The main new finding of this study is, that after exercise training the PRKAG3 mutation influences metabolic and fibre characteristics to a varying degree among muscles with different functions. Fibre type composition and the physical activity level of the muscle are factors that may contribute to the differences seen in glycogen content and enzyme activities between muscles. In agreement with earlier studies on untrained pigs, the pigs carrying the PRKAG3 mutation had in comparison to the non-carriers, higher content of glycogen in both m. longissimus dorsi and in m. biceps femoris [1, 18, 19]. Previous studies have shown that the mutation does mainly affect white glycolytic muscles such as m. longissimus dorsi and has no effect on a red muscle such as m. semispinalis capitis . M. masseter is considered to be a red muscle based on a high CS activity and low glycolytic potential whereas m.longissimus is a glycolytic muscle based on a low CS activity and high glycolytic potential . M. semitendinosus of non-carriers had similar metabolic and fibre characteristics as seen in m. longissimus dorsi and is thus considered to be a white glycolytic muscle. As expected the carriers of the PRKAG3 mutation had higher glycogen content also in this muscle. The fact that the total glycogen content seemed to be somewhat lower in m. semitendinosus than in m. longissimus dorsi is in agreement with earlier observations of non-carriers of the PRKAG3 mutation . Notable was that the carriers of the PRKAG3 mutation had higher glycogen content than the non-carriers also in m. masseter, which is considered to be a red oxidative muscle. However, as seen in the present study, some glycolytic type II fibres exist in this muscle. These may be influenced by the mutation, resulting in overall higher glycogen content. The higher synthesis of glycogen in the muscles of the carriers of the PRKAG3 mutation is likely related to a higher capacity for phosphorylation of glucose as indicated by the higher HK activity observed in the muscles. The PRKAG3 mutation may also have an effect on glycogenolysis in association with high muscle glycogen storage as indicated by the higher phosphorylase activity found in both m. longissimus dorsi and m. biceps femoris in the carriers. The higher phosphorylase and HK activity observed in m. biceps femoris of the exercise trained carriers is in agreement with results on young untrained carriers . This indicates that the PRKAG3 mutation has a great influence on these enzymes and may suggest that the carriers of the mutation have an increased glycogen turnover. The increased oxidative capacity (indicated by the higher CS activity) and the decreased glycolytic capacity (indicated by lower LDH activity) in m. longissimus dorsi of the carriers of the PRKAG3 mutation, is also in agreement with earlier studies of untrained pigs [4, 9]. In a previous study the HAD activity was higher in m.longissimus dorsi  but this was not seen in any of the muscles in the present study. A study with transgenic mice models showed that mice with a chronically AMPK-activating mutation caused a shift from fibre type B to IIA/X fibres . These mice had higher activity of CS and increased hexokinase protein expression regardless if they had exercised or not. AMPK signalling was suggested to play an important role for transforming skeletal muscle fibre types as well as for increasing hexokinase II protein expression and oxidative capacity. These findings are in agreement with effects of the PRKAG3 mutation on muscle characteristics in the present study especially in m. longissimus dorsi. Studies on transgenic mice (Tg-Prkag3225Q) have shown that the PRKAG3 mutation is associated with a greater basal AMPK activity . Previous studies of fibre characteristics in m. longissimus dorsi in pigs that carry the PRKAG3 mutation indicate that alterations may occur in the subgroups of type II fibres [4, 23]. This is also in agreement with the findings of the present study. Notable was that the carriers of the PRKAG3 mutation had less IIB fibres, not only in m. longissimus dorsi, but also in m. biceps femoris, compared to non-carriers. The fact that the oxidative capacity evaluated by the CS activity in the present study did not differ between genotypes in m. biceps femoris but differed in m. longissimus dorsi, may be related to these muscles being differently involved during locomotion . It has earlier been indicated that adaptations to training differ between muscles . Endurance trained pigs had in comparison to non-trained pigs an increased oxidative capacity and a higher glycogen content in m. biceps femoris, but no differences were seen in m. longissimus dorsi and in m. semitendinosus, muscles thus considered to be less involved during training on a treadmill .
In both genotypes training adaptations in the fibres of m. biceps femoris may have caused a similar oxidative capacity in response to the increased energy demand during locomotion. A previous study of pigs has shown that glycogen is lowered in both genotypes in type I, IIA and in some IIB fibres in m. biceps femoris during the same type of exercise as used in this study, which indicates that these fibres have been recruited . Adaptations to exercise training in this muscle may have decreased the effects of the PRKAG3 mutation on muscle metabolic and contractile properties.
The carriers had less type IIB fibres in m. longissimus dorsi which indicates that one effect of the PRKAG3 mutation may be associated with transformation of type IIB towards type IIX and IIA fibres, as carriers also had more type IIAX fibres. The muscle fibres that are classified as MHCIIAX may be a mixture of pure IIX and/or hybrid IIA+IIX and IIX+IIB as the antibody A4-74 identifies both IIA and IIX fibres [24, 25]. Transition of myosin heavy chains is said to follow a sequential, yet reversible, pathway: I↔IIA↔IIAX↔IIX↔IIB [26, 27]. Interestingly, genetic selection for growth performance in pigs, shifts fibre type towards type IIB fibres [28, 29] whereas endurance exercise training has been shown to shift the fibre type towards type IIA in rats  and in man . Studies in pigs also indicate that fibre type shifts from type IIB to IIA may occur with training [32, 33]. Oxidative capacity is known to increase with training and among fibre types oxidative metabolism is high in type I fibres and decreases in the rank order from type I to type IIA to type IIX to type IIB fibres . Intensive selection for a higher meat content and lean muscle growth in modern pigs has not only caused shifts in contractile fibre types, but also induced a change in muscle metabolism towards a more glycolytic and less oxidative fibre type . In contrast, the PRKAG3 mutation has been shown to decrease IIB and increase IIA and IIX mRNA expression, which also implies that the genotype promotes a more oxidative phenotype . The changes seen in muscle characteristics in the carriers with the PRKAG3 mutation thus resemble those seen when muscles in pigs adapt to an increased physical activity level. In rabbits contractile activity induces a fast-to-slow and glycolytic-to-oxidative fibre transition in skeletal muscle . In the present study the pigs with the mutation in the γ-subunit of AMPK seem to have developed a more oxidative phenotype independent of contractile activity. This is supported by the higher CS activity and the higher oxidative capacity of type IIB muscle fibre types according to the NADH-tetrazolium reductase staining found in m. longissimus dorsi of the carriers. Notable, many type IIAX fibres in the carriers were classified as having a low oxidative capacity. However, these IIAX fibres in the carriers probably also had an overall higher oxidative capacity as they were larger in size. As seen from Figure 1, the staining intensity for NADH-tetrazolium reductase is usually homogeneous within a fibre, but more intense at the periphery due to a higher density of mitochondria there.
The muscle fibre composition of m. masseter, m. semitendinosus, m. biceps femoris and m. longissimus dorsi identified according to the ATPase stains is in good agreement with earlier studies [10, 37]. If differences among subgroups of type II fibres (including hybrids) also occurred in m. semitendinosus and m. masseter between the two genotypes is not known, since only the ATPase staining technique was used to identify fibre types in these muscles. If fibre types had been identified only from the ATPase stains in m. longissimus dorsi and m. biceps femoris, no changes in subgroups of type II fibres between the two genotypes would have been revealed. This clearly shows that the use of antibodies against the different myosin heavy chains will give a more detailed picture of the fibre type composition in a muscle. When pure IIX and hybrid IIA+IIX and IIX+IIB fibres cannot be detected alterations in muscle fibre types might be overlooked.
The MHCIIB isoform was previously said to exist only in small animals such as mouse, rat, guinea pig and rabbit [38, 12]. However, studies have shown that large animals i.e. pigs and llamas do exhibit MHCIIB fibres and mostly in glycolytic muscles [34, 39, 40]. In fact the m. longissimus dorsi has been shown to contain 51% of type MHCIIB in pigs . This is in good agreement with 47% type IIB fibres observed in the m. longissimus dorsi of non-carriers in the present study. As seen from Figure 1 the NADH-staining intensity showed marked differences in oxidative capacity among the fibre types and as expected type IIB fibres had mainly a low oxidative capacity. Whether type IIAX fibres with low staining intensity for oxidative capacity correspond to pure type IIX and/or hybrid type IIX + IIB needs to be investigated in future studies using antibodies that can separate MHCIIA and MHCIIX fibres.