Alterations in mitochondrial respiratory function in response to endurance training and endurance racing




Reasons for performing study: Limited information exists about the muscle mitochondrial respiratory function changes that occur in horses during an endurance season.

Objectives: To determine effects of training and racing on muscle oxidative phosphorylation (OXPHOS) and electron transport system (ETS) capacities in horses with high resolution respirometry (HRR).

Methods: Mitochondrial respiration was measured in microbiopsies taken from the triceps brachii (tb) and gluteus medius (gm) muscles in 8 endurance horses (7 purebred Arabians and 1 crossbred Arabian) before training (T0), after two 10 week training periods (T1, T2) and after 2 CEI** endurance races (R1, R2). Muscle OXPHOS capacity was determined using 2 titration protocols without (SUIT 1) or with pyruvate (SUIT 2) as substrate. Electrons enter at the level of Complex I, Complex II or both complexes simultaneously (Complexes I+II). Muscle ETS capacity was obtained by uncoupling Complexes I+II sustained respiration.

Results: T1 improved OXPHOS and ETS capacities in the tb as demonstrated by the significant increase of oxygen fluxes vs. T0 (Complex I: +67%; ETS: +37%). Training improved only OXPHOS in the gm (Complex I: +34%). Among horses that completed the race, a significant decrease in OXPHOS (Complex I: ∼−35%) and ETS (−22%) capacities was found in the tb with SUIT 2 indicating a reduced aerobic glycolysis. Significant correlations between CK activities and changes in OXPHOS were found suggesting a relationship between exercise-induced muscle damage and depression of mitochondrial respiration.

Conclusions: For the first time, OXPHOS and ETS capacities in equine muscle at different steps of an endurance season have been determined by HRR. Significant alterations in mitochondrial respiratory function in response to endurance training and endurance racing have been observed although these changes appeared to be muscle group specific.


Muscle fibre type composition, muscle fibre area and enzymes of aerobic and anaerobic metabolism vary among different muscles. These characteristics are not fixed and change with training as demonstrated by histochemical and immunohistochemical procedures (Rivero 2007; Votion et al. 2007). In horses, these studies are based on samples mainly obtained with percutaneous needle biopsy technique introduced by Lindholm and Piehl (1974) and Snow and Guy (1976). Biochemical assays of activities of enzyme and (immuno) histochemical procedures applied to biopsies are the bases of our knowledge of muscle energetics in exercising horses. Mostly, adaptations to training in endurance horses have been evaluated by the activities of mitochondrial enzymes and/or muscle fibre types characterisation (Lopez-Rivero et al. 1991, 1992; Rivero et al. 1995). Aerobic conditioning programmes in endurance-type horses induce firstly an increase in the activity of oxidative metabolism and a decrease of the anaerobic metabolism. Following 3 months of a low intensity training programme in Andalusian horses, activity of markers of oxidative capacity was increased (+22% for citrate synthase, +15% for 3-hydroxy-acyl-CoA-dehydrogenase and up to +15% for succinate dehydrogenase enzyme activity depending on fibre type) and activity of markers of glycolytic capacity decreased (−12% for phosphofructokinase activity) in the gluteus medius (gm) muscle (Serrano et al. 2000). These changes were associated with a decrease of the percentage of pure IIX fibres, whereas the percentage of pure IIA fibres increased in the deep region of the muscle (i.e. 60 mm sampling depth; Serrano et al. 2000) was also found with Arabian horses (D'Angelis et al. 2005). A longer period of training (i.e. 8 months) induces a significant remodelling of fibre type with conversion of fast (II) to slow (I) fibre type, further increase of aerobic capacity and further decrease of anaerobic capacity (Serrano et al. 2000). Effect of endurance training on skeletal muscle histochemistry is strongly influenced by breed of the horse, intensity of training and sample collection site and depth (Lindner et al. 2002).

Until now, no quantitative data is available on oxidative phosphorylation (OXPHOS) and mitochondrial electron transport system (ETS) capacities during an endurance season. Furthermore, little is known about the involvement of mitochondrial dysfunction in the development of exercise-induced muscle damage and/or exercise intolerance in endurance horses. With a percutaneous needle biopsy, 50–200 mg of muscle is obtained and although this procedure does not induce untoward effects (Snow 1982), owners of horses may be reluctant to give permission for a biopsy especially in healthy horses and/or repeated biopsies for longitudinal follow-up.

High-resolution respirometry (HRR) enables the determination of OXPHOS and ETS capacities in small samples (2 mg) of permeabilised tissues. This new advanced technology allows continuous monitoring of tissues respiration in response to titrated compounds. Recently, multiple substrate-uncoupler-inhibitor titration (SUIT) protocols were devised to screen the mitochondrial function of equine muscles (D.M. Votion et al. unpublished data). With these protocols, the mitochondrial respiratory function is evaluated by measuring the oxygen fluxes when electrons are furnished by different substrate combination to the ETS.

The present study aimed to determine the quantitative changes in muscle mitochondrial respiration of endurance horses with training and after racing over the endurance season using a minimally invasive technique: the muscle microbiopsy technique combined with HRR. Many studies on muscle energetics utilise biopsy from the gm to study adaptations to exercise and training because of its importance in locomotion. In this study, HRR was applied also to triceps brachii (tb) muscle samples. The tb also plays a role in the locomotion and practically, the tb is safer to access for the investigator than the propulsive muscles of the hindlimb. This study also aimed to compare gm and tb muscles to define whether this forelimb muscle might be used in further research to follow effect of training and/or exercise on mitochondrial function.

Materials and methods

Horses and study protocol

Mitochondrial respiration was measured on 5 occasions in muscle samples obtained by microbiopsies in 8 endurance horses (Horses A–H). Information about horses and racing performances may be found in Table 1. Additional information regarding this group of research horses may be found in a concurrent paper (Robert et al. 2010).

Table 1. Characteristics of horses used in the study and their racing performances
R1 (km)a
Mean speed (km/h)R2Distance
R2 (km)a
Mean speed (km/h)
  1. (a) Total length of the ride or distance run before elimination for lameness; (b) Horse A participated in only one race; (c) Horse C did not participate to R1 because he was lame; (d) Horse D did not participate to R2 because he was lame. G, gelding; F, female; Bwt, body weight (before training); R1 and R2, location of the first and second race, respectively.

AArabianG10399Ghlin132 km15.2bbb
BArabianG11375Vittel117 km14.0Monpazier119 km14.6
CArabianF13399cccMonpazier89/119 km
DCrossbred ArabianG13486St Galmier119 km17.6ddd
EArabianF7430St Galmier89/119 kmGhlin132 km15.2
FArabianF11443Vittel117 km14.9Monpazier89/119 km
GArabianG8487Vittel117 km14.0Monpazier62/119 km
HArabianG9446St Galmier119 km16.9Ghlin132 km15.2
Mean ± s.d.  10.3 ± 2.2433 ± 41  15.4 ± 1.5  15.0 ± 0.3

Horses were sampled before training (T0; i.e. after the period of winter rest), after 2 periods of 10 weeks of endurance training (T1 and T2) and each one followed, one week later, by a 117–132 km endurance ride (CEI**). Horses were also sampled 2 h after each ride (R1 and R2). Each muscle was sampled 5 times using the left and right side alternately (starting from the left for a group of 4 horses and from the right with the others).

Horse A joined the group of experimental horses after the beginning of the trial and was already in training but there is no T0 data for this horse and he raced only once. Consequently, Horse A was not considered for studying the effect of training over the endurance season. Horses C and D did not participate in R1 and R2, respectively, because they were lame. Five horses participated in R1 and R2 and 3 horses in a single race.

Blood samples for determination of serum creatine kinase (CK) activities were taken 2 h after the end of the rides. After clotting, serum tubes were centrifuged and harvested serum transferred to plastic tubes and stored on dry ice until measurement.

The conditioning programme consisted of 20 km outdoor-rides at low speeds (walk and trot mainly) every 2 days. One week before T1 and T2, the level of physical training of horses was evaluated by performing a standardised exercise test (SET1 and SET2, respectively) on a 800 m track. After a 30 min warm-up at walk and trot, the horses galloped twice (18 km/h) for 30 min followed by 20 min recovery. Lactate was measured after each step of gallop using a portable Accusport analyser1 (Evans and Golland 1996) whereas CK activities were measured 2 h after completion of the SET.

Muscle microbiopsy

Muscle tb and gm samples (± 20 mg) were obtained by needle biopsy after local mepivacaine anaesthesia (scandicaine 2%)2 using a 14 G Pro-Mag Ultra Biopsy Needle3 mounted on an automatic Pro-Mag Ultra Biopsy Instrument3. The sampling depth was fixed at 50 mm depth for both muscles. For the tb, muscle microbiopsy specimens were obtained at a standardised site within the long head of the muscle (D.M. Votion et al. unpublished data) determined as the intersection between a vertical line raised from the tricipital line and a line running from the point of the shoulder to the elbow. For the gm, the microbiopsy was made according to Lindholm and Piehl (1974), one-third of the distance along a line running from the tuber coxae to the root of the tail.

Muscle samples were immediately transferred into ice-cold relaxing solution BIOPS (Veksler et al. 1987; Letellier et al. 1992) containing 10 mmol/l CaK2-EGTA, 7.23 mmol/l K2-EGTA, 20 mmol/l imidazole, 20 mmol/l taurine, 50 mmol/l K-MES, 0.5 mmol/l dithiothreitol, 6.56 mmol/l MgCl2, 5.77 mmol/l ATP and 15 mmol/l phosphocreatine adjusted to pH 7.1. The samples were kept around 4°C until preparation which was delayed by 1–3 days (i.e. horses were sampled in France and HRR was performed in Belgium).

High-resolution respirometry

The microbiopsy material was processed as indicated by D.M. Votion et al. (unpublished data). In summary, muscle samples were permeabilised by dissection into ice-cold BIOPS using forceps with sharp tips followed by saponisation at 4°C in BIOPS (50 µg/ml) for 30 min. The permeabilised fibres (Pfi) were then rinsed for 10 min at 4°C in ice-cold mitochondrial respiration medium (MiR054; Gnaiger et al. 2000).

About 2 mg wet weight (Ww; AB204-S5) of Pfi was added to each Oxygraph-2k4 chamber containing 2 ml of MiR05 maintained at 37°C. The oxygen concentration (mol/l) and the oxygen flux (pmol O2/( Ww)) were recorded online in the closed chambers using the DatLab software4. Oxygen level in the chambers was maintained between 200 and 500 µmol/lO2 to avoid oxygen restriction due to diffusion limitation in Pfi (Gnaiger 2003).

The OXPHOS and ETS capacities per muscle mass were determined by sequential addition of different substrate combinations that furnish electrons upstream of the Q-junction i.e. at the level of Complex I (CI), Complex II (CII) or simultaneously to CI and CII (CI+II), in the presence of a saturating amount of ADP (2.5 mmol/l). The 2 protocols performed differed critically by the introduction of pyruvate thus studying mitochondrial respiration with (SUIT 2) or without (SUIT 1) assessment of aerobic glycolysis through the activity of the pyruvate dehydrogenase (PDH) complex. After CI respiration was initiated, cytochrome c (10 µmol/l) was added to test the preservation of the outer mitochondrial membrane following the permeabilisation procedure (Kuznetsov et al. 2004). Injury of the membrane would lead to release of cytochrome c from the mitochondria and to significant stimulation of respiration following addition of exogenous cytochrome c in the respiration medium. Maximal OXPHOS capacity (i.e. CI+II respiration) was reached following addition of succinate oxidised at the level of CII. The muscle ETS capacity was obtained by uncoupling the system with stepwise addition of the uncoupler carbonyl cyanide p-trifluoromethoxyphenylhydrazone (0.05 µmol/l at first step followed by 0.025 µmol/l steps). Subsequently, CI was inhibited by rotenone to obtain CII respiration alone. In a previous study, it was showed that respiration sustained by CII was not influenced significantly by uncoupling (D.M. Votion et al. unpublished data). Finally, residual oxygen consumption which was subtracted from oxygen fluxes recorded online was obtained by addition of antimycin A (2.5 µmol/l) to block electron transfer to Complex III (CIII). The SUIT 2 protocol was not performed at T0 but only for T1 and T2 samples when a second Oxygraph-2k4 was available that enabled additional titration protocols to be run.

The steady-state oxygen fluxes (in the presence of ADP and cytochrome c) considered for statistical analysis according to the sequence of titrated compounds are (enclosed in brackets according to D.M. Votion et al. unpublished data):



The subscripts P and E are the OXPHOS and ETS capacities with the following substrates and inhibitors: glutamate (G; 10 mmol/l), malate (M; 2 mmol/l), succinate (S; 10 mmol/l), rotenone (Rot; 0.5 µmol/l) and pyruvate (P; 5 mmol/l).

Statistical analysis

Data are presented as mean ± s.d. for the different sampling times. A one-way ANOVA for repeated measurements was done within each respirometric parameter among the different sampling time (i.e. T0, T1, T2 for training effects; T1, R1 and T2, R2 for racing effects) and post hoc comparisons of means provided by Turkey's least significant difference test. Respirometric parameters obtained with tb and gm samples at T0, T1, T2, R1 and R2 were compared using paired t test analysis. The significance of associations between the percentage of changes induced by racing among horses that successfully completed the race and serum CK activities 2 h post racing was determined by linear regressions using GraphPad Instat (3.05)6. Significance was set at P<0.05.


Effects of training over the endurance season are summarised in Table 2 as well as effect of racing. Although, there was no significant difference between the tb and gm for OXPHOS and ETS capacities at all sampling periods, training effect differed between both muscles.

Table 2. Tissue-oxidative phosphorylation (OXPHOS) and electron transport system (ETS) capacities (mean ± s.d.) in triceps brachii (tb) and gluteus medius (gm) muscles during an endurance season with the substrate-uncoupler-inhibitor titration protocol no.1 (SUIT 1) and no.2 (SUIT 2); n = 8 horses. Oxygen flux is expressed per wet weight (Ww) of muscle tissue (pmol O2/(
Protocol  T01T12,3R14T2R25,6,7
  1. Comments: (1): there is no T0 for Horse A because this horse was already in training when he joined the group of experimental horses; (2): values of Horse F in the gm were considered as pathological at T1 but neither at T0 nor at T2; (3): data of Horse H at T1 (tb, SUIT 2) were lost because of a technical problem; (4): Horse C did not participate in R1 because he was lame; (5): there is no R2 data for Horse A because he participated in only one race; (6): Horse D did not participate in R2 because he was lame; (7): cytochrome c assays (SUIT 2) indicated that data of Horse B (tb and gm) at R2 were not valid. Individual values have been taken into consideration for statistical analysis, accordingly. Effect of training (T0, T1, T2): values with different superscripts are significantly different from the other (P<0.05). Effect of racing (T1 vs. R1 and T2 vs. R2): values with an asterisk are significantly different (P<0.05). T0: before training, T1 and T2: after the first and second training period, respectively; R1 and R2: after the first and the second race, respectively; CI: Complex I; CII: Complex II; CI+II: Complexes I+II; G: glutamate, M: malate; S: succinate; Rot: rotenone; the subscripts P and E are the OXPHOS and ETS capacities, respectively.

SUIT 1 tbOXPHOS capacity[CI-GM]P42 ± 14a71 ± 12b68 ± 1865 ± 12b59 ± 20
[CI+II-GMS]P98 ± 32a133 ± 16a147 ± 27129 ± 27a133 ± 42
ETS capacity[CI+II-GMS]E108 ± 30a148 ± 15b165 ± 26150 ± 21b146 ± 35
[CII-S(Rot)]E81 ± 24a100 ± 10a125 ± 22*109 ± 21b100 ± 26
SUIT 1 gmOXPHOS capacity[CI-GM]P51 ± 12a72 ± 14b57 ± 1376 ± 20b64 ± 8
[CI+II-GMS]P109 ± 33a128 ± 12a121 ± 24149 ± 44a136 ± 21
ETS capacity[CI+II-GMS]E123 ± 32a148 ± 12a137 ± 30164 ± 40a154 ± 11
[CII-S(Rot)]E85 ± 19a103 ± 7a95 ± 17112 ± 30a103 ± 13
SUIT 2 tbOXPHOS capacity[CI-PM]P85 ± 18a56 ± 2380 ± 26a43 ± 13*
[CI-PMG]P96 ± 20a64 ± 2791 ± 33a51 ± 13
[CI+II-PMGS]P152 ± 39a118 ± 39134 ± 47a111 ± 38
ETS capacity[CI+II-PMGS]E188 ± 36a150 ± 41185 ± 51a141 ± 33
[CII-S(Rot)]E100 ± 24a104 ± 3295 ± 25a89 ± 25
SUIT 2 gmOXPHOS capacity[CI-PM]P77 ± 30a64 ± 2175 ± 21a45 ± 17*
[CI-PMG]P91 ± 39a72 ± 2582 ± 31a51 ± 18*
[CI+II-PMGS]P154 ± 64a133 ± 44113 ± 28a114 ± 36
ETS capacity[CI+II-PMGS]E185 ± 63a166 ± 33168 ± 48a139 ± 35
[CII-S(Rot)]E105 ± 27a94 ± 2584 ± 19a90 ± 28

Training effect on triceps brachii

In tb, OXPHOS and ETS capacities improved with training as demonstrated by the significant increase of oxygen fluxes with SUIT 1 at T1 vs. T0 ([CI-GM]P: +67% and [CI+II-GMS]E: +36%). No further significant increases of OXPHOS and ETS capacities were observed from T1 to T2 neither with SUIT 1 nor SUIT 2. However, [CII-S(Rot)]E in SUIT 1 was significantly increased in the tb at T2 vs. T0 (+35%).

Training effect on gluteus medius

In gm, OXPHOS capacity with SUIT 1 was increased at T1 vs. T0 only for [CI-GM]P (+34%) excepted in Horse F that presented abnormally low fluxes at each step of SUIT 1 (see alterations in individual horse). No further significant increase of OXPHOS capacity was observed from T1 to T2 neither with SUIT 1 nor SUIT 2.

Effect of racing

Following completion of R1 [CII-S(Rot)]E in SUIT 1 was increased significantly in the tb (+25%). Although no other significant effects were found following R1 and R2 in the tb and the gm, oxygen fluxes of some horses were severely depressed after racing whereas others showed increased OXPHOS and ETS capacities. Following completion of R2, significant changes were only observed in SUIT 2: [CI-PM]P was decreased in the tb (−40%) and the gm (46%) whereas [CI-PMG]P was decreased only in the gm (−38%).

When considering only horses that successfully completed the race (Table 3), significant decreases of [CI-PM]P (−37%), [CI-PMG]P (−35%) and [CI+II-PMGS]E (−22%) were found in tb following R1. No statistical analysis was performed regarding T2 vs. R2 because only 3 horses successfully completed the race.

Table 3a,b. Changes (in %) induced by racing in triceps brachii (tb) and gluteus medius (gm) muscles on each steps of the substrate-uncoupler-inhibitor titration (SUIT 1 and SUIT 2) protocols in horses that successfully completed the race
(a) SUIT 1Horsestb OXPHOS capacitytb ETS capacitygm OXPHOS capacitygm ETS capacityCK (u/l)
T1 vs. R1Mean5%13%16%25%−16%−4%−7%−9% 
T2 vs. R2Mean−17%−5%−7%−16%−9%−2%0%−4% 
Correlation with CK activitiesR20.37790.55640.48660.51770.17930.04640.072210.09107 
(all horses)PNSP = 0.021P = 0.0367P = 0.0289NSNSNSNS 
(b) SUIT 2Horsestb OXPHOS capacitytb ETS capacitygm OXPHOS capacitygm ETS capacityCK (u/l)
  1. Comments: (a): values of Horse F in the gm were considered as pathological at T1; (b): not enough data for statistical analysis; (c): data of Horse H at T1 were lost because of a technical problem; (c): cytochrome c assays indicated that data of Horse B (tb and gm) at R2 were not valid. OXPHOS: oxidative phosphorylation; ETS: electron transport system; T1 and T2: after the first and second training periods, respectively; R1 and R2: after the first and the second race, respectively; CI: Complex I; CII: Complex II; CI+II: Complexes I+II; G: glutamate, M: malate; S: succinate; Rot: rotenone; the subscripts P and E are the OXPHOS and ETS capacities, respectively; R2: coefficient of determination; P: value of significance; NS: not statistically significant.

T1 vs. R1Mean−37%−35%−17%−22%8%−9%−7%9%12%5% 
 P = 0.0408P = 0.0365NSP = 0.0086NSNSNSNSNSNS 
T2 vs. R2 bbbbbbbbbb 
Correlation with CK activitiesR20.49130.24290.003090.030820.58130.02620.013670.055440.040180.07156 
(all horses)PNSNSNSNSP = 0.0463NSNSNSNSNS 

Significant correlations were found between CK activities and the percentage of changes induced by R1 particularly in SUIT 1 protocol: the mitochondrial respiration was more depressed in horses with high serum CK activities whereas increased oxygen fluxes were found in horses with the lowest CK activities. No statistical analysis was performed regarding T2 vs. R2 because only 3 horses successfully completed the race.

Alterations in individual horse

At T1, Horse F presented abnormally low fluxes at each step of SUIT 1 in the gm (below minus 2 s.d. from the mean group value) whereas other respirometric parameters were within the range of other horses. In this horse, the lactataemia during the first step of the SET1 was abnormally high (8.3 vs. 2.2 ± 1.0 mmol/l; n = 7) indicating impaired aerobic metabolism. Measurements of lactate during SET2 were within the normal range for all horses including Horse F (data not shown). Serum CK activities post SET1 and SET2 remained within normal range for all horses.


This study is the first quantitative determination of muscle OXPHOS and ETS capacities in horses at different steps of an endurance season. We showed that endurance training resulted in marked increases in muscle mitochondrial respiration although adaptations appeared to be muscle group specific. We also observed exercise-induced mitochondrial functional changes.


The microbiopsy technique was well tolerated by all horses with no untoward effect. No suture was required at the sampling site. This study demonstrated that muscle samples can be collected in field conditions and the normal training programme can be continued immediately after sampling.

In publications studying oxidative changes induced by training, the gm is the most frequently biopsied muscle (Rivero 1996; Serrano et al. 2000). Biochemical assays of mitochondrial enzymes showed that increase aerobic capacity with training may also be followed in the tb (Kim et al. 2005). It has been previously demonstrated that the gm is a heterogeneous muscle (van den Hoven et al. 1985; Serrano et al. 1996) with effects of training particularly noticeable in the deepest region of the muscle (Serrano et al. 2000). For that reason, we fixed the sampling site at 5 cm depth.

Mitochondrial pathways

The OXPHOS and ETS capacities per muscle mass were determined by sequential addition of substrate combination that furnishes electrons upstream of the Q-junction. The ETS should not be conceptualised as a linear chain, instead CI and CII should be viewed as convergent input branches to the Q-junction (Gnaiger 2009). In vivo, electron flow joins the Q-junction of the ETS from various metabolic pathways. Electrons from aerobic glycolysis (via the pyruvate decarboxylation by the PDH complex located in the mitochondria), from the tricarboxylic acid cycle (TCA) cycle and from the β oxidation of fatty acids (via the dehydrogenation of (-hydroxyacyl-CoA) join the Q-junction via CI and/or CII. Other mitochondrial dehydrogenases pass electrons into the ETS directly at the level of the Q-junction: 1) the mitochondrial glycerol 3-phosphate dehydrogenase (GPDH), which is a flavoprotein participating to the transfer of electrons from the NADH+H+ formed during the glycolysis and 2) the electron-transferring flavoprotein (ETF), which supplies electrons via β oxidation. Within the SUIT protocols proposed, there was no contribution of these 2 important dehydrogenases (i.e. GPDH and EFT) to the measured fluxes, neither was the assessment of dehydrogenases contributing to degradation of amino acids. Catabolism of amino acids may contribute to energy production during extremely prolonged low-intensity exercise, such as an endurance race (Essen-Gustavsson and Jensen-Waern 2002). Permeabilisation of the muscle cells induces a loss of cytoplasm and its soluble components, preventing the contribution of cytoplasmic pathways in the measured respiration. Measurements are fully dependent on what is supplied at each step of the SUIT protocols performed with specific experimental conditions (i.e. 37°C, high ADP concentration, no substrate limitation and control of the level of oxygen in the milieu) to achieve maximal flux (D.M. Votion et al. unpublished data). Therefore, the SUIT protocols applied in this study enabled us to measure mitochondrial respiration following electron entry at 2 specific sites: CI (with or without involvement of PDH; i.e. SUIT 2 and SUIT 1, respectively). Addition of succinate together with NADH-linked substrates reconstitutes the TCA and enables determination of maximal oxidative capacity ([CI+II-GMS]P and [CI+II-PMGS]P in SUIT 1 and 2, respectively).

Effect of training

Training effects were more pronounced in the tb than in the gm. A higher increase of oxidative capacity with training in the tb vs. a locomotor muscle (i.e. semimembranosus) has already been documented (Kim et al. 2005). The greater proportion of type I fibres in tb might result in a greater enhancement of the mitochondrial oxidative machinery and/or mitochondrial mass. However, results of the SUIT 1 protocol indicated not only quantitative but also sequential qualitative functional changes: OXPHOS capacity supported by NADH-linked substrates (i.e. electron entry at the level of CI) was significantly increased at T1 (in both the tb and gm muscles) whereas OXPHOS capacity at the level of CII (i.e. where succinate is oxidised with reduction of FAD into FADH2 and subsequent electron flow) required a longer time to be significantly increased (only at T2 in the tb).

Absence of significant changes with training in the gm regarding CII-sustained respiration was unexpected. The most commonly reported adaptation to training in the gm based on histological procedures applied to biopsy is an increase of the activity of succinate dehydrogenase (Serrano et al. 2000), the mitochondrial enzyme which is an integrate part of CII. Different muscle responses to training within a specific muscle have been found according to breed of horse, depth of sampling, type and length of training (Rivero et al. 1995; Serrano et al. 2000). The lack of significant changes at the level of CII-sustained respiration in our study might be explained by an insufficient training programme and/or an inadequate sampling site (Serrano et al. 2000). All horses underwent the same training regimen (see Robert et al. 2010) with no individual adjustment of intensity of exercise. The mean values for [CII-S(Rot)] showed a trend toward increase in SUIT 1 with training which remained nonsignificant over time. It is possible that for some horses (the level of qualification differed among horses), the training programme was too light to induce significant changes. Evaluation of OXPHOS and ETS capacities was based upon the analysis of a single biopsy but previous studies have highlighted that the metabolic profile changes dramatically with depth (van den Hoven et al. 1985; Kline and Bechtel 1988; Serrano et al. 1996, 2000) and, furthermore, training-induced modifications are not homogeneous among the gm (Serrano et al. 2000). It would be of interest to follow the effect of training in the gm with HRR at several sampling depths (e.g. 2, 4 and 6 cm) with concomitant determination of muscle fibre types. Unfortunately, in our study, we were technically limited to manage a larger number of samples. Last, but not least, HRR determines OXPHOS and ETS capacities in small samples with a mixed unknown composition. The HRR does not discriminate fibre types and might be less sensitive to specific changes as opposed to the commonly used histochemical methods which determine oxidative capacity with serial sections of the sample for determination of fibre types. However, this study demonstrated that the sampling site of the tb as described in this paper enabled us to follow a conditioning programme in endurance horses.

Alterations in individual horse

In one horse (Horse F), OXPHOS and ETS capacities were severely depressed at T1 in the gm. It is not known whether training failed to improve mitochondrial function of this horse specifically in the gm, a muscle particularly prone to cramping in endurance race. The decreased fluxes were associated with high lactataemia without an increase in serum CK activities that remained within normal ranges after SET1. No dysfunction was found with SUIT 2, indicating a satisfactorily PDH activity. Whatever the cause, the problem resolved with time since at T2, respirometric parameters of Horse H were within the range of other horses. This field study also suggests that abnormally high lactataemia after exercise may be associated with mitochondrial dysfunction in locomotor muscles.

Effect of racing

Endurance racing is one of the most demanding equestrian disciplines. The inability of the muscle cell to adequately supply ATP leads to energy depletion and muscle fatigue (Essén-Gustavsson et al. 1999). The TCA cycle is the major final common pathway for oxidation of carbohydrates, lipids and some amino acids that result in production of large amounts of ATP via OXPHOS. In the course of an endurance ride, free fatty acids become the major source of energy (Essen-Gustavsson et al. 1984). Despite the fact that it was not significant (R2= 0.76; P = 0.0517), the [CI-PM] decrease induced by R1 showed a clear trend with [CI-GM] changes: the most severe decreases of [CI-PM] were observed in horses that had a high increase (>40%) of [CI-GM] (seeTable 3). This tendency suggests that during the time of prolonged exercise, mitochondrial adaptation promotes the enhancement of mitochondrial capacities to oxidise fatty acids and depresses PDH activity.

The correlation found between CK and several respirometric parameters in SUIT 1 suggests a possible relationship between impaired mitochondrial function and exercise-induced muscle alterations. Interestingly, this correlation was only significant in the tb most probably due to its larger composition in type I fibres (that have a greater mitochondrial mass as compared to type II fibres). Up and downregulationsof OXPHOS and ETS capacities were temporary since T1 did not significantly differ from T2. It is not known if an excessive number of endurance races might result in permanent changes. Previously, Kim et al. (2005) showed in mature horses that increased muscle oxidative capacity strengthens resistance to exercise-induced ultrastructural muscle cell damage. Therefore, appropriate aerobic exercise training not only improves the metabolic properties but may also contribute to improving resistance to injury (Kim et al. 2005). The microbiopsy technique offers the opportunity to study mitochondrial respiration and to evaluate muscle fibres by light or electron microscopic evaluation. Therefore, this technique may be of great help in improving training methods and controlling integrity of muscle following racing or to ensure proper muscle recovery before engaging in a new race.


The training-induced changes in mitochondrial respiration were mainly observed in the tb muscle. Also, the correlation between serum CK activities after racing and muscle respiratory dysfunction was only significant in the tb muscle. Therefore, effect of training and racing may be followed in the tb but results of this study emphasised that SUIT 1 and SUIT 2 protocols are both necessary to comprehend the modification induced by racing. Quantitative assessment of mitochondrial respiration by HRR may contribute to a better understanding of the apparent relationship between impaired mitochondrial function, exercise intolerance and/or exercise-induced muscle damage.


The authors wish to express their sincere gratitude to the riders and owner of horses. This work was supported by the Haras Nationaux (France) and by the ‘Ministère de l'Agriculture et de la Ruralité de la Région Wallonne’ of Belgium. The authors thank the Aliments Etienne, Audevard, Michel Vaillant and Virbac for their contribution to the maintenance of horses. They also acknowledge Oroboros Instruments for their technical support and scientific advice regarding HRR.

Conflicts of interest

The authors declare no potential conflicts.

Manufacturers' addresses

1 Boehringer, Mannheim, Germany.

2 AstraZeneca, Brussels, Belgium.

3 Angiotech, Gainesville, Fl, USA.

4 Oroboros Instruments, Innsbruck, Austria.

5 Mettler Toledo, Zaventem, Belgium.

6 GraphPad Software, San Diego, California, USA.