Purification of sarcoplasmic reticulum vesicles from horse gluteal muscle

Abstract

We have analyzed protein expression and enzyme activity of the sarcoplasmic reticulum Ca²⁺-transporting ATPase (SERCA) in horse gluteal muscle. Horses exhibit a high incidence of recurrent exertional rhabdomyolysis, with myosolic Ca²⁺ proposed, but yet to be established, as the underlying cause. To better assess Ca²⁺ regulatory mechanisms, we developed an improved protocol for isolating sarcoplasmic reticulum (SR) vesicles from horse skeletal muscle, based on mechanical homogenization and optimized parameters for differential centrifugation. Immunoblotting identified the peak subcellular fraction containing the SERCA1 protein (fast-twitch isoform). Gel analysis using the Stains-all dye demonstrated that calsequestrin (CASQ) and phospholipids are highly enriched in the SERCA-containing subcellular fraction isolated from horse gluteus. Immunoblotting also demonstrated that these horse SR vesicles show low content of glycogen phosphorylase (GP), which is likely an abundant contaminating protein of traditional horse SR preps. The maximal Ca²⁺-activated ATPase activity (V_max) of SERCA in horse SR vesicles isolated using this protocol is 5–25-fold greater than previously-reported SERCA activity in SR preps from horse skeletal muscle. We propose that this new protocol for isolating SR vesicles will be useful for determining enzymatic parameters of horse SERCA with high fidelity, plus assessing regulatory effect of SERCA peptide subunit(s) expressed in horse muscle.

GRAPHICAL ABSTRACT

1. Introduction

Calcium is the signal that controls muscle contraction [1]. Sarcoplasmic reticulum (SR)¹ is a dual-structure/function intracellular membrane system comprising (i) junctional SR where the ryanodine receptor Ca²⁺ release channel (RYR) initiates contraction, and (ii) longitudinal SR where the Ca²⁺-transporting ATPase (SERCA) induces relaxation [2, 3]. For muscle contraction, the bolus of released Ca²⁺ originates almost solely from SR through the RYR channel, and this released Ca²⁺ is subsequently re-sequestered in the SR lumen by SERCA [4]. The strength of muscle contraction is controlled by the amount of Ca²⁺ released from SR, which in turn is modulated by the Ca²⁺ load in the SR lumen via Ca²⁺ uptake activity by SERCA, an ion-motive ATPase that is regulated by transmembrane peptide subunit(s) such as sarcolipin (SLN) and phospholamban (PLB) [5, 6]. Calsequestrin (CASQ), which binds up to 80 Ca²⁺ ions (mol/mol), is the high-capacity Ca²⁺ storage protein localized to junctional SR [7, 8]. Per its luminal Ca²⁺ load, CASQ regulates RYR opening and closing [9, 10].

Microsomal membranes enriched in SR vesicles have been used since the 1960s to study SERCA Ca²⁺ transport and ATPase activities in vitro. These SR vesicles are purified from muscle homogenates using differential centrifugation, yielding sealed right-side-out SR vesicles, albeit with varying degrees of purity. Rabbit fast-twitch muscle is a standard model for purification of SR vesicles from skeletal muscle. early preparations of rabbit muscle SR vesicles were contaminated by significant amounts of glycogen granules, nucleic acids, and cellular debris (“glycogen-sarcovesicular fraction”) [11], yet further protocol development have yielded purified SR vesicles that have been well characterized and widely used [11–15]. Protocols for preparation of SR vesicles have also been developed for skeletal muscles of mouse [16], rat [17], squirrel [17], chicken [18], dog [19], pig [20], and cow [21], resulting in varied degrees of SR purity and SERCA activity; thus, SR vesicles from rabbit muscle have been the gold standard for use in biochemical, biophysical, and structural studies [22–24]. Purification of SR vesicles from horse skeletal muscle [25–29] and horse left ventricle [30] has been achieved, and these SR preparations have been used to assess the activity of horse SERCA and RYR. However, SR vesicles purified from horse skeletal muscle (using previously-reported protocols) show low SERCA activity compared to SR vesicles from rabbit muscle, which seems inconsistent with the high fast-twitch fiber composition of horse breeds. For example, horse myofibers exhibit intrinsic shortening velocities greater than that predicted from animals of comparable body size, thereby suggesting a cellular mechanism for the high performance of equine muscle.

Horses are highly susceptible to exertional rhabdomyolysis from a variety of causes, including glycogen storage disorders, malignant hyperthermia, and other proposed abnormalities in myosolic Ca²⁺ regulation [31]. Recurrent exertional rhabdomyolysis (RER) is one of the most common causes of poor performance and economic loss in racehorses [32, 33]. It has been proposed that the cellular etiology of RER in Thoroughbred racehorses involves aberrant SR Ca²⁺ release and excitation-contraction coupling [29, 31, 34, 35]. Mammalian adult skeletal muscles express predominantly the CASQ1 gene transcript, particularly fast-twitch myocytes (~100% CASQ1 protein) and also slow-twitch myocytes (≥75% CASQ1 protein) [36, 37]. We determined recently that gene transcription of CASQ1, the main protein isoform expressed in fast-twitch and slow-twitch skeletal muscle (i.e., cardiac CASQ2 gene and proteins expressed in heart), is downregulated in gluteal muscle of male and female horses diagnosed with RER, as compared to healthy male and female controls [38]. We determined also that gene transcription of CASQ1 is downregulated to a greater extent in the gluteal muscle of male horses with RER compared to female horses with RER [38].

To further resolve the role of myosolic Ca²⁺ regulation in horse exertional rhabdomyolysis, optimal SR membrane preparations are needed. The goal of the present study was to optimize SR membrane preps and examine Ca²⁺ transport regulation in horse SR. We hypothesized that horse, through natural selection as prey animals and subsequent selective breeding, have highly-adapted mechanisms for Ca²⁺ transport regulation that contribute to powerful muscular performance. Here we report a greatly-improved protocol for isolating SR vesicles from horse muscle, thereby allowing quantitation of SERCA expression and activity from gluteus medius, a primary locomotive muscle.

2. Results

2.1. A new fractionation protocol to purify SR vesicles from horse muscle

To investigate SERCA activity and regulation in horse gluteal muscle, we first reviewed previous reports on the isolation of SR vesicles from horse muscle [25–29], as summarized in Table 1. These studies prepared unfractionated SR vesicles composed of longitudinal and junctional SR vesicles, as purified using high-speed centrifugation pelleting (ranging typically from 38,000 to 165,000 × g_max), which was preceded by a clarification step of moderate-speed centrifugation (ranging typically from 8,000 to 12,000 × g_max) (Table 1). None of these previous studies have reported electrophoretic analysis of the protein profile of horse SR vesicles or demonstrated immunoblot analysis of SERCA protein content [25–29]. Instead, these previously studies relied on measurement of Ca²⁺-activated ATPase activity as an indicator of SERCA protein content in the horse SR vesicles, which exhibited only 2–28% the activity of rabbit SERCA in unfractionated SR vesicles purified from rabbit skeletal muscle (Table 1). Similarly, previously-reported preparations of horse SR vesicles showed low activity compared to unfractionated SR vesicles isolated from rat skeletal muscle, with horse SR vesicles showing apparent 12% activities (Ca²⁺-activated ATPase and Ca²⁺ transport) compared to SERCA in rat SR vesicles [27]. The mechanistic origin of low enzymatic activity of SERCA in horse SR has been unclear [27]. Horse gluteal muscle comprise 69 ± 8% fast-twitch myocytes [39]. In comparison, muscles from rabbit back and hind legs comprise 94–100% fast-twitch myocytes, for example, psoas, vastus lateralis, tibialis anterior, and extensor digitorum longus [40]. Horse gluteus, via its primary locomotive role, has been examined in equine exertional rhabdomyolysis [31, 41, 42]. We propose that studies of SERCA function and Ca²⁺ transport in SR vesicles from horse gluteus will provide key information on equine muscular function, and that these equine-based studies will also provide additional insights via a large-animal system relevant to human muscular health and disease.

Table 1. Reported Ca²⁺-activated ATPase activity by SERCA in SR vesicles purified from horse muscle.

Published results of horse SERCA activity are listed from studies utilizing unfractionated SR vesicles purified from horse muscle, as assessed using an ionophore-facilitated Ca²⁺-ATPase assay [25–29]. Our improved protocol for isolating horse SR vesicles from gluteal muscle (this study) provided a Ca²⁺-activated ATPase activity of 4.0 ± 0.4 IU at 37 °C (Fig.4). Prior to this study, the maximum reported Ca²⁺-ATPase activity of horse SR vesicles was 0.71‒0.73 IU at 37 °C [25, 29]. For comparison, the Ca ²⁺-ATPase activity of unfractionated SR vesicles purified from rabbit fast-twitch muscle is also listed (7.6 ± 0.5 IU at 37 °C), assayed under the same conditions utilized for horse SR vesicles in this study.

Horse SR vesicles reported protocol	Centrifugal g-force (×1K) a pre-clear SR collection		Ca2+-ATPase (IU) activity + A23187b
this studyc	4	10	4.0 ± 0.4
Am J Vet Res 2000d	2.6	140	0.71 ± 0.42
J Anim Sci 1998e	10	38	0.16 ± 0.01
Equine Vet J 1995f	12	165	0.16 ± 0.01
Equine Vet J 1995f	7	160	unsuccessful
Res Vet Sci 1993g	8	12	n.r.h
J Appl Physiol 1989i	10	38	0.73 ± 0.14
*rabbit SR: this studyj	12	23	7.6 ± 0.5

^aThe g-force of centrifugation spins (pre-clear and SR collection steps) are reported in relative centrifugal force (g-force × 1000).

^bA23187, a Ca²⁺ ionophore, was added to the ATPase assay in order to release accumulation of intravesicular Ca²⁺ and prevent product inhibition of SERCA.

^cSR vesicles from horse gluteal muscle (Fig. 2, Fig. 4) purified by the protocol outlined in Fig. 1.

^dWard et al., Am. J. Vet. Res. 2000 [29]

^eWilson et al., J. Anim. Sci. 1998 [28]

^fWilson et al., Equine. Vet. J. 1995 [27]

^gBeech et al., Res. Vet. Sci. 1993 [26]

^hn.r. = not reported.

ⁱByrd et al., J. Appl. Physiol. 1989 [25]

^jSR vesicles from rabbit fast-twitch skeletal muscle (Fig. 2) purified by the protocol of Ikemoto et al., J Biol Chem 1971 [14], with slight adaptation as described in Experimental Section 5.4 and Fig. S1.

Our present study aimed to develop an improved protocol for purifying a subcellular microsomal fraction enriched in SR vesicles [29] using muscle samples from the middle gluteus of four horses: three Quarter Horses and one Thoroughbred (see “ 5. Experimental Section”). Validation included key criteria such as a high content of SR-specific proteins (e.g., SERCA and CASQ), a high level of SERCA Ca²⁺-ATPase activity, a high level of phospholipid content, and a low content of contaminating proteins from non-SR membrane organelles and subcellular compartments (e.g., glycogen phosphorylase).

We utilized gel-based assays to examine subcellular fractions isolated from horse gluteal muscle, in the search for a microsomal membrane preparation enriched in longitudinal and junctional SR vesicles (i.e., crude unfractionated SR) [29]. In particular, we examined muscle fractions isolated at two common centrifugal speeds that yield microsomal membranes, albeit with a combination of vesicular types derived from organelles such as SR, ER, nucleus, and mitochondria, with the composition of muscle microsomes depending on a variety of factors, such as mammalian species, tissue type, and prep solutions. Here, we first isolated a horse muscle fraction using a previously-utilized centrifugation step at 100,000 × g_max, which was preceded by the centrifugation step at 10,000 × g_max, thereby yielding a pellet which we termed the 100K fraction (100KP) (see red box in Fig. 1). The 100KP pellet from horse muscle fractionation was colorless and translucent, indicating a centrifugal pellet with a high content of glycogen granules [43]. The yield of the 100KP fraction from 200 g of horse gluteus was ~50 mg protein, i.e., ~0.25 mg protein/g tissue, which may be considered relatively low (~25%) per standard prep of crude SR vesicles from muscle and heart. Secondly, we isolated a horse muscle fraction using a centrifugation step at 10,000 × g_max, which was preceded by a homogenate clarification step using centrifugation at 4,000 × g_max, thereby yielding a pellet which we termed the 10K fraction (10KP) (see green box in Fig. 1). The 10KP pellet from horse muscle fractionation was beige-colored, without the bright-red color of heme-enriched microsomes (e.g., liver ER vesicles with P450 system proteins or mitochondrial membranes with electron transport complexes), thus suggesting that the 10KP centrifugal pellet from horse skeletal muscle is enriched in SR vesicles, including additional vesicles from less-abundant systems such as sarcolemma, transverse tubules, and nuclei. The yield of the 10KP fraction from horse gluteal muscle was ~1 mg protein/g tissue, using a standard weight of 200 g (wet tissue) per preparation, i.e., muscle excised from horse middle gluteus.

Fig. 1: Improved protocol for purification of SR vesicles from horse muscle.

Five fractions from horse muscle were isolated and characterized biochemically, as identified by boxed text: homogenate 1 (H1), homogenate 2 (H2), 10,000 × g_max pellet (10KP), 100,000 × g_max pellet (100KP), and 100,000 × g_max supernatant (SOL). Intermediate fractions include supernatants 1–3 (blue text). NOTE: This is a 1-column fitting image.

For comparative biochemical analysis of SERCA activity in SR vesicles from two species, mechanical homogenization and differential centrifugation of rabbit muscle were used to purify a microsomal fraction enriched in longitudinal and junctional SR vesicles (unfractionated SR vesicles) from rabbit fast-twitch (white) muscle. The method of Ikemoto et al. (1971) was used to purify SR vesicles from pooled skeletal muscles from rabbit back and hind legs [14], with slight modifications (see Experimental Section 5.4 and Fig. S1²). Rabbit SR vesicles were collected using centrifugation at 23,000 × g_max, as preceded by a homogenate clarification step via centrifugation at 12,000 × g_max, thereby yielding a pellet which we termed here the 23KP fraction, i.e., unfractionated SR vesicles (see green box in Fig. S1A and the gel lanes labeled RAB/SR in Fig. 2, Fig. 3, and Fig. S1B). The 23KP pellet comprising SR vesicles from rabbit muscle was beige, as commonly-observed for a centrifugal pellet enriched in SR membranes. The yield of unfractionated SR vesicles from rabbit muscle was ~1 mg of SR protein per g of tissue (N = 5 preps). Rabbit SR vesicles isolated by this method are highly purified (Fig. 2, Fig. S1), plus well characterized and widely utilized [12, 22–24, 44, 45], with robust SERCA activity (Table 1).

Fig. 2: Immunoblot and gel-stain analyses of horse muscle fractions.

Five horse muscle fractions were assayed: H1, H2, 10KP, 100KP, and SOL. Samples were electrophoresed through 4–15% Laemmli gel for SERCA and GP blots, and 4–20% Laemmli gel for SLN blots. The protein load (μg) in each lane is indicated in white text. Protein molecular mass markers are indicated on the right. (A) SERCA immunoblot using mAb VE12₁G9 (primary) and goat anti-mouse-IgG pAb labeled with 800-nm fluorophore (secondary). SR vesicles isolated from rabbit skeletal muscle (RAB) were used as a quantitative positive control (see the prep flowchart for isolating rabbit SR vesicles, plus gel analyses of rabbit muscle fraction, in Fig. S1). (B) GP immunoblot using mAb ab88078 (primary) and goat anti-mouse-IgG pAb labeled with 680-nm fluorophore (secondary). Rabbit SR was also assayed (RAB). (C) SLN immunoblot using anti-horse-SLN pAb GS3379 (primary) and goat anti-rabbit-IgG pAb labeled with 800-nm fluorophore (secondary). Synthetic horse SLN was used as a positive control. Molecular species of SLN (monomer, dimer, and trimer) are indicated on the left. (D) Gel-stain analysis of horse muscle fractions. Samples were electrophoresed through 4–15% Laemmli gels and stained with Coomassie blue (left) or Stains-all (right). The amount of horse protein loaded per lane is 30 μg for homogenate 1 (H1) and homogenate 2 (H2), 20 μg for the 10,000 × g_max pellet (10K), 10 μg for the 100,000 × g_max pellet (100K), and 30 μg for the 100,000 × g_max supernatant (SOL). Rabbit SR vesicles (RAB) were used as positive control (10 μg protein load). Gel bands are identified by arrows to the right, along with the predominant muscle fraction for each. Protein molecular mass markers are indicated on the left. NOTE: This is a 2-column fitting image.

Fig. 3: Semi-quantitative immunoblot analysis of SERCA expression in horse muscle fractions.

A, SERCA immunoblots analyzed horse muscle fractions (H1, H2, 10KP, 100KP, SOL) from four preps, and SR vesicles from rabbit muscle were used as a quantitative standard. The protein load (μg) in each lane is listed. B, samples were electrophoresed through 4–15% Laemmli gels. Four horse preps were analyzed in duplicate on each blot (left, right), along with rabbit SR. Immunoblotting was performed using the mAb VE12₁G9 (primary) [46] and goat anti-mouse-IgG pAb labeled with a 800-nm fluorophore (secondary). Immunolabeling was quantified using the LI-COR laser scanner system in the near-infrared fluorescence mode. The intensity scaling of each fluorescence image is the same for all four immunoblots. The molecular mass of protein gel markers are indicated on the left. C, linear regression analysis (open circles: average ± SEM) of rabbit SR standard (0.025, 0.05, and 0.1 μg loads) on immunoblots 1–4. Semi-quantitative analysis (filled circle: duplicate average ± SEM) of SERCA content in the 10KP fraction (0.1 μg load) of horse preps 1–4. In some cases, the error bars (SEM) are smaller than the symbol (open or filled circles). NOTE: This is a 1.5-column fitting image.

2.2. Immunoblotting and Coomassie staining identify SERCA and glycogen phosphorylase as co-migrating gel-bands that are differentially enriched in the 10KP and 100KP fractions from horse muscle, respectively

SDS-PAGE and immunoblot analysis were used for initial characterization of five subcellular fractions from horse muscle: homogenate 1 (H1), homogenate 2 (H2), membrane pellet 10KP (4,000–10,000 × g_max), membrane pellet 100KP (10,000–100,000 × g_max), and the 100,000 × g_max supernatant (SOL) (Fig. 1, Fig. 2). Immunoblotting for SERCA detection utilized mouse monoclonal antibody (mAb) anti-SERCA1 VE12₁G9, which identified SERCA1 protein in the five fractions isolated from horse gluteal muscle, with the 10KP fraction showing many-fold greater SERCA1 content than the other four horse muscle fractions. Semi-quantitative immunoblotting using mAb VE12₁G9 (Fig. 3) determined that the level of SERCA1 protein in the 10KP fraction is enriched 10.5 ± 1.1 fold compared to the initial H1 homogenate of horse gluteal muscle, by defining the SERCA expression of the H1 fraction as 1.0 (Fig. 2A, Fig. 3). The 10KP fraction also showed 5.0 ± 0.7 fold greater level of SERCA1 protein than the 100KP fraction, and the 10KP fraction showed ≥ 10-fold level of SERCA1 protein than the H2 and SOL fractions (Fig. 4). Thus, immunoblot analysis of horse muscle fractions demonstrated high enrichment of SERCA1 protein in the 10KP fraction; a unique observation, since the majority of reported horse SR preps did not include this centrifugal cut, but instead collected SR vesicles that pelleted at ≥ 10,000 × g_max (Fig. 2, Table 1). SLN, a SERCA regulatory peptide commonly expressed in a myriad of mammalian fast-twitch muscles, was also detected by Western blotting (Fig. 2C). Thus, immunoblotting and Coomassie gels indicate that the 10KP fraction (4,000–10,000 × g_max cut) from horse muscle is an improved membrane vesicle preparation for functional and structural assays of SERCA.

Fig. 4: Comparison of Ca²⁺-activated ATPase activity and SERCA protein content in horse muscle fractions.

A, Immunoblotting was used to measure the relative SERCA content in horse muscle fractions and rabbit SR vesicles. The amount of SERCA protein in horse muscle fraction was calculated relative to the immunoreactivity of the H1 fraction, as determined per individual preparation with the initial H1 fraction set to an arbitrary unit (AU) of 1.0. Data are compiled from experiments performed as in Fig. 2A and Fig. 3, i.e., samples were assayed in a single-lane for relative comparison, with the per-sample protein loading amount adjusted to give a similar immunoreactive signal of SERCA/lane. B, Ca²⁺-activated ATPase activity was measured in horse muscle fractions and rabbit SR. The ATPase assay was run at 37 °C in the presence of the Ca²⁺ ionophore A23187, with saturating substrate concentration of Ca²⁺ and Mg•ATP (i.e., V_max assay). For comparison, the Ca²⁺-activated ATPase activity of rabbit SR vesicles is 7.6 ± 0.5 IU. NOTE: This is a 1-column fitting image.

Rabbit muscle expresses almost exclusively the SERCA1 isoform, with undetectable amounts of SERCA2 or SERCA3 proteins [45], while horse gluteal muscle is composed predominantly of fast-twitch fibers expressing the SERCA1 isoform [38, 41]. Thus, the collection of SERCA proteins (mostly SERCA1) expressed in horse and rabbit muscle will hereafter be referred to as “SERCA” per consensus.

Immunoblot analysis of glycogen phosphorylase (GP) in horse muscle was also performed (Fig. 2B), because GP is a common contaminant in SR preparations [11], and because of the apparent nature of the 100KP pellet, which was clear, glassy, and sticky, i.e., reminiscent of glycogen granules, instead of membrane vesicles. GP is a soluble protein that remains bound to glycogen granules during muscle homogenization and fractionation by differential centrifugation. Horse muscle is rich in glycogen, comprising 1.5–2.0% of wet weight muscle [41]. GP and SERCA co-migrate at 100 kDa on SDS-PAGE, thus leading to possible misidentification of GP as SERCA in muscle fractions, and/or not accounting for GP protein contribution at 100 kDa when performing Coomassie densitometry for SERCA in various muscle fractions. This problem was recognized early in the rabbit SR field, and protocols were developed to separate SR membranes from glycogen granules from an initial glycogen-sarcovesicular fraction [11]. Our Western blotting indicates that the 100KP fraction is highly enriched in GP, as compared to initial homogenates H1 and H2 of horse gluteal muscle (Fig. 2B). Thus, GP immunoblotting provides an additional indicator, along with the appearance and physical nature of the 100KP pellet, that glycogen granules are a considerable component of the 100KP fraction from horse gluteal muscle.

Horse muscle fractions obtained by differential centrifugation were also analysed using electrophoretic analysis and in-gel staining by Coomassie blue (Fig. 2D). Gluteus homogenates H1 and H2 express abundant Coomassie-stained bands that migrated at 200 kDa (myosin) and 40 kDa (actin), plus a moderately-abundant band at 100 kDa band, likely a mix of SERCA, GP, and other proteins. The 10KP fraction (4,000–10,000 × g_max pellet) and the 100KP fraction (10,000–100,000 × g_max) from horse muscle express abundant band(s) at ~100 kDa, which appear to be an unresolved doublet comprising one or two proteins. Although Coomassie blue is utilized frequently to identify protein bands by gel mobility, this dye cannot be utilized to identify protein bands that co-migrate on SDS-PAGE. On Laemmli-type gradient gels, SERCA runs at 100 kDa, as does GP, a common contaminant in SR vesicles [11]. Thus, traditionally-studied preparations of horse SR which are isolated at higher centrifugal forces (between 10,000 and 165,000 × g_max) (Table 1) are probably non-optimal for examining Ca²⁺-ATPase and Ca²⁺ transport activities due to enrichment of GP and glycogen granules, plus de-enrichment of SERCA protein (Fig. 2). Interestingly, the 10KP fraction of horse muscle expresses a Coomassie-stained band that migrates faster, i.e., lower molecular mass, than any Coomassie-stained band in rabbit SR vesicles (Fig. 2, Fig. S1). The identity of this low molecular mass protein (< 10 kDa) in horse 10KP is unknown. We propose that the 10KP fraction (4,000–10,000 × g_max cut) from horse muscle is an improved membrane vesicle preparation enriched in SR vesicles that will provide more effective assays of SERCA function, structure, and regulation.

2.3. Stains-all identifies calsequestrin and phospholipids as SR markers enriched in the 10KP fraction, but not the 100KP fraction

Additional electrophoretic analysis was performed to corroborate the identification of SR distribution in horse muscle fractions. CASQ is a high-capacity Ca²⁺-binding protein localized to junctional SR with RYR; the CASQ1 isoform is expressed in horse gluteus and rabbit muscles [38, 47]. A useful tool to identify CASQ and phospholipids on SDS-PAGE is the cationic carbocyanine dye Stains-all (Fig. 2D). Stains-all is a metachromatic dye (i.e., having an environmentally-sensitive absorbance) which stains most proteins pink or red, yet also provides unique staining of proteoglycans as purple, phospholipids as yellow, and highly-acidic proteins as blue (e.g., high-capacity Ca²⁺-binding proteins such as CASQ) [8, 47, 48]. Electrophoretic analysis with in-gel staining by Stains-all illustrates the protein profile of horse muscle fractions (Fig. 2). In H1 and H2, actin stains as an abundant pink band at 40 kDa. In the 10KP, there are high abundance band(s) staining pink at 90–100 kDa (SERCA and/or GP), two low abundance bands staining faint purple at >250 kDa in the 10KP fraction (possibly ‘CASQ-like’ proteins) [8], and one abundant band staining intensely-blue at 50 kDa, i.e., CASQ (Fig. 2D). SDS-PAGE with Stains-all detection (Fig. 2D) illustrated that the CASQ protein expressed in horse SR vesicles co-migrates with the CASQ1 protein expressed in rabbit SR vesicles (containing ~100% CASQ1 protein [37, 40]), with similar molecular mass via SDS-denatured electrophoresis, suggesting that CASQ in SR vesicles from horse skeletal muscle is predominantly the CASQ1 protein isoform. In the 100KP, there are high abundance band(s) staining pink at 85–95 kDa (GP and/or SERCA) and one low abundance band staining blue at 50 kDa (CASQ). The abundant CASQ band in the 10KP fraction serves as a subcellular marker for SR vesicles. Stains-all also detected the expression of CASQ at 55 kDa in fractions H1 and H2 from horse muscle, demonstrating dye sensitivity and specificity.

Phospholipids, which are stained yellow by Stains-all, migrate near the bromophenol blue dye-front on SDS-PAGE [47]. For rabbit muscle subcellular fractions, SDS-PAGE and the Stains-all dye demonstrated that phospholipids are most enriched in the P5 fraction (washed 23KP = rabbit SR vesicles) (Fig. S1, Fig. 2D). For horse muscle, Stains-all demonstrated that phospholipids are enriched in the 10KP fraction, serving as an additional indicator for the abundance of SR vesicles in this fraction (Fig. 2D). Stains-all also demonstrated that phospholipids are de-enriched in the 100KP fraction from horse muscle, indicating the deficit of SR vesicles in this fraction. Thus, Stains-all gel identification of CASQ and phospholipids in horse muscle fractions corroborates immunoblotting results demonstrating that SERCA protein expression is enriched in the 10KP fraction. Furthermore, Stains-all gel and Coomassie analyses are consistent with immunoblotting results that identified the 100 kDa band in the horse 100KP fraction is predominantly GP, instead of SERCA. Thus, the 10KP fraction from horse muscle homogenate shows the most enrichment of three distinct SR markers (SERCA, CASQ, and phospholipids), so we define the 10KP fraction isolated from horse gluteal muscle as “SR vesicles”.

2.4. Semi-quantitative immunoblotting of SERCA in horse muscle SR

We utilized quantitative immunoblotting and Coomassie densitometry to compare the relative amount of SERCA protein in SR vesicles from horse versus rabbit muscle. For standardization, the SERCA content of SR vesicles from rabbit fast-twitch (white) muscle has been determined to be 55‒70% of total protein (weight/weight), i.e., 5.0–6.4 nmol SERCA/mg rabbit SR protein [17, 44, 49, 50]. Immunoblotting with anti-SERCA1 mAb VE12₁G9 demonstrated that SR vesicles from horse gluteal muscle express ~35% the relative amount of SERCA1 as compared to rabbit SR (Fig. 2A, Fig. 3), although this is probably a slight underestimate of total SERCA content, since horse gluteal muscle also expresses a small amount of ATP2A2 (SERCA2) [38, 41]. Assuming that (i) the expression level of SERCA in rabbit SR is ~6 nmol mg SERCA/mg SR proteins, a common value, and (ii) horse SR contains ~35% the amount of SERCA protein as rabbit SR, then the density of SERCA expressed in horse SR is 2.1 nmol SERCA/mg SR protein. Thus, our improved protocol for isolating horse SR vesicles greatly enhances Ca²⁺ -activated ATPase activity as compared to prior reports of SERCA Ca²⁺-ATPase activity in horse muscle SR (Table 1). Our next steps are to (i) increase the purity of horse SR vesicles via further improvement in the isolation protocol, and (ii) purify addition SR preps from horse gluteal muscle in order to assess individual variations in SERCA activity and regulatory peptide expression.

2.5. Ca²⁺-activated ATPase activity by SERCA is enriched in the 10KP fraction, but not the 100KP fraction, from horse muscle

SERCA activity in horse muscle fractions was characterized by ATPase assay and compared to rabbit SR (Fig. 4). ATP hydrolysis was measured using a spectrophotometric linked-enzyme assay, whereby production of ADP was coupled stoichiometrically to oxidation of NADH, as detected by a decrease in absorbance at 340 nm. ATPase activity was measured at 37 °C with 5 mM Mg•ATP in the presence and absence of 100 μM Ca²⁺, the difference being SERCA activity (Ca²⁺-activated ATPase = total ATPase minus basal ATPase), to provide a readout of SERCA maximal enzyme velocity (V_max) with saturating concentration of substrates (Ca²⁺, Mg²⁺, and ATP). The Ca²⁺ ionophore A23187 was added to prevent back-inhibition of SERCA by high luminal Ca²⁺ concentration. ATPase assays were performed in the presence of 5 mM azide, which inhibits the ATP synthetase running in reverse mode (proton pumping ATPase activity) in disrupted mitochondrial fragments present in muscle homogenates and fractions. Horse 10KP had a Ca²⁺-activated ATPase activity of 4.0 ± 0.4 IU (V_max), as attributed to SERCA. Addition of the SERCA inhibitor thapsigargin (1 μM TG) eliminated Ca²⁺-activated ATPase activity in horse fractions H1, H2, 10KP, and 100KP, attributing the predominant contribution of SERCA to Ca²⁺-ATPase activity in these fractions. In comparison, rabbit SR had a Ca²⁺ activated ATPase activity of 7.6 ± 0.5 IU (V_max), as attributed to rabbit SERCA using the same assay conditions (i.e., rabbit SR vesicles have ~1.9-fold greater Ca²⁺-ATPase activity than horse SR vesicles in the 10KP fraction from horse gluteal muscle). The Ca²⁺-activated ATPase of the 100KP fraction from horse muscle reported here (1.4 ± 0.3 IU in Fig. 4B) is equal to or greater than previously-reported Ca²⁺-activated ATPase activity of SR vesicles isolated from horse skeletal muscle (Table 1), as identified by extensive online searches using Google and PubMed [25–29]. Thus, the 100KP fraction is adequate for assays of SR protein composition and enzyme activity. We conclude that the 10KP fraction of horse muscle exhibits robust levels of SERCA protein expression and Ca²⁺-activated ATPase activity.

3. Discussion

This study is a quantitative biochemical analysis of the expression of Ca²⁺ transport proteins in horse SR vesicles, in comparison to previously-reported horse studies and widely-utilized experimental models from rabbit skeletal muscles. Horse gluteal muscle is composed mostly of fast-twitch skeletal myocytes, similar to the back and hind-leg muscles of rabbit [39, 40]. Comparative assessment of SR vesicles isolated from horse and rabbit muscle included measurement of SERCA content and Ca²⁺-ATPase activity, thereby providing fundamental quantitative measurements of SERCA function at the molecular ensemble level. Our specific purpose was to optimize in vitro parameters used to assess SR Ca²⁺ transport regulation from horse, a species that suffers from a high incidence of exertional rhabdomyolysis potentially related to dysregulation of Ca²⁺ cycling.

To date, six protocols from five articles have been reported for isolating SR vesicles from horse skeletal muscle with measurement of steady-state SERCA activity (Table 1) [25–29]. The most recent report for isolation of SR vesicles from horse muscle utilized a 100KP membrane fraction, thereby providing successful analysis of RYR and Ca²⁺-ATPase activities in muscle membranes from control horses and horses susceptible to recurrent exertional rhabdomyolysis [29]. No differences in muscle membrane assays were identified between healthy and myopathic horses [29]. Previous reports for isolation of SR vesicles from horse skeletal muscle did not show electrophoretic analysis of muscle fractions using gel stains (e.g., Coomassie blue or Stains-all), nor immunoblotting for muscle marker proteins (e.g., SERCA, regulatory peptides, CASQ, or GP). Here we further characterized subcellular fractions from horse skeletal muscle, finding that the 10,000 × g_max pellet (10KP fraction) has 5–25-fold greater SERCA activity than the 100KP fraction previously utilized as SR vesicles (Table 1). With such low SERCA activity and probable high glycogen contamination in other reported SR preps (Fig. 2, Table 1), the SR preparations isolated by high-speed centrifugation may comprise only a small subfraction of SR vesicles from horse muscle and thus may not be representative of the SR system as a whole.

Here we report that the 10KP fraction from horse skeletal muscle (i) expresses 10-fold greater SERCA content than the gluteal muscle homogenate, and (ii) is enriched in SR markers such as CASQ and phospholipids. This 10KP fraction from horse muscle has enabled characterization of the Ca²⁺-activated ATPase activity by horse SERCA, with robust enzymatic level. Our results further demonstrate that isolation of SR vesicles requires species-specific and muscle-specific protocols; for example, in agreement with Sacchetto et al. reports that utilize SR vesicles successfully for (i) biochemical studies of SERCA in the 100KP fraction from rabbit fast-twitch skeletal muscle, (ii) structural studies of SERCA purified from the 150KP fraction of cow fast-twitch skeletal muscle, and (iii) enzymatic assay and electrophoretic analysis of SERCA2a in the 100KP vesicles from horse heart [15, 21, 30]. Per the purity of SR vesicles, crude preps isolated by differential centrifugation may also show enrichment of non-SR membrane components. Thus, protocol development with subcellular characterization is the key to enhancing the purity and activity of SR preps [15, 21, 30, 48, 51, 52].

Here we demonstrate that SR vesicles containing horse SERCA and CASQ are enriched in the muscle 10KP pellet, yet this fraction may also show enrichment in other membrane components, such as mitochondrial and sarcolemmal vesicles. The Stains-All dye suggested that SR vesicles from horse gluteal muscle express predominantly the CASQ1 protein isoform, via comparative SDS-PAGE analysis with SR vesicles from rabbit fast-twitch muscle. For future studies, we propose that our newly-reported protocol for isolating SR vesicles from horse gluteal muscle (i.e., the 10KP fraction) will provide a useful resource for further subfractionation of SR vesicles from the primary 10KP fraction, including the use of density gradient and/or rate-zonal centrifugation techniques to increase the purity of SR vesicles from horse gluteal muscle. We propose that our current protocol, plus future improvements, will enable more-precise quantitation of the expression level and correlated function of SR Ca²⁺ regulatory proteins, including SERCA [2, 53–55], peptide regulatory subunit(s) [56–59], and CASQ [8, 47, 60].

4. Summary

The reported protocol for isolation of horse SR vesicles is useful for assessing the protein content and Ca²⁺-ATPase activity of SERCA with high fidelity. We propose that this improved SR prep will facilitate biochemical and biophysical analyses of horse SERCA, along with measurement of the expression level and enzymatic effect of potential SERCA peptide regulator(s) [61]. Further biochemical analyses of these SR vesicles will probably provide key information on Ca²⁺ transport regulation in horse muscle, including new insights on proposed mechanisms for susceptibility to exertional rhabdomyolysis.

5. Experimental Section

5.1. Materials and Reagents

The Enzyme Commission (EC) number of SERCA is 7.2.2.10³. The EC of GP is 2.4.1.1. Two enzymes for the linked-enzyme ATPase assay – pyruvate kinase (EC 2.7.1.40) and L-lactate dehydrogenase (EC 1.1.1.27) – were purchased from Sigma-Aldrich Corp. Standard chemicals and Stains-all dye were purchased from Sigma-Aldrich. Laemmli-type SDS-PAGE gels were purchased from Bio-Rad Laboratories, Inc. The BCA Protein Assay Kit was purchased from Pierce Biotechnology, Inc. Primary and secondary antibodies for immunoblotting were purchased from specific companies, as described below.

5.2. Animals

The Research Animal Resources Facility at the University of Minnesota (U of MN) complies with the USDA Animal Welfare Act Regulations and the NIH Public Health Service Policy on Humane Care and Use of Laboratory Animals. U of MN has an approved Animal Welfare Assurance (A3456–01) on file with the NIH Office of Laboratory Animal Welfare. U of MN received accreditation renewal from the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) in November 2015. All animal research was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) at U of MN, with IACUC protocol # 1511–33199A for horses and IACUC protocol # 1611–34327A for rabbits. No experiments were performed on live animals.

Muscle samples were obtained from four horses (ages 10–18 yr) that were donated to U of MN for euthanasia due to orthopedic disease and were otherwise healthy: three castrated males (two Quarter Horses and one Thoroughbred) and one female (Quarter Horse). Horse owners provided written consent for obtaining samples for this research.

New Zealand white female rabbits (junior does less than six months of age) were provided by the Research Animal Resources Facility at U of MN. Standard protocol ensures that rabbits were housed for a minimal time. RAR staff performed euthanasia of rabbits with appropriate palliative care consistent with American Veterinary Medical Association guidelines. Rabbits were anesthetized prior to euthanasia to ensure minimal pain and distress; specifically, rabbits were given a subcutaneous injection of two sedatives: Acepromazine (1 mg/kg body weight) and Torbugesic (1 mg/kg body weight). After sedation, the rabbit’s ear was dabbed with wintergreen oil as a topical anesthetic, and a lethal dose of pentobarbital (100 mg/kg body weight) was injected through an ear vein catheter.

5.3. Purification of SR vesicles from horse muscle

Horses were humanely euthanized, and middle gluteal muscle was harvested, rinsed with ice-cold phosphate-buffered saline (PBS) at pH 7.4, and transported in chilled PBS from the veterinary clinic to the biochemistry laboratory within 1 h of euthanasia. Processing steps were performed in a 4 °C cold-room, with samples, tubes, and solutions on ice. Muscle tissue was trimmed of connective tissue, cut into 1 g chunks, and minced into small pieces. Minced muscle (200 g wet weight) was further disrupted using a 2-step Waring blender procedure: (i) blend 100 g tissue in 100 mL of 100 mM KCl, 20 mM MOPS (pH 7.0) for 15 s at 100% full-speed, and (ii) add 200 mL of 100 mM KCl, 20 mM MOPS (pH 7.0) and blend for 35 s at 100% full-speed. Next, the Waring blender procedure was repeated with another 100 g of tissue. The Waring-blended muscle homogenate samples were pooled, split into six SLA-1500 centrifuge bottles, and homogenized in each bottle using a Polytron homogenizer (Kinematica PT1300 with a 1.5-cm sawtooth generator) for 30 s at 50% of full-speed, yielding muscle fraction homogenate 1 (H1) (see horse SR protocol flow-chart in Fig. 1). The H1 fraction was spun at 4,000 × g_max for 20 min at 4 °C, yielding muscle fractions supernatant 1 (S1) and pellet 1 (4KP1). The 4KP1 fraction (6 pellets in 6 bottles) was re-homogenized as before (30 s at 50% of full-speed) in 300 mL solution of 100 mM KCl, 20 mM MOPS (pH 7.0) per bottle, yielding muscle homogenate 2 (H2). The H2 fraction was spun at 4,000 × g_max for 20 min at 4 °C, yielding muscle fractions supernatant 2 (S2) and pellet 2 (4KP2). The S1 and S2 fractions were pooled, filtered through 5 layers of cheesecloth, and spun at 10,000 × g_max for 20 min at 4 °C, yielding muscle fractions supernatant 3 (S3) and pellet 3 (10KP). The S3 fraction was spun at 100,000 × g_max for 30 min at 4 °C, yielding muscle fractions supernatant 4 (SOL) and pellet 4 (100KP). The 10KP and 100KP pellets were resuspended in a solution of 250 mM sucrose, 20 mM MOPS (pH 7.0) using a Potter-Elvehjem glass-teflon homogenizer. For further detail, we performed five steps to resuspend centrifugal pellets, i.e., release the subcellular muscle membrane fraction into solution: (i) scrape pelleted material from centrifuge tubes using a spatula, (ii) add pelleted material to resuspension solution (sucrose and MOPS) in a 15-mL polystyrene tube, (iii) pass pelleted material in solution through a 10-gauge needle and 10-mL syringe, i.e., cannulation, with 20 strokes, (iv) transfer membrane pellet solution into a 10-mL Potter-Elvehjem homogenizer, i.e., a glass mortar with Teflon pestle, and (v) further resuspend vesicles using twenty up/down strokes in the glass mortar, including constant hand-rotation of the pestle in circular fashion. Thus, we used standard biochemical methods for resuspending a centrifugal pellet, ala membrane microsomes [8, 48, 51, 52, 62]. The 10KP fraction was easily resuspended using this methodology, as assessed by visual inspection); however, the 100KP remained goopy, with macroscopic chunks, via visual inspection. Five horse muscle fractions were collected for biochemical analysis: H1, H2, 10KP, 100KP, and SOL (Fig. 1, Fig. 2, Fig. 3, Fig. 4). For storage, horse muscle fractions were aliquoted, snap-frozen in liquid nitrogen, and kept in a −80 °C freezer. The protein concentration of horse muscle fractions was determined using a colorometric BCA assay in microplate format with bovine serum albumin (BSA) as the standard; fractions were assayed in triplicate by n ≥ 2 assays.

5.4. Purification of SR vesicles from rabbit muscle

SR vesicles were isolated from rabbit fast-twitch skeletal muscle using mechanical homogenization and differential centrifugation [14] (see also protocol flow-chart for rabbit SR purification in Fig. S1A). White, fast-twitch muscles from the back and hind legs were harvested (~300–500 g), minced, and homogenized. A blender was used for homogenization (Waring Products, Inc.), with 100 g of tissue in 300 mL solution of 100 mM KCl and 50 mM MOPS, pH 7.0: 30 s at 10% of full-speed, 15 s off, and 20 s at 10% of low-speed. Rabbit muscle homogenate 1 (H1) was centrifuged at 4,000 × g_max, yielding muscle fractions supernatant (S1) and pellet 1 (4KP1). The 4KP1 fraction was homogenized as before, yielding homogenate 2 (H2). The H2 fraction was centrifuged as before, yielding muscle fractions supernatant 2 (S2) and pellet 2 (4KP2). Supernatants 1 and 2 (S1 + S2) were pooled, filtered through cheesecloth, and centrifuged at 11,750 × g_max for 30 min at 4 °C, yielding muscle fractions supernatant 3 (S3) and pellet 3 (12KP). To release myofibrillar and glycolytic proteins from rabbit SR vesicles in the S3 fraction, KC1 salt was slowly added to S3 to give a final concentration of 0.6 M, and then the S3 fraction was stirred at 4 °C for 15 min. The S3 fraction was then centrifuged at 23,400 × g_max for 60 min at 4 °C, yielding muscle fractions supernatant 4 (SOL) and pellet 4 (P4 = unwashed rabbit SR vesicles). The P4 fraction was resuspended in a wash solution of 250 mM sucrose, 20 mM MOPS (pH 7.0) and centrifuged at 55,300 × g_max, yielding pellet 5 (23KP = washed rabbit SR vesicles), which was resuspended in a solution of 250 mM sucrose and 20 mM MOPS (pH 7.0) using a Potter-Elvehjem glass-teflon homogenizer. For further detail, we performed five steps to resuspend centrifugal pellets, i.e., release the subcellular muscle membrane fraction into solution: (i) scrape pelleted material from centrifuge tubes using a spatula, (ii) add pelleted material to resuspension solution (sucrose and MOPS) in a 15-mL polystyrene tube, (iii) pass pelleted material in solution through a 10-gauge needle and 10-mL syringe, i.e., cannulation, with 20 strokes, (iv) transfer membrane pellet solution into a 10-mL Potter-Elvehjem homogenizer, i.e., a glass mortar with Teflon pestle, and (v) further resuspend vesicles using twenty up/down strokes in the glass mortar, including constant hand-rotation of the pestle in circular fashion. Thus, we used standard biochemical methods for resuspending a centrifugal pellet, ala membrane microsomes [8, 48, 51, 52, 62]. The washed 23KP fraction was easily resuspended using this methodology, as assessed by visual inspection. Five rabbit muscle fractions were collected for biochemical analysis: H1, H2, P3 (8KP), P5 (23KP = washed SR vesicles), and SOL (S4 = 23K supernatant) (Fig. S1). For electrophoretic analysis of rabbit muscle fraction, see Coomassie and Stains-all gels in Fig. S1B. For storage, rabbit muscle fractions were aliquoted, snap-frozen in liquid nitrogen, and kept at −80 °C. The protein concentration of rabbit SR vesicles was determined using a colorimetric BCA assay in microplate format with bovine serum albumin (BSA) as the standard; fractions were assayed in triplicate by n ≥ 2 assays.

5.5. SDS-PAGE and colorimetric gel-stain analysis

Muscle fractions were solubilized for 10 min at 23 °C in Laemmli sample buffer (69 mm Tris, pH 6.8, 11% glycerol, 1.1% lithium dodecyl sulfate, 144 mm β-mercaptoethanol, and 0.005% bromophenol blue), electrophoresed through Laemmli gels, and stained with the dye Coomassie blue R-250 or Stains-all dye. Coomassie-stained gels were imaged using a Bio-Rad ChemiDoc MP imaging system, and exposure time was optimized to prevent camera pixel saturation. Coomassie densitometry was performed using a Bio-Rad ChemiDoc MP imaging system. Stains-all was used for in-gel identification of CASQ (stained blue) and phospholipids (stained yellow) in a mix of total proteins (stained pink) [8, 47, 48]. Stains-all gels were imaged using a hand-held color camera.

5.6. Quantitative immunoblotting of muscle proteins

Immunoblotting was performed as previously described [58, 59, 63, 64]. Proteins were transferred to PVDF membrane (0.2-μm pore Immobilon-FL) in transfer solution of 25 mM Tris, 192 mM glycine, pH 8.3. Blots were blocked using Odyssey-TBS blocking buffer (LI-COR Biosciences, Inc.). Primary antibodies are described below. Primary antibodies were used typically at 1:1000 dilution, with incubation for 16–18 h at 4 °C. Secondary antibodies (fluorescently-labeled) were used at 1:15,000 dilution, with incubation for 20 min at 4 °C. After incubation with primary or secondary antibody, blots were washed three times with Tris-buffered saline (pH 7.4) with 0.05% Tween-20, and then once with Tris-buffered saline (pH 7.4). Sandwich-immunolabeling of the target protein with primary plus secondary antibodies was detected using an Odyssey laser scanner in near-infrared fluorescence mode, and bands were quantified using Image Studio Lite software (LI-COR Biosciences, Inc.).

Anti-SERCA1 antibody VE12₁G9 is a mouse mAb (IgG1-kappa isotype) developed by Dr. Kevin P. Campbell [46]. The immunogen for mAb VE12₁G9 was rabbit SR vesicles, and the epitope has been mapped to residues 506‒994 [46] of rabbit SERCA1a (GenBank ABW96358.1⁴ [65, 66]), a sequence that is 94% identical to horse SERCA1a (GenBank XM_001502262.6). mAb VE12₁G9 was purchased as Ascites fluid diluted in PBS (Abcam, P.L.C.). Abcam reports that mAb VE12₁G9 reacts on immunoblot with rabbit, mouse, human, dog, pig, and amphibian SERCA1. mAb VE12₁G9 has been validated for SERCA1 immunoblot using human vastus lateralis homogenate [67], human diaphragm homogenate [68], mouse gastrocnemius homogenate [69]. Here mAb VE12₁G9 was used for primary labeling at 1:1000 dilution and validated for horse SERCA1 immunoblot using horse gluteal homogenate (H1) and SR vesicles (100KP) (Fig. 2, Fig. 3).

Anti-GP ab88078 is a mouse mAb purified by protein-A chromatography (Abcam, P.L.C.). The immunogen was residues 734–843 of human muscle GP (UniProt reference sequence P11217), which is 95% identical to the amino acid sequence of horse muscle GP (GenBank NP_001138725.1) and 96% identical to rabbit GP (GenBank NP_001075653.1). Abcam reported that mAb ab88078 reacts with human and mouse muscle GP on immunoblot using mAb ab88078 at a concentration of 1 μg/mL. mAb ab88078 has been validated on GP immunoblot using human muscle homogenate [70], mouse muscle membrane vesicles [71], rat brain homogenate [72], and rat muscle homogenate [73]. Here mAb ab88078 was used for primary labeling at 1:1000 dilution (0.43 μg/mL), with successful detection of horse GP (Fig. 2B). mAb ab88078 has been discontinued by Abcam.

Secondary antibodies labeled with near-infrared fluorophore (700 nm or 800 nm emission) were purchased from LI-COR Biosciences, including affinity-purified goat antisera against mouse IgG (goat anti-mouse pAb) and affinity-purified goat antisera against rabbit IgG (goat anti-rabbit pAb). Secondary antibodies were used at 1:15,000 dilution (20 min at 4 °C) for sandwich-labeling of target protein through primary antibody binding. Fluorescent bands were detected using an Odyssey laser scanner in near-infrared mode and quantified using Image Studio Lite software (LI-COR Biosciences, Inc.).

5.7. Ionophore-facilitated Ca²⁺-activated ATPase assay

ATP hydrolysis was determined using a spectrophotometric assay of phosphate release. The assay was started in a 2 mL solution of 100 mM KCl, 0.1 mM EGTA, 5 mM MgCl₂, 5 mM Na₂ATP, 5 mM NaN₃, and 50 mM MOPS (pH 7.0) at 37 °C. The Ca ²⁺ ionophore A23187 was added (3 μg/mL = 5.7 μM) to eliminate the accumulation of Ca²⁺ inside SR vesicles, i.e., to relieve product inhibition [51]. ADP production was coupled to NADH oxidation by a linked-enzyme ATP-regenerating system: 10 U/mL pyruvate kinase, 0.6 mM phosphoenol pyruvate, 10 U/mL lactate dehydrogenase, and 0.2 mM NADH [74]. Assays were started with the addition of muscle fractions (5–20 μg protein/mL). After 1-min equilibration, Ca²⁺-independent ATPase activity was measured for 2–3 min, and then 100 μM CaCl ₂ was added (20 μL of 100 mM stock) to activate Ca ²⁺-dependent ATP hydrolysis by SERCA. The rate of ATP hydrolysis was calculated as the rate of NADH oxidation by measuring the decrease in NADH absorbance at 340 nm using an extinction coefficient of 6220 M⁻¹ cm⁻¹, as detected by a Spectramax 384 spectrophotometer (Molecular Dynamics, Inc.) in cuvette mode. SERCA activity is defined as the Ca²⁺-dependent ATPase activity (i.e., the difference in ATP hydrolysis rate measured in the presence and absence of Ca²⁺) and expressed in international units (1 IU = 1 μmol ADP/mg protein/min) [75].

5.8. Experimental design, statistical analysis, and data presentation

Biochemical assays were performed using independent SR vesicle preparations from N = 1–4 horses. Scientists were not blinded to sample identity during data acquisition or analysis. Data are reported as mean ± standard deviation (SD), which were calculated typically from n ≥ 2 independent experiments for each SR prep. Graphs were generated using OriginLab 9.2 (Northampton, MA).

HIGHLIGHTS.

• Devise a protocol to purify sarcoplasmic reticulum (SR) vesicles from horse muscle

• Analyze the new horse SR prep by gel stains, immunoblots, and enzymatic activity

• New horse SR prep shows greatly increased Ca²⁺-ATPase activity by SERCA Ca²⁺ pump

• New horse SR prep shows greatly decreased contamination by glycogen phosphorylase

• Rabbit muscle SR vesicles were used as comparative standard for biochemical analyses