Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Undernutrition has been reported for a long time in 20 to 70% of all chronic obstructive pulmonary patients (1), as well as in children with chronic diseases like cystic fibrosis (CF) (4) or Duchenne muscular dystrophy after the age of 14 yr (5, 6). Several studies have shown the association between undernutrition and increased morbidity and mortality (7).

In hindlimb skeletal muscles, undernutrition results in a loss of muscle mass and a decrease in strength (8). These alterations are associated with a decrease not only in glycolytic but also in oxidative enzyme activities (9), as well as a reduction in high-energy phosphate stores and an increase in the calculated free ADP (12). Furthermore, a decrease in total cytochrome c associated with changes in oxidative enzyme activity in skeletal muscle has been shown after total or partial dietary restriction (9). These changes suggest deficient oxidative phosphorylation and an alteration in mitochondrial oxygen uptake.

Undernutrition also induces respiratory muscle weakness, particularly in the diaphragm, with a decrease in strength (13) and endurance (13, 17) partially related to a loss of muscle mass (13). This muscle dysfunction seems to play an important role in the genesis of respiratory muscle failure. In vitro animal studies have shown significant reduction in peak twitch and tetanic tensions of the diaphragm (18) that seem related to decreased diaphragmatic weight and decreased cross-sectional area (CSA) of all muscle fibers (19).

Contrary to the hindlimb muscle, the effect of undernutrition on oxidative metabolism is not known, except a preserved activity of succinate dehydrogenase in individual muscle fibers (21). However, respiratory muscle function is reduced more dramatically than weight loss after undernutrition (22, 23) and is reversed with refeeding before weight is restored, suggesting an additional myopathic effect (17). On account of clinical observations and results about enzymatic activities in hindlimb muscles, we hypothesized that undernutrition could decrease oxidative metabolism in the diaphragm. To test this, we examined the effects of prolonged undernutrition on rat diaphragm mitochondrial oxygen uptake. This assessment may provide a more integrative index of metabolic capacity than assays of individual enzyme activities. The aim of this study was thus to evaluate the effect of prolonged undernutrition on diaphragm mitochondrial oxygen uptake with pyruvate and palmitate as substrate coming from glycolysis and lipolysis, respectively, and to determine at what level of mitochondrial metabolism undernutrition may act.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals and Nutritional Protocol

Twenty adult male Wistar rats (Janvier, France) were studied (initial body weight ~ 250 g). Daily food consumption for each rat was measured over a 1-wk period. The animals were then divided into two groups: (1) control (CTL) animals (n = 10) that received food and water ad libitum (Purina rat chow, Pietrement, Provins, France: 56% carbohydrate, 23% protein, 4.5% fat, 6% fiber, and 10.5% ash and minerals) and (2) nutritionally-deprived (ND) animals (n = 10). These animals received water ad libitum and the same Purina rat chow, but daily food access was restricted to 40% of their estimated daily consumption. Nutritional deprivation was continued until the ND animals weighed ~ 60% of the controls (40% below control body weight). All animals were housed in individual cages without physical activities and were maintained on a 12:12 h inverted light-dark cycle at normal ambient temperature. Body weight was measured twice a week. After ~ 5 wk, animals were killed by cervical dislocation, and the entire diaphragm was rapidly removed. The protocol was accepted by our local ethics committee for animal care.

Mitochondrial Isolation

The diaphragm was immediately placed in ice-cold isolation medium (IM) containing 225 mM mannitol and 75 mM sucrose at pH 7.4 to prevent breakdown of mitochondrial enzymes. The muscles were cleared of fat and connective tissue and weighed. Wet muscle weight was determined as the difference between a given volume of cold IM with and without muscle sample. A total of 250 mg of each muscle were removed and transferred into fresh IM for mitochondrial respiration. For mitochondrial preparation we used a method for small samples described by Rasmussen and colleagues. (24). Briefly, each sample was cut into five pieces and incubated for 2 min at 4°C with 500 µl proteinase medium (PM): ATP medium (100 mM KCL, 50 mM Tris, 5 mM MgSO4, 1mM EDTA, 1 mM ATP, 0.5% BSA, pH 7.40) with the addition of 2 mg/ml trypsine (bovine pancreas type III; Sigma Chemical Co., St Louis, MO). All further procedures were performed at 0-5°C. The PM was then diluted with 3 ml ATP medium and the liquid was sucked away. ATP medium (6 ml) was added. Samples were homogenized using an Ultra-Turrex tissue homogenizer (Prolabo, France) with five slow passes of the pestle at 40% power. Muscle homogenates were centrifuged at 300 × g (KR 422, Jouan, France) for 5 min. The supernatants were centrifuged 10 min at 4000 × g. Pellets was resuspended in 1,050 µl cold isolation KCL medium (100 mM KCL, 50 mM Tris, 5 mM MgSO4, 1 mM EDTA, pH 7.4) and centrifuged for 10 min at 7000 × g. The pellets were then resuspended in 700 µl IM. The final mitochondrial protein concentration was determined by the Bradford protein assay (Biorad, Hercules, CA) with bovine -globuline as a standard. The mitochondrial protein yield (Bradford assay) from the diaphragm, which represents the amount of mitochondrial protein isolated per gram of muscle, was calculated.

Mitochondrial Respiration

The order of mitochondrial oxygen consumption measurement was randomized (control and denutrition samples). Rates of oxygen consumption by mitochondria were measured with a complete oxygen-measurement system composed of an oxygen meter (model 781, Strathkelvin System, Glasgow, Scotland), a microcathode oxygen electrode (Clark-type polarographic electrode, Glasgow, Scotland), and a respiration chamber (3 ml) set at 37°C. When the electrode was assembled, it was calibrated. The zero value was obtained with a sodium borate solution (2.0 ml of a 0.01 M solution) added to the chamber with crystals of sodium sulfite. The maximal value was calibrated with 1.87 ml respiration medium alone containing 780 × 10⁹ nmol atoms of oxygen. Before measurement of oxygen consumption, the respiration chamber was filled with 1.87 ml of respiratory medium (RM) containing 15 mM KCL, 30mM K₂HPO₄, 25 mM Tris base, 45 mM sucrose, 12 mM mannitol, 5 mM MgCL₂, 7 mM EDTA, 0.2% free fatty acid bovine albumin, and 20 mM glucose (pH 7.4). One of three substrates, 10 mM pyruvate plus 2.5 mM malate, 40 µM L-carnitine palmityl plus 1 mM malate, or 10 mM succinate plus 5.0 µM rotenone, was added to the respiration chamber to inhibit complex I (NADH reductase) of the electron-transport chain. A capillary-shaped opening in the electrode allowed sequential addition of substrate with no risk of air diffusing into the chamber. Then, 100 µl of the isolated mitochondria, corresponding approximately to 1 mg of mitochondrial protein resuspended in 700 µl IM was loaded into the chamber with a magnetic stirring apparatus to ensure complete mixing of mitochondria and substrate. ADP at a final concentration of 250 µM was added to a 2-ml final volume chamber to initiate state 3 (ADP-stimulated) respiration. The respiratory control ratio (RCR) was obtained by dividing state 3 respiration by the recovery rate after completion of ATP synthesis, i.e., state 4 respiration, which is defined as oxygen consumption of mitochondria after the depletion of exogenous ADP (5). Oxygen consumption was determined as the amount of oxygen disappearing from the respiration chamber over time per 1 mg of mitochondrial protein. The ratio of the amount of phosphorylated ADP to oxygen consumed (ADP/O) was calculated by dividing the nanomole of added ADP (500 nmol) by the nanomole of oxygen consumed during state 3 respiration. All measures of every state were performed at least in duplicate.

Enzymatic Activities and Glycogen Measurement

Citrate synthase (CS) activity was measured in isolated mitochondria in a reaction medium consisting of 100 µM dinitrothiobenzoic acid, 100 mM tris(hydroxymethyl)aminomethane (Tris), and 15 µM acetylcoenzyme A, according to the method of Srere (25). Isolated mitochondria were defrosted at room temperature and 10 µg of mitochondrial protein were added to the reaction medium. The absorbance at 412 nm (DU 640 spectrophotometer; Beckman, Roissy, France) was checked for 3 min to determine the nonspecific activity. Reactions were then initiated by addition of 500 µM of oxaloacetate, and the change in absorbance was recorded for at least 3 min at 412 nm.

NADH dehydrogenase activity was assessed with a method described by Holloszy (26). The reduction of potassium ferricyanide was measured at 420 nm. The reaction medium consisted of 0.15 mM DPNH, 0.6 mM potassium ferricyanide, and 40 mM potassium phosphate buffer (pH 7.4). Reactions were then initiated by addition of 500 µl of mitochondrial suspension. Proteins were estimated with the same method described above. Results were expressed in micromoles per minute per gram protein.

The glycogen concentration in diaphragm was determined using the method of Lo and colleagues (27). Briefly, muscle samples were boiled in 30% KOH saturated with Na₂SO₄ until homogenized. Homogenates were kept on ice, and glycogen was precipitated by addition of a 1.2 vol of 95% ethanol. Samples were centrifuged for 30 min at 840 × g and pellets were resuspended in H₂O. Assays were conducted on aliquots against appropriate blanks at 490 nm. Results were determined from a standard curve generated at the same time and expressed in micromoles of glycogen unit per gram of tissue.

Extraction of Myofibrils and Separation of the Myosin Heavy Chain Isoforms

Separating and stacking gels were prepared according to the detailed procedure of Talmadge and Roy (28) using an electrophoresis system (Miniprotean II; Bio-Rad). A sample of muscle (50 mg) was homogenized in 10 vol of 250 mM sucrose, 20 mM Tris, 5 mM ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'- tetraacetic acid (EGTA), and 100 mM KCL, pH 6.8, with the use of a glass tissue homogenizer on ice. Homogenates were then centrifuged for 10 min at 4°C at 1500 × g. The myofibril pellets were subsequently washed twice with 175 mM KCL, 20 mM Tris, 5 mM EGTA, and 0.5% Triton X-100 (pH 6.8), as described previously (29). Two final washes with 150 mM KCL, 5 mM EDTA, and 20 mM Tris (pH 7.0) gave purified myofibril pellets that were diluted to 1 mg/ml in 50 mM pyrophosphate, 50% glycerol, 5 mM EGTA, and 2 mM mercaptoethanol (pH 8.8) and incubated in 1 vol of denaturing buffer for 5 min at 100°C. Then, 1 µg of protein was loaded in each well. The run was at 90 V for 27 h in a cold room. Overnight staining by Coomassie blue in 40% methanol-10% acetic acid, followed by destaining in 10% methanol-10% acetic acid, revealed several types of MHC isoforms. The gel was scanned (Studioscan II SI; AGFA, France), and the MHC isotypes were quantified with National Institutes of Health software. To do this, we took into account the area and density of each band in the gel. We then calculated the total density of each well; this was normalized to represent 100% so that the MHC subtypes were expressed as percentages.

Statistical Analysis

Statistical analysis was done using Sigma Stat statistical software (Jandel Scientific, Erkrath, Germany). Data are reported as means ± SE. Data concerning body and muscle weights, mitochondria isolated from single muscle, and enzyme activity were analyzed using Student's t test, with P = 0.05 designated as significant. When data distribution was not normal, a Mann-Whitney test was used. The mitochondria were suspended in a final volume of 700 µl, and 100 µl of mitochondria were used for each mitochondrial respiration measurement. Thus, we were able to obtain two assays for each substrate. Each retained value corresponds to the mean of these two assays. We report the mean ± SE of these values from the ten rats of both groups.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of Denutrition on Body Weight, Muscle Weight, and Diaphragm Glycogen Content

The initial body weights of the CTL and ND animals were similar. After nutritional deprivation for ~ 5 wk, body weight in the ND group was reduced to ~ 60% of that of the CTL group. During this period, the body weight of the CTL animals increased steadily by 49% of initial weight (292 ± 21 versus 435 ± 38 g), whereas the body weight of the ND animals decreased by ~ 10% of initial weight (295 ± 29 versus 276 ± 25 g).

The diaphragm weight in the ND group was 48% of that in the CTL group (Table 1). Moreover, diaphragm glycogen was significantly lower in the ND group than in CTL without significant difference in the mitochondrial protein yields that represent the total mitochondrial protein per gram of muscle (Table 1).

Effect of Denutrition on MHC Composition

In ND compared with CTL animals, we found a significant change in diaphragm muscle MHC isoform with undernutrition. With undernutrition, the MHC IIx isoform was decreased (33 versus 37%) with no difference in the percentage of type MHC IIa isoforms (18 versus 19%) and in type IIb isoform (18 versus 20%). On the other hand, we found an increase in the percentage of type MHC I isoforms with denutrition (31 versus 24%). The results are shown in Table 2.

Effect of Denutrition on Mitochondrial Respiration

Figure 1 shows a significant decrease in state 3 mitochondrial respiration in the ND group compared with CTL group for pyruvate plus malate as substrate (993 ± 171 versus 1488 ± 167 nmol atomic O/mg/min, P < 0.01). We also found a significant decrease in state 3 mitochondrial respiration in ND compared with CTL for palmitate plus malate (516 ± 89 versus 850 ± 165 nmol atomic O/mg/min, P < 0.05).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Mitochondrial oxygen consumption during state 3 and state 4 rates in diaphragm for control and undernourished rats with pyruvate and malate, palmitate and malate, or succinate and rotenone as substrates, respectively. Open bars are data from control animals and dark bars are data from undernourished rats. Values are means ± SEM. *P < 0.05, **P < 0.01.

Mean mitochondrial state 3 respiration with succinate plus rotenone added to block complex I (NADH reductase) of the electron transport chain was not significantly different between the ND and CTL groups (906 ± 204 versus 1,106 ± 262 nmol atomic O/mg/min).

In CTL and ND groups, the mean rates of oxygen consumption under state 4 respiration with pyruvate plus malate (270 ± 27 versus 222 ± 35 nmol atomic O/mg/min), palmitate plus malate (222 ± 38 versus 168 ± 38 nmol atomic O/mg/min), or succinate plus rotenone (528 ± 124 versus 432 ± 90 nmol atomic O/mg/min) as substrate were not significantly different between groups.

The mitochondrial isolation protocol used for this experiment resulted in well-coupled mitochondrial oxidation. The RCR, i.e., the ratio of state 3/state 4 respiration determined for the different conditions, was high. As shown in Figure 2, the value of RCR in diaphragm with pyruvate plus malate as substrate was significantly lower (P < 0.05) in ND (4.5 ± 1) compared with CTL (5.5 ± 0.8). For palmitate (3.3 ± 0.42 versus 3.6 ± 0.28) or succinate (2.0 ± 0.09 versus 2.1 ± 0.11) as substrate, the differences between ND and CTL group were not significant.

View larger version (13K): [in this window] [in a new window]   Figure 2.   Effect of undernutrition on the ADP/O ratio and respiratory control ratio (RCR) in diaphragm for control and undernourished rats with pyruvate and malate, palmitate and malate, or succinate and rotenone as substrates, respectively. Open bars are data from control animals and dark bars are data from undernourished rats. Values are means ± SEM. *P < 0.05, **P < 0.01.

Moreover, with the ND and CTL groups, the ratio of the amount of phosphorylated ADP to amount of oxygen (ADP/O) was 2.70 ± 0.1 versus 2.92 ± 0.09 for pyruvate plus malate and 2.46 ± 0.1 versus 2.65 ± 0.1 for palmitate plus malate as substrate. When succinate plus rotenone was used as substrate, ADP/O decreased for both groups (1.48 ± 0.06 versus 1.52 ± 0.05), as would be expected for reducing equivalents entering the electron transport chain at complex II. As shown in Figure 2, ADP/O ratios were not significantly different between the ND and CTL groups.

Effect of Denutrition on Citrate Synthase and NADH Dehydrogenase Activities

The CS activity was significantly lower in the ND group compared with CTL (11.3 ± 1.08 versus 18.0 ± 2.14 µmol/ min/g protein; P < 0.01). The value of NADH dehydrogenase activity was also significantly lower in the ND group (1.3 ± 0.17 versus 1.8 ± 0.26 µmol/min/g protein, P < 0.05). Results are shown in Figure 3.

View larger version (16K): [in this window] [in a new window]   Figure 3.   Histogram comparing diaphragmatic citrate synthase activity and NADH dehydrogenase activity in control rats (open bars) and undernourished rats (dark bars). Values are means ± SEM. *P < 0.05, **P < 0.01.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The results of this study showed that prolonged undernutrition induced a significant decrease in mitochondrial oxygen consumption. The state 3 respiratory rate of the diaphragm decreased with pyruvate plus malate and palmitate plus malate as substrates. No significant difference was noted with succinate as substrate plus rotenone, which inhibits complex I of the electron chain transport. Furthermore, the mitochondria from the diaphragm showed significant decrease in the activity of citrate synthase, one of the oxidative enzymes, and in NADH dehydrogenase activity, which represents complex I.

We used the undernutrition protocol described by Lewis and colleagues (19). After five weeks of undernutrition, we obtained an ~ 10% loss of body weight in ND rats, which is less than the 22% described by these authors. In the control rats, we obtained a 49% increase in weight. Thus, secondary to undernutrition, the ND rats weighed 60% of the control rats. Undernutrition also caused a significant reduction in diaphragm muscle weight. As previously noted in the hindlimb muscles of undernourished rats (30), we observed a decrease in the glycogen content in the diaphragm of ND rats as compared with controls, which is a metabolic index of undernutrition. In previous studies in our laboratory (31, 32), we used a method of mitochondrial isolation that requires several grams of tissue. As our sample of ND rat diaphragm was small, in this study we used a method described by Rasmussen and colleagues (24) that is appropriate for tissue amounts of ~ 100 mg. The values of RCR were high and indicate that the isolation procedure allowed us to obtain well-coupled mitochondria.

The main result of this study was a significant decrease, secondary to undernutrition, in mitochondrial respiration in the diaphragm with pyruvate plus malate and palmitate plus malate as substrate. Bissonnette and colleagues (11) found a lower glycolitic flux in the soleus muscle of underfed rats, suggesting a preferential shift toward an alternate exogenous substrate, such as free fatty acids. Although the same shift toward free fatty acid substrate could also occur in diaphragm, our results showed a decreased mitochondrial respiration in the undernourished rats for both carbohydrate and lipid utilization. This seems to be in agreement with a recent previous study (14), which found with undernutrition an imbalance in the energy flux between ATP production and utilization, with a decrease of the ratio ATP to ADP. However, the low ATP level with undernutrition, taken in isolation, is a poor index of tissue metabolic function, as alterations at a given time may reflect alterations in ATP synthesis, creatine phosphate store, or decreased ATP utilization. The present study, with measurement of mitochondrial respiration, allowed us to have a more direct examination of ATP synthesis by oxidative phosphorylation, which is one of the determinants of overall cellular energy balance.

As noted by Amederes and colleagues (33, 34), we found with undernutrition a change in diaphragm isoform composition with an increase in the percentage of type I MHC isoforms and a decrease in the percentage of type IIx. It is important to note that coexpression of MHC isoforms in a same muscular fiber exists. Thus MHC isoform composition of the mixed fiber biopsy does not represent a clear picture of the fiber type composition. However, our result concerning MHC composition seems in agreement with previous studies (18, 19) showing a decrease in percentage of type II fiber, which has the lowest oxidative capacity, and an increase in type I fiber, which has the highest oxidative capacity. Thus, as the percentage of type I MHC isoforms increases, alterations in mitochondrial state 3 rates seem to reflect a true reduction in mitochondrial function rather than a switch in fiber type. This increase in type I MHC isoforms should have been accompanied by a higher mitochondrial protein yield in the undernourished group compared with the control group. Although we observed this tendency, the differences were not statistically different. This is probably due to our low number of samples.

It is possible that the decrease in mitochondrial oxygen consumption observed in rat diaphragm was due to reduced activity of rats (32) secondary to undernutrition. However, our groups of rats were kept in small, individual metabolic boxes and their physical activities were both reduced. It thus seems that a difference in physical activity level between groups could not be involved in the substantial decrease in state 3 mitochondrial respiration. Moreover, it has been shown that exercise prevents the negative effect on oxidative muscular enzyme activity induced by undernutrition (35). From our results, however, it seems that the chronic contractile activity exerted by the diaphragm does not prevent the negative effect of denutrition on mitochondrial respiration.

Although an in vitro test of mitochondrial oxygen consumption cannot represent the whole phenomenon induced in vivo by undernutrition, alteration in diaphragmatic mitochondrial respiration with carbohydrates and lipids as substrates could be involved in the negative effect of undernutrition on respiratory muscle function in patients. Indeed, acute respiratory failure in COPD with undernutrition has been partly explained by an alteration in muscle contractile properties secondary to decreased diaphragmatic lean mass (36, 37). However, weight loss does not seem to be the only mechanism involved. Based on clinical evaluation, endurance, which is the main function of respiratory muscle, seems to be reduced more dramatically than weight loss (22, 23), and its reversibility occurs with renutrition before weight is restored (17).

The observed decrease in mitochondrial respiration, which reflects the alteration in oxidative phosphorylation, may be explained by a negative effect of undernutrition on a number of factors. These include the provision of reducing equivalents with their oxidation through the electron transport chain, the provision of ADP and Pi, and the phosphorylation state, i.e., the ratio of ATP to ADP. The electron transport chain is composed of four complexes: complex I, NADH-ubiquinone reductase; complex II, succinate-ubiquinone reductase, which is not a proton pump; complex III, ubiquinone-cytochrome c reductase, also called cytochrome bc₁ complex, the second of the three proton pumps; and complex IV, cytochrome c oxydase (38). These complexes in the mitochondrial inner membrane are responsible for the process of oxidative phosphorylation, i.e., ATP production. Because we observed the negative effect of undernutrition with all substrates except succinate plus rotenone, which is a complex I inhibitor, our results indicate that complexes II, III, and IV may not be involved and that undenutrition acts essentially on complex I or on prior sites. We observed a significant decrease in NADH dehydrogenase activity in the ND group, which represents complex I. However this alteration is relatively weak and could not be primarily responsible for the observed reduction in state 3 mitochondrial respiration rate. Indeed, to produce such a reduction in overall state 3 mitochondrial respiratory rates with specific complex inhibitor, a greater reduction in the individual complex is necessary (39, 40). Thus, although undernutrition may have a negative effect on individual complex I activity, this alteration does not represent its main effect and may not explain the alterations observed on the oxidative phosphorylation pathway.

Undernutrition may thus have deleterious effects on other sites, i.e., (i) on substrate entry through the mitochondrial membrane, (ii) on the generation of NADH through all enzyme complexes of the tricarboxylic acid cycle or through NADH transport with the -glycerophosphate shuttle, (iii) on the integrity of the inner mitochondrial membrane, or (iv) on the F₁F₀ ATPase and the adenine nucleotide transporter. In the diaphragm, the ADP/O ratio with pyruvate plus malate as substrate was near 3.0 for control and malnourished rats. Theoretically, NADH-linked substrates allow three molecules of ATP formed per passage of pair of electrons from NADH along the electron transport chain. Identically, the oxidation of succinate allows for the transfer of electrons to FADH, which enters the chain at complex II and induces two molecules of ATP formed per passage of pair of electrons from NADH. For succinate as substrate and rotenone as a complex I inhibitor, the ADP/O ratio was 2.0 for both groups, with no difference in state 3 respiration rates. As control and undernutrition samples were treated at the same time in randomized order by groups of two, we thus could hypothesize that undernutrition alone does not affect the phosphorylative coupling efficiency nor the chemiosmotic gradient across the inner membrane or the F complex containing the mitochondrial ATPase.

The influx of ADP, under the regulation of the adenine nucleotide transporter that is located in the inner mitochondrial membrane, is a rate controller for oxidative phosphorylation (41). The lack of difference between the control and undernourished groups for state 3 respiration rates with succinate plus rotenone suggests no negative effect induced by undernutrition on this transporter.

Similarly, a deleterious effect of NADH transport does not seem to be involved in our study. Indeed, respiratory rates were evaluated on isolated mitochondria, and the role of NADH transport at the inner membrane could not have been predominant as the initial source of NADH is the mitochondrial matrice. Moreover, on isolated mitochondria, a decrease in state 3 oxygen consumption, by modification of the NADH to NAD ratio in the matrice, is associated not with a decrease but on contrary with an increase in the -glycerophosphate dehydrogenase activity, which represents NADH transport (42).

Then, the observed reduction in state 3 respiration rate is not due primarly to a reduction in electron transport chain, or a reduction in NADH or adenine nucleotide transporter. The main result of this study was that prolonged undernutrition induced a significant reduction in NADH generation by the Krebs cycle. This may be due to a decrease of enzyme activities of the Krebs cycle. The significant decrease in citrate synthase activity in the diaphragm of ND rats as compared with CTL rats may reflect this phenomenon. Sieck and colleagues (21) showed a relative preservation of succinate dehydrogenase activity in undernourished rats. As suspected by these authors, the difference in the two results could be explained by the fact that in our study enzyme activities were assessed not on isolated muscle fiber, but from homogenates of muscle that provided an overall estimate of enzyme activity in the muscle. Our result was similar to those of previous investigators (9, 10, 12), who reported that undernutrition reduced the activity of a number of oxidative enzymes in hindlimb skeletal muscles. Indeed, Layman and colleagues (9) showed that denutrition induces a glycolitic alteration in slow-twitch fibers, reducing the availability of energy from glycolysis during contraction. Russell and colleagues (10) reported that phosphofructokinase, hydroxyacyl-CoA-dehydrogenase, and succinate dehydrogenase activities were reduced in skeletal muscle homogenates of undernourished rats. The decrease in oxidative enzymatic activities in the diaphragm after undernutrition could be related to the changes in the metabolic mixture of fuels being metabolized. Indeed, it has been shown that the metabolic state of an animal (41) and the composition of the diet can influence oxydative phosphorylation (43). Undernutrition does not provide enough substrate for oxidation that in turn provide not enough reducing equivalents available to the respiratory chain to make water and generate ATP. This, in addition to a shortage of amino acids required to synthesize new protein, creates a shortfall of ATP and a reduced protein synthesis (41).

In conclusion, the main result of this study was that prolonged undernutrition induced a decrease in mitochondrial respiration secondary to a significant reduction in NADH generation by the Krebs cycle. This alteration in mitochondrial respiration, induced by denutrition, may aggravate the inspiratory muscle dysfunction related to disease, which may lead to respiratory failure. It thus seems very important to follow the nutritional status of patients. Indeed, an early refeeding program, perhaps associated with growth hormone administration, can be expected to improve respiratory muscle function compromised by malnutrition, as suggested by studies in human (16, 22, 37) and animal (34). Further studies are needed to verify if an improvement in mitochondrial respiration contributes to the beneficial effect of a refeeding program on respiratory muscle function.