small number of peroxisomal proteins contain a conserved peptide recognition sequence, peroxisome targeting signal 2 (PTS2-N’), including catalase, located on the amino terminus. Again, various heat shock chaperones bind to the newly synthesized protein with evidence indicating the interaction of Hsc70 and HSP40 with PTS2-containing peptide, to ensure proper folding. Pex7, which recognizes the PTS2-containing sequence, along with the cofactors Pex18 and Pex21, brings the protein to the peroxisome, where it interacts with the docking complex, comprised of Pex13 and Pex14. Upon import, the PTS2 sequence is cleaved off.

Usually, ubiquitinylation of peroxisomal proteins occurs to maintain the peroxiosmal import cycle. Several yeast peroxins (Pex4, Pex5, Pex7, and Pex20) are ubiquitinylated during the peroxisomal protein import cycle (139). In S. cerevisiae, ScPex5 ubiquitinylation signals for both recycling and degradation. It is not clear how ubiquitinylation directs both recycling of peroxins as well as a signaling for degradation, and this is an area that warrants further investigation. Under normal physiological conditions, mono-ubiquitinylation of ScPex5 on C6 (in mammals at C11) by the E2 enzyme, Pex4, and the E3 ligases, Pex12 and Pex10, is required for receptor recycling to the cytosol. This receptor recycling process also requires the AAA-ATPases, Pex1 and Pex6, to export Pex5 in an ATP-dependent fashion, from peroxisomes to the cytosol, where Pex5 is de-ubiquitinylated by Ubp15 to enable another round of import. Malfunction of Pex5 recycling induces RADAR (receptor accumulation and degradation in the absence of recycling), wherein Pex5 becomes poly-ubiquitinylated on lysines (Lys18 and Lys24 of ScPex5) by the E2 enzymes, Ubc4 or Ubc5, and the E3 ligase, Pex2. Poly-ubiquitinylated Pex5 is then extracted from the peroxisomal membrane and degraded via the ubiquitin-26S proteasome system (180, 181).

Mechanisms that may participate in peroxisome biogenesis have been a topic of debate for many years (123). Development of new technology, especially live cell imaging with fluorescent reporters, has revealed that two pathways exist for peroxisome biogenesis: 1) “de novo” synthesis from the ER, and 2) growth and fission of preexisting peroxisomes (FIGURE 4). Both de novo and preexisting peroxisomes grow by importing new proteins and lipids formed in the cytosol at the ER (154).

FIGURE 4.

Peroxisome biogenesis. Details of peroxisome generation by de novo synthesis, and by growth and fission, are shown here. In the de novo pathway, peroxisomes are formed via budding from the endoplasmic reticulum (ER) and pairwise heterotypic fusion of two vesicles. Initiation begins by the peroxisomal membrane proteins (PMPs) entering into the ER via Sec61 or GET3 and traffic through the ER to get into the vesicles. One of the vesicles contains a set of RING finger proteins (PEX2, PEX10, and PEX12), while the other vesicle contains a set of docking group proteins. Following heterotypic fusion, both vesicles merge and form an import complex group. PEX1 and PEX6 are found on separate preperoxisomal vesicles and are necessary for heterotypic fusion. Once peroxisomes are formed, they grow by import of matrix and membrane-targeted proteins via the import receptor PEX5/PEX7 and PEX19, respectively, to form mature peroxisomes. In turn, mature peroxisomes grow by fission which begins with membrane remodeling and elongation by PEX11. The elongated extension grows and dynamin-related proteins (DRPs), which are also involved in fission of mitochondria, including the Fis1 (or DLP1; dynamin-like proteins in mammalian cells) are recruited from the cytosol to help with the fission and formation of new peroxisomes.

A. Biogenesis of Peroxisome via the De Novo Pathway

This pathway was first discovered in Yarrowia lipolytica (239); however, the details of this mechanism were unraveled in Saccharomyces cerevisiae only recently. The de novo synthesis of peroxisomes begins when two forms of pre-peroxisomal vesicle buds are made by the ER, which bud off and subsequently fuse with each other to form the “mature peroxisome” (2, 92, 235, 247). These two classes of pre-peroxisomal vesicles, each containing different subsets of peroxisomal proteins, fuse (heterotypic fusion) with each other to form mature peroxisomes (247). This de novo pathway occurs via the following steps.

1. Incorporation of peroxins/PMPs in the ER membrane through Sec 61/GET

In this step, peroxisome membrane proteins (PMPs) are translated in the cytosol on free ribosomes or on the ER-associated ribosomes, and incorporated posttranslationally (Pex2, Pex10, Pex12, Pex4 and others) or cotranslationally (Pex3 and Pex16) into the ER membrane. The ER-translocon Sec61 plays a role in these PMPs incorporation process (147). Furthermore, the GET complex mediates the insertion of the tail anchor proteins such as Pex15 (yeast) and PEX26 (mammals) in ER (216). In yeast, the insertion of tail anchor protein Pex15 in the ER is GET-dependent, whereas in mammals, tail anchor protein PEX26 insertion is GET-independent (23).

2. Intra-ER sorting of peroxins/PMPs

A survey of the literature reveals that specific domains such as “pre-peroxisomal domain” or “templates” or “peroxisomal ER (pER)” exist in the ER where PMPs are localized before targeting to the nascent peroxisome (77, 92, 234, 235). In mammalian cells, such specialized ER domains nucleate the formation of peroxisomes by acquiring other PMPs and matrix proteins, while still attached to the ER lamellae (2, 261). In the process of intra-ER sorting, the two subcomplexes of the import machinery, the docking complex and ring ubiquitnylation complex, are distributed on two separate pre-peroxisomal vesicles.

3. Heterotypic fusion of vesicles and formation of peroxisome

Once pre-peroxisomal vesicles are loaded with PMPs, detachment of these vesicles from the ER occurs. After detachment, these vesicles fuse to each other in heterotypic fashion to form an import-competent peroxisome. These fusion events are selective and specific inside the cytosol, although how specificity is obtained during the formation and fusion of these vesicles is not fully known. However, some studies have shown that Pex1 and Pex6 belong to the class of AAA-type ATPase peroxins, which play a major role in the heterotypic fusion process (240, 247). Such observations led to the hypothesis that selective enrichment of Pex1 and Pex6, depending on the class of vesicles involved, provides specificity for the fusion events.

B. Biogenesis of Peroxisomes via the Growth and Fission Pathway

Mature and metabolically active peroxisomes replicate autonomously by fission. Two proteins, Pex11 and dynamin-related proteins [DRPs; also known as dynamin-like proteins (DLPs) in mammals], are well characterized for their roles in peroxisome fission, and DRPs are also known to play a major role in mitochondrial fission. The molecular mechanism of peroxisomal fission is similar in both yeast and mammals. The fission process starts with activation of Pex11, which mediates elongation and tubulation of the peroxisomal membrane. The membrane-anchored DRP-interacting proteins are subsequently enriched on the elongated membrane. After elongation, the membrane is constricted and DRPs are recruited from the cytosol by DRP-interacting proteins to promote division to form new peroxisomes. The regulation of these molecular events is not well characterized; however, a report from Knoblach and Rachubinski (117) suggests that phosphorylation of Pex11 is required to activate peroxisome proliferator protein Pex11 and regulate the number and size of peroxisomes. Furthermore, Marshall et al. (145) reported that Pex11 is a redox-sensitive protein, which exists as a monomer or dimer depending on the levels of various reactive oxygen/nitrogen species, with the potential to regulate peroxisome division.

VIII. PEROXISOME PROLIFERATION

The peroxisome’s role as a redox mediator becomes especially evident during H₂O₂ dysregulation and peroxisomal proliferation, frequently experienced in various metabolic diseases and in aging. Loss of peroxisome redox balance occurs upon sustained peroxisome proliferation and has been well-characterized in the liver. This is in part because of the large number (350–400 per cell) and size of peroxisomes in hepatocytes compared with other cell types (8).

Under homeostatic conditions, peroxisomes represent only a small percentage (~2%) of the cytosol. Yet in rodent studies, when peroxisomes were stimulated to proliferate, their numbers actually expanded to occupy as much as 25% of the cytosol (59, 149). The peroxisome proliferation response has been divided into two phases: immediate and sustained. The short-term increase in peroxisome number leads to a 30-fold increase in enzymatic activity, necessary for β-oxidation (primarily from increased transcription of fatty acyl-CoA oxidase). In contrast, sustained, long-term activation is carcinogenic in rodent models, with tumor prevalence predicated to be nearly 100%.

The three members of the peroxisome proliferator-activated receptors (PPAR) are part of the nuclear receptor family of transcription factors that collectively activate various genes involved in lipid homeostasis (208). The peroxisome proliferator-activated receptor alpha (PPARα), was first discovered for its role in promoting fatty acid breakdown and has been dubbed the master regulator of β-oxidation (189). During transient conditions, the rapid upregulation of PPARα is necessary for the conversion of excess long-chain fatty acids into substrates used by the mitochondria. This is highly relevant during the opposing conditions of the fed versus fasted states. During the fed state, once glycogen stores are filled, lipogenesis occurs. Conversely, during fasting, as fatty acids are liberated from the adipose tissue, they interact and possibly activate PPARα to promote their oxidation into ketone bodies, such as acetoacetate and β-hydroxybutyrate (109, 110, 115). In turn, ketone bodies are taken up by the cells and converted into the usable form of acetyl CoA for the TCA cycle. The importance of PPARα during fasting is evident in PPARα-null mice, which show a defect in fatty acid oxidation, resulting in the elevation of plasma free fatty acids, hypoketonemia, and hypothermia (134).

The hyperactivation of PPARα, however, causes heightened and sustained activation of the basal β-oxidation rate. This increases the catabolism of fatty acids and elevates H₂O₂ generation by several enzymes, including acyl-CoA oxidase-1 (ACOX1) and a cytochrome P-450 family protein (CYP4A), at the expense of downregulating H₂O₂ removing enzymes such as catalase and glutathione peroxidase. As a consequence, excess H₂O₂ production, coupled with decreased removal, promotes chronic oxidative stress as evidenced by increased lipid peroxidation and the accumulation of lipofuscin in liver tissue (127). Studies have also shown that chronic peroxisome proliferation activates proinflammatory cytokines and blocks apoptosis (104, 177).

To date, the best studied chronic activators of PPARα are the fibrate class of hypolipidemic drugs. Though relatively weak activators of PPARα at low concentrations, these drugs have been found to increase the expression of peroxisome lipid metabolizing enzymes at pharmacologically relevant levels. Rodents appear to be especially sensitive to the detrimental consequences of sustained exposure to activators of PPARα (195). Specifically, in murine models, treatment with potent lipid-reducing drugs (Wy-14,643, ciprofibrate, or nafenopin) induces liver tumors in 100% of the animals treated with these drugs for sustained periods of time (50–60 wk) (192, 193). Less potent drugs, such as clofibrate, trigger tumorgenesis in ~70% of the animals with sustained treatment (70–104 wk) (7, 191). Of note, these types of drugs form a unique carcinogenic phenotype because they do not cause genotoxicity or mutagenicity (260), or activate the classical liver carcinogenic marks such as γ-glutamyl transpeptidase (190). Long-term studies in humans exposed to these drugs, however, have failed to show increased liver toxicity or tumorigenesis. As the majority of these drugs are prescribed to treat chronic diseases, humans are predicted to have sustained exposures over years to decades (81). Long-term epidemiological studies found that individuals consuming fibrate drugs were at no greater risk for developing carcinomas (80), nor has there been a marked increase in peroxisome proliferation in patients on sustained hypolipederimic drug treatment. Thus the vastly different response variation between rodents and humans may be explained by several species-specific factors, including basal differences in PPARα expression, as human livers have only 1/10 the amount of functional PPARα as found in mouse livers (173).

IX. PEROXISOME DEGRADATION (PEXOPHAGY)

The half-life of mammalian peroxisomes is 1.3–2.2 days (96, 133, 185), with the balance between formation and degradation essential to the maintenance of peroxisome, redox, and cellular homeostasis. The mode of selective degradation of excess peroxisomes in mammalian cells occurs via three main, independent, degradation systems: the LonP2 protease system, 15-lipoxygenase-mediated membrane disruption, and the selective autophagy-related system (pexophagy). Previous studies using Atg7 conditional knockout mice revealed that 70–80% of liver peroxisomes are degraded via pexophagy, while the remaining 20–30% are degraded by the peroxisome Lon protease and 15-lipoxygenase systems (274). Disruption of peroxisome homeostasis, including peroxisome degradation defects, underlies several human diseases (220, 238).

Autophagy is a self-digestion process by which eukaryotic cells degrade and recycle aggregation-prone proteins, macromolecules, and organelles. Pexophagy prevents the accumulation of functionally compromised peroxisomes, maintains redox balance by removing excess or damaged peroxisomes, and protects against renal damage in human and mouse vascular endothelial cells exposed to endotoxin and lipopolysaccharides, respectively (250, 251). Genetic screening has identified more than 40 ATG (autophagy-related) genes required for autophagy in yeast, and the list is still increasing (93, 116). There are three main types of autophagy: macroautophagy, microautophagy, and chaperone-mediated autophagy. During macroautophagy, which is the primary pathway for directed degradation of peroxisomes, also known as pexophagy (FIGURE 5), cytoplasmic contents are sequestered in double-membrane-bound vesicles called autophagosomes and delivered to lysosomes for degradation (135, 150). In microautophagy, lysosomes engulf the cytoplasmic material by inward invagination of the lysosomal membrane, while chaperone-mediated autophagy is facilitated by HSC70, co-chaperones, and the lysosomal-associated membrane protein type 2A.

FIGURE 5.

Three mechanisms of pexophagy in mammalian cells. Three different forms of ubiquitin (Ub)-dependent pexophagy have been shown in mammals. 1) First, the ubiquitinylation of the overexpressed peroxisomal membrane-associated protein, PEX3, serves as a signal for the adaptor complex NBR1/p62, causing an induction of pexophagy. 2) The second method of pexophagy is triggered by high amounts of reactive oxygen and nitrogen species (ROS, RNS), which cause the activation of the ATM-TSC2 signaling node, which resides on the peroxisome membrane. ATM performs two functions: a) activation of the AMPK-TSC2 signaling node and the suppression of the mTORC1, together inducing autophagy via the activation of ULK1. b) ATM phosphorylates PEX5 at S141 position, which serves as a signal for the Ub E3 ligase (PEX2, PEX10, and PEX12) and the ubiquitinylation at K209. Following ubiquitinylation, the autophagic receptor/adaptor p62 is activated, resulting in the induction of pexophagy by binding the targeted peroxisome to a phagophore. 3) The third pathway begins by the mono-ubiquitinylation of the peroxisome matrix protein import receptor, PEX5, on Cys-11 and blocks the recycling of this mono-ubiquitinylated PEX5 (due to the failure of the cargo delivery into the matrix of the peroxisome) from the peroxisomal cytosol via the receptor recycling complex (AAA-ATPases, PEX1, and PEX6 anchored at peroxisome membrane via PEX26), which serves as a signal for pexophagy.

In mammals, an important cell signaling pathway involved in regulation of autophagy engages the serine/threonine kinase mTOR, the kinase component of the TORC1 complex. mTORC1 is a negative regulator of autophagy that phosphorylates and inhibits ULK1, a key kinase involved in autophagy initiation (112). The tuberous sclerosis complex 1 and 2 (TSC1 and TSC2) tumor suppressors, which form the active tuberous sclerosis complex (TSC) heterodimer, are negative regulators of mTORC1. Both TSC2 and mTORC1 are regulated by upstream kinases, including AMPK and AKT (26). AMPK regulates several metabolic processes and activates TSC to repress mTORC1 under conditions of energy stress and can also directly repress mTORC1 signaling (219, 221).

The kinase ULK1 binds the ATG13 and FIP200 proteins to form the ‟pre-initiation complexˮ for autophagy (75). This pre-initiation complex induces autophagy by phosphorylating Beclin-1 and activating VPS34 lipid kinase (initiation complex-Beclin1, ATG 14L, VPS34, and VPS 15 also called class III PI3K complex) (200). The class III PI3K complex generates phosphatidylinositol 3-phosphate (PI3P) at the site of nucleation of the isolation membrane (phagophore), leading to the recruitment of PI3P-binding proteins, and proteins involved in the elongation reaction, to the isolation membrane to form the double membrane autophagosome (136). mTORC1 negatively regulates ULK1 by phosphorylating it at S757, while AMPK positively regulates ULK1 by phosphorylating it at S317 (112). As an inhibitor of autophagy, suppression of mTORC1 by TSC actually stimulates autophagy by inhibiting repressive mTORC1 phosphorylation of ULK1, and activating the ULK1 kinase.

In addition to metabolic-dependent regulation of autophagy, the mTOR pathway also regulates autophagy, and pexophagy, in response to reactive oxygen or nitrogen species (65, 79, 205) in a variety of settings (32, 206). Generation of free radicals such as O₂^·− can be selectively induced by glucose starvation, whereas amino acid and serum starvation can induce production of both O₂^·− and H₂O₂. Interestingly, autophagy induced by the addition of exogenous H₂O₂ also appeared to occur through a subsequent increase in intracellular levels of O₂^·− and not through H₂O₂ itself (32). Both reactive oxygen and reactive nitrogen species can induce autophagy via activation of ATM and subsequent suppression of mTORC1 via LKB1-AMPK-TSC2 signaling to mTOR (5, 242). ATM is a well-known DNA damage response protein (14, 106, 222), which only recently has become appreciated for its equally important role in the cytoplasm, where this kinase becomes activated in response to both reactive oxygen and reactive nitrogen species. Guo and colleagues (86, 87) have described the mechanisms of ATM activation under two different (DNA damage and H₂O₂) circumstances in the cell. ATM is generally present in an inactive dimer form, but is activated in the form of a monomer in response to DNA damage. In contrast, H₂O₂ directly activates ATM in its dimer form, via the formation of an intermolecular covalent linkage between two key cysteine residues. As peroxisomes are a very good source of O₂^·−, H₂O₂, and NO^· via the activity of multiple NADPH oxidases (NOX), peroxide-generating oxidases, and an inducible nitric oxide synthase (iNOS) (6), it is perhaps not surprising that several recent reports have identified both reactive oxygen and reactive nitrogen species as triggers for activation of ATM and directed autophagy of peroxisomes (pexophagy) (5, 242, 279, 280).

Autophagy/pexophagy occurs via the action of specific cargo receptor(s) and/or adaptors. These proteins contain modular domains that recognize cargo to be targeted by the autophagosome and autophagy effectors belonging to LC3 or Atg-8 like proteins (FIGURE 6) (174). These receptor(s)/adaptors provide a bridge between selected substrates and the autophagy machinery. Many such receptors/adaptors are found in mammals (12) and, based on their ubiquitin (Ub) binding characteristics, they can be divided into two groups: 1) receptors/adaptors that bind ubiquitinylated cargo through defined ubiquitin-binding domains (UBDs), e.g., NBR1 (45) and p62 (270); or 2) receptors/adaptors that contain a transmembrane or other domain/motif that mediates their interaction with the target substrate, rather than using a defined ubiquitin-binding domain, e.g., Atg30 in Pichia pastoris (PpAtg30) (61) and Atg36 in Saccharomyces cerevisiae (ScAtg36) (155, 156).

FIGURE 6.

Domains and motifs of the pexophagy receptors in yeast and mammals. In Saccharomyces cerevisiae, Atg36, and Pichia pastoris, Atg30, serve as pexophagy receptors. They contain a potential globular domain (GlobDom) as defined by Russell and Linding. Both Atg36 and Atg30 directly recognize proteins on the peroxisome surface and interact with the autophagy machinery via their LIR motifs and Atg11 binding sites. The mammalian systems have two pexophagy receptors, p62 and NBR1, also called adaptors. Both receptors/adaptors contain several conserved domains and motifs which include the following: PB1 (Phox/Bem1p) domain at the amino terminus, which is involved in the interaction with another p62 molecule; ZZ-type zinc finger domain in the middle; LIR domain for interacting with autophagic machinery; and the UBA domain at the carboxy terminus, which binds to ubiquitinated cargo. In addition, the NBR1 adaptor contains the CC1 and CC2 domains for homodimerization, the ZZ domain, and an amphipathic JUBA domain involved in lipid binding (allowing for direct peroxisome binding). Additionally, p62 has a KIR motif (Keap 1-interacting) nuclear localization signal (NLS1 and NLS2), nuclear export signal (NES), and TRAF6 binding domain (TBS).

While pexophagy is well studied in yeast, fewer studies have been performed in mammalian systems. In yeast, the Slt2 mitogen-activated protein kinase (MAPK) and several other upstream components of this signaling pathway have been recognized to regulate pexophagy (142). Phosphorylation of the pexophagy receptor PpAtg30 or ScAtg36 is the key event to start the pexophagy (FIGURE 7). The kinase for PpAtg30 is not known; however, Hrr25 kinase has been recognized to phosphorylate ScAtg36 (236). The receptor proteins PpAtg30 and ScAtg36 proteins have very low sequence similarity; however, both are pexophagy specific and required for the formation of the pexophagic receptor protein complex (RPC). PpAtg30 and ScAtg36 localized on the peroxisomal membrane via interaction with Pex3 and Pex14 (24, 156). Both PpAtg30 and ScAtg36 proteins interact with at least one peroxin of peroxisomes undergoing pexophagy. After peroxin binding, they recruit the autophagy machinery via interaction with autophagy-related (Atg) proteins. ScAtg36 is recruited to peroxisomes in a Pex3-dependent manner (156) and interacts with Atg11 or Atg8, while PpAtg30 interacts with Pex3 and Pex14 on the peroxisome membrane, as well as with Atg11, Atg8, and Atg17 to recruit the core autophagic machinery responsible for pre-autophagosomal structural (PAS) organization (24, 61, 159). In P. pastoris, another protein Atg37 which is evolutionarily conserved and involved in the formation of RPC complex has been identified recently (159). PpAtg37 interacts with peroxisomal membrane protein PpPex3 as well as PpAtg30 and enables recruitment of PpAtg11 to the RPC by positive regulation of PpAtg30 phosphorylation status (158, 159). No homologs of PpAtg30 or ScAtg36 exist in mammalian cells, where the mechanism of pexophagy is more complex and generally one of the peroxins is ubiquitinylated to serve as a signal for pexophagy (281a).

FIGURE 7.

Mechanisms of pexophagy in yeast. In yeast, pexophagy signaling is dependent on mitogen-activated protein kinase (MAPK) pathways (Mid2-Slt2 cascade). It may also be triggered by environmental (external or internal) factors such as signals related to the status of metabolic needs, or to high amount of reactive oxygen species generated by damaged or superfluous peroxisomes. In contrast to ubiquitinylation as observed in mammalian systems, phosphorylation of pexophagy receptors S. cerevisiae Atg36 or P. pastoris Atg30 in yeast serve as signals to induce pexophagy.

In mammals, NBR1 and p62 are the key adapter proteins for pexophagy. In contrast to PpAtg30 and ScAtg36, NBR1 and p62 are multifunctional proteins involved in other physiological processes (125). NBR1 is involved in endocytic trafficking as well as degradation of protein aggregates and midbody rings (100, 114, 143, 144). In addition to pexophagy, p62 is also involved in degradation of soluble proteins and aggregates, midbody rings, mitochondria, as well as bacteria (76, 101, 121, 161, 182, 276). Both p62 and NBR1 possess a similar Ub binding domain (UBA) (FIGURE 8). However, NBR1 additionally contains a JUBA domain that recognizes phosphatidylinositol phosphates (PIPs) and phosphatidic acid (PA) lipids and is essential for the localization of NBR1 to peroxisomes (45). Thus NBR1 is considered an endogenous pexophagy adaptor, as it can directly bind to both native (via JUBA) and ubiquitinylated (via UBA) peroxisomes. However, binding of the p62 adapter is pexophagy-specific, binding only ubiquitinylated peroxisomes via its UBA domain (280). In addition, both p62 and NBR1 also have LC3 interacting regions (LIR), with p62 possessing one and NBR1 possessing two LIRs (103) that specifically interact with autophagosome membranes (FIGURE 8).

FIGURE 8.

Binding of NBR1 and p62 to peroxisomes during pexophagy. During pexophagy, p62 and NBR1 bind with LC3 on the phagophore via the LIR domain and with ubiquitinylated peroxins. Together, these serve as signals for pexophagy via the UBA domain. Additionally, NBR1 has a JUBA domain so it can bind directly to the peroxisome membrane, and it is considered to be an endogenous receptor of pexophagy.

The mechanism(s) by which dysfunctional peroxisomes are recognized and selected for pexophagy in mammals is not well understood. Kim et al. (113) first demonstrated that pexophagy in mammalian cells is enhanced by ubiquitinylation of peroxisomal membrane proteins, PMP34 and PEX3, with the ubiquitinylated peroxins/PMPs facing the cytosol serving as a signal for recognition by ubiquitin-binding autophagy adaptors (NBR1 or p62) to trigger pexophagy. In mammals, PEX5 is ubiquitinylated during both the peroxisome import cycle, where it is ubiquitinylated at Cys11 (170), and during pexophagy. The ATM kinase shuttles between the nucleus and cytoplasm (5), where it localizes to peroxisomes and is activated in response to reactive oxygen species (263). ATM localization to peroxisomes is mediated by PEX5 binding to an internal PTS1-like sequence in the extreme carboxy terminus. In response to reactive oxygen species, ATM performs two functions: it phosphorylates PEX5 at Ser141 and suppresses mTORC1 via activation of AMPK-TSC2 signaling. Suppression of mTORC1 relieves its repression of ULK1 and promotes autophagy, while phosphorylation of PEX5 by ATM triggers ubiquitinylation of PEX5 and recognition and binding of p62 to target peroxisomes for pexophagy (243, 244, 279, 280). PEX5 phosphorylation by ATM occurs at S141 and is required for ubiquitinylation of PEX5 at Lys209 via the peroxisomal E3 ligase (PEX2, PEX10, PEX12). If normal recycling is prevented, PEX5 is monoubiquitinylated at Cys11 to trigger pexophagy in SV40 large-T antigen-transformed mouse embryonic fibroblasts (165), but whether this involves additional ubiquitinylation of PEX5 at Lys209 is not known.

X. MITOCHONDRIAL-PEROXISOME COMMUNICATION

Recent evidence has highlighted the strong reliance of mitochondria on peroxisomes. Both organelles communicate with one another to maintain lipid homeostasis, via β-oxidation. Coordinated growth is also carefully orchestrated by cell division machinery proteins from both organelles (210), vesicular trafficking between the organelles (160), and coupled redox sensitivity (70). Both organelles also share key proteins of their division machinery (210). Inability of either of these organelles to divide can have deleterious effects, manifested through developmental delays and early death (262). In addition, due to the necessity for coordinated function, both organelles rely on peroxisome proliferator-activated receptor gamma, coactivator 1 (PGC1α) for biogenesis (141, 266).

Unlike yeast or fungi, where all β-oxidation occurs in the peroxisome, mammals utilize both mitochondria and peroxisomes to degrade fatty acids. Each organelle harbors a unique set of catabolic enzymes necessary for processing these essential fats, each with unique substrate specificities. For example, the peroxisome is only capable of breaking down very-long-chain fatty acids and certain dietary fatty acids (primarily dairy-derived) that require α-oxidation (259). In turn, the shortening of these fatty acids allows for their transport and complete degradation within the mitochondria, as the precursor fuel molecule, acyl-CoA, which drives cellular ATP production (212).

Mitochondria necessarily rely on the acyl-CoA synthesized from fatty acids to maintain cellular homeostasis, and slowing or blocking their import decreases electron transport chain (ETC) function that relies on acyl-CoA, and ATP generation. The inability to process TCA intermediates stops the enzymatic activity of succinate dehydrogenase, eventually blocking the conversion of pyruvate and acetyl-CoA to α-ketoglutarate, a TCA substrate. The rippling effect results in the inability to convert α-ketoglutarate to succinyl-CoA, and the loss of NAD⁺ conversion into NADH, a precursor molecule for complex I of the ETC (64), promoting mitochondrial stress. Additionally, unique findings in yeast have revealed that the inability to form α-ketoglutarate stops the formation of their primary source of nitrogen: glutamate (140). To fulfill this need, yeast activate the mitochondrial retrograde (RTG) signaling pathway, causing the migration of the heterodimeric transcription factor Rtg1p-Rtg3p into the nucleus and the subsequent upregulation of RTG-targeted genes (102). These include the carnitine-dependent mitochondrial transporter, for increased transport of acetyl-CoA from peroxisomes, and genes necessary for peroxisome biogenesis and function (30, 241). The net effect of RTG signaling is to increase the available intermediates for the TCA cycle and ensure viability. To further facilitate this goal, mitochondria rely on mitochondrial-derived vesicles to facilitate transport of needed lipids from the peroxisomes (232). These mitochondrial-derived vesicles contain a mitochondrial-anchored protein ligase, dubbed ‟MAPL,ˮ which targets to the peroxisome (137), but further study will be necessary to fully elucidate what these vesicles actually contain, and to detail their roles in mitochondrial/peroxisome communication.

Peroxisomes are also sensitive to declines in the activity of the various mitochondrial electron transport chain complexes, several components of which are encoded by the mitochondrial DNA (mtDNA). Studies have revealed that rho-low (ρ^o) yeast cells, i.e., those lacking mtDNA, experience an increase in peroxisome biogenesis (58). This upregulation of peroxisomes allows for the additional production of Krebs cycle intermediates that can be utilized by the mitochondria. In the same study, yeast with intact mtDNA (ρ⁺), but treated with the complex III inhibitor antimycin, also triggered peroxisome growth, mirroring that found in ρ^o yeast cells (58).

Lastly, changes in the redox state of either mitochondria or peroxisomes can adversely impact the other organelle. Under homeostatic conditions, peroxisomes rely on a wide array of detoxification enzymes, such as catalase, to quickly eliminate H₂O₂ (as detailed in sect. VB). However, catalase lacks a strong peroxisome-targeting signal, which decreases its import efficiency (188), and this problem is only compounded with age (132). As a result, catalase mislocalization diminishes its enzymatic activity in peroxisomes. Excess accumulation of H₂O₂ transitions peroxisomes into an intracellular source of reactive oxygen species. Cell culture studies in which the enzymatic activity of catalase was blocked found a marked increase in H₂O₂ levels, not only within peroxisomes, but also in mitochondria. These mitochondria were also shown to have decreased aconitase activity, and lower transmembrane electrical potential, indicative of electron transport chain dysfunction (255) and/or membrane leakage. Conversely, the overexpression of catalase was found to prevent these effects (120). Additionally, in PEX5 knockout cells, deficient in peroxisome biogenesis, an abnormal abundance of mitochondria and a decline in mitochondrial respiratory rate occurs (11). Overall, these findings show that not only are peroxisomes and mitochondria closely metabolically linked, but they are also highly sensitive to one another’s redox state.

XI. HEREDITARY DISEASES OF PEROXISOMES

Peroxisome-specific enzymes account for <3% of the total cellular proteome (175). However, dysregulation of peroxisome biogenesis or peroxisomal enzymes can result in hereditary diseases that in the extreme are fatal or, in less severe cases, result in the development of neurological impairments such as blindness, hearing loss, and ataxia (8). The extent of disease severity is dependent on enzymatic function. Mutations that result in partial loss of function are typically milder in disease manifestation, whereas complete loss of enzymatic activity results in severe disease phenotypes. As stated earlier, peroxins are critical for proper formation of mature peroxisomes. Due to their critical role in peroxisome maintenance, defects found within 14 of the PEX genes have been directly linked to a wide array of disorders that are classified as peroxisome biogenesis disorders (PBD) (261). Detailed clinical presentation of PBDs has been extensively reviewed (152, 230), so it will not be focused on within this review.

The majority of patients affected (60–70%) are identified as having either a missense mutation or frameshift mutation in the PEX1 gene (186, 196). Normally, PEX1 along with PEX6 are anchored to the peroxisome membrane and are important in the release of ubiquitinylated-PEX5 from the peroxisome membrane (84). Defects in PEX1 can limit or block the import of peroxisomal matrix proteins. Specifically, one of the most common mutational variants is the relatively mild missense mutation C.2528G>A, which triggers temperature-dependent inactivation, blocking peroxisomal protein import in fibroblasts from these patients (98). The second most common form is a frameshift mutation resulting in a truncated form of PEX1 (34), vastly limiting peroxisomal protein import.

Mutations in the PEX2 gene can also cause peroxisome dysfunction. Specifically, 17 different mutations, including 4 missense, 5 nonsense, 6 deletions, and 2 insertions, have been identified (52), with the most common generating a truncated protein and resulting in partial loss of catalase import (99). Additionally, mutations in the PEX5 gene, key for cytosolic transport of peroxisome-destined proteins, has been found to have 11 different mutations, including 3 missense, 4 nonsense, 1 indel, and 2 splice site variants, which dampen PTS1 protein import (53).

With the discovery of the TSC signaling node at the peroxisome (279) in mammals, it is likely that dysfunctional peroxisomes also contribute to TSC, a hereditary hamartoma syndrome caused by defects in either the TSC1 or TSC2 genes. The TSC tumor suppressor is a heterodimer comprised of tuberin (TSC2), and its activation partner hamartin (TSC1), which localizes the TSC tumor suppressor to endomembranes and protects TSC2 from proteasomal degradation. TSC1 and TSC2 are localized to peroxisomes via peroxisome import receptors PEX19 and PEX5, respectively, and an ARL sequence was identified that is required for PEX5 binding and localization of TSC2 to peroxisomal membranes. In response to growth signals, TSC2 is phosphorylated by AKT and translocated to the cytosol by 14-3-3 proteins. Thus, when bound by 14-3-3 in response to AKT phosphorylation, TSC2 is sequestered away from its membrane-bound activation partner (TSC1) and its target GTPase (Rheb) to relieve the growth inhibitory effects of this tumor suppressor (26). The localization of the TSC to the peroxisomal membrane revealed that this tumor suppressor acts to suppress mTORC1 signaling and induce autophagy in response to peroxisomal reactive oxygen species. Certainly, it is at least plausible that defects in other components of signaling pathways that reside at the peroxisome, such as cancer syndromes associated with defects in the ATM and LKB1 tumor suppressors, may have peroxisome dysfunction as a contributor to disease pathogenesis.

XII. PEROXISOMES AND AGING

Similar to other organelles, peroxisomes manifest age-associated changes to their structure and functions (FIGURE 9). Indeed, very late-passage senescent cells try to mitigate decreased enzymatic activity within peroxisomes by stimulating peroxisome proliferation, which nearly doubles the amount of peroxisomes compared with cells of early or middle passage (131). Age-dependent changes in peroxisomal function were first noted in rat liver peroxisomes, with aged tissues showing a decrease in enzymatic activities (175). This was later shown to include a major decline in catalase activity (13). As mentioned previously, low import efficiency of catalase is further exacerbated by a decline in PEX5 affinity for peroxisome targeting signals with age (237). Consequently, the import of catalase is further dampened by an age-associated decline in the peroxisome import machinery (237). Declines in the import of de novo synthesized catalase, coupled with its already decreased activity, limits its ability to remove H₂O₂ (131), potentially tilting peroxisomes into disruptors of the cellular redox balance (FIGURE 10). The consequences of diminished catalase activity have been well-documented in individuals with hypocatalasemia disease. These individuals have only one-quarter of normal catalase levels (37), and their cells show increased H₂O₂ accumulation, protein oxidation, DNA damage, and decreased cellular growth rates, coupled with reduced peroxisome function (264). As a result, their cells undergo accelerated deterioration due to elevated peroxisomal reactive oxygen species.

FIGURE 9.

Decline in peroxisome function with age and/or senescence. With age, peroxisome activity declines. Peroxisomes from low-passage cells show efficient fatty acid oxidation (green). However, upon increasing passages, peroxisome efficiency declines (yellow), until peroxisome function is highly inefficient (red). To compensate, peroxisome proliferation occurs in late-passage cells. However, increasing peroxisome number is not enough to compensate for the high amount of damage that has already occurred, especially as many of the additional peroxisomes will also be damaged.

FIGURE 10.

Decline in catalase with age and/or senescence. Catalase is one of the primary peroxisomal enzymes to breakdown hydrogen peroxide. However, catalase contains a noncanonical peroxisomal-targeting signal, that decreases its recognition affinity by Pex5, compared with the stronger recognition elicited by the traditional signal. The decreased catalase targeting to the peroxisome is further compounded with age. Unlike low-passage cells, senescent cells show a decreased amount of Pex5 present in the cell, combined with further decreased recognition. This in turn lowers the ability of catalase to be targeted to the peroxisome. In addition, high-passage cells show decreased import efficiency by peroxisome import proteins, thus further limiting catalase import. As a result, peroxisomes from senescent cells have decreased function, and the organelle has a higher propensity to cause oxidative damage to itself and to surrounding cellular structures.

Age-associated declines in peroxisome function may also impact membrane integrity throughout the cell. Due to the importance of peroxisomes in lipid metabolism, their generation of usable fatty acids may restore lipid membranes that have been damaged by oxidative stress (175). In aged animals, the decrease in both catabolic and anabolic peroxisomal lipid metabolism may have direct consequences for membrane composition and function. Aged mice show a decline in both the synthesis of fatty acids, and in rates of fatty acid oxidative degradation, implying inefficiencies in these complementary processes. These effects appear to be partly mitigated in rodents by administration of the drug Clofibrate, which induces peroxisome proliferation (176).

Peroxisomes metabolize very-long-chain fatty acids (VLCFA) into usable components that are further metabolized by mitochondria, and if not eliminated, these VLCFA are actually toxic to the cell. Acyl-CoA oxidase is the primary enzyme for the degradation of VLCFA. Yet with age, the enzymatic activity of peroxisomal acyl-CoA oxidase decreases (176). Conversely, VLCFA can accumulate in the body, a hallmark common to peroxisomal diseases and a sign of β-oxidation dysfunction (153). Declines in VLCFA processing also dramatically lower the formation of docosahexaenoic acid (DHA), an essential polyunsaturated fatty acid (PUFA) (248). Furthermore, at least some studies suggest that accumulation of VCLFA can trigger the expression of inflammatory markers, which may downregulate peroxisome activity, thus further decreasing cellular ability to remove VLCFA and perpetuating a vicious cycle of VLCFA toxic accumulation (172, 227).

Nowhere is loss of DHA more detrimental than in the brain, as DHA constitutes ~30–40% of the phospholipids found in the grey matter of the cerebral cortex, with the greatest amount localized to synaptosomal membranes and synaptic vesicles (48, 217). Since neuronal tissues lack the necessary enzymes to synthesize DHA, they rely on DHA synthesized in the liver or DHA precursors derived from the diet. For optimal function, neuronal membranes require just the right degree of membrane fluidity. Increased accumulation of membrane cholesterol results in increased rigidity, which can be overcome by DHA supplementation (95, 272). Furthermore, declining levels of PUFA can be detrimental to essential dopaminergic signaling in the prefrontal cortex. Chronic deficiency of DHA is also an indication of decreased dopamine processing and signaling (282).

During normal aging, DHA levels have been found to decrease in neuronal membranes, with some speculation that suboptimal DHA levels are partially responsible for the decline in memory and learning that often occurs with age (128). Promising experimental results have indicated that DHA dietary supplementation in aged rodents was sufficient to restore DHA levels. Animals on DHA supplements showed improved maze-learning ability, increased dendritic spine numbers, and increased choline and acetylcholine levels (63, 74, 148), suggesting at least the possibility of dietary intervention in humans.

Interestingly, DHA levels are found to be significantly lower in the hippocampus of brains from (postmortem) Alzheimer's patients, compared with age-matched controls (229). Additionally, epidemiological studies found that high intake of dietary saturated fats was associated with an increased risk of developing Alzheimer's disease, while high dietary intake of PUFAs lowered the apparent risk (151). DHA supplementation is also beneficial in rodent transgenic Alzheimer's disease models, with results showing a 40–50% decrease in beta amyloid accumulation and improved learning (138). More importantly, in rodents, interventions of DHA supplementation before disease manifestation prevented decline in memory function (90).

XIII. CONCLUSIONS AND FUTURE DIRECTIONS

Peroxisomes execute a wide array of metabolic and detoxification functions, with the most well-characterized being β-oxidation of fatty acids and the breakdown of reactive oxygen and nitrogen species. Moreover, due to peroxisomes’ critical role in metabolism, it is not surprising that growing evidence suggests the interdependence between mitochondria, the ER, and peroxisomes, in coordinating peroxisome proliferation, turnover, and metabolic functions.

Peroxisome dysfunction has widespread consequences for cellular metabolism, as evidenced by multiple genetic peroxisome-linked disorders. Indeed, inability to turn over proteins, disruption of peroxisome protein import, and decreased efficiency in detoxification all contribute to reduced peroxisome function with age. Nowhere is peroxisome dysfunction more apparent than during the aging process, which is characterized by decreased peroxisome turnover, protein degradation, and the reduction in fatty acid breakdown. Moreover, aged peroxisomes become increasingly greater intracellular sources of reactive oxygen and nitrogen species, with decreased ability to neutralize these potentially harmful molecules.

Over the past several years, the role of peroxisomes has expanded far beyond that of a simple detoxifying organelle, such that it is now recognized as a central hub for multiple cellular signaling pathways. Originally, the most well-known cell signaling pathway linked to peroxisomes involved transcriptional activation of enzymes necessary for peroxisome biogenesis (201). More recent evidence points to the important link between peroxisomes and cellular redox status. Specifically, in response to oxidative stress, the TSC signaling node localizes to peroxisomes, where it is important to regulate the mTOR response (279). Additionally, ATM, which was also shown to be activated in response to oxidative stress, was found to localize in peroxisomes, where it is involved in signaling pexophagy (244). Damaged peroxisomes are internal sources of reactive oxygen and nitrogen species, which in turn can signal for their own degradation. Moreover, these new lines of evidence indicate the growing importance of peroxisomes in the redox homeostasis of the cell.

Despite significant progress, a full understanding of the mechanisms behind peroxisome biogenesis and degradation remains to be elucidated. Indeed, more understanding is still required to fully explain the roles of individual peroxisomal proteins. Nowhere is this more evident than in the case of LonP2, the peroxisome-specific protease, originally identified for its degradation of misfolded catalase (9), and whose role in peroxisome proteostasis largely inferred from parallel studies conducted in the mitochondrial LonP1 (184). Clearly far more research will be required to achieve a full understanding LonP2’s role(s) in the peroxisome.

Overall then, while great progress has been made in the past few years, the peroxisome remains perhaps the least well-understood of all eukaryotic intracellular organelles. This deficiency is especially apparent in our limited understanding of the role(s) of peroxisomes in redox regulation of multiple cellular functions, in various metabolic diseases, in the normal aging process, in age-related chronic diseases, and in the possible amelioration (or accidental exacerbation) of aging and disease by diet. Clearly, study of peroxisomal function and dysfunction will continue to be a productive and important area of research for many years to come.

GRANTS

C. L. Walker was supported by National Cancer Institute Grant CA143811 and National Institute of Environmental Health Sciences Grant ES023512. L. C. D. Pomatto was supported by National Science Foundation Grant DGE-1418060. K. J. A. Davies was supported by National Institute of Environmental Health Sciences Grant ES003598 and National Institute on Aging Grant AG052374.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.