Pulse Electrophoresis of Muscle Myosin Heavy Chains in Sodium Dodecyl Sulfate–Polyacrylamide Gels

Analytical Biochemistry 291, 229 –236 (2001) doi:10.1006/abio.2001.5018, available online at http://www.idealibrary.com on

Pulse Electrophoresis of Muscle Myosin Heavy Chains in Sodium Dodecyl Sulfate–Polyacrylamide Gels Jose´ A. A. Sant’Ana Pereira, Marion Greaser, and Richard L. Moss 1 Department of Physiology, University of Wisconsin Medical School, and Muscle Biology Laboratory, Madison, Wisconsin, 53706

Received August 8, 2000; published online March 9, 2001

We have developed a new method that provides enhanced resolution of myosin heavy chain (MHC) isoforms by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). The key feature of this protocol involves the application of current to slab SDS gels in a pulsatile, repetitive manner rather than continuously as in standard gel systems. This protocol, designated pulse electrophoresis, was achieved by means of a device that intermittently gates the output of a conventional power supply. When used in long (32 cm) separating gels, pulse electrophoresis not only significantly improves the resolution of MHC isoforms compared to conventional systems, but also reduces common artifacts associated with long running times, such as blurred bands and comingling of closely spaced bands. In addition to the increased resolution of protein bands, pulse electrophoresis also allows detection of bands corresponding to previously unidentified MHC isoforms in mammalian and avian tissue. In rat myocardium, for example, pulse electrophoresis revealed three MHC isoform bands, two of which appeared to correspond to two alpha-MHC subspecies. Alternative splicing of the rat alpha-MHC gene is known to generate two isoform species differing by inclusion (or exclusion) of a single glutamine residue, whose relative levels of expression correspond nicely with the amounts of each band identified in this study. Therefore, we cannot rule out that the system presented here may be sufficiently sensitive to differentiate between high molecular weight proteins differing in a single amino acid. © 2001 Academic Press Key Words: SDS-PAGE; pulse electrophoresis; myosin heavy chains.

It is widely recognized that gel electrophoresis is among the most powerful and convenient methods for 1

To whom correspondence should be addressed.

0003-2697/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

separation of macromolecules. With significant improvements such as the introduction of discontinuous pH, polyacrylamide gel (PAGE) 2 systems in presence of amphipathic molecules like sodium dodecyl sulfate (SDS), electrophoretic methods have become routinely used in the vast majority of laboratories. Even though electrophoretic protocols have been optimized in various ways, it is recognized that the sensitivity of SDSPAGE is still inadequate for the resolution of some classes of proteins, such as muscle myosins, which are encoded by closely linked isogenes. Muscle myosins, also called type II myosins, are the basic molecular motors involved in muscular contraction and consist of two heavy chains (MHCs) and two pairs of light chains referred to as essential and regulatory light chains, respectively. MHC, the predominant protein of the thick filament, is highly conserved across vertebrates (5) and consists of a large globular catalytic domain that contains the sites for binding both actin and nucleotides and a more slender neck region that stretches from the catalytic domain to the junction with the rod (12). Nine MHC isoforms have so far been identified in mammalian muscles (for review, see 15) and in chicken, as many as 30 different MHC isoforms are believed to exist (13). The amino acid sequences of sarcomeric MHC isoforms are highly homologous among mammalian muscles and few regions of sequence divergence have been identified. Two such regions consist of loops located in the vicinity of the ATP-binding and actin-binding sites of the myosin head, which have been implicated with the regulation the mechanochemical properties of the molecular motors (14, 18, 19, 22). All mammalian MHC isoforms are encoded by distinct genes, which are expressed in a tissue-specific 2 Abbreviations used: PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate; MHC, myosin heavy chain; TA, tibialis anterior; ALD, latssimus dorsi; PAT, patagialis; DTT, dithiothreitol; APS, ammonium persulfate.

229

230

SANT’ANA PEREIRA, GREASER, AND MOSS

and developmentally regulated manner (15). The ␤-cardiac/slow type I, and fast type IIA, IIX, and IIB MHCs are the predominant isoforms in adult skeletal muscles; embryonic and perinatal are predominant in developing tissue; and beta-MHC (␤) and alpha-MHC (␣) are the major cardiac isoforms which are differently expressed in atria and ventricles of distinct mammalian species. Although all of the above protein isoforms have been resolved in one way or another by existing SDS-PAGE protocols, it is widely recognized that significant limitations in SDS-PAGE persist. One limitation concerns the reproducible separation of already identified MHC isoforms, particularly the skeletal type IIA and type IIX MHCs and the ␣- and ␤-cardiac MHC isoforms, which exhibit very similar migration mobilities among mammalian species. Additional limitations arise from various observations that there are more MHC isoforms than those detected by conventional electrophoresis. First, even though only two cardiac MHC isoforms have been resolved to date by existing SDS-PAGE protocols, there is now evidence that the rat ␣-MHC gene generates two variants by alternative splicing (16). Second, molecular genetic analysis of the chromosomal localization of human MHC genes has shown that the skeletal MHC cluster on chromosome 17 comprises more MHC genes than the number of MHC isoforms yet characterized (23). Third, the MHC polymorphism in chicken muscles has been shown to be greater than in mammalian muscles and yet from the 30 or so possible isoforms detected with molecular biological techniques, PAGE has identified considerably fewer protein isoforms. In an attempt to resolve some of the limitations listed above we have developed an ultrasensitive SDSPAGE protocol that is based on the principle initially introduced by Fauteck and Kandarian (7) and which we refer to as “pulse electrophoresis.” This method consists of applying intermittent bursts of current to polyacrylamide gels by means of a pulsing unit that is connected to a conventional power supply. We found this maneuver to improve significantly upon the resolution of MHC isoforms obtained with conventional SDS-gels, especially when using long separating slab gels. MATERIALS AND METHODS

Tissue Collection and Processing All of the animals used in this study were housed, cared for, and sacrificed in accordance to the guidelines of animal use committees at the University of Wisconsin. Skeletal, smooth, and heart muscle samples were collected and immediately frozen in liquid nitrogen. Frozen tissue (100 – 600 mg) from rat tibialis anterior muscle (TA) and ventricles (n ⫽ 4), pig ventricles (n ⫽ 1), chicken ventricles and atria (n ⫽ 2), and latissimus dorsi muscle (ALD) and patagialis muscle (PAT), as

well as from gizzard (n ⫽ 2), were homogenized (Polytron) separately in 10 ml of ice-cold relaxing solution (100 mM KCl, 10 mM imidazole, 5 mM MgCl 2, 2 mM EGTA, 4 mM ATP, pH 7.0) and centrifuged for 10 min (12,000 rpm). The pellet was resuspended in a skinning solution consisting of relaxing solution containing 1% Triton X-100. Both relaxing and skinning solutions contained 2 mM Pefabloc (Boehringer Mannheim, Germany), a protease inhibitor. The samples were placed for 6 min in ice with occasional mixing and then centrifuged. The supernatants were subsequently decanted and the pellets resuspended in 2 vol of sample buffer consisting of 15% glycerol, 15 mM dithiothreitol (DTT), 62.5 mM Tris/HCl, pH 6.8, and 2.5% bromphenol blue. The samples were finally ultrasonicated (5 min) and centrifuged (1 min at 12,000g) to pellet any undissolved material. Additional rat hearts (n ⫽ 2) were enzymatically digested to obtain cardiac myocytes using a protocol modified from that previously published by Strang and Moss (17). In brief, the hearts of anesthetized (methoxyfluorane) rats were excised and the aorta was cannulated for subsequent retrograde perfusion via the coronary circulation using a modified Langendorff apparatus. The hearts were initially perfused with a Ca 2⫹-containing Ringer solution (1.2 mM MgCl 2, 1.0 mM CaCl 2, 4.8 mM KCl, 118.0 mM NaCl, 2.0 mM KH 2PO 4, 5.0 mM pyruvate, 11 mM glucose, and 25.0 mM Hepes, pH 7.4), a Ca 2⫹-free Ringer solution (4.5 min), and finally a collagenase-containing (1.0 mg/ml) Ringer solution (11 min) which also contained 1.0 mg/ml hyaluronidase and 0.05 mM CaCl 2. The hearts were subsequently minced and incubated for 20 min with the collagenase-containing Ringer solution, plus 0.25% trypsin. Immediately after the incubation period trypsin inhibitor (Sigma) was added to the solution (10.0 mg/ml). Undigested material was then removed by filtration (300 ␮m Teflon mesh) and the cells were resuspended in a Ringer solution containing 1.0 mM Ca 2⫹. All solutions were maintained at 37°C and bubbled with 100% O 2. Prior to skinning (6 min at 22°C in 1.0 mM free Mg 2⫹ 100 mM KCl, 2.0 mM EGTA, 4.0 mM ATP, 10 mM imidazole, pH 7.0, and 0.3% Triton X-100), the cells were aliquoted and resuspended in 1.0 mM Ca 2⫹ Ringer solution. The cells were finally washed in a solution identical to skinning solution but without Triton X-100 and stored frozen at ⫺80°C until analyzed. Composition and Preparation of the Gels All gels were run in vertical slab gel units (Hoefer SE 600, Pharmacia) using ECPS 3000/150 power supplies (Pharmacia). Two types of gels were used to achieve optimal separation of skeletal and cardiac MHC isoforms.

PULSE ELECTROPHORESIS OF MYOSIN HEAVY CHAINS

Skeletal myosins (modified from Talmadge and Roy (21). Stacking gels consisted of 4.0% (w/v) acrylamide with an acrylamide:N,N⬘-methylene-bisacrylamide (bis) ratio of 50:1, 70 mM Tris, pH 6.8, 0.4% (w/v) SDS, 30% (v/v) glycerol, 4.0 mM EDTA, 0.04% (w/v) ammonium persulfate (APS), and 0.36% (v/v) TEMED. Separating gels consisted of 9.0% (w/v) acrylamide with bis cross-linking of 1.5% (ratio 67:1), 200 mM Tris, pH 8.8, 0.4% (w/v) SDS, 30% (v/v) glycerol, 100 mM glycine, 0.03% APS, and 0.15% TEMED. Cardiac myosins (modified from Sweitzer and Moss (20). Stacking gels consisted of 3.7% acrylamide with a acrylamide:bis ratio of 20:1, 125 mM Tris, pH 6.8, 0.1% SDS, 10% glycerol, 0.03% APS, and 0.3% TEMED. The separating gels consisted of 12% acrylamide with an acrylamide:bis cross-linking of 200:1, 750 mM Tris, pH 9.3, 0.1% SDS, 10% glycerol, 0.025% APS, and 0.14% TEMED. All of the separating gels were cast the day before electrophoresis in order to allow overnight polymerization, whereas the stacking gels were allowed to polymerize for 1 h prior to sample loading. Composition of Buffers The electrode buffer used in this study was the same for both the upper and the lower reservoirs of the two types of gels used and consisted of 0.38 M glycine, 0.05 M (w/v) Tris, and 0.01% (w/v) SDS (21). In addition, the upper buffer was supplemented with 0.2 mM DTT immediately before the start of the electrophoresis (8). Pulse Unit and Running Conditions A custom-built electronic timer device (pulse unit) connected to a power supply for the gel system was used to phasically switch on and off the running current (Fig. 1). The pulse unit has an adjustable timer connected to an output flip/flop that drives two relays in parallel, one of which functioning as a safety circuit that senses ground-current leakage. The pulse unit only adjusts the duration of the on/off cycles, with the remaining conditions for SDS-PAGE set in the power supply as for standard electrophoresis. All of our protocols were performed under standardized current (13 mA) and temperature (10°C) with no additional restrictions in either voltage or watts. The effects of increased current (16 mA) were also investigated. The performance of this new system was examined by comparing the resolution of MHC isoforms in gels that were pulse-electrophoresed and in gels run at constant current (controls), under identical conditions with respect to duration of electrophoresis and length of the gels. The control samples were run at constant current (no pulse) for 7 h in 16-cm-long gels, for 12 h in 16-cm gels, or for 32 h in 32-cm gels. Identical samples were then pulse-electrophoresed in 16-cm-long gels for

231

24 h and in 32-cm gels for 64 h, using continuous on/off pulse cycles of 20 s each. Gels Staining All the gels were silver-stained using the method of Giulian et al. (9). RESULTS

The migrations of pig cardiac, rat cardiac, and rat skeletal MHC isoforms in 9% separating gels (first type of gel) under conditions of constant electrophoresis for 7 h and pulse electrophoresis for 24 and 64 h are shown in Fig. 2. Two observations can be made from the figure. The first is that pulse electrophoresis resolved MHC isoforms identical to those recognized by standard electrophoresis, i.e., one isoform in pig myocardium corresponding to the ␤-MHC, two MHC isoforms in rat myocardium designated ␣-MHC (predominant band) and ␤-MHC, and four MHC isoforms in rat skeletal muscle designated IIA, IIX/D, IIB, and I (in order from least to greatest migration mobility). The second observation is that the relative separation of the four skeletal and the two cardiac MHC isoforms increased by a factor of 2 with 24 h of pulsed current and by a factor of 4 when pulsed current was applied for 64 h in 32-cm-long gels. Increased spacing between MHC isoform bands was also observed in gels run at constant current for 32 h, indicating that both the duration of the electrophoresis and the length of the separating gel are important factors for the separating MHC isoforms (Fig. 3). However, as shown in Figs. 3B and 3C, longduration electrophoresis under constant current invariably resulted in profound artifacts such as severe blurring of bands and/or comingling of closely spaced bands, which increased when current was increased (Fig. 3C). These artifacts made it impossible to characterize most of the protein isoform bands in gels run at constant current, and therefore, control gels were run for just 7 h. Pulse electrophoresis of heart samples in gels optimized for cardiac MHC isoforms resolved additional bands that were not detected by standard electrophoresis. Rat heart samples (Fig. 4, lanes 1, 3 and 5) pulseelectrophoresed for 64 h yielded 3 MHC bands, whereas samples from pig myocardium yielded just one. While the relative migration of the fastest band from rat myocardium (␤-MHC) was comparable to the ␤-MHC isoform expressed in pig ventricular myocardium (lanes 2, 4, and 6), the two slower migrating and predominant bands appear to have originated from the single ␣-MHC band resolved in the remaining gels (Fig. 4). The electrophoretic migration of heart MHC isoforms presented in Fig. 4 also shows that the mobilities of the rat ␤-MHC and the pig ␤-MHC isoforms were similar but not identical.

232

SANT’ANA PEREIRA, GREASER, AND MOSS

FIG. 1.

Electronic circuit diagram of the pulse unit.

The relative mobilities of chicken skeletal (lanes 2 and 3), cardiac (lanes 4 and 5), and smooth muscle (lane 6) MHC isoforms under conditions of standard electrophoresis (constant current for 7 h, panel A), pulse electrophoresis for 24 h (panel B), and pulse electrophoresis for 64 h (panel C) are shown in Fig. 5. The results indicate that pulse-electrophoretic analysis of chicken heart yielded two bands in ventricular myocardium (lane 4) and three in atrial myocardium (lane 5), exceeding the number identified by standard elec-

trophoresis. The resolution of additional MHC isoforms by pulse electrophoresis was also observed in chicken skeletal muscle samples. The ALD muscle (lane 2) for example, which according to standard electrophoresis yielded 4 bands, may in fact contain as many as 6 or 7. Likewise, pulse electrophoresis of the PAT muscle (lane 3) yielded 5 distinct bands, 3 of which were also expressed in equal proportions in ALD muscle. Because rat skeletal muscle (4 MHC isoforms bands, Fig. 5, lane 1) and chicken gizzard muscle (two predomi-

FIG. 2. Electrophoretic migration of pig cardiac (left ventricle) (lanes 1, 4, and 7), rat cardiac (atria and ventricles) (lanes 2, 5, and 8), and rat skeletal (tibialis anterior) (lanes 3, 6, and 9) MHC isoforms in 9% SDS separating gels under standard conditions of current (13 mA) and temperature (10°C). (Left panel) Gel (16 cm long) was run at constant current (no pulse) for 7 h. (Middle panel) Gel (16 cm long) was pulse-electrophoresed for 24 h using on/off cycles of 20-s duration. (Right panel) Gel (32 cm long) was pulse-electrophoresed for 64 h using on/off cycles of 20-s duration. All the bands in the three gel panels are of the same magnification. Although the three systems resolved the same number of protein isoform bands, the degree of separation was increased when pulse electrophoresis was applied.

PULSE ELECTROPHORESIS OF MYOSIN HEAVY CHAINS

233

FIG. 4. Electrophoretic migrations of rat cardiac (atria and ventricles, lanes 1, 3, and 5) and pig cardiac (left ventricle, lanes 2, 4, and 6) MHC isoforms in 12% SDS separating gels under standard conditions of current (13 mA) and temperature (10°C). (Left panel) Lanes 1 and 2. Gel (16 cm long) was run at constant current (no pulse) for 7 h. (Middle panel) Gel (16 cm long) was pulse-electrophoresed for 24 h using an on/off cycles of 20 s. (Right panel) Gel (32 cm long) was pulse-electrophoresed for 64 h using an on/off cycle of 20 s. All the gel bands in the three panels are of the same magnification. The figure shows that application of pulse electrophoresis for 64 h in 32-cm-long gels resolves three MHC protein bands in rat myocardium, two of which appear to correspond to two distinct ␣-MHC species.

The comparative approach we used in this study indicated that greater effects were observed in long slab gels (32 cm) pulse-electrophoresed for 64 h. In

FIG. 3. Effects of the duration of electrophoresis and current in the resolution of MHC isoforms in both standard gels run at constant current and in gels pulse-electrophoresed. The three panels depict the migration of rat skeletal (TA, lanes 1, 4, and 7) and chicken (ALD, lanes 2, 5, and 8; PAT, lanes 3, 6, and 9) MHCs in 9% polyacrylamide separating gels cast in 32-cm-long slabs. (A) Gel was pulse-electrophoresed for 64 h using on/off cycles of 20-s duration, 13 mA current, and a temperature of 10°C. (B) Gel was run for 32 h at a constant current (no pulse) of 13 mA and temperature of 10°C. (C) Gel was run for 32 h at a constant current (no pulse) of 16 mA and temperature of 10°C. There was a marked decrease in artifacts in gels that were pulse-electrophoresed relative to those run at constant current.

nant bands, Fig. 5, lane 6) samples display the same number of bands with and without pulse electrophoresis, they were always loaded in the gels flanking the chicken muscle samples and used as controls.

FIG. 5. Electrophoretic migration of rat skeletal (TA, lane 1), chicken skeletal (ADL, lane 2; PAT, lane 3), chicken heart (ventricular, lane 4; atrial, lane 5), and chicken smooth muscle (gizzard, lane 6) MHC isoforms in 9% SDS separating gels under standard conditions of current (1 mA) and temperature (10°C). (A) Gel (16 cm long) was run at constant current (no pulse) for 7 h. (B) Gel (32 cm long) was pulse-electrophoresed for 64 h using an on/off cycle of 20 s. All the gel bands in the panels are of the same magnification. Thus, while pulse electrophoresis does not resolve additional protein bands in rat and chicken gizzard samples, it does show greater than expected polymorphism in both chicken skeletal and chicken cardiac samples.

234

SANT’ANA PEREIRA, GREASER, AND MOSS

order to understand the basis of such effects, we examined the relative mobilities of MHCs (apparent MW, 220 kDa), C-protein (apparent MW 150 kDa), and actin (apparent MW 45 kDa) referenced to the beginning of the separating gel and relative to space in gels run either with constant current or with pulsed current. Figure 6 highlights three important observations: (1) for the same running time, the migration of the MHCs relative to the beginning of the separating gel was greater in gels run at constant current than in pulsed gels; (2) the migration of C-protein relative to the beginning of the separating gels was greater in the gels run at constant current, although its migration relative to the MHC isoforms was comparable in both systems; (3) the migration of actin relative to MHC and C-protein was reduced in gels run at constant current compared to pulsed gels. This indicates that the relative migration of proteins of high and low molecular weight is altered with pulse electrophoresis. In order to understand these phenomena we examined the voltage traces and the variation of voltage during the course of the electrophoresis in both systems. Our observations revealed that both systems are very similar in both the absolute voltage values (760 – 780 mV) and voltage fluctuations during the course of the electrophoresis. The only dissimilarity between the two systems was observed in the pulse electrophoresis voltage recordings at the beginning of each pulse cycle, which exhibited a voltage spike in excess of 1500 mV, lasting approximately 500 ms. DISCUSSION

In recent years several studies have reported improved electrophoretic resolution of MHC isoforms as a result of increased glycerol concentrations (3), different acrylamide:bisacrylamide cross-linking (3, 20, 21), increased ionic strength of the running buffers (21), gradient gels (2), and intermittent running conditions (7). In the present study we demonstrate that the application of current to SDS-slab gels in a repetitive, pulsatile manner can dramatically improve the resolution of MHC isoforms compared to existing protocols. The comparative approach we used has revealed that there are at least two improvements due to pulse electrophoresis under the conditions used in this study. The first is the reduction of gel artifacts, which is well illustrated in Fig. 3. This enabled us to take full advantage of the improvements in resolution that result from significantly increasing the duration of the electrophoresis in long separating slab gels and thereby unequivocally characterize the two MHC isoforms (␣ and ␤) of rat myocardium and the four MHCs of rat skeletal muscle, in particular, fast MHCs IIA and IIX/D (Fig. 2). Although the precise mechanism by which pulse electrophoresis reduces artifacts is not understood, some explanations can be inferred on the

basis of the results shown in Fig. 6. This figure shows that although the migration of MHCs was faster in gels run at constant current, the migration of actin relative to the MHCs was slower. This indicates that proteins of smaller molecular weight ran closer together in the gels run at constant current than in those that were pulse-electrophoresed, even though the duration of the electrophoresis and the application of voltage during the course of the electrophoresis were similar in both systems. This suggests that the application of current in a pulsatile manner slows the migration of all proteins, and probably because of the gel filtration characteristics of the polyacrylamide matrix, proteins of higher molecular weight are likely to be slowed to a greater extent and/or reaccelerated at a slower rate in every pulse cycle. Such differences in migration would be likely to originate from differences in temperature within the gels and, therefore, be directly related with the extent of the artifacts observed. This idea is supported by the observations of Fig. 3C, which shows that the artifacts became more marked when the current was increased from 12 to 16 mA. The second beneficial effect of pulse electrophoresis appears to be an increased resolution, which allowed us to detect bands that appear to correspond to previously unidentified protein isoforms. This is well illustrated in the samples of rat myocardium that yielded three bands in long-length pulsed gels (Fig. 4). Previous molecular biological studies have shown that the rat ␣-MHC gene generates two different protein isoforms by alternative splicing, corresponding respectively to 40 and 60% of the total ␣-MHC pool (16). Scanning of our gels indicate the proportional amounts of the two predominant bands to match precisely those published by Sindhwani et al. (16), thus supporting the notion that the two bands identified in our gels may correspond to the two ␣-MHC species. Moreover, the fact that samples from frozen material (i.e., not chemically digested myocytes) also displayed three bands (results not shown) and that porcine myocardium yielded only one band in all of the gel protocols strengthens our contention that the three bands are resolved due to the greater sensitivity of the system. Since splicing of the rat ␣-MHC gene generates two species by inclusion or exclusion of a single glutamine residue, our results could suggest that the pulse electrophoresis system may be sufficiently sensitive to resolve two isoforms differing in only a single amino acid. However, the results may also reflect other posttranscriptional events in rat heart yet unidentified at the gene level. Finally, it should be noted that the system clearly differentiates the homologous ␤-cardiac MHC isoforms of pig and rat, which on the basis of their full-length cDNAs are predicted to exhibit few regions of sequence divergence but exactly the same number of residues (Sant’Ana Pereira et al., in preparation).

PULSE ELECTROPHORESIS OF MYOSIN HEAVY CHAINS

FIG. 6. Protein electrophoretic profile of rat skeletal muscle samples (TA) in two 32-cm-long, 9% SDS separating gels, (1) pulse electrophoresed for 64 h and (2) run at a constant current for 32 h. In both gels, current was set at 13 mA and temperature at 10°C. While protein migration was faster for all the proteins in the slab gel ran at constant current, there was greater separation between proteins of lower and higher molecular weights in pulse-electrophoresed gels.

235

The improved performance due to pulse electrophoresis was also evident in the analysis of chicken MHCs. Even though the number of bands resolved by our standard gels (Fig. 5A) is comparable to the number reported previously in the literature (1), the number of bands resolved in the pulse gels was dramatically higher, especially in the chicken skeletal muscle samples (for comparison of ALD and PAT muscles, see (1). Analysis of chicken atrial myocardium and ventricular myocardium in pulsed gels revealed three and two bands, respectively. Such high MHC polymorphism in chicken myocardium is consistent with studies at the gene level which to present have recognized at least two atrial MHC isoform genes (11, 24) and two ventricular MHC genes (4, 25) in both developmental and adult chicken hearts as well as with immunological studies (6). Although we cannot state with absolute certainty that all the bands resolved by our gels correspond to distinct MHC isoforms, we must emphasize that we used rat and smooth muscle gizzard samples as controls (flanking the chicken samples), and both displayed precisely the same number of bands in all of the protocols used. Moreover, the fact that some of the extra bands recognized were present in more than one tissue sample (e.g., the three slowest migrating bands of PAT and ALD muscles, Fig. 5C, lanes 2 and 3), despite differing considerably in their relative amounts, rules out the possibility that the extra bands could have resulted from the application of pulsatile current to bands with greater protein content. These observations suggest that the increased number of bands in chicken muscle is a biological rather than a methodological phenomenon and so, the basis for the enhanced resolution of bands in our results must be found elsewhere. Clearly, there is high polymorphism in chicken MHC isoforms (13) presumably due to still poorly understood mechanisms of gene expression and regulation or to posttranscriptional events. The performance of pulse electrophoresis in detecting additional bands has surpassed our own expectations, and as yet we are unsure as to the precise mechanism involved. From our comparative analysis, we recognize that unlike standard electrophoresis, pulse electrophoresis is characterized by a very brief but significant increase in voltage at the beginning of each pulse cycle. This was the only quantifiable difference we observed between the two systems, and therefore, we expect it to be a major determinant of the increased performance of the system. Although future manufacture of new systems that would allow variations in duration and strength of this initial voltage spike could be used to test our interpretation, other factors may also play a role. For example, one possibility is that the on/off cycles, possibly through diffusion, may rearrange the orientation of some protein isoform species, from a more horizontal to a more vertical position or vice versa, speeding up or slowing down their migrations.

236

SANT’ANA PEREIRA, GREASER, AND MOSS

This would be particularly important in high-glycerolcontaining gels, like most of our gels, in which the localization of hydrophilic regions in the proteins may be of significance. In fact, we observed (results not shown) that altering the glycerol concentration in our protocols clearly changes the relative migration and resolution of skeletal and cardiac MHC isoforms. REFERENCES 1. Alway, S. E., Carson, J. A., and Roman, W. J. (1995). J. Muscle Res. Cell Motil. 16, 111–122. 2. Bar, A., and Pette, D. (1988). FEBS Lett. 235, 153–155. 3. Biral, D., Betto, R., Danieli-Betto, D., and Salviati, G. (1988). Biochem J. 250, 307–308. 4. Bisaha, J. G., and Bader, D. (1991). Dev. Biol. 148, 355–364. 5. Cope, J. T. V., Whisstock, J., Rayment, I., and Kendrick-Jones, J. (1996). Structure 4, 969 –987. 6. Evans, D., Miller, J. B., and Stockdale, F. E. (1988). Dev. Biol. 127, 376 –383. 7. Fauteck, S. P., and Kandarian, S. (1995). Am. J. Physiol. 268, C419 – 424. 8. Fritz, J. D., Swartz, D. R., and Greaser, M. L. (1989). Anal. Biochem. 208, 205–210. 9. Giulian, G. G., Moss, R. L., and Greaser, M. L. (1983). Anal. Biochem. 129, 277–287. 10. La Framboise, W. A., Daood, M. J., Guthrie, R. D., Schiaffino, S., Moretti, P., Brozanski, B., Ontell, M. P., Butler-Browne, G. S., Whalen, R. G., and Ontell, M. (1991). Dev. Biol. 144, 1–15.

11. Oana, S., Matsuoka, R., Nakajima, H., Hiratsuka, E, Furutani, Y., Takao, A., and Momma, K. (1995). Eur. J. Cell Biol. 67, 42– 49. 12. Rayment, I., Rypniewski, W. R., Schmidt-Base, K., Smith, R., Tomchick, D. R., Benning, M. M., Winkelmann, D. A., Wessenberg, G., and Holden, H. M. (1993). Science 261, 50 –58. 13. Robbins, J., Horan, T., Gulick, J., and Kropp, K. (1986). J. Biol. Chem. 14, 6606 – 6612. 14. Rovner, A. S., Freyzon, Y., and Trybus, K. M. (1997). J. Muscle Res. Cell Motil. 18, 103–110. 15. Schiaffino, S., and Reggiani, C. (1996). Physiol. Rev. 76, 371– 423. 16. Sindhwani, R., Ismail-Beigi, F., and Leinwand, L. (1994). J. Biol. Chem. 269, 3272–3276. 17. Strang, K. T., and Moss, R. L. (1995). Circ. Res. 77, 114 –20. 18. Spudich, J. (1994). Nature 372, 515–518. 19. Sweeney, H. L., Rosenfeld, S. S., Brown, F., Faust, L., Smith, J., Xing, J., Stein, L. A., and Sellers, J. (1998). J. Biol. Chem. 273, 6262– 6270. 20. Sweitzer, N. K., and Moss, R. L. (1993). Circ. Res. 73, 1150 –1162. 21. Talmadge, R., and Roy, R. (1993). J. Appl. Physiol. 75, 2337– 2340. 22. Uyeda, T. Q., Ruppel, K. M., and Spudich, J. A. (1994). Nature 368, 567–569. 23. Yoon, S. J., Seiler, S. H., Kucherlapatir, R., and Leinwand, L. A. (1992). Proc. Natl. Acad. Sci. USA 89, 12078 –12082. 24. Yutzey, K. E., Rhee, J. T., and Bader, D. (1994). Development 120, 871– 883. 25. Zhang, Y. Shafiq, S. A., and Bader, D. (1986). J. Cell Biol. 102, 1480 –1484.