THE ANATOMICAL RECORD 2461195-207 (1996) Correlation Between Myofibrillar ATPase Activity and Myosin Heavy Chain Composition in Equine Skeletal Muscle and the Influence of Training JOSlbLUIS L. RIVERO, ROBERT J. TALMADGE, AND V. REGGIE EDGERTON Department of Comparative Anatomy and Pathological Anatomy, Faculty of Veterinary Science, University of Cordoba, Spain (J.-L.L.R.); and Department of Physiological Science and Brain Research Institute (R.J.T., V.RX.1, University of California, Los Angeles, California ABSTRACT Background: The histochemical myofibrillar ATPase (mATPase) method is used routinely for identification of equine skeletal muscle fiber types, but important problems have been observed with the subdivision of fast fiber population when using this method. To verify the use of this qualitative method, a number of equine muscle biopsies were analyzed with a combination of histochemical, immunohistochemical, electrophoretic, and morphometric techniques. The influence of training on these interrelations was also evaluated. Methods: Five young (2-3 years old) thoroughbred horses were intensively trained for 8 months on a high-speed treadmill. Biopsies were taken from the gluteus medius muscle at the beginning, after 4 months, and at the end of the training program. Serial sections of the samples were stained by mATPase histochemistry and immunohistochemistry by using a number of monoclonal antibodies specific to selected myosin heavy chain (MyHC)isoforms. The histochemical and immunohistochemical categorization of a large number of fibers (N = 2,078) was compared fiber by fiber. The MyHC content of homogenates of the same biopsies were quantified by densitometry of a sensitive gel electrophoretic technique and compared with histochemical and immunohistochemical fiber types. Results: A large proportion of fibers examined (-20%) were misclassified by traditional mATPase histochemistry. Many fibers histochemically identified as type IIB displayed both type IIa and type IIb MyHC isoforms, and nearly all type IIAB fibers in mATPase contained only the type IIa MyHC isoform by immunohistochemistry. Correlation analyses suggested a weak relation between the histochemically assessed relative cross-sectional area occupied by the three major fiber types (I, IIA, and IIB) and the electrophoretically assessed MyHC content, whereas a stronger relation was found between immunohistochemically defined fiber types and electrophoretic data. The four fiber type populations delineated according to MyHC content (1, IIA, IIAB, and IIB) had sizes and oxidative capacities significantly different from each other. No adaptation of any parameter measured to training was found. Training had no significant effect on the number of fibers misclassified by mATPase histochemistry. ConcZusions:These data demonstrate a significant limitation in mATPase histochemistry for assessing fibers containing fast MyHC isoforms. The use of monoclonal antibodies against specific MyHC isoforms seems to be a more sensitive and less subjective method. Q 1% Wiley-Liss, Inc. Key words: Horse, Guteus medius muscle, Histochemistry, Immunohistochemistry, Electrophoresis, Muscle fiber types Over the past 2o years, enzyme-histochemica1 methhave provided much information about the plasticity of equine skeletal muscle in relation to growth, training, and performance potential (for review, see Ods 0 1996 WILEY-LISS, INC. Received October 25, 1995; accepted February 27, 1996, Address reprint requests to Jose-Luis L. Rivero, Ph.D., Department of Veterinary Anatomy, Faculty of Veterinary Science, University of Cordoba, Medina Azahara 9, 14005 Cordoba, Spain. 196 J.-L.L. RIVER0 ET AL. Snow and Valberg, 1994). In horses, as in other mammals, muscle fibers have been routinely categorized into three major types, designated I, IIA, and IIB, and the minor IIC, based on the myofibrillar actomyosin adenosine triphosphatase (mATPase) histochemical reaction proposed by Brooke and Kaiser (1970). In addition to these four fiber types, some fibers with staining characteristics intermediate between types IIA and IIB (type IIAB; Ingjer, 1979) have frequently been observed in horses (White and Snow, 1985; Rivero et al., 1993). Although the use of mATPase methods t o identify slow (type I) and fast (type 11)contracting fibers is highly reproducible, the further subdivision of type I1 fibers is controversial. The subjective evaluation, the multiplicity of factors affecting reaction patterns, and, mainly, molecular diversities of mammalian skeletal muscle fibers have all been adduced in criticism of this qualitative method (for review, see Pette and Staron, 1990). Using immunohistochemical methods, the histochemical staining intensity of the mATPase reaction in a given muscle fiber may be determined by its myosin heavy chain (MyHC)content (Billeter et al., 1981;Pierobon-Bormioli et al., 1981).Moreover, now there is sufficient evidence that the MyHC protein expressed in a single muscle fiber is correlated to its contractile properties (Reiser et al., 1985; Bottinelli et al., 1994a). To date, four major MyHC isoforms have been identified in adult skeletal muscle in a number of mammalian species: the p-, slow, or type I MyHC and the three fast IIa, IIx or IId, and IIb- MyHCs (Bar and Pette, 1988; Schiaffino et al., 1989; Gorza, 1990; Aigner et al., 1993).In contrast, in many other species including primates and carnivores, only two fast MyHCs have been identified; whereas one of these two fast MyHCs has been unequivocally identified as type IIa-MyHC, there are serious doubts regarding the identity of the second fast MyHC isoform. Because it is mainly present in histochemical type IIB muscle fibers, it has been termed type IIb-MyHC. However, two recent studies using molecular biological techniques (Smerdu et al., 1994; Ennion et al., 1995) have reported the identification of one human skeletal MyHC gene as fast IIx, whose transcripts are more abundant in histochemical type IIB fibers. Similarly, in a companion study, we have identified one slow and two fast (IIa and IIx or IIb) MyHC isoforms in the equine skeletal muscle by using monoclonal antibodies and gel electrophoresis techniques. The differential distribution of these MyHCs defines three main fiber types containing a single MyHC isoform (I, IIA, and IIX or IIB) and a number of intermediate hybrid fiber populations containing both I and IIa MyHCs (type C fibers) and type IIa and IIx or IIb MyHCs (type IIAX or IIAB fibers). Although a direct correlation between the histochemical reactivity for mATPase and the MyHC content of a given fiber has been established in rabbit single fibers (Staron and Pette, 19861, the disadvantages of the mATPase technique have recently been highlighted in humans, where gel electrophoresis of MyHC in single fibers was used in conjunction with mATPase staining (Klitgaard et al., 1990a,b; Staron, 1991; Staron and Hikida, 1992; Andersen et al., 1994a,b). These studies suggest that many fibers histochemically classified as type I or type IIB contain to some degree the fast IIa MyHC, particularly in trained muscles. Moreover, the coexpression of multiple MyHC isoforms within a single fiber also occurs under normal conditions (Biral et al., 1988).In summary, because the dominant MyHC isoform determines the histochemical mATPase reaction of a fiber (Danielli-Beto et al., 1986), this method does not reveal subtle alterations in the expression of MyHCs in muscle fibers and, therefore, may not adequately characterize muscle fiber distribution in control or in trained horses. No previous studies in horses have made qualitative and quantitative comparisons between myofiber distribution and MyHC content in equine skeletal muscle by combining histochemical, immunohistochemical, and electrophoretic techniques. This relation would lend validity to the use of routine histochemical techniques in distinguishing muscle fiber types in horse muscle. The main purposes of the present investigation were (1)t o ascertain the degree of association between the MyHC content of muscle fibers and histochemical muscle fiber type distribution by combining classical qualitative mATPase histochemistry and immunohistochemical and electrophoretic analyses of MyHC; and (2) to estimate the effect of training on the relations among these variables. The study also provides quantitative information regarding fiber size and oxidative capacity of the various muscle fiber types in horses. MATERIAL AND METHODS Horses Six clinically healthy thoroughbred racehorses (five mares and one gelding) were used in this study. Two horses were 3 years old and the other four were 2 years old a t the beginning of the study. They were all of comparable size and body weight (range = 440-465 kg). All horses were kept under identical environmental conditions and fed the same diet throughout the experiment. The horses were housed individually in 3-X-3-mboxes a t a temperature ranging from 5 to 25°C. All exercise workouts were made at a temperature ranging between 16 and 26°C. In general, the animals were fed daily with 4.5 kg of oats and 5 kg of grass-mix hay, and water and minerals were provided ad libitum. The days on which the horses were exercised, the horses consumed one more kilogram of oats. Training Program Five of the horses were included in a training program aimed at investigating the influence of intensity and duration of exercise on several physiological variables. The sixth horse was used as a control to evaluate changes associated with age. All the exercise workouts and standardized tests were done on a high-speed treadmill (Mustang 2200@, Kagra SA Fahrwangen, Switzerland). The experimental period was extended for about 8 months. After a 1-month acclimatization period, all five trained horses carried out six phases of exercise, varying in intensity and duration. Each phase consisted of 11 exercise workouts (once a day every second day), so each phase was extended for 21 days. Between two consecutive phases, horses were allowed to rest for 1 week. During this week, a standardized exercise test (SET) was performed for each horse to determine VLA2.5and VLA4or speeds that run over a defined period of time and produce a concentration of MYOSIN ATPase AND MYOSIN HEAVY CHAIN IN HORSES WITH TRAINING lactate in blood of 2.5 and 4 mmol/l, respectively (Lindner et al., 1992). This test was the parameter on which the intensity of exercise made in each phase of training was based. Of the six phases of training, three were made at an intensity of work of -VLA2.5and the other three in VLAI.In both sets of phases, the duration of each exercise session was 5, 15, and 25 min. SET and Estimation of ,V ,, and V ,, The SET consisted of several gallop workouts of 5 min duration each. Between two consecutive gallops, a resting period of 60 sec was allowed. The velocity in the first step was 6 d s e c and was increased by 0.5 d s e c in each consecutive gallop. The test was finished when the blood lactate concentration of a given horse was close to or above 4 mmol/l. Before the test and after warm up, the horses walked for 5 min at 1.5 d s e c and trotted for 5 min at 4 d s e c . Immediately after each gallop in the SET, blood samples were collected from pectoral skin (Lindner et al., 1992) to determine blood lactate concentration. Analysis of lactate was made with a EPOS 5060 lactate analyzer (Eppendorff-Netheler-Hinz GmbH) using an enzymatic test kit (Boeringer Mannheim hr 1178 750). To compare blood lactate concentrations from horse to horse, the individual blood lactate concentration-running speed relation were calculated by plotting lactate concentrate vs. running speed and drawing the curve of each horse by linear interpolation. 197 Serial sections were reacted with seven different monoclonal antibodies specific to rat MyHC isoforms. Individual cross sections were labeled for fast, slow, SC-71, N2.261, BF-35, BF-G6, and BF-B6 monoclonal antibodies. The source and specificity of these monoclonal antibodies and the immunohistochemical procedure employed are described in the companion paper (Rivero et al., 1996). Quantitative Tissue Analysis After histochemical and immunohistochemical staining of the tissue sections, a region of the cross sections containing 100-135 fibers per muscle biopsy was randomly selected for further analysis. These fibers were numbered and classified at random with histochemical and immunohistochemical methods. Myofibers were classified into types I, IIC, IIA, IIAB, and IIB (Brooke and Kaiser, 1970; Ingjer 1979) according to the mATPase staining characteristics a t the different levels of preincubation acidity. The same fibers were allotted to types I, I + IIA, IIA, IIA + IIB (or IIA + 11x1and IIB (or 11x1according to the MyHC content as revealed by reactivity against monoclonal antibodies (see Rivero et al., 1996). To compare histochemical vs. immunohistochemical data, these five fiber types were also renamed types I, IIC, IIA, IIAB, and IIB, respectively. The fiber type distribution of each muscle biopsy was established by counting and typing the relative frequency of the various fiber types in each sample. The cross-sectional areas (CSAs), and optical densities Muscle Biopsies and Tissue Preparation (ODs) for SDH activity of the same fibers were deterMuscle biopsies (75-150 mg) were obtained from the mined by using a computer-enhanced image processing right gluteus medius muscle of each horse according to system as described in detail elsewhere (Martin et al., Lindholm and Piehl(1974). Biopsies were taken before, 1985). Tissue sections were digitized within the first 2 4 months after training, and a t the end of the 8-month days after staining as gray-level (range of gray levels training program. Control horse biopsies were taken a t was 0-255) images. The gray level for each picture the same time periods. Attempts were made to extract element was converted to an OD by using a calibration tissue from approximately the same relative location of curve, which was linear up to 1.0 OD unit. Succinate the muscle by using the prebiopsy scar and depth dehydrogenase activity was interpolated from a single markings on the needle. All the muscle biopsies were endpoint measurement after a 30-min incubation, and removed a t 2-cm sampling depth. After collection, mus- the values were reported as final OD. For each muscle cle samples were frozen, stored, and transported as de- biopsy, CSA and SDH activity were averaged for each scribed elsewhere (Rivero et al., 1996). fiber type. Frozen biopsy samples were thawed to -20°C in a In addition, the relative CSA that a fiber type occucryostat and serially sectioned for histochemistry (10- pied in a muscle sample was calculated by dividing the pm-thick sections placed on cover slips), immunohis- product of the fiber type percentage and the mean CSA tochemistry (10-pm-thick sections placed on gelatin of the fiber type by the sum of these products for all coated slides), and gel electrophoresis of MyHC (20- fiber types (Sullivan and Armstrong, 1978). pm-thick sections placed in precooled microcentrifuge Myosin Heavy Chain Electrophoresis tubes). MyHC analysis was performed on the biopsy samples Histochemistry and lmrnunohistochernistry using the 8% sodium dodecyl sulfate-polyacrylamide Serial cross sections were stained qualitatively for electrophoretic (SDS-PAGE) protocol of Talmadge and the demonstration of mATPase activity after alkaline Roy (1993) and detailed in the companion paper (Riv(pH 8.75) and acid (pH 4.2 and pH 4.5) preincubation ero et al., 1996). Three MyHC isoforms were clearly by using a modification (Nwoye et al., 1982) of the separated in the gels and identified as types I, IIa, and Brooke and Kaiser (1970) method. Serial sections were IIb. The gels were scanned with an Alpha Innotech stained for quantitative succinate dehydrogenase IS-1,000 videoscanning densitometric system, and a (SDH) activity (Blanco et al., 1988). Briefly, tissue sec- quantification of MyHC isoforms was obtained in reltions were incubated in a medium containing 48 mM ative terms for each sample. succinate, 1.5 mM NBT, 5 mM EDTA, 1mM l-methoAlthough five fiber types were distinguished by both syphenazine methosulfate (mPMS; Dojindo Laborato- histochemical and immunohistochemical methods, the ries, Japan), 0.75 mM sodium azide, and 100 mM hybrid fiber types (IIC and IIAB) made up a minor sodium phosphate buffer (pH 7.60) for 30 min (qualita- portion of the entire biopsy. To correlate quantitative tive procedure) a t room temperature. information from MyHC electrophoresis vs. histochem- 198 J.-L.L. RIVER0 ET AL. ical and immunohistochemical data, the type IIC fibers (always fewer than 1%) were split so that one-half of these were combined with type I and one half with type IIA fibers. Similarly, one-half of the type IIAB fiber population was combined with type IIA fibers and the other half with type IIB fibers. These revised data for fiber type distribution (which gives a more accurate representation of the three major fiber types) were then used in combination with the CSA data to obtain the relative CSA of the three major fiber types. Statistical Analyses The statistical package STAT 5.2 SYS developed for Macintosh was used for all statistical analyses. Descriptive statistics were used to derive means (and standard deviations) for all variables. One-way analysis of variance (ANOVA) with repeated measurements was used to determine whether significant effects of training existed. In the presence of a significant F ratio, post hoc comparisons of means were provided by Tukey’s t procedure. Correlation analysis was performed to compare the histochemistry, immunohistochemistry, and electrophoretic information. In addition, coefficients of variation between histochemical and immunohistochemical data were calculated according to the formula c v = V E-XZ F i where CV is the coefficient of variation, d is the difference between duplicate determinations, n is the number of repeated analyses, and 5i is the mean value. Finally, because there were significant variations in both fiber CSA and SDH between muscle biopsy specimens, the CSA and SDH activity of each fiber was standardized to Z scores based on the following algorithm: -x z = -xi , S where Xiis a n individual CSA or SDH measure, X is the mean for all fibers within the biopsy, and s is the standard deviation (Sieck et al., 1986). Differences among fiber types for the CSA and SDH activity were then examined by a one-way ANOVA for all data standardized to Z scores. Statistical significance was accepted at P < 0.05. RESULTS Fiber Typing In all of the 18 muscle cross sections examined, histochemical and immunohistochemical analyses of muscle revealed a classical mosaic pattern, and the shape and size of fibers appeared normal. Based on histochemical analysis (mATPase reaction), the muscle fibers could be divided into five categories: I, IIC, IIA, IIAB, and IIB (Fig. 1B-D). A continuum in the staining intensity for mATPase after acid preincubation (pH 4.5) was observed between the type IIA and type IIB fiber population (Fig. 1D). All of these intermediate fibers were identified as type IIAB. Five different fiber populations were also demonstrated immunohistochemically by using specific monoclonal antibodies (Fig. 1F-L). Fibers that reacted exclusively with the slow MyHC antibody were termed I. Fibers th a t were unreactive with the slow MyHC antibody and reacted with fast, SC-71, N2.261, BF.35, and BF-G6 MyHC antibodies were identified as IIA, and those fibers that did not react with all monoclonal antibodies except the fast MyHC antibody were identified as IIB. Two subgroups of fibers coexpressing two different MyHCs were also identified, and these were designated a s IIC (fibers coexpressing types I and IIa MyHCs) and IIAB (fibers coexpressing both fast MyHCs). No fibers were labeled with the BF-B6 monoclonal antibody, indicating th a t no embryonic or neonatal MyHCs were present in the muscle biopsies examined. A more detailed description of these five immunohistochemical fiber types has been shown in the previous study (River0 et al., 1996). Combined Histochemical and lmmunohistochemical Analyses Combined histochemical and immunohistochemical analyses demonstrated a certain degree of correlation, albeit not unequivocal, between mATPase staining intensities (pH 4.2, 4.5, and 8.75) and MyHC content. Type I fibers were histochemically uniform and reacted with the slow, N2.261, BF-35, and BF-G6 MyHC mAbs (e.g., fibers labeled 1 in Fig. 1). However, the fasttwitch fiber population was histochemically and immunohistochemically heterogeneous. Most fibers histochemically classified a s IIA contained only IIa MyHC (e.g., fiber labeled 3 in Fig. 1). However, about 2% of the IIA fibers were composed of both fast IIa and IIb MyHCs (and thus named IIAB immunohistochemically; not shown in Fig. 1). Another 1%of the IIA mATPase fibers contained only type IIb MyHC (IIB fibers immunohistochemically; not shown). Similarly, a high percentage of fibers histochemically classified as IIB contained a mixture of IIa and IIb MyHCs (-24% of IIB mATPase fibers) and were classified as type IIAB by immunohistochemistry (e.g., fiber labeled 5 in Fig. 1). A small amount of fibers histochemically identified as IIB contained only the IIa MyHC (-2%; fibers not shown). Those fibers histochemically identified as type IIAB frequently contained both IIa and 1% MyHCs, but a high proportion were composed exclusively of IIa MyHC (-61%), so they were immunohistochemically classified as IIA (e.g., fiber labeled 4 in Fig. 1).A few fibers histochemically classified as type IIAB contained only IIb MyHC (-4%; not shown). Finally, there was no optimal correlation between mATPase activity and MyHC content for the C-fiber population. Some fibers stained more like IIC fibers by histochemistry contained only the slow-MyHC (not shown); conversely, a few fibers coexpressing both the slow and the fast IIa MyHCs were typed a s I for mATPase (e.g., fiber labeled 2 in Fig. 1). Overall, 18.5% of the muscle fibers (N = 2,078) were classified differently by using qualitative mATPase activity and the MyHC content as shown immunohistochemically. Effects of Training No significant changes in either body weight or performance was observed over the training period. V, values were similar a t all three times in which muscle samples were collected: pretraining (7.33 * 0.35 d s e c , n = 5 horses), middle of training (7.76 * 0.57 d s e c ) , and posttraining (7.76 2 0.39 d s e c , P > 0.05). MYOSIN ATPase AND MYOSIN HEAVY CHAIN IN HORSES WITH TRAINING 199 Fig. 1. Identification of fiber types by enzyme histochemistry and immunohistochemistry. Serial sections of horse gluteus medius muscle were processed for the demonstration of myosin ATPase and SUCcinate dehydrogenase (SDH) activities and stained with monoclonal antibodies against MyHC isoforms. A. Scheme showing the pattern of reactivity of the six most common fiber types identified by enzyme histochemistry and immunohistochemistry. E D : Myosin ATPase after preincubations at pH 8.75 (B), pH 4.2 (C), and pH 4.5 (D). E: Qualitative SDH activity. F-L Immunohistochemical staining with monoclonal antibodies fast (F), slow ( G ) ,SC-71 (HI, N2.261 (I), BF-35 (J),BF-G6 (K), and BF-B6 (L). The fibers labeled 1-6 correspond to those fibers illustrated in A. Bar = 100 Fm. Almost no significant changes (P > 0.05) in fiber type composition, mean CSA, relative CSA, or SDH activity of muscle fiber types were recorded as a consequence of training (Table 1).A decrease in the percentage of histochemically identified type IIAB fibers during the second half of the training program was the only significant (P < 0.001) change recorded after training. Nevertheless, a tendency for both hypertrophy and an increase of the oxidative capacity of all fiber types was suggested (Table 1).No changes in percentages, CSA, relative CSA, and SDH activity were observed for any of the four histochemical and immunohistochemical fiber types in the control horse (Table 1). Training had no significant effect on the number of muscle fibers misclassified by qualitative mATPase activity in correspondence with their MyHC content as revealed by immunohistochemistry (Table 2). 200 J.-L.L. RIVER0 ET AL. TABLE 1. Percentage, cross-sectional area (CSA), relative CSA, and SDH activity of the various fiber types identifed immunohistochemically and by qualitative myosin ATPase of five horses (training) and one control horse (control) dona the experimental Deriod' Month Training (N = 5 ) Fiber composition (%) I IIA IIAB IIB CSA (pm') I IIA IIAB IIB Relative area (%) I IIA IIAB IIB SDH activity (OD) I IIA IIAB IIB Control (N = 1) Fiber composition (%) I IIA IIAB IIB CSA (pm') I IIA IIAB IIB Relative area (%) I IIA IIAB IIB SDH activity (OD) I 0 Immunohistochemistry 4 0 Histochemistry 4 8 13.2 (4.2) 42 (7) 9 (3.8) 35.8 (4.8) 8.6 (6.1) 43.2 (7.2) 13.4 (4.2) 34.8 (5.4) 13.8 (11.4) 40.8 (13.7) 11.8 (3.7) 33.6 (11.5) 13.2 (4.2) 34.4 (8.5) 10.2 (2.2) 42.2 (6.4) 8.6 (6.1) 34.6 (5.3) 15 (4.2) 41.8 (4) 13.8 (11.4) 34.2 (10.2) 6.2 (1.5)* 45.8 (8.2) 1,349 (268) 1,995 (223) 2,605 (607) 3,676 (1,059) 1,922 (615) 2,180 (719) 2,765 (922) 4,386 (1,176) 1,848 (354) 2,339 (440) 2,953 (450) 3,954 (758) 1,349 (268) 1,918 (292) 2,069 (226) 3,575 (988) 1,922 (615) 2,003 (608) 3,309 (1,197) 4,115 (1,131) 1848 (354) 2232 (547) 2719 (230) 3817 (651) 7 (3.1) 33 (4.2) 9.6 (4.3) 50.4 (7.1) 5.2 (4) 31.4 (6.5) 12.6 (4.2) 50.8 (6.3) 9 (7.1) 33.4 (13.3) 12.2 (3.8) 45.4 (13.5) 7 (3.1) 26.2 (6.6) 8.6 (0.9) 58.2 (7.2) 5.2 (4) 22.8 (6.8) 16.2 (6.9) 55.8 (2.7) 9 (7.1) 25.6 (7.8) 6 (1.6) 59.4 (8.2) 1.23 (0.35) 1.06 (0.27) 0.92 (0.24) 0.69 (0.22) 1.31 (0.1) 1.17 (0.07) 1.05 (0.1) 0.72 (0.08) 1.32 (0.08) 1.15 (0.06) 1.03 (0.07) 0.81 (0.06) 1.23 (0.35) 1.08 (0.28) 0.98 (0.26) 0.73 (0.23) 1.31 (0.1) 1.19 (0.06) 1.08 (0.08) 0.77 (0.09) 1.32 (0.08) 1.18 (0.06) 1.07 (0.05) 0.85 (0.05) 8 12 38 15 35 7 42 16 35 5 35 27 33 12 32 8 48 7 34 12 47 5 27 23 45 990 1,960 2,158 3,614 923 1,707 2,247 2,750 846 846 1,631 2,547 990 1,918 2,335 3,190 923 2,669 2,719 2,709 846 1,218 1,608 2,340 5 30 13 52 3 34 17 46 2 19 27 52 5 25 7 63 3 27 10 60 2 18 21 59 1.15 0.99 0.83 0.64 1.16 1.08 0.96 0.67 1.17 1.17 0.89 0.62 1.15 1 0.95 0.68 1.16 1.1 1.03 0.72 1.17 1.03 0.92 0.69 'Values are mean (SD). N = number of horses. 'Significantly different fmm second muscle biopsy (P < 0.001).No significant (P > 0.05) variations along the experimental period were found for the other muscle parameters. Quantitative Histochemical and Immunohistochemical Interrelationshipsof Fast-Twitch fibers showed a histochemical reaction between IIA and IIB (type IIAB) fibers was similar in average to those fibers The mean percentage of type I and type I1 fibers cal- identified immunohistochemically as fibers containing culated on serial sections stained histochemically and both IIa and IIb MyHCs. Statistical comparisons of the immunohistochemically were identical. However, im- fiber composition, mean CSA, relative CSA, and SDH portant variations existed between the various sub- activity of the histochemically and immunohistochemtypes of fast-twitch fibers. A comparison of quantita- ically identified fast-twitch fibers for all muscle bioptive immunohistochemical vs. histochemical data of sies revealed significant correlations and relatively low the three fast-twitch muscle fibers is summarized in covariation coefficients between duplicate determinaTable 3. As shown, the number of pure IIA fibers im- tions for IIA and IIB fibers. However, correlation coefmunohistochemically identified as larger than ex- ficients were not significant (P > 0.05) for the hybrid pected based on the histochemical analysis (P < 0.001). IIAB fiber population. Similarly, coefficients of variaIn contrast, the proportion of pure IIB fibers was lower tion between immunohistochemical and histochemical by histochemical analysis in comparison with the information for these fibers were dramatically higher. MyHC content (P < 0.001). The number of fibers th a t Measurements of fiber size and SDH activity of fast 201 MYOSIN ATPase AND MYOSIN HEAVY CHAIN I N HORSES WITH TRAINING TABLE 2. Number and percent of muscle fibers misclassified based on qualitative mATPase histochemistry and MyHC immunohistochemistry in all 18 muscle bioasies Month Histochemistry Immunohistochemistry N IIA IIB IIAB IIAB IIB IIA IIC N = iiAB IIAB 0 4 684 698 8 IIB 3 49 46 3 IIA 10 IIB I 2 3 1 3 0 IIA Percent 8 696 Total 5 16 173 160 45 42 3 4 11 3 17 0 0 3 4 1 3 1 59 65 65 5 49 2,078 4 number of fibers examined. fibers were not significantly different between the two methods, but the mean CSA was lower for each of the three fast fibers histochemically identified. Electrophoretic Data Electrophoresis of MyHCs separated three bands in all the 18 muscle biopsies examined from the training group and the control horse (Fig. 2). These three bands were in the order of mobility I > IIb > IIa. Training did not cause significant changes (P > 0.05) in the relative proportion of the three MyHC isoforms of the gluteus medius muscle (Table 4). However, the percentage of type I MyHC showed a clear tendency to increase and that of type IIB MyHC to decrease in the biopsy samples after 8 months of training (P < 0.1 in both). No significant changes in the MyHC composition of the gluteus medius muscle were observed after 4 and 8 months in the control horse (Table 4). Electrophoretic, Immunohistochemical, and Histochemical Interrelation To compare the MyHC content with the fiber type composition properly, the relative CSA of the three major fiber types (I, IIA, and IIB) was determined (Table 4). The mean percentage of type I fibers derived electrophoretically, immunohistochemically, and histochemically was similar (P > 0.05). Once again, the percentage of relative CSA occupied by IIA fibers immunohistochemically classified was higher than that based on the histochemical analysis (P< 0.001 in all the samples). Conversely, the relative CSA area occupied by IIB MyHC fibers was lower than that of IIB mATPase fibers (P< 0.01). The percentage of fast IIa and IIb MyHC isoforms based on gel electrophoresis was intermediate between those of the relative CSA occupied by these fibers identified by immunohistochemistry and those identified by qualitative mATPase, but this difference was not significant (P> 0.05; Table 4). The derived relative CSA of the immunohistochemically and histochemically major fiber types (see Materials and Methods) were again compared with the MyHC content measured in each biopsy sample by gel electrophoresis (Table 5). Comparisons over the experimental period for all the 18 muscle biopsies examined revealed low but significant (P < 0.05) correlations between the percentage fiber-type CSA and MyHC content measured from gels, the exception being the rela- tion between the relative CSA of histochemically identified IIA fibers and the IIa MyHC (r = 0.07, P > 0.05). When all fiber types and MyHC were combined, the correlation coefficients ranged from 0.94 to 0.96. Differences in Fiber Size and SDH Activity Among Fiber Types Figure 3 shows the proportional distributions of Z scores of the mean CSA and SDH activity for the four fiber types immunohistochemically identified from the 18 muscle biopsy specimens examined in the present study (N = 2,078 fibers). Unimodal distributions of CSA for the total populations of type I, IIA, and IIAB fibers were generally evident for most of the fiber types, but a generally wider distribution was observed for type IIB fibers. Despite the overlap in the distributions of CSA among the four fiber types, significant differences among fiber types were observed (P < 0.0001). Thus, the fiber size was in the order of IIB > IIAB > IIA I. Moreover, in addition to being larger, fast type IIB fibers also showed a significantly greater range of CSAs than the other fiber types (Fig. 3A). Similarly, distributions of relative SDH activities were unimodal for the total fiber populations. Despite the overlap in the distribution of SDH activities among fiber types, significant variations were recorded (P < 0.0001) for the overall population among fiber types, with the type I fibers being more oxidative than fast fibers. Within the fast fibers, the order of oxidative capacity was type IIA > type IIAB > type IIB. In the total population of muscle fibers, there was a significant (P < 0.001) negative correlation (r = -0.53) between CSA and SDH activity. This negative correlation between CSA and SDH was evident within each population of fiber classifications except for type I: I, r = -0.08, P > 0.05; IIA, r = -0.25, P < 0.001; IIAB, r = -0.27, P < 0.001; and IIB, r = -0.35, P < 0.001. DISCUSSION The interrelations of histochemical, immunohistochemical, electrophoretic, and morphometric analyses of muscle biopsies from trained and untrained horses are described in the present study. As in humans (Staron, 1991; Staron and Hikida, 1992), there is a high level of predictability between mATPase activity and MyHC content among most, but not all, fibers in equine skeletal muscle. Comparisons between the histochemical and immunohistochemical data show that 202 J.-L.L. RIVER0 ET AL. TABLE 3. Mean values and correlation and variation coefficient (CV)between immunohistochemical and histochemical data of the various fast-twitch fiber types for training and control horses (N = 18 biopsies) Immunohistochemistry Fiber composition (%) IIA IIAB IIB CSA (pm2) IIA IIAB IIB Relative CSA (%) IIA IIAB IIB SDH activity (OD) IIA IIAB IIB Histochemistry 33.8 10.4 43.8 41.4 11.9 34.7 2,060 2,647 3,833 1,976 2,564 3,654 32.5 11.9 48.9 24.6 10.7 58.3 1.12 0.98 0.73 1.13 1.03 0.77 r CV 0.86 0.5" 0.73 26.2 50.5 29.6 0.95 0.56* 0.99 10.3 30.2 6.84 0.79 0.39* 0.69 24.6 64.5 23.2 0.96 0.93 0.99 4.31 8.33 7.23 *Correlation values not significant (P > 0.05). 1A B 4A B C 2A B C 3A B C 5 A B C 6A B C IIAIIBI - C Fig. 2. Eight-percent sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel of myosin heavy chain (MyHC) of the training group (horses 1-5) and the control horse (horse 6). A-C are the three muscle biopsies removed from each horse. The MyHC isoforms are identified as types I (I),IIA (Ira), and IIB (IIb). -18% of the fibers were misclassified by traditional mATPase histochemistry. Each MyHC isoform can be associated rather consistently with a specific mATPase-based muscle fiber type (Staron and Pette, 1986), except when two or more MyHCs coexist within the same fiber. A large proportion of the fast fibers in the equine gluteus medius muscle coexpress the two fast MyHC isoforms in varying ratios. Likewise, a few 203 MYOSIN ATPase AND MYOSIN HEAVY CHAIN I N HORSES WITH TRAINING TABLE 4. Percentage of MyHC based on gel electrophoresis and percentage of the relative cross-sectional area occupied by the three major fiber types identified by immunohistochemistryand qualitative myosin ATPase of five horses (training)and one control horse (control)' Month Training Electrophoresis (N) I IIA IIB Immunohistochemistry I IIA IIB Histochemistry I IIA IIB Control Electrophoresis (N) I IIA IIB Immunohistochemistry I IIA IIB Histochemistry I IIA IIB All horses Electrophoresis (N) I IIA IIB Immunohistochemistry I IIA IIB Histochemistry I IIA IIR All samples 0 4 8 5 6.2 (2.5) 33.8 (4.6) 60.0 (5.9) 5 5.4 (2.0) 35.4 (2.1) 59.0 (2.7) 5 11.8 (6.1) 34.6 (2.9) 53.4 (5.1) 15 7.8 (4.9) 34.6 (3.5) 57.6 (5.6) 7.0 (2.8) 37.8 (3.9) 55.2 (5.9) 5.2 (3.6) 37.7 (5.9) 57.1 (4.5) 9.0 (6.3) 39.5 (12.4) 51.5 (11.0) 7.1 (4.7) 38.3 (3.5) 54.6 (7.8) 7.0 (2.8) 30.5 (5.8) 62.5 (6.5) 5.2 (3.6) 30.9 (4.3) 63.9 (3.3) 9.0 (6.3) 28.6 (7.2) 62.4 (7.2) 7.1 (4.7) 30.0 (6.0)** 62.9 (6.0)** 1 8.0 30.0 62.0 1 4.0 27.0 69.0 1 6.0 28 66.0 3 6.0 (2.0) 28.0 (1.4) 65.7 (2.9) 5.0 36.5 58.5 3.0 42.5 52.5 2 32.5 65.5 3.3 (1.2) 38.2 (4.1) 58.5 (5.3) 5.0 28.5 66.5 3.0 32.0 65.0 2.0 28.5 67.5 3.3 (1.2) 30.4 (1.6)* 66.3 (1.0)* 6 6.5 (2.3) 33.2 (4.4) 60.3 (5.4) 6 5.2 (1.9) 34.0 (3.7) 60.8 (4.5) 6 10.8 (6.0) 33.3 (3.9) 55.9 (6.6) 18 7.5 (4.6) 33.5 (4.0) 59.0 (6.0) 6.7 (2.6) 37.6 (3.6) 55.7 (5.5) 4.8 (3.4) 38.5 (5.7) 56.7 (4.4) 7.8 (6.3) 38.3 (11.6) 53.9 (11.3) 6.5 (4.6) 38.2 (7.7) 55.3 (7.8) 6.7 (2.6) 30.2 (5.5) 63.1 (6.2) 4.8 (3.4) 31.1 (4.0) 64.1 (3.1) 7.8 (6.34) 28.6 (6.5) 63.6 (6.9) 6.5 (4.6) 30.0 (5.5)*** 63.5 (5.6)** 'Values are mean (SD); N = number of horses or samples. Training had no significant effect (P > 0.05) on muscle characteristics analyzed. *P < 0.05, **P< 0.01, and ***P < 0.001 compared with the same fiber type immunohistochemically classified. TABLE 5. Correlation analysis between percentages of fiber type area and myosin heavy chain for training and control horses (N = 18 biopsies)' Immunohistochemistry vs. histochemistry Immunohistochemistry vs. electrophoresis Histochemistry vs. electrophoresis I 1.00 0.57 0.57 IIA 0.82 0.51 0.07* IIB 0.80 0.48 0.43 All 0.95 0.94 0.96 'All values given are correlation coefficients. *Only correlation value not significant (P> 0.05). fibers (designated IIC) coexpress the I and IIa MyHC. Fibers coexpressing two or more MyHC isoforms may react histochemically according to the dominant isoform (Danieli-Betto et al., 1986; Klitgaard et al., 199Oa). The finding that many fibers histochemically classified as type IIB coexpressed IIa and IIb MyHCs supports this assumption. Furthermore, some fibers histochemically identified as type I coexpressed type I and type IIa MyHCs based on immunohistochemistry . However, nearly all fibers histochemically classified as type IIAB contained only IIa MyHC. The observation that a large number of histochemically determined IIB fibers actually contain a varying amount of type IIa MyHC is in agreement with previous studies on human single fibers (Staron, 1991; Staron and Hikida, 1992; Andersen et al., 1994a,b). In the two latter studies, the researchers reported that nearly all histochemical type IIB fibers of sprinters J.-L.L. RIVER0 ET AL. 204 B. SDH Activity A. Cross-Sectional Type I fibers (N=235) . . . . 35 I Type ZIA fibers ( N = 8 6 1 ) ""T & 35T- ~ Type IIAB fibers ( N = 2 6 2 ) . . . . 0 1 Type ZIB fibers (N=721) 30 T 35 1 Z Score Fig. 3. Histogram of frequencies showing population distributions of cross-sectional area (A) and SDH activities (B)(standardized to 2 scores) for fibers containing types I (I), IIa (IIA), IIa + IIb (IIAB), and only IIb MyHCs when identified by immunohistochemistry. All fibers examined (N = 2,078) are computed; all fiber types were statistically different (P < 0.0001). MYOSIN ATPase AND MYOSIN HEAVY CHAIN IN HORSES WITH TRAINING (Andersen et al., 1994a) and soccer players (Andersen et al., 199413) coexpressed both IIa and IIb MyHC isoforms. The explanation for the finding that practically all histochemically typed IIAB fibers contained only IIa MyHC is unclear. The mATPase histochemical staining may be affected by factors other than the relative content of MyHC. For example, the myosin light chain isoform composition of a fiber can, independently of the MyHC, affect the maximal velocity of shortening of that fiber (Bottinelli et al., 1994b).Thus, the myosin light chain may also influence the myofibrillar ATPase staining characteristics of a fiber because ATPase activity and shortening velocity are highly correlated (Barany, 1967). However, in a recent study (Sant’Ana Pereira et al., 1995), the combined quantitative histochemical, immunohistochemical, electrophoretic, and immunoblotting analyses of single human skeletal muscle fibers reported that the mATPase staining patterns usually correlated with the expression of distinct MyHC isoforms. A continuum exists from pure IIA to pure IIB in the staining density for mATPase after preincubation at pH 4.5-4.6 (White and Snow, 1985; Rivero et al., 1993). The results of a previous study of the mATPase activity of equine muscle fiber types by quantitative histochemistry (White and Snow, 1985) clearly showed a continuous range within the type I1 fibers, with two or three overlapping peaks. Because this continuum of fiber types is difficult to quantify by using a qualitative technique such as mATPase histochemistry , this method of classification of muscle fibers into four major types (five if the type IIC is included) may have major limitations in objectively differentiating among the three fast subtypes (IIA, IIAB, and IIB) based on MyHC content. An alternative explanation is that histochemical type IIAB fibers might contain a third distinct fast MyHC isoform, which was not revealed by the monoclonal antibodies tested in the present study. This possibility was not suggested, however, by the monoclonal antibodies used or by the SDS-PAGE assayed in the present study. Our observation of some fibers histochemically identified as type I, but coexpressing both I and IIa MyHC, is similar to that reported for human muscle (Klitgaard et al., 1990a). Klitgaard et al. (1990a) reported that in the 95%of fibers in endurance athletes that contained both I and IIa MyHCs the major fraction was the type I MyHC, and this perhaps explains the histochemical staining profile as type I fibers. Another striking observation of the present study was that training had no significant effect on the correlation between mATPase activity and MyHC content of muscle fibers. In some studies, it has been argued that the “misclassification” of many fibers by mATPase histochemistry seems to be especially pronounced in trained individuals because of the high number of fibers coexpressing multiple MyHCs (Klitgaard et al., 1990a; Staron, 1991). Although this assumption would be supported by the larger number of fibers coexpressing types I and IIa MyHCs in endurance athletes than in sedentary men reported by Klitgaard et al. (1990a), it is not confirmed by either the present or previous results in humans (Klitgaard et al., 1990b;Andersen et al., 1994a,b). Our results suggest that the coexpression of the two fast MyHCs within the same fiber is not 205 necessarily a sign of fiber type transformation because these fibers did not increase in number with training. The assumption is also supported by the results of two previous studies in humans, where the percentage of fibers with coexisting IIa and IIb MyHCs decreased as a result of training (Klitgaard et al., 1990b; Andersen et al., 1994a). The correlations between the relative CSA of the three major fiber types, as determined by immunohistochemistry and histochemistry, and the percentage MyHC content derived by SDS-PAGE coincide with the results reported by Fry et al. (1994) in humans. These low correlations do not give credence to the use of mATPase histochemistry for assessing fiber type composition, as was concluded in that study. When the five immunohistochemically and histochemically delineated equine fiber types were compressed into three major types (I,IIA, and IIB), the percentage of the CSA occupied by type IIA fibers histochemically determined was underestimated and that of type IIB fibers was overestimated in comparison with the relative CSA occupied by the same fibers classified by immunohistochemistry. The stronger correlations found between immunohistochemical vs. electrophoretic data when compared with histochemistry vs. electrophoresis (Table 5) give more validity to the use of monoclonal antibodies against selected MyHCs as a means of fiber typing in horses. Surprisingly, no significant changes in muscle fiber type characteristics were observed in the current study after 4 and 8 months of training. In horses as in humans and other mammals, the most common adaptation to training is an increase of the type IIA to type IIB histochemical fiber ratio and an increase of both the oxidative capacity and capillary supply of muscle fibers (for review, see Snow and Valberg, 1994). In the current study, a higher, albeit not significant, SDH activity of all fiber types as a result of training was also recorded. The data of the present study for the mean CSA were similar in range to those reported for the gluteus medius muscle in young thoroughbred racehorses (Lindholm et al., 1983; Serrano et al., 1996).The effects of training on muscle fiber size in horses have been inconsistent but may depend on the nature, intensity, and duration of the exercise involved (Snow and Valberg, 1994). An important observation of the present study is that even after 8 months of intensive training no significant modifications in fiber size occurred. In other studies on horses, little change or a general or specific type increase was found in fiber size after different training programs (see Snow and Valberg, 1994). As the intensity and duration (both daily and in total) of exercise involved in the current study was greater and longer than in other previous studies in horses (Snow and Valberg, 19941, the limited training-related changes in muscle fiber type characteristics recorded in the present study is difficult to explain. However, the limited changes may be related to the relative superficial sampling site of the gluteus medius muscle from which the samples were taken. In a recent study (Rivero et al., 19951, the adaptation of the gluteus medius muscle fibers to training was not homogeneous in superficial and deep sampling regions of this muscle. Because these training-linked modifications were more marked in the deep region of the muscle 206 J-L.L. RIV‘ERO ET AL. than in the superficial one, different functional demands must be placed on different depths of the gluteus medius muscle. This lack of muscular response to training may also be explained in part by the fact that no real metabolic adaptation occurred at the systemic level in response t o training. For example, the performance capacity did not improve in any of the horses throughout the training program, as shown by the constant values of ITLAI. The most common approach to classifying horse skeletal muscle fibers has been dividing them into three main fiber types (I, IIA, and IIB) based on myofibrillar ATPase histochemistry (Snow and Valberg, 1994). Type I and IIA fibers have a high oxidative capacity, whereas type IIB fibers have a heterogeneous oxidative potential (high or low). Histochemical analysis of these three fiber types also reveals important differences in sizes, the order generally being I < IIA < IIB. The current study confirms these concepts. However, our results, based on the MyHC content of muscle fibers (immunohistochemistry), support the separation of equine skeletal myofibers into four major types (I, IIA, IIAB, and IIB) because the hybrid IIAB fibers appear to be a biologically important type in horses (- 10-15%). In conclusion, histochemical, immunohistochemical, and electrophoretic techniques applied to equine skeletal muscle establish a relation between qualitative myofibrillar mATPase activity and MyHC content. However, a high percentage of fibers of the equine gluteus medius muscle typed according to the traditional mATPase histochemistry does not always correspond to a classification based on MyHC isoforms. Because the immunohistochemical techniques used are specific to MyHC isoforms, these results suggest that the qualitative mATPase histochemical technique is an indication of not only the MyHC isoform but also other characteristics of the myofibrillar protein. The use of monoclonal antibodies against specific MyHC isoforms seems to be a more sensitive, more reproducible, and less subjective method for categorizing muscle fibers. Finally, because a significant metabolic adaptation to training did not occur in our horses, the results from the present study are inconclusive to suggest that the misclassification of a large number of fibers by mATPase histochemistry may be magnified in trained individuals. ACKNOWLEDGMENTS This study was completed while Jos6-Luis L. Rivero was working at the Department of Physiological Sciences, University of California at Los Angeles; his work was supported by grants and scholarships from the Spanish D.G.C.Y.T. (PR94-202) and the University of Cordoba, Spain. We thank Dr. S. Schiaffino (University of Padova, Italy) for the generous gift of MAbs. The MAb developed by Dr. H. Blau (N2.261) was obtained from the Developmental Studies Hybridoma Bank, maintained by the Department of Pharmacology and Molecular Sciences, Johns Hopkins University School of Medicine, Baltimore, MD, and the Department of Biology, University of Iowa, Iowa City, IA, under contract N01-HD-2-3144 from the NICHD. LITERATURE CITED Aigner, S., B. Gohlsch, N. HBmalainen, R.S. Staron, A. Uber, U. Wehrle, and D. Pette 1993 Fast myosin heavy chain diversity in skeletal muscles of the rabbit: Heavy chain IId, not IIb, predominates. Eur. J. Biochem., 211,367-372. Andersen, J.L., H. Klitgaard, and B. Saltin 1994aMyosin heavy chain isoforms in single fibres from m. vastus lateralis of sprinters: Influence of training. Acta Physiol. Scand., 151:135-142. Andersen, J.L., H. Klitgaard, J. Bangsbo, and B. Saltin 1994b Myosin heavy chain isoforms in single fibres from m. vastus lateralis of soccer players: Effects of strength-training. Acta Physiol. Scand., 150:2 1-26. Bar, A,, and D. Pette 1988 Three fast myosin heavy chein in adult rat skeletal muscle. FEBS Lett., 235:153-155. Barany, M. 1967 ATPase activity of myosin correlated with speed of muscle shortening. J . Gen. Physiol., 50t197-216. Billeter. R.. C.W. Heizmann. H. Howard. and E. Jennv 1981 Analvsis of myosin light and heavy chain t&es in single human skelktal muscle fibres. Eur. J. Biochem., 116:389-395. Biral, D., R. Betto, D.D. Betto, and G. Salviati 1988 Myosin heavy chain composition of single fibers from normal human muscle. Biochem. J.,250:307-308. Blanco, C.E., G.C. Sieek, and V.R. Edgerton 1988 Quantitative histochemical determination of succinic dehydrogenase activity in skeletal muscle fibres. Histochem. J., 20.230-243. Bottinelli, R., R. Betto, S. SchiafTino, and C. Reggiani 1994a Maximum shortening velocity and coexistence of myosin heavy chain isoforms in single skinned fast fibres of rat skeletal muscle. J. Muscle Res. Cell Motil., 15:413-419. Bottinelli, R., R. Betto, S. Schiaffino, and C. Reggiani 1994b Unloaded shortening velocity and myosin heavy chain and alkali light chain isoform composition in rat skeletal muscle fibres. J . Physiol., 478:341-349. Brooke, M.M., and K.K. Kaiser 1970 Muscle fiber types: . _ How many and what kind? Arch. Neurol., 23r369-379. Danieli-Betto, D., E. Zerbato, and R. Betto 1986 Type 1, 2A and 2B myosin heavy chain electrophoretic analysis of rat muscle fibers. Biochem. Biophys. Res. Comm., I38:981-987. Ennion, S., J. Sant’Ana Pereira, A.J. Sargeant, A. Young, and G. Goldspink 1995 Characterization of human skeletal muscle fibres according to the myosin heavy chains they express. J. Muscle Res. Cell Motil., 16:35-43. Fry, A.C., C.A. Allemeier, and R.S. Staron 1994 Correlation between percentage fiber type area and myosin heavy chain content in human skeletal muscle. Eur. J. Appl. Physiol., 68:246-251. Gorza, L. 1990 Identification of a novel type 2 fiber population in mammalian skeletal muscle bv combined use of histochemical myosin ATPase and anti-myosin monoclonal antibodies. J. Histochem. Cytochem., 38:257-265. Ingjer, F. 1979 Effects of endurance training on muscle fibre ATPase activity, capillary supply and mitochondria1 content in man. J. Physiol. (Lond.) 2941419-432. Klitgaard, H., 0. Bergman, R. Betto, G. Salviati, S. Schiafho, R. Clausen, and B. Saltin 1990a Co-existence of myosin heavy chain I and IIa isoforms in human skeletal muscle fibres with endurance training. Pliigers Archiv., 41 6:470-472. Klitgaard, H., M. Zhou, and E.A. Richter 1990b Myosin heavy chain composition of single fibres from m. biceps brachii of male body builders. Acta Physiol. Scand., 140:175-180. Lindner, A., P. von Wittke, M. Schmald, J. Kusserov, and H. Sommer 1992 Maximal lactate concentrations in horses after exercise of different duration and intensity. J. Equine Vet. Sci., 12t30-33. Lindholm, A., and K. Piehl 1974 Fibre composition, enzyme activity and concentration of metabolites and electrolytes in muscle of standardbred horses. Acta Vet. Scand., 155’87-309. Lindholm, A,, B. EssBn-Gustavsson, D. McMiken, S. Persson, and J.R. Thornton 1983 Muscle histochemistry and biochemistry of thoroughbred horses during growth and training. In: Equine Exercise Physiology. D.H. Snow, S.G.B. Persson, and R.J. Rose, eds. Granta Editions, Cambridge, UK, pp. 211-217. Martin, T.P., A.C. Vailas, J.B. Durivage, V.R. Edgerton, and K.R. Castleman 1985 Quantitative histochemical determination of muscle enzymes: Biochemical verification. J. Histochem. Cytochem., 33t1053-1059. Nwoye, L., W.F.H.M. Mommaerts, D.R. Simpson, K. Serayderian, and M. Marusich 1982 Evidence for a direct action of thyroid hormone in specifying muscle properties. Am. J. Physiol., 242:R401-R408. Pette, C., and S. Staron 1990 Cellular and molecular diversities of mammalian skeletal muscle fibers. Rev. Physiol. Biochem. Pharmacol., 116:1-76. MYOSIN ATPase AND MYOSIN HEAVY CHAIN IN HORSES WITH TRAINING Pierobon-Bormioli,S.,S. Sartore, L. Dalla Libera, M. Vitadello, and S. Schiafho 1981 Fast isomyosins and fiber types in mammalian skeletal muscle. J. Histochem. Cytochem., 29t1179-1188. Rivero, J.-L.L., A.L. Serrano, P. Henckel, and E. Agiiera 1993 Muscle fiber type composition and fiber size in successfully and unsuccessfully endurance-raced horses. J. Appl. Physiol., 75:17581766. Rivero, J.-L.L., M.C. Ruz, A.L. Serrano, and A.M. Diz 1995 Effects of a 3-month endurance training programme on skeletal muscle histochemistry in Andalusian, Arabian and Anglo-Arabian horses. Equine Vet., 27.51-59. Rivero, J.-L.L., R.J. Talmadge, and V.R. Edgerton 1996 Myosin heavy chain isoforms in adult equine skeletal muscle: an immunohistochemical and electrophoretic study. Anat. Rec., 246r185-194. Reiser, P.J., R.L. Moss, G.G. Giulian, and M.L. Greaser 1985 Shortening velocity in single fibres from adult rabbit soleus muscle is correlated with myosin heavy chain composition. J . Biol. Chem., 260r9077-9080. Sant'Ana Pereira, J.,A. Wessels, L. Nijtmans, A.F.M. Moorman, and A.J. Sargeant 1995 New method for the accurate characterization of single human skeletal muscle fibres demonstrates a relation between mATPase and MyHC expression in pure and hybrid fibre types. J. Muscle Res. Cell Motil., 16r21-34. Schiaffino, S., L. Gorza, S. Sartore, L. Saggin, S. Ausoni, M. Vianello, K. Gundersen, and T. Lemo 1989 Three myosin heavy chain isoforms in type 2 skeletal muscle fibers. J. Muscle Res. Cell. Motil., 1Ot197-205. Serrano, A.L., J.L. Petrie, J-L.L. Rivero, and J.W. Hermanson 1996 Myosin isoforms and muscle fiber characteristics in equine gluteus medius muscle. Anat. Rec., 244:444-451. Sieck, G.C., R.D. Sacks, C.E. Blanco, and V.R. Edgerton 1986 SDH 207 activity and cross-sectional area of muscle fibers in cat diaphragm. J. Appl. Physiol., 60r1284-1292. Smerdu, V., I. Karsh-Mizrachi, M. Campione, L. Leinwand, and S. Schiaffino 1994 Type IIx myosin heavy chain transcripts are expressed in type IIb fibers of human skeletal muscle. Am. J. Physiol., 267:(Cell Physiol, 36)C1723-C1728. Snow, D.H., and Valberg, S.J. 1994 Muscle anatomy, physiology and adaptations to exercise and training. In: The Athletic Horse. D.R. Hodgson and R.J. Rose, eds. WB Saunders, Philadelphia, pp. 145-179. Staron, R.S. 1991 Correlation between myofibrillar ATPase activity and myosin heavy chain composition in single human muscle fibers. Histochemistry, 9621-24. Staron, R.S., and R.S. Hikida 1992 Histochemical, biochemical, and ultrastructural analyses of single human muscle fibers, with special reference to the C-fiber population. J . Histochem. Cytochem., 40563-568. Staron, R.S., and D. Pette 1986 Correlation between myofibrillar ATPASE activity and myosin heavy chain composition in rabbit muscle fibres. Histochemistry, 86.19-23. Sullivan, T.E., and R.B. Armstrong 1978 Rat locomotory muscle fiber activity during trotting and galloping. J . Appl. Physiol., 44:358363. Talmadge, R.J., and R.R. Roy 1993 Electrophoretic separation of rat skeletal muscle myosin heavy-chain isoforms. J. Appl. Physiol., 75t2337-2340. White, M.G., and D.H. Snow 1985 Quantitative histochemistry of myosin ATPase activity after acid preincubation, and succinate dehydrogenase activity in equine skeletal muscle. Acta Histochem. Cytochem., 18:483-493.