The aim of this study was to determine if adaptations to skeletal muscle mitochondrial FAO, enzyme activity, morphology, and transcriptional responses were differentially regulated in skeletal muscle from rats that are protected or susceptible to HFD-induced insulin resistance due to selective breeding for high (HCR) or low endurance running capacity (LCR), respectively. As expected, the HFD worsened insulin resistance and increased adiposity in the LCR, while the HCR were protected from these changes; however, several unexpected findings in skeletal muscle emerged. On a NC diet, the HCR had a 3-fold higher mitochondrial FAO in skeletal muscle despite the LCR having significantly higher PGC-1α gene expression and increased mitochondrial area measured by electron microscopy. Both groups increased mitochondrial FAO on the HFD eliminating between strain differences; however, only the LCR had a decrease in PGC-1α gene expression. Contrary to our hypothesis, the HCR did not display a change in PGC-1α or PPARδ expression nor did our other enzyme and electron microscopy mitochondrial measures capture a significant difference between strains on a NC diet or in response to the HFD.
Previous studies have shown that insulin resistant obese humans have reduced skeletal muscle and whole body FAO (Kelley, Goodpaster et al. 1999
; Kim, Hickner et al. 2000
; Hulver, Berggren et al. 2003
; Thyfault, Kraus et al. 2004
; Hulver et al. 2005
) which may contribute to both expanding adiposity and an increased storage of lipids in muscle. These data in combination with the intensely studied link between insulin resistance and elevated lipids storage in skeletal muscle led to the lipotoxicity hypothesis, which contends that the increased storage of intramuscular lipid metabolites (diacylglycerol, fatty acyl-CoAs, and ceramides) impairs insulin action (Muoio et al. 2006
). In contrast to the reduced FAO that has been reported in human obesity, rodent models of obesity and insulin resistance have higher skeletal muscle FAO than non-obese controls but still possess elevated intramuscular lipids (Turcotte, Swenberger et al. 2001
; Thyfault et al. 2007
; Holloway, Benton et al. 2009
Newer studies have utilized HFD-induced insulin resistance to examine changes in skeletal muscle mitochondrial function and FAO as insulin resistance develops. Turner et al (Turner, Bruce et al. 2007
) found that both 5 and 20 weeks of a HFD-induced both increased FAO in muscle homogenate and reduced insulin sensitivity with a glucose tolerance test in mice. In addition, Hancock et al (Hancock et al. 2008
) found that a HFD increased several markers of mitochondrial content and function in parallel with worsening insulin resistance in rats. In relation this, Koves and Muoio et al (Koves, Li et al. 2005
; Koves, Ussher et al. 2008
) have developed an alternative hypothesis that insulin resistance may be linked to an excess entry of fatty acids into the mitochondria and increased incomplete fatty acid oxidation which results in oxidative stress or other mitochondrial derived metabolites which impair insulin action. This paradigm is supported by their work showing that blocking fatty acid entry into the mitochondria by knocking out malonyl CoA decarboxylase activity in mice or in cultured muscle cells prevents HFD or lipid-induced insulin resistance, respectively (Koves, Ussher et al. 2008
). In contrast to the report from Turner et al (Turner, Bruce et al. 2007
) we found that the HCR rats maintained insulin sensitivity, as measured by a glucose tolerance test. It could be hypothesized that the elevated FAO in the HCR prior to the initiations of the HFD provided protection against HFD-induced insulin resistance and adiposity. This association was clearly shown in the HCR rats in this study and in a previous report in the model (Noland, Thyfault et al. 2007
). Importantly, both the HCR and LCR groups increased both incomplete (acid soluble metabolite production) and complete FAO to CO2
(data not shown) in response to the HFD suggesting that mismatch between mitochondrial fatty acid entry and complete oxidation did not play a role in the development of insulin resistance in the LCR. In support of the concept that high mitochondrial FAO may provide protection it is documented that animals given daily exercise are protected against HFD-induced insulin resistance assumedly due in part to enhanced lipid handling and utilization (Koves, Li et al. 2005
; Bradley et al. 2008
). Additional studies in which FAO is elevated by overexpression of CPT-1 in skeletal muscles of rats (Bruce et al. 2009
), in muscle cells, (Perdomo et al. 2004
) or by knockout of specific genes (ACC−/−
) (Abu-Elheiga et al. 2003
) has also been shown to protect against lipid induced insulin resistance.
An outcome that is difficult to interpret is the mismatched response between mitochondrial FAO and mitochondrial enzymes in response to the HFD. Both groups increased mitochondrial FAO in response to the HFD but neither group showed a significant increase in the activity of mitochondrial enzymes. This could be attributed to FAO being measured in equal contents (per ug protein) of isolated mitochondria while the mitochondrial enzyme assays were studied in muscle homogenates. A more likely rationale is that FAO is a biological measure that calls on a series of mitochondrial enzymes to work in unison to take fatty acids through transport into the mitochondria, beta oxidation, and the TCA cycle, while enzyme assays are measuring one part of the pathway in maximal conditions. Thus, it is possible that the function of enzymes not measured (acyl-CoA-synthetase or CPT-1 for example) may have increased in activity and contributed to the increased FAO response, while the mitochondrial enzymes measured in this report were already working at a high enough activity and did not need to increase in response to the HFD.
The effects of high fat feeding on skeletal muscle transcription factors and co-activators that mediate mitochondrial biogenesis, function, and lipid metabolism have also been intensively examined. The most highly studied factors include PGC-1α, a vital co-activator of nuclear receptors, that plays a large role in the control of mitochondrial biogenesis and oxidative genes (Lin et al. 2005
). Importantly, reduced PGC-1α expression in muscle has been previously linked to insulin resistance and type 2 diabetes (Mootha, Lindgren et al. 2003
; Patti, Butte et al. 2003
; Mensink, Hesselink et al. 2007
) while newer studies have refuted this connection (Hancock, Han et al. 2008
). The role of the PPAR family of nuclear receptors, which are activated by fatty acids and exert transcription control of lipid oxidation and synthesis pathways have also been heavily examined. PPARδ is believed to provide a coordinated response to high fat feeding that will increase the disposal of fatty acids (Muoio et al. 2002
; Barish et al. 2006
). Surprisingly in this study we witnessed a decline in PGC-1α gene expression in the LCR after the HFD, and we also found no change in PPARδ expression between strains or diets. Previous studies examining PGC-1α mRNA have been unequivocal with a HFD causing no change (McAinch et al. 2003
; Hancock, Han et al. 2008
), or a decrease in muscle PGC-1α (Koves, Li et al. 2005
; Sparks, Xie et al. 2005
; Crunkhorn et al. 2007
). Many studies including this one, have only assessed PGC-1α gene expression because of difficulty with commercially available antibodies, making comprehension of these findings more difficult due to transcriptional and post-translational regulation of PGC-1α. A recent study found that a 4 week HFD doubled skeletal muscle PGC-1α protein in rats (Hancock, Han et al. 2008
), while Sparks et al (Sparks, Xie et al. 2005
) showed a 3 week HFD reduced PGC-1α protein in skeletal muscle of mice by 40%. The discrepancies between studies could be due to different types of high fat feeding (varying macro-nutrient composition and diverse lipid sources), different species, different types of skeletal muscle tissues, and perhaps different antibody specificity. In addition, studies are needed to determine if a HFD alters the acetylation of PGC-1α, which has been shown to significantly change its co-activating capacities towards nuclear receptors (Rodgers et al. 2008
). Thus, the meaning of the PGC-1α mRNA data reported here should be interpreted with caution. As for the lack of a change in PPARδ responses, it is possible that initial changes in expression occur early on but are lost as the HFD continues and a steady state condition ensues, but there is no evidence to support that in the current study. In addition, we did not detect a significant difference in plasma free fatty acids as a result of the HFD in either group in fasting conditions, although it is still probable that increased circulating free fatty acids would have been detected in HFD animals during the fed state.
There are limitations to this study including the probability that our analysis of only a red-oxidative muscle (red gastrocnemius) and not skeletal muscle with a mixed or white-glycolytic fiber type significantly impacted many of our results. Because red muscle fiber already has high oxidative capacity its ability to have a HFD induced transcriptional response or increase in mitochondrial enzyme activities may be blunted due to a ceiling effect. However, the ability for both groups to increase mitochondrial FAO in the red gastrocnemius in response to the HFD refutes this concept. That being said, it is certainly possible that different responses would have been witnessed in white or mixed muscle fibers and thus it should be emphasized that the results reported here are specific to an oxidative muscle fiber type. An additional limitation is that we did not attempt to monitor or control for the estrous cycle in the female rats which could have had an impact on circulating estrogen levels which is known to modulate lipid metabolism.
Mitochondrial reactive oxygen species (ROS) production also has been linked to skeletal muscle insulin resistance (Abdul-Ghani, Jani et al. 2009
; Anderson et al. 2009
). The m-SOD enzyme putatively provides protection against ROS induced cellular damage by converting superoxide into hydrogen peroxide. As such, m-SOD activity typically increases when tissues are subjected to stressors such as intensive exercise or a HFD (Powers et al. 1999
; Anderson, Lustig et al. 2009
). In addition, SOD expression is putatively controlled by PGC-1α (St-Pierre, Lin et al. 2003
), thus we sought to determine how SOD activity measured in isolated mitochondria would respond to a HFD in the HCR and LCR skeletal muscle. On the NC, the LCR displayed higher m-SOD activity than the HCR, suggesting that the LCR mitochondria are already fending against a higher level of oxidative stress; however, m-SOD activity significantly declined in the LCR rats after the HFD. The reduction in SOD activity in the LCR would likely lead to a greater exposure to mitochondrial derived ROS production which has been linked to reduced insulin sensitivity (Anderson, Lustig et al. 2009
) and the propagation of mitochondrial dysfunction due to mitochondrial DNA mutations (Hiona et al. 2008
). In contrast, the HFD induced increase in m-SOD in the HCR would act to protect against cellular damage and insulin resistance.
The LCR unexpectedly showed a trend for greater mitochondrial size than the HCR despite being bred over several generations for lower endurance running capacity and possessing lower significantly lower FAO per mitochondrial protein (on NC diet), lower cytochrome c protein content, and a trend for lower citrate synthase and β-had enzyme activity on the normal chow diet. This led us to speculate that the LCR muscle may possess swollen, less functional mitochondria than their HCR counterparts. Previous studies have shown that reduced mitochondrial function is associated with round, swollen morphology, and less visible mitochondrial matrix (Ibdah et al. 2005
; Lim et al. 2009
). Electron microscopy images of skeletal muscle mitochondrial in the LCR rats did not reveal an overt morphological phenotype of swollen or round mitochondria. However, the ratio of cytochrome c to total mitochondrial area, an estimate of mitochondrial protein per mitochondrial area and thus an estimate of functional quality, was significantly higher in the HCR rats than the LCR rats on both the HF and NC diets. Additionally, this ratio increased in the HCR rats with high fat feeding. This ratio of mitochondrial protein to area suggests that a slightly larger mitochondrial area in the LCR is not associated with improved function. These association, particularly after the HFD should be taken with caution as the mitochondrial enzyme measures of citrate synthase and β-had tended to decrease in the HCR group after the HFD, which go in the opposite direction of the HFD induced effect to increase the cytochrome c to mitochondrial area ratio. Mitochondrial respiration studies are needed to confirm if the HCR do possess a higher quality mitochondria than the LCR.
In conclusion, our results show that both HCR and LCR rats have elevated mitochondrial FAO after a HFD, despite the HCR being protected and the LCR being susceptible to HFD induced insulin resistance. In addition, the HCR skeletal muscle did not show significant adaptations in mitochondrial content or morphology measures, or in transcriptional responses of PGC-1α or PPARδ as hypothesized. However, the HCR did show a possible response for higher mitochondrial quality as evidenced by an increased ratio of cytochrome C per mitochondrial area. It is possible that higher skeletal muscle mitochondrial content and FAO prior to initiation of the HFD in addition to HFD-induced increase in m-SOD and indices of mitochondrial area played a role in protecting against insulin resistance. Moreover, it is also plausible that the HCR protected insulin sensitivity because they did not have a HFD induced increase in adiposity like the LCR. All told, these results point to the powerful health implications of intrinsic aerobic capacity on both skeletal muscle and whole body metabolic health and function.