digitoxin nor monensin inhibited Ca2+ uptake, indicat- ing that Ca2+ accumulation did not occur through a sodium/calcium exchange. Conditions for half-maxi-.
THEJOURNAL
OF BIOLOGICAL CHEMISTRY Vnl. 255, No. 13, Issue of July 10. pp. 6290-6298, 1980 Prrnled I n U S A .
ATP-energized Ca2+Pump in Isolated Transverse Tubulesof Skeletal Muscle* (Received for publication, July 5, 1979, and in revised form, March 21, 1980)
Neil R. Brand&$Anthony H. Caswell, and Jean-PierreBrunschwig From the Departmentof Pharmacology, University of Miami, Schoolof Medicine, Miami, Florida33101
A modified protocol for isolation of transverse tubules incorporated an extra stage of purification. The existence of an ATP-energized Ca2+pump in transverse tubules isolated from rabbit skeletal muscle has been demonstrated. Isolated transverse tubules had a CaATPase activity of 0.78 pmol/min-mg; this was 300% in excess of that activity attributable to sarcoplasmic reticulum contamination. The distribution of part of the CaATPase activity and ATP-energized Ca" uptake coincided with the distributionof transverse tubules in isopycnic sucrose gradients loaded with mechanically disrupted triad junctions. Transverse tubules accumulated over 70 nmol of Ca2+/mgof protein; this uptake was abolished by the Ca2' ionophore A23187. Neither digitoxin nor monensin inhibited Ca2+uptake, indicating that Ca2+accumulation did not occur through a sodium/calcium exchange. Conditions for half-maximal Ca2+ uptakewere 5 p~ free Ca2+and 10 PM ATP. The Ca2+pump of isolated transverse tubules was distinguished from the Caz+pump of sarcoplasmic reticulum and sarcolemma in that thetransverse tubule Caz+ pump: 1) was not enhanced by oxalate; 2) was not energized by acetyl phosphate, p-nitrophenyl phosphate, or 3-0-methylfluorescein phosphate; and 3) did not hydrolyze p-nitrophenyl phosphateor 3-0-methylfluorescein phosphate. Using Caz+-dependent 3-0methylfluorescein phosphatase as a marker forsarcoplasmic reticulum, the contamination of the transverse tubule preparation was calculated to be 6%.This agreed with a contamination level of 5% estimated by freezefracture electron microscopy.
pump exists in the transverse tubulein order toreplenish the Ca2' in this specialized region. Ca*'-stimdated ATPases havebeen reported toexist inthe plasma membranes of several cell types (3-5), including skeletal muscle (6-8). The presence of such an enzyme in the sarcolemma, however, does not necessarily imply that a Ca2+ pump exists in the highly specialized region of the transverse tubule.Earlypreparations of externalmembrane did not differentiatebetween transversetubuleand plasma membrane. Themethod of preparation inwhich a low speed precipitate was employed probably gave a preparation low in transverse tubules (8). We have reported previously finding CaATPase activityin our preparations of transverse tubules isolated from rabbit skeletalmuscle (9). Because of the method of isolation in which intact triad junctions are first prepared, these preparations have low or negligible contamination arising from the bulk sarcolemma. The level of sarcoplasmic reticulum contamination, however, is of considerableconcernsince the CaATPase activity in the isolated transverse tubule preparations could arise from a small amount of the ATPase-rich longitudinal reticulum. Thus, in earlier communications we were unable to state positively that the CaATPase in the transversetubule fractionwas anintegral protein of this organelle. Inthisreport, wewill establish that the Ca"-pumping activity found in the transverse tubule preparation is indeed a property of the transverse tubule. CaATPase activity COpurifies with markers for the transverse tubule and iswell in excess of the activity ascribable to thesarcoplasmic reticulum contamination of the preparation. Furthermore, Ca'+-pumping activity can be determined in transverse tubule vesicles Although the intracellular movementsof Ca2+ during mus- with little interference from the sarcoplasmic reticulum and cle contraction andrelaxation have been studied extensively, the transverse tubule pump can be distinguished biochemithe role that Ca2' flux across the external membrane plays in cally from thesarcoplasmic reticulum enzyme. excitation-contraction coupling is not well defined. One hyMETHODS pothesis states thata small amountof Ca2+ enteringfrom the from transverse tubule actsas a trigger to induce Ca2+ release Preparation of Organelles-Purified transverse tubules were pretheterminalcisternae of the sarcoplasmic reticulum ina pared by a modification of the method previously described by Lau cascade process (1). This CaZ+ must be extruded from thecell et al. (9). We introduced an intermediary purification stage which against, its concentration gradient during relaxation in order took advantage of the ability of transverse tubules and terminal to maintain a constant level of intracellular Ca'+. Because of cisternae to recombine when incubated in the presence of potassium cacodylate (10). The full method is as follows: of the restricted diffusion of extracellular fluid into the lumen Two rabbits(1.5 to 2 kg) were stunned by a blow to theback of the transverse tubule (2), it may be proposed that an activeCa'+ head and bled. The back muscles were exposed and each muscle was * This work was supported by research grants from the National Institutes of Health (I-ROl-AM 21601). T h e costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. +Postdoctoral fellow supported by theNationalInstitutes of Health Cardiovascular Pharmacology Training Grant T 32 HL 07188 and the American Heart Association (Florida affiliate) Postdoctoral Fellowship 7/79 RF2.
injected at multiple sites with a total of 10 ml of medium containing 120 mM NaCl, 4.8 mM KCI, 1.5 mMCaC12, 11 mM glucose, and 1 mM sodium phosphate, pH 7.4. For some preparations, ["Hlouabain (20 to 50 pCi) was introduced into the muscle in the above medium in order to track the fate of the transverse tubules. Aftera 30-min incubation at room temperature, themuscles were excised and placed in 10 volumes of cold 250 mM sucrose, 0.5 mM EDTA,pH 7.4. Homogenization was effected for 1%min (3 X 30-s periods with 30-s rests) in a Waring Blendor in the cold. The homogenate was centrifuged a t 10,OOO X g for 20 min to sediment large cell elements. The
6290
6291
ea2' Pump in T-tubules supernatant was passed through eight layers of cheesecloth toremove floating material and centrifuged at 125,000 x g for 60 min. The pellet was resuspended using a Teflon-glass homogenizer in 400 ml of 250 mM sucrose, 2 mM histidine, pH 7.3, and recentrifuged a t 125,000 X g for 60 rnin to wash the vesicles. The washed vesicles were resuspended in 60 ml of 250 mM sucrose, 2 mM histidine, pH7.3, and layered on a continuous sucrose gradient. The density gradient was formed by passing 1250 ml of65% (w/w) sucrose in 0.01% NaN:, into a constant volume mixing chamber containing 1200 ml of 12.5%sucrose (w/w) in 0.01% NaN I and pumped into a Sorvall TZ-28 zonal rotor. Centrifugationwas carried out for at least 5 h at 80,000 X g. Terminal cisternae/Triads (38 to 43% sucrose) and longitudinal reticulum (28 to 33% sucrose) bands were then collected and diluted with an equal volume of water. The vesicles were concentrated by centrifugation a t 125,000 X g for 45 min. Terminal cisternae/Triads were resuspended in 8 ml of 250 mM sucrose, 2 mM histidine, pH 7.3, and then passed through a French Press a t 8,000 p.s.i. The extrudate was mixed with an equal volume of 0.6 M potassium cacodylate and then centrifugedon continuous sucrose gradients.Gradients were formed by passing 35 ml of 65% (w/w) sucrose through 35 ml of 12.54 sucrose (w/w) in a constant volume mixing chamber, and 4 ml of the potassium cacodylate-treated vesicles were placed on each gradient. Centrifugation was carried out at 150,000 X g for 1 h in a Sorvall TV850 vertical rotor. Two bandsof vesicles obtained were the "light" terminal cisternae band at 32% sucrose and a distinctly sharp dense band at 40% sucrose. Material with the buoyant densityof transverse tubules in the 22 to 268 region of the gradient was absent. The vesicles banding a t 40% sucrose have beenshownpreviously to contair. terminal cisternae and transverse tubules reattached to the terminal cisternae. This band was collected, diluted with an equal volume of water, and centrifuged a t 125,000 X g for 45 min. The pellet was resuspended in 4 to 8 ml of 250 mM sucrose, 2 mM histidine, and was subjected to a second mechanical disruption in the French Press at 4,000 p.s.i. and separated on a 35-ml continuous sucrose density gradient formed as described above. The transverse tubule band was clearly separated from the terminal cisternae and the amount of light terminalcisternaebanding a t 32% sucrose had been dramatically reduced compared to the gradients of the unmodified method of Lau et al. (9). This preparation routinelyyielded 4 mgof transverse tubules from the back muscles of two rabbits. In initial preparations using swinging bucket rotors, the rejoining step added 1 day to the total isolation procedure. Our recent use of the Sorvall TZ28 reorientating zonal rotor and the Sorvall TV-850 vertical rotor has allowed the isolation to be completed within the same time span as theunmodified procedure of Lau et al. (9). Protein was estimated by the Folin method. In some experiments, the fateof the transverse tubules was tracked throughout the isolation protocol. The introduction of ['"Hlouabain into the intactmuscle and its employment for tracking transverse tubules has been described previously (9, 11). Crude sarcoplasmic reticulum vesicles for Ca"+ uptake studieswere isolated by a modification of the method of Palmer and Posey (12). One sacrospinalis muscle was immediately excised from a 3%pound rabbit, homogenized in a Waring Blendor (3 X 30 s with 30-s rest intervals) in cold, 250 mM sucrose, 2 mM histidine, and then centrifuged at 10,OOO X g for 20 min. The supernatant was filtered through eight lagers of cheesecloth and centrifuged for 80 min at 30,OOO x g. The microsome pellet was washed once and resuspended in a low volume of the incubation medium. The sarcoplasmic reticulum vesicles were either used immediately or stored at -20°C for furtheruse. CaATPase Assays-ATPase activity was assayed in a basic medium consisting of 100 mM KC1, 10 mM histidine, and 5 mM MgCl, at pH 7.3 and temperature 37°C. Five to twenty micrograms of protein were added to 1 ml of medium and the reaction initiated by addition of Tris/ATI' to 2.5 mM. After 2 % min, the reaction was stopped by dilution of0.5-1111 aliquots into 2.5 ml of assay solution to give final concentrations of 0.69 M HPSOlr0.02 M NH, molybdate, and 0.025; Triton X-100. Inorganic phosphate was then assayed by the method of Eibl andLands(13). TheCaATPaseactivityrepresentedthe difference in the activity in the basicmedium plus 0.1 mM CaCI, compared with the basic medium plus 1 mM Tris/EGTA.'This concentration of Ca" was found to be optimal for the sarcoplasmic reticulum CaATPase. ~ _ _ . " _ _ _ _
' The abbreviations used are: EGTA, ethyleneglycol bis([j-aminoethyl ether) N.N,N',N'-tetraaceticacid; SDS, sodium dodecvl sulfate.
3-0-Methylfluorescein Phosphatase-Basal activity was assayed by addition of 5 to 50 pg of protein to a medium consisting of 50 mM irnidazole/HCl, pH 7.5.3 m~ MgClz, 1 mM EGTA, and 1.6 pM A23187 at temperature of 24OC. The reaction was initiated by addition of 30-methylfluorescein phosphate to 0.02 m ~ After . a constant slope had been obtained, Ca"-stimulatedhydrolysis was determined by addition of CaCIL to 0.96 mM (40 PM free Ca"). The fluorescence . standard was 3-0-methylfluorescein as described by Hill et ~ l (14). Reactions were carried out in the Perkin Elmer Fluorescence Spectrometer MPF-3Lat anexcitation wavelength of 470 nm andemission wavelength of 510 nm. Ca"-dependent 3-0-methylfluorescein phosphatase activity was calculated from the difference between the slope in the presence of Ca'+ and slope in the absence of added Ca'+. p-Nitrophenyl Phosphatase-p-Nitrophenyl phosphatase activity was estimated by a slight modificationof the procedureof Sulakhe et al. (15). Basal activitywas assayed in a medium consisting of 50 mM imidazole/HCl, pH 7.5, 3 m~ MgC12, 1 mM EGTA,and 5 mM pnitrophenyl phosphate. Ca"-stimulated activity was determined in the basal medium plus 0.96 mM CaC1.L (40 p~ free Ca") and 4.8 PM A23187. These conditions have been reported by Sulakhe and Sulakhe (16) to be optimal for the sarcoplasmic reticulum and sarcolemma CaATPases. Reactions were initiated by addition of 100 to 200 pgof protein to 1 ml of medium a t 37"C, and quenched after 10 min by addition of 0.5 ml of 12% trichloroaceticacid.Ca"-dependent pnitrophenyl phosphatase activity was calculated from the difference between the activity in thepresence of Ca2+ and thebasal activity. ATP-energized ea2' Uptake-ATP-energized Ca2+uptake was assayed by Millipore filtration. Theassay medium consisted of 50 mM KCI, 30 mM histidine, 5 m~ NaCI, 3 mM MgCls, and 3 pCi of ""CaCI,/ ml a t pH 6.8. Unless specified in the figure legends, the totalCa" was 0.1 mM and the free Ca" was adjusted to 20 p~ by addition of Tris/ EGTA. Free Ca2+ concentrations were calculated using the affinity constants recently reported by Kim and Padilla (17). These differ materially from earlier estimates since these authors employed for their estimation media which are closer to those employed in studies of Ca" movement in isolated organelles. Thirty to ninetymicrograms of protein was added to 1.5 ml of medium at 37"C, and the reaction was initiated by addition of Tris/ATP to 2.5 mM. The reaction was stopped by filtration of 0.5-ml aliquots through Millipore GSWP (0.22 p ) filters, and thefilters washedtwice with ice-cold medium containing no '%a'+. Radioactivity remaining on the filters was determined by scintillation counting. Filters retained less than 1 nmol/mg ofCa" when ATP was excluded from the reaction medium. The ATP-energized Ca" uptake of crude sarcoplasmic reticulum vesicles was assayed under slightly modified conditions. The MgCI, concentration was raised to 5.5 mM and sucrose was added to 100 mM final concentration. The reaction was initiated by addition of Tris/ ATP to 5 mM. PreliminaryexperimentsdemonstratedthatATPdependent Ca" uptake did not alter when the filters were washed, but more consistent results were obtained when the filters were not washed. By this procedure, about 5 mmol/mg of Ca" were retained on the fdters when ATP was excluded from the reaction mixture. Electron Microscopy-The preparation of freeze-fracture replicas of the transverse tubule band have been previously described (18). Fifteenfields of the replicasurfaceswerechosen atrandomfor determining the sarcoplasmic reticulum contamination of the preparation. The vesicle count was converted to contamination on milligram protein basis using the membrane area andlipid/protein values for sarcoplasmic reticulum and transverse tubules reportedby Lau et al. (18). SDS-Gel Electrophoresis-SDS-gel electrophoresis of the different organelles was carried out in the gel system as described by Laemmli (19) using 129 acr,ylamide gels. RESULTS
Purity of T r a n s v e r s e T u b u l e Preparation-We have estiby mated the purity of our transverse tubule preparation morphological techniques previously described by Lau et al. (18). Freeze-fracture replicas of the materialin the transverse tubule band were prepared and examined by freeze-fracture electron microscopy. The vesicles of 15 random fields were identified and counted. We estimated the number of sarcoplasmic reticulum vesicles to be 3.2% of the total vesicle population. This value was then converted to a protein mass using estimation of membrane area and lipid/protein ratios
6292
Ca 2+ Pump in T-tubules
for transverse tubules and sarcoplasmic reticulum. Contami- portion of the remaining band can be attributed to residual nation of our transverse tubule preparation was calculated tocontamination from longitudinal reticulum or light terminal cisternae. Calsequestrin is negligible in both transverse tubule be 5.05% on a milligram protein basis. This compared with preparations. The purified transversetubulepreparation the estimate of Lau et al. (18) of12.5%. Thus, the extra purification step has decreased the contamination by a factor shows an enrichmentin other proteinsincluding major bands of molecular weights 112,000,87,000,77,000,68,000 and oneof of approximately 2.5. all more intensein the The sodium dodecyl sulfate gel electrophoresis patterns of low molecular weight. These bands are the proteins of transverse tubules prepared by the two meth- more purified transverse tubule preparation. Some further of terminal cisternae/triads bands are present which are observed alsoin the sarcoplasmic ods as well as the protein patterns are identical proteins, and of sarcoplasmic reticulum subfractions are shown in Fig. reticulum. It is not clear whether these 1. The Laemrnli gel system was employedbecause of its although itis possible that theyreside at the junctionbetween resolving power. Track A shows the proteinsof longitudinal thetransversetubuleandtheterminalcisternaeandare reticulum in which the 102,000-dalton Ca” pump protein is extracted into either organelle on separation in the French predominant. A protein with the same migration pattern as press. The comparison of the patterns for the transverse tubules phosphorylase is present immediately below it. Tracks B, E, and F are of terminal cisternae/triads and light and heavy prepared by the twodifferentprotocolssuggests that our of Lau et al. (9) terminal cisternae, respectively. The Ca2+ pump and calse- preparation (Track D) is more pure than that questrin which runs in this gel systemwithanapparent (Track C). This contentionis supported by the fact that the molecular weight of 6 0 , 0 0 0 are clearly discerned. Other pro- band a t 6 8 , 0 0 0 daltons which appears to be specific to transa teins include ones a t 85,000, 32,000, and 27,000 daltons. The verse tubuleshas been enriched in ourpreparation.In heavy terminal ,cisternae are most highly enriched in calse- complementary fashion, several other bands have been apquestrin in accordwith theirhigh content of internal electron- parently decreased in staining intensity. The loss of protein dense material. Tracks C and D are of transverse tubules. bands, however, must be conservatively interpreted as such Track D is obtainedfromthemore purified preparation decreases could arise by solubilization of extrinsic proteins in employed in this paper. The main differences between the the potassium cacodylate treatment as well as by removal of proteins from the two preparations is the reductionin content contaminating organelles. Lau et al. (9) observed a decreased in Track D of 102,000-dalton protein which is found in the staining intensity for some protein bands when their transsarcoplasmic reticulum. T h e 102,000-dalton protein is seen to verse tubules were treated with KCI. Most noticeable is the be a minor proteih in the purified transverse tubules and a reduction of a band a t 102,000. This is the molecular weight of the Ca” pump of sarcoplasmic reticulum. The molecular A B C D E F weights of the ATPases of the transverse tubule have yet to be determined. CaATPase Activity-Table I shows the Ca”-stimulated ATPase activity of the transverse tubules and sarcoplasmic 1 reticulumsubfraction as obtained with our new isolation protocol. The control activity represents the ATPase detected in intact organelles in which ion gradients may be formed 2 while the maximal activity is obtained when permeability to 3 Ca” is enhanced by additions ofA23187 or Triton X-100. Considerable variation among the CaATPase activities of the sarcoplasmic reticulum subfractions is observed. These may 4 reflect the differing contents of Ca2+pumpprotein on a milligram of protein basis as described by Meissner (20) and Lau et al. (9). The heavy terminal cisternae exhibit a low of Cap’ CaATPase activityas expected from the small amount pump protein content.The enhancementof activity when the membranes are rendered leaky to Ca“ is most pronounced with the longitudinal reticulum. We will show evidence later FIG.1. Laemmli sodium dodecyl sulfate slabgel electropho- that the terminal cisternae in this preparation have an intrinmicrosomes. All organ- sic high permeability toCa2’. resis of organelles isolated from rabbit elleswereisolated from the same microsome preparation by the method of Lau et al. (9) or as described under “Methods” and concentrated by centrifugation a t 95,000 X g for 45 min. The pellets were resuspended in 250 mM sucrose, 2 mM histidine, and diluted a t least 5-fold toobtain a concentration of 1 mgof protein/ml in solubilizing medium (10% glycerol, 5% 2mercaptoethano1, 4% SDS, 0.0625 M Tris-HCI, pH 6.5, with bromphenol as tracking dye). After disruption of the vesicles by heating in boiling water for 3 min, 20 pg of protein/lane were run on a 128 polyacrylamide slab under conditions described by Laemmli (19). Longitudinal reticulum (Track A) and terminal cisternae/triads (Track B ) were prepared as described under “Methods.” The terminal cisternae/triad preparation was then divided: one portion was used to prepare transverse tubules by the method of Lau et al. (9) (Track0,the remainder wasused to prepare transverse tubules (TrackD ) , light terminal cisternae (TrackE ) , and heavy terminal cisternae (Track F) as described under “Methods.” The bars to the right on Track F indicate the mobility of protein standards in an adjacent lane. Phosphorylase A ( 1 ) . bovine serum albumin (2),human y-globulin subfragment (3).and ovalbumin ( 4 ) of molecular weights 94,WM. 68,OOO, 50,000, and 43,000 were employed.
TABLE I
ea2’ ATPase
activity in microsomal subfractions Control CaATPase activity was taken as the ATPase activity in the presence of0.1 mM CaClz minus the activity in 1 m~ ECTA. Maximal ATPase was determined in the presence of 0.03% Triton X100 or 1.9 p~ A23187. ATPase wasassayed as describedunder “Methods.” Each value represents the average of four different preparations. The basal (MgATPase) activityfor all fractions was 0.42 f 0.05 pmol/min. mg. CaATPaw Fraction ControlactivityMaximalactivity Longitudinal reticulum Light terminal cisternae Heavy terminal cisternae Transverse tubules
*
pmol/rnin.rng
2.47 0.39 2.13 f 0.33 0.86 f 0.15 0.78 f 0.30
3.95 f 1.04 2.42 f 0.45 1.02 0.34 0.70 0.40
* *
Ca
‘+Pump in T-tubules
The CaATPase activity of the transverse tubules was determined to be 0.78 pmol/min.mg protein. This activity is only slightly lower than the value of 0.92 pmol/min.mg protein reported by Lau et al. (9). However, the CaATPase activity of the transverse tubule is almost constant relative to the activity of the light terminal cisternae for the two preparative methods. The CaATPase activity is still manifest in the more purified transverse tubules, suggesting that theenzyme may be an intrinsic component of this organelle. The CaATPase in the transverse tubule is not enhanced by reagents that render the transverse tubule leaky. The explanation for this will be discussed later in the context of the kinetics of Ca2+accumulation in transverse tubules. The morphological data presented above show the sarcoplasmic reticulum contamination of the transverse tubule preparation to be 5.05%on a milligram protein basis. This particular preparation was also employed for one of the experiments described in Table I. From Table I, the highest CaATPase activity is that of the longitudinal reticulum with a value of 3.96 pmol/min mg. If all the sarcoplasmic reticulum contamination of the transverse tubules were longitudinal reticulum, then the CaATPase activity contributed from extraneous sources wouldbe 0.20 pmol/min.mg. Hence, the transverse tubule CaATPase activity of 0.78 pmol/min mg is 300% in excess of that maximally attributable to the sarcoplasmic reticulum contamination, which implies thatthe transverse tubules contain an intrinsic CaATPase activity. The location of the CaATPase on the fiialdensity gradient of the isolation protocol is reported in Fig. 2. As described under “Methods,” transverse tubules and terminal cisternae, recombined in potassium cacodylate, were subjected to mechanical disruption in the French Press at 4,000 p s i . and the extrudate fractionated on a continuous sucrose gradient. The distribution of protein, CaATPase, and [3H]ouabain as afunction of sucrose density were determined. The [3H]ouabainwas employed as a transverse tubule marker as described previously. Fig.1 shows that theouabain and hence the transverse tubules have an isopycnic point of around 24%sucrose. The bulk of the protein represents the terminal cisternae and is
-
a
6293
found in the dense regions of the gradient. As expected, most of the CaATPase activity is also found in the terminal cisternae region. There is, however, a distinct shoulder in the CaATPase distribution, the leading edge of which corresponds to thedistribution of the [3H]ouabain marker and the leading edge of the protein profile. Because of the high CaATPase activity in the light terminal cisternae, a well defined peak of enzymatic activity is not found and would not be expected in the transverse tubule region of the gradient. Despite the modification of the procedure of Lau et al. (9), the protein profile of Fig. 1 shows that some light terminal cisternae are still present in the gradient and partially mask the transverse tubular CaATPase activity. When light terminal cisternae alone were centrifuged to isopycnic equilibration, the resultant scan of CaATPase activity showed a nearly symmetrical steep distribution with a median buoyant density of 32% sucrose (data not shown) and little activity at the density of transverse tubules. Therefore, the CaATPase activity in the 24%region of the gradient appears to be intrinsic to vesicles of that buoyant density, and hence to the transverse tubules. ATP-energized Ca2+ Uptake-Fig. 3 shows CaZ+accumulation in a representative sucrose gradient loaded with mechanically disrupted transverse tubules-terminal Cisternae. Transverse tubules, as shown by the [3H]ouabain distribution have a median buoyant density of 26% sucrose while the bulk of the protein is found in the denser regions of the gradient with a peak at 38% sucrose corresponding to the terminal cisternae. In the absence of oxalate (solid line) the ATPenergized CaZ+uptake activity demonstrates a sharp peak at a median buoyant density of 24% sucrose and alsosome activity at the denser region of the gradient. Negligible Ca2+ uptake is detectable in the region of the light terminal cisternae around 32%sucrose. The band at 24%sucrose corresponds in shape and position to the ouabain marker for transverse tubules indicating Ca2+pumping activity by these organelles.
R sucrose w/w
FIG. 2 . CaATPase, [‘I-IIouabain binding, and protein profile from a sucrose gradient of the rejoined transverse tubuleterminalcisternaebandwhichwassubsequentlypassed through a French Press.Organelles were prepared and rejoinedas described under “Methods.” Ca’+-stimulatedATPase (upper panel) was calculated fromthe difference in ATPase activity assayed in the presence of 100 p~ CaClz comparedwith the presence of 1 m~ EGTA. The injection of and assay of [3H]ouabain (lower panel, solid line) has been described previously(9). Protein (Lowerpanel, dashed line) was estimated using the Folin reagent.
FIG. 3. ATP-energized Ca2+ uptake, [‘@ouabain binding, and protein profile from a sucrose gradient of the rejoined transverse tubule-terminal cisternae band which was subeequently passed through a French Press. ATP-energized Ca’+ uptake (upper panel, solid line) was assayed via Millipore filtration in the medium described under “Methods.” Total [Ca”] was 20 p~ and total [EGTA] was 10 p ~ Free . Ca’+was calculated as 8.6 p ~ . Oxalate-enhanced Ca’+uptake (upper panel, dashed line) was assayed in the samemedia to which 3 mM potassiumoxalate(final concentration) was added. The scale, right side (oxalate), is 10-fold smaller than the scale for the solid line. [3H]Ouabain binding (lower panel, solid line) and protein(dashed line)were assayed as described in Fig. 1.
6294
Ca2+Pump in T-tubules
The evidence of Figs. 2 and 3 indicates a high Ca2+ permea- either in free solution or is only weakly bound to sites on the bility in the light terminal cisternae; therefore, theproblem of interior of the transverse tubule membrane. Thus the Cadistinguishing transverse tubular activity of the Ca" pump ATPase of the transverse tubule is capable of moving Ca" from light terminal cisternae does not arise in the experiment across the vesicle membrane against the Ca" activity graof Fig. 3. The high correlation of the Ca"-pumping activity dient. with the distribution of transverse tubules in these sucrose Mechanisms of Ca" transport other than by a CaATPase gradients further supports the contention that a CaATPase is have been proposed. Ca2+might be carried by the Na' pump an intrinsic component of these organelles. directly (22) or might exchange with Na' as described in the When oxalate was added to the assay medium in order to squid axon (23). In these circumstances, ATP would be reprecipitate Ca2' within the vesicles, a different distribution of sponsible for generating the Na' gradient through the Na' Ca2+-pumping activity wasobserved (dashed line).Ca2' up- pump (24) and exchange of Ca" for Na+ subsequentlywould take occurred in regions of the gradient corresponding to the account for the Ca2' accumulation. Lau et al. (24) have distribution of terminal cisternae in the gradient, with a peak recently demonstrated that the Na' ionophore, monensin,and at 32% sucrose. As expected, thelightterminalcisternae the Na+ pump inhibitor, digitoxin, prevented formation of a showed higher activity on a protein basis than the denser Na' gradient across themembrane of isolated transverse vesicles, in accordance with higher specific activity of Ca- tubules. The cardiac glycoside, ouabain, does not inhibit Na' ATPase in the light terminal cisternae (Table I). The lack of uptake, since the transverse tubule membrane is impermeable Ca'+ uptake activity by the terminal cisternae in the absence to this drug and thebinding site is on theluminal surface of of oxalate is therefore not due to an uncoupling of ATPase the vesicle. Thus, if Cas+transport were coupled to Na' activity from Ca2+ transport but appears to indicate a higher movement, digitoxin and monensin would beexpected to intrinsic permeability of the vesicles to Ca". This may indi- inhibit Ca2' uptake by transverse tubuleswhile ouabain would cate eithera nonspecific leakiness of the terminal cisternae or exert no effect. that the Ca2' release channels of the sarcoplasmic reticulum Table I1 shows that Ca'+ uptake by transverse tubules is have been activated. not inhibited by digitoxin or monensin, but theremay even be In contradistinction to the 200-fold enhancement of oxalate a slight facilitation of uptake by these reagents. The signifiof Ca2' uptake by sarcoplasmic reticulum, Ca" uptake of the cance of these observationsis being further investigated. The vesicles banding at 26% sucrose only doubled from 28 nmol/ data presented hereshow that theCa2' taken up by transverse mg to 55 nmol/mg in the presence of oxalate. Sarcoplasmic by an exchange tubules is pumped via the CaATPase and not reticulum contamination could account for 15 nmol/mg of the of Ca" for Na'. increase,assuming 5% contamination and 300 nmol/mg of Neither 1.0 mM NaNs nor 0.1 mM dinitrophenol influenced uptake by the light terminal cisternae. The enhancement in Ca2' accumulation (datanotshown). Lau et al. (9) have uptake of Ca" by oxalate in transverse tubules is notsignifi- previously demonstrated thatmitochondrial contamination of cant and is in marked contrast to thatobserved in the sarco- transverse tubules is negligible. Fig. 4 shows the time course of ATP-energized Ca2' uptake plasmic reticulum.Sulakhe et al. (8) havereportedthat by transverse tubules in comparison with sarcoplasmic reticvesicles formed from skeletal muscle sarcolemma havea Ca" uptake capacity enhanced several hundred-fold by oxalate. ulum. Since the longitudinal reticulumand light terminal Therefore thelow ability of transverse tubules to accumulate cisternae in our preparation were partly leaky, an alternate calcium oxalate can be employedto distinguish them from the procedure for the isolation of sarcoplasmic reticulum vesicles was employed. The method of Palmer and Posey (12) was sarcoplasmic reticulum and plasma membrane. The ATP-energized Ca2+ uptake by the transverse tubule chosen since it is rapid, involves little handling of the tissue, vesicles could be the result of binding of Ca2' to the vesicle and does not involve washing the vesicles with hypertonic membrane or an actual transmembrane transport of the cat- KCl. In agreement with the results of others (25-27), our ion. These two possibilities could be distinguished by employ- crude sarcoplasmic reticulum preparation rapidlybound Ca'+ ment of the Ca2+ionophore, A23187. Caswell and Pressman with an initial velocity in excess of200 nmol/min. mg; full (21) have demonstrated that A23187 causes rapid equilibration of the Ca2' sequestered in the lumen of sarcoplasmic reticulum vesicles with the externalmedium and hence ATPenergized Cap+ uptakeis no longer supported. Table I1 shows "_"" I I that A23187 abolishes Ca" uptake by transverse tubules. This observation demonstrates that the Ca2+ within the vesicles is TABLE I1 Effect of drugs on eaz' uptake by transverse tubules ATP-energized Ca2+uptake was assayed by Millipore filtration as described under "Methods" except that in the digitoxin and ouabain experiments KHzP04was added to 1 m~ to facilitate drug binding prior to ATP addition. Free Ca2+was 20 p ~ Uptake . in the presence of the drug vehicle, 0.1% ethanol, was identical withthat of no addition. Each value represents the average of at least three preparations. Drug added
Ca" uptake nmol/mg protein (mean
No addition 1.9 p~ A23187 2 pM digitoxin 10 p~ ouabain 1 p~ monensin
59.4 k 5.6 0.33 k 0.33 71.2 '. 7.2 61.3 f 4.6 70.5 f 2.0
=tS.E.)
V
I
I
I
lo
I
mins
IS
FIG. 4. Time course of ATP-energized Caz+uptake by transverse tubules(T-tubule,solid line) and sarcoplasmic reticulum (SR, doshed line).ATP was added at time zero and Ca2+uptake was assayed by Millipore filtration as described under "Methods." Free Ca" was 10 p~ for transverse tubules and 20 PM for sarcoplasmic reticulum experiments.
Ca2+Pump in T-tubules saturation was approached more slowly. The transverse tubules, on the other hand, show an initial rate of uptake of 30 nmol/min.mg, and saturation was not approached until after 10 min of incubation. Although we may not have obtained optimal conditions for Ca2+ uptake by the transverse tubules, the data presented here indicate this thatenzyme hasa slower reaction velocity per mg of membrane protein than that of the sarcoplasmic reticulum. The slow rate atwhich the transverse tubules approach equilibriumwith respect to Ca2+ uptake may explain the absence of activation of the transverse tubule CaATPase byA23187. Fig. 5 shows pH dependence of the Ca2+ uptakeby transverse tubules. Maximal activity was found at pH 6.5. This compares with the pH optimum of 6.0 for Ca2+ uptakein both sarcoplasmic reticulum and sarcolemma(8).Fig. 6 shows the Ca2‘ dependence of Ca2+ uptakeby transverse tubules. In this experiment, the total Ca2+ concentration was set at 0.1 mM and varying amounts of EGTA was added tobuffer free Ca2+. Ca2+ activity was calculated using the Ca2+ EGTA affinity constant recently reported by Kim and PadiLla (17) which is 7-fold lower than those previously reported. It is appropriate to use this new constant since it was determined in media similar to thoseused for ATP-energized Ca’+ uptake studies. Half-maximal Ca2’ uptake occurs at approximately 5 PM free Ca2+ ion. Fig. 7 shows the ATP dependence of transverse tubule Ca2+ uptake. The curve is biphasic, with a steep increase in Ca2+uptake activity between 1 and 50 PM ATP and
I
J 5.0
I
1
I
6.0
7.0
1.0
6295
then a shallow increase to 2.5 mM ATP. The concentrationof ATP required for half-maximal activity, as estimated from the steep portion of the curve, was 10 w. Thus, the transverse tubule Caz+ pump appears to be qualitatively similar to the sarcoplasmic reticulum enzyme system in respect to pH optimumandrequirements for naturalsubstrateunderthese quasi-equilibrium conditions. Sulakhe et al. ( 8 ) have reported that some artificial high energy phosphatesources which supportCa2+uptake by sarcoplasmic reticulum areless effective in stimulating transport of Ca2+by a preparation of sarcolemma. Table I11 shows the influence of some phosphate donors on Ca2+ accumulation by transversetubulesand sarcoplasmicreticulum. Acetyl phosphate, p-nitrophenyl phosphate, and 3-0-methylfluorescein phosphate p a r t i d y support Ca” uptake by sarcoplasmic reticulum .vesicles as shown previously (8, 28, 29). The transverse tubules, however, sequester less than 1 nmol/mg when these compounds are substituted for ATP. Thus the ATP site of the Ca” pump of transverse tubules appear to be more specific for ATP than the equivalent site on sarcoplasmic the reticulum enzyme. Biochemical Estimation of Purity of Transverse TubulesConsidering the high specificity of the transverse tubule Ca2+ pump for ATP as the energysource, we investigated the possibility that chromogenic high energy phosphate donors could be employed to assay for sarcoplasmic reticulum contamination in our transverse tubule preparations. Table IV shows the basal and Ca2+-dependent rates of hydrolysis of 30-methylfluorescein phosphate and p-nitrophenyl phosphate by transverse tubules and sarcoplasmic reticulum subfractions. Transverse tubuleswere isolated byboth the methodof
PH FIG. 5. ATP-energized Ca“ uptake by transverse tubules as a function of pH. Ca2’ uptake assayed by Millipore filtration as described under “Methods.” Free Ca2+was 10 PM.
m y M ATP FIG. 7. ATP dependence of ATP-energized Caz+ uptake by transverse tubules. CaZ+uptake assayed by Millipore filtration as Io)
I
m M EGTA
10.0
3.0
1.0
I
I
1
(0.lOmM .M
IO
I
I
03 0 ,
8
described under “Methods.” Free Ca2+was 20 p ~ All . assays were conducted in the presence of 2.5 mM phosphocreatine and 250 p g of creatinekinase.
TABLE 111 High energyphosphate donors for Ca2+ uptake Ca2’ uptake was assayed by Millipore filtration as described under “Methods.” Reactions were initiated by addition of ATP to 2.5 mM (transverse tubules) or 5.0 m (sarcoplasmic reticulum), acetyl phosphate to 2.5 m,p-nitrophenyl phosphate to 2.5 mM, or 3-0-methylfluorescein phosphate to 0.1 m. Free CaZ’ was 20 p ~ . Ca2+ uptake
10”
lo-?
FIG. 6 . Ca2+dependence of ATP-energized Ca2+uptake by transverse tubules. Ca” uptake was assayed by Millipore filtration as described under “Methods.” All assays were conducted in the presence of 0.1 nm total Ca2+and varying amounts of EGTA.
Phosphate donor
Transverse tuSarcoplasmic reticulum bules
nmol/mg
ATP Acetyl phosphate p-Nitrophenyl phosphate 3-0-Methylfluorescein phosphate
72 0.5 0.2 0
125 86 24 19
6296
Ca2’ Pump in T-tubules
TABLEIV Rates of hydrolysis of chromogenic highenergyphosphates (1) Organelles prepared by method of Lau et al. (9). (2) Organelles from the same triad prepared by protocol described under “Methods” preparation. 3-0-Methylfluorescein phosphatase and p-nitrophenyl phosphatase assayed as described under “Methods.” Ca“-dependent activity was calculated fromthe activity in the presence of 20 PM free ea2+minus that in its absence. 3-0-Methylfluorescein phosphatase activity was calculated from the reaction slope. Each value for thepnitrophenylphosphataseactivityrepresents the mean of 6 to 12 determinations k S.E.
fracture electron microscopy. similarly, the 18% contamination estimate from the data presented in Table IV for the transverse tubules prepared by the protocol of t a u et al. (9) is not in discrepancy with the published estimate of 12.5% sarcoplasmic reticulum contamination (18). Thep-nitrophenyl phosphatase data presented in Table IV support the contention that the Ca2’-dependent hydrolysis of artificial high energy phosphates found in transverse tubule preparations arises from sarcoplasmic reticulum contamination. In the presence of the Ca2+ ionophore A23187, the p 3-0-Methylflunitrophenyl phosphatase activity of the longitudinal reticulum orescein phosp-Nitrophenyl phosphate is stimulated 2%-fold by Ca”, as compared to the 2.8-fold phate stimulation reported by Sulakhe and Sulakhe(16). We found Ca2+a 1.2-fold stimulation of p-nitrophenyl phosphatase by Ca2+ depenCa’*.de. for the transverse tubules prepared by our new protocol. This Microsomal fraction Basal dent Basal pendent compares witha reported &fold stimulation of the sarconmol/rnin. mg lemma p-nitrophenyl phosphatase by Ca2’. Thus, the highly Longitudinal reticu2.3 77.4 12.1 +- 0.6 17.7 +- 2.1 active Ca2’-dependent p-nitrophenyl phosphataseof the sarlum colemma is not found in the isolated transverse tubules. On Triads 0.94 35.4 5.5 2 0.7 10.1 f 1.3 the other hand, if all of the Ca”-dependent p-nitrophenyl Transverse tubules 0.86 13.7 7.5 f 1.0 5.2 f 2.0 phosphatase activity found in our transverse tubule prepara(1) Light tkrminal cister- 1.38.1 72.4 LO 16.5 3.4 tion arises from longitudinal reticulum contamination, thena nae (1) contamination level of 8% can be estimated from the datain Heavyterminal cis- 0.35 5.7 28.5 f 0.7 8.8 f 2.2 Table IV. This is in good agreement with the contamination ternae (1) estimates by the 3-0-methylfluorescein phosphatase activity Transverse tubules 1.7 7.64.7 f 0.6 1.4 1.3 and by the freeze-fracture electron microscopy considering (2) Light terminal cister- 1.7 the lack of precision in the p-nitrophenyl phosphatase assays. 62.6 7.9 f 1.2 14.4 5 3.9 nae (2) The data presented in Table IV indicate that the Ca”dependent 3-0-methylfluorescein phosphatase assay is the Lau et al. (9) andour new protocol from thesametriad preferred method of estimating the sarcoplasmic reticulum junction preparation. contamination in transverse tubule preparations. Ca2’ stimuCa2’-dependent 3-0-methylfluorescein phosphatase activ- lates thebasal activity of the longitudinal reticulum 30-fold as ity was detected in all microsomal fractions. As expected from compared to the2.8-fold Ca” stimulation of thep-nitrophenyl the CaATPase activities, the sarcoplasmic reticulum subfrac- phosphatase activity. The basal 3-0-methylfluorescein phostionsshoweddifferentactivities. The Ca2’-dependent 3-0- phatase activity of all microsomal fractions is considerably methylfluorescein phosphatase activity of the light terminal lower than the basal p-nitrophenyl phosphatase activity and cisternae, however, was almost equal to thatof the longitudi- the fluorescent assay requires only a tenth of the amount of nal reticulum while the maximal CaATPase activitywas only protein. Thegood agreement of the purity estimatesby the 360% of the longitudinal reticulum (Table I).The Ca”-depend- 0-methylfluorescein phosphatase assay and the freeze-fracent 3-0-methylfluorescein phosphatase activity of the heavy ture electron microscopy data suggests that the biochemical terminal cisternaeis lower than thatof the longitudinal retic- technique mayreplace themore laboriousmorphological ulum aswas found inthe CaATPase activity. Triad junctionsmethod. Furthermore, considering that a CaATPase is an showed 3-0-methylfluorescein phosphatase activity slightly intrinsic component of the isolated transverse tubules, it aphigher than the heavy terminal cisternae which agrees with pears that the Ca’+-stimulated 3-0-methylfluorescein phosthe fact that the triad junctions contain both the heavy and phatase may be the most reliable marker enzyme for sarcolight terminal cisternae. plasmic reticulum in microsomal preparations. Both transverse tubule preparations show a low level of 3DISCUSSION 0-methylfluorescein phosphataseactivity. The activity of the The datain this report demonstrate that ATP-energized an transverse tubules isolated by the method of Lau et al. (9) is about &fold higher than theorganelles prepared by our new Ca” pump is an intrinsic component of the isolated transverse protocoI. This correlates with the ZE-foId higher sarcoplasmic tubule membrane. Thisconclusion is supported by the followreticulum contentas estimated by the freeze-fracture electron ing observations: 1. The total CaATPase activity of purified transverse tubule microscopy. Otherexperiments(notshown)demonstrated that the activities of the transverse tubules and longitudinal preparations is 300% in excess of the maximum activity asreticulum had the same dependency upon free Ca’+ and the cribable to sarcoplasmic reticulum contamination. 2. CaATPase activity is associatedwith vesicles derived ratio of the two activities remained constant between 10 pM from skeletal muscle triads which band at 24% sucrosein and 250 PM free Ca”. Although we cannotruleoutthe isopycnic gradients. We have previously established by biopossibility that the transverse tubules have an intrinsic Ca2+chemical,pharmacological, and morphological techniques dependent 3-0-methylfluorescein phosphatase, it appears that all the Ca2’-dependent activity arises from sarcoplasmic retic- that isolated transverse tubules have a similar buoyant denulum contamination.The Ca2‘-dependent 3-0-methylfluores- sity, while the terminal cisternae vesicles are found at 32% cein phosphatase activity of the transverse tubules prepared and 41%sucrose. 3. The activity profiie of Ca2+ uptake in the absence of a by our protocol is about 6% of the longitudinal reticulum activity and7.5% ofthe activityof the light terminal cisternae,precipitated anion corresponds bothin position and in distriwhich are thetwo most likely sources of contamination of the bution to the [3H]ouabain markerfortransversetubules. transverse tubule band. This agrees well with a 5% contami- Furthermore, the light terminal cisternae whose distribution nation level as estimated ona separate preparation by freeze- in isopycnic sucrose gradients partiallyoverlaps the transverse
*
~
~
~~~~~
Ca 2+ Pump in T-tubules
6297
tubule distribution, do not show significantCaS+ uptake in the reticulum readilyhydrolyzes p-nitrophenyl phosphate. Although Sulakhe et al. (15) originally reported that the Ca" absence of oxalate. 4. CaZ+ uptakeby isolated transverse tubule vesicles is not stimulation of the Mg""-dependentp-nitrophenylphosphatase (16) have enhanced by the Ca"-precipitating agent, oxalate, in contra- of the sarcolemma was slight, Sulakhe and Sulakhe 8-fold Cai+ stimulationwhen the vesicles distinction to the enhancementby this anion of Ca2+ uptake recently reported an are incubated with 2 pg/ml of A23187. Under the same concapacity into sarcoplasmic reticulum and sarcolemma. 5. The artificial substrates p-nitrophenyl phosphate, acetyl ditions, we have been unable to detect significant Ca2+ stimphosphate, and 3-0-methylfluorescein phosphate do not sup- ulation of that enzyme (Table IV). Thus, the CaATPase of be incapable of accepting pport Ca2+ uptakeby transverse tubules yet serve as effective the transverse tubule appears to energy donors for the sarcoplasmic reticulum enzyme thus nitrophenyl phosphate asa substrate while the Ca2+ pumpof indicating a biochemical difference between the Ca2+ accu- plasma membrane is partiallyactivated by p-nitrophenyl mulation in transverse tubules and that in sarcoplasmic retic- phosphate. Two findings in this paper suggest that the Ca2+ pumpof ulum. protein from that of That theCa2+ uptakeby the transverse tubules appears to transverse tubules may be a different be a true transmembrane transportof Caz+via the CaATPase sarcoplasmic reticulum: 1)the qualitativelydifferent specificwas demonstrated by the use of ionophores and inhibitorsof ities for artificial high energy phosphate donors; 2) the low (Na-K)-ATPase. Ca2+ uptakewas abolished by the Ca" ion- extent of 102,000-dalton protein in the purified transverse ophore A23187, suggesting that the Ca2' taken up by the tubules. However, some cautions should be employed in indifferent responseto artificial phosphate vesicles remained as a free ion in the lumen of the vesicle as terpretation since the opposed to being tightly bound to the exterior of the mem- donors could occur through a different lipid or protein envibrane. Inhibition of formation of a Na+ gradient via the Na+ ronment. We are currently investigating this matter. Suitable markers which give a quantitative assay of the pump by digitoxin did not decrease Ca2' uptake. Similarly, dissipation of the Na' gradient by the Na' ionophore, monen- contamination of plasma membranein our transverse tubular sin, also did not decreaseCa2' uptake. Thesetwo observations preparation have yet to be devised. However, we now have a is a required substratesuggest that the number of indications that theplasma membrane content is and the fact that ATP Ca2' uptake by transverse tubulesis driven by the CaATPase very slight or negligible. 1. The transverse tubules are intially isolated from triad itself rather than via a n A T Pfacilitated Na+/Ca" exchange junctions which have a buoyant density of 40% sucrose while system. The demonstration of a Ca2+ pump in isolated transverse that of plasma membrane is less than 30% sucrose (31, 32). In tubules forces a reconsideration of the employment of Ca- our new three-stage purification schedule, we rejoin transverse of our protocol. This will ATPase as a marker for the sarcoplasmic reticulum. Estima- tubules to terminal cisternae as part separate anyvesicles which do notrejoin to terminalcisternae. tion of enhancement of Ca2+ uptake by Ca'+-precipitating 2. The ,&adrenergic receptorandadenylate cyclase are agents mayalso be an invalid assay forsarcoplasmic reticulum since Sulakhe et al. ( 8 ) have demonstrated that sarcolemma foundin our microsome preparation only at the isopycnic indication of a band with vesicles can trap Ca2+oxalate. We have demonstrated previ- point of the triad junctions with no ously that transverse tubulescould be distinguished morpho- an isopycnic point similar to plasma membrane indicating logically from sarcoplasmic reticulum in freeze-fracture repli- that most of the sarcolemma is precipitated in the intial low cas. This technique,however, is too laborious for routine use. speed differential centrifugation (33). The data in this reportsuggest that Ca"-stimulated hydrol3. The electron microscopic evidence presented previously ysis of chromogenic higher energy phosphates such as p- shows that many or most of the preparation contains small nitrophenyl phosphate or 3-0-methylfluorescein phosphate elongated vesicles which bear a physical resemblanceto intact may be readily used as a marker for sarcoplasmic reticulum. transverse tubules. Freeze-fracture analysis fails to reveal the tetragonal array of intercalated particles found on the faceof It is possible that this assay will be useful for distinguishing endoplasmic reticulum from external membrane in other mus- the intactplasma membrane. Also the densityof intercalated particles on theP face of plasma membrane ishigh (34) while cles and in non-muscle cells. In comparison of the Ca" pumpactivity of transverse the concave face of the transverse tubule vesicles which cortubules with that of plasma membrane we observe two areas responds to the P face has a very low content of intercalated of distinction. We findlittleornoenhancement of Ca2+ particles. accumulation by oxalate, whilein theplasmamembrane 4. The vesicles consist mainly of intact organelles with the preparation, Sulakhe et al. (8) have found little Ca" accu- cytoplasmic face oriented toward the outside(24). This is the mulation in the absenceof oxalate andconsiderable enhance- conformation of theintacttransversetubule in situ. The ment by the precipitating anion. This may reflect a biochem- plasma membrane would not be expected to form vesicles ical differencebetween thetransversetubuleandplasma with such a high asymmetry of orientation. The indications membrane. Alternatively, the plasma membranevesicles may are that many sarcolemma vesicles are not intact since they beleaky. If this is the case then oxalate may be able to do not effectively accumulate ions in the absence of trapping penetrate the interior of the vesicle and by precipitating the agents. ca2' prevent it from egressing from the leaky vesicle. Thus 5 . The transverse tubular preparation has a much higher the difference in the effect of oxalate on isolated transverse lipid content than that reported for plasma membrane (1.57 tubules and plasma membrane may reside largely in the high pmol of phosphorus/mg of protein uersus 0.60 in plasma integrity of the transverse tubulevesicle. The isolated trans- membrane). verse tubule systemwill afford a unique opportunity to study 6. Ca2+ uptake by isolated transverse tubules is not enthe Caz+ pump since these organelles are of uniform sidedness hanced by oxalate while that of plasma membraneis enhanced and retain their permeability barriers throughout isolation the up to 30-fold by this anion. procedure (24). 7. In contradistinction to the plasma membrane system, the The second area of distinction is in the employment of transverse tubule Ca2' uptake is not supported by p-nitroartikial phosphatedonors. As shown inTable I1 and reported phenyl phosphate. Ca2'-stimulated p-nitrophenyl phosphaby others (15, 16, 30), the Ca2+ pump of the sarcoplasmic tase is insignificant in transverse tubule preparations.
Ca '+ Pump in T-tubules
6298
The correlation between Ca"-pumping activity in isolated 8. Sulakhe, P. V., Drummond, G. I., and Ng, D. C. (1973) J. Biol. Chem. 248,4150-4157 of physiological experiments transverse tubules and the results 9. Lau, Y. H., Casell, A. H., and Brunschwig, J.-P. (1977) J. Biol. in which Ca2+ flux across the total external membrane is Chem. 252,5565-5574 described must be made with some caution. Efflux of Ca2+ 10. Caswell, A. H.,Lau, Y. H., Garcia, M., and Brunschwig, J.-P. from muscles exposed previously to 45Ca2+has been investi(1979) J. Biol. Chem. 254, 202-208 gated by several groups (35-39). Although there is a discrep- 11. Caswell, A. H., Lau, Y. H., and Brunschwig, J.-P. (1976) Arch. ancy concerning the number and time constantsof the comBiochem. Biophys. 176,417-430 ponents of such ion flux, it appears thata slow component of 12. Palmer, R. F., and Posey, V. A. (1967) J. Gen. Physiol. 50,20852095 Ca"+ efflux with a T in the range 165 to 300 min is stimulated 13. Eibl, H., and Lands, W. E. M. (1969) Anal. Biochem. 30, 51-57 by mechanical activity. Caputo and Balanos(39) have further 14. Hill, H. D., Summer, G. K., and Waters, M. D. (1968) Anal. dissected this slow component into three parts: a component Biochem. 24,9-17 dependent upon external Na+, a component dependent upon 15. Sulakhe, P. V., Drummond, G . I., and Ng, D. C. (1973) J.Biol. Chem. 248,4158-4162 external ea2+, and a residual efflux. Our experimental results have indicated that Ca'+ accumulation in the lumen of the 16. Sulakhe, S . J., and Sulakhe, P. V. (1979) Gen. Pharrnacol. 10, 103-113 transverse tubule is not decreased when the Na+ gradient is 17. Kim, Y. S., and Padilla, G. M. (1978) Anal. Biochem. 89,521-528 diminished. This argues againsta direct exchange of Na+ for 18. Lau, Y. H., Caswell, A. H., Brunschwig, J.-P., Baenvald, R. J., Ca"' representing a major portionof the Ca"' efflux. We have and Garcia, M. (1979) J . Biol. Chem. 254, 540-546 not yet determined theeffect of the luminal Ca2+ concentra- 19. Laemmli, U. K. (1970) Nature 227,680-685 tion on Ca'+ uptake. Our data suggest that such ATP-ener- 20. Meissner, G. (1975) Biochim. Biophys. Acta 389.51-68 gized Ca"' uptake maybe the majorif not thesole mechanism 21. Caswell, A. H., and Pressman, B. C. (1972) Biochem. Biophys. Res. Commun. 49, 292-298 of Ca2+efflux across the transverse tubule membrane. On the 22. Schwartz, A. (1976) Circ. Res. 39,40-45 other hand, the transverse tubule Ca" pump may be involved 23. Baker, P. F., Blaustein, M. P., Hodgkin, A. L., and Steinhardt, R. in one of the components of Ca2+efflux with a shorter T and A. (1967) J . Physwl. 192,43P serve to maintaina high Ca2+ concentrationwithin the lumen 24. Lau, Y. H., Caswell, A. H., Garcia, M., and Letellier, L.(1979) J . Gen. Physiol. 74,335-349 of the transverse tubule. The relative importance of the transverse tubule Cat' pump 25. Weber, A., H e n , R., and Reiss, I. (1966) Biochemistry 345, 329369 depends upon the role of Ca2+in excitation-contraction cou26. Inesi, G., and Scarpa, A. (1972) Biochemistry 11, 356-359 pling. Ca'+ efflux (40) and an inward Ca2+ current (41)have 27. Hasselbach, W., and Makinose, M. (1962) Biochem. Biophys. Res. been detected duringmuscle activity, although such events do Commun. 7, 132-136 not appear to be well correlated withthe action potential. Our 28. DeMeis, L. (1969) J. Biol. Chem. 244,3733-3739 experiments at this time neither support nor refute the Ca"+- 29. Inesi, G. (1971) Science 171,901-903 induced regenerative Ca2' release hypothesis. We have, how- 30. Rossi, B., deAssis Leone, F., Gache, G., and Lazdunski, M. (1979) J.Biol. Chem. 254,2302-2307 ever, demonstrated that Ca2+ canbe sequestered by the iso31. Kidwai, M., Radcliffe, M. A., Lee, E. Y., and Daniel, E. E. (1973) latedtransversetubuleand a Na+potentialcan also be Biochim. Biophys. Acta 298,593-607 imposed on this organelle (24). Such findings are the fwst 32. Barchi, R. L., Weigele, J. B., Chalikiam, D. M., and Murphy, L. steps toward developing a system to investigate the mechaE. (1979) Biochim. Biophys. Acta 550,59-76 nism of excitation-contraction coupling in the isolated organ- 33. Caswell, A. H., Baker, S . P., Boyd, H., Potter, L. T.,and Garcia, M. (1978) J.Biol. Chem. 253,3049-3054 elle system. REFERENCES 1. Ford, L. E., and Podalsky, R. J. (1970) Science 167,58-59 2. Barrett, J. N., and Barrett, E. F. (1978) Science 200, 1270-1272 3. DiPolo, R., and Beauge, L. (1979) Nature 278,271-273 4. Lee, K. S., and Shin, B. C. (1969) J.Gen. Physiol. 54, 712-729 5. Sulakhe, P. V., Leung, N. L.-K., and St. Louis, P. J . (1976) Can. J. Biochem. 54,438-445 6 . Peter, J. B. (1970) Biochem. Biophys. Res. Commun. 40, 13621367 7. McNamara, D. B., Sulakhe,P. V., and Dhalla,N. S. (1971) Biochem. J. 125, 525-530
34. Smith, D. S., Baenvald, R. J., and Hart, M. A. (1975) Tissue Cell 7,369-382 35. Winegard, S. (1968) J . Gen. Physiol. 51, 65-83 36. Shanes, A. M., and Bianchi, C. P. (1959) J. Gen.Physiol. 42, 1123-1137 37. Curtis, B. A. (1970) J. Gen. Physiol. 55,243-253 38. Kirby, A. C., Lindley, B. D., and Picken, J. R. (1975) J. Physiol. 254,37-52 39. Caputo, D., and Balanos, P. (1978) J. Membr. Biol. 41, 1-14 40. Bianchi, C. P., and Shanes,A. M. (1959) J.Gen. Physiol. 42.803815 41. Sanchez, J. A,, and Stefani, E. (1978) J. Physiol. 283, 197-209
