| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Biochemistry, Microbiology and Immunology, Faculty of Medicine, University of Ottawa, Ottawa, Ontario, Canada
2Correspondence: Department of Biochemistry, Microbiology and Immunology, Faculty of Medicine, University of Ottawa, 451 Smyth Rd., Ottawa, ON, Canada K1H 8M5. E-mail: maryellen.harper{at}uottawa.ca
 |
ABSTRACT |
|---|
Key Words: proton leak • skeletal muscle • thermogenesis • ROS • fatty acid metabolism
|
INTRODUCTION |
|---|
35 members. The original UCP, UCP1, or "thermogenin", is highly selectively expressed in brown adipose tissue. The physiological role of UCP1 in thermogenesis is well characterized and has been well described (1
, 2)
. The novel uncoupling proteins (UCPs 2–5) discovered in the late 1990s have high homology to UCP1, but it is unclear whether they in fact can uncouple oxidative phosphorylation. UCP2 mRNA is quite ubiquitously expressed, and its protein may have functional importance in neuronal and immune system cells, and in pancreatic beta-cells. Its roles will not be discussed here, but a number of recent reviews are available [see for example, (3
4
5
6)
]. UCP3, which has 60% homology to UCP1, is highly expressed in skeletal muscle and, to a lesser extent, in brown adipose tissue (BAT) and heart. That this UCP is the only one to be expressed at the level of protein in skeletal muscle makes it of particular interest, e.g., does UCP3 play an important role in muscle energetics? The homology between UCP3 and UCP1 hinted indeed to a potential role in the uncoupling of oxidative phosphorylation in skeletal muscle, with implications for energy balance, and the development of obesity. But paradoxical physiological triggers for up-regulated UCP3 expression, such as fasting, argue against such a function. With no other obvious leads, a wide variety of experimental approaches and systems were explored with the goal of identifying the true function of UCP3. To date there are well over 600 UCP3 publications, and still the function of UCP3 is unclear. Hypothesized functions currently being studied include those relating to mitigated ROS production and modified FA handling. This review will focus on mechanistic aspects of UCP3 function and the potential metabolic implications. The specific hypotheses of UCP3 function are introduced immediately below and are discussed later in the review in the context of the various experimental approaches used.
|
STATEMENT OF HYPOTHESES |
|---|
). Accordingly UCP3 would dissipate the proton gradient across the mitochondrial inner membrane, thereby dissociating respiration from ATP synthesis. In other words, oxygen consumption would be uncoupled from ATP production.
|
The uncoupling caused by UCP1 in BAT is a tightly regulated process, with adrenergic activation and thyroid hormones playing important roles (1)
. For UCP1-mediated uncoupling to occur, activation of UCP1 is required; FA are thought to be the key activating species. In tissues other than BAT, including muscle, another form of uncoupling is known to occur, and, since its identification in the early 1990s, it has been referred to as "proton leak" (7)
. It does not appear to be acutely regulated by intracellular signaling molecules but is increased when protonmotive force is high. When UCP3 was identified in 1997, accordingly proton leak was proposed as its mechanism of action. However, many subsequent studies have demonstrated that UCP3 does not cause this basal proton leak [e.g., (8
, 9)
]. Indeed it appears that the adenine nucleotide transporter accounts for the majority of proton leak (10)
.
Thus, it rapidly became apparent that UCP3 did not cause basal proton leak in muscle. A mechanism involving an activated proton leak similar to UCP1 has also been explored. It is important to acknowledge, however, that the molecular mechanism of UCP1 action remains unresolved. Two mechanisms have been proposed. Either directly or indirectly, both mechanisms result in the movement of protons from the mitochondrial intermembrane space to the matrix. These include the FA protonophore mechanism (Fig. 1-1
) (11)
and the FA proton-buffering mechanism (Fig. 1-2
) (12)
. Indeed, the nature of the molecular species transported by UCP1 (protons vs. FA anions) and the mode of transport across the mitochondrial inner membrane remain unclear. Thus, while the homology to UCP1 biased opinions about the physiological function of UCP3, the controversy surrounding the mechanism of UCP1 action has provided a less restrictive milieu for hypothesized mechanisms of UCP3 action.
FA anion exporter: increasing FA oxidation (FAO) capacity
Much evidence, including that from many UCP3 mRNA expression pattern studies, points to a role for UCP3 in FA metabolism. The first specific role for UCP3 in FA metabolism was hypothesized by Himms-Hagen and Harper (13)
, which essentially puts forth the idea that UCP3 provides an overflow pathway for FA in conditions of excess mitochondrial FA supply. When mitochondria are faced with greatly elevated FA flux, LCFA-coenzyme A units accumulate in the matrix sequestering the limited coenzyme A (CoA) pool in the matrix. Limited CoA levels could thus prevent further oxidation of partially metabolized FA. It is hypothesized that UCP3 acts in conjunction with a mitochondrial thioesterase (MTE-1) to liberate CoA and export FA anions (Fig. 2
-1). Export of FA anions from the matrix by UCP3 would not only allow greater rates of FAO but may also reduce membrane potential and thereby mitigate ROS production.
|
FA anion exporter: protection from lipid peroxidation
Schrauwen et al. independently proposed an hypothesis that is somewhat similar to the extent that UCP3 is proposed to translocate FA (14)
. FA anion export during conditions of elevated FA flux is also posited, but it is hypothesized as a mechanism to protect from lipid peroxidation (Fig. 2-2
). The overall rationale is as follows: when FA supply exceeds oxidation, a larger fraction of FA enters the matrix via a transmembrane FA flip-flop mechanism (as opposed to the traditional CPT system) (15)
. Further, the imported FA would be nonmetabolizable due to the absence of acyl-coenzyme A synthase (ACS) in the matrix. As the accumulation of FA in the matrix increases the risk of mitochondrial lipid peroxidation, UCP3-mediated FA efflux would protect against such damage. The mechanism of action proposed by Schrauwen et al. is essentially the FA cycling model proposed by Skulachev for UCPs in general (16)
; however, Schrauwen et al. suggest that the FA cycling is limited to conditions of elevated FA supply.
Activated uncoupling: protection from ROS
This hypothesis, proposed by Brand and colleagues, suggests an important role for UCP3 in the mitigation of ROS production. It is estimated that, of the oxygen consumed by mammalian cells, 0.2–2% of it is converted to ROS, and most of the ROS have mitochondrial origins (17
, 18)
. Superoxide is quantitatively the most important species, and is produced at complexes I and III of the electron transport chain (ETC) (19
20
21)
. There is also evidence for matrix sources of O2– not produced by the ETC (22
, 23)
. Superoxide production by the ETC is favored by high cellular oxygen content and/or a highly reduced state of the ETC. Therefore, mechanisms that reduce membrane potential lower O2– production (24)
. Indeed, mitochondrial H2O2 production is abolished by low concentrations of chemical uncoupling agents (25)
. While there are many other well-recognized ROS protective mechanisms (e.g., the superoxide dismutases; glutathione peroxidase, catalase), this hypothesis proposes that UCP3 mitigates the actual production of ROS. Following a series of elegantly designed in vitro studies (further described below), Brand and colleagues proposed a mechanism (Fig. 2-3
) involving the activation of a UCP3-mediated proton leak by 4-hydroxynonenal (4-HNE), a byproduct of lipid peroxidation (26)
. This is suggested as a negative feedback loop to reduce transmembrane protonmotive force and, thereby, mitigate ROS production.
These specific hypotheses have provoked pronounced interest in the potential physiological and pathological implications of UCP3 function. For example, many studies have been conducted in the context of energy balance and obesity, in lipotoxicity and insulin resistance, and in oxidative stress. In the next sections we discuss evidence for and against the above hypotheses in this wide contextual spectrum.
|
Nonshivering thermogenesis |
|---|
, 27)
. Based on the putative role of UCP3 as an uncoupler, an early hypothesis for its physiological function was that it contributes to NST in skeletal muscle. As UCP3 is also expressed in BAT (28
, 29)
, it is possible that it may play some ancillary role in BAT NST. However, evidence argues against such a function. Injection of triiodothyronine or a selective ß3 agonist increases oxygen consumption and body temperature to the same extent in wild-type (WT) and UCP3 (–/–) mice (30)
. The thermogenic response to ß3 agonist is blunted in UCP1 (–/–) mice (2
, 30)
, and no phenotypic differences have been found between UCP1 (–/–) mice and UCP1,3 double knockouts (30)
. Recently it was demonstrated that hamsters with a mutation leading to loss of UCP3 in BAT had the same maximal thermogenic response to noradrenaline as WT hamsters (29)
.
It is important to recognize that the level of UCP1 protein in BAT is 200- to 700-fold greater than UCP3 protein levels in skeletal muscle or BAT (28)
. Thus, the low amount of UCP3 protein may be one explanation as to why UCP3 does not contribute to NST. Another explanation may relate to the state of activation of UCP3. In this regard, UCP3-associated thermogenesis was recently demonstrated with intense cellular adrenergic activation induced by 3,4-methylenedioxymethamphetamine ("ecstasy"). Mice treated with ecstasy undergo rapid increases in rectal and muscle temperature, whereas UCP3 (–/–) mice do not (31)
. A proximal mechanism for this effect is a massive release of norepinephrine with subsequent signaling via 
1 and ß3 adrenergic receptors on skeletal muscle [reviewed in (32)
]. Thus, it can be concluded that UCP3 has the capacity to mediate heat production in vivo. However, a significant thermogenic function of UCP3 would be restricted only to situations of supraphysiological muscle stimulation. That the rise in muscle temperature in WT mice is accompanied by rhabdomyolysis (33)
suggests that significant activation of UCP3 could also compromise skeletal muscle integrity.
|
Mitigation of ROS production |
|---|
. It has been further hypothesized that UCP3 is specifically activated by ROS or byproducts of lipid peroxidation, which in turn would lead to a mild uncoupling, membrane depolarization, and subsequently mitigated ROS production. Supportive evidence for this hypothesis was first obtained in isolated skeletal muscle mitochondria in which exogenously produced O2– (via addition of xanthine and xanthine oxidase) induced a proton leak that could be inhibited by GDP ("GDP-sensitive conductance"; Fig. 3
). This conductance could not be activated in mitochondria from UCP3 (–/–) mice or in liver mitochondria (which do not express UCP3 protein) (35)
. In a subsequent study, exogenously supplied 4-HNE, a reactive aldehyde derivative of lipid peroxidation, also activated a GDP-sensitive conductance in skeletal muscle mitochondria from WT but not UCP3 (–/–) mice (26)
. These observations causally linked UCP3 to the stimulation of uncoupled respiration in skeletal muscle mitochondria by exogenously supplied ROS or their lipid peroxide metabolites, using GDP inhibition to expose the conductance (see below). However, one concern was the use of supraphysiological amounts of O2– or lipid peroxide metabolites. As well, it was intriguing that GDP inhibition occurred even at low membrane potentials, where ROS production would be expected to be negligible. Both matters were addressed in a follow-up study using rat skeletal muscle mitochondria. Endogenously generated O2– (using conditions that favor O2– production at complex I of the ETC) was associated with increased nonphosphorylating O2 consumption that could be blocked by GDP; this effect was greatly attenuated when membrane potential was not allowed to reach high levels and thus oxidative stress would not occur (36)
.
|
While the above-described observations are consistent with the idea that ROS can induce UCP3-mediated uncoupled respiration, several questions remain. The mechanistic basis for this association is unclear; it remains possible that a process correlated with the presence of UCP3 is the true determinant of the effects attributed to UCP3. A more fundamental issue relating to this series of studies is the use of GDP to elucidate a potential physiological role of UCP3. UCP3, like UCP1 (and UCP2), contains a nucleotide binding domain (37
, 38)
. Although it is clear that a GDP-sensitive conductance can be activated in skeletal muscle mitochondria from WT but not UCP3 (–/–) mice (26
, 35
, 39)
the Km for inhibition of UCP3 is below intracellular levels of free GDP (40)
. Thus, while GDP has been a useful tool, it is difficult at this time to draw conclusions regarding the physiological significance of the GDP-sensitive conductance. It was recently suggested that physiological conductance changes (i.e., those occurring in the presence of physiological concentrations of purine nucleotides) mediated by UCP3 may not be detectable using current methods (41)
.
Recently, data obtained by another group have challenged the hypothesis that O2– induces mild uncoupling catalyzed by UCP3. Silva et al. investigated the hypothesis using mice overexpressing superoxide dismutase (SOD)-2 (SOD2) (42)
, the isoform found in the mitochondrial matrix. Isolated skeletal muscle mitochondria from SOD2 mice produce less ROS and have higher aconitase activity (suggestive of lower oxidative stress) than WT mice. It was expected that if UCP3-mediated uncoupling were playing a role in mitigating ROS production, then mitochondria from SOD2 mice would have a smaller GDP-sensitive proton conductance than WT mice under conditions of high endogenous ROS production. Under such conditions a GDP-sensitive conductance was detected in the two genotypes, suggestive of uncoupling presumed to be mediated by UCP3. However, the magnitude of the GDP-sensitive conductance was similar in SOD2 and WT mice. These observations indicate that the mechanism proposed by Brand and colleagues was not operative, despite the greater ROS production in the mitochondria from the WT compared to the SOD2 mice, and despite the presence of a GDP-sensitive conductance in both genotypes. Possibly, the difference in ROS production between the two genotypes was too small to result in any discernable difference in the GDP-sensitive conductance. Alternatively, the lower amount of ROS produced in the mitochondria from SOD2 mice may have been sufficient to produce a maximal GDP-sensitive (and presumably UCP3-mediated) conductance. A better understanding of the quantitative relationships between ROS or HNE and uncoupled respiration or GDP-sensitive conductance would provide greater insight into these possibilities.
Mechanistic details notwithstanding, there are several reports of an association between UCP3 and a reduction in oxidative stress. Isolated mitochondria from UCP3 (–/–) mice produce more ROS (43)
, show an increase in markers of oxidative damage to proteins (44)
, and have lower aconitase activity (40
, 43)
. It is noteworthy that the knockout mice used by Brand and colleagues were backcrossed six generations (40
, 44)
. ROS production is also reduced in L6 myotubes overexpressing UCP3 at levels that did not uncouple cells under basal conditions (45)
. In contrast, however, mice overexpressing UCP3 do not have lower levels of oxidative damage to proteins compared to WT mice (44)
, and Mozo et al. (46)
found no evidence for lower ROS production in Chinese hamster ovary (CHO) cells expressing UCP3 at levels found in skeletal muscle (and which caused no respiratory uncoupling) compared to control CHO cells expressing no detectable UCP3. These discrepant results may be explained by differences in the basal levels of ROS; levels of oxidative stress in WT mice and nontransfected CHO cells may be below the subthreshold for detection by UCP3 so that no further protection could be afforded in UCP3 overexpressing mice.
In summary, the in vitro evidence indicates oxidative stress can lead to activation of mild-uncoupling mediated by UCP3. It is not yet clear, however, whether such a mechanism could operate in vivo. Moreover, given that skeletal muscle mitochondria already possess robust antioxidant systems, some might question the "need" for another such mechanism. The in vitro studies described above predict that UCP3 would have an antioxidant function selectively at times when protonmotive force is high, such as in resting muscle (low ATP demand), or when activity of ETC enzyme complexes is inhibited, such as during hypoxia (47)
.
|
Whole body energy balance |
|---|
, 49)
.
Evidence supporting a role for UCP3 in body energy balance comes primarily from transgenic studies. Mice overexpressing human UCP3 have lower body weights than WT mice (50
, 51)
. A leaner phenotype despite hyperphagia is prominent at supraphysiological UCP3 overexpression (50)
but is at least partially due to artifactual uncoupling from improper insertion of UCP3 in the mitochondrial membrane (28)
. However, reduced body weights, in the absence of hyperphagia, are also detectable in older (51)
, but not younger (8)
, mice overexpressing UCP3 at a physiological 2-fold levels. UCP3 overexpressing mice have smaller fat pads and livers, indicating less energy storage in those organs when compared to congenic (backcrossed 10 generations) WT mice (51)
. Leaner phenotypes with UCP3 overexpression have not always been apparent and have required conditions challenging energy balance to become noticeable (52)
. Interestingly, UCP3 (–/–) are not obese (8
, 30
, 43
, 51
, 53
, 54)
. When challenged with a high-fat diet, UCP3 (–/–) and WT mice gain weight at a similar rate (30
, 54)
. When faced with a fast, reports vary: Both comparable (30)
and decreased (8)
rates of weight loss have been observed when compared to WT mice. It is fair to state that the mouse phenotype resulting from either physiological UCP3 overexpression or ablation is quite subtle. In many cases, energy balance needs to be challenged (fasting, exercise, high-fat feeding) to detect differences in body weight and energy storage. Some discrepant results may be related to the varying degree of UCP3 overexpression (e.g., 2- vs. 20-fold) in the UCP3-tg mice, as well as the mixed genetic background of UCP3 (–/–) mice used in many of the studies. Mouse strain can strongly affect metabolic phenotype (55
56
57)
. As a result, potential differences between groups may be masked, or differences may be revealed that are in fact related to genetic background. Therefore, results obtained using UCP3 (–/–) and UCP3-tg mice should be interpreted cautiously.
The role of UCP3 in human energetics has been examined by correlating UCP3 expression in energy balance. Largely, weight loss is associated with decreased UCP3 gene expression and protein (58
, 59)
. It is highly likely that UCP3 levels are reduced in response to diminished FA exposure post weight reduction. Mingrone and colleagues found a 35% reduction in UCP3 protein levels in subjects having undergone bilio-pancreatic diversion (BPD) surgery (60)
. But since BPD results in large fecal lipid loss, which drastically minimizes dietary fat availability (61)
, it is difficult to dissociate the effect of weight loss alone, from FA levels, on UCP3 expression.
The possibility of a causal relationship between UCP3 and weight reduction has only been briefly examined. In type II diabetics, Schrauwen et al. found a 40% reduction of UCP3 expression and protein levels following weight loss (59)
. Interestingly, a negative correlation was identified between the change in UCP3 protein levels and BMI, which suggests that UCP3 abundance facilitates weight loss. This observation is in line with findings obtained in nondiabetic obese subjects on a very low caloric-restriction diet (62)
. Obese, otherwise healthy, female subjects placed on a 900 Kcal diet for 6 months lost weight at very different rates. Diet-sensitive subjects, who lost weight at a greater rate than diet-resistant subjects, had 25% higher UCP3 expression levels than diet-resistant subjects. The absence of pre-weight loss UCP3 quantification prevents a true comparison between this study and the above-mentioned study by Schrauwen et al.
The human cases described above allow for no more than an association between UCP3 expression/protein levels and body weight or weight loss success. Whether the change in UCP3 levels is a cause or an effect of weight loss is also unknown. Moreover, these findings do not indicate which of the putative UCP3 functions could be responsible for weight loss success. Increased uncoupling or increased energy expenditure (via FAO) could both be implicated in UCP3-derived whole-body energy balance.
|
Genetic variants of human UCP3 |
|---|
, increased BMI in a French Caucasian population (64)
, increased hip-to-waist ratio in female, but not male South Indians (65)
, and in a German population when adjusted for gender, age, BMI, and diabetes mellitus (66)
. Recently, Gable et al. convincingly demonstrated in a prospective study that homozygosity of the T allele in the –55C > T polymorphism accelerates the onset of diabetes when examined after 10 years (67)
. The diabetes hazard ratio was 2.6 times higher in men when the UCP3T allele was combined with the UCP2 –866 G > A, and 12 times higher in subjects with a BMI greater than 30 kg/m2, thereby providing strong evidence for a link between these UCP polymorphisms and energy balance. Conversely, the –55C > T polymorphism has been found to be less prevalent in type 2 diabetics than nondiabetic French subjects but was simultaneously associated with an atherogenic lipid profile (68)
. It is also associated with lower BMI in UK (69)
, US Caucasian (70)
, and Spanish (71)
populations, with the latter conditional on correction for recreational energy expenditure. Lastly, two long-term follow-up studies have found no association between the –55C > T polymorphism and BMI or percent body fat in obese and control Danish subjects (72
, 73)
.
Another intriguing polymorphism is the exon 6 splice donor IVS6 + 1G > A, which results in a truncated version of UCP3. Functional analysis of human UCP3 mutants with the IVS6 + 1G > A mutation has shown that it results in a truncated protein identical to the short form of UCP3 (which lacks the sixth transmembrane domain) but with equal ability to alter membrane potential when tested in yeast systems (74
, 75)
. It is, however, unknown to what extent artifactual uncoupling may have contributed to the latter findings in the yeast expression model. The IVS6 + 1G > A is particularly intriguing since it has been associated with elevated respiratory quotient in Gullah African-Americans (76)
but not in African-Americans from Maywood, IL, where genetic admixture is approximately four-fold greater than in the former population (77)
.
Several other mutations/polymorphisms have been identified in UCP3, but most appear to be isolated cases with no consistent family history of obesity or diabetes. Although of interest, UCP3 genetic variants studies are mainly correlative, or at most, allele frequencies are linked with crude measures of energy expenditure. For population studies to become more powerful, measurements of the proposed functions of UCP3 must be made in human subpopulations with specific allele frequencies for a given polymorphism.
|
Patterns of modulated UCP3 expression |
|---|
, 79)
, acute exercise (80
, 81)
, and high-fat feeding (82
, 83)
have all demonstrated increased UCP3 expression in rodents and humans. These observations are consistent with the identification of a PPAR response element in the promoter region of UCP3 (84)
and the demonstration that treatment of L6 myotubes with PPAR
agonists increases UCP3 mRNA levels (85)
. Recently, Pedraza et al. demonstrated that in addition to PPAR, PPAR is also an important regulator of UCP3 mRNA expression, particularly in the heart and during the neonatal period (86)
. Altogether these observations are consistent with the general line of thought that UCP3 plays some role in FA metabolism.
However, it is quite clear that chronic exercise, which leads to an accentuated reliance on FAO, is associated with decreased UCP3 mRNA and protein levels in rats and humans (87
, 88)
. Similarly, levels are lower in oxidative than in glycolytic fibers of rat and human skeletal muscle (89
90
91)
. Beyond the fact that oxidative fibers, compared to glycolytic fibers, have greater mitochondrial contents, they have a much greater capacity for elevated FAO rates. Thus, oxidative fibers and their resident mitochondria have an increased capacity to cope with dramatically elevated FA fluxes. Therefore, it could be proposed that fibers already equipped with this increased capacity for FA handling would not require UCP3-mediated FA anion export pathways (Figs. 2.1, 2.2
). Chronic exercise, which recruits oxidative fibers, would thus be expected to result in decreased UCP3 content. However, one could also argue that the hypothesized "mild uncoupling to protect against ROS" mechanism (Fig. 2.3
) would also be consistent with the decreases in UCP3 expression with chronic exercise. Primary ROS handling enzymes such as glutathione peroxidase, catalase, and SOD increase with chronic exercise (92)
. Hence, a role for UCP3 in the mitigation of ROS production may become less important.
Patterns of UCP3 expression have led Dulloo and colleagues to propose that UCP3 does not facilitate FAO per se but rather is involved in the adaptation to the increased supply of FA (48)
. Greater fasting-induced up-regulation of UCP3 in glycolytic than in oxidative fibers supports this notion. It is proposed that the larger increase in UCP3 expression in glycolytic fibers is due to the fact that glycolytic fibers are required to change their "fuel mix" to a greater extent than are oxidative fibers (91)
. UCP3 expression levels have been extensively examined in tissues and fiber types across a range of oxidative capacities. It is, however, still unknown if the overexpression and ablation of UCP3 in murine models affects the relative proportions of muscle fiber types. Specifically it would be of interest to determine whether long-term adaptation to the presence or absence of UCP3 causes a shift in fiber type proportions. If UCP3 facilitates FAO, one might expect increases in oxidative fibers in skeletal muscle of UCP3 overexpressing mice. However, the presence of a FA "overflow" pathway may also hinder the normal adaptation and fiber type switch observed in mitochondria normally adapting to FAO.
|
Lipid metabolism |
|---|
, 93)
, skeletal muscle (94)
, or isolated mitochondria from muscle of mice overexpressing UCP3 (53)
. Labeled palmitate yielded greater 14CO2 production rates in soleus muscle of mice overexpressing UCP3 (20-fold protein) when compared with WT littermates (94)
. To obtain more clues as to how FAO may be augmented in physiological conditions up-regulating UCP3 expression, Bezaire et al. assessed FA transport and oxidation capacity in a physiological (2-fold) UCP3-tg mouse model (53)
. The concerted up-regulation of FA binding protein levels, and carnitine palmitoyltransferase I, citrate synthase, and ß-hydroxyacyl-coenzyme A dehydrogenase activity, independent of mitochondrial volume, supported increased uptake of FA into the cell and into the mitochondria for oxidation. Indeed, intramuscular triglyceride (IMTG) storage in the cytosol was decreased in UCP3 overexpressing mice. Given the originally proposed function of UCP3 in uncoupling, it is reasonable to question whether FAO is increased in a specific manner or as result of uncoupling-driven increased oxygen consumption. Two studies have found the effects of UCP3 overexpression on oxidation, membrane potential, and energy status not mimicked by chemical uncoupler treatment of cultured cells, thereby providing support against a role for UCP3 as a true uncoupler of oxidative phosphorylation (45
, 93)
.
Increased FAO rates and decreased IMTG stores in UCP3 overexpression models provide convincing evidence to support the UCP3 FA anion export model proposed by Himms-Hagen and Harper (Fig. 2.1
). It is difficult to reconcile these findings with Schrauwen’s hypothesis (Fig. 2.2
), which also proposes FA anion export but without a link to FAO, and thereby muscle lipid stores. Importantly, no impairment in FAO capacity has been detected in the absence of UCP3 (43
, 53
, 54)
. Conflicting results have been found with regard to muscle lipid pool size in high fat-fed UCP3 (–/–) when compared to WT littermates (51
, 54)
. Schrauwen et al. found that lipid stores were smaller in UCP3 (–/–) than WT mice, while lipid peroxidation was more pronounced in this genotype on a control chow diet but was reduced to equal levels in both genotypes on a high-fat diet. The authors speculated that lipid stores may be reduced in UCP3 (–/–) mice to minimize mitochondrial damage from lipid peroxidation, pointing to adaptation as an explanation for their peculiar findings obtained in UCP3 ablated mice. Why overexpression, but not ablation of UCP3, supports a role for UCP3 in FAO is indeed perplexing. Adaptation and/or compensation in the UCP3 (–/–) mice are possible explanations, given the existing functional redundancy across the mitochondrial anion carrier proteins. However, to date, no one has attempted to convincingly address this issue in skeletal muscle of UCP3 (–/–) mice. Gene array analyses in skeletal muscle from UCP3 (–/–) mice or the use of a conditional UCP3 (–/–) mouse could provide important insights.
Measuring rates of 14C-FAO has been a popular approach to assess the role of UCP3 in FAO capacity. One must remember, however, that a large fraction of FA will be trapped as acid soluble product (ASP – also called "products of incomplete oxidation") and not fully oxidized to 14CO2 (95)
. The respective contributions of CO2 and ASP to total oxidation can vary as a function of limitations at any step of the TCA cycle or ß-oxidation spiral. It is, therefore, important to assess both 14C-CO2 and 14C-ASP to obtain an accurate status of ß-oxidation capacity in the genotypes studied. One could also perform a more detailed assessment of oxidation capacity employing metabolomics analysis. Identifying and quantifying specific TCA and ß-oxidation intermediates accumulating over time during FAO in WT, UCP3 overexpressing, and UCP3 (–/–) mice would certainly highlight a facilitated or impaired oxidation pathway in a highly specific manner.
Recently, Gerber et al. provided the first evidence of FA export from heart mitochondria (96)
. Using labeled palmitoylcarnitine as a substrate, the authors quantified labeled palmitate that was apparently produced and exported by mitochondria. They provide strong evidence that palmitate is generated and exported in heart mitochondria under basal conditions, and more so in a diabetic rat model. UCP3 and MTE-1 protein levels were also found to be increased in diabetic rat heart mitochondria. Mitochondrial palmitate export in heart, a tissue containing very low UCP3 levels, particularly under basal conditions is somewhat surprising, but we suspect that, shortly, new evidence will allow comparisons of the palmitate export rates in heart mitochondria obtained by Gerber et al. to rates in skeletal muscle under various conditions, including UCP3 ablation and overexpression. At this point, evidence for UCP3-dependent FA export is only correlative but these findings may be very important.
|
Glucose metabolism and diabetes |
|---|
, 51)
. In both cases, glucose clearance was also improved, which may suggest a role for UCP3 in insulin sensitivity. Two in vitro reports have indicated peculiarities in glucose oxidation and sensitivity to FA-induced inhibition of glucose oxidation (93
, 94)
while another found that glucose oxidation was not increased with UCP3 overexpression, whereas FAO was (45)
. These conflicting results emphasize the ambiguity of a role for UCP3 in glucose metabolism.
Nonetheless, the cellular mechanisms responsible for potential UCP3-derived improvements in glucose homeostasis have been investigated in UCP3 overexpressing mice and cultured myotubes. Huppertz et al. demonstrated that overexpression of UCP3 in cultured myotubes stimulates glucose transport and GLUT4 recruitment to the cell surface via a PI3K-dependent mechanism (97)
. However, the nonspecificity of this response was exposed by the demonstration that a low-dose, long-term chemical uncoupler treatment of cultured myotubes similarly increased glucose uptake via a PI3K-mediated pathway. Therefore, improved glucose homeostasis in this model may simply be an artifact of UCP3 protein overexpression. The plausibility that UCP3 exerts its effect on glucose metabolism via AMP-activated protein kinase (AMPK), the metabolic "master switch" of the cell, has also been investigated. While no change in phosphorylated AMPK to total AMPK ratio was detected in a 2-fold overexpression model when compared with WT and UCP3 (–/–) congenic mice (53)
, others have found decreased ATP/AMP ratio and increased 1 AMPK activity in a slightly more pronounced overexpression mouse model (5.8-fold protein), (98)
. The authors suggest that AMPK may be partially responsible for improved glucose handling in UCP3 overexpressing mice despite evidence that the 2 subunit, and not 1, is most involved in glucose homeostasis. Unfortunately, serum fasting glucose was not measured in these mice specifically, thereby weakening the link between the two observations. Nevertheless, a secondary function for UCP3 in glucose homeostasis via mild uncoupling is proposed. This secondary function is more limited than the one proposed by Himms-Hagen and Harper, for which one could envision improvements in glucose homeostasis via both mild uncoupling, and decreased lipotoxicity through increased FAO. In light of these findings, it is apparent that functional analyses are warranted to better assess the plausible role of UCP3 in glucose homeostasis.
Similarly to FAO, to our knowledge no evidence indicates impaired glucose uptake or clearance in the absence of UCP3. Both original characterization studies of UCP3 ablated mice found no indication of reduced glucose uptake or abnormal insulin levels (30
, 43)
. In fact, decreased glucose levels were even found in female UCP3 (–/–) when compared to WT mice (30)
. Similarly, Costford et al. found that UCP3 (–/–) were just as protected from glucose intolerance and insulin resistance as UCP3-tg mice when compared to WT mice from congenic background (51)
. These findings demonstrate again the complexity of the UCP3 (–/–) mouse model. Therefore, it is difficult to conclude whether UCP3 is involved in glucose homeostasis. It is highly likely that the subtle improvements in glucose homeostasis could be secondary to mild uncoupling from UCP3 overexpression or reduced lipotoxicity as a result of enhanced skeletal muscle FAO.
|
PROSPECTS AND PREDICTIONS |
|---|
). An association between UCP3 and lower oxidative stress seems strong; although the mechanistic details are unresolved (see Mitigation of ROS production). Oxidative stress is widely implicated in the processes underlying aging (21
, 99
100
101)
. Further information regarding the effects of absence and increased expression of UCP3 on aging and life span is needed. Oxidative stress has been causally linked to reduced insulin sensitivity in adipocytes (102)
; whether this also occurs in skeletal muscle has not been directly investigated. However, it can be suggested that increases in UCP3 may improve insulin sensitivity in skeletal muscle; supportive evidence for this possibility exists (51)
. Further study of this hypothesis seems worthwhile, especially in light of the associations between UCP3 polymorphisms and perturbed metabolic phenotypes (see Genetic variants of human UCP3). One approach could be to cross mice that overexpress UCP3 (at physiological levels) with an insulin-resistant mouse model, to test whether increased UCP3 expression can rescue one or more of the metabolic abnormalities. In the context of oxidative stress and the role of UCP3, it is clear that oxidative stress alters cell function through signaling cell death pathways and may ultimately compromise cell viability. Evidence exists, in the ischemia-reperfusion literature, that UCP3 enhances cell viability (103)
. Moreover, in endothelial cells, hyperglycemia increases ROS production, which then activates several pathways implicated in cell damage, such as the NF
B pathway via PKC activation (104)
. Interestingly, overexpression of UCP1 in endothelial cells exposed to hyperglycemia lowers ROS production and prevents PKC activation. Whether increased ROS production in skeletal muscle initiates these processes remains to be determined. If such pathways are activated in skeletal muscle, it would be of interest to investigate whether UCP3 plays a role in decreasing or preventing their activation. An additional and as-yet-inadequately-explored possibility is a role for UCP3 in the context of hyperoxia, which is associated with oxidative stress and modified tissue function and tissue damage. UCP3 protein is higher in rats exposed to hyperoxia (105)
; however, the implications of this increase were not examined. The hyperoxic stimulus is of interest for at least two reasons: it is widely seen in the clinical setting, and it could be a useful tool to test hypotheses related to oxidative stress and UCP3.
|
Although evidence for the FA handling hypotheses is thus far correlative, it supports a specific association between UCP3 and improved FAO (see Lipid metabolism). An important implication is reduced levels of cytoplasmic lipid intermediates; to date only an association between UCP3 expression and reduced IMTG has been demonstrated (51
, 53
, 54)
. While increased IMTG is correlated with reduced insulin sensitivity (106)
, long-chain FA-coenzyme A units involved in IMTG breakdown may be more directly related to insulin resistance (106)
. Long-chain FA-coenzyme A interferes with insulin signaling via PKC activation (107)
. Therefore, a reduction in cytoplasmic lipids through enhanced FAO could be another process whereby UCP3 indirectly enhances insulin sensitivity and/or protects against insulin resistance. An increased capacity to oxidize FA could also lead to improved metabolic flexibility; metabolic flexibility refers to the ability of skeletal muscle to rapidly switch from predominantly carbohydrate oxidation during which FAO is suppressed to predominantly FAO (i.e., during fasting) (108)
. The hypothesis proposing UCP3-mediated export of FA anions from the matrix with the potential for reactivation and oxidation (13)
would provide a mechanism for improved metabolic flexibility. The UCP3 expression profile, with greater expression in type II fibers (see Patterns of modulated UCP3 expression), is also in line with this idea.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication August 10, 2006. Accepted for publication September 22, 2006.
|
REFERENCES |
|---|
TOP
ABSTRACT
INTRODUCTION
