THEJOURNAL OF BIOLOGICAL CHEMISTRY
Vol. 267, No. 1, Issue of January 5, p : 577-585,1992 rrnted m U.S.A.
Processing, Intracellular Transport, and Functional Expression of Endogenous and Exogenousa-@s Na,K-ATPase Complexes in Xenopus Oocytes* (Received for publication, July 15, 1991)
Philippe JauninS, Jean-Daniel HorisbergerS, Klaus Richter& Peter J. Goodq, Bernard C. RossierS, and KathiGeeringS 11 From the $Institute of Pharmacology and Toxicology, University of Lausanne, CH-1005 Luusanne, Switzerland, the $Genetics Department, University of Salzburg, A-5020 Salzburg, Austria, and the llhboratory of Molecular Genetics, Nationnl Institute of Child Health and Human Development, Bethesda, Mayland 20892
The minimal functional Na,K-ATPase unit is com- They show a tissue-specific distribution, but the functional posed of a catalytic a-subunit and a glycosylated B- differences still are poorly understood (for review, see Ref.4). subunit. So far three putative B-isoforms have been Putative isoforms also have been identified for the @-subunit described, but only B1-isoforrns have been identified (5-7). The pl-isoform is the most ubiquitous and is so far the clearly as part of a purified active enzyme complex. In only isoform for which it has clearly been shown that it is this study we provide evidence that a putative B3- part of a purified, active kidney enzyme complex (8).Nothing isoform might bethe functional component of Xenopus is known about the regulatory role of the @-subunit in the oocyte Na,K-ATPase. B3-isoformsare expressed in the catalytic cycle, and the fact that p-isoforms exist makes this oocyte plasma membrane together with a-subunits, butquestion even more puzzling. B3-isoforms are synthesized to a lesser extent than aWhile thedistinct role of @-subunitsas modulators of subunits. The unassembled oocyte a-subunits accumulate in an immature trypsin-sensitive form most likely transport activity or as cellular recognition signals remains to in the endoplasmic reticulum (ER).Injection of both be elucidated, we recently have proposed that @-subunitshave first of all a more general function; namely, they are needed fll- and Bs-cRNA into oocytes abolishes the transport for the structural and functional maturation of the enzyme constraint of the oocyte a-subunit, renders it trypsinresistant, and finally leads to an increased number of (for review, see Ref. 9). Assembly to the &-subunit indeed functional pumpsat theplasma membrane.In addition, leads to a stabilization (10) and an increased trypsin resistBs-isoforms as B1-isoforrns depend on the concomitant ance (11) of newly synthesized a-subunits which reflects a synthesis of a-subunits to be able to leave the ER and structural change in thecatalytic subunit likely to be needed to become fully glycosylated. Finally, a-B1 and a-B3 for the acquisition of functional properties(11) and/or its exit complexes expressed at the plasma membrane appear from the ER’ (10). as assessed by to have similar transport properties To learn more about the functional significance of p-isoouabain binding, rubidium uptake,and electrophysio- forms as part of active Na,K-ATPase, in this study, we have logical measurements in oocytes coexpressing exoge- investigated whether a putative @3-isoformrecently cloned nous al- and Dl- or B3-isoforms. Thus our data indicate from a cDNA library of Xenopus neurula stages (7) can that @3-isoformshave functional qualities similarB1to support functions similarto pl-isoforms with respect to their isoforms. They can assemble and impose a structural ability to associate with a-subunits and to promote the strucreorganization to newly synthesized a-subunits which tural maturation, theintracellular transport, and theexprespermits the exit from the ER and the expression of sion of functional Na,K-pumps at theplasma membrane. functional Na,K-pumps at the plasma membrane. As an experimentalsystem to express a’-,PI-, and p3cRNAs, we have used the Xenopus oocyte which, according to previous observations, displays several advantages for the Na,K-ATPase is an ubiquitous plasma membrane trans- proposed studies. We could show that, in contrast to other porter in animal cells which mainly contributes to cellular cellular systems studied, Xenopus oocytes do not synthesize (11).The overexpressed, ionic homeostasis (for review, see Refs. 1and 2). The minimal equal amounts of a- and @-subunits functional enzyme unit is a heterodimer composed of a large unassembled a-subunit is trypsin-sensitive and thus has charmultimembrane-spanning a-subunit and a smaller glycosy- acteristics of a structurally immature a-subunit population lated @-subunit.The a-subunitbears the functional domains (12). Significantly, the sole injection ofP1-cRNA renders oocyte a-subunits trypsin-resistant and leads to the expresof the enzyme such as cation, ATP, phosphate, andthe specific binding sites for cardiac glycosides. Three a-isoforms sion of more functional pumps at the plasma membrane as have been identified and cloned so far (forreview, see Ref. 3). assessed by ouabain binding (11). In this study, we have again exploited the unusual biosyn* This work was supported by the Swiss National Fund for Scien- thesis mode of Na,K-ATPase of Xenopus oocytes to study the tific Research Grant 31-26241-89. The costs of publication of this functional propertiesof p3-isoforms. For this purpose, we first article were defrayed in part by the payment of page charges. This have characterized in more detail the isoform composition of article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 11 TOwhom correspondence should be sent: Institut de PharmacolThe abbreviations used are: ER, endoplasmic reticulum; MBS, ogie et de Toxicologie, Universite de Lausanne, rue du Bugnon 27, modified Barth’s medium; SDS, sodium dodecylsulfate; PAGE, polyCH-1005 Lausanne, Switzerland. Tel.: 41-21-313-27-00 or 313-27-44. acrylamide gel electrophoresis; HEPES, 4-(2-hydroxyethyl)-l-piperFax: 41-21-313-27-75. azineethanesulfonic acid.
577
578
a-ps Na,K-ATPase Complexes
in Xenopus Oocytes
trichloroacetic acid and resolubilized with 1 N NaOH. Sucrose densities of fractions were determined by refractometry, using a Zeiss refractometer. Different cellular compartments were characterized by fractionation of control oocytes and by measuring the distribution in the various fractions of specific cellular markers? Plasma membraneenriched microsomes were recovered in fractions of sucrose densities around 1.10 according to therecovery of highest Na,K-ATPase activity and a concentration of labeled membrane fragments obtained from oocytes which were either exposed to [3H]ouabain or radioiodinated in an intact state. Golgi membrane-enriched microsomes were recovered infractions of densities around 1.12 according to the presence of highest galactosyltransferase activity and endoplasmic reticulum membranes in fractions of densities >1.12 according to concentration of an ER resident protein, the Bip protein (for review, see Ref. 18). MATERIALS ANDMETHODS Controlled Trypsinolysisof a-Subunits-Microsomes of labeled ooOocyte Injection withcRNA-StageV-VI oocytes were obtained cytes (see above) were subjected to controlled trypsinolysis after from Xenopus females (Noordhoek, Republic of South Africa) by freezing and thawing the samples twice in liquid nitrogen. Aliquots removal of ovary segments from anesthetized frogs as described (11). were incubated in the absence or presence of trypsin (type XI, Sigma) Oocyteswere defolliculated with collagenase (type IA, Sigma) in at a trypsin/protein ratio of 0.1:l for 3 min on ice. Trypsinolysis was modified Barth's medium (MBS, 13) without Ca'+. After overnight stopped with a &fold excess (w/w) of soybean trypsin inhibitor incubation in MBS at 19 "C, oocytes were injected with HzO, or a ] - , (Sigma) and themixture was left on ice for 10 min before addition of PI-, p3-, al-and &-,or al- and P3-cRNAsat different concentrations SDS a t a final concentrationof 3.7%. Samples were stored at -70 "C as indicated in the figure legends. cRNAs were obtained by in vitro before immunoprecipitation (see below). transcription of linearized templates with sP6 polymerase according Immunoprecipitation and Endoglycosiduse H Treatment-Immuto Melton et al. (14). For templates, cDNAs from a1 and cloned noprecipitations of Triton extracts, microsomes, and sucrose gradient from Xenopus kidney derived A, cells (15) or from p3 cloned from a fractionsdenatured with SDS(final concentration 3.7%), SDSXenopus neurula library (7) were recloned into the plasmid pSD5 PAGE, and fluorography were performed as previously described (19, which allows for synthesis of capped, full-length, poly(A+)RNA (16). 20). Quantification of the immunoprecipitated material was done by Metabolic Labeling of Oocytes and Preparation of Microsomes or laser densitometry using an LKB 2202 ultrascan densitometer. ImTriton Extracts-Noninjected or cRNA-injected oocytes were incu- munoprecipitation of the a-subunitwas done with a polyclonal antibated in MBS containing 2 mCi/ml of [35S]methionine(Amersham) serum prepared against the purified a-subunit from Bufo murinus for 4 h at 19 "C. In some instances, oocytes were washed three times (21) which cross-reacts with Xenopus a-subunits (11, 20). Immunowith MBS containing 10 mM cold methionine and further incubated precipitation of Xenopus @l-isoformswas performed with a polyclonal for a 21-72-h chase period in the same solution. antibody prepared against the purified @-subunitfrom Xenopus kidAfter the pulse or the pulse-chase period, oocytes were either ney (20). Finally, Xenopus 03-isoforms wereimmunoprecipitated with subjected to cellular fractionation on sucrose gradients (see below), an antibody prepared against a trp E fusion protein containing the or Triton extracts or microsomal fractions were prepared. For the extracellular domain of the C terminus of the Xenopus P3-subunit preparation of Triton extracts, oocytes were washed three times with (7). In preliminary experiments, we determined the concentration of MBS, and after removalof all liquid, 20 pl/oocyte of a solution anti &-serum needed for quantitative immunoprecipitation. In addicontaining 100 mM NaCl, 20mM Tris-HC1 (pH 7.6), 10 mM methio- tion, recent results indicate that anti-@3-serumdoes not cross-react nine, 1%(v/v) Triton, 1 mM phenylmethylsulfonyl fluoride, and 5 with pl-isoforms made in vitro from cRNA.' pg/ml (final concentration) of each leupeptin, pepstatin, and antipain In some instances, immunoprecipitated and @a-isoformseluted (Sigma) were added. After vortexing with intermittent cooling, the with sample buffer from protein A-Sepharose beads were precipitated extracts were centrifuged for 10 min a t 12,000 rpm at 4 "C in a Sigma with cold acetone for 1 h on ice and centrifuged at 10,000 rpm in a 202 MK centrifuge. After centrifugation, the supernatant was sepa- HB-4 rotor for 10 min at 0 "2. Acetone was removed and thesample rated from the pelleted yolk granules and stored a t -70 'C. dried in a Speed Vac concentrator (Savant). The samples were then For the preparation of microsomes, labeled oocytes were washed taken up in a solution containing 50 mM sodium acetate (pH 5.5), 1 three times with MBS and homogenized by 15 strokes in a glass mM phenylmethylsulfonyl fluoride, and 2 milliunits of endoglycosiTeflon homogenizer. Yolk granules were removed by centrifugating dase H (Calbiochem) and incubated overnight a t 37 "C. After reprethe homogenate twice at 2500 rpm in a HB-4rotor for 10 min a t 4 "C. cipitation and resolubilization in sample buffer, the samples were Floating lipids were removed with a cotton stick and the supernatant loaded on SDS-PAGE. recentrifuged at 42,000 rpm in a Ti-55 rotor for 90 min at 4 "C. The Ouabain Binding-Ouabain binding to oocytes was essentially done microsomal pellet (P) was taken up in6 pl/oocyte of a solution as described by Schmalzing et al. (22). Oocytes were loaded with containing 30 mM DL-histidine, 5 mM EDTA, 18 mM Tris-HC1 (pH sodium for 1 h at room temperature with potassium-free solution 1 7.4). (110 mM NaC1, 10 mM Tris-HC1 (pH 7.4)) before incubation with 0.4 Cell Fractionation of Xenopus Oocytes-Metabolically labeled oo- p~ [21,22-3H]ouabain(Amersham, specific activity, 18 Ci/mmol) in cytes were fractionated by density centrifugation on a 12-50% (w/v) solution 2 containing 90 mM NaC1, 1mM CaC12,1 mM MgC12,lO mM linear sucrose gradients according to a recently developed protocol.' Hepes (pH 7.4) for 20 min. Nonspecific ouabain binding was deterOocytes were homogenized by 15 strokes in a Teflon-glass homoge- mined in parallel experiments by including a 1000-fold excessof cold nizer in buffer A (10 pl/oocyte) containing 250 mM sucrose, 1 mM ouabain in the reaction mixture. Nonspecific binding amounted to 3EDTA, and 10 mM Tris-HC1 (pH 7.5), supplemented with 0.5 mM 7% of total binding. All experimental data shown represent total phenylmethylsulfonyl fluoride and 5 pg/ml (final concentration) of oubain binding. Oocytes were washed extensively after incubation, each antipain, pepstatin, andleupeptin. The homogenate was centri- individually transferred to Eppendorf tubes and solubilized with 100 fuged twice at 1000 X g for 10 min to remove yolk granules. The p1 of 5% SDS. Individual solubilized oocyteswere counted after supernatant was layered on a 9.8-ml linear sucrose gradient (12-50%) addition of 10 ml of Scintillator 299 (Packard). (w/v), in 1 mM EDTA and 10 mM Tris-HC1 (pH 7.5) and centrifuged 86RbFlux Meusurements--86Rb uptake into oocytes was measured in a SW40 Ti rotor for 4 h at 165,000 X g. Membrane fractions were after loading the oocytes with sodium for 1 h in solution 1. Oocytes collected from the bottom of the gradients as follows: one fraction of were then washed once with solution 2 supplemented with 5 mM 2.1 ml,12 fractions of 0.7 ml, and the top of the gradient. Microsomes BaC12to inhibit rubidium uptake throughK+ channels and incubated were collected from the 12 fractions of 0.7 ml after a 10-fold dilution 12 min a t room temperature insolution 2 containing5 mM BaCL and with buffer A by centrifugation in a Ti-50 rotor a t 228,000 X g for 3 5 mM 86RbC1(1mCi/mmol, =RbCl from Amersham, specific activity, h. Membrane pellets were taken up in buffer A minus sucrose (0.4 1-8 mCi/mg rubidium). Nonspecific 86Rbuptake was determined in pl/oocyte) and kept at -70 'C. Protein content (17) and radioactivity parallel experiments by including 1 mM ouabain in the reaction of each fraction were determined on aliquots precipitated with 10% mixture. Nonspecific ffiRb uptake amounted to 7-14% of total =Rb uptake. All experimental data shown represent total 86Rbuptake. The 'D. Pralong-Zamofing,Y. Qi-Han, G. Schmalzing, P. J. Good, and process was stopped by washing the oocytes twice in solution 2 complemented with 5 mM cold RbCI, and then individually centriK. Geering, manuscript in preparation.
Xenopus oocyte Na,K-ATPase, the stoichiometry of subunit synthesis, and thecellular localization of unassembled oocyte a-subunits. Our data indicate that thep3-isoform might itself be the functional component of Xenopus oocyte Na,K-ATPase expressed at theplasma membrane but that is it a limiting factor for the number of functional pumpsexpressed. Finally, we provide evidence that p3-isoforms as P,-isoforms can associate and impose a structural reorganization to newly synthesized &-subunitswhich permits the exit of the ER and the cell surface expression of functional a-p3and a-p, complexes with similar transport properties.
CY-& Na,K-ATPase Complexes in Xenopus Oocytes .-
Ln
a-subunit
fuged through 500 pl of an oil mixture made of 3 parts of dibutylphthalate and 2 part?, of dionylphthalate (23) for 5 s in an Eppendorf S 414 table centrifuge. The pelleted but intact oocytes were dissolved in 100 pl of 5% SDS and counted with 10 ml of scintillator in a Packard 8-counter. Na,K-pump Current Measurements-Active Na,K-ATPase a t the oocyte surface was estimated by measurement of the current generated by the Na,K-pump under VmaX conditions. Oocytes were loaded with sodium and kept in a potassium-free medium, similar to the conditions used for ouabain binding (see above). Using the twoelectrode voltage clamp technique, with a Dagan 8500 voltage-clamp apparatus (Dagan Corporation, Minneapolis, MN) the membrane potential was set a t -50 mV and the Na,K-pump current was measured as the outward current induced by the addition of 10 mM potassium, in the presence of 5 mM barium. We have shown that the potassium-activated current measured under these conditions is equal to the ouabain-sensitive current (24).
h
density a endo
-59,tg -42,Cg
83endo
579
RESULTS
Expression and Cellular Localization of Endogenous Na,KATPase Subunits in Xenopus Oocytes-In a previous study we have reported that Xenopus oocytes synthesize an excess of trypsin-sensitive a-subunits over &subunits. In addition, the P-subunit immunoprecipitated with a &-antibody from the oocyte plasma membrane did not have a molecular mass of 49 kDa as expected for a B1-subunit but migrated instead at 59 kDa on SDS-PAGE(11). To learn more about theisoform biogenesis of Na,K-ATPase in Xenopus oocytes, we applied in this study a cell fractionation protocol which permitted us to determine more precisely 1)the stoichiometry of Na,K-ATPase subunit synthesis and 2) the cellular localization of the unassembled a1-04 1.06 1.08 1.10 1.12 1.14 1.16 1.18 subunit. In addition, by checking the immunoreactivity of a density specific O3-antibody (see “Materials and Methods”), we tested kRa whether the oocyte 0-subunit component detected in a prea endo “88 vious study with a OI-antibody might be a p3-isoform. Fig. 1A shows the distribution of immunoprecipitated aand B-subunits in sucrose gradient fractions prepared from Xenopus oocytes labeled with [3sS]methionine for 4 h. As D3endo r) -59 visualized in thefluorograms, the anti-al-serumrecognizes as expected the 98-kDa a-subunit (a endo) which is concentrated FIG. 1. Biosynthesis and cellular distribution of a- and Bsisoforms of Xenopus oocyte Na,K-ATPase. A, unequal biosyn- mainly in fractions of high densities around 1.12-1.14. These thesis of a-and &isofomsin Xenopus oocytes. Two hundred oocytes fractions are likely to be enriched in ER microsomes since were injected with [3sS]methionine(50 nl, specific activity >lo00 Ci/ Bip, an ER resident protein, codistributes in these fractions mmol) and incubated for 4 h in MBS at 19 “C. Oocytes were then (Fig. 1A). washed, homogenized, and fractionated on sucrose gradients as deThe anti-f13-serumspecifically detects two polypeptides: on scribed under “Materials and Methods.” Aliquots of the 12 fractions the one hand, a 42-kDa polypeptide which is predominantly collected were immunoprecipitated with anti-a-serum oranti-& serum, and immunoprecipitates were subjected to SDS-PAGE and located in heavy fractions together with a-subunits and, on fluorography as described under “Materials and Methods.” Fluoro- the other hand, a 59-kDa polypeptide in lighter fractions grams (shown below line drawing) of a endogenous (a endo) and 183 which has a molecular mass similar to thepolypeptide previendogenous (83 endo) were exposed for 17 and 122 days, respectively, ously immunoprecipitated withan anti-pl-serurn from plasma quantitated by laser densitometry, and corrected for differences be- membranes of radioiodinated oocytes (11).The 59-kDa polytween a- and &-isoforms in exposure time, protein content, and the peptide detected by the anti-&serum is likely to be the fully number of methionines in al-and p3-isoforms (&/a1 = 1:3.25). The corrected values are expressed as total arbitrary units per fraction, glycosylated form which has passed a trans Golgi compartand the amount of core glycosylated 83-isoform in the fraction of ment while the 42-kDa polypeptide is the core glycosylated sucrose density of 1.13 was arbitrarily set to1. cg = core glycosylated; form residing in the ER (see below). fg = fully glycosylated;Bip = binding protein, an ERresident protein Significantly, quantification of the fluorograms and comwhich was revealed with an anti Bip-serum in a parallel experiment parison of the absolute amount of a- and B3-isoformsclearly on Western blots (see “MaterialandMethods”). B, a- and fully reveal that between 10 and 20 times fewer B3-isoformsthan glycosylated @3-isoformscodistribute in plasma membrane-enriched fractions, but an excess a-subunits over &isoforms accumulates in a-subunits aresynthesized duringa 4-h pulse (Fig. 1A). Since we recently provided experimental evidence that athe ER of Xenopus oocytes. One hundred and fifty oocytes were incubated in MBS containing 1.5 mCi/ml of [35S]methioninefor 21 subunits expressed in theoocyte from injected cRNA can only h at 19 “C. Oocytes were then washed, homogenized, and fractionated leave the ER if synthesized concomitantly with 8-subunits on sucrose gradients as described under “Materials and Methods.” (lo), we wondered whether the excess endogenous a-subunit Aliquots of the 12 fractions collected were immunoprecipitated with synthesized in the oocyte is subjected to the same transport n”“
-
anti-a- or anti-&-serum and subjected to SDS-PAGE and fluorographyas described under“Materials and Methods.” Fluorograms (shown below line drawing) of a endo and B3 endo were exposed for 3 and 28 days, respectively, quantitated by laser densitometry, and corrected for differences in a- and &-isoforms as in A. Shown is
the percent of a-subunits (0)and fully glycosylated &-isofoms (fg, 0)found in each fraction compared to thetotal amount recovered in the 12 fractions. Shown is 1 out of 3 similar experiments.
580
a-P3Na,K-ATPase
-
Complexes in Xenopus Oocytes
--
constraints andwould be retained in the ER. To answer this A. control plcRNA Injected question we subjected Xenopus oocytes to a 21-h pulse and 5: followed the distribution of a- and @-subunits in sucrose 4gradient fractions. As shown in Fig. lB, aftera 21-h pulse we 0 . find the &-subunit nearly exclusively in its fully glycosylated 2 3C I . 59-kDa form concentrated in lighter fractions of densities 2around 1.1. In parallel, part of the a-subunit which after a 43 . P h pulse was nearly entirely found in heavier ER fractions 3 1(Fig. lA),redistributes now to the same fractions. It can be 7 . t3 0 0 calculated that theratio of a and fully glycosylated /3-subunits 1.04 1.06 1.08 1.10 1.12 1.14 1.16 1.18 approaches 1 in these fractions (data not shown). Together density withourrecentobservation that fractionswithdensities around l.!. are enriched in plasma membranes (see “Materials and Methods”), our data suggest that within the 21-h pulse control period some newly synthesized a-63 complexes have left the ER, have been subjected to full glycosylation in a trans Golgi compartment, and have reached the cell surface. Interestingly, ” ” plcRNA however, after the 21-h pulse, a large proportion of the total a-subunit population remains concentrated in ER-enriched fractions nearly devoid of &subunits.From three similar experiments, it canbe roughly estimated that the total oocyte a-subunit pool is about 10 times larger than the 8-subunit pool, and it appears that the unassembled a-subunits are located in the ER. Mobilization of Oocyte a-Subunits fromthe ER by Exogenous Dl- and &Isoforms-ER retention of unassembled oocyte asubunits provided a convenient experimentalsituation to test the competence of pl- and &subunits to assemble to asubunits and topermit their exitfrom the ER. Fig. 2A again shows that after a 4-h pulse and a 21-h chase, the oocyte a1.04 1.06 1.08 1.10 1.12 1.14 1.16 1.18 subunit is mainly concentrated in ER-enriched fractionsand density littlein plasma membrane-enriched fractions. Ifhowever, oocytes are injected with &-cRNA just before the 4-h pulse control which permits the concomitant synthesisof oocyte a-subunits and exogenous P1-subunits,the oocyte a-subunits redistribute . .” . to a greatextent to the plasma-membrane fractions(Fig. 2A). Interestingly, the same phenomenon is observed when 8 3 cRNA instead of &-cRNA is injected into oocytes (Fig. 2B). Thus, these data provide a first argument that 0 3 - as B1FIG.2. Exogenous Dl- and Bs-isoforms can mobilize oocyte isoforms can associate to andabolish the transport constraint a-subunits from the ER. A, cellular redistribution of oocyte aof unassembled a-subunits. subunits after injection of oocytes with P1-cRNA. Oocytes were inInduction of the Structural Maturation of a-Subunits by pl- jected with 50 nl of H20(0,control) or 50 nl of H20containing 12 ng of 01 cRNA (0,&-cRNA injected) and incubated for a 4-h pulse and /33-Isoforms-Previously we have reported that the unassembled oocyte a-subunit is highly trypsin-sensitive and period with 3.5 mCi/ml [3sS]methionine and a 21-h chase period at 19 “C.Oocytes were then homogenized, fractionated on sucrose gracan be rendered trypsin-resistant by the concomitant synthe- dients, and the oocyte a-subunit was immunoprecipitated from the sis of exogenous P1-subunits (11). In addition, we observed 12fractions collected as describedin Fig. 1 and“Materialsand that exogenous a-subunits arerapidly degraded in theoocyte Methods.” Immunoprecipitated a-subunits revealed on fluorograms (shown below the line drawing) of control and &-cRNA-injected unless they are synthesized together with pl-subunits (10). In theexperiment shown in Fig. 3A, we tested whetherboth oocytes were quantitated and expressed as a-subunit (arbitary units) pg of protein in each fraction.B , cellular redistribution of oocyte PI-and &isoforms are able to stabilize exogenous al-subunits per a-subunits after injection of oocytes with &-cRNA. Oocytes were and to render oocyte a or exogenous al-subunits trypsin- injected with H20(control) or with 12 ng of Ps-cRNA (&cRNAresistant. For this purpose, oocytes were injected with H20or injected) and processed exactly as described in A. the indicated cRNA, labeled for 4 h with[35S]methionine,and incubated for a 72-h chase period before a microsomal fraction 20% resist tryptic attack under the experimental conditions was prepared and aliquots were subjected or not to a trypsi- used) (Fig. 3, A and B, lanes 2 and 4 ) . Injection of &-cRNA nolysis assay. First, this experiment confirms our previous (Fig. 3, A and B, lane 6) and &-cRNA (Fig. 3, A and B, lane observation (11) that a-subunits synthesizedalone in the 8)increases the trypsin-resistantoocyte a-subunit population oocyte from injected al-cRNA arerapidly degraded since the to about 60 and 50%, respectively. A similar proportion of alamount recovered after the 72-h chase period is not signifi- subunits becomes trypsin-resistant when p1-isoforms (Fig. 3, cantly different from H20injected oocytes (Fig. 3, A and B, A and B, lane 10) and P3-isoforms (Fig. 3, A and B, lane 12) compare lanes 1 and 3). On the other hand, cosynthesis of are coexpressed with exogenous al-subunits. Altogether these both pl-isoforms (Fig. 3, A and B, lane 9) and @s-isoforms datasupportthe hypothesis that assembly of as of PI(Fig. 3, A and B, lane 11) with exogenous al-subunits leads to isoforms can impose a structural change on the newly synthethe stabilization of a-subunits as reflected by their cellular sized a-subunit which is reflected in its increased trypsin accumulation. Second, as expected from previous studies ( l l ) , resistance and in itsstabilization. the oocyte a-subunit is highly trypsin-sensitive (only about Transport Competence and Glycosylation Processing of PI-
5
-
- “ I
”
“
”
a-& Na,K-ATPase Complexes in Xenopus Oocytes
A
581
1 2 3 4 5 6 7 8 9101112
" " " 1 4h pulse
trypsin
-
+
injection H20
B
2o 1
-
+ a1
-
+ pl
-
+ p3
-
chase
-
48
-
48
-
48
-
48
48 I
+
al+Pl 100
IendoHI-(+I-I+(
FIG.4. Glycosylation processing of &- and Bs-isoforms.00cytes were injected with 0.5 ng of P1-cRNA (lanes 1 and 2), 5 ng of at- and 0.5 ng of pI-cRNA (lanes3 and 4 ) , 0.5 ng of &cRNA (lanes 5 and 6 ) , 5 ng of al-and 0.5 ng of B-cRNA (lanes 7 and 8 ) , 5 ng of al-and 3.5 ng of P1-cRNA (lanes9 and IO), 5 ng of al- and 3.5 ng of &-cRNA (lanes 11 and 12) and incubated for a 4-h pulse with [35S] methionine (2 mCi/ml, lanes 1, 3, 5, 7) or a 4-h pulse and a 48-h chase (lanes2, 4, 6, and 8-12) at 19 "C. After preparation of Triton extracts (see "Materials and Methods"), aliquots containing 1 X 10' cpm were immunoprecipitated with anti-pl-serum (lanes 1-4, 9 and 10) or anti-&-serum (lanes 5-8, 11 and 12). Aliquots of samples shown in lanes 9-12 were subjected (lanes10 and 12)or not (lanes 9 and 11) to treatment with endoglycosidase H (endo H)after immunoprecipitation as described under "Materials and Methods." The black points in lanes 10 and 12 indicate the three (lane 10) andfour (lane 12) glycosylation sites of &- and &-isoforms, respectively.
- FIG.3. Cellular accumulation and acquisition of trypsin resistance of oocyte and exogenous al-subunits by & and B3isoforms. A, trypsinolysis assay of oocyte and exogenous a-subunits. Oocytes were injected with H 2 0 (lanes 1 and 2), 5 ng of al-cRNA (lanes 3 and 4 ) , 1 ng of &-cRNA (lanes 5 and 6 ) , 1 ng of &-cRNA (lanes 7 and 8),5 ng of a- and 1 ng of &-cRNA (lanes 9 and lo), 5 ng of aI-and 1ng of j33-cRNA (lanes11 and 12), and incubated for a 4-h pulse with 2 mCi/ml [3SS]methionineand a 72-h chase period at 19 "C. Microsomes of the oocytes were prepared and subjected (lanes 2, 4, 6, 8, 10, 12) or not (lanes 1, 3, 5, 7, 9, 11) to controlled trypsinolysis as described under "Materials and Methods." &-Subunits were immunoprecipitated, subjected to SDS-PAGE, and revealed by fluorography as described under "Materials and Methods." One representative experiment is shown. B, quantification of the cellular accumulation and thetrypsin resistance of a-subunits. Shown is the quantification of one to four experiments similar to the one shown in A. Values have been corrected for differences in proteinand counts in the different samples. The amount of a-subunits found in al-cRNA-injected oocytes was arbitrarily set to 1. n = number of different experiments. The numbers above the columns represent the percentage of a-subunits recovered in trypsin-treated samples compared to nontreated controls in each injection protocol.
with al-subunits (Fig. 4, lane 4 ) . After a 48-h chase, only about 6%of the core glycosylatedb3-subunitsynthesized alone during a 4-h pulse can be recovered (Fig. 4, compare lanes 5 and 6 ) , indicating that @3-isoformsnot assembled to a-subunits might even be morerapidly degraded than unassembled @l-isoforms. Clearly,a small population of @3-isoformssynthesized during a 4-h pulse in P3-cRNA injected oocytes become fully glycosylated during the 48-h chase period. (Fig. 4, lane 6). This population likely represents P3-isoforms which associated to endogenous oocytea-subunits and were able to leave the ER. Significantly, however, only when @3-isoforms are synthesized concomitantly with stoichiometric amounts of a-subunit, the total @3-isoform population synthesized during a 4-h pulse is recovered and completely processed to the fully glycosylated form after a 48-h chase (Fig. 4, compare lanes 7 and 8). To further characterize the different glycosylated speciesof B1- and @3-isoformsand to determine the actual number of glycosylation sites used in the two isoforms, we tested their endoglycosidase H sensitivity after coinjection of al- and @3.5 ng ofD-cRNA) and ~~-Isoforms-Previously we have reported that B1-sub- cRNAs at a ratio (5 ng of al-cRNA units need to be synthesized concomitantly with a-subunits which results in an overexpression of 8-isoforms compared to to become fully glycosylated in a trans Golgi compartment al-subunits and makes it possible to reveal both the fully (10). /31-isoforms synthesized alone remain in their core gly- glycosylated (associated to a-subunits) and the core glycosylated (in excess overa-subunits) forms after a 48-h chase.As cosylated ER form and are slowly degraded. shown in Fig. 4 (lanes9-12), the 49-kDa P1-forrn (Fig. 4, lanes Fig. 4A documents again these results and illustrates that 9 and 10) and the 59-kDa &-form (Fig. 4, lanes 11 and 12) similar conditions apply to
[email protected] a 4-h pulse, are not attacked by endoglycosidase H treatment and thus the pl- and the &isoforms are mostly recovered as 40-kDa represent the fully glycosylated species. On the other hand, (Fig. 4, lanes 1 and 3 ) and 42-kDa (Fig. 4, lanes 5 and 7) core the 40-kDa pl-forrn (Fig. 4, lanes 9 and 10) and the 42-kDa glycosylatedspecies,respectively, both in oocytes injected &form (Fig. 4, lanes 11 and 12) are digested to a 32-kDa and with B1 (Fig. 4, lane 1 ) and p3 (Fig. 4, lane 5)cRNA alone or a 30-kDa core peptide, respectively. Partial digestion by enin combination with aI-cRNA (Fig. 4, lanes 3 and 7). After a doglycosidase H reveals that pl-isoforms are core glycosylated 48-h chase, &subunits synthesized alone are found mostly in on three sites (Fig.4, lane 10) while@3-isoforms are core their core glycosylated form,but the total amount recovered glycosylated on four sites (Fig. 4 C,lane 12). Examination of decreases by about 70% (Fig. 4, lane 2). On the other hand, the amino acid sequence of Dl- and p3-isoforms reveals that about 75% of the /31-isoform is processed to the fully glyco- &-isoforms are indeed glycosylated on all three potential sylated form after a 48-h chase if it is synthesized in parallel glycosylation sites. Interestingly, however, P3-isoforms reveal
+
a-PsNa,K-ATPase Complexes in Xenopus Oocytes
582
only three accepted consensus sequences forN-linked glycosylation (NRT at amino acid 147, NCT at amino acid 190, and NFT at amino acid 240). Apparently, an additional unusual asparagine residue (perhaps NDS at amino acid 117)is glycosylated in j334soforms. In general, it is accepted that glycoproteins which appear in their core glycosylated form resideat the level of the ER. However, at least theoretically, it is possible that core glycosylated proteins, e.g. 8-subunits of Na,K-ATPase which are not associated witha-subunits, have a structural defect which indeed permits exit from the ER but does not permit complex type glycosylation at thelevel of the distal Golgi. To clearly determine the cellular localization of the B-isoforms synthesized alonein 8-cRNA-injected oocytes, we again fractionated homogenates of Xenopus oocytes which were subjected to a pulse or a pulse-chase protocolafter injection of 8-cRNA. Fig. 5A shows that after a 4-h pulse in B1-cRNA-injected oocytes, the @l-isoformis as expectedmainly in its core glycosylated form concentrated in ER-enriched fractions as defined in Fig. L4. After a 4-h pulse and a 21-h chase, a small amount of fully glycosylated 49-kDa P1-forms appear which distribute in plasma membrane-enriched fractions, but the main proportion of the &subunit population is in its core glycosylated form still residing in ER fractions. From these
results we conclude that after a 21-h chase a small percentage (about 15%) of the total &-subunit synthesized during a 4-h pulse can leave the ER andbecome fully glycosylated, probably due to association with endogenous oocyte a-subunits (see Fig. 2A), but that the unassembled &isofoms remain in the ER in their core glycosylated form. Fig. 5Billustrates that thesame holds true if f13-isoformsinstead of &-isoforms are expressed in the oocyte. Indeed, after a 4-h pulse and a 21-h chase, the &-isoform is mainly in its core glycosylated form and resides at the level of the ER. Finally, confirming the results shown in Fig. 4, ifal-and &-cRNAs are coinjected, nearly the total &-isoform population is processed into the 49-kDafullyglycosylatedform after a 21-h chase and it distributes in plasma membrane-enrichedfractions (Fig. 5C). Functionul Expression of a-O1and 4 3 Na,K-ATPase Complexes-So far the present results indicate the following. 1) Xenopus oocytes express an excess trypsin-sensitive a-subunits over &subunits; 2) according to immunochemical criteria, the 8-subunit of Xenopus oocytesis not a PI- but a 8 3 isoform; and 3) exogenous Dl- as well as /33-isoforms are able to associate with endogenous oocyte a- and exogenous alsubunits. Oligomerizationof PI- or j33-isoforms and &-subunits render a-subunits trypsin-resistant and permit the transport out of the ERof the two subunits. Although these latter data
A. 401
O
1.W
A
L
1.W 1.018.110.112.114.16
118
density
bPI
0 . i41.016.018.110.112.114.116.18 density
ma
G 4 0 ”
4h pulse
4 8 -40
”58
1-42
1.01 4.01 6.01 8.11 0.11 2.11 4.11 6.18
density 41pulm
21h chase
-
”
-m
FIG. 5. ER retention of unassembled B1- and Bs-isoforms. A, cellular distribution of core glycosylated 01isoforms. Oocytes were injected with 5 ng (0)or 12 ng 8, cRNA (0)and incubated for a 4-h pulse period with 3.5 mCi/ml of [35S]methionine(0)or for a 4-h pulse and a 21-h chase period (0)at 19 “C. Oocyteswere then homogenized and fractionated on sucrose gradients and the B1-subunit was immunoprecipitated from the 12 fractions collected as described in Fig. lA and “Materials and Methods.” Immunoprecipitates of &-isoforms revealed on fluorograms (shown below the line drawing) after a 4-h pulse or a 4-h pulse and a 21-h chase were quantitated and expressed as thepercent of &-subunits (40-kDacore plus 49-kDa fullyglycosylatedspecies) found in each fraction compared to the total amount recovered in the 12 fractions. B, cellular distribution of core glycosylated &-isoforms. Oocytes wereinjected with 3 ng of &cRNA and incubated for a 4-h pulse with 3.5 mCi/ ml [3sS]methionineand a 21-h chase period at 19 “C. Oocytes were then homogenized and fractionated on sucrose gradients, and the&-isoform was immmunoprecipitated. Quantifications of immunoprecipitates and representation of data as in A. 42 kDa = core glycosylated and 59 kDa = fully glycosylated p3-isoform. C, ER exit and full glycosylation of fll-isoforms by concomitant synthesis of al-subunits. Oocytes were injected with 9 ng of ax-and 3 ng of &cRNA and incubated for a 4-h pulse with 3.5 mCi/ml [%]methionine and a 21-h chase period at 19 “C. Oocyte fractionation, immunoprecipitation of &-isoforms, and quantification of data as in A and “Materials and Methods.”
a-Pa
Na,K-ATPase Complexes in Xenopus Oocytes
583
support the notion that P3-isoforms have similar basic properties to &-isoforms, we tested in a final set of experiments whether putativeP3-isoformsare indeed able to support Na,KATPase activity at the plasma membrane. For this purpose, we expressed B3-isoforms alone or together with a1-subunits in Xenopus oocytes and measured on the one hand theouabain binding capacity of intact oocytes and on the otherhand, %Rb uptake or Na,K-pump currents. Fig. 6A shows that compared to injection of al-cRNA alone, injection of al-and &-cRNAs Y = - 5.677'2 + 2.43811 R"2 0.0% produces a significant increase of both ouabain binding capacity and %Rb uptake. The increase in the two parameters 0 10 20 30 40 is similar if not more important than the one observed in ouabain binding(fmoUoocyte) oocytes injected with a>-and &-cRNA (Fig. 6A). Similar observations are made if pump currents are correlated with ouabain binding capacities (Fig. 6B). Inthese experiments it becomes apparent that the two parameters measured do not significantly differ in H20-injected and al-cRNA-injected oocytes supporting the notion that al-subunits alone cannot 4-1 / produce functional Na,K-pumps in addition to the endogenous oocytes Na,K-pumps. On the other hand, injection of Dl- and &-cRNA alone leads to a small but significant increase in both the number of pumps expressed at theplasma membrane and thepump current produced. This result reinforces 0 4' , , , , , , , , the observation that &- and p3-isoforms can associate to 0 1 2 3 4 5 6 7 6 oocyte a-subunits synthesized in excess over @-subunitsand ouabaln binding (miativa valuer) form functionalpumps at the plasma membrane. Finally, FIG. 6. Cell surface expression and functional characteriza- injection of al-and Dl- or al-and &cRNAs leads to an tion of a-81and a-& complexes. A, ouabain binding and ffiRbflux measurements in oocytes expressing a1-& and o(1-63 complexes. 00- important increase of functional pumps expressed, which cytes were injected with 10 ng of al-cRNA, 10 ng of al-cRNA, and 4 varies in different batches of oocytes between 2.5- and 7-fold ng of B1-cRNA or 10 ng of al-and 4 ng of &-cRNA and incubated in over controls. Significantly, the increased number of pumps MBS for 48 h a t 19 "C. Ouabain binding was determined as described is paralleled by a similar increase in the pump current prounder "Materials and Methods" on 15, 14, and 13 individual oocytes duced in all injection protocols (Fig. 6B). From the mean injected with al-cRNA, al- and &-cRNA, and al-and &cRNA, number of ouabain binding sites and values of Na,K-pump respectively. On another set of oocytes, 86Rbuptake was determined as described under "Materials and Methods" on 9,9, and7 individual current, we could calculate a turnoverrate of the Na,K-pump, oocytes injected with al-cRNA, al-and P1-cRNA, and al- and p3- assuming thetransfer of one charge per cycle. Inthe 8 cRNA, respectively. Represented is the mean f S.E. of =Rb (pico- experimental groups of Fig. 6B, the value of this turnover rate moles/oocyte/min) taken up into oocytes as a function of the mean ranged from 30 to 58 s-' (mean f S.E.: 48 f 3.2).
f"
& S.E. ouabain (femtomoles/oocyte) bound tointact oocytes. B, ouabain binding and pump current measurements in oocytes expressing oocyte a-& oocyte a&, exogenous al-P1, and exogenous al$3 complexes. Oocytes were injected with HzO, P1-cRNA,P3-cRNA,aland pl-cRNA, or al-and &-cRNA and incubated for 48 h at 19 "C. Ouabain binding was determined on one set of oocytes and K+induced pump current was measured in another set of oocytes as described under "Materials and Methods." The graph represents the pump current as a function of ouabain binding and is a compilation of three experiments performed on threedifferent animals. 0, oocytes were injected with H20, 4 ng of pl-cRNA, 4 ng of P3-cRNA,or 10 ng of a,-cRNA and ouabain binding was measured on 14-15 oocytes and pump current on 8-11 oocytes in each injection protocol. aI-cRNAinjected oocytes bound 20.4 f 2.1 fmol of ouabain per oocyte and produced a pump current of 96.6 nA f 6.6. Unpaired t test; A, ouabain binding: a1 versus H20,p > 0.6; a1versus Dl, p < 0.004; a1versus p3; p < 0.017; B, pump current: a1versus HzO, p > 0.1; a1versus pl, p < 0.001; a1versus p3, p < 0.001. 0, oocytes were injected with 10 ng of aland 7 ng of P1-cRNA, 10 ng of al-and 7 ng of p3-cRNAor 10 ng of al-cRNA. Ouabain binding and pump current was measured on 13-15 oocytes in each injection protocol. al-cRNA injected oocytes bound 14.5 f 0.7 fmol of ouabain per oocyte and produced a pump current of70.4 f 5.9 nA. Unpaired t test, A, ouabain binding: a1 versus a1 and Dl, p < 0.002; a1 versus a1and p3, p < 0.001; al and p1 versus a1 and p3,p > 0.5. B, pump current: a1versus a1and pl, p < 0.001; al v e r s u a1and p3, p < 0.001, a1and p1versus a1and p3,p > 0.4. 0, oocytes were injected with 10 ng of al and 7 ng of & cRNA, 10 ng of 011- and 7 ng of &-cRNA or 10 ng of al-cRNA. Ouabain binding was measured on 25 oocytes and pump current on 10-16 oocytes in each injection protocol. al-cRNA-injected oocytes bound 6.8 f 0.2 fmol of ouabain per oocyte and produced a pump current of 19.5 f 4.7 nA. Unpaired t test, A, ouabain binding: a1 versus a1and 01, p < 0.001; a1 versus a1 and p3,p < 0.001, a1and p1versus a1 and p3, p < 0.001, B, pump current: a1 versus a1 and Dl, p < 0.001, a1 versus a1 and p3 p < 0.001; a1and p1versus a1and p3, p > 0.1. In all
DISCUSSION
In thisstudy, we have characterized the unequal expression of a- and P-subunits of Na,K-ATPase in Xenopus oocytes and have taken advantage of the unusual biosynthesis modeof the enzyme to elucidate whether a putative B3-isoform might be a functional component of Na,K-ATPase in general and of oocyte Na,K-ATPase in particular. Immunochemical Evidence Suggests That Xenopus Oocytes Express a Putatiue Ps-Isoform of Na,K-ATPase-In this study we have applied a cell fractionation protocol and were able to follow the synthesis and theintracellular transport of a polypeptide immunoreactive with a specific anti-p3-serum. This polypeptide has a molecular mass when recovered in plasma membrane-enriched fractions on sucrose gradients similar to that of the immunoreactive material detected previously in the plasma membrane of radioiodinated oocytes with an anti&-serum (11). Since we could recently show that anti-/&serum does not cross-react with the &-subunit but that anti&-serum recognizes &-isoforms synthesized in vitro from tRNA: we believe that Xenopus oocytes express a p3- and not a pl-isoform. This hypothesis is also supported by the observation that thepolypeptides detected in Xenopus oocytes are processed froma 42-kDa species to a 59-kDa species during apulse-chase experiment similarly to thepolypeptides three experiments, ouabain binding and pump currents measured in a,-cRNA-injected oocytes were arbitrarily setto one (W) and the fold increase in the two parameters was calculated in each injection protocol.
584
(~$33
Na,K-ATPase Complexes in Xenopus Oocytes
synthesized in the oocytes from injected P3-cRNA (Figs. 1,A and B, and 4). The expression of @3-isoformsin Xenopus oocytes is also consistent with the presence of a maternal pool of &,-cRNA (7). Finally, @l-isoformsexpressed in oocytes from injected P1-cRNA are processed from a 40-kDa core glycosylated to a 49-kDa fully glycosylated species as expected from data obtained in aXenopus kidney cell line (19). The purpose of the present study was to provide evidence that this putative Bs-isoform can support similar functions than @l-isoformsand thuscould be part of a functional Na,KATPase, e.g. of the Xenopus oocyte Na,K-pump. Xenopus Oocytes Accumulate an Excess of a-Subunits over @3-Isoformsin the ER-It is now established that botha- and @-subunits are needed to form functional Na,K-pumps (11, 25, 26) and that unassembled subunits are in a structurally immature form (11) and are retained in the ER (10). We previously have provided some evidence that Xenopus oocytes indeed express @-immunoreactive material atthe plasma membrane but, as shown also in this study, synthesize the putative p3-isoformto a lesser extent than the a-subunit. The question arises whether still another 6-isoform might exist in the oocyte which we were not able to detect or whether the a-subunit is synthesized in excess over @-subunits.To answer this question we tested whether part of the oocyte a-subunit shows characteristics of unassembled immature a-subunits. We confirm in this study our previous observation (11) that a large proportion of oocyte a-subunits is highly trypsinsensitive similar to a-subunits synthesized in the oocyte in excess over @-subunitsfrom injected mRNA (11). In addition, by using a cell fractionation protocol, we provide evidence that an importantpopulation of oocyte a-subunits remain in the ER or a nearby compartment. While no @-subunitscodistribute with this ER a-subunit population, the stoichiometry of a- and fi3-isoformsrecovered in plasma membrane-enriched fraction is on the other handclose to 1. PI- as Well as &-Isoforms Can Stabilize the a-Subunit and Induce Its Exit from the ER-Previously we reported that injection of &-cRNA into oocytes leads to an increase in trypsin-resistant oocyte a-subunits and anincreased number of functional pumps at theplasma membrane as assessed by ouabain binding (11). In this study we could confirm this observation and by the use of a cell fractionation protocol were able to show that theconcomitant synthesisof exogenous &-subunits withoocyte a-subunits provokes the exit from the ER of the naturally ER-retained oocyte a-subunit population. These data thus confirm in a physiologically relevant situation, and not only when exogenous a-subunits areartificially overexpressed in a cell, that a-subunits only become transport-competent when cosynthesized with @-subunits. These resultsreflect that, as in other multisubunit proteins (for review, see Ref. 27), oligomerization of @-subunitsto asubunits is a prerequisite for the structural maturationof the protein and its ability to leave the ER. Significantly, in this study we show that putative @3-isoformscan take over similar functions than pl-isoforms. The fact that B3-isoformstoo can mobilize oocyte a-subunits from the ER, can stabilize exogenous a-subunits andrender oocyte and exogenous a-subunits trypsin-resistant provides the first evidence that P3-isoforms potentially can associate with a-subunits andbecome part of the functional enzyme. @3-Isoformsas P1-Isoforms Can Only Leave the ER When Coexpressed with a-Subunits-On the basis of the observation that B1-subunits are not fully glycosylated in the absence of concomitant synthesisof a-subunits, we previously suggested that P1-subunits not assembled with a-subunits cannotleave the ER similar to a-subunits not assembled with @-subunits
(10). In this study we support this hypothesis by localizing the core glycosylated unassembled &-subunitin the ER. Indeed, unassembled P1-subunits are mostly recovered in sucrose gradient fractions which contain Bip, an ER resident protein (Figs. lA and 5 A ) while the peak of galactosyltransferase activity, a Golgi marker, is found in lighter fractions (see “Materials and Methods”). We therefore exclude the possibility that unassembled @-subunitsmight perhaps be transported to a distal Golgi compartment but might not be fully glycosylated due to a structural default. In addition, by following the intracellular localization and the glycosylation processing of the P3-isoformsynthesized alone or in the presence of exogenous al-subunits we could show that p3-isoforrns exhibit similar properties than @l-isoforms.This result provides a further argument for a functional relationship of &isoforms with Na,K-ATPase. ffl-83and a-& Complexes Have Similar Transport Properties-Apart from the characterization of its basic functional properties, namely its implication in the structural maturation of the a-subunit and the intracellular transport of the enzyme to the plasma membrane, we finally tackled the question whether &-isoforms can form functional pumpsat theplasma membrane. By using two functional assays, rubidium uptake and electrophysiological Na,K-pump current measurements combined with ouabain binding studies, we show that P3isoforms can form functional pumpsat theplasma membrane with bothoocyte a-subunits andexogenous al-subunits. Since the increase in pump current produced in B1-, p3-, al- and @ I - , or al- and &cRNA-injected oocytes compared to H20or al-cRNA-injected controlsclosely parallels the increase in the number of pumps expressed at theplasma membrane, we can conclude that at leastunder the actual experimental conditions which provide high sodium and potassium concentrations to the enzyme, a-P3enzyme complexes have transport properties similar to a-P1complexes. This hypothesis is supported further by our recent observation that the potassium affinity and thevoltage dependence of the electrogenic activity of the Na,K-pump expressed in oocytes is similar with alp1and a1-@3complexes (24). A still fineranalysis is needed to uncover potentialfunctional differences inNa,K-ATPase complexes composed of different subunit isoforms. In conclusion, in this study we have attempted to identify @3-isoformsas functional components of Na,K-ATPase and in particular of the Na,K-pump of Xenopus oocytes in which b3-isoforms are the only @-subunitspecies detected. For this purpose, we have further characterized basic functional properties of Pl-isoforms and have been able to show that @3isoforms share similar characteristics to pl-isoforrns with respect to their ability to associate with and to stabilize asubunits, to render the a-subunit transport competent, and to form functional pumpsat theplasma membrane. Together with the recent observation that @3-isoformsare down-regulated at theplasma membrane of Xenopus oocytes along with the a-subunitduring progesterone-induced maturation: these data strongly suggest that Ps-isoforms which so far have been thoughtto be brain-specific inadult Xenopus (7)arein addition part of the functional Na,K-pump of oocytes. Acknowledgments-Wewould like to thank E. Schneeberger, S. Roy, and D. Schaer for their skillful technical assistance. Our thanks also go to E. Casadio for typing the manuscript and to N. Narbel for preparing the bibliography. REFERENCES 1. Jrgensen, P. L., and Andersen, J. P. (1988) J.Membr. Biol. 103, 95-120 2. Skou, J. C. (1990) FEBS Lett. 268, 314-324 3. Lingrel, J. B., Orlowski, J., Shull, M. M., and Price, E. M. (1990)
.-Pa Nu,K-ATPase Complexes in Xenopus Oocytes Prog. Nucleic Acid Res. Mol. Biol. 38,37-89 4. Sweadner, K. J. (1989) Biochim. Biophys. Acta 9 8 8 , 185-220 5. Gloor, S., Antonicek, H., Sweadner, K. J., Pagliusi, S., Frank, R., Moos, M., and Schachner, M. (1990) J. Cell Bwl. 110,165-174 6. Martin-Vasallo, P., Dackowski, W., Emanuel, J. R., and Levenson, R. (1989) J. Biol. Chem. 264,4613-4618 7. Good, P. J., Richter, K., and Dawid, I. B. (1990) Proc Natl. Acad. Sci. U. S. A. 87,9088-9092 8. Jrgensen, P. L. (1974) Biochim. Biophys. Acta 3 5 6 , 36-52 9. Geering, K. (1990) J. Membr. Biol. 1 1 5 , 109-121 10. Ackermann, U., and Geering, K. (1990) FEBS Lett. 2 6 9 , 105108 11. Geering, K., Theulaz, I., Verrey, F., HgUptle, M. T., and hssier, B. C. (1989) Am. J. Physiol. 2 6 7 , C851-C858 12. Geering, K., Kraehenbuhl, J. P., and Rossier, B. C (1987) J. Cell Biol. 105,2613-2619 13. Colman, A. (1984) in Transcription and Translution: A Practical Approach (Hame%B. D., and s. J., PP. 271-302,
IRL Press, Oxford
14. Melton, D. A., Krieg, P. A., Rebagliati, M. R., Maniatis, T.,Zinn, K.9 and Green, M.R. (1984) Nucleic Acids Res. 1297035-7056 15. Verrey,F., Schaerer, E., Zoerkler, P., Paccolat, M. P., Geering, K., Kraehenbuhl, J. P., and Rossier, B. C. (1987) J . Cell Biol. 104,1231-1237
585
16. Good, P. J., Welch, R. C., Barkan, A., Somasekhar, M. B., and Mertz, J. E. (1988) J. Virol. 62,944-953 17. Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275 18. Pelham, H. R. B. (1989) Annu. Rev. Cell B i d . 5 , 1-23 19. Geering, K., Girardet, M., Bron, C., Kraehenbuhl, J. P., and Rossier, B. C. (1982) J. Biol. Chem. 2 5 7 , 10338-10343 20. Geering, K., Meyer, D. I., Paccolat, M. P., Kraehenbuhl, J. P., and Rossier, B. C. (1985) J. Bwl. Chem. 260,5154-5160 21. Girardet, M., Geering, K., Franks, J. M., Geser, D., Rossier, B. C., Kraehenbuhl, J. P., and Bron, C. (1981) Biochemistry 2 0 , 6684-6691 22. Schmalzing, G., Omay? H.?Kroner, sv G100r7 s . 9 Appelhans, H., and Schwarz, W. (1991) Biochem. J. 279,329-336 23. Sepulveda, F. V., and Pearson, J. D. (1984) J. Cell. Physiol. 118, 211-217 24. Horisberger, J.-D., Jaunin, P., Good,P. J., Rossier, B. C., Geering, K. (1991) Proc. Natl. Acad. Sci.U. S. A,, in press 25. Horowitz, B., Eakle, K. A., Scheiner-Bobis, G., Randolph, G.R., Chen, C. Y., Hitzemann, R. A., and Farley, R.A. (1990) J. Biol. Chem. 265,4189-4192 26. Noguchi, S., Mishina, M., Kawamura, M., and Numa, S. (1987) FEBSLett. 225,27-32 27. Hurtley, S. M., and Helenius, A. (1989) Annu. Rev. Cell Biol. 5 , 277-307
