Reduced skeletal muscle mitochondrial content and fatty acid oxidation are associated with obesity and insulin resistance. Although the exact mechanisms remain elusive, this may result from impaired mitochondrial biogenesis or reductions in the mitochondrial reticulum network. Therefore, the purpose of this study was to determine whether the protein contents of various transcription factors, including PGC-1α and PGC-1β and proteins associated with mitochondrial fusion events, were reduced in skeletal muscle of nine obese (BMI = 37.6 ± 2.2 kg/m−2) compared with nine age-matched lean (BMI = 23.3 ± 0.7 kg/m−2) women. The protein contents of PGC-1α, PGC-1β, PPARα, and tFAM were not reduced with obesity. In contrast, PPARγ was increased (+22%, P < 0.05) with obesity, and there was a trend toward an increase (+31%, P = 0.13) in PPARδ/β. In lean individuals, PGC-1α protein correlated with citrate synthase (CS; r = 0.67) and rates of palmitate oxidation (r = 0.87), whereas PGC-1β correlated with PPARγ (r = 0.90), PPARδ/β (r = 0.63), and cytochrome c oxidase IV (COX-IV; r = 0.63). In obese individuals, the relationship between PGC-1α and CS was maintained (r = 0.65); however, the associations between PGC-1α and palmitate oxidation (r = −0.38) and PGC-1β with PPARγ (r = 0.14), PPARδ/β (r = 0.21), and COX-IV (r = 0.01) were lost. In addition, mitofusin-1 (MFN-1), MFN-2, and dynamin-related protein-1 (DRP-1) total protein contents were not altered with obesity (P > 0.05). These data suggest that altered regulation, and not reductions in the protein contents of transcription factors, is associated with insulin resistance. Also, it does not appear that alterations in the proteins associated with mitochondrial network formation and degradation can account for the observed decrease in mitochondrial content.
- transcription factors
- fatty acid oxidation
- mitochondrial fusion/fission
skeletal muscle, due to its mass and ability to dispose of glucose, represents an important tissue in the development of insulin resistance. In recent years, impaired skeletal muscle fatty acid oxidation, and the concomitant increase in intramuscular diacylglycerol and ceramide contents, has been implicated in the development of insulin resistance and diabetes (reviewed in Ref. 45). Recent evidence indicates that the obesity-related impairments in fatty acid oxidation are associated with reductions in muscle mitochondrial content (14, 21), not with a dysfunction within mitochondria (5, 8, 14, 29) as has been suggested (19, 39). Although the exact mechanisms responsible for the observed reductions in mitochondria remain unknown, it is possible that decreases in the general transcription factors responsible for mitochondrial biogenesis may contribute to the diabetic phenotype as a result of a genetic predisposition and/or a sedentary lifestyle.
The coordination between mitochondrial DNA (which contains genes for 13 mRNA, 22 tRNA, and 2 rRNA) and nuclear DNA transcription is a key component of mitochondrial biogenesis. Peroxisome proliferator-activated receptor (PPAR)γ coactivator-1α (PGC-1α) has been considered the “master regulator” coordinating these events (38), because overexpressing PGC-1α resulted in increases in both nuclear and mitochondrial proteins (23). In addition, it has become apparent that PGC-1α coordinates these events by activating downstream transcription factors such as nuclear respiratory factor-1 (13, 49) and -2 (20, 40). PGC-1β, although less studied than PGC-1α, has also been implicated in the proliferation of mitochondria in the liver (26) and skeletal muscle (2), and its overexpression can prevent diet-induced insulin resistance (18).
Reductions in PGC-1α and PGC-1β have been linked to the development of insulin resistance in skeletal muscle given that 1) their mRNAs are decreased with diabetes (30, 34) and 2) single-nucleotide polymorphisms in PGC-1α (42) and PGC-1β (1) are prevalent in obesity and diabetes. However, reductions in PGC-1α and PGC-1β expression are not always observed with insulin resistance (31), and research in humans has relied largely on information obtained from microarray and/or gene chips and is therefore limited to mRNA, not protein measurements. In addition, the role of PGC-1α protein in obesity-related insulin resistance is also unclear. For example, 1) ablating PGC-1α unexpectedly improved glucose tolerance and insulin sensitivity in mice consuming a high-fat diet (24, 27), and 2) overexpressing PGC-1α in mice unexpectedly induced insulin resistance (28). Given the prominent roles of PGC-1α and PGC-1β, and their downstream targets [e.g., PPARs and mitochondrial transcription factor A (tFAM)], in regulating mitochondrial biogenesis in muscle, and therefore the capacity to oxidize fatty acids, it is important to further our understanding of the roles of these proteins in the development of insulin resistance.
Although reductions in mitochondrial content reduce the capacity of skeletal muscle to oxidize fatty acids (14), the capacity to oxidize fatty acids could also be compromised as a result of alterations in the mitochondrial morphology. Recently, the notion that mitochondria exist as a complex tubular network continually undergoing fission and fusion reactions has gained support. Several important regulatory proteins have been identified in these processes; dynamin-related protein-1 (DRP-1) and fission-1 (Fis-1) have been implicated in fission (41, 43), whereas mitofusin-1 (MFN-1), MFN-2, and autosomal dominant optical atrophy-1 have been suggested to coordinate fusion events (4, 50). Importantly, reductions in the mitochondrial network have previously been shown to decrease the cellular oxidation capacity (4). Thus, obesity-related reductions in the capacity of skeletal muscle to oxidize fatty acids could be due to a shift toward mitochondrial fission.
Given that obesity is associated with a decreased capacity of skeletal muscle to oxidize fatty acids (8, 14), the purpose of the current study was to determine whether 1) PGC-1α and PGC-1β protein contents were reduced in skeletal muscle of obese individuals, 2) the protein content of other transcription factors that are located downstream of PGC-1α and -β were reduced in the skeletal muscle of obese individuals, and 3) proteins that are suspected to influence mitochondrial reticulum formation were altered with obesity. Specifically, we hypothesized that obesity would be associated with reductions in the skeletal muscle protein contents of PGC-1α, PGC-1β, PPARγ, PPARδ/β, PPARα, and tFAM. In addition, we hypothesized that relationships between these transcription factors and a functional measure of skeletal muscle fatty acid oxidation would be lost with obesity. We also hypothesized that MFN-1 and MFN-2 fusion proteins would be reduced with obesity, whereas obesity would be associated with an increase in the content of the fission protein DRP-1.
MATERIALS AND METHODS
Nine lean [body mass index (BMI): 23.3 ± 0.7 kg/m−2] and nine obese (BMI: mean 37.6 ± 2.2 kg/m−2), age-matched (47 ± 3 vs. 45 ± 3 yr), sedentary, nondiabetic women participated in this study (fasting glucose: lean 4.1 ± 0.1, obese 4.5 ± 0.4 mM, P > 0.05; fasting insulin: lean 24.8 ± 1.9, obese 35.9 ± 4.9 pM; homeostasis model assessment: lean 0.66 ± 0.05, obese 1.00 ± 0.13, P < 0.05). We have previously published information on the same participants showing that palmitate oxidation is reduced in mitochondria extracted from skeletal muscle of obese individuals as a result of decreases in mitochondrial content (14). Subjects were admitted to McMaster Health Sciences Centre for abdominal surgery and gave informed written consent prior to participation. All experimental procedures were performed according to the declaration of Helsinki, and ethical approval was obtained from the McMaster University and University of Guelph Research Ethics Boards prior to recruitment. Individuals were screened and excluded from the study if any known diseases (other than obesity) or medications were reported.
Following an overnight fast (12–18 h), general anesthesia was induced with a short-acting barbiturate and maintained as required by a fentanyl and rocuronium volatile anesthetic mixture. A portion of the rectus abdominus muscle was sampled and immediately placed in ice-cold oxygenated buffer (modified Krebs-Henseleit buffer, containing 8 mM glucose) for transport to the laboratory.
Whole muscle crude homogenates were generated as described previously (17) and analyzed for total protein (BCA protein assay), and 40 μg of denatured protein was loaded for Western blotting. All samples were separated by electrophoresis on 8–12% SDS-polyacrylamide gels and transferred to polyvinylidene difluoride membranes. Commercially available monoclonal antibodies were used to detect PPARγ (R&D Systems, Minneapolis, MN), PPARα, PPARδ/β, MFN-1, MFN-2 (all from Abnova, Hornby, ON, Canada), DRP-1 (BD Biosciences, Mississauga, ON, Canada), citrate synthase (CS; Chemicon International, Temecula, CA), and cytochrome c oxidase complex IV (COX-IV; Invitrogen, Burlington, ON, Canada). Commercially available polyclonal antibodies were used to detect PGC-1α (Calbiochem, La Jolla, CA), PGC-1β (Abnova), and tFAM (Santa Cruz Biotechnology, Santa Cruz, CA). All samples for a given protein were developed on the same membrane to limit variation, and Ponceau staining was used to ensure consistent loading and transferring. Samples were run for each protein, and therefore, stripping of membranes was not required. Blots were quantified using chemiluminescence and the ChemiGenius 2 Bioimaging system (SynGene, Cambridge, UK).
Whole muscle palmitate oxidation.
We previously reported the mean whole muscle palmitate oxidation rates on the same nine lean and nine obese participants (14). However, in the current study we have used the individual values for the purposes of examining relationships with various transcription factors.
All data are presented as means ± SE. Differences between lean and obese participants were analyzed with unpaired t-tests. Associations between variables were investigated using Pearson product moment correlation analyses and multiple regression analysis as appropriate. Statistical significance was accepted at P < 0.05.
Effect of obesity on general transcription factor expression.
Obesity did not alter the skeletal muscle protein content of PGC-1α, PGC-1β (Fig. 1), PPARα, or tFAM (Fig. 2). In contrast, there was a 22% increase (P < 0.05) in PPARγ and a trend toward an increase in PPARδ/β (31%, P = 0.13) with obesity (Fig. 2).
Associations between PGC-1α, PGC-1β, and mitochondrial content.
When lean and obese individuals were grouped together, PGC-1α protein was significantly (r = 0.65, P < 0.01) correlated with CS, a common representative marker of mitochondrial content (Table 1). Independently, PGC-1α was also significantly correlated with CS protein in lean (r = 0.67, P = 0.05) and obese (r = 0.65, P = 0.05) individuals (Table 1). In contrast, PGC-1β protein did not correlate with CS when lean and obese participants were grouped together or when they were analyzed independently (Table 1). PGC-1α protein did not correlate with COX-IV when lean and obese participants were grouped together or when they were analyzed independently (Table 1). In contrast, PGC-1β protein correlated with COX-IV when lean and obese individuals were grouped together (r = 0.46, P = 0.05) and when lean individuals were analyzed independently (r = 0.63, P = 0.05; Table 1). However, this relationship was lost with obesity (r = 0.01, P = 0.98; Table 1).
PGC-1α and PGC-1β associations with palmitate oxidation.
When lean and obese individuals were grouped together, PGC-1α protein did not correlate with palmitate oxidation rates (Fig. 3). Independently, PGC-1α was significantly (r = 0.87, P < 0.01) correlated with palmitate oxidation in lean participants; however, no relationship was observed in obese individuals (Fig. 3). In contrast, PGC-1β protein content did not correlate with palmitate oxidation when lean and obese individuals were grouped together or when lean and obese groups were analyzed independently (Fig. 3).
Multiple associations between PGC-1α, PGC-1β, CS, and palmitate oxidation.
Multiple regression analysis did not reveal a significant relationship between PGC-1α, CS, and palmitate oxidation (r = 0.13, P = 0.88) when lean and obese individuals were analyzed together. In contrast, a significant relationship was observed in lean participants (r = 0.93, P < 0.01) that was lost with obesity (r = 0.42, P = 0.55). Multiple regression analysis did not reveal a relationship between PGC-1β, CS, and palmitate oxidation when lean and obese participants were analyzed together (r = 0.16, P = 84); however, lean participants again displayed a significant relationship (r = 0.96, P < 0.01) that was lost with obesity (r = 0.55, P = 0.33).
Associations between general transcription factors.
PGC-1α protein content did not correlate with PGC-1β when lean and obese individuals were grouped together (r = −0.15) or when lean (r = −0.23) and obese (r = −0.11) individuals were analyzed independently (Table 1). When lean and obese individuals were grouped together, PGC-1α protein did not correlate with the protein expression of PPARγ (r = −0.05) or PPARα (r = 0.35) but did correlate with the content of PPARδ/β (r = 0.48; Table 1). Independently, PGC-1α protein expression did not correlate with PPARγ, PPARα, or PPARδ/β in lean participants (Fig. 4 and Table 1). However, although PGC-1α remained unassociated with PPARγ in obese participants, PGC-1α protein did significantly correlate with both PPARα (r = 0.71) and PPARδ/β (r = 0.65) in these individuals (Fig. 4 and Table 1).
When lean and obese individuals were analyzed together, PGC-1β significantly correlated with PPARγ (r = 0.60), but not with PPARα (r = 0.48), whereas there was a trend toward a correlation with PPARδ/β (r = 0.43, P = 0.07; Table 1). Independently, PGC-1β protein significantly correlated with PPARγ (r = 0.90) and PPARδ/β (r = 0.63) in lean individuals, whereas there was a trend toward an association with PPARα (r = 0.48, P = 0.09) (Fig. 5 and Table 1). In contrast, PGC-1β did not correlate with PPARγ, -α, or -δ/β in obese individuals (Fig. 5 and Table 1).
Effect of obesity on proteins associated with mitochondrial fission/fusion events.
Obesity did not alter the expression of MFN-1, MFN-2, or DRP-1 (Fig. 6). The net difference between DRP-1 and MFN-2 protein content, an indication of the overall net mitochondrial reticulum fragmentation, showed only a trend (−28%, P = 0.13) toward a decrease with obesity (47.1 ± 9.5 vs. 34.1 ± 3.9 for lean and obese, respectively).
When lean and obese individuals were analyzed together, MFN-2 significantly correlated with COX-IV (r = 0.50). Independently, in lean individuals MFN-2 protein significantly (P < 0.01) correlated with COX-IV protein (r = 0.82), and this relationship was lost with obesity (r = 0.42).
Impairments in fatty acid oxidation have been associated with obesity and the development of insulin resistance and are believed to result from decreases in either the content or the function of mitochondria. Although the exact mechanisms remain elusive, it has been hypothesized that reductions in general transcription factors associated with mitochondrial biogenesis may be associated with these observations. The novel findings of this study are that 1) obesity did not decrease the protein content of PGC-1α, PGC-1β, PPARα, PPARδ/β, or tFAM, 2) compensatory increases in PPARγ occurred with obesity, 3) PGC-1α correlated with palmitate oxidation in lean individuals only, and 4) PGC-1β correlated with PPARγ, PPARδ/β, and COX-IV in lean individuals only. The latter two findings suggest that the coordinated regulation of selected transcription factors involved in mitochondrial biogenesis was lost with obesity. A final novel finding was that 5) MFN-1, MFN-2, and DRP-1 were not altered with obesity, suggesting that reductions in the formation of a mitochondrial reticulum were not responsible for reductions in mitochondrial function and fatty acid oxidation in this population.
PGC-1α and PGC-1β antibodies.
Recent experiments in our laboratory indicate that we are accurately measuring PGC-1α. Specifically, 1) we can detect proportional changes in PGC-1α mRNA and protein following electrotransfection of PGC-1α cDNA into mammalian muscle cells (Supplemental Fig. 1; Supplemental Material for this article is available at the AJP-Endocrinology and Metabolism web site) (6); 2) following electrotransfection, we observe proportional increases in downstream targets of this cofactor, including increases in CS enzymatic activity and mitochondrial DNA (6); 3) we observe fiber-type differences in PGC-1α protein (red muscle > white muscle) (6); and finally, 4) we observe positive correlations between PGC-1α protein and various downstream targets [carnitine palmitoyltransferase I (r = 0.96) and tFAM (r = 0.99)] in skeletal muscle (6). In addition, in the present study, PGC-1α and PGC-1β proteins did not correlate with each other (r = −0.15; Supplemental Fig. 2). Taken together, our work strongly indicates that our measurements are specific for PGC-1α and PGC-1β.
PGC-1α relationship with obesity.
For several years it has been known that many transcription factors regulate genes involved in lipid metabolism, and PGC-1α has been considered the master regulator coordinating these genes (38). As a result, the concept of reduced PGC-1α expression as an explanation for the observed reduction in fatty acid oxidation with insulin resistance has gained attention in recent years, especially following the reports of reduced PGC-1α mRNA expression in this population (30, 34). The current data do not support this interpretation, given that PGC-1α protein was not reduced in the skeletal muscle of obese individuals, whereas the muscle mitochondrial content was markedly reduced in these individuals (14). Although these data are contrary to the majority of the literature regarding mRNA, PGC-1α mRNA changes are only weakly associated with changes in PGC-1α protein (47), and there is only a very modest relationship between skeletal muscle insulin sensitivity and PGC-1α mRNA (33). However, in support of our findings, a recent publication reported no change in either PGC-1α mRNA or total protein with obesity (10). It is possible that part of the controversy may revolve around the confounding variable of physical activity, because a sedentary lifestyle typically associated with obesity could account for the previously reported reductions in PGC-1α. Although information on the fitness of the current participants was not known, when fitness levels are accounted for, an impairment in the activation of PGC-1α is still reported (10).
In the current study, we chose to examine relationships between PGC-1α and whole muscle palmitate oxidation and not oxidation rates normalized to mitochondrial protein or CS for two reasons: 1) the various transcription factors measured exert their effects by inducing mitochondrial proliferation, and therefore, mitochondrial content is an important variable that would be negated if palmitate oxidation values were normalized to CS or mitochondrial protein; and 2) we (14) and others (8, 29) have shown previously that mitochondria isolated from these individuals contain no dysfunction in terms of the ability to oxidize palmitate; they simply have less mitochondria. The impairment in fatty acid oxidation is present only at the whole muscle level, and this represents the more physiological representation of overall muscle fatty acid oxidation. Therefore, we believe that the ability of the whole muscle to oxidize fatty acids is more important than “normalized” mitochondrial oxidation rates.
The relationship between PGC-1α and palmitate oxidation that we observed in lean participants was not present with obesity. This suggests that the regulation of PGC-1α may be altered with obesity. De Filippis et al. (10) have recently reported an attenuation in the rise of PGC-1α mRNA and protein following an acute bout of exercise in insulin-resistant individuals, further suggesting that the regulation of PGC-1α is altered in this population. These observations underscore the complexity of mechanisms that regulate protein expression, including the influence of phosphorylation on increasing PGC-1α mRNA stability (37) or activation (16, 25), the ability of other proteins to suppress PGC-1α activity (11), and the importance of nuclear import (48). Puigserver et al. (36) have proposed a three-step model of activation that includes interacting and docking with transcription factors, undergoing a conformational change enabling binding of additional cofactors, and subsequent induction of transcription. Clearly, this is a complex process, and caution should be taken when interpreting data that provide insight into only one of these three steps. However, the current data suggest that the reduction in mitochondrial content associated with obesity does not involve diminished PGC-1α protein and may be located downstream of PGC-1α and involve activation or repression.
PGC-1β relationship with obesity.
PGC-1β, although slightly larger than PGC-1α (1,023 vs. 798 amino acids), contains conserved regions, and as such the two proteins share many target genes. Although less is known about PGC-1β, recent “gain of function” studies have reported the ability of PGC-1β to induce mitochondrial proliferation similarly to PGC-1α (18). The present data suggest that PGC-1β protein is not reduced in skeletal muscle of obese individuals. However, like PGC-1α, PGC-1β may contribute to the obesity phenotype as a result of a dysfunction in the regulation of the activity of this cofactor. For example, although PGC-1β had significantly positive relationships with PPARγ, PPARδ/β, and COX-IV in lean individuals, these relationships were lost with obesity. Together, these data may suggest that PGC-1β is regulated posttranslationally, but the mechanisms have not yet been defined. Alternately, mutations within PGC-1β may explain these results, because constant protein content would not reflect activity or transcriptional capacity. Importantly, PGC-1β mutations have been associated with reductions in the expression of oxidative phosphorylation genes and mitochondrial dysfunction in skeletal muscle of mice (46). In addition, single-nucleotide polymorphisms within PGC-1β have been associated with diabetes in human skeletal muscle (1).
The present data suggest that skeletal muscle PGC-1α and PGC-1β are regulated differently, since they do not correlate with each other in either lean or obese individuals. This is consistent with the literature, since skeletal muscle PGC-1α mRNA increases in response to exercise and decreases in response to denervation, whereas PGC-1β mRNA is unaltered in either condition (22). Interestingly, in the present study PGC-1α correlated with PPARα and PPARδ/β in obese but not lean individuals. In marked contrast, PGC-1β correlated with PPARγ and PPARδ/β in lean but not obese individuals. These data further suggest that the regulation of PGC-1α and PGC-1β is different; however, the importance of their distinct relationships with the various PPARs in obesity remains unknown. Importantly, the relationship between PGC-1α and one of its target proteins, CS (23), is retained with obesity, suggesting that only specific associations are altered.
PPAR relationships with obesity.
It was hypothesized that all three PPARs would be reduced with obesity; however, PPARγ was increased +22%, and although not significant, PPARδ/β was increased +31% with obesity, whereas PPARα remained unchanged. Although these observations are contrary to our hypothesis, it has been shown recently that raising plasma fatty acid concentration for 4 wk can increase PPARα mRNA (12). These observations suggest that high plasma fatty acid concentrations, typically associated with obesity, can increase the content of various PPARs. Although this remains a plausible explanation, in the current study the fasting free fatty acid concentrations were not higher in the obese individuals, suggesting that other unknown mechanisms may account for the upregulation of PPARs. However, it should be mentioned that fasting free fatty acid concentrations may not accurately reflect skeletal muscle free fatty acid exposure rates during the day, since this measurement does not account for diet, nor does it account for the greater rate of palmitate uptake into skeletal muscle of obese individuals (7). As with PPAR cofactors, PPARs can also be externally regulated (15). Raising plasma fatty acid concentrations for 4 wk increased the binding of PPARδ/β to the promoter of carnitine palmitoyltransferase I, suggesting an increased activation (12). The trend toward an increase PPARδ/β protein content observed in the current study, and the increased activation previously reported with high fatty acid concentrations (12), suggests that PPARδ/β in skeletal muscle of obese individuals increases to compensate for impairments in lipid oxidation. In support of this suggestion, when PPARδ/β was pharmacologically activated, it attenuated the metabolic syndrome by increasing fatty acid oxidation and basal metabolic rate, decreasing weight gain and adipose tissue size, and ultimately improving glucose and insulin tolerance (44).
In addition, a compensatory increase in PPARγ was observed in the present study. However, the relevance of this finding is unknown because the expression of PPARγ in skeletal muscle has previously been shown to be low in relation to the other PPARs (32), and this transcription factor is considered to play only a primary role in adipose tissue (reviewed in Ref. 9). Nevertheless, PPARγ may play a prominent role in skeletal muscle. Interestingly, when PPARα is knocked down and fatty acid oxidation is compromised, a compensatory increase in PPARγ is observed in cardiac muscle (32). In the current study, PPARγ was the only PPAR that did not correlate with either PGC-1α or PGC-1β in obesity, further suggesting that dysregulation of this transcription factor may be important. Therefore, it appears that, in skeletal muscle, the PPARs may be upregulated in response to the impairments in lipid oxidation associated with obesity.
Mitochondrial fission/fusion proteins with obesity.
In addition to the apparent altered coordination of mitochondrial transcription factors, reductions in the capacity of mitochondria to oxidize substrates could result from decreases in the mitochondrial reticulum network. Mitochondria appear to exist as complex tubular networks, as opposed to individual organelles, that continually undergo fission and fusion reactions. Although the exact mechanisms responsible for these events remain elusive, fission appears to be promoted by DRP-1 and Fis-1 (41, 43), whereas fusion is likely facilitated by MFN-1, MFN-2, and autosomal dominant optical atrophy-1 (4, 50). Genetic modifications that alter the content of MFN-2 have been shown to reduce the rate of mitochondrial fusion and decrease the ability of mitochondria to oxidize substrates (4). It is currently believed that an intrinsic dysfunction within mitochondria is not responsible for reductions in whole muscle fatty acid oxidation rates in obese individuals (14, 29). However, it is possible that the isolation techniques used in these studies altered the reticulum network and thus may not reflect the capacity of mitochondria to oxidize fatty acids in vivo. It has previously been reported that MFN-2 mRNA and protein are reduced with obesity (3, 4), although this may not reflect a movement away from fusion since reductions in fission proteins could result in no change in the balance between these dynamic processes.
This study reported that the total muscle protein contents of MFN-1, MFN-2, and DRP-1 were not different with obesity compared with age-matched lean controls. Although we used a monoclonal antibody to detect MFN-1, we detected a doublet with an upper band at ∼86 kDa and a lower band at ∼70 kDa. Based on the known amino acid sequence of MFN-1, the predicted molecular weight is ∼86 kDa. Therefore, we have quantified the upper band for the purposes of this study. However, it should be noted that when the lower band was quantified there was also no difference with obesity. These data also clearly showed that MFN-2 was not reduced with obesity, since neither the mean nor the range (lean: 32.1–49.6 vs. obese: 30.8–50.9 arbitrary units) of MFN-2 protein content was different in lean and obese individuals. Although this is in contrast to a previous report showing a negative relationship between BMI and MFN-2 mRNA (3), this likely represents a difference between mRNA and protein responses. However, Bach et al. (4) reported that MFN-2 protein is significantly reduced (∼40%) with obesity, which is more difficult to explain. However, this may reflect a technical difference since MFN-2 protein was measured in muscle homogenates in the current study, whereas Bach et al. conducted measurements in isolated mitochondria. Although differences between individual fission/fusion proteins were not evident in the current study, collectively, fusion proteins increased nonsignificantly (∼12%), whereas fission proteins decreased slightly (−11%) with obesity. Since mitochondrial tubular networks are continually changing, a shift in the balance between fission and fusion is required to infer alterations in a reticulum. The net difference between DRP-1 and MFN-2, which provides a measure of the balance between fission and fusion, showed a trend toward a decrease with obesity, possibly indicating compensation toward creating a larger mitochondrial reticulum, decreasing mitochondrial degradation, and improving fuel oxidation.
In addition, independent of its role in fusion, MFN-2 has been shown to function as a signal that induces the expression of oxidative phosphorylation genes (35). Importantly, the current data found that MFN-2 was associated with COX-IV protein in lean individuals, a relationship that was lost with obesity. These data further suggest a dysregulation of transcription factors with obesity.
In summary, the novel findings of the current study are that obesity is not associated with reductions in the protein contents of PGC-1α or PGC-1β in skeletal muscle and that the PPARs (PPARγ and potentially PPARδ/β) may be upregulated in response to the impairments in lipid oxidation associated with obesity. The relationship between PGC-1α and palmitate oxidation as well as the relationships between PGC-1β and PPARγ, PPARδ/β, and COX-IV were also lost with obesity. These data suggest that the regulation of PGC-1α and PGC-1β, and not total protein, is altered with obesity. In addition, MFN-1 and MFN-2 proteins were not reduced with obesity, and DRP-1 was not increased with obesity, suggesting that increased mitochondrial reticulum fragmentation does not precede the development of insulin resistance.
This research was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC; to A. Bonen, D. J. Dyck, and L. L. Spriet) and by the Canadian Institutes of Health Research (to A. Bonen, D. J. Dyck, G. J. F. Heigenhauser, and L. L. Spriet). A. Bonen is also the recipient of a Canada Research Chair in Metabolism and Health, and G. P. Holloway is the recipient of an NSERC Canadian Graduate Scholarship.
We acknowledge the contributions of the surgeons Dr. Margaret Lightheart, Dr. Caroline Sibley, and particularly the late Dr. Lawrence Loopstra.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 by American Physiological Society