Chiung-Ying Ho*, Marvin H. Stromer*,3, and Richard M. Robson*, Introduction

Transcription

1 Effect of Electrical Stimulation on Postmortem Titin, Nebulin, Desmin, and Troponin-T Degradation and Ultrastructural Changes in Bovine Longissimus Muscle 1,2 Chiung-Ying Ho*, Marvin H. Stromer*,3, and Richard M. Robson*, Muscle Biology Group, *Departments of Animal Science and Biochemistry and Biophysics, Iowa State University, Ames ABSTRACT: Electrical stimulation (ES) of bovine carcasses is usually done to increase tenderness and has been hypothesized to increase the activity of proteolytic enzymes that may degrade structural proteins in muscle cells and cause fractures and breaks in muscle fibers, thus enhancing meat tenderness. Our objective was to compare postmortem (PM) changes in the muscle proteins, titin, nebulin, a- actinin, desmin, and troponin-t and in myofibrillar structure in nonstimulated (NS) and ES bovine skeletal muscle. One side of eight beef carcasses was stimulated within 1 h of death, and the other side was the NS control. Myofibrils for SDS-PAGE and samples for transmission electron microscopy were prepared from the longissimus muscle at 0, 1, 3, 7, 14, and 28 d PM. In SDS-PAGE, titin migrated as three bands in both NS and ES 0-d samples. The slowest migrating band, T1 (intact titin), decreased slightly faster in ES samples from five animals. The fastest migrating band, T2 (degraded titin), increased in amount through 3 d and was still present at 28 d. A titin monoclonal antibody (mab) identified a large family of degradation products that migrated faster than myosin heavy chains and that was more heavily labeled in Western blots of ES samples than in NS samples. In SDS-PAGE of NS samples, intact nebulin disappeared by 3 d in three animals, by 7 d in four animals, and by 14 d in one animal, but in ES samples the nebulin band was absent by 3 d in three animals and by 7 d in five animals. SDS-PAGE showed that the amount of intact desmin decreased slightly sooner in two ES samples and was absent earlier in one ES sample than in the corresponding NS control samples. Blots labeled with a polyclonal antibody to desmin showed that a more heavily labeled 38-kDa desmin degradation product was present in ES than in NS samples. Postmortem degradation of a-actinin was not detected. Contraction node (CN) formation, stretching of conjoined sarcomeres adjacent to the nodes, increased frequency of I-band fractures and accelerated appearance of wide I-band fractures adjacent to the Z- line, and, in some animals, slightly accelerated degradation of titin, nebulin, and troponin-t were characteristics of ES muscle. Key Words: Titin, Nebulin, Desmin, Troponin-T, Electrical Stimulation, Electron Microscopy J. Anim. Sci : Introduction 1 We thank Gene Rouse for obtaining the animals, Mary Sue Mayes for assistance in preparing TEM samples, ISU Meat Lab personnel for slaughtering the animals, and Darrel E. Goll for generously providing polyclonal antibodies to bovine a-actinin. This research was supported in part by the Beef Industry Council of the National Live Stock and Meat Board and the USDA National Research Initiative Competitive Grants Program Award Journal paper no. J of the Iowa Agric. and Home Econ. Exp. Sta., Ames. Projects 3349, 2127, and To whom correspondence should be addressed. Received October 26, Accepted February 7, Early postmortem ( PM) electrical stimulation ( ES) is usually used to increase meat tenderness and promotes the activity of some endogenous proteolytic enzymes, including m-calpain (Dransfield et al., 1992; Smulders and van Laack, 1992; Uytterhaegen et al., 1992). In myofibrils, titin is a very long filamentous protein that extends from the Z-line to the M-line and links thick filaments to the Z-line; nebulin comprises inextensible filaments that are closely associated with, or part of, thin filaments; a-actinin is an integral Z- line protein; and desmin is the major component of intermediate filaments that associate with Z-lines (for reviews, see Robson et al., 1981, 1991; Robson and Huiatt, 1983). Degradation of some of these structural 1563

2 1564 HO ET AL. proteins has been implicated in the loss of myofibrillar integrity, thus, increasing meat tenderness (Huff- Lonergan et al., 1995; Taylor et al., 1995). Robson et al. (1980) and Koohmaraie et al. (1986), for example, suggested that PM meat tenderness may be attributed, in part, to desmin degradation. Uytterhaegen et al. (1992) reported that aging and(or) ES increased the degradation of titin and troponin-t but that ES had no effect on sarcomere length. We have shown by SDS-PAGE that troponin-t decreased more rapidly PM in ES than in nonstimulated ( NS) bovine muscle and by Western blots that the accumulation of the 30-kDa degradation product of troponin-t increased more rapidly in ES than in NS samples (Ho et al., 1994). Electron microscopy studies have shown that ES caused the formation of contraction nodes ( CN), and that sarcomeres in the internodal zones were stretched or fractured in bovine muscle (Takahashi et al., 1987). Sarcomere shortening and both Z- line and M-line degradation were characteristics of PM bovine skeletal muscle (Stromer et al., 1967, 1974). Changes in proteins that compose cytoskeletal filaments in muscle, or anchoring structures for these filaments, therefore may be very important in determining the properties of PM muscle. The objectives of this study were 1) to investigate the extent of degradation of titin, nebulin, a-actinin, desmin, and troponin-t during PM storage of NS and ES bovine skeletal muscle by using SDS-PAGE and Western blots and 2) to compare the degradation of these proteins with structural changes in NS and ES samples by using transmission electron microscopy ( TEM). Sample Preparation Materials and Methods We used one line of crossbred steers (Angus Jersey), which is different from most previous studies that have used typically heterogeneous feedlot cattle. Previous studies at Iowa State University (Reiling et al., 1991) have shown that cattle from this line are genetically predisposed to produce tender muscle. The left side of eight market-weight Angus Jersey steers was electrically stimulated with 200 V, 20 Hz for 15 s for three stimulation periods with a 30-s interval between stimuli. The right side of each carcass was a NS control. The electrical stimulation and sampling procedures were previously described for these eight animals in Ho et al. (1994). Myofibrils were prepared from the longissimus muscle ( LM) immediately after stimulation (0 d) and at 1, 3, 7, 14, and 28 d PM by using the method of Fritz et al. (1989). Muscle samples for TEM also were taken from the LM immediately after stimulation (0 d) and at 1, 3, 7, 14, and 28 d PM. Thin strips of muscle were cut parallel to the longitudinal fiber axis and were isometrically clamped before the ends of the strips were severed. Tissue samples were doubly fixed in Karnovsky s fixative (Karnovsky, 1965) at 2 C for 3 h and then in 1% osmium tetroxide for 1.5 h. Dehydration in graded acetones was followed by embedding in Epon-Araldite resin. Two blocks were selected at random from each sample for sectioning. Thin sections were mounted on 300-mesh grids and were positively stained with methanolic uranyl acetate and Reynold s lead citrate (Stromer and Bendayan, 1988). All sections were examined at 80 kv in a JEOL JEM-100CXII electron microscope. Images were recorded on Kodak SO-163 electron image film. The frequency of I-band fractures was determined by counting, in alternate rows of grid openings, both the total number of myofibrils and the total number of each fracture type in sections that covered 15 to 20 grid openings per grid. Two grids were counted per sample treatment and time for each animal. Frequencies reported are averages over all animals for each treatment and sampling time. Statistical analysis was done by using the GLM (General Linear Model) procedure of SAS (1993). Sarcomere lengths were measured on at least eight randomly selected negatives from each sample. Length of sarcomeres in CN and stretched sarcomeres adjacent to CN were not included in the average sarcomere length for each sample. Sarcomere lengths in these two regions were measured and reported separately. Preparation of Selected Myofibrillar Proteins Used as Markers/Standards Titin and nebulin were prepared from bovine LM and were purified by the methods of Wang (1982) and Wang and Wright (1988). Myosin, desmin, tropomyosin and troponin-t, -I, and -C standards were prepared as we previously described (Ho et al., 1994). SDS-PAGE Analysis Three gel systems were used to analyze both NS and ES muscle samples in order to resolve proteins that varied widely in molecular mass: 1) an 8% Trisglycine slab gel (8 cm high 10 cm wide 1.5 mm thick) (acrylamide:n, N -methylenebisacrylamide = 200:1, wt/wt) with a 2.57% stacking gel (acrylamide: N, N -methylenebisacrylamide = 5.7:1, wt/wt) was run at 8 ma for 6 h and was used to detect PM changes in titin and nebulin; 2) an 18% Tris-glycine slab gel (16 cm high 18 cm wide 1.5 mm thick) (acrylamide:n, N -methylenebisacrylamide = 100:1, wt/wt) with a 5% stacking gel (acrylamide:n, N -methylenebisacrylamide = 37:1, wt/wt) was run at 10 ma for 17 h and was used to monitor the titin and nebulin degradation products, which migrate faster than myosin heavy chain ( MHC); 3) a 15% Tris-glycine slab gel (16 cm high 18 cm wide 1.5 mm thick) with a 5% stacking gel (both with an acrylamide:n, N -methylenebisacrylamide = 37:1, wt/wt) was run at 10 ma for 17

3 PROTEIN DEGRADATION AND STRUCTURAL CHANGES 1565 Table 1. Days postmortem on which protein changes occur Animal and treatment T1 T kda Nebulin Desmin TN-T 130 kda 39 kda 30 kda TN-I 1-NS 7 a >28 e 3 a 3 a 14 c >28 e 7 b 1 b 3 b 14 d 1-ES 3 a >28 e 1 a 3 af 14 c 14 a 7 b 1 b 1 b 14 d 2-NS 3 a >28 e 3 a 3 a 14 c 28 a 3 b 1 b 0 b 14 d 2-ES 3 af >28 e 1 a 3 af 14 c 14 a 3 b 1 b 0 b 14 d 3-NS 14 a >28 e 14 a 7 a 28 c >28 e 14 b 0 b 1 b 14 d 3-ES 7 a >28 e 7 a 7 a 28 cf >28 ef 14 b 0 b 1 b 14 d 4-NS 7 a >28 e 7 a 7 a 14 c >28 a 7 b 3 b 1 b 14 d 4-ES 7 a >28 e 7 a 7 a 14 c >28 e 7 b 3 b 0 b 28 d 5-NS 7 a >28 e 7 a 7 a 28 c 28 a 7 b 0 b 1 b 28 d 5-ES 7 a >28 e 7 a 7 a 14 c 28 a 7 b 0 b 0 b 14 d 6-NS 7 a >28 e 3 a 7 a 28 c 14 a 7 b 1 b 3 b 14 d 6-ES 3 a >28 e 1 a 7 af 28 cf 14 af 7 b 1 b 3 b 14 d 7-NS 14 a >28 e 7 a 14 a 28 c 28 a 14 b 1 b 0 b 28 d 7-ES 14 af >28 e 3 a 7 a 28 c 28 a 14 b 1 b 0 b 28 d 8-NS 3 a >28 e 3 a 3 a 14 c 28 a 3 b 1 b 3 b 14 d 8-ES 3 a >28 e 1 a 3 af 14 c 14 a 3 b 1 b 3 b 14 d a b Numbers in a column are the first sampling day when a protein was absent by SDS-PAGE. Numbers in a column are the first sampling day when a polypeptide appeared by SDS-PAGE. Numbers in this column are the first sampling day when desmin was absent on Western blots. Numbers in this column are the first sampling day when TN-I migrated further into the gel. Protein band still present at 28 d PM by SDS-PAGE. The protein band decreased slightly sooner (became lighter) but ultimate disappearance was on same day. h and was used to detect desmin and a-actinin degradation (Laemmli, 1970; Fritz et al., 1989). Western Blot Analysis Proteins were transferred from the slab gel to a polyvinylidene fluoride ( PVDF) membrane (Immobilon-P, Millipore, Bedford, MA) by a wet transfer method (Towbin et al., 1979; Ho et al., 1994). After transfer, the membrane was incubated in a 5% (wt/ vol) bovine serum albumin-phosphate buffer solution ( BSA-PBS) for 30 min at 37 C and then washed three times in a.1% (wt/vol) BSA-PBS solution for 5 min at 25 C. A monoclonal antibody ( mab) to titin, 9D10 (Greaser antibody, Developmental Studies Hybridoma Bank) and monospecific polyclonal antibodies ( pab) to bovine a-actinin and bovine skeletal muscle desmin were used as primary antibodies. Each membrane was incubated with a primary antibody for 2hat25 Cand then washed three times in.1% (wt/ vol) BSA-PBS for 5 min. Incubation with the immunogold-labeled second antibody was done for 2 h at 25 C. After immunogold labeling, membranes were washed twice in.1% (wt/vol) BSA-PBS for 5 min each, twice in deionized water for 1 min each, and were silver-enhanced by the method of Moeremans et al. (1989). Myofibril samples (NS and ES) were prepared from all eight animals, and all samples were examined by both SDS-PAGE and Western blotting with regard to overall qualitative differences (i.e., no attempt was made to quantify the amount of a given protein band or degradation product, only visual observation of relative intensity or complete absence of a band was used to determine the values in Table 1. Results The days on which protein changes occurred are shown for each animal by treatment in Table 1. All SDS-PAGE gels and Western blots shown herein are from animal number six. Titin SDS-PAGE analysis of both NS and ES samples with an 8% gel (Figure 1a, b) showed that three closely spaced titin bands are present at 0 d. The slowest migrating band is T1 (intact titin) and the fastest band is T2 ( a major, 2,400-kDa degradation product of titin; Robson et al., 1991). The band migrating between T1 and T2 has been referred to as T1-2, and also may be a transient degradation product of titin (Huff-Lonergan et al., 1995). The titin standard consists predominantly of T1 with a lesser amount of the faster migrating T1-2. The T1 band decreases slightly faster in ES samples from five animals (1, 2, 3, 6, and 7) and disappears by either 3 d or 7 d in all but three samples (3 NS, 7 NS, and 7 ES) (Table 1). The T2 band increases in amount through 3 d and is still present at 28 d. In NS samples, an 1,200-kDa band is present at 0 d and 1 d and is barely detectable by 3 d. In ES samples, the 1,200-kDa band decreases slightly faster than in NS

4 1566 HO ET AL. samples and is nearly absent by 1 d (cf. Figure 1a and 1b). Examination of 18% gels of both NS and ES samples (Figure 2a, b) showed that 130- and 39- kda bands increase from 3 d to 7 d and remain as prominent bands at 28 d. A Western blot (Figure 3a, b) produced by 9D10 labeling of a single PVDF membrane that contained proteins simultaneously transferred from one 18% SDS-PAGE gel shows that a more heavily labeled family of degradation products between MHC and 67 kda is present in the 3-, 7-, 14-, and 28-d ES samples. An 39-kDa band that is also labeled by 9D10 is present at 1 d PM, increases by 3 d PM, persists through 28 d PM, and is a more prominent band in the ES samples. Nebulin The nebulin band began to decrease slightly sooner in ES samples from four animals (1, 2, 6, and 8) and was absent earlier in ES samples from one animal (7, Table 1). In SDS-PAGE of 0-d and 1-d NS samples (Figure 1a), nebulin migrates predominantly as a single band that decreases in prominence by 3 d and is absent by 7 d. In ES samples (Figure 1b), nebulin migrates predominantly as a doublet at 1 d that also is absent by 7 d. Desmin and α-actinin The desmin band began to decrease sooner in intensity in ES samples from two animals (3 and 6) and was absent earlier in ES samples from one animal (5, Table 1) than in the samples from the corresponding NS animals. The desmin band observed by 15% SDS-PAGE of NS samples (Figure 4a) also is a light band by 3 d, is not detectable at 7 d, and follows a similar pattern by analysis with 18% SDS-PAGE (Figure 2a). In ES samples (Figure 4b), the desmin band is not detectable by 3 d. In 15% gels (Figure 4) an unidentified prominent band migrates faster than desmin, but in 18% gels (Figure 2) a similar band migrates slower than desmin. This unidentified band in each gel system is unaffected by PM storage and stains with Stains-all, which indicates that it is a glycoprotein and(or) a phosphoprotein (Cutting, 1984). Glycoproteins and(or) phosphoproteins frequently migrate anomalously in SDS-PAGE conditions, e.g., neurofilaments M and H (Kaufmann et al., 1984) and triadin (Knudson et al., 1993a,b). A pab to desmin recognized both the desmin band and an 43- kda desmin degradation product in the desminenriched fraction used as the standard in the Western blot from the 15% gel (Figure 5). After 1 d PM, there is a steady decrease in the desmin band until 28 d, when the desmin band can hardly be seen in both NS and ES samples. An 38-kDa band is very lightly labeled in NS samples (Figure 5a) but is present as a more heavily labeled desmin degradation product from 7 to 28 d in ES samples (Figure 5b). SDS-PAGE (Figures 2a,b and 4a,b) and Western blot (Figure 6a,b) analysis of both NS and ES samples indicated that there is little a-actinin change PM. Troponin-T and Troponin-I Troponin-T degraded sooner in ES samples from five animals (Table 1, Figure 4). There was, however, little or no affect of ES on the migration distance of troponin-i into the gel (Table 1, Figure 4). Troponin-I is degraded in PM muscle as indicated by increased migration distances seen in 14- or 28-d samples. Ultrastructural Changes in Nonstimulated and Electrically Stimulated Muscle Sarcomere Length and Appearance of Contraction Nodes. Immediately after death, sarcomeres in NS muscle samples are well ordered (Figure 7a) and have an average length of 2.5 ±.1 mm (n = 66). Immediately after ES, contraction nodes (CN) (Figure 7b,c) and stretched sarcomeres (Figure 7d) around the CN are formed in seven animals, and all sarcomeres were moderately shortened in the remaining animal (results not shown). The CN, and stretched sarcomeres around the CN, were seen in all samples through 28 d PM (Figure 7e) from four animals but could not be detected at 1 d PM or in subsequent samples in the other three animals that had CN at 0 d. The number of CN ranges from 0 to 4 per grid opening in 0-d ES samples and did not change through 28 d PM in the four animals still exhibiting CN at 1 d PM. The number of sarcomeres per myofibril in a CN ranges from 2 to 10 and averages 4. These extremely shortened sarcomeres are 1.1 ±.1 mm long (n = 82) and contain curved Z-lines, no I-bands, and thick filaments that are in contact with Z-lines (Figure 7c). Stretched sarcomeres in the region adjacent to a CN within the same myofibril are 3.2 ±.2 mm long (n = 73) and contain wide I-bands, crooked Z-lines and M-lines, and A-bands with uneven edges (Figure 7d). In 0-d ES samples, sarcomeres at greater distances from a CN and the associated stretched sarcomeres within the same myofibril, and in myofibrils without a CN (4 to 6 myofibrils away ), are similar to the sarcomeres observed in the 0-d NS samples and have a sarcomere length of 2.6 ±.1 mm (n = 72). At 1 d PM, NS samples have a sarcomere length of 1.8 ±.1 mm (n = 48), compared with a sarcomere length of 2.0 ±.2 mm (n= 47) in ES samples. No further change in sarcomere lengths occurred in both NS and ES samples from 1 to 28 d. Changes in the I-Band/Z-Line Region. Degradation of Z-lines per se progresses similarly in both NS and ES samples as time PM increases. At 1 d PM, one to three discontinuities or gaps in the Z-line, narrower I-bands, double overlap of thin filaments at the center of sarcomeres, and increased intermyofibrillar spaces are

5 PROTEIN DEGRADATION AND STRUCTURAL CHANGES 1567 Figure 1. Eight percent gels of myofibrils prepared at different times postmortem (PM) from (a) nonstimulated (NS) and from (b) electrically stimulated (ES) bovine muscle. Lane Nb = nebulin standard. Lane T = titin standard. T1 = intact titin. T2 = an 2,400-kDa titin degradation product. MHC = myosin heavy chain. T1 and intact nebulin both disappear slightly faster in ES myofibrils. An 1,200-kDa band (arrow) present at 0 d also decreases faster in ES samples. Figure 2. Eighteen percent gels of myofibrils prepared at different times postmortem (PM) from (a) nonstimulated (NS) and from (b) electrically stimulated (ES) bovine muscle. Lane TM = tropomyosin standard. Lane TN-T = troponin-t standard. Lane D = desmin-enriched standard. Lane M = partially purified myosin standard. Lane TN-I = troponin-i standard. Lane TN-C = troponin-c standard. MHC = myosin heavy chain. MLC-1, 2, 3 = myosin light chains 1, 2 and 3. These higher percent gels show that 130-kDa (asterisk) and 39-kDa (arrow head) components increase in amount as PM time increases. When compared with TN-C, which is immediately below TN-I, TN-I decreases somewhat with increasing PM time.

6 1568 HO ET AL. Figure 3. Western blots prepared from a single 18% gel of (a) nonstimulated (NS) and (b) electrically stimulated (ES) samples and labeled with 9D10 titin mab. MHC = myosin heavy chain. A large family of lower molecular weight bands is labeled more heavily in ES samples than in NS samples. Figure 4. Fifteen percent gels of myofibrils prepared at different times postmortem from (a) nonstimulated (NS) and from (b) electrically stimulated (ES) bovine muscle. Lane M = partially purified myosin. Lane TN-T = troponin-t standard. Lane D = desmin-enriched standard. Lane TM = tropomyosin standard. Lane TN-I = troponin-i standard. Lane TN-C = troponin-c standard. MHC = myosin heavy chain. MLC-1, 2, 3 = myosin light chains 1, 2 and 3. Desmin decreases as PM time increases, but a-actinin is not affected by stimulation or PM time.

7 PROTEIN DEGRADATION AND STRUCTURAL CHANGES 1569 Figure 5. Western blots prepared from 15% gels of (a) nonstimulated (NS) and (b) electrically stimulated (ES) samples were labeled with polyclonal anti-desmin. Lane D = desmin-enriched standard. The desmin band decreases similarly in NS and ES samples. Increased labeling of lower molecular weight bands, especially a 38-kDa band in ES samples, is seen as PM time increases. evident in both NS (Figure 8a-1) and ES samples. The extent of Z-line structural degradation that typifies 3-d PM samples (Figure 8a-2) is evident when compared with a typical Z-line in 0-d samples (Figure 8a-3). Overlapping, interconnected filaments characterize Z-lines seen in longitudinal sections of 0-d samples (Figure 8a-3). By 3 d PM, the fibrillar structure has been replaced by an amorphous band that is not uniform in density and that is adjacent to I- band fractures (Figure 8a-2). By 28 d PM, Z-lines are Figure 6. Western blots prepared from a single 15% gel of (a) nonstimulated (NS) and (b) electrically stimulated (ES) samples were labeled with polyclonal anti-a-actinin. Electrical stimulation and PM time do not alter the a-actinin content of myofibrils.

8 1570 HO ET AL. wider, are amorphous with more irregular edges, are less dense and often have 2 to 5 gaps in a typical Z- line (results not shown). In addition to these characteristic PM changes in Z- line structure, three types of I-band fractures were observed adjacent to Z-lines. The frequency of these I- Figure 7. Structure of nonstimulated (NS) and electrically stimulated (ES) bovine skeletal muscle. (a) NS 0-d muscle shows the regular banding pattern characteristic of relaxed skeletal muscle. Bar = 2 mm. (b) An ES 0-d muscle fiber contains a contraction node (CN) and stretched sarcomeres (S), but an adjacent fiber contains relaxed sarcomeres. Bar = 4 mm. (c) Higher magnification of part of a contraction node in a 0-d stimulated fiber. Thick filaments of the A-band are in contact with Z-lines (Z). Bar =.5 mm. (d) Myofibrils that contain contraction nodes also contain stretched sarcomeres such as this in adjacent regions. Note the wide I-band centered on the Z-line (Z) and the irregular edges of the A-band (A). Bar = 1 mm. (e) Contraction nodes (CN) often are still present at 28 d in ES samples. Bar = 4 mm.

9 PROTEIN DEGRADATION AND STRUCTURAL CHANGES 1571 Figure 8. The Z-line and I-band changes in postmortem (PM) bovine skeletal muscle. (a-1) Nonstimulated (NS) muscle at 1 d PM contains shortened sarcomeres and Z-lines that have discontinuities (arrows). Bar = 1 mm. (a-2) Z- lines lose their fibrillar structure and become amorphous by 3 d in ES samples. I-band fractures adjacent to Z-lines (arrow) are readily seen. Bar =.2 mm. (a-3) Z-lines in 0-d NS samples contain overlapping and interconnected filaments that give Z-lines a distinct fibrillar structure. Bar =.2 mm. (b) Narrow I-band fractures (arrows) adjacent to the Z-line with filaments spanning the fracture zone are present in ES samples at 1 d. Bar = 1 mm. (c) Intermediate width I-band fracture (arrow) in NS 7-d muscle samples. Bar = 2 mm. (d) Wide I-band fractures (arrow) in NS 28-d muscle samples frequently contain sarcomeres that are randomly oriented (double arrow head). This orientation implies that sarcomeres released from adjacent myofibrils are present in the fracture zone. Bar = 2 mm. band fractures was influenced by both ES and storage time (Table 2). Narrow I-band fractures (Figure 8b), ranging from.1 mm to.8mm wide and involving 1 to 6 contiguous myofibrils, are present in 1.8% of 1-d NS myofibrils and in 7.2% of ES myofibrils. These fractures in 1-d samples are spanned by a few filaments that have diameters less than or equal to the diameter of thin filaments. At 3 d PM, narrow I- band fractures can be detected in 5.8% of myofibrils in NS samples and in 9.5% of myofibrils in ES samples. At 7 d PM, narrow I-band fractures, with or without filaments spanning the fracture zone, were observed in 6.5% of the myofibrils in NS samples and in 11.3% of myofibrils in ES samples. At 14 d PM, narrow I-

10 1572 HO ET AL. Table 2. Frequency of three types of I-band fractures Days postmortem Fracture type (treatment) Frequency expressed as % of all myofibrils Narrow (NS) 0 a 1.8 ±.5 b 5.8 ± 1.3 d 6.5 ±.9 f 10.5 ± 2.0 g 11.6 ± 2.7 h Narrow (ES) 0 a 7.2 ± 1.2 c 9.5 ± 1.3 e 11.3 ± 1.6 g 13.8 ± 1.8 g 14.0 ± 2.4 h Intermediate and wide (NS) 0 a 0 a 0 a 22.5 ± 4.2 j 43.3 ± 9.0 l 63.6 ± 9.5 m Intermediate and wide (ES) 0 a 0 a 15.0 ± 3.4 i 30.7 ± 3.7 k 45.7 ± 4.9 l 70.9 ± 6.4 m a m Means within a column and within a fracture type lacking a common superscript differ ( P <.05). band fractures with or without spanning filaments increased in frequency and can be seen in 10.5% of the myofibrils in NS samples and in 13.8% of the myofibrils in ES samples. At 28 d PM, all narrow I- band fractures have no filaments spanning the fracture zone and are present in 11.6% of the myofibrils in NS samples and in 14.0% of the myofibrils in ES samples. The frequency of narrow I-band fractures in ES samples was significantly higher ( P <.05) at 1, 3, and 7 d and was also higher at 14 and 28 d. Neither intermediate width I-band fractures (Figure 8c), ranging from.8 to 3.5 mm wide and involving 3 to 12 contiguous myofibrils, nor wide I- band fractures (Figure 8d), which range from 4 to more than 20 mm wide, involve 6 to 35 contiguous myofibrils and often contain released sarcomeres, can be detected in 3-d NS samples but are present in 15% of the myofibrils in 3-d ES samples. At 7 d PM, both intermediate and wide I-band fractures were found in 22.5% of myofibrils in NS samples and in 30.7% of myofibrils in ES samples. The combined frequency of intermediate and wide I-band fractures increased to 43.3% of myofibrils in 14-d NS samples and to 45.7% in 14-d ES samples. Wide I-band fractures were seen in 63.6% of myofibrils at 28 d in NS samples and in 70.9% of the 28 d ES samples. The frequency of intermediate and wide I-band fractures was significantly higher ( P <.05) in ES samples at 3 and 7 d and was also higher at 14 and 28 d. Discussion Effect of Electrical Stimulation on Degradation of Titin, Nebulin, α-actinin, and Desmin During Postmortem Storage Several investigators (e.g., Dransfield, 1992; Dransfield et al., 1992; Koohmaraie, 1992; Uytterhaegen et al., 1994) have suggested that m-calpain activity in PM muscle causes many of the proteolytic alterations in key structural proteins and that this, in turn, is related to tenderization. It also has been suggested that one of the effects of ES is to increase the amount of proteolytic enzyme activity in bovine muscle (Dransfield, 1992; Dransfield et al., 1992; Smulders and van Laack, 1992; Uytterhaegen et al., 1992). One objective of our experiments was to determine the effect of ES on proteins believed to be involved in muscle integrity. Reports from several laboratories have indicated that titin was degraded during PM storage (Lusby et al., 1983; Bandman and Zdanis, 1988; Fritz and Greaser, 1991; Huff-Lonergan et al., 1995; Taylor et al., 1995). Association of the degree of titin degradation with increasing meat tenderness has been reported by Paterson and Parrish (1987) and Huff-Lonergan et al. (1995); however, Fritz et al. (1993) found no relationship between titin content on SDS gels and tenderness. Our SDS-PAGE results showed that the T1 band of titin degraded during PM storage, and that for five of the eight animals studied, the degradation rate was slightly faster in ES than in NS samples. The T2 titin band increased in amount and was still present at 28 d in both ES and NS samples (Figure 1a,b). The small amount of the 1,200-kDa band (Figure 1b), reported to be a fragment of T1 as described by Takahashi et al. (1992), is also, in turn, further degraded slightly faster in ES samples from six animals. However, ES does not accelerate the appearance of either a 130- or a 39-kDa band (cf. Figure 2a and b). The source of the 130-kDa band is not known, but it is possible that the 130-kDa band is a fragment of higher molecular weight proteins such as titin, nebulin or M- line components. Western blots labeled with mab 9D10 show that a large family of degradation products (below MHC to 22-kDa) is labeled (Figure 3a, b), one of which is a 39-kDa band that increases during the first 7 d PM. The 39-kDa band, which migrates slightly faster than actin, has previously been hypothesized to be creatine phosphokinase (Pommier et al., 1987) and seemingly was identified as a 43-kDa peptide by Uytterhaegen et al. (1992). Lusby et al. (1983), Locker and Wild (1984), and Fritz and Greaser (1991) have reported that nebulin degrades somewhat faster than titin in PM bovine muscle. The intact nebulin band in SDS-PAGE (Figure 1) decreases slightly faster in ES than in NS samples from five animals (Table 1). Intact nebulin disappears either at the same time (six animals) or

11 PROTEIN DEGRADATION AND STRUCTURAL CHANGES 1573 slightly faster (two animals) than titin in NS samples (Table 1). Because desmin constitutes <1% by weight of myofibrillar protein in skeletal muscle, it is seen as a very minor band on SDS-PAGE even in 0-d samples and in samples taken 1 and 3 d PM (Figure 4). To determine whether intact desmin persisted throughout the sampling period and whether desmin was affected by ES, it was necessary also to use Western blots as a detection system (Figure 5). The intact desmin band decreased at comparable rates in NS and ES samples with little or no intact desmin present at 14 or 28 d (Table 1). A 38-kDa band that was first recognized by the antibody in the 7 d samples was consistently more heavily labeled in the 7-, 14-, and 28-d PM ES samples than in the NS samples. The intensity of the labeling of this band in the ES samples suggests that desmin is degraded more rapidly in ES samples than in NS samples. Our blot results with the pab are consistent with those of Hwan and Bandman (1989), who reported that desmin fragments existed by 4 d in muscle stored at 4 C, and that little intact desmin remained at 3 wk PM. In both SDS-PAGE and Western blots, the amount of intact a-actinin does not change in NS or ES samples during PM storage, which is consistent with the results of Stromer et al. (1974) and Uytterhaegen et al. (1992). This also is consistent with the results of Goll et al. (1991), who showed that calpain digestion of myofibrils releases intact a- actinin. Our observation that ES caused earlier PM degradation of troponin-t in five animals supports our previous report (Ho et al., 1994) that ES caused earlier degradation of troponin-t. Effect of Electrical Stimulation on Ultrastructural Changes During Postmortem Storage Electrical stimulation improves meat tenderness (Savell et al., 1977) and has been hypothesized to do so by preventing cold-shorting (Takahashi et al., 1984), by producing fractures and breaks in muscle fibers (Savell et al., 1978; Takahashi et al., 1987), by producing an intermediate rate of glycolysis (Smulders et al., 1990), or by a combination of these factors (for review, see Smulders and van Laack, 1992). Seven animals had prominent CN immediately after stimulation (Figure 7b). These CN consisted of 2 to 10 highly contracted sarcomeres per myofibril similar to those described by Savell et al. (1978) and Sorinmade et al. (1982). George et al. (1980), however, suggested, on the basis of histological examination, that the formation of CN was due to denaturation of sarcoplasmic proteins and not to muscle fiber contraction or disruption. The observation by George et al. (1980) that CN were birefringent in polarized light indicates that CN are sites of accumulation of thick filaments, which are responsible for A-band birefringence in muscle. Our thin, longitudinal sections through a CN clearly show densely packed thick filaments in the CN (Figure 7c). Muscle from four animals still contained CN at 28 d PM, but three animals had no CN at 1 d PM. Cassens et al. (1963 a,b) also found that CN were absent in porcine muscle at the onset of rigor mortis and 24 h PM and suggested that if the stimulus is not too intense, the process of CN formation is reversible, but when the stimulus is intense, CN are formed before rigor onset, and the formation of CN becomes irreversible. The existence of CN in muscle at 1 d PM has been reported by Savell et al. (1978) and Takahashi et al. (1987), at 2 d PM by Sorinmade et al. (1982), at 4 d PM by McKeith et al. (1980), and at 5 d PM by George et al. (1980). None of these groups sampled at longer PM time than 5 d, so it is not known whether CN persisted in their samples. Sarcomere lengths in 0-d ES muscle CN were 44% of rest length sarcomeres in control muscle, but stretched sarcomeres adjacent to a CN within the same myofibril were 128% of rest length sarcomeres. At 1 d PM, there was little difference in sarcomere length between NS and ES muscle samples. Although Savell et al. (1978) and Uytterhaegen et al. (1992) both reported that ES had no effect on sarcomere length in 20 to 24 h PM muscle samples, sarcomeres were measured by Savell et al. (1978) in isolated myofibrils and by Uytterhaegen et al. (1992) in ground muscle samples. Because of the very different sample preparation techniques, comparisons with our data are, at best, problematic. Alterations near, or changes in, Z-lines and increased intermyofibrillar spaces with the loss of lateral attachments between Z-lines of adjacent myofibrils during PM storage reported here are consistent with previous observations on bovine muscle (Stromer et al., 1967; Davey and Gilbert, 1967, 1969; Davey and Dickson, 1970; Henderson et al., 1970; Sorinmade et al., 1982). Our results also showed that Z-lines were no longer continuous structures by 1 d (Figure 8a-1), and many became amorphous by 3 d (Figure 8a-2). Taylor et al. (1995), however, reported that Z-lines remained almost unchanged ultrastructurally after 16 d PM. It is possible that the section thickness, the staining method, and the relatively low magnification chosen by Taylor et al. (1995) may have obscured important ultrastructural changes in Z- line structure in their samples. The extent or rate of progression of Z-line changes per se and the development of intermyofibrillar spaces were very similar in NS and ES samples. Fractures in the I-band adjacent to the Z-line result in Z-lines still attached to thin filaments on one side of the fracture. I-band fractures adjacent to the Z-line have also been noted by Taylor et al (1995) in nonhomogenized bovine biceps femoris muscle that was stored 4 d or longer at 4 C and in homogenized bovine biceps muscle used for determining the myofibrillar

12 1574 fragmentation index. Davey and Dickson (1970) also reported that breaking at the I-band/Z-line junction and occasionally at the edge of the A-band occurred in bovine muscle during aging. Ouali (1990) described I- band fractures in bovine muscle after 15 d PM as being mainly at the N 2 line, but occasionally near the A-I junction. These observations suggest that the I- band adjacent to the Z-line is a structurally labile site. We characterized these fractures as narrow, intermediate, or wide, depending on the axial width of the fracture. The frequency of these fractures is affected by ES and by time PM. Narrow fractures occur at 1 d PM in both NS and ES samples but with significantly ( P <.05) greater frequency at 1, 3, and 7 d PM in ES samples. The frequency of narrow fractures in ES muscle gradually increases from 7.2% at 1 d PM to 14% at 28 d PM. Intermediate and wide I-band fractures are seen sooner PM and with greater frequency in ES samples. Taylor et al. (1995) suggested that degradation of titin and nebulin may contribute to the increased I-band fractures during PM storage. Our results also indicated that the disappearance of T1 and nebulin are closely related temporally to the appearance of both intermediate and wide I-band fractures. In NS samples, T1 and intact nebulin (Figure 1a) disappeared at 7 d, and concomitantly both intermediate and wide I-band fractures were first seen at 7 d. In ES samples, T1 and nebulin (Figure 1b) were nearly gone at 3 d, and both types of I-band fractures also appeared earlier at 3 d. Despite the uncertainty of the causative agent(s), the progressive increase in both I-band fracture frequency and width in ES samples compared with NS samples and during PM storage suggests that these fractures may contribute to improved tenderness. Because we have observed major ultrastructural changes both inside the Z-line and beside it, we believe it may be premature to conclude that only those changes near the Z-line are important (Taylor et al., 1995). It seems reasonable to us that alterations occurring in the Z-line itself (e.g., degradation or loss of function of Z-line components), which are many (Robson et al., 1991), may release titin and(or) nebulin from their Z- line attachment sites and cause the fractures. From our results, we conclude that titin, nebulin, and desmin, but not a-actinin, are degraded PM. Our data show that titin and nebulin are degraded much more rapidly in PM muscle than desmin or troponin-t and that nebulin may be degraded more rapidly than titin. Although treatment effects on PM proteolysis varied somewhat among animals, analysis by SDS- PAGE showed accelerated degradation of titin, nebulin, and troponin-t in ES samples from five animals. Western blots showed that ES enhanced the degradation of desmin in three animals and had no detectable affect in the other animals. Electrical stimulation produced CN and variable sarcomere lengths in 0-d samples. Electrical stimulation also increased the HO ET AL. frequency of I-band fractures in ES muscle samples. Wide I-band fractures adjacent to the Z-line were seen sooner PM in ES than in NS bovine muscle. Degradation of Z-lines occurred at similar rates in NS and ES samples. Formation of both I-band fractures and CN are the two principal structural alterations caused by ES. Implications One result of electrical stimulation is an increase in meat tenderness. We have shown that electrical stimulation slightly accelerates the degradation of the cytoskeletal proteins titin and nebulin in postmortem muscle. We also found that electrical stimulation increased the frequency of three types of myofibrillar I-band fractures (narrow, intermediate, and wide) and caused wide I-band fractures to appear sooner postmortem. Electrical stimulation may improve meat tenderness by a combination of mechanical disruption and enhanced proteolysis, but mechanical disruption seems to be the predominant mechanism. Literature Cited Bandman, E., and D. Zdanis An immunological method to assess protein degradation in post-mortem muscle. Meat Sci. 22:1. Cassens, R. G., E. J. Briskey, and W. G. Hoekstra. 1963a. Similarity in the contracture bands occurring in thaw-rigor and in other violent treatments of muscle. Biodynamica 9:165. Cassens, R. G., E. J. Briskey, and W. G. Hoekstra. 1963b. Electron microscopic observations of a dense, irregularly banded material occurring in some porcine muscle fibres. Nature (Lond.) 198:1004. Cutting, J. A Gel protein stains: Phosphoproteins. Methods Enzymol. 104:451. Davey, C. L., and M. R. Dickson Studies in meat tenderness. 8. Ultra-structural changes in meat during aging. J. Food Sci. 35:56. Davey, C. L., and K. V. Gilbert Structural changes in meat during ageing. J. Food Technol. 2:57. Davey, C. L., and K. V. Gilbert Studies in meat tenderness. 7. Changes in the fine structure of meat during aging. J. Food Sci. 34:69. Dransfield, E Modelling post-mortem tenderisation-iii: Role of calpain I in conditioning. Meat Sci. 31:85. Dransfield, E., D. J. Etherington, and M.A.J. Taylor Modelling post-mortem tenderisation-ii: Enzyme changes during storage of electrically stimulated and non-stimulated beef. Meat Sci. 31:75. Fritz, J. D., and M. L. Greaser Changes in titin and nebulin in postmortem bovine muscle revealed by gel electrophoresis, western blotting and immunofluorescence microscopy. J. Food Sci. 56:607. Fritz, J. D., M. C. Mitchell, B. B. Marsh, and M. L. Greaser Titin content of beef in relation to tenderness. Meat Sci. 33:41. Fritz, J. D., D. R. Swartz, and M. L. Greaser Factors affecting polyacrylamide gel electrophoresis and electroblotting of highmolecular-weight myofibrillar proteins. Anal. Biochem. 180: 205. George, A. R., J. R. Bendall, and R.C.D. Jones The tenderizing effect of electrical stimulation of beef carcasses. Meat Sci. 4:51.

13 PROTEIN DEGRADATION AND STRUCTURAL CHANGES 1575 Goll, D. E., W. R. Dayton, I. Singh, and R. M. Robson Studies of the a-actinin/actin interaction in the Z-disk by using calpain. J. Biol. Chem. 266:8501. Henderson, D. W., D. E. Goll, and M. H. Stromer A comparison of shortening and Z line degradation in post-mortem bovine, porcine, and rabbit muscle. Am. J. Anat. 128:117. Ho, C.-Y., M. H. Stromer, and R. M. Robson Identification of the 30-kDa polypeptide in post mortem skeletal muscle as a degradation product of troponin-t. Biochimie 76:369. Huff-Lonergan, E., F. C. Parrish, Jr., and R. M. Robson Effects of postmortem aging time, animal age, and sex on degradation of titin and nebulin in bovine longissimus muscle. J. Anim. Sci. 73:1064. Hwan, S.-F., and E. Bandman Studies of desmin and a- actinin degradation in bovine semitendinosus muscle. J. Food Sci. 54:1426. Karnovsky, M. J A formaldehyde-glutaraldehyde fixative of high osmolality for use in electron microscopy. J. Cell Biol. 27: 137A. Kaufmann, E., N. Geisler, and K. Weber SDS-PAGE strongly overestimates the molecular masses of the neurofilament proteins. FEBS Lett. 170:81. Knudson, C. M., K. K. Stang, A. O. Jorgensen, and K. P. Campbell. 1993a. Biochemical characterization and ultrastructural localization of a major junctional sarcoplasmic reticulum glycoprotein (triadin). J. Biol. Chem. 268: Knudson, C. M., K. K. Stang, C. R. Moomaw, C. A. Slaughter, and K. P. Campbell. 1993b. Primary structure and topological analysis of a skeletal muscle-specific junctional sarcoplasmic reticulum glycoprotein (triadin). J. Biol. Chem. 268: Koohmaraie, M The role of Ca 2+ dependent proteases (calpains) in post mortem proteolysis and meat tenderness. Biochimie 74:239. Koohmaraie, M., J. E. Schollmeyer, and T. R. Dutson Effect of low-calcium-requiring calcium activated factor on myofibrils under varying ph and temperature conditions. J. Food Sci. 51: 28. Laemmli, U. K Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond.) 227: 680. Locker, R. H., and D.J.C. Wild The fate of the large proteins of the myofibril during tenderizing treatments. Meat Sci. 11:89. Lusby, M. L., J. F. Ridpath, F. C. Parrish, Jr., and R. M. Robson Effect of postmortem storage on degradation of the myofibrillar protein titin in bovine Longissimus muscle. J. Food Sci. 48:1787. McKeith, F. C., C. G. Smith, T. R. Dutson, J. W. Savell, R. L. Hostetler, and Z. L. Carpenter Electrical stimulation of intact or split steer or cow carcasses. J. Food Prot. 43:795. Moeremans, M., G. Daneels, M. De Raeymaeker, B. De Wever, and J. De Mey The use of colloidal gold particles for testing the specificity of antibodies and/or the presence of antigen. In: A.J.J. Verkleij and J.L.M. Leunissen (Ed.) Immuno-Gold Labeling in Cell Biology. p 17. CRC Press, Boca Raton, FL. Ouali, A Meat tenderization: Possible causes and mechanisms. A review. J. Muscle Foods 1:129. Paterson, B. C., and F. C. Parrish, Jr SDS-PAGE conditions for detection of titin and nebulin in tender and tough bovine muscles. J. Food Sci. 52:509. Pommier, S. A., L. M. Poste, and G. Butler Effect of low voltage electrical stimulation on the distribution of cathepsin D and the palatability of Longissimus dorsi from Holstein veal calves fed a corn or barley diet. Meat Sci. 21:203. Reiling, B., G. Rouse, D. G. Olson, D. Duello, and D. Maxwell A performance and carcass evaluation of extended feeding of small-framed steers to meet the Japanese beef demand. ISU Beef/Sheep Report (A. S. Leaflet R 813). p 71. Ames, IA. Robson, R. M., and T. W. Huiatt Roles of the cytoskeletal proteins desmin, titin and nebulin in muscle. Proc. Recip. Meat. Conf. 36:116. Robson, R. M., T. W. Huiatt, and F. C. Parrish, Jr Biochemical and structural properties of titin, nebulin and intermediate filaments in muscle. Proc. Recip. Meat. Conf. 44:7. Robson, R. M., J. M. O Shea, M. H. Stromer, M. Yamaguchi, M. K. Hartzer, Y.-Y.S. Pang, T. W. Huiatt, W. E. Rathbun, R. R. Evans, F. L. Richardson, P. J. Schreiner, J. F. Ridpath, and L. E. Kasang Biochemistry and molecular architecture of muscle cell 10-nm filaments and Z-line: Roles of desmin and a- actinin. Proc. Recip. Meat. Conf. 34:5. Robson, R. M., M. H. Stromer, T. W. Huiatt, J. M. O Shea, M. K. Hartzer, F. L. Richardson, and W. E. Rathbun Biochemistry and structure of desmin and the recently discovered muscle cell cytoskeleton. Proc. 26th Eur. Mtg. of Meat Res. Workers, Colorado Springs, CO. 1:22. SAS SAS Language and Procedures. SAS Inst. Inc., Cary, NC. Savell, J. W., T. R. Dutson, G. C. Smith, and Z. L. Carpenter Structural changes in electrically stimulated beef muscle. J. Food Sci. 43:1606. Savell, J. W., G. C. Smith, T. R. Dutson, Z. L. Carpenter, and D. A. Suter Effect of electrical stimulation on palatability of beef, lamb and goat meat. J. Food Sci. 42:702. Smulders, F.J.M., B. B. Marsh, D. R. Swartz, R. L. Russell, and M. E. Hoenecke Beef tenderness and sarcomere length. Meat Sci. 28:349. Smulders, R.J.M., and L.J.M. van Laack Accelerated processing to improve the ageing response of meat. In: F.J.M. Smulders, F. Toldrá, J. Flores, and M. Prieto (Ed.) New Technologies for Meat and Meat Products. p 181. ECCEAMST, Audet Tijdschriften B. V., Kroonstaadt, The Netherlands. Sorinmade, S. O., H. R. Cross, K. Ono, and W. P. Wergin Mechanisms of ultrastructural changes in electrically stimulated beef longissimus muscle. Meat Sci. 6:71. Stromer, M. H., and M. Bendayan Arrangement of desmin intermediate filaments in smooth muscle cells as shown by high-resolution immunocytochemistry. Cell Motil. Cytoskel. 11: 117. Stromer, M. H., D. E. Goll, W. J. Reville, D. G. Olson, W. R. Dayton, and R. M. Robson Structural changes in muscle during postmortem storage. Proc. IV Int. Cong. Food Sci. Technol. 1: 401. Stromer, M. H., D. E. Goll, and L. E. Roth Morphology of rigor-shortened bovine and the effect of trypsin on pre-and postrigor myofibrils. J. Cell Biol. 34:431. Takahashi, G., J. V. Lochner, and B. B. Marsh Effect of lowfrequency electrical stimulation on beef tenderness. Meat Sci. 11:207. Takahashi, G., S.-M. Wang, J. V. Lochner, and B. B. Marsh Effects of 2-Hz and 60-Hz electrical stimulation on the microstructure of beef. Meat Sci. 19:65. Takahashi, K., A. Hattori, R. Tatsumi, and K. Takai Calciuminduced splitting of connectin filaments into b-connectin and a 1,200-kDa subfragment. J. Biochem. 111:778. Taylor, R. G., G. H. Geesink, V. F. Thompson, M. Koohmaraie, and D. E. Goll Is Z-disk degradation responsible for postmortem tenderization? J. Anim. Sci. 73:1351. Towbin, H., T. Staehelin, and J. Gordon Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci. USA 76:4350. Uytterhaegen, L., E. Claeys, and D. Demeyer The effect of electrical stimulation on beef tenderness, protease activity and myofibrillar protein fragmentation. Biochimie 74:275. Uytterhaegen, L., E. Claeys, and D. Demeyer Effects of exogenous protease effectors on beef tenderness development and myofibrillar degradation and solubility. J. Anim. Sci. 72: Wang, K Purification of titin and nebulin. Methods Enzymol. 85:264. Wang, K. and J. Wright Architecture of the sarcomere matrix of skeletal muscle: Immunoelectron microscopic evidence that suggests a set of parallel inextensible nebulin filaments anchored at the Z line. J. Cell Biol. 107:2199.