Biochimica et Biophysica Acta, 1160(1992) 293-300
293
© 1992 Elsevier Science Publishers B.V. All rights reserved 0167-4838/92/$05.00
BBAPRO 34335
Characterization of cardiac Na + / C a 2 +-exchange by site-directed polyclonal antibodies Calvin C. Hale, Steven B. Kleiboeker and Victoria Burton Ochoa Department of Veterinary Biomedical Sciences and the John M. Dalton Research Center, Unicersity of Missouri, Columbia, MO (USA)
(Received 25 February 1992) (Revised manuscript received 24 June 1992)
Key words: Na+/Ca2+-exchange;Cardiac sarcolemmalvesicle; Site-directed polyclonalantibody; Protein structure Cardiac Na+/Ca2+-exchange is an integral membrane protein consisting of approx. 970 amino acids with as many as 12 membrane-spanning and 11 extramembranal regions (Nicoll, D.A., Lognoni, S. and Philipson, K.D. (1985) Science 250, 562-565). Based upon primary sequence information, 3 amino-acid sequences located in either extramembranal segment a or f, consisting largely of acidic amino-acids, were selected for the production of synthetic peptides. The peptides were cross-linked to carrier ovalbumin and used to generate site-directed polyclonal antibodies (sd-Ab). Western blot analysis of bovine cardiac sarcolemmal (SL) proteins demonstrated that sd-Ab against segment a and 1 against loop f recognized a 70 kDa protein and a lower molecular mass band at 55 kDa under reducing conditions. A different loop f sd-Ab failed to recognize the 70 kDa protein but did associate with a 120, 65 and 55 kDa protein under the same conditions. Under non-reducing conditions, antibodies to all three peptides recognized the 65 kDa protein. All sd-Ab were blocked by addition of their respective peptides and were not inhibited by either of the other peptides. A sd-Ab against loop f was immobilized to an affinity support matrix and used to immunoprecipitate detergent solubilized cardiac SL vesicle protein. Immunoprecipitated protein was reconstituted into proteoliposomes which demonstrated Na+/CaZ+-exchange activity. Immunoprecipitated protein cross-reacted with sd-Ab against all three peptides with bands at 120, 70 and 55 kDa on Western blots. Tryptic digests of native SL vesicles abolished recognition of segment a sd-Ab for SL proteins while having little or no affect on reactivity to the protein by both sd-Ab against loop f. Digestion of the SL vesicle protein with endoproteinase Arg C did not alter sd-Ab recognition. The results suggest that specific domains of the cardiac Na+/CaZ+-exchanger depending upon the conformation of the protein, may not be available for antibody binding. The 70 kDa polypeptide appears to include the N-terminal region of the protein and what is believed to be a large cytoplasmic extramembranal loop. Limited proteolysis by trypsin and endoproteinase Arg C yielded results consistent with the model which places the N-terminus of the protein on the extracellular surface and a large extramembranal segment (loop f) on the cytoplasmic side of the SL membrane.
Introduction In cardiac myocytes, the plasma membrane or sarcolemma (SL) contains a number of Ca2+-shuttling mechanisms that play central roles in Ca 2÷ homeostasis and, thus, cardiac contractility. An important regulator of cardiac contractility during diastole is Na+/Ca2+-exchange. Recent work by Philipson and co-workers described the cloning and functional expression of cardiac Na+/Ca2+-exchange [6]. This report included information on the primary sequence of the Na+/Ca2+-exchange protein and a preliminary structural model.
Correspondence to: C.C. Hale, Department of Veterinary Biomedical Sciences and the John M. Dalton Research Center, University of Missouri, Columbia, MO 65211, USA.
A hydropathy plot of the primary sequence reported by Philipson and co-workers suggested that cardiac Na+/CaZ+-exchange may have as many as 12 transmembranal regions and 11 or 12 extramembranal loops and segments. A large extramembranal loop consisting of approx. 520 amino acids (over 1 / 2 of the entire protein), is located between transmembranal segments 5 and 6. Thus, the 2-dimensional picture of the cardiac Na+/CaZ+-exchanger based upon these data suggest a nearly symmetrical protein, in terms of membranespanning regions, separated by a large extramembranal loop. The function of these domains and their role in cation binding and transmembranal ion-transport is not known. Using this structural model as the basis for the synthesis of antigenic peptides, we have developed site-directed polyclonal antibodies (sd-Ab) against three extramembranal regions of the cardiac Na+/CaZ+-ex-
294 changer. In this report, we describe the initial characterization of these sd-Ab. Our data suggest that antibody recognition of various regions of the cardiac Na+/Ca2+-exchange protein can vary with the state of reduction and thus physical conformation of the protein. The previously described 70 kDa protein may be generated from the N-terminus of the protein and include the majority of the large cytoplasmic loop. Sd-Ab analysis of trypsin- and endoproteinase Arg C-treated SL vesicle protein is consistent with the structural model of Philipson and co-workers in which the N-terminus of the protein is located on the opposite side of the membrane from what appears to be a large cytoplasmic loop. Materials and Methods
Preparation of cardiac sarcolemmal vesicles Bovine cardiac sarcolemmal (SL) vesicles were prepared by the procedure of Kuwayama and Kanazawa [2], as modified by Slaughter, Sutko and Reeves [9]. Bovine ventricular tissue was obtained fresh from a local abattoir, trimmed to remove endocardium and epicardium prior to homogenization. The final vesicle product was suspended in 160 mM NaCl, 20 mM Mops adjusted to a final pH of 7.4 with Tris (buffer A). Vesicles were maintained at -70°C prior to use. Protein values were determined by the method of Lowry et al. [4]. Na+/Ca2+-exchange was measured as previously described [8].
Preparation of site-directed polyclonal antibodies For the preparation of site-directed polyclonal antibodies (sd-Ab) to the cardiac Na+/CaZ+-exchange protein, peptides based upon the canine amino-acid sequence of Nicoll, Longoni and Philipson [6] were synthesized (Fig. 1). All peptides used in this study were synthesized and sequenced to assure purity by the Peptide Synthesis Laboratory at the University of Kentucky. Peptides were cross-linked to ovalbumin (20:1 molar ratio) using bis(sulfosuccinimidyl) suberate (BS 3) by the following procedure: Peptide was added to a 1 ml solution of ovalbumin (1 mg/ml) in 150 mM NaC1, 10 mM sodium phosphate (pH 7.4) (PBS) at room temperature. BS 3 was added to a final concentration of 10 mM followed by mixing. After 2 h, the mixture was dialyzed in three 2.0-1 changes of PBS at 4°C for a minimum of 36 h. The cross-linked peptide-ovalbumin complex was monitored by SDS-PAGE to confirm a decreased mobility compared to non-cross-linked ovalbumin. The peptide-ovalbumin complex was mixed 1 : 1 (v/v) with Freund's complete adjuvant and injected subcutaneously into multiple dorsal sites in female New Zealand white rabbits. Rabbits received booster inoculations at 4, 6, 8, 10 and 12 weeks as described above using Freund's incomplete adjuvant. Whole
blood was collected by ear bleeds and allowed to clot. Serum was separated by low speed centrifugation. Serum was heated to 56°C for 15 min to destroy compliment activity and stored either at 4°C or -70°C until use.
Immunoprecipitation and reconstitution of cardiac Na +/Ca 2 +-exchange actiuity The IgG fraction of crude serum against peptide 2 was immobilized to a support matrix which was then used to immunoprecipitate cardiac SL protein which in turn were reconstituted into proteoliposomes. Purified immune IgG was prepared by fractionating 5 ml of crude serum on a 1.5 x 25 cm CM Affigel blue column according to the manufacturers instructions (Bio-Rad; Bulletin 1092). The purity of the IgG preparation was verified by SDS-PAGE. Purified IgG was mixed 1:1 (v/v) with a saturated ammonium sulfate solution and stirred overnight at 4°C. The protein fraction was pelleted (48 000 X g, 1 h) and dialyzed (2 x 1 litre) against 150 mM NaC1, 100 mM sodium acetate (pH 5.5), coupling buffer). IgG was coupled to Affigei Hz Hydrazide immunoaffinity support matrix according to the manufacturers instructions (Bio-Rad; Bulletin 1424). The immobilized IgG matrix was stored at 4°C in buffer A containing 0.05% sodium azide. Approx. 3 mg of cardiac SL vesicle protein was pelleted by centrifugation and resuspended in 2 ml of buffer A containing 1% (v/v) Triton X-100. After 0.5 h on ice, this mixture was subjected to centrifugation at 200000 x g for 0.5 h at 4°C. The resulting supernatant fluid, containing detergent solubilized SL vesicle proteins, was removed and separated into 2 fractions (control and immunoprecipitate fraction). The immunoprecipitate fraction was combined with 0.5 ml of immobilized IgG matrix that had been previously washed in buffer A containing 1% Triton X-100. This mixture and control sample (no additions) were gently rotated for 2 h at room temperature. The immobilized IgG matrix was separated from unbound (and therefore unprecipitated) SL proteins by a low-speed centrifugation in a clinical centrifuge (250 X g). The IgG matrix pellet was washed twice by resuspension in 15 ml of buffer A containing 1% Triton X-100. The final pellet was resuspended in 2 ml of elution buffer which consisted of 2% sodium cholate in 1 M ammonium hydroxide (pH 10.0), containing 25 mg/ml soybean phospholipids (Asolectin; Associated Concentrates) which eluted bound SL proteins from the Ab matrix. After 10 min of gently rotating this mixture at room temperature, the matrix was again pelleted by low-speed centrifugation. The supernatant fluid was collected and diluted into 10 ml of ice-cold buffer A followed by centrifugation overnight at 200000 x g at 4°C. The resulting proteoliposome pellet was washed 2 more times in buffer A (200000 x g, 1 h). The final pellet
295 was resuspended in 1.0 ml of buffer A. Control and unprecipitated SL proteins (that had not bound to the IgG matrix) were reconstituted as follows: Soybean phospholipids in buffer A containing 1% Triton X-100 were added to each sample to a final concentration of 25 mg/ml. After 0.5 h on ice, each sample received the addition of approximately 0.1 g of detergent absorbent beads (Bio-Beads SM-2, Bio-Rad). This mixture was gently rocked overnight at 4°C. After removing the absorbent beads by low speed centrifugation, the samples were diluted 10-fold into ice-cold buffer A and centrifuged at 200 000 × g at 4°C for 1 h. The resulting proteoliposome pellet was washed two more times as described above and finally resuspended in 1.0 ml of buffer A. Proteoliposomes were maintained at 4°C until assayed. Protein concentration of reconstituted proteoliposomes was determined by an Amido-black staining procedure as previously reported [1].
Electrophoretic and blotting conditions The Tris-glycine system of Laemmli was used for electrophoresis in 0.1% SDS with 8.5% acrylamide in the resolving gel. Each lane was loaded with 30 Izg of vesicle protein. Following SDS-PAGE, SL proteins were transblotted (Western blot) onto nitrocellulose for 1 h at 100 V in a buffer containing 25 mM Tris, 192 mM glycine, 20% methanol (v/v) (pH 8.3). Visualization of SL proteins recognized by sd-Ab was accomplished use of a second antibody conjugated to alkaline phosphatase (Bio-Rad). Blotted SL proteins were rinsed briefly in 500 mM NaC1, 20 mM Tris (pH 7.5) (TBS) and blocked for 30 min in TBS containing 3% gelatin followed by a 5 min wash in 0.05% Tween 20 in TBS (TTBS). Sd-Ab (crude serum) was diluted 1 : 1000 in 1% gelatin-TTBS and incubated with blot strips for 1 h with gentle agitation followed by two 5-min washes with TTBS. Blot strips were incubated with the second antibody (1 h) and subjected to alkaline phosphatase color development according to the manufacturers instructions (Bio-Rad). Chemicals and reagents Endoproteinase Arg-C (mouse submaxillary gland) was purchased from Calbiochem. Trypsin (bovine pancreas) was purchased from Sigma. Soybean trypsin inhibitor was purchased from United States Biochemical Corporation. Results
Three non-overlapping peptides were synthesized based upon the reported primary sequence of canine cardiac Na+/CaZ+-exchange [6]. Peptide sequences are shown in Fig. 1. The three peptides represent regions of the protein which include the N-terminus (extramembranal segment a) and 2 areas of a large pur-
peptide
G
-
1
E
-
peptide I
-
2 D
-
peptide
A
-
E
D
D
3
-
T
-
-
loop
-
D
-
loop
-
D
-
loop
E
-
F
-
D
D
-
D
-
E
-
C
-
G
-
E
-
F
-
E
-
E
-
D
-
E
-
N
E
- G - E -
-
E
F
-
M
-
I
A
-
G - N -
E - T
Fig. 1. Synthetic peptides used for the production of site-directed polyclonal antibodies. Shown are amino-acid sequences used for the synthesis of peptides. Peptides were conjugated to ovalbumin and injected into rabbits for antibody production as described in Materials and Methods.
ported cytoplasmic region (loop f). The approximate location of these peptides on the cardiac Na+/Ca 2+exchange protein is indicated in Fig. 2. Site-directed polyclonal antibodies (sd-Ab) were raised against ovalbumin conjugated peptides as described in Materials and Methods. Rabbit sera were screened using bovine cardiac SL proteins on Western blots. Pre-immune sera failed to recognize any bovine cardiac SL protein (not shown). As shown on Fig. 3, antibodies capable of recognizing bovine cardiac SL protein were raised against each peptide. When SDSPAGE was performed under reducing conditions (5% (v/v)/3-mercaptoethanol and boiling samples for 3 min prior to electrophoresis), sd-Ab against peptides 2 and 3 (Fig. 3, lanes 4 and 6, respectively) recognized a protein at approx. 70 kDa and a lower band at 55 kDa. A polyclonal antibody raised against purified canine cardiac Na+/Ca2+-exchange recognized a 70 kDa bovine protein as well as 2 higher M r proteins of 120 and > 200 kDa (not shown). In contrast, sd-Ab against peptide 1 weakly recognized a protein slightly lower than 70 kDa as well as at 120 and 55 kDa (Fig. 3, lane 2). The relationship between the 70 kDa protein recognized by sd-Ab in Fig. 3 (lanes 4 and 6) and the slightly smaller 65 kDa protein recognized by sd-Ab to peptide 1 (lane 2) was further explored. Antibody recognition was tested under non-reducing SDS-PAGE conditions. In this experiment, equal amounts of bovine cardiac SL vesicle protein were identically prepared, except that the reducing agent /3-mercaptoethanol was excluded and samples were not boiled prior to SDS-PAGE. Sd-Ab against peptides 2 and 3 (Fig. 3, lanes 3 and 5) detected a downward shift in the molecular mass of the 70 kDa protein to that of the protein recognized by sd-Ab against peptide 1. Removal of reducing agent also resulted in an increase in intensity of the 65 kDa band recognized by sd-Ab against peptide 1 and lack of recognition of the 120 kDa band (Fig. 3, lane 1). These data suggest that the three sd-Ab recognize the same protein, that the ability to migrate during SDS-PAGE
296
kDa
is affected by the proteins conformation (reduced vs. non-reduced), and that antibody recognition by Western blot analysis is influenced by the proteins conformational state. In other experiments, peptides used to raise sd-Ab blocked recognition of their respective sd-Ab whereas either of the other 2 peptides had little or no affect at the same concentration. The specificity of each sd-Ab for its respective peptide is shown in Fig. 4. For all three sd-Ab, the peptide used to generate the antibody response blocked recognition to both the 70 and 55 kDa protein, whereas the other 2 peptides failed to completely block recognition at the same concentration (1 mg/ml peptide). These data demon-
-200
-92.5 -69
-46
lane
1
2
3
4
5
6
Fig. 3. Western blot analysis of bovine cardiac SL proteins with antibodies raised to synthetic peptides. 30/zg of bovine cardiac SL vesicle protein were subject to SDS-PAGE followed by electrophoretic transfer (Western blot) to nitrocellulose. Blotted proteins were probed with sera and visualized as described in Materials and Methods. Antibody to peptide 1, lanes 1-2; Antibody to peptide 2, lanes 3-4; Antibody to peptide 3, lanes 5-6. SL proteins in lanes 1, 3, and 5 were run under nonreducing SDS-PAGE conditions. SL proteins in lanes 2, 4 and 6 were run under reducing SDS-PAGE conditions as described in the text.
A
a =
r,d~b
Ex,tracellular
~
.,
..
~ r ~
B Fig. 2. Structural model of cardiac Na+/Ca2+-exchange showing position of synthesized peptides and proteolytic cleavage sites. A model of the cardiac Na+/Ca2+-exchange protein based upon the work of Nicoll, Longoni and Philipson [61 is shown. The size of loop f is drawn to correspond to the total number of amino acids located in this segment. The positions of areas corresponding to the respective synthesized peptides for sd-Ab production (Fig. 1) are indicated. Hatched lines indicate extramembranal proteolytic sites for (A) trypsin and (B) endoproteinase Arg C.
strate the high specificity of these antibodies and suggest the 55 kDa protein is related to the 70 kDa protein. Experiments such as those shown in Figs. 3 and 4, while useful in demonstrating that the three sd-Ab appear to recognize the same protein, do not provide direct evidence that the 70 and 55 kDa proteins are Na+/CaZ+-exchange derived proteins. In order to ascertain if this was the case, we prepared an immobilized antibody matrix for the purpose of immunoprecipitation and reconstitution of activity. For these experiments, we utilized sd-Ab against peptide 2. A typical immunoprecipitation and reconstitution experiment is shown in Table I. Immunoprecipitated SL proteins were eluted from the IgG matrix with a sodium cholate, high-pH buffer and reconstituted by detergent dilution. Control and unprecipitated SL proteins, which were in Triton X-100, were reconstituted through removal of detergent by absorbent beads. Notwithstanding differences in technique, compared to control values, the Ab matrix precipitated and facilitated the reconstitution of 80-90% of the Na+/Ca2+-exchange activity. Ab matrix prepared from pre-immune serum did not immunoprecipitate Na+/CaZ+-exchange activity. Proteins precipitated by the Ab matrix described in Table I were visualized on Western blots by all three
297
kDa
A
B
C
D
92.569463021.5lane
1
2
3
1
2
3
1
2
3
1
2
3
Fig. 4. Specificity of site-directed polyclonal antibodies. The ability of sd-Ab to recognize bovine cardiac SL proteins on Western blots was performed as in Fig. 3, except that synthetic peptides (Fig. 1) were included during the primary antibody incubation. For panels A, B, and C, lanes 1, 2, and 3 contained 0, 0.25 and 1.0 mg/ml, respectively, of the indicated peptide. Blot strips in panel D contained 1.0 mg/ml of the indicated peptide. Panel A, sd-Ab 1 plus peptide 1; panel B, sd-Ab 2 plus peptide 2; panel C, sd-Ab 3 plus peptide 3; panel D, lane 1, sd-Ab 1 plus peptide 3; panel D, lane 2, sd-Ab 2 plus peptide 1; panel D, lane 3, sd-Ab 3 plus peptide 2.
sd-Ab. These results, shown in Fig. 5, indicate that all three sd-Ab reacted to the same i m m u n o p r e c i p i t a t e d proteins. F u r t h e r m o r e , following immunoprecipitation and reconstitution, all three sd-Ab recognized a b a n d at approx. 120 kDa, in addition to proteins at 70 and 55 kDa. Bands at 120 k D a and 55 k D a were markedly c o n c e n t r a t e d by immunoprecipitation. In other experiments, the I g G matrix Specifically i m m u n o p r e c i p i t a t e d a 70 k D a protein that had b e e n crosslinked to radiolabeled exchange inhibitory peptide [5], a specific in-
kDa
hibitor of cardiac N a + / C a 2 + - e x c h a n g e (unpublished data). W e further tested the structural model shown in Fig. 2 by subjecting SL vesicles to mild digestion with either trypsin or e n d o p r o t e i n a s e Arg C. As shown in Fig. 2, the primary sequence of cardiac N a + / C a 2 + - e x change has many potential extramembranal trypsin cleavage sites, most of which are located on what has been postulated to be the cytoplasmic side of the protein. Prior to loop i, however, there is only one
A
B
C
200
92.5 69
46
30
lane
1
2
1
2
1
2
Fig. 5. Western blot analysis of immunoprecipitated SL proteins. Bovine cardiac SL proteins were immunoprecipitated with sd-Ab against peptide 2 that had been immobilized to a support matrix as described in Materials and Methods. Immunoprecipitated proteins were visualized by Western blot analysis. Panels A, B and C were probed with sd-Ab against peptides 1, 2, and 3, respectively. Lane 1 contained reconstituted bovine SL vesicle protein (control). Lane 2 contained reconstituted, immunoprecipitated proteins.
298 C
A
kDa
kDa
-200
-200
-92.5 -69
C
0
2
4
6
....
-200 -92.5 -69
-92.5 -69
.46
.46
Time (min)
kDa
-
-46
Time (min)
C
0
2
4
6
Time (min)
C
0
2
4
6
Fig. 6. Time-dependent effect of trypsin digestion on antibody recognition of cardiac SL protein. Bovine cardiac SL vesicles were subjected to trypsin-treatment (0.25 mg trypsin/mg SL protein) at 37°C for the times indicated. Trypsin inhibitor (0.5 mg/mg trypsin) was added at the endpoint of each reaction. Antibody recognition of SL protein was determined under reducing SDS-PAGE conditions as described in Fig. 3 and the text. Panel A, antibody to peptide 1; Panel B, antibody to peptide 2; Panel C, antibody to peptide 3. In each panel, lane C represents a trypsin- and trypsin inhibitor-free control.
extramembranal trypsin site available on the external surface, which is located on segment a. T h e segment-a trypsin site is downstream to the sd-Ab site on that segment. It is possible, therefore, that a mild trypsin digestion of right side out vesicles would result in a single cleavage point in segment a with no other trypsin-mediated cleavages prior to loop i. O n the other hand, N a + / C a Z + - e x c h a n g e protein in inside out vesicles would be d e g r a d e d into n u m e r o u s small fragments by trypsin. In contrast, the model would predict that e n d o p r o t e i n a s e A r g C, a proteinase that cleaves polypeptides at the carboxy-side of arginine, would have 'external' cleavage sites only in loop k. Western blot analysis using sd-Ab as in Fig. 3 was p e r f o r m e d on trypsin-treated and e n d o p r o t e i n a s e A r g C-treated SL vesicles with the results shown in Figs. 6 and 7. Sd-Ab against loop f reacted with trypsin-treated exchange protein while recognition by sd-Ab against segment a
A
was rapidly and totally abolished. Digestion by endoproteinase Arg C did not affect recognition of N a + / C a 2 + - e x c h a n g e protein by any of the 3 sd-Ab. T h e results are therefore consistent with the model shown in Fig. 2 which places the N-terminus of the protein and the large extramembranal loop f on opposite sides of the membrane.
Discussion T h e cloning and identification of the cardiac N a + / C a 2 + - e x c h a n g e was made possible in part by the use of a polyclonal antibody that was raised by inoculating rabbits with reconstituted proteoliposomes containing highly purified N a + / C a 2 + - e x c h a n g e protein [6,7]. That antibody, which recognized related canine SL proteins of 70, 120 and 160 kDa, was used to identify a c D N A clone coding for the canine cardiac
B
C
kDa -200
iii
-92.5 -69
C 0.33
1 4.5
kDa
-200
-200 -92.5 -69
-92.5 -69
-46
.46
Time (hrs)
kDa
-46
Time (hr$)
C 0.33
1 4.5
Time (hrs)
C 0.33
1 4.5
Fig. 7. Time-dependent effect of endoproteinase Arg C on antibody recognition of cardiac SL protein. Bovine cardiac SL vesicles were subjected to endoproteinase Arg C-treatment (0.1 unit endoproteinase Arg C//xg SL protein) at 37°C for the times indicated. Antibody recognition of SL protein was determined under reducing SDS-PAGE conditions as described in Fig. 3 and the text. Panel A, antibody to peptide 1; Panel B, antibody to peptide 2; Panel C, antibody to peptide 3. In each panel, lane C represents an endoproteinase Arg C-free control.
299 TABLE I Immunoprecipitation and reconstitution exchange
of cardiac
Na +/ Ca 2 +-
SL vesicle proteins were detergent solubilized, immunoprecipitated by sd-Ab immobilized to a support matrix and reconstituted into proteoliposomes as described in Materials and Methods. Na+/Ca 2+exchange specific activity (S.A.) is nmol CaZ+/mg protein per s. Sample
Protein Protein Na+/Ca 2+ Na+/Ca 2+ recovered recovered exch. exch. (/~g) (% control) (S.A.) recovered (% control)
Reconstituted SL Protein (control)
129
100
1.06
100
Reconstituted Immunoprecipitated SL Protein
34
26
3.76
93
Reconstituted Unprecipitated 100
78
0.28
20
Na+/Ca2+-exchange. RNA transcribed from the isolated clone was shown to code for Na+/CaZ+-exchange by expression in Xenopus oocytes. We have developed sd-Ab based upon the deduced amino acid sequence of that clone. In the present report, we describe the initial characterization of three sd-Ab with cardiac N a + / Ca 2+-exchange. Peptides were synthesized based upon the aminoacid sequence of the canine cardiac Na+/CaZ+-exchanger. Sd-Ab raised against the primary sequence of canine cardiac Na+/CaZ+-exchange react with what are apparently comparable proteins from the bovine heart, suggesting that sequence and size homology exists for this transport protein between these two species. Sd-Ab against peptide 2 was immobilized to a bead matrix and used to precipitate SL proteins. Precipitated proteins were reconstituted and assayed for Na+/Ca2+-exchange activity. Results of these experiments, such as those shown in Table I, are a direct demonstration that the sd-Ab are specific for and can interact with the cardiac Na+/CaZ+-exchanger. While we were successful in precipitating Na+/Ca2+-exchange activity in the presence of the nonionic detergent Triton X-100, we were unsuccessful in reconstituting activity from the Ab matrix, except in the presence of a sodium cholate, high-pH buffer. Therefore, while immunoprecipitation and reconstitution experiments were useful in making the connection between the sd-Ab and the Na+/CaZ+-exchanger, differences in the reconstitution procedures employed made quantitative analysis difficult. Different reconstitution precedures were utilized out of necessity. A nonionic detergent was best sutited for the precipitation of activity because it did not interfere with the antibody-antigen
complex. Once precipitated, however, we used an ionic detergent to help facilitate the disruption of the complex, so that the released membrane protein could be reconstituted. Changing detergents in order to manipulate the Na+/Ca2+-exchange protein has been previously reported [7]. Low immunoprecipitate protein recoveries (that were near the limits of the protein assay) coupled with variable protein recoveries associated with the two different reconstitution methods used also made comparative quantitative analysis difficult. Under reducing conditions, sd-Ab against peptides 2 and 3 appear to recognize the same 70 kDa protein that is recognized by polyclonal antibodies produced by Philipson and co-workers [7] against canine cardiac Na+/Ca2+-exchange (data not shown). Sd-Ab against peptide 1 on the other hand recognize a protein with an apparent molecular mass of 65 kDa under reducing conditions. When a conformational difference is imposed by excluding the reducing agent from the electrophoresis procedure, all three sd-Ab react with the 65 kDa protein. Under nonreducing conditions, sd-Ab against peptide 1 react more strongly with the 65 kDa protein. This may reflect the increased amount of nonreduced protein present. The weak recognition observed under reducing conditions by this sd-Ab may be due to relatively small amount of Na+/CaZ+-exchange protein that either is not entirely reduced or for some unknown reason can not be reduced. In any event, it appears that antigenic regions recognized by sd-Ab against peptide 1 may not be available when the protein is fully reduced. Sd-Ab against peptide 2 and 3 inconsistently recognized higher molecular mass forms of the Na+/Ca 2+exchanger. The 70 kDa protein has been described as an active proteolytic fragment of the 120 kDa protein [7]. If the SL vesicle preparations used in the present study were substantially proteolytically degraded, there should be minimal recognition of proteins higher than 70 kDa. Proteolytic degradation of the SL preparations seems unlikely however as canine polyclonal antibody (not shown) and sd-Ab against peptide 1 did recognize higher M r proteins (Fig. 3, lane 2). Furthermore, immunoprecipitation and reconstitution of SL proteins, a process that facilitates concentrating specific proteins, revealed that all three sd-Ab recognized a 120 kDa protein (Fig. 5), a protein first reported by Philipson and co-workers [7] to be associated with cardiac Na+/CaZ+-exchange. Thus, the avidity of individual antibodies toward the antigen coupled with relatively lower concentrations of the higher M r species may contribute to inconsistent recognition. All three sd-Ab recognize a protein at 55 kDa that has not been previously described as a Na+/CaZ+-exchange protein. Antibody association with the 55 kDa protein was inhibited by addition of their respective peptides and was present in immunoprecipitates, sug-
300 gesting this protein is related to the higher M r proteins. Due to the location of sd-Ab sites on the Na+/Ca2+-exchange (as shown in Fig. 2), it would appear that this protein must contain at a minimum, the portion of the protein from the N-terminus through loop f. Knowledge of the primary sequence of cardiac Na+/Ca2÷-exchange allowed Philipson and his coworkers to develop a preliminary structural model of the protein with regards to its orientation in the membrane [6]. The model shown in Fig. 2 differs in that the area originally thought to be m e m b r a n e spanning region 1 is missing in the mature protein and therefore has been deleted (Philipson, K.D., personal communication) and loop f has been proportionately increased to more closely represent its size in relationship with the entire protein. As shown in Fig. 2, sd-Ab to peptide 3 is directed to the N-terminus of the protein and sd-Ab to peptide 1 is directed to the C-terminus side of cytoplasmic loop f. As these sd-Ab recognize the same protein, then the 70 kDa proteolytic fragment is likely to be generated from the N-terminus side of a cleavage point distal to the peptide 1 region. This would further suggest that the proteolytic cleavage(s) that generates the 70 kDa protein is not likely to result from a 'symmetrical' cutting the protein in the middle of loop f. The structural model in Fig. 2 is further supported by the trypsin and endoproteinase Arg C digestion experiments shown in Figs. 6 and 7. The model predicts that trypsin digestion of right side out vesicles would generate a large fragment which would include all but a small portion of the N-terminus region and all of loop f. As a consequence, sd-Ab recognition of segment a would be abolished while the 2 sd-Ab raised to loop f sequences would still detect Na+/Ca2+-ex change protein, which is what was observed. In the case of digestion with endoproteinase Arg C, all three sd-Ab should continue to react with right-side out vesicle protein because the antibody recognition site on segment a would remain intact. This is also what was observed. These data, therefore, suggest that segment a and loop f are on opposite sides of the m e m b r a n e and as we and others have implied, that the orientation or sidedness is as shown in Fig. 2. The alternative possibility is that loop f faces out of the m e m b r a n e and that segment a is located on the cytoplasmic side. While the results in the present study do not directly refute this possibility, we feel that this is unlikely in part because of the location of a potential calmodulin
binding site on loop f and glycosylation sites on loops a, c and i [6]. It is important to note that the strength of the signal observed by both loop f sd-Ab was decreased following proteolytic digestion. This is an expected result, as any recognition (antibody binding) present due to inside-out vesicles in these preparations would not be observed following proteinase treatment. Because there was little, if any, size reduction in the 70 kDa protein recognized by sd-Ab to loop f, it is possible that the proteolytic site which generates the 70 kDa protein from the 120 kDa protein is located in or near loop i. A problem with this interpretation is that the cleaved C-terminus remainder of the protein would not appear sufficiently large enough to compensate for the additional apparent size of the 120 kDa protein observed on SDS-PAGE. One possibility is that the protein in question migrates anomalously under different conditions a n d / o r conformations. For example, as shown in Fig. 3, under nonreducing conditions, sd-Ab against peptide 1 does not recognize the 120 kDa protein recognized in the presence of a reducing agent. These issues can unfortunately not be totally resolved by the present study.
Acknowledgements The authors wish to thank T.F. Thompson, II and Scott Bliler for expert technical assistance, Mike Russ for peptide synthesis, and Dr. K.D. Philipson for the gift of polyclonal antibody used for comparative purposes at the onset of this study. This work was funded by the American Heart Association - Missouri Affiliate.
References 1 Hale, C.C., Slaughter, R.S., Ahrens, D.C. and Reeves, J.P. (1984) Proc. Natl. Acad. Sci. 81, 6569-6573. 2 Kuwayama,H. and Kanazawa, T. (1982) J. Biochem. 91, 1419-1426. 3 Laemmli, U.K. (1970) Nature 227, 680-685. 4 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 5 Li, Z., Nicoll, D.A., Collins, A., Hilgemann, D.W., Filoteo, A.G., Penniston, J.T., Weiss, J.N., Tomich, J.M. and Philipson, K.D. (1991) J. Biol. Chem. 266, 1014-1020. 6 Nicoll, D.A., Longoni, S. and Philipson, K.D. (1990) Science 250, 562-565. 7 Philipson, K.D., Longoni, S. and Ward, R. (1988) Biochim. Biophys. Acta 945, 298-306. 8 Reeves, J.P. and Sutko, J.L. (1983) J. Biol. Chem. 258, 3178-3182. 9 Slaughter, R.S., Sutko, J.L. and Reeves, J.P. (1983) J. Biol. Chem. 258, 3183-3190.
293
© 1992 Elsevier Science Publishers B.V. All rights reserved 0167-4838/92/$05.00
BBAPRO 34335
Characterization of cardiac Na + / C a 2 +-exchange by site-directed polyclonal antibodies Calvin C. Hale, Steven B. Kleiboeker and Victoria Burton Ochoa Department of Veterinary Biomedical Sciences and the John M. Dalton Research Center, Unicersity of Missouri, Columbia, MO (USA)
(Received 25 February 1992) (Revised manuscript received 24 June 1992)
Key words: Na+/Ca2+-exchange;Cardiac sarcolemmalvesicle; Site-directed polyclonalantibody; Protein structure Cardiac Na+/Ca2+-exchange is an integral membrane protein consisting of approx. 970 amino acids with as many as 12 membrane-spanning and 11 extramembranal regions (Nicoll, D.A., Lognoni, S. and Philipson, K.D. (1985) Science 250, 562-565). Based upon primary sequence information, 3 amino-acid sequences located in either extramembranal segment a or f, consisting largely of acidic amino-acids, were selected for the production of synthetic peptides. The peptides were cross-linked to carrier ovalbumin and used to generate site-directed polyclonal antibodies (sd-Ab). Western blot analysis of bovine cardiac sarcolemmal (SL) proteins demonstrated that sd-Ab against segment a and 1 against loop f recognized a 70 kDa protein and a lower molecular mass band at 55 kDa under reducing conditions. A different loop f sd-Ab failed to recognize the 70 kDa protein but did associate with a 120, 65 and 55 kDa protein under the same conditions. Under non-reducing conditions, antibodies to all three peptides recognized the 65 kDa protein. All sd-Ab were blocked by addition of their respective peptides and were not inhibited by either of the other peptides. A sd-Ab against loop f was immobilized to an affinity support matrix and used to immunoprecipitate detergent solubilized cardiac SL vesicle protein. Immunoprecipitated protein was reconstituted into proteoliposomes which demonstrated Na+/CaZ+-exchange activity. Immunoprecipitated protein cross-reacted with sd-Ab against all three peptides with bands at 120, 70 and 55 kDa on Western blots. Tryptic digests of native SL vesicles abolished recognition of segment a sd-Ab for SL proteins while having little or no affect on reactivity to the protein by both sd-Ab against loop f. Digestion of the SL vesicle protein with endoproteinase Arg C did not alter sd-Ab recognition. The results suggest that specific domains of the cardiac Na+/CaZ+-exchanger depending upon the conformation of the protein, may not be available for antibody binding. The 70 kDa polypeptide appears to include the N-terminal region of the protein and what is believed to be a large cytoplasmic extramembranal loop. Limited proteolysis by trypsin and endoproteinase Arg C yielded results consistent with the model which places the N-terminus of the protein on the extracellular surface and a large extramembranal segment (loop f) on the cytoplasmic side of the SL membrane.
Introduction In cardiac myocytes, the plasma membrane or sarcolemma (SL) contains a number of Ca2+-shuttling mechanisms that play central roles in Ca 2÷ homeostasis and, thus, cardiac contractility. An important regulator of cardiac contractility during diastole is Na+/Ca2+-exchange. Recent work by Philipson and co-workers described the cloning and functional expression of cardiac Na+/Ca2+-exchange [6]. This report included information on the primary sequence of the Na+/Ca2+-exchange protein and a preliminary structural model.
Correspondence to: C.C. Hale, Department of Veterinary Biomedical Sciences and the John M. Dalton Research Center, University of Missouri, Columbia, MO 65211, USA.
A hydropathy plot of the primary sequence reported by Philipson and co-workers suggested that cardiac Na+/CaZ+-exchange may have as many as 12 transmembranal regions and 11 or 12 extramembranal loops and segments. A large extramembranal loop consisting of approx. 520 amino acids (over 1 / 2 of the entire protein), is located between transmembranal segments 5 and 6. Thus, the 2-dimensional picture of the cardiac Na+/CaZ+-exchanger based upon these data suggest a nearly symmetrical protein, in terms of membranespanning regions, separated by a large extramembranal loop. The function of these domains and their role in cation binding and transmembranal ion-transport is not known. Using this structural model as the basis for the synthesis of antigenic peptides, we have developed site-directed polyclonal antibodies (sd-Ab) against three extramembranal regions of the cardiac Na+/CaZ+-ex-
294 changer. In this report, we describe the initial characterization of these sd-Ab. Our data suggest that antibody recognition of various regions of the cardiac Na+/Ca2+-exchange protein can vary with the state of reduction and thus physical conformation of the protein. The previously described 70 kDa protein may be generated from the N-terminus of the protein and include the majority of the large cytoplasmic loop. Sd-Ab analysis of trypsin- and endoproteinase Arg C-treated SL vesicle protein is consistent with the structural model of Philipson and co-workers in which the N-terminus of the protein is located on the opposite side of the membrane from what appears to be a large cytoplasmic loop. Materials and Methods
Preparation of cardiac sarcolemmal vesicles Bovine cardiac sarcolemmal (SL) vesicles were prepared by the procedure of Kuwayama and Kanazawa [2], as modified by Slaughter, Sutko and Reeves [9]. Bovine ventricular tissue was obtained fresh from a local abattoir, trimmed to remove endocardium and epicardium prior to homogenization. The final vesicle product was suspended in 160 mM NaCl, 20 mM Mops adjusted to a final pH of 7.4 with Tris (buffer A). Vesicles were maintained at -70°C prior to use. Protein values were determined by the method of Lowry et al. [4]. Na+/Ca2+-exchange was measured as previously described [8].
Preparation of site-directed polyclonal antibodies For the preparation of site-directed polyclonal antibodies (sd-Ab) to the cardiac Na+/CaZ+-exchange protein, peptides based upon the canine amino-acid sequence of Nicoll, Longoni and Philipson [6] were synthesized (Fig. 1). All peptides used in this study were synthesized and sequenced to assure purity by the Peptide Synthesis Laboratory at the University of Kentucky. Peptides were cross-linked to ovalbumin (20:1 molar ratio) using bis(sulfosuccinimidyl) suberate (BS 3) by the following procedure: Peptide was added to a 1 ml solution of ovalbumin (1 mg/ml) in 150 mM NaC1, 10 mM sodium phosphate (pH 7.4) (PBS) at room temperature. BS 3 was added to a final concentration of 10 mM followed by mixing. After 2 h, the mixture was dialyzed in three 2.0-1 changes of PBS at 4°C for a minimum of 36 h. The cross-linked peptide-ovalbumin complex was monitored by SDS-PAGE to confirm a decreased mobility compared to non-cross-linked ovalbumin. The peptide-ovalbumin complex was mixed 1 : 1 (v/v) with Freund's complete adjuvant and injected subcutaneously into multiple dorsal sites in female New Zealand white rabbits. Rabbits received booster inoculations at 4, 6, 8, 10 and 12 weeks as described above using Freund's incomplete adjuvant. Whole
blood was collected by ear bleeds and allowed to clot. Serum was separated by low speed centrifugation. Serum was heated to 56°C for 15 min to destroy compliment activity and stored either at 4°C or -70°C until use.
Immunoprecipitation and reconstitution of cardiac Na +/Ca 2 +-exchange actiuity The IgG fraction of crude serum against peptide 2 was immobilized to a support matrix which was then used to immunoprecipitate cardiac SL protein which in turn were reconstituted into proteoliposomes. Purified immune IgG was prepared by fractionating 5 ml of crude serum on a 1.5 x 25 cm CM Affigel blue column according to the manufacturers instructions (Bio-Rad; Bulletin 1092). The purity of the IgG preparation was verified by SDS-PAGE. Purified IgG was mixed 1:1 (v/v) with a saturated ammonium sulfate solution and stirred overnight at 4°C. The protein fraction was pelleted (48 000 X g, 1 h) and dialyzed (2 x 1 litre) against 150 mM NaC1, 100 mM sodium acetate (pH 5.5), coupling buffer). IgG was coupled to Affigei Hz Hydrazide immunoaffinity support matrix according to the manufacturers instructions (Bio-Rad; Bulletin 1424). The immobilized IgG matrix was stored at 4°C in buffer A containing 0.05% sodium azide. Approx. 3 mg of cardiac SL vesicle protein was pelleted by centrifugation and resuspended in 2 ml of buffer A containing 1% (v/v) Triton X-100. After 0.5 h on ice, this mixture was subjected to centrifugation at 200000 x g for 0.5 h at 4°C. The resulting supernatant fluid, containing detergent solubilized SL vesicle proteins, was removed and separated into 2 fractions (control and immunoprecipitate fraction). The immunoprecipitate fraction was combined with 0.5 ml of immobilized IgG matrix that had been previously washed in buffer A containing 1% Triton X-100. This mixture and control sample (no additions) were gently rotated for 2 h at room temperature. The immobilized IgG matrix was separated from unbound (and therefore unprecipitated) SL proteins by a low-speed centrifugation in a clinical centrifuge (250 X g). The IgG matrix pellet was washed twice by resuspension in 15 ml of buffer A containing 1% Triton X-100. The final pellet was resuspended in 2 ml of elution buffer which consisted of 2% sodium cholate in 1 M ammonium hydroxide (pH 10.0), containing 25 mg/ml soybean phospholipids (Asolectin; Associated Concentrates) which eluted bound SL proteins from the Ab matrix. After 10 min of gently rotating this mixture at room temperature, the matrix was again pelleted by low-speed centrifugation. The supernatant fluid was collected and diluted into 10 ml of ice-cold buffer A followed by centrifugation overnight at 200000 x g at 4°C. The resulting proteoliposome pellet was washed 2 more times in buffer A (200000 x g, 1 h). The final pellet
295 was resuspended in 1.0 ml of buffer A. Control and unprecipitated SL proteins (that had not bound to the IgG matrix) were reconstituted as follows: Soybean phospholipids in buffer A containing 1% Triton X-100 were added to each sample to a final concentration of 25 mg/ml. After 0.5 h on ice, each sample received the addition of approximately 0.1 g of detergent absorbent beads (Bio-Beads SM-2, Bio-Rad). This mixture was gently rocked overnight at 4°C. After removing the absorbent beads by low speed centrifugation, the samples were diluted 10-fold into ice-cold buffer A and centrifuged at 200 000 × g at 4°C for 1 h. The resulting proteoliposome pellet was washed two more times as described above and finally resuspended in 1.0 ml of buffer A. Proteoliposomes were maintained at 4°C until assayed. Protein concentration of reconstituted proteoliposomes was determined by an Amido-black staining procedure as previously reported [1].
Electrophoretic and blotting conditions The Tris-glycine system of Laemmli was used for electrophoresis in 0.1% SDS with 8.5% acrylamide in the resolving gel. Each lane was loaded with 30 Izg of vesicle protein. Following SDS-PAGE, SL proteins were transblotted (Western blot) onto nitrocellulose for 1 h at 100 V in a buffer containing 25 mM Tris, 192 mM glycine, 20% methanol (v/v) (pH 8.3). Visualization of SL proteins recognized by sd-Ab was accomplished use of a second antibody conjugated to alkaline phosphatase (Bio-Rad). Blotted SL proteins were rinsed briefly in 500 mM NaC1, 20 mM Tris (pH 7.5) (TBS) and blocked for 30 min in TBS containing 3% gelatin followed by a 5 min wash in 0.05% Tween 20 in TBS (TTBS). Sd-Ab (crude serum) was diluted 1 : 1000 in 1% gelatin-TTBS and incubated with blot strips for 1 h with gentle agitation followed by two 5-min washes with TTBS. Blot strips were incubated with the second antibody (1 h) and subjected to alkaline phosphatase color development according to the manufacturers instructions (Bio-Rad). Chemicals and reagents Endoproteinase Arg-C (mouse submaxillary gland) was purchased from Calbiochem. Trypsin (bovine pancreas) was purchased from Sigma. Soybean trypsin inhibitor was purchased from United States Biochemical Corporation. Results
Three non-overlapping peptides were synthesized based upon the reported primary sequence of canine cardiac Na+/CaZ+-exchange [6]. Peptide sequences are shown in Fig. 1. The three peptides represent regions of the protein which include the N-terminus (extramembranal segment a) and 2 areas of a large pur-
peptide
G
-
1
E
-
peptide I
-
2 D
-
peptide
A
-
E
D
D
3
-
T
-
-
loop
-
D
-
loop
-
D
-
loop
E
-
F
-
D
D
-
D
-
E
-
C
-
G
-
E
-
F
-
E
-
E
-
D
-
E
-
N
E
- G - E -
-
E
F
-
M
-
I
A
-
G - N -
E - T
Fig. 1. Synthetic peptides used for the production of site-directed polyclonal antibodies. Shown are amino-acid sequences used for the synthesis of peptides. Peptides were conjugated to ovalbumin and injected into rabbits for antibody production as described in Materials and Methods.
ported cytoplasmic region (loop f). The approximate location of these peptides on the cardiac Na+/Ca 2+exchange protein is indicated in Fig. 2. Site-directed polyclonal antibodies (sd-Ab) were raised against ovalbumin conjugated peptides as described in Materials and Methods. Rabbit sera were screened using bovine cardiac SL proteins on Western blots. Pre-immune sera failed to recognize any bovine cardiac SL protein (not shown). As shown on Fig. 3, antibodies capable of recognizing bovine cardiac SL protein were raised against each peptide. When SDSPAGE was performed under reducing conditions (5% (v/v)/3-mercaptoethanol and boiling samples for 3 min prior to electrophoresis), sd-Ab against peptides 2 and 3 (Fig. 3, lanes 4 and 6, respectively) recognized a protein at approx. 70 kDa and a lower band at 55 kDa. A polyclonal antibody raised against purified canine cardiac Na+/Ca2+-exchange recognized a 70 kDa bovine protein as well as 2 higher M r proteins of 120 and > 200 kDa (not shown). In contrast, sd-Ab against peptide 1 weakly recognized a protein slightly lower than 70 kDa as well as at 120 and 55 kDa (Fig. 3, lane 2). The relationship between the 70 kDa protein recognized by sd-Ab in Fig. 3 (lanes 4 and 6) and the slightly smaller 65 kDa protein recognized by sd-Ab to peptide 1 (lane 2) was further explored. Antibody recognition was tested under non-reducing SDS-PAGE conditions. In this experiment, equal amounts of bovine cardiac SL vesicle protein were identically prepared, except that the reducing agent /3-mercaptoethanol was excluded and samples were not boiled prior to SDS-PAGE. Sd-Ab against peptides 2 and 3 (Fig. 3, lanes 3 and 5) detected a downward shift in the molecular mass of the 70 kDa protein to that of the protein recognized by sd-Ab against peptide 1. Removal of reducing agent also resulted in an increase in intensity of the 65 kDa band recognized by sd-Ab against peptide 1 and lack of recognition of the 120 kDa band (Fig. 3, lane 1). These data suggest that the three sd-Ab recognize the same protein, that the ability to migrate during SDS-PAGE
296
kDa
is affected by the proteins conformation (reduced vs. non-reduced), and that antibody recognition by Western blot analysis is influenced by the proteins conformational state. In other experiments, peptides used to raise sd-Ab blocked recognition of their respective sd-Ab whereas either of the other 2 peptides had little or no affect at the same concentration. The specificity of each sd-Ab for its respective peptide is shown in Fig. 4. For all three sd-Ab, the peptide used to generate the antibody response blocked recognition to both the 70 and 55 kDa protein, whereas the other 2 peptides failed to completely block recognition at the same concentration (1 mg/ml peptide). These data demon-
-200
-92.5 -69
-46
lane
1
2
3
4
5
6
Fig. 3. Western blot analysis of bovine cardiac SL proteins with antibodies raised to synthetic peptides. 30/zg of bovine cardiac SL vesicle protein were subject to SDS-PAGE followed by electrophoretic transfer (Western blot) to nitrocellulose. Blotted proteins were probed with sera and visualized as described in Materials and Methods. Antibody to peptide 1, lanes 1-2; Antibody to peptide 2, lanes 3-4; Antibody to peptide 3, lanes 5-6. SL proteins in lanes 1, 3, and 5 were run under nonreducing SDS-PAGE conditions. SL proteins in lanes 2, 4 and 6 were run under reducing SDS-PAGE conditions as described in the text.
A
a =
r,d~b
Ex,tracellular
~
.,
..
~ r ~
B Fig. 2. Structural model of cardiac Na+/Ca2+-exchange showing position of synthesized peptides and proteolytic cleavage sites. A model of the cardiac Na+/Ca2+-exchange protein based upon the work of Nicoll, Longoni and Philipson [61 is shown. The size of loop f is drawn to correspond to the total number of amino acids located in this segment. The positions of areas corresponding to the respective synthesized peptides for sd-Ab production (Fig. 1) are indicated. Hatched lines indicate extramembranal proteolytic sites for (A) trypsin and (B) endoproteinase Arg C.
strate the high specificity of these antibodies and suggest the 55 kDa protein is related to the 70 kDa protein. Experiments such as those shown in Figs. 3 and 4, while useful in demonstrating that the three sd-Ab appear to recognize the same protein, do not provide direct evidence that the 70 and 55 kDa proteins are Na+/CaZ+-exchange derived proteins. In order to ascertain if this was the case, we prepared an immobilized antibody matrix for the purpose of immunoprecipitation and reconstitution of activity. For these experiments, we utilized sd-Ab against peptide 2. A typical immunoprecipitation and reconstitution experiment is shown in Table I. Immunoprecipitated SL proteins were eluted from the IgG matrix with a sodium cholate, high-pH buffer and reconstituted by detergent dilution. Control and unprecipitated SL proteins, which were in Triton X-100, were reconstituted through removal of detergent by absorbent beads. Notwithstanding differences in technique, compared to control values, the Ab matrix precipitated and facilitated the reconstitution of 80-90% of the Na+/Ca2+-exchange activity. Ab matrix prepared from pre-immune serum did not immunoprecipitate Na+/CaZ+-exchange activity. Proteins precipitated by the Ab matrix described in Table I were visualized on Western blots by all three
297
kDa
A
B
C
D
92.569463021.5lane
1
2
3
1
2
3
1
2
3
1
2
3
Fig. 4. Specificity of site-directed polyclonal antibodies. The ability of sd-Ab to recognize bovine cardiac SL proteins on Western blots was performed as in Fig. 3, except that synthetic peptides (Fig. 1) were included during the primary antibody incubation. For panels A, B, and C, lanes 1, 2, and 3 contained 0, 0.25 and 1.0 mg/ml, respectively, of the indicated peptide. Blot strips in panel D contained 1.0 mg/ml of the indicated peptide. Panel A, sd-Ab 1 plus peptide 1; panel B, sd-Ab 2 plus peptide 2; panel C, sd-Ab 3 plus peptide 3; panel D, lane 1, sd-Ab 1 plus peptide 3; panel D, lane 2, sd-Ab 2 plus peptide 1; panel D, lane 3, sd-Ab 3 plus peptide 2.
sd-Ab. These results, shown in Fig. 5, indicate that all three sd-Ab reacted to the same i m m u n o p r e c i p i t a t e d proteins. F u r t h e r m o r e , following immunoprecipitation and reconstitution, all three sd-Ab recognized a b a n d at approx. 120 kDa, in addition to proteins at 70 and 55 kDa. Bands at 120 k D a and 55 k D a were markedly c o n c e n t r a t e d by immunoprecipitation. In other experiments, the I g G matrix Specifically i m m u n o p r e c i p i t a t e d a 70 k D a protein that had b e e n crosslinked to radiolabeled exchange inhibitory peptide [5], a specific in-
kDa
hibitor of cardiac N a + / C a 2 + - e x c h a n g e (unpublished data). W e further tested the structural model shown in Fig. 2 by subjecting SL vesicles to mild digestion with either trypsin or e n d o p r o t e i n a s e Arg C. As shown in Fig. 2, the primary sequence of cardiac N a + / C a 2 + - e x change has many potential extramembranal trypsin cleavage sites, most of which are located on what has been postulated to be the cytoplasmic side of the protein. Prior to loop i, however, there is only one
A
B
C
200
92.5 69
46
30
lane
1
2
1
2
1
2
Fig. 5. Western blot analysis of immunoprecipitated SL proteins. Bovine cardiac SL proteins were immunoprecipitated with sd-Ab against peptide 2 that had been immobilized to a support matrix as described in Materials and Methods. Immunoprecipitated proteins were visualized by Western blot analysis. Panels A, B and C were probed with sd-Ab against peptides 1, 2, and 3, respectively. Lane 1 contained reconstituted bovine SL vesicle protein (control). Lane 2 contained reconstituted, immunoprecipitated proteins.
298 C
A
kDa
kDa
-200
-200
-92.5 -69
C
0
2
4
6
....
-200 -92.5 -69
-92.5 -69
.46
.46
Time (min)
kDa
-
-46
Time (min)
C
0
2
4
6
Time (min)
C
0
2
4
6
Fig. 6. Time-dependent effect of trypsin digestion on antibody recognition of cardiac SL protein. Bovine cardiac SL vesicles were subjected to trypsin-treatment (0.25 mg trypsin/mg SL protein) at 37°C for the times indicated. Trypsin inhibitor (0.5 mg/mg trypsin) was added at the endpoint of each reaction. Antibody recognition of SL protein was determined under reducing SDS-PAGE conditions as described in Fig. 3 and the text. Panel A, antibody to peptide 1; Panel B, antibody to peptide 2; Panel C, antibody to peptide 3. In each panel, lane C represents a trypsin- and trypsin inhibitor-free control.
extramembranal trypsin site available on the external surface, which is located on segment a. T h e segment-a trypsin site is downstream to the sd-Ab site on that segment. It is possible, therefore, that a mild trypsin digestion of right side out vesicles would result in a single cleavage point in segment a with no other trypsin-mediated cleavages prior to loop i. O n the other hand, N a + / C a Z + - e x c h a n g e protein in inside out vesicles would be d e g r a d e d into n u m e r o u s small fragments by trypsin. In contrast, the model would predict that e n d o p r o t e i n a s e A r g C, a proteinase that cleaves polypeptides at the carboxy-side of arginine, would have 'external' cleavage sites only in loop k. Western blot analysis using sd-Ab as in Fig. 3 was p e r f o r m e d on trypsin-treated and e n d o p r o t e i n a s e A r g C-treated SL vesicles with the results shown in Figs. 6 and 7. Sd-Ab against loop f reacted with trypsin-treated exchange protein while recognition by sd-Ab against segment a
A
was rapidly and totally abolished. Digestion by endoproteinase Arg C did not affect recognition of N a + / C a 2 + - e x c h a n g e protein by any of the 3 sd-Ab. T h e results are therefore consistent with the model shown in Fig. 2 which places the N-terminus of the protein and the large extramembranal loop f on opposite sides of the membrane.
Discussion T h e cloning and identification of the cardiac N a + / C a 2 + - e x c h a n g e was made possible in part by the use of a polyclonal antibody that was raised by inoculating rabbits with reconstituted proteoliposomes containing highly purified N a + / C a 2 + - e x c h a n g e protein [6,7]. That antibody, which recognized related canine SL proteins of 70, 120 and 160 kDa, was used to identify a c D N A clone coding for the canine cardiac
B
C
kDa -200
iii
-92.5 -69
C 0.33
1 4.5
kDa
-200
-200 -92.5 -69
-92.5 -69
-46
.46
Time (hrs)
kDa
-46
Time (hr$)
C 0.33
1 4.5
Time (hrs)
C 0.33
1 4.5
Fig. 7. Time-dependent effect of endoproteinase Arg C on antibody recognition of cardiac SL protein. Bovine cardiac SL vesicles were subjected to endoproteinase Arg C-treatment (0.1 unit endoproteinase Arg C//xg SL protein) at 37°C for the times indicated. Antibody recognition of SL protein was determined under reducing SDS-PAGE conditions as described in Fig. 3 and the text. Panel A, antibody to peptide 1; Panel B, antibody to peptide 2; Panel C, antibody to peptide 3. In each panel, lane C represents an endoproteinase Arg C-free control.
299 TABLE I Immunoprecipitation and reconstitution exchange
of cardiac
Na +/ Ca 2 +-
SL vesicle proteins were detergent solubilized, immunoprecipitated by sd-Ab immobilized to a support matrix and reconstituted into proteoliposomes as described in Materials and Methods. Na+/Ca 2+exchange specific activity (S.A.) is nmol CaZ+/mg protein per s. Sample
Protein Protein Na+/Ca 2+ Na+/Ca 2+ recovered recovered exch. exch. (/~g) (% control) (S.A.) recovered (% control)
Reconstituted SL Protein (control)
129
100
1.06
100
Reconstituted Immunoprecipitated SL Protein
34
26
3.76
93
Reconstituted Unprecipitated 100
78
0.28
20
Na+/Ca2+-exchange. RNA transcribed from the isolated clone was shown to code for Na+/CaZ+-exchange by expression in Xenopus oocytes. We have developed sd-Ab based upon the deduced amino acid sequence of that clone. In the present report, we describe the initial characterization of three sd-Ab with cardiac N a + / Ca 2+-exchange. Peptides were synthesized based upon the aminoacid sequence of the canine cardiac Na+/CaZ+-exchanger. Sd-Ab raised against the primary sequence of canine cardiac Na+/CaZ+-exchange react with what are apparently comparable proteins from the bovine heart, suggesting that sequence and size homology exists for this transport protein between these two species. Sd-Ab against peptide 2 was immobilized to a bead matrix and used to precipitate SL proteins. Precipitated proteins were reconstituted and assayed for Na+/Ca2+-exchange activity. Results of these experiments, such as those shown in Table I, are a direct demonstration that the sd-Ab are specific for and can interact with the cardiac Na+/CaZ+-exchanger. While we were successful in precipitating Na+/Ca2+-exchange activity in the presence of the nonionic detergent Triton X-100, we were unsuccessful in reconstituting activity from the Ab matrix, except in the presence of a sodium cholate, high-pH buffer. Therefore, while immunoprecipitation and reconstitution experiments were useful in making the connection between the sd-Ab and the Na+/CaZ+-exchanger, differences in the reconstitution procedures employed made quantitative analysis difficult. Different reconstitution precedures were utilized out of necessity. A nonionic detergent was best sutited for the precipitation of activity because it did not interfere with the antibody-antigen
complex. Once precipitated, however, we used an ionic detergent to help facilitate the disruption of the complex, so that the released membrane protein could be reconstituted. Changing detergents in order to manipulate the Na+/Ca2+-exchange protein has been previously reported [7]. Low immunoprecipitate protein recoveries (that were near the limits of the protein assay) coupled with variable protein recoveries associated with the two different reconstitution methods used also made comparative quantitative analysis difficult. Under reducing conditions, sd-Ab against peptides 2 and 3 appear to recognize the same 70 kDa protein that is recognized by polyclonal antibodies produced by Philipson and co-workers [7] against canine cardiac Na+/Ca2+-exchange (data not shown). Sd-Ab against peptide 1 on the other hand recognize a protein with an apparent molecular mass of 65 kDa under reducing conditions. When a conformational difference is imposed by excluding the reducing agent from the electrophoresis procedure, all three sd-Ab react with the 65 kDa protein. Under nonreducing conditions, sd-Ab against peptide 1 react more strongly with the 65 kDa protein. This may reflect the increased amount of nonreduced protein present. The weak recognition observed under reducing conditions by this sd-Ab may be due to relatively small amount of Na+/CaZ+-exchange protein that either is not entirely reduced or for some unknown reason can not be reduced. In any event, it appears that antigenic regions recognized by sd-Ab against peptide 1 may not be available when the protein is fully reduced. Sd-Ab against peptide 2 and 3 inconsistently recognized higher molecular mass forms of the Na+/Ca 2+exchanger. The 70 kDa protein has been described as an active proteolytic fragment of the 120 kDa protein [7]. If the SL vesicle preparations used in the present study were substantially proteolytically degraded, there should be minimal recognition of proteins higher than 70 kDa. Proteolytic degradation of the SL preparations seems unlikely however as canine polyclonal antibody (not shown) and sd-Ab against peptide 1 did recognize higher M r proteins (Fig. 3, lane 2). Furthermore, immunoprecipitation and reconstitution of SL proteins, a process that facilitates concentrating specific proteins, revealed that all three sd-Ab recognized a 120 kDa protein (Fig. 5), a protein first reported by Philipson and co-workers [7] to be associated with cardiac Na+/CaZ+-exchange. Thus, the avidity of individual antibodies toward the antigen coupled with relatively lower concentrations of the higher M r species may contribute to inconsistent recognition. All three sd-Ab recognize a protein at 55 kDa that has not been previously described as a Na+/CaZ+-exchange protein. Antibody association with the 55 kDa protein was inhibited by addition of their respective peptides and was present in immunoprecipitates, sug-
300 gesting this protein is related to the higher M r proteins. Due to the location of sd-Ab sites on the Na+/Ca2+-exchange (as shown in Fig. 2), it would appear that this protein must contain at a minimum, the portion of the protein from the N-terminus through loop f. Knowledge of the primary sequence of cardiac Na+/Ca2÷-exchange allowed Philipson and his coworkers to develop a preliminary structural model of the protein with regards to its orientation in the membrane [6]. The model shown in Fig. 2 differs in that the area originally thought to be m e m b r a n e spanning region 1 is missing in the mature protein and therefore has been deleted (Philipson, K.D., personal communication) and loop f has been proportionately increased to more closely represent its size in relationship with the entire protein. As shown in Fig. 2, sd-Ab to peptide 3 is directed to the N-terminus of the protein and sd-Ab to peptide 1 is directed to the C-terminus side of cytoplasmic loop f. As these sd-Ab recognize the same protein, then the 70 kDa proteolytic fragment is likely to be generated from the N-terminus side of a cleavage point distal to the peptide 1 region. This would further suggest that the proteolytic cleavage(s) that generates the 70 kDa protein is not likely to result from a 'symmetrical' cutting the protein in the middle of loop f. The structural model in Fig. 2 is further supported by the trypsin and endoproteinase Arg C digestion experiments shown in Figs. 6 and 7. The model predicts that trypsin digestion of right side out vesicles would generate a large fragment which would include all but a small portion of the N-terminus region and all of loop f. As a consequence, sd-Ab recognition of segment a would be abolished while the 2 sd-Ab raised to loop f sequences would still detect Na+/Ca2+-ex change protein, which is what was observed. In the case of digestion with endoproteinase Arg C, all three sd-Ab should continue to react with right-side out vesicle protein because the antibody recognition site on segment a would remain intact. This is also what was observed. These data, therefore, suggest that segment a and loop f are on opposite sides of the m e m b r a n e and as we and others have implied, that the orientation or sidedness is as shown in Fig. 2. The alternative possibility is that loop f faces out of the m e m b r a n e and that segment a is located on the cytoplasmic side. While the results in the present study do not directly refute this possibility, we feel that this is unlikely in part because of the location of a potential calmodulin
binding site on loop f and glycosylation sites on loops a, c and i [6]. It is important to note that the strength of the signal observed by both loop f sd-Ab was decreased following proteolytic digestion. This is an expected result, as any recognition (antibody binding) present due to inside-out vesicles in these preparations would not be observed following proteinase treatment. Because there was little, if any, size reduction in the 70 kDa protein recognized by sd-Ab to loop f, it is possible that the proteolytic site which generates the 70 kDa protein from the 120 kDa protein is located in or near loop i. A problem with this interpretation is that the cleaved C-terminus remainder of the protein would not appear sufficiently large enough to compensate for the additional apparent size of the 120 kDa protein observed on SDS-PAGE. One possibility is that the protein in question migrates anomalously under different conditions a n d / o r conformations. For example, as shown in Fig. 3, under nonreducing conditions, sd-Ab against peptide 1 does not recognize the 120 kDa protein recognized in the presence of a reducing agent. These issues can unfortunately not be totally resolved by the present study.
Acknowledgements The authors wish to thank T.F. Thompson, II and Scott Bliler for expert technical assistance, Mike Russ for peptide synthesis, and Dr. K.D. Philipson for the gift of polyclonal antibody used for comparative purposes at the onset of this study. This work was funded by the American Heart Association - Missouri Affiliate.
References 1 Hale, C.C., Slaughter, R.S., Ahrens, D.C. and Reeves, J.P. (1984) Proc. Natl. Acad. Sci. 81, 6569-6573. 2 Kuwayama,H. and Kanazawa, T. (1982) J. Biochem. 91, 1419-1426. 3 Laemmli, U.K. (1970) Nature 227, 680-685. 4 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 5 Li, Z., Nicoll, D.A., Collins, A., Hilgemann, D.W., Filoteo, A.G., Penniston, J.T., Weiss, J.N., Tomich, J.M. and Philipson, K.D. (1991) J. Biol. Chem. 266, 1014-1020. 6 Nicoll, D.A., Longoni, S. and Philipson, K.D. (1990) Science 250, 562-565. 7 Philipson, K.D., Longoni, S. and Ward, R. (1988) Biochim. Biophys. Acta 945, 298-306. 8 Reeves, J.P. and Sutko, J.L. (1983) J. Biol. Chem. 258, 3178-3182. 9 Slaughter, R.S., Sutko, J.L. and Reeves, J.P. (1983) J. Biol. Chem. 258, 3183-3190.
