| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1Sanofi-Aventis Deutschland, Frankfurt am Main, Germany; 2Medical Department, Hanusch Hospital; and 3Division of Endocrinology and Metabolism, Department of Internal Medicine III, Medical University of Vienna, Vienna, Austria
Submitted 21 September 2005 ; accepted in final form 18 December 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT RESEARCH DESIGN AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
insulin resistance; intramyocellular lipids; fatty acids; muscle fiber type; Zucker diabetic fatty rat; lipid metabolism
The molecular basis for the interaction of glucose and fatty acid metabolism is still not completely elucidated. In addition to Randle et al.'s (42) glucose-fatty acid cycle, whose entire validity has been repeatedly questioned (46, 47, 53), other mechanisms were proposed to explain how fatty acids impair glucose metabolism. Muscle triglycerides themselves do not seem to interfere directly with insulin action but reflect rather some other lipid-dependent changes resulting in IR (9, 35). Good candidates are presently metabolites derived from intramyocellular fatty acids, like long-chain fatty acyl-CoAs (LCACoAs; see Refs. 9, 18, 61), which are thought to activate directly or indirectly via diacylglycerol (19, 38)-specific protein kinase C isoforms (19, 61), ultimately resulting in the inhibition of insulin signaling molecules by serine phosphorylation and impairment of glucose uptake (53).
Whether skeletal IMCL levels and IR are causally connected remains unclear. The failure to clarify this relationship is probably also related with the often used generic concept "skeletal muscle," which fails to account for the large functional and metabolic variability of muscle tissue. The rat has muscles with highly specific fiber types and thus offers the chance to study their specific response.
Fatty acid metabolism in skeletal muscle is far from being homogenous when different fiber types are compared with respect to biochemical and physiological properties. Skeletal muscle comprises the following four different fiber types: one slow-twitch fiber type (type I, oxidative) and two fast-twitch fiber types (type IIA, oxidative/glycolytic and type IIB, nonoxidative/glycolytic), as well as an additional fiber type named IID/X, which is in between type IIA and IIB fibers (11). During a 5-day fed-starvation-refed study in Wistar rats, the elevation of plasma free fatty acids (FFA) led to muscle type-specific responses in intramyocellular fatty acid metabolism (39); in nonoxidative muscles, IMCL levels increased significantly during starvation, whereas in the oxidative soleus muscle IMCL levels remained constant. Similar muscle type-specific differences in starvation-induced dynamics of IMCL have recently been demonstrated also in humans (58). In our study with Wistar rats (39), in the glycolytic longissimus dorsi muscle, LCACoA levels remained constant during the starvation period. From these findings it was concluded that, during starvation-induced adipocytic lipolysis, oxidative muscle disposes elevated FFA by oxidation, whereas nonoxidative muscles neutralize FFA by reesterification. Both mechanisms might prevent the impairment of insulin signaling by keeping the levels of LCACoAs low.
The present study was performed to characterize dynamics of fatty acid metabolites, in relationship with IR. We chose the adolescent male obese Zucker diabetic fatty (ZDF) rat as the animal model displaying an insulin-resistant prediabetic condition. The muscles white tibialis anterior (wTA, predominantly type IIB fibers), extensor digitorum longus (EDL, mainly type IIB fibers), and soleus (predominantly type I fibers) were selected to cover the spectrum from glycolytic to oxidative fiber type. In these three muscles, IMCL levels and dynamics during fed-to-starved-to-refed transition were investigated and correlated with levels of malonyl-CoA, with individual types and levels of LCACoAs, and with relevant plasma parameters (glucose, insulin, triglycerides, FFA, and ketone bodies) in an integrated effort to identify key differences between insulin-resistant and -sensitive metabolism.
|
RESEARCH DESIGN AND METHODS |
|---|
|
TOPABSTRACTRESEARCH DESIGN AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental protocol. All experimental procedures were conducted according to the German Animal Protection Law. The study protocol included groups of lean and obese ZDF rats; the study ran for 5 days and consisted of the following study points: normal fed; 24, 48, and 72 h starvation; and refed. At each study point, body weight and metabolic serum parameters were measured.
MRS groups. IMCL values were determined repeatedly in all animals (n = 8 lean and obese) by in vivo 1H-MRS in wTA (glycolytic), EDL (intermediary), and soleus (oxidative).
Satellite groups. In parallel groups (n = 8 lean and obese for each study point), defined tissue samples were collected as follows: wTA, EDL, and soleus muscles of both hindlegs were isolated and immediately frozen in liquid nitrogen. In these samples, malonyl-CoA and total creatine (tCr) levels were determined. LCACoA levels in these three muscles were measured in additional rats at the study points mentioned in Experimental protocol (except the 48-h time point). Furthermore, insulin sensitivity was determined with the euglycemic-hyperinsulinemic glucose clamp in additional rats (n = 6 lean and obese) both after 24 and 72 h of starvation.
MRS.
In vivo MRS studies in anaesthetized rats were performed on a 7 Tesla Biospec system (Bruker BioSpin, Ettlingen, Germany) as described previously (27). Briefly, IMCL was observed using a single voxel spectroscopy sequence. IMCL was determined as IMCL-to-tCr-ratio in the three muscles. The validity of this approach was checked by biochemical determination of tCr in muscle tissue samples. There was a difference of 
35% in the tCr content of soleus vs. wTA. Therefore, IMCL values were normalized to biochemically determined tCr in wTA.
Euglycemic-hyperinsulinemic glucose clamp study. To investigate insulin sensitivity in lean and obese rats, euglycemic-hyperinsulinemic glucose clamp studies with infusion rates of 9.6 mU·kg–1·min– insulin and 1 mg·kg–1·min–1 [U-13C]glucose were performed under pentobarbital sodium anesthesia as described previously (39) after both a 24-h and a 72-h period of starvation.
Monitoring of metabolic serum parameters. Blood samples were taken from the tip of the tail for determination of blood glucose. Samples for monitoring of FFA, ketone bodies, triglycerides, and insulin were obtained by puncture of the sternal venous plexus during short-term isoflurane anesthesia. Blood samples were obtained from all groups (MRS and satellite groups). Representative values are shown from the MRS groups.
Analytical procedures of blood parameters. Standard procedures were used to determine blood glucose, FFA, ketone bodies, and triglycerides (3). Plasma insulin concentrations were assayed by RIA (Linco).
HPLC analysis of malonyl-CoA, tCr, and LCACoAs. HPLC analysis was performed using a Waters Alliance 2690 system (2487 detector; Millennium 2010 chromatographic manager).
Measurements of malonyl-CoA and tCr in muscle were performed in freeze-dried tissue as described previously, whereas analysis of LCACoAs was slightly modified (39).
Tissue was homogenized, and C17:0-CoA was added as an internal standard. LCACoAs were then extracted from the tissue. After centrifugation, the supernatant was applied to a solid-phase extraction column. An aliquot of the resulting eluate was then applied to the HPLC column. The gradient system included three mobile phases, as follows: KH2PO4 (25 mmol/l, pH 4.9), acetonitrile (ACN), and methanol. The starting conditions were 40% KH2PO4, 30% ACN, and 30% methanol. The flow rate was set at 0.4 ml/min. Over 30 min, ACN increased to 50%, whereas KH2PO4 decreased to 20% and methanol remained constant at 30%. The sum of the major species (C16:0, C16:1, C18:0, C18:1, C18:2, and C20:4) was referred to as total LCACoAs, the sum of C16:0, C16:1, C18:0, and C18:1 was referred to as saturated plus monounsaturated (SAT + MUFA), and the sum of C18:2 and C20:4 was referred to as polyunsaturated (PUFA) LCACoAs.
Enzymatic characterization of muscle tissue.
To characterize the individual muscles biochemically, the maximal activity of four enzymes involved in energy metabolism was determined in tissue extracts of fed rats: glycogen phosphorylase (GPase, EC 2.4.1.1
[EC]
) as an indicator for glycogen breakdown; hexokinase (HK, EC 2.7.1.1
[EC]
), as such for glucose transport and phosphorylation; glyceraldehyde-3-phosphate dehydrogenase (GAPDH, EC 1.2.1.12
[EC]
) and 
-hydroxyacyl-CoA dehydrogenase (HOAD, EC 1.1.1.35
[EC]
) as markers for glycolysis and mitochondrial fatty acid oxidation. Enzymatic activities were assayed according to standard procedures (3).
Statistical analysis. Data are presented as means ± SE. Statistical differences between lean and obese rats at the different study points were determined using a one-way ANOVA followed by a post hoc analysis with Bonferroni's correction. When testing for differences between the different study points, a two-factor ANOVA (for repeated measures) followed by a post hoc analysis with Bonferroni's correction was used (software package SigmaStat; Jandel, Erkrath, Germany). The differences in the relationship of GAPDH-to-HOAD and GPase-to-HK-ratio between lean and obese animals were tested using a one-way ANCOVA (SPSS; SPSS, Chicago, IL). P < 0.05 was considered to be statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTRESEARCH DESIGN AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
Malonyl-CoA. In the fed state, malonyl-CoA levels for every muscle type were similar in both strains (Fig. 2). In the obese, malonyl-CoA levels were significantly elevated in their wTA but reduced in their EDL. In soleus, malonyl-CoA levels did not vary throughout the study in either group. During starvation, the lean group showed larger malonyl-CoA changes than the obese, and the changes increased with rising glycolytic capacity. In the lean rats, malonyl-CoA in wTA and EDL decreased significantly during the first 48 h in the obese only in wTA, reaching a minimum at 24 h, whereas EDL showed no variation.
|
|
|
In the lean group, a decrease of total LCACoA levels was observed at 72 h of starvation (Table 3). This is consistent with the incipient drop of FFA, ketone bodies, and IMCL observed in this group already at 48 h of starvation and reflects the depletion of lipid stores.
|
DISCUSSION |
|---|
|
TOPABSTRACTRESEARCH DESIGN AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
10 wk. At the age of 8 wk, the obese ZDF rats showed markedly reduced insulin sensitivity (GIR, clamp Rd) compared with their lean littermates. According to the prevailing view that starvation induces IR (33), we measured insulin sensitivity by the glucose clamp technique after 24 and 72 h of starvation in both rat groups. Consistent with the constant levels of FFA, IMCL, and LCACoA in muscle of obese rats after a fast of 24 and 72 h, there was no change in insulin-induced muscle Rd detectable after prolonged starvation. After 72 h of starvation, lean rats demonstrated lower FFA, IMCL, and LCACoA compared with the values after 24 h of starvation, which might reflect a different state of intermediary lipid metabolism after nearly complete mobilization of fat stores, as mentioned below. In fact, muscle Rd during basal conditions (i.e., basal EGP) was increased after 72 h of starvation, which was consistent with the reduced serum and tissue lipid parameters. However, whole body insulin sensitivity (GIR) was reduced in lean rats after prolonged starvation, which was based on a nonsuppressable EGP by insulin, reflecting hepatic IR after 72 h of starvation. The biochemical characterization of the different muscle fiber types used for this study confirmed the classification of wTA, EDL, and soleus muscle as "fast twitch, glycolytic," "intermediary," and "slow twitch, oxidative," respectively (Table 1). Specifically, the reduction of maximal HK activity in all three muscles of the obese rats was consistent with their degree of IR (Table 2). As expected, mitochondrial fatty acid oxidation capacity (HOAD activity) was low in glycolytic and high in oxidative muscles. Additionally, HOAD activity was reduced (P < 0.05) in wTA but not in EDL or soleus of the obese strain. For Sprague-Dawley rats, histochemically determined muscle fiber compositions of 76 muscles have been reported and demonstrated that 89% of total skeletal muscle mass was comprised of glycolytic fiber types (71% for type IIB and 18% for type IID/X fibers) and only 11% oxidative fiber types (6% for type I and 5% for type IIA fibers; see Ref. 11). Assuming a similar proportion of fiber type composition of total skeletal mass in ZDF rats, the importance of glycolytic muscles for insulin sensitivity in rats is evident (28).
Fed state.
The higher IMCL values of the obese rats were consistent with their degree of IR. The relative IMCL interstrain differences were most pronounced in wTA (394%) followed by EDL and soleus (333 and 304%, respectively) and predicted the relative responses of IMCL resulting from starvation described later. In spite of these important differences, the levels of plasma FFA, as well as the total LCACoA levels, were identical in both strains. The markedly increased serum triglyceride levels during fed conditions seem to determine IMCL values rather than FFA. Additionally, the obese rats exhibited higher body weight, which has been shown to be a result of a reduced lean-to-fat mass ratio (13). All of these factors have been shown to correlate with higher IMCL values and IR in rats (24) and in humans. Malonyl-CoA levels in soleus were not significantly different between the two strains as opposed to previous reports (50, 51). It seems questionable whether the small, although statistically significant, differences of malonyl-CoA levels detected in wTA and EDL were sufficient to influence -oxidation.
Transition to starvation and to refeeding.
In lean rats starved for 72 h, starvation-induced elevated levels of FFA, ketone bodies, and IMCL as well as total LCACoAs decreased. These decreases were consistent with a nearly complete mobilization of fat stores in lean animals after 72 h of food deprivation in contrast to obese rats. Therefore, the results for the lean rats from this time point should be interpreted separately, since this long-lasting food deprivation might reflect a different state of starvation with a subsequent different impact on intermediary metabolism compared with that of obese rats.
Some of our recently reported findings on lipid metabolism during starvation in normal, insulin-sensitive Wistar rats (39) were confirmed by the present study: muscles with different fiber types coped with starvation-induced elevated FFA levels differently (glycolytic wTA and intermediary EDL by increasing FFA reesterification and storage as IMCL and oxidative soleus probably by -oxidation).
In wTA, the scarce lipid oxidative capacity available (Table 1) was strongly or even maximally activated by low malonyl-CoA levels at 48 h (Fig. 2A) in lean animals, whereas in the obese rats the decrease in malonyl-CoA levels was less pronounced. In this group of animals, the drop in malonyl-CoA was obviously insufficient for an adequate carnitine palmitoyltransferase (CPT) I function and subsequent efficient fatty acid oxidation, and FFA had to be stored as IMCL (Fig. 3A). In line with this interpretation were the EDL figures. In the obese, there was no significant change in malonyl-CoA and a rise for IMCL up to a level of 6 arbitrary units (AU; Fig. 3B), and in the lean there was a significant reduction of malonyl-CoA by 0.6 nmol/g together with an IMCL increase up to a level of <3 AU. In the oxidative soleus, however, neither malonyl-CoA nor IMCL values experienced any change at any time. In glycolytic muscle, cytosolic malonyl-CoA levels depend primarily on the production of acetyl-CoA via glycolysis, whose activity will be strongly diminished during starvation. The concomitant increase in IMCL and fall of malonyl-CoA values support the concept that, as a reaction to increased FFA availability during starvation, glycolytic muscle (wTA and EDL) will activate lipid oxidation more the lower its maximal oxidative capacity and then store the rest as IMCL. The smaller malonyl-CoA changes observed in the muscles of obese rats probably reflected their reduced HK activity and their inherent IR. Referring to data from the literature, the concentrations of malonyl-CoA measured in our study (wTA 600 nM, EDL 400 nM) should inhibit CPT I at all times (37, 59). But we cannot assume, however, that the concentration required for 50% inhibition (IC50) determined in isolated mitochondria [34 nM according to McGarry et al. (37)] necessarily is the IC50 of the intact muscle. To solve this question McGarry and Brown (36) suggested the possible distribution of total muscle malonyl-CoA to subcellular compartments unavailable to the CPT I enzyme. Yet, the muscle fatty acyl-CoA concentration might also be high enough during starvation (reflected by increased LCACoA levels in starved animals) to displace a significant fraction of malonyl-CoA from the enzyme (60). In liver, the sensitivity of CPT I to malonyl-CoA inhibition has been reported to decrease in fasted rats (44). This may also be true for CPT I in rat muscle.
In line with this interpretation, malonyl-CoA concentrations rose on refeeding in glycolytic wTA and EDL of the lean group. Because IMCL declined already at 72 h in the lean rat muscles because of their exhausted fat stores, no additional drop in IMCL was detected when refeeding the animals. Chien et al. (8) reported rapid increases in malonyl-CoA levels in Sprague-Dawley rats on refeeding for their gastrocnemius (60% within 1 h), presumably white gastrocnemius, and their soleus muscle (100% within 1 h). However, the initial fed-state value of 2 nmol/g was not obtained before 12 h of refeeding. These findings agree with our data. We observed increases in malonyl-CoA in wTA and EDL of lean insulin-sensitive rats after refeeding them overnight. The soleus muscle, however, showed no such dynamics during fed-to-starved-to-refed transition.
Because we did not observe any starvation-induced changes in the malonyl-CoA content of the soleus muscle, we conclude that, in contrast to glycolytic muscles, the oxidative soleus in both strains did obviously not require an activated -oxidation via reduced malonyl-CoA levels nor an enhanced IMCL storage to cope with the starvation-induced increased FFA supply. Actually, according to all biochemical and physiological parameters measured, soleus was only affected a little by the feeding status or the insulin-resistant condition. The question arises whether, during starvation, malonyl-CoA plays the same role in oxidative as in glycolytic muscle. In soleus with its low glycolytic capacity, malonyl-CoA could partly be derived from mitochondrial lipid oxidation and export of citrate. This supports the additional existence of a malonyl-CoA-independent mechanism for transfer of LCACoAs to mitochondria, which has recently been described (23, 25). Furthermore, a higher concentration of LCACoA-binding protein (ACBP) has been reported for soleus vs. EDL muscle and for obese vs. lean Zucker rats (14). In our study, the levels of total LCACoA increased from wTA over EDL to soleus. The increase in ACBP could be an attempt to protect the cell against too high concentrations of LCACoA in the cytosol while maintaining the availability of fatty acids as energy substrate in the oxidative muscle.
The decrement in malonyl-CoA in the glycolytic muscles was larger in the lean animals compared with the insulin-resistant obese, hinting at a higher metabolic flexibility in the former. The obese animals, although having much higher IMCL levels, appeared to be metabolically "inflexible" due to a relatively smaller decrement in malonyl-CoA, whose levels were obviously not permissive for appropriate CPT I function and subsequent adequate fatty acid oxidation. Besides the diminished capacity for lipid oxidation, total lipid (FFA plus triglyceride) concentration in plasma is elevated in the obese rats. Therefore, obese rats have to store lipids as IMCL for two reasons. First, the smaller decrement of malonyl-CoA that prevents appropriate lipid oxidation and, second, the greater amount of lipids entering the muscle cell. Therefore, our observations support the regulatory fuel-sensing role of malonyl-CoA to control fatty acid oxidation in glycolytic muscles in the fed-to-starved and in the starved-to-refed transition in lean insulin-sensitive rats, whereas in the obese insulin-resistant animals this fuel-sensing system is obviously defective, reflecting an impaired metabolic flexibility (35, 43, 48).
On absolute terms, the starvation-induced changes in IMCL were far larger in the obese animals (for wTA and EDL, respectively, 3.2 and 1.2 AU in lean vs. 6.2 and 2.3 AU in obese rats within 24 h of starvation). These changes occurred in parallel to those observed for plasma FFA levels (Figs. 1 and 3), confirming the influence of FFA on the dynamics of IMCL (7). In both rat strains, plasma FFA levels were identical in the fed state and up to 48 h of starvation despite highly different IMCL values. These findings suggest larger esterification rates fueled by increased fluxes of FFA in the obese rats, finally leading to accumulation of IMCL. The higher FFA levels in obese rats measured during the glucose clamp study might be related to the long-term pentobarbital sodium anesthesia in contrast to the short-term isoflurane anesthesia for blood collections in all other studies.
During starvation, skeletal muscle lipoprotein lipase (LPL) activity is increased compared with adipose tissue LPL, leading to an increased uptake of triglyceride-derived fatty acids in the skeletal muscle cell (30). Because the obese animals display higher serum triglyceride levels compared with lean animals, an increased uptake of triglyceride-derived fatty acids might be an additional cause for the elevated absolute IMCL values in obese compared with lean animals during starvation.
LCACoA.
The dissociation of changes in LCACoA and IMCL levels in the presence of starvation-induced elevated FFA supports the concept that intramyocellular reesterification of fatty acids is a physiological mechanism to control LCACoA levels and thereby efficiently prevents impairment of insulin signaling, at least in glycolytic muscles. That triglyceride accumulation protects against fatty acid-induced lipotoxicity in the context of apoptosis has also been reported from in vitro studies with cells (31).
The lower insulin sensitivity (GIR, Rd) in insulin-resistant obese ZDF rats compared with their lean littermates occurred in the presence of comparable levels of total LCACoAs in all three muscle types. Elevated LCACoA levels in muscle have been identified in rats and mice being fed a high-fat diet and demonstrating IR (12, 22, 40). The elevated saturated and monounsaturated LCACoAs (SAT + MUFA) in obese rats compared with lean rats might be consistent with the notion of the metabolically more deleterious effects of SAT + MUFA compared with PUFA (10, 55).
It is concluded that IMCL is the most robust noninvasive biomarker for IR in ZDF rats compared with any other metabolic tissue or blood parameter determined in this study; however, muscle type-specific differences regarding enzymatic capacity and thus fatty acid metabolism have to be considered when applying this biomarker. Our study supports the concept of "metabolic inflexibility" in the obese ZDF rats, which is reflected by markedly increased IMCL levels and a smaller drop in malonyl-CoA levels during starvation in wTA and EDL than in the lean littermates. The concomitant increase in IMCL with the fall of malonyl-CoA supports the concept that, as a reaction to starvation-induced FFA availability, muscle will activate lipid oxidation more the lower its oxidative capacity (wTA > EDL > soleus) and then store the rest as IMCL.
|
FOOTNOTES |
|---|
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.
|
REFERENCES |
|---|
|
TOPABSTRACTRESEARCH DESIGN AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. F. Ndisang, N. Lane, and A. Jadhav Upregulation of the heme oxygenase system ameliorates postprandial and fasting hyperglycemia in type 2 diabetes Am J Physiol Endocrinol Metab, May 1, 2009; 296(5): E1029 - E1041. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Ndisang and A. Jadhav Heme oxygenase system enhances insulin sensitivity and glucose metabolism in streptozotocin-induced diabetes Am J Physiol Endocrinol Metab, April 1, 2009; 296(4): E829 - E841. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Schummer, U. Werner, N. Tennagels, D. Schmoll, G. Haschke, H.-P. Juretschke, M. S. Patel, M. Gerl, W. Kramer, and A. W. Herling Dysregulated pyruvate dehydrogenase complex in Zucker diabetic fatty rats Am J Physiol Endocrinol Metab, January 1, 2008; 294(1): E88 - E96. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) and obese (
) Zucker diabetic fatty (ZDF) rats. All values are from one group of animals (magnetic resonance spectroscopy group). Values are means ± SE; n = 8 rats. *P < 0.05 and **P < 0.001 vs. fed condition; 
P < 0.05 and 
