INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

UNLIKE MOST EUKARYOTIC CELLS, skeletal muscle fibers are multinucleated and contain hundreds or even thousands of myonuclei within a common cytoplasm. The reasons for having so many nuclei within a single cell are not well understood but may be related to the enormous size of these cells. The presence of multiple nuclei, therefore, may be a mechanism to avoid diffusion and translocation limitations of nuclear products. Thus each myonucleus may be responsible for the maintenance of a limited volume of cytoplasm, a concept referred to as the "DNA unit" (8) or "myonuclear domain" (15), which is defined as the quantity of cytoplasm regulated by a single myonucleus. The observation that newly transcribed cytoplasmic proteins can be localized to the space around individual myonuclei and neighboring myonuclei in vitro (23, 25) and in vivo (32) gives functional significance to the concept of a myonuclear domain.

Myonuclear domain size appears to differ between different fiber types in control animals. Studies in which stereological and morphological techniques were used have shown that, in general, fibers from predominantly slow muscles have a greater myonuclear density than do fibers from predominantly fast muscles in rats (4, 6), rabbits (6), and chickens (19). More recently, confocal microscopy was used to demonstrate that slow fibers from the plantaris muscle of control rats (37) and cats (2) have 25-50% more myonuclei per millimeter and possess a smaller average myonuclear domain size than do fast fibers. The reason for the difference in myonuclear domain size between slow and fast fibers is not currently known but may be related to the greater quantity of enzyme synthesis necessary in more highly oxidative fibers (37) and/or the greater protein turnover level of slow vs. fast muscles (5).

In addition, it is also unclear whether these relationships are maintained within distinct adult fiber types during muscle adaptation. During muscle fiber hypertrophy, one cellular strategy for increasing protein synthesis would be to increase RNA synthesis from existing myonuclei. In this scenario, myonuclear domain size would increase as each myonucleus encoded more proteins for a larger volume of cytoplasm. An alternative strategy would be to maintain the level of RNA synthesis per myonuclei at a normal level but to increase the amount of DNA available to the cell by adding more myonuclei. In this scenario, myonuclear domain size would remain unchanged or even decrease. Evidence to date has strongly suggested that fiber hypertrophy is accompanied by commensurate changes in myonuclear number such that myonuclear domain size remains unchanged (2, 20). However, whether this is true for different types of fibers within a muscle that has a mixture of fiber types is not currently known.

It is generally assumed that satellite cells provide a source for new myonuclei in adult skeletal muscle (9, 22, 33), and it has been suggested that their fusion with the fiber may be crucial to the hypertrophic response (26). In particular, these latter studies have demonstrated that, if cell proliferative activity is ablated by using irradiation, increases in myonuclear number and fiber hypertrophy do not occur (26, 27). These data indicate that the increase in fiber volume during fiber hypertrophy appears to be dependent on the modulation of myonuclear number via satellite cell activation. What is less clear is whether there is any limit to the response of muscle stem cells in producing new myonuclei. In a previous study, myonuclear number increased two- and fourfold in slow and fast cat plantaris fibers, respectively, in response to ablation of both the gastrocnemius and the soleus (2), suggesting that a dramatic increase in myonuclear number is possible under conditions of extreme loading. In this functional overload (FO) model, high levels of loading of the plantaris muscle during locomotion have been documented (28, 29). This load-induced increase in myonuclei also is supported by other studies; e.g., administration of daily injections of growth hormone/insulin-like growth factor has been shown to further augment the hypertrophy of the soleus in response to synergist ablation (1), and treadmill exercise enhances the hypertrophic response of the FO rat plantaris (31).

Although it is clear that myonuclear modulation is an important cellular strategy in skeletal muscle hypertrophy, how the changes in myonuclear number relate to myosin heavy chain (MHC) isoform expression, including coexpression, within individual fibers is not known. Adult rat skeletal muscles contain at least three fast (IIa, IIx, and IIb) and one slow (I) MHC isoforms that are responsive to alterations in the functional demands placed on the muscle (28, 35). The interrelationships among these specific isoforms, myonuclear number, and muscle fiber size in control and hypertrophied muscles are unknown. The purposes of the present study, therefore, were to determine 1) whether myonuclear number is related to the specific MHC isoform profile expressed within rat muscle fibers, 2) how the MHC composition of a muscle fiber is related to the magnitude of change in myonuclear number and cell volume during FO, and 3) whether changes in myonuclear number and fiber size during FO can be induced by daily exercise to increase further than by routine cage activity alone. The rat plantaris was chosen because of its responsiveness to FO and because it has a full complement of myosin phenotypes (31), thus allowing the identification of differences in the potential of fibers to generate additional myonuclei on the basis of myosin phenotype. The treadmill training protocol was used in the present study because it "maximizes" the activation of the plantaris and induces a larger fiber hypertrophic response than does FO alone (31).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experimental design and surgical procedures. Adult female Sprague-Dawley rats (initial mean body weight of ~180 g) were used. These rats were a subset of those used for the determination of MHC and fiber-size adaptations in a previous study (31). All rats were run on a motor-driven rodent treadmill at various speeds and grades for a few minutes per day for 1 wk to acclimate them to the treadmill. Rats that would not run on the treadmill or that had high or low body weights were eliminated from the study before group assignment. The rats then were assigned randomly to one of three groups: 1) control (Con), 2) FO, or 3) FO plus treadmill training (FO_Tr). The plantaris muscle in both legs of the rats in the FO groups was functionally overloaded by surgical removal of its major synergists, i.e., the soleus and both heads of the gastrocnemius muscles, under anesthesia (ketamine, 75 mg/kg body wt; and xylazine, 10 mg/kg body wt) and aseptic conditions as described previously (30, 31). Rats were allowed 4 days of recovery before treadmill training was initiated. The rats then were maintained for 10 wk because this represents the period at which both the hypertrophic and myosin adaptations appear to plateau (39). All rats were given water and food ad libitum. The rats were housed in groups of three to four in standard rodent cages. This study was approved by the Animal Use Committee at University of California, Los Angeles, and followed the American Physiological Society Animal Care Guidelines.

Single-fiber isolation. Individual muscle fiber segments were isolated as previously described (1-3, 37). Briefly, the proximal one-third of the left plantaris muscle was gradually thawed to room temperature and placed in 100% relaxing solution (11) in a Sylgard-coated culture dish. Single-fiber segments were mechanically dissected by using fine forceps under a dissection microscope, placed on gelatin-coated slides, and stored in a 5°C freezer until analysis. Approximately 30-50 fiber segments, 5-10 mm long, were plucked from each muscle. Care was taken to isolate fibers from all areas of the muscle block and to place fibers on the slide as close to resting length (2.0- to 2.5-µm sarcomere length) as possible. This mechanical isolation technique strips off any nonmuscle nuclei and satellite cells because the basal lamina is removed (16). To ensure that basal lamina was not present, 16 fibers were chosen at random and immunostained with rabbit anti-laminin (Sigma Chemical, St. Louis, MO). Only one fiber had a small, 300-µm-long section covered with tattered laminin remains.

Confocal microscopy. Slides were removed from the 5°C freezer, air dried, and stained for both cytoplasm and myonuclei by using propidium iodide and acridine orange as described previously (1-3, 37). After staining, fibers were analyzed by using a Sarastro 2000 confocal microscope (Molecular Dynamics, Sunnyvale, CA) through three areas (173-µm field of view) near the middle of each fiber segment. Values for myonuclear number, fiber cross-sectional area (CSA), and sarcomere length were recorded for the three areas and averaged to obtain one value for each fiber. A cross-sectional view of the fiber was generated from the series of optical sections and measured by using imaging software (Molecular Dynamics). Any damaged areas, torn or distorted parts of the fiber, were not analyzed. To correct for any differences in the state of stretch of the fibers, values for myonuclei per millimeter and CSA were normalized to a 2.5-µm sarcomere length because each has been shown to have a linear relationship with sarcomere length (37). This relationship was less predictable in fibers excessively stretched or contracted; thus fibers with a sarcomere length of >3.3 µm or <1.75 µm were not used in the analyses. Domain size was calculated as fiber volume for the field of view divided by the number of nuclei per field of view: (CSA × 173 µm)/(number of myonuclei in field).

Gel electrophoresis. High-resolution gel electrophoresis on single-muscle fiber segments was carried out as previously described (1, 3) by using the technique of Talmadge and Roy (36). After confocal microscopy, fibers were destained and dehydrated in 50% ethanol and then scraped off the slide and placed into a microcentrifuge tube containing gel sample buffer (17). The centrifuge tubes were heated at 80-95°C for 2-3 min before 7-8.5 µl of sample buffer were pipetted into gel lanes. By using a Bio-Rad Mini Protean II Dual Slab electrophoresis unit (Bio-Rad, Richmond, CA), samples were run at 80 V and for 24 h in a separating gel containing 8% acrylamide and 30% glycerol, which yielded four distinct bands corresponding to MHC isoforms I, IIb, IIx, and IIa (Fig. 1). Gels were stained with Rapid Coomassie (Diversified Biotech, Boston, MA) as per the vendor's instructions to visualize the protein bands. The fiber was considered to be coexpressing if at least two bands of any intensity were observed in the lane using quantitative densitometry (see below). On the basis of the MHC expression profile, each fiber was placed into one of three groups: type I⁺ (containing some type I MHC with or without any combination of fast MHCs), type IIa⁺ (containing some type IIa with or without some type IIx and/or IIb, but no type I MHC), or type IIx/b (containing only type IIx and/or IIb MHCs). One fiber expressed type IIa, IIx, and IIb MHCs, but it was not used in the statistical analysis.

View larger version (35K): [in this window] [in a new window]   Fig. 1.   SDS-PAGE of myosin heavy chain (MHC) isoforms in representative single-fiber segments (lanes 2-12) of control (Con; A) and functionally overloaded and trained (FOTr; B) rats. Lane 1, homogenate of rat soleus and medial gastrocnemius. All 4 adult rat MHC isoforms (I, IIa, IIx, and IIb) were present in both groups.

Statistical procedures. All values are reported as means ± SE unless indicated differently. Two-way ANOVA was used to assess the main effects of MHC type and experimental group by using an alpha level of 0.05 for statistical significance. The Fisher post hoc test was used to determine differences between individual groups. Pearson product correlations were calculated to determine the relationships between fiber CSA and myonuclear number.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Body and muscle weights. The mean body weights were similar among the three groups (Table 1). The mean absolute plantaris weights were 65 and 96% larger in the FO and FO_Tr compared with Con, and the FO_Tr muscles were significantly larger (~20%) than the FO muscles. The differences in mean relative muscle weights [absolute weight (mg)/body weight (g)] paralleled the differences in absolute weights.

MHC composition of single fibers. Representative single-fiber MHC gels are depicted in Fig. 1. Note that there are multiple combinations of expression of two MHCs in fibers in both the control and the FO_Tr groups. The number of type I⁺, IIa⁺, and IIx/b fibers, identified as described in METHODS, in each group are listed in Fig. 3A.

Myonuclear number. Representative confocal microscope images of a segment from a type IIx/b fiber from each group are shown in Fig. 2. Note the increase in size and myonuclear number in the fibers from the FO groups compared with control. Type I⁺ fibers in the Con group had 42 and 21% more myonuclei than did type IIa⁺ and type IIx/b fibers, respectively (Fig. 3A). Type I⁺ fibers in the FO group also contained significantly more myonuclei than did type IIa⁺ (39%) and type IIx/b (23%) fibers. In contrast, myonuclear number was similar across fiber types within the FO_Tr group. All fiber types of both FO groups had significantly more myonuclei per millimeter than did the Con group. Specifically, the FO group had 64, 68, and 61% and the FO_Tr group 64, 109, and 79% more myonuclei in types I⁺, IIa⁺ and IIx/b fibers than did Con, respectively. No significant differences were observed between the FO and FO_Tr groups.

View larger version (36K): [in this window] [in a new window]   Fig. 2.   Representative confocal microscope composite images of segments of type IIx/b fibers from the Con (A), functional overload (FO; B), and FOTr (C) groups. Maximum intensity values for each series of confocal images were used to produce the images. Fibers were stained with a combination of propidium iodide and acridine orange to label fiber myonuclei and cytoplasm, respectively. Scale bar in the Con panel, 20 µm. Abbreviations are same as in Fig. 1.

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Bar graph of myonuclei/mm (A), cross-sectional area (B), and cytoplasmic volume-to-myonucleus ratio (C) for rat plantaris fiber segments in Con (open bars), FO (hatched bars), and FOTr (filled bars) groups. Fibers are grouped by MHC content. Values are means ± SE. Abbreviations are same as in Fig. 1. a, No. of fibers. *, , I, and A Significant difference from Con, FO, type I+, or type IIa+ groups, respectively, P  0.05.

CSAs. In the Con group, the type IIx/b fibers had a significantly larger (41%) CSA than did the type I⁺ fibers (Fig. 3B). Furthermore, in the FO_Tr group the type IIx/b fibers also were significantly larger than the type I⁺ (54%) and type IIa⁺ (49%) fibers. In contrast, there was no significant difference in fiber CSA among the MHC categories in the FO group (Fig. 3B). Compared with Con, the mean CSAs of type I⁺, type IIa⁺, and type IIx/b fibers were 90, 54, and 36% and 60, 51, and 75% larger in the FO and FO_Tr groups, respectively. Furthermore, the CSAs of the type IIx/b fibers in the FO_Tr group were significantly larger (28%) than those in the FO group.

Myonuclear domain. Cytoplasmic volume-to-myonucleus ratios were not significantly different across groups for a given fiber type, except for the type IIx/b fibers in the FO group being smaller than in the control group (Fig. 3C). Between fiber types, all fibers from Con and FO rats had similar ratios, although there was a tendency toward larger ratios in IIx/b fibers in Con and FO animals (P > 0.05). In the FO_Tr group, the cytoplasmic volume-to-myonucleus ratio of the type IIx/b fibers was significantly larger than that for the type I⁺ (77%) and type IIa⁺ (38%) fibers.

Relationships between myonuclear number and CSA. After FO, there was a shift toward fibers with larger CSAs and higher myonuclear numbers (Fig. 4). The relationship between fiber size and myonuclear number within a group was not very strong (r values ranging from 0.29 to 0.42). This relationship, however, was more evident when all fibers across groups were considered (r = 0.51, P < 0.05). Fiber size and myonuclear number were highly correlated within each fiber type across groups (r values ranging from 0.56 to 0.60; P < 0.05; Fig. 5).

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Relationship between fiber cross-sectional area and myonuclei/mm for all 229 fibers examined. Equations for regression lines and correlation coefficients were the following: y = 708 + 7.13x, y = 981 + 7.18x, and y = 1,136 + 6.25x and r = 0.29, 0.42, and 0.31 for Con, FO, and FOTr groups, respectively. Overall regression equation and correlation for all of fibers sampled were the following: y = 730 + 9.00x and r = 0.51, respectively. Inset, plot of mean (± SD) values for the 3 groups.

View larger version (11K): [in this window] [in a new window]   Fig. 5.   Relationship between myonuclei/mm and cross-sectional area for type I+ (A), type IIa+ (B), and type IIx/b fibers (C) in all 3 groups. Equations for regression lines and correlation coefficients are shown for each plot. All correlations were significant at P  0.05.

View larger version (25K): [in this window] [in a new window]   Fig. 6.   Relationship between MHC isoform composition of a fiber and myonuclei/mm, cross-sectional area, or cytoplasmic volume-to-myonucleus ratio for all fibers sampled. Thin vertical lines represent a MHC composition of 100% type I, IIa, IIx, and IIb (from left to right). Values between thin vertical lines indicate percent composition of faster MHC of fibers expressing 2 MHC isoforms. Con (), FO (×), and FOTr () fibers are shown. Arrows, mean values for all fibers expressing only the indicated MHC.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Relationship between fiber type and myonuclear number. Previous studies have reported that slow muscle fibers have more myonuclei than do fast fibers (2, 37). The present study supports this finding in that control fibers containing type I MHC, i.e., the type I⁺ fibers, had significantly more myonuclei per millimeter than did the other fiber types. Similarly, type I⁺ fibers in the FO group had significantly more myonuclei than did type IIa⁺ or IIx/b fibers. However, exercise in addition to functional overload (FO_Tr) affected myonuclear content in such a way that all the fiber types had similar myonuclei per millimeter after FO_Tr. Thus it appears that exercise in FO rats enhanced myonuclear number specifically in IIa⁺ and IIx/b fibers. In a previous study, Allen et al. (2) reported a two- and threefold increase in myonuclear number in slow and fast MHC fibers in the cat plantaris after 3 mo of FO plus exercise (daily walking, jumping, and playing in the colony room). Together, these data indicate that the potential for myonuclear incorporation (and fiber growth) may be greater in fast compared with slow muscle fibers, under conditions in which the fast fibers are likely to be recruited. This contention was supported by the presence of extremely large fast fibers in the FO plus exercise cat plantaris muscles (7).

Relationship between fiber type and size. As expected from previous reports (2, 10, 37), the mean fiber CSA of the type IIx/b fibers tended to be larger than both the type I⁺ and IIa⁺ fibers in the Con group (Fig. 3B). This tendency also was apparent in the Fo_Tr but not the FO group where the mean fiber size was similar across all fiber types. The mean CSA of each fiber type was significantly larger in the FO and FO_Tr groups compared with Con. There was an additional effect of exercise only for the type IIx/b fibers. Together, these findings are consistent with the idea that FO alone has a preferential effect on the "slower" and more easily recruited fibers, whereas the addition of exercise to an FO muscle involves the larger and less-frequently recruited fibers.

Relationship between fiber type and myonuclear domain. In control rats, the myonuclear domain size was similar in type I⁺ and type IIa⁺ fibers, whereas the type IIx/b fibers had a larger domain size than did the type I⁺ fibers (Fig. 3C). Qualitatively, these data are consistent with previous reports comparing slow and fast fibers (2, 13, 37). This relationship was maintained in the FO_Tr group because of proportional changes in myonuclear number and fiber CSA. However, in the FO group all fiber types had a similar myonuclear domain size (Fig. 3C) because of the relatively greater increase in myonuclear number (Fig. 3A) compared with fiber CSA (Fig. 3B) in the type IIx/b fibers. Thus, whereas myonuclear number and fiber CSA generally increase in proportion, these two variables can change independently of one another under certain circumstances.

The role of the metabolic status of a fiber in determining myonuclear number, size, and myonuclear domain. Previous studies (13, 37) and the present results (Fig. 3) indicate that the average cytoplasmic volume-to-myonucleus ratio is somewhat larger in fast (presumably low oxidative) compared with slow and/or high oxidative fibers. The principal metabolic pathways used within a fiber appear to be an important determinant of cell morphology and myonuclear number as well. Roy et al. (28) reported an inverse relationship between succinate dehydrogenase (SDH) activity, a marker for mitochondrial density, and CSA. Tseng et al. (37) extended these findings, demonstrating that fiber size is proportional to the number of myonuclei and indirectly correlated with SDH activity. Higher myonuclear density, or smaller nuclear domains, may reflect in part the higher metabolic demands and perhaps the higher demands on protein expression in slow than in fast fibers. A myonucleus in a fast, glycolytic fiber appears to be able to support a greater volume of cytoplasm than can a myonucleus in a slow fiber. The higher mean nuclear density in slow fibers also may be related to the faster contractile and metabolic protein turnover rates in predominantly slow compared with predominantly fast muscles (18, 37).

Myonuclear domain functional limits. In general, the fibers that had the largest CSAs also had the highest number of myonuclei per millimeter (Fig. 4). Also, the slope of the relationship between fiber CSA and myonuclear number was in the following order: I⁺ < IIa⁺ < IIx/b. Thus, although fiber size generally increased with myonuclear number, it was not a strictly linear relationship. In both FO groups, differences in mitochondrial content in transforming fibers may explain why CSA is not more tightly coupled with myonuclear number.

Myonuclear domains and expression of multiple MHC isoforms within single-fiber segments. Single fibers expressing two MHC isoforms did not show a consistent preference for either of the MHC isoforms (Fig. 6). The heterogeneities in MHC composition, i.e., multiple/different isoforms, found along the length of a fiber may differ from that observed in immunohistochemical data derived from a single 10-µm cross section. Thus one might expect a higher amount of expression of multiple isoforms in these single-fiber segment studies than in the cross-sectional analyses. In the present study, 36, 44, and 48% of the total number of fibers examined in the Con, FO and FO_Tr groups, respectively, expressed at least two MHC isoforms (Fig. 6). The percentage of fibers expressing multiple MHC isoforms observed in the control muscles was somewhat surprising, although recent reports have shown expression of multiple MHC isoforms in a variety of muscles from adult control rats (24).