ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 294, No. 2, May 1, pp. 550-556,1992

Rat Skeletal Muscle Membrane Associated Carbonic Anhydrase Is 39-kDa, Glycosylated, GPI-Anchored CA IV’

Abdul Waheed,* Xin Liang Zhu,* William S. Sly, *J Petra Wetzel,? and Gerolf Grost *Edward A. Daisy Department of Biochemistry and Molecular Biology, St. Louis University School of Medicine, 1402 South Grand Boulevard, St. Louis, Missouri 63104; and tMedizinische Hochschub Hannover, Zentrum Physiologic, Hannover, Germany

Received September 9,1991, and in revised form January 8,1992

Sarcolemmal membrane vesicle preparations from white and red muscles of rat were found to contain a carbonic anhydrase which was indistinguishable from carbonic anhydrase IV from rat lung. This isozyme ap- pears to account for all of the carbonic anhydrase activity in the sarcolemmal vesicle preparations. Digestion of 39, kDa CA IV with endoglycosidase F reduced the M, to 36 kDa, suggesting that it contains one N-linked oligosac- charide. Treatment of sarcolemmal vesicles with phos- phatidylinositol-specific phospholipase C released all of the activity, indicating that the enzyme is anchored to membranes by a phosphatidylinositol-glycan linkage. White muscle sarcoplasmic reticulum vesicles also con- tain a small amount of 39-kDa CA IV-type enzyme. A 62-kDa polypeptide in sarcoplasmic reticulum mem- branes cross-reacts with anti-human CA II and anti-rat CA II antisera, but does not bind to the sulfonamide af- finity column. This cross-reacting polypeptide has no de- tectable CA activity. Q 1~s~ ACP~~C m. IUC.

Several lines of physiological and histochemical evi- dence suggest that one of the carbonic anhydrases (CAS)~ in skeletal muscle cells is associated with the sarcolemmal membrane (l-7). Recently, we presented a systematic characterization of sarcolemmal-associated CA from rab- bit skeletal muscle (8). Sarcolemmal vesicle preparations

i This research was supported by the Deutsche Forschungsgemein- schaft Gr 489/4 and by Grants GM34182 and DK40163 from the Na- tional Institutes of Health.

’ To whom correspondence should be addressed. 3 Abbreviations used: CA, carbonic anhydrase; SR, sarcoplasmic re-

ticulum; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel elec- trophoresis; PI-PLC, phosphoinositol-specific phospholipase C; PVDF, polyvinylidene difluoride; GPI, glycosyl-pbosphatidylinositol.

550

were shown to contain a membrane-associated CA which partitions into the detergent phase upon Triton X-114 extraction. Inhibition characteristics of the sarcolemmal vesicle-associated CA were different from those of known soluble CA isozymes, and we set out to determine whether the CA in skeletal muscle is CA IV, the membrane- associated isozyme recently characterized in lung and kidney.

Membrane-associated CA IV has been purified from bovine lung (9), human lung, and human kidney (10,ll). CA IV has also been purified from rat lung and used to raise a monospecific antiserum in rabbits (12, 13) which was useful for immunolocalization of CA IV-type enzyme in rat kidney epithelial cells (12). The availability of the rat lung CA IV antiserum and preparations of rat skeletal muscle sarcolemmal vesicles allowed us to characterize the sarcolemmal vesicle-associated CA. In this study we present evidence that sarcolemmal vesicle-associated CA is definitely a CA IV-type enzyme which is anchored to membranes by a phosphatidylinositol-glycan linkage. This isozyme plays a critical role in COP and HCO; trans- port in other tissues. The presumed function of this iso- zyme in skeletal muscle would be to catalyze the rapid hydration of COz to HCO; as the COx produced in exer- cising muscle diffuses across the membrane to the extra- cellular space.

MATERIALS AND METHODS Preparation of sarcokmmal vesicles. Sarcolemmal vesicles were pre-

pared as described previously (8). The white muscle preparation was from the hind limbs of five female Wistar rats (30 g). In the case of the red muscle preparation, the solei of 50 rats were employed. Sarcolemmal vesicle fractions IIA and IIIA banded at the sample/8.7% dextran in- terface and the sarcolemmal vesicle fractions IIB and IIIB at the 8.7%/ 18% dextran interface. White muscle sarcoplasmic reticulum (SR) vesicle fractions SRl and SR2 were obtained from the 35% sucrose phase.

Analytical methods. Ouabain-sensitive Na+,K+-ATPase and Ca’+- dependent ATPase were measured as described by Seiler and Fleischer

0003-9861/92 $3.00 Copyright 0 1992 by Academic Press, Inc.

All rights of reproduction in any form reserved.

551 RAT MUSCLE MEMBRANE CARBONIC ANHYDRASE IV

TABLE I

Sarcolemmal and Sarcoplasmic Reticulum Vesicle Fractions from White and Red Muscles

Fraction Protein bdml)

Ouabain-sensitive Me-ATPase NA+,K+-ATPase km01 Pi/ bmol PAmg- h)l (mg . min)]

Cholesterol (wnol/mg)

Cax+-ATPase bmol Pi/

(mg. min)]

Carbonic anhydrase NJ * ml/md

White muscle SL” IIA IIB IIIA IIIB

Red muscle SL IIA IIB IIIA IIIB

White muscle SR SRl SR2

2.82 58.3 0.67 0.90 0.11 24.4 (47)* 5.20 27.2 0.49 0.51 0.09 12.2 (44) 1.33 46.5 0.40 0.55 0.13 25.8 (42) 3.55 45.5 0.81 0.48 0.13 15.7 (45)

1.63 16.3 1.46 1.31 0.09 18.5 4.63 5.9 0.45 0.39 0 5.6 1.22 13.9 1.25 0.89 0.02 18.6 5.21 10.2 0.98 0.51 0.05 6.9

12.20 3.2 0.09 0.01 1.45 0.5 (54) 11.25 2.5 0.06 0.01 1.37 0.5 (52)

Note. All enzyme activity measurements were in duplicate on one preparation of vesicles. ’ Sarcolemma. * % of carbonic anhydrase activity sensitive to 0.2% SDS for 30 min is given in parentheses.

(14). The cholesterol content was determined enzymatically according to the method of Allain et al. (15) employing the A-Gent cholesterol test (Abbott Laboratories, Diagnostics Division). Carbonic anhydrase (CA) activity measurements were done in the presence and absence of SDS (16). To determine the SDS-resistant CA activity, the enzyme preparation was exposed to 0.2% SDS at room temperature for 30 min before the enzyme assay was performed. CA inhibitor affinity chroma- tography on sarcolemmal and SR CAs was carried out as described (10). Protein was measured according to the method of Lowry et al. (17) as modified by Peterson (18). Bovine serum albumin was used as a standard. SDS-PAGE under reducing conditions was carried out according to Laemmli (19). Low molecular weight standard proteins were from Bio-Rad.

After SDS-PAGE, protein polypeptides were electrophoretically transferred on an Immobiion-P, PVDF membrane (from Millipore) using a polyblot assembly (American Bionetics Co). Following peptide transfer, membranes were subjected to immunostaining as described (10) using 1:lOOO diluted rabbit anti-rat lung CA IV or rabbit anti-human CA II antiserum as primary antibodies and goat anti-rabbit antibodies con- jugated to alkaline phosphatase, 1:lOOO dilution, or goat anti-rabbit an- tibodies conjugated to peroxidase, 1:500 dilution, as a second antibody. In order to visualize all polypeptides, the membranes wem stained with 0.5% amino black or 0.1% Coomassie blue in 50% methanol and 10% acetic acid for 1 min.

For endoglycosidase F treatment, sarcolemmal membrane fractions equivalent to 30 pg protein, or solubilized enzyme from 30 I.cg membrane proteins, were denatured and treated with 150 mU endoglycosidase F in a 20-pl reaction volume as described (lo), except the denaturation buffer used was 0.1 M sodium phosphate, pH 8.6, containing 1.2% NP- 40, 0.2% SDS, and 1% &mercaptoethanol. Before addition of endogly- cosidase F 1 mM each of phenylmethylsulfonyl fluoride and EDTA and 5 mM iodoacetate were added to the reaction mixture as protease in- hibitors. The deglycosylation reaction was analyzed by SDS-PAGE and immunoblotting.

To measure release of CA by treatment with phosphatidylinositol- specific phospholipase C, sarcolemmal membrane vesicles equivalent to 30 pg protein were treated with 10 mEU of PI-PLC in 100 mM Tris- sulfate, pH 7.5, in the presence and absence of 0.1% sodium deoxycholate at 37°C. Soluble and membrane forms of the enzyme were recovered by

centrifugation of the reaction mixture (10). The soluble enzyme (su- pernatant) was precipitated using 96% cold acetone at -2O’C.

For immunoprecipitation of carbonic anhydrase released by phos- phatidylinositol-specific phospholipase C, PI-PLC-solubilized enzyme (20 ~1) was first mixed with an equal volume of 50 mM Tris-SO4 buffer, pH 7.5, and 20 fig of rabbit anti-rat lung CA IV IgG or rabbit preimmune IgG for 4 h at 4’C. The immune complex was recovered by incubation with 20 pl of 10% suspension of IgGSorb after mixing for 1 h at 4°C and centrifugation. The clear supematants were used for a CA assay to measure the released CA remaining after precipitation by immune and preimmune IgG.

Antiserum against homogeneous rat lung CA IV (12,13) and recom- binant human CA II (D. Roth and W. S. Sly, unpublished) was raised in rabbits. Anti-human CA II antiserum cross-reacts with rat CA II on Western blots. Goat anti-rabbit antibodies conjugated with alkaline phosphatase or peroxidase were from Sigma Chemical Co. 4-Chloro-l- naphthol was purchased from Aldrich Chemical Co. PI-PLC from Ba- ciUu.s thuringiensis was the generous gift of Dr. Martin Low (Columbia University, New York). p-(Aminomethyl)benzenesulfonamide coupled to Affi-Gel-10 resin was used as described (10). All other reagents used were of analytical grade.

RESULTS AND DISCUSSION

Characterization of Sarcolemmal and Sarcoplasmic Reticulum Membrane Vesicles from White and Red Muscles

Table I presents a summary of the characterization of sarcolemmal and SR membrane vesicle preparations. Ouabain-sensitive Na+,K+-ATPase and Mg’+-ATPase, markers for sarcolemma (14,20,21), are 9- to 23-fold and 4- to 14-fold higher, respectively, in white muscle sarco- lemma1 fractions IIA (8.7% dextran interface) and IIIB (8.7%/H% dextran interface) than in white muscle sar- coplasmic reticulum fractions SRl and SR2. Similarly, cholesterol, a component of the sarcolemmal membrane

552 WAHEED ET AL.

(14,20,21), is 48- to go-fold higher in sarcolemmal vesicles from white muscles than in SR fractions 1 and 2. However, Ca’+-ATPase, a marker for SR, is lo- to 16-fold higher in fractions SRl and SR2 than in white muscle sarcolem- ma1 membrane fractions IIA-IIIB. Mgzf-ATPase activ- ities and cholesterol content in red muscle sarcolemmal fractions IIA-IIIB are comparable to those of white mus- cle sarcolemmal vesicles, whereas ouabain-sensitive Na+,K+-ATPase activities are 3-fold lower. Ca2+-ATPase activities in red muscle sarcolemmal fractions are similar in magnitude to those in sarcolemmal membrane fractions from white muscles. These results show that, judging from the sarcolemmal and SR markers, sarcolemmal membrane vesicles used in these studies are highly enriched in surface membranes and display a negligible contamination by SR. Further characterization was done by analyzing the pro- tein composition of the sarcolemmal and SR membrane fractions from white muscles. Identical amounts of mem- brane protein, 30 pg from each membrane fraction, were subjected to SDS-PAGE under reducing conditions. After electrophoretic transfer to PVDF membranes, polypep- tides were stained with Coomassie blue, both on PVDF membranes and in polyacrylamide gels. There was no staining in the polyacrylamide gels, suggesting that elec- trophoretic transfer of polypeptides was complete (results not shown). The results of staining of polypeptides on a PVDF membrane are shown in Fig. 1. A variety of poly- peptides are present in sarcolemmal and SR membrane preparations. No marked quantitative differences between different preparations of sarcolemmal membranes IIA, IIB, IIIA, and IIIB were seen, with the exception of a polypeptide of 32 kDa which is relatively more abundant in IIB and IIIB (marked by an open arrowhead). The polypeptide pattern of SR was significantly different from sarcolemmal membranes and contained higher amounts of 52- and 60-kDa polypeptides and 92- to 95-kDa poly- peptides as indicated by a closed arrowhead.

Carbonic Anhydrase Activity in Vesicles

The specific activities for CA shown in Table I for white muscle sarcolemmal vesicles are 24-52 times higher than those of the SR membrane fractions SRl and SR2. The CA activities in the red muscle sarcolemmal fractions are 30-50% lower than those in white muscle sarcolemma. Since membrane-associated CA IV from human lung and kidney (10, ll), bovine lung (9), and human urinary membrane (16) is resistant to 0.2 or 1% SDS at room temperature for 15-30 min (a treatment which inactivates soluble CAs), we checked SDS resistance of the CA in sarcolemma and SR vesicles. We found that about 50% of the total CA activity was SDS-resistant (0.2% SDS at room temperature for 30 min, Table I). The fact that about 50% of the membrane-associated CA is sensitive to SDS means either that membrane-associated CA from rat is more sensitive to SDS than human CA IV or that a por-

FIG. 1. Electrophoretic pattern of polypeptides of sarcolemmal and sarcoplasmic reticulum membrane vesicles. Different sarcolemmal (IIA, IIB, IIIA, and IIIB) and SR preparations (SRl and SR2) (30 pg each) were subjected to 10% SDS-PAGE. After electrophoretic transfer of the polypeptides to a PVDF membrane, the proteins were stained with Coomassie blue. Open (32 kDa) and closed (52, 60, and 90-95 kDa) arrowheads indicate uncommon polypeptides. Apparent molecular masses of standard proteins are marked in kilodaltons.

tion of the rat enzyme is partially degraded or otherwise modified to make it sensitive to SDS.

We previously reported that urinary membranes con- tain both CA IV and vesicle-enclosed CA II (16). The vesicle-enclosed CA II activity was stimulated when 0.1% saponin was included in the assay to permeabilize the membrane vesicles, presumably because it allowed escape of HCO;, relieving the product inhibition by HCO; ac- cumulation in sealed vesicles. We examined the effects of saponin on the sarcolemmal vesicle CA activity from white muscles and found no stimulation by 0.1% saponin (Table II, “Leakiness of vesicles”), suggesting that the mem- brane-associated CA is accessible to the substrate and the reaction is not limited by product accumulation.

Vesicle Carbonic Anhydrase Has a Molecular Mass of 39 kDa and Is Immunoreactive with Antibody to CA IV

Presence of immunoreactive CA IV-type peptides. Sarcolemmal fractions IIA-IIIB and sarcoplasmic retic- ulum fractions SRl and SR2 were subjected to SDS- PAGE. After electrophoretic transfer of polypeptides from SDS-polyacrylamide gel to PVDF membranes, mem- branes were treated with rabbit anti-rat lung CA IV anti- serum followed with goat anti-rabbit IgG peroxidase con- jugate. A polypeptide of 39 kDa was observed in white muscle sarcolemmal membrane fractions IIA-IIIB, as in- dicated by an open arrowhead (see Fig. 2A). In SR frac- tions SRl and SR2 immunologically active polypeptides of 39 kDa were not observed. In order to check whether a CA II-type CA is present in white muscle sarcolemma and SR, immunoblots were developed using rabbit anti- CA II antiserum followed with goat anti-rabbit IgG-per- oxidase conjugate. The results are shown in Fig. 2B. There

RAT MUSCLE MEMBRANE CARBONIC ANHYDRASE IV 553

TABLE II

Topology of Carbonic Anhydrase in Purified Sarcolemmal Vesicles and Characterization of Solubilized Enzyme

EU ml/mg

Leakiness of vesicles” No saponin + saponin

Accessibility to PI-PLC * No PI-PLC + PI-PLC

Immunoprecipitable CA’ Solubilized CA + preimmune

serum Solubilized CA + antiserum

9.2

10.6

1.1

8.2

% latency -

15 % accessibility

-

88 % precipitable

a.4 5 0.4 95

“Total CA activity measured in the absence and presence of 0.1% saponin.

* Sarcolemmal vesicles before and after PI-PLC treatment for 4 h were centrifuged and the supernatant was assayed for CA activity. Total CA activity was 9.3 EU ml/mg.

’ CA solubilized by PI-PLC was treated with preimmune IgG or anti- IgG followed with Staph A. The immunocomplex was recovered by cen- trifugation. The enzyme activities in the supernatant were determined.

was no immunologically cross-reacting CA II-type poly- peptide (29 kDa) in any membrane preparation. However, a polypeptide of 52 kDa was observed in SR membranes and a small amount also in sarcolemmal fraction IIIB, indicated by the open arrowhead. These results indicate that either the small amount of CA activity in the SR is due to a different CA or the amount of CA IV-type enzyme is too little to be detected by Western blot. The amount of 0.2% SDS-resistant CA in SR is 20- to 40-fold less than that in sarcolemma membrane fractions (see Table I). The minimum amount of CA IV required to detect it on immunoblot under our experimental conditions is around 50 mEU. Therefore, to see a signal for CA IV in SR, one would need to add loo-150 pg of membrane proteins (50- 75 mEU CA). The comparative results of immunoblots of sarcolemmal vesicles (30 pg) and SR (150 and 100 1.18) are shown in Fig. 2C. It can be seen that the signal for CA IV in SR at 150 pg membrane protein is detectable, but several-fold weaker (lane 2) than that produced by 30 pg of sarcolemmal membrane protein (lane 1). The signal for the CA IV polypeptide in SR is barely detectable at 100 pg membrane protein (lane 3). Thus, the failure to detect any cross-reacting polypeptides for CA IV for 30 pg of SR is understandable, and the level of CA IV de- tected in SR may be adequate to explain the low level of CA activity.

Binding to sulfonamide-inhibitor affinity column. Ex- tracts of white muscle sarcolemmal (IIB and IIIB) and SR (SRl and SR2) fractions in 0.2 M Tris-SO1, pH 9.0, containing 1% SDS were applied separately to 0.5 ml of the affinity column. Flow through, high-salt wash, and

A IIA IIB IIIA IIIB SR, SR,

C IIA SRl Protein -

aI) 30 150 100

FIG. 2. Immunochemical detection of CA IV and CA II from rat sar- colemmal and sarcoplasmic reticulum membrane vesicles. Duplicate sarcolemmal vesicles (IIA-IIIB) and SR fractions SRl and SR2 (30 pg of each) were subjected to SDS-PAGE. (A) After electrophoretic transfer of the polypeptides to PVDF membranes one PVDF membrane was incubated with rabbit anti-rat lung CA IV antiserum followed with a second antibody conjugated with peroxidase. Immunoreactive CA IV- type polypeptides of 39 kDa are marked by an open arrowhead. A minor, nonspecific cross-reacting polypeptide of 76 kDa was observed in IIA and IIIA. (B) The other PVDF membrane was incubated with rabbit anti-human CA II antiserum followed with a second antibody conjugated with peroxidase. A 52-kDa polypeptide observed in SR fractions SRl and SR2 and a small amount of the polypeptide are seen in sarcolemmal vesicles (IIIB) which are marked by an open arrowhead. Authentic CA II polypeptides of 29 kDa are not seen in any membrane vesicles. (C) Sarcolemmal vesicles containing 30 fig of membrane proteins (IIA) and SR fraction SRl containing 100 and 150 pg of proteins were subjected to SDS-PAGE followed by immunoblotting using rabbit anti-lung CA IV antiserum and a second antibody conjugated with peroxidase. Im- munoreactive polypeptides of 39 kDa are marked.

554 WAHEED ET AL.

0.5 M sodium perchlorate eluate fractions were analyzed by SDS-PAGE followed with immunoblotting. The re- sults are shown in Fig. 3. From white muscle sarcolemmal vesicles, CA IV-type 39-kDa polypeptide was retained on the affinity column (left panel). The wash with high salt eluted 40% of the 39-kDa polypeptide. The remaining enzyme was eluted with specific inhibitor.

The 52-kDa polypeptide from the SR, which cross- reacts immunologically with rabbit anti-human CA II an- tibodies, was not retained on the affinity column in the presence of 1% SDS. It appeared in the flow-through fractions and wash (right panel). Since 1% SDS inacti- vates several CAs (though not CA IV), sarcoplasmic re- ticulum (SRl and SR2) extract in 0.1% Triton X-100 (which does not inhibit CAs) was applied to the affinity column. The 52-kDa polypeptide was again not retained on the column (results not shown). The facts that the 52- kDa polypeptide protein does not bind to the inhibitor- affinity column to which most CAs bind and that it has no measurable CA activity suggest that it is not a CA.

Red muscle sarcolemmal carbonic anhydrase. The im- munological results in the red muscle sarcolemmal mem- brane are depicted in Fig. 4. As in white muscle mem- branes, the sarcolemmal membrane from red muscle contained an immunoreactive protein with an apparent molecular mass of 39 kDa (see Fig. 4A). Enzyme solubi- lized from red muscle sarcolemmal membranes was re- tained on the sulfonamide-affinity gel and was specifically eluted with CA inhibitor (Fig. 4B), showing that it is a CA.

Vesicle Carbonic Anhydrase Is a GPI-Anchored Membrane Protein

We next studied whether CA could be released from the membranes by PI-PLC, which had been found to re- lease CA IV from human lung and kidney membranes (10). The results are shown in Fig. 5. When 200 pg mem- brane protein of white muscle sarcolemmal vesicle prep- aration IIB was treated with 10 mEU PI-PLC in 200 ~1 reaction mixture at 37°C for 2 h, most of the membrane- associated enzyme was solubilized (see Fig. 5A). The ki- netics of release of CA IV from fraction IIA are depicted in Fig. 5B. After 1 h, more than 50% of the enzyme was solubilized. By 2 h, over 80% was released. By 4 h, essen- tially all of the CA IV was released from the membrane. These results suggest that essentially all of the membrane- associated enzyme which cross-reacts to the rat anti-CA IV antiserum is anchored via a phosphatidylinositol-gly- cosyl linkage.

The nature of the anchor of CA IV to the red muscle sarcolemmal membranes was also characterized by ex- amining solubilization of the enzyme by PI-PLC in the presence and absence of detergent. The results are shown in Fig. 4C. It is apparent from the results that PI-PLC treatment of the red muscle sarcolemmal membranes caused release of the enzyme from the membrane, both

III A+ III 6 SR,+SR*

FIG. 3. Affinity column chromatography of sarcolemmal and sarco- plasmic reticulum membrane extracts. (left) Sarcolemmal vesicle extract was applied to the affinity column; flow through, high-salt wash (wash), and inhibitor eluates (eluate) were pooled and subjected to SDS-PAGE followed by immunoblotting using rabbit anti-rat lung CA IV antiserum. (right) SR membrane extract was applied to the affinity column; flow through, high-salt wash, and inhibitor eluates were pooled and subjected to SDS-PAGE followed by immunoblotting using rabbit anti-human CA II antiserum.

in the presence and in the absence of detergent. These results suggest that CA IV in the red muscle sarcolemma is anchored through a phosphatidylinositol-glycan linkage and that the enzyme is accessible to added PLC without added detergent.

Vesicle Carbonic Anhydrase Is an N-Linked Glycoprotein

Membrane-associated enzyme and the enzyme solu- bilized by PI-PLC were treated with endoglycosidase F to analyze for the presence of N-linked oligosaccharide in the CA IV-type enzyme. The results are shown in Fig. 6. Upon endoglycosidase F treatment, the apparent mo- lecular mass of the membrane-associated or soluble en- zyme from white muscle sarcolemmal fraction IIIA was reduced from 39 to 36 kDa. When the muscle sarcolemma fraction was treated with 0-glycanase which specifically removes O-linked sugar, there was no apparent change in the molecular weight of the polypeptide (results not shown). This result suggested that rat muscle CA IV does not contain O-linked sugar as shown for human lung CA IV (10). A trace amount of 35-kDa polypeptide in the soluble fraction was seen which was not affected by en- doglycosidase F treatment (see Fig. 6, left panel, second and fourth lanes). The 35-kDa fraction may be due to proteolytic cleavage and oligosaccharide removal during isolation and solubilization of the membrane preparation, resulting in removal of the oligosaccharide moiety together with 1 kDa of polypeptide. When soluble and membrane-

RAT MUSCLE MEMBRANE

A. B. II III Fr W ELUATES

--- AB AB

C. + PI-PLC D. CONTROL - DOC + DOC PN Gase F

SM SM s M - +

FIG. 4. Immunological detection of CA IV from rat red muscle sar- colemmal vesicles. (A) Sarcolemmal membrane vesicles (IIA-IIIB) (30 ng of each) were subjected to SDS-PAGE followed by immunoblotting using anti-rat lung CA IV antiserum and a second antibody conjugated with peroxidase. The apparent molecular mass of immunoreactive CA IV-type polypeptides is marked. (B) Affinity column chromatography of sarcolemmal membrane extracts. Membrane extracts from red muscle sarcolemmal vesicles were applied to the affinity column; flow through (FT), high-salt wash (W), and inhibitor eluates (eluates) were subjected to SDS-PAGE followed by immunoblotting. (C) Treatment of sarco- lemma1 membrane vesicles of PI-PLC. Sarcolemmal vesicle membranes were treated with buffer alone (control) and with PI-PLC in the absence (-) and presence (+) of 0.1% deoxycholate (DOC) at 37°C for 2 h. Soluble (S) and membrane-associated (M) enzymes were recovered by centrifugation and were analyzed by SDS-PAGE followed by immu- noblotting. (D) Deglycosylation of CA IV-type enzyme from sarcolemmal membranes. Sarcolemmal membrane vesicles were treated with buffer alone (-) and PNGase F (+). Deglycosylation was analyzed by SDS- PAGE followed with immunoblotting. The apparent molecular mass of the deglycosylated polypeptide is marked.

bound enzymes from sarcolemmal vesicle fraction IIIB were treated with endoglycosidase F, the apparent mo- lecular mass of 39 kDa of the polypeptide was reduced into two polypeptides of 36 and 35 kDa, respectively. The PI-PLC-released soluble form of the enzyme from fraction IIIB contained more than 80% of the polypeptide as 35 kDa. The membrane-associated enzyme was reduced to 36- and 35.kDa polypeptides in equal amounts by endo- glycosidase treatment. These results suggest that proteo- lytic nicking of the polypeptides is more apparent in sar- colemmal fraction IIIB.

The presence of N-linked carbohydrate chains in the red mu.& sarcolemmal enzyme is shown in Fig. 4D. En-

CARBONIC ANHYDRASE IV 555

doglycosidase PNGase F treatment resulted in a decrease of apparent molecular mass of the CA IV polypeptide from 39 to 35 kDa, suggesting that the CA IV-type enzyme in red muscle sarcolemma contains one oligosaccharide chain as was found with white muscle sarcolemmal en- zyme.

Nearly All Vesicle-Associated Carbonic Anhydrase Is Due to CA IV

Essentially all of the white muscle sarcolemmal vesicle- associated enzyme was solubilized by PI-PLC (Table II, PI-PLC treatment), suggesting that (a) nearly all of the CA activity is anchored to the membrane by PI-PLC- sensitive linkage and (b) the enzyme is accessible to added PI-PLC without added detergent. This means either that the CA IV-type enzyme is present on the surface of the vesicles or that the vesicles are freely permeable to the PI-PLC and the released CA IV. Since all of the mem- brane-associated CA can be solubilized with PI-PLC, we immunoprecipitated the solubilized enzyme using rat lung CA IV antiserum. Enzyme activity remaining in the su- pernatant after immunoprecipitation would be evidence for an immunologically nonidentical CA. Bat lung CA IV antiserum was able to immunoprecipitate essentially all

A. CONTROL PI-PLC

2h lh 2h

SM SMSM

B. S M

TlME(min) 0 15 30 60 120 240 240

FIG. 5. Treatment of sarcolemmal membrane vesicles with phos- phoinositol specific phospholipase C. Sarcolemmal membrane vesicles IIB (A) or IIIB (B) were treated with PI-PLC for different times as indicated. Soluble (S) and membrane-associated (M) enzymes were re- covered by centrifugation. Both soluble and membrane-bound enzymes were analyzed by SDS-PAGE followed by immunoblotting.

556 WAHEED ET AL.

III A III s 3.

MS MS SM SM 4.

_ _ + + _ _ + + 5.

6.

-39 7. 136

35 8.

9.

FIG. 6. Deglycosylation of CA IV-type enzyme from sarcolemmal vesicles. Membrane-associated (M) or soluble (S) enzymes from low- 10. density sarcolemmal membranes IIIA (left) and high-density sarcolem- ma1 membranes IIIB (right) were treated with buffer alone (-) and

l1 ’

PNGase F (+). The deglycosylation reaction was analyzed by SDS- PAGE followed by immunoblotting. Apparent molecular masses of the 12. glycosylated and deglycosylated polypeptides are marked.

13.

of the CA activity. Only 5% was precipitated by preim- 14. mune serum (Table II, “Immunoprecipitable CA”). This 15. result demonstrates that most of the CA activity is due to a CA IV-type enzyme. 16.

ACKNOWLEDGMENTS 17.

We thank Dr. Seiji Sato for helpful discussion and Elizabeth Torno for typing and editing the manuscript. 18.

19. REFERENCES 20.

1. Geers, C., Gros, G., and Gartner, A. (1985) J. Appl. Physiol. 69, 548-558. 21.

2. Ridderstrale, Y. (1979) Actu Physiol. Scud. 106, 239-240.

Lonnerholm, G. (1980) J. Histochem. Cytochem. 28,427-433.

Riley, D. A., Ellis, S., and Bain, J. (1982) J. Histochem. Cytochem. 30,1275-1288.

Moyle, S., Jeffery, S., and Carter, N. D. (1984) J. Histochem. Cy- to&em. 32,1262-1264.

Dermietzel, R., Leibstein, A., Siffert, W., Zamboglou, N., and Gros, G. (1985) J. Histochem. Cytochem. 33,93-98.

Bruns, W., Dermietzel, R., and Gros, G. (1986) J. Physiol. 371, 351-364.

Wetzel, P., and Gros, G. (1990) Arch. Biochem. Biophys. 279,345- 354.

Whitney, P. L., and Briggle, T. V. (1982) J. Biol. Chem. 267,12,056- 12,059.

Zhu, X. L., and Sly, W. S. (1990) J. Bid. Chem. 266, 87958801.

Wistrand, P. J., and Knuuttila, K. G. (1989) Kidney Int. 35, 851- 859.

Brown, D., Zhu, X. L., and Sly, W. S. (1990) Proc. Nutl. Acad. Sci. USA 87,7457-7461.

Waheed, A., Zhu, X. L., and Sly, W. S. (1992) J. Biol. Chem., in press.

Seiler, S., and Fleischer, S. (1982) J. BioL Chem. 257,13,862-13,871.

Allain, C. C., Poon, L. S., Chan, C. S. G., Richmond, W., and Fu, P. C. (1974) Clin. Chem. 20,470-475.

Sato, S., Zhu, X. L., and Sly, W. S. (1990) Proc. Natl. Acad. Sci. USA 87,6073-6076.

Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275.

Peterson, G. L. (1979) And. Biochem. 100, 201-220.

Laemmli, U. K. (1970) Nature 227,680-685.

Rosenblatt, M., Hidalgo, C., Vergara, C., and Ikemoto, N. (1981) J. Biol. Chem. 256,8140-8148.

Hidalgo, C., Gonzalez, M. E., and Lagos, R. (1983) J. Biol. Chem. 268, 13,937-13,945.