Immunochemical comparison of cardiac glycoside-sensitive (lamb) and -insensitive (rat) kidney (Na+ + K+)-ATPase

Biochimica et Biophysica Acta 873 (1986) 79-87 Elsevier

79

BBA 32602

Immunochemical comparison of cardiac glycoside-sensitive (lamb) and - i n s e n s i t i v e (rat) k i d n e y ( N a + + K + ) - A T P a s e W i l l i a m J. Ball, Jr. a n d L o i s K . L a n e Department of Pharmacology and Cell Biophysics, University of Cincinnati College of Medicine, 231 Bethesda A venue, Cincinnati, OH 45267-0575 (U.S.A.)

(Received May 13th, 1986)

Key words: ( N a + + K +)-ATPase; Cardiac glycoside; lmmunochemistry; Monoclonal antibody

The immunological cross-reactivity of the ouabain-sensitive lamb kidney and the ouabain-insensitive rat kidney ( N a + + K+)-ATPase (EC 3.6.1.37) was examined using polyclonal and monoclonal antibodies. Studies using rabbit antisera prepared against both the lamb kidney and rat kidney holoenzymes showed the existence of substantial antigenic differences as well as similarities between the holoenzymes and the respective denatured a and fl subunits of these two enzymes. Quantitation of the extent of cross-reactivity using holoenzyme-directed antibodies showed a 40-60% cross-reactivity. In addition, rabbit antisera monospecific to the purified, denatured a and /~ subunits of the lamb kidney enzyme showed about a 50% cross-reactivity towards the respective subunit of the rat enzyme. In contrast to the cross-reactivity observed using the polyclonal antibodies, six monoclonal antibodies specific for the a subunit of the lamb holoenzyme exhibited no cross-reactivity with the rat holoenzyme. Four of these monoclonal antibodies, however, showed substantial cross-reactivity with rat a subunit as resolved by SDS-polyacrylamide gel electrophoresis. A fifth antibody did not bind to the denatured a subunit of either the lamb or the rat enzyme. Another monoclonal antibody (M7-PB-E9), which is specific for an epitope previously implicated in the regulation of both ATP and ouabain binding to (Na + + K +)-ATPase (Ball, W.J., Jr. (1984) Biochemistry 2275-2281) was found to bind to the denatured lamb a but not to the rat a. This antibody has identified a region of the lamb a that has an altered amino acid sequence in the ouabain-insensitive rat enzyme. These immunological studies indicate that there are substantial antigenic differences between the lamb and rat kidney ( N a + + K+) ATPases. The majority of these antigenic differences appear to be due to variations in the tertiary structures rather than to variations in the primary structures of the a subunits.

Introduction

The (Na + + K+)-ATPase (ATP phosphohydrolase, Na+/K+-transporting, EC 3.6.1.37) is an intrinsic enzyme of the plasma membrane that regulates the active transport of Na + and K + across the cell membrane. This enzyme is also specifically inhibited by cardiac glycosides such as ouabain, and it is generally believed to be the pharmacological receptor for these drugs. It consists of catalytic (a) and glycoprotein (13) sub-

units, with both the ATP hydrolysis and ouabainbinding sites residing on the a subunit (see reviews by Glynn and Karlish [1], Wallick et al. [2], Robinson and Flashner [3] and Jorgenson [4]). It appears that the mechanism of action and the gross molecular structure of the ( N a + + K + ) ATPase is basically the same in all vertebrates [1,5]. Immunological studies have also suggested considerable immunological cross-reactivity between these enzymes [6-8]. Sweadner [9], however, has shown that two forms of this enzyme may

0167-4838/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)

80

occur in the brain tissue of various mammals. The rat astrocyte ( N a + + K+)-ATPase contains an c~ subunit that resembles that of non-neural tissues as cardiac muscle and kidney. The plasma membrane of myelinated axons contains an ( N a + + K ' )-ATPase with a higher molecular weight a + subunit which is more sensitive to cardiac glycoside inhibition than is c~. These ~ and a + forms of the brain enzyme show immunological cross-reactivity with each other and with the a form of dog, rat and lamb kidney [7]. The a and ~ + subunits appear to be structurally similar as determined by one-dimensional peptide maps [9], but they also appear to have some antigenic differences [10]. Kidney tissue appears to contain a single form of the a-subunit, but there is considerable species variation in the sensitivity of this enzyme to inhibition by cardiac glycosides. Human, dog, cat and lamb kidney enzymes are very sensitive, while mouse and rat kidney enzymes are relatively insensitive to inhibition [11]. For example, the rat kidney enzyme is approx, 2000-fold less sensitive to inhibition by ouabain than is the lamb kidney enzyme, yet in most respects these enzymes are similar. Their c~ and /3 subunits comigrate on gel electrophoresis, the ratios of their ( N a + + K+) ATPase to p-nitrophenyl phosphatase activity and their affinities for Na +, K +, ATP, vanadate and oligomycin are essentially the same [12,13]. The cause of their difference in sensitivity to inhibition by ouabain is not known, but an as yet undemonstrated structural variation in the a subunit appears to be the most likely explanation [9], although Akera et al. [14] have suggested that an alteration in enzyme boundary lipids may be implicated. In the present study, the extent of immunological cross-reactivity between the lamb and rat kidney (Na + + K +)-ATPases has been determined using various preparations of rabbit polyclonal antibodies. In addition, six monoclonal antibodies raised to the lamb c~ subunit have been used to demonstrate specific conformational and structural differences between the a subunits of these two enzymes. One monoclonal antibody (M7-PBE9) has detected a difference in the primary structure of the two enzymes at a functionally important region of the c~ subunit.

Experimental procedures Purification of (Na ~ + K +)-ATPase and subunits. The (Na + + K+)-ATPase was purified from the outer medulla of frozen lamb kidneys as previously described by Lane et al. [15]. ( N a ~ + K+) ATPase was isolated from whole rat kidney using essentially the same procedures as for the lamb enzyme, as described in Periyasamy et al. [12]. The and /3 subunits of the lamb enzyme were purified by gel filtration on Bio-Gel A-5m in 0.1% SDS as previously described with Sepharose CL-6B by Reeves et al. [16]. Eno'me assays. The (Na ~ + K+)-ATPase activity was determined by using a coupled enzyme spectrophotometric assay [17] in a medium containing 30 mM histidine, 2.5 mM N a 2 A T P (P-E Biochemicals), 5.0 mM MgCI 2, 95 mM NaC1, 10 mM KCI, 1.0 mM EGTA-Tris, 0.36 mM N A D H , 2 mM phosphoenolpyruvate and 10 /~1 pyruvate kinase/lactate dehydrogenase (Sigma), p H 7.2. The antisera, preimmune sera, sheep globulin fractions, and albumin solutions were extensively dialyzed against a 25 mM Tris buffer, pH 7.2, before being used. The effects of the antibody fraction on ( N a + + K + ) - A T P a s e activity were recorded after prior incubation of 0.75-1.0 /xg enzyme with varying amounts of antisera at 37°C for 20 rain. The ATPase reaction was initiated by the addition of ATP. The enzyme activity data are corrected for the occasional endogenous serum (Na + + K+)-ATPase activity, and for the (5-15%) increase in activity observed with high concentrations of added globulin or albumin fractions. Immunization of animals. Antibodies to the lamb and rat holoenzymes and to the lamb r~ and /3 subunits were raised in rabbits by injecting an emulsified suspension of complete Freunds adjuvant and the antigen (1 mg/ml). The suspension (1 ml) was initially injected subcutaneously (multiple sites) on the backs of the rabbits weekly for 4 weeks followed by 0.2 ml injected intramuscularly into each haunch for 2 weeks. Animals were bled bi-monthly, with booster injections being administered on the weeks between bleedings. The antisera complement was heat-inactivated for 30 min at 57°C, and then centrifuged, filtered through (0.45 and 0.22 p,m) Millipore ''~ filters and dialyzed against the 25 mM Tris buffer before use. The

81 binding specificity of the antibodies to the lamb kidney ( N a + + K + ) - A T P a s e proteins was confirmed by double diffusion immunoprecipitation and agar immunoelectrophoresis [18,19] and by electrophoretic transfer immunoblot binding.

Hybridoma cell lines and monoclonal antibodies. The anti-a subunit antibody secreting hybridoma cell lines were obtained by the fusion of mouse splenic lymphocytes from immunized mice (CB6F1/J), from Jackson Laboratory, with a Sp2/0-Ag14 myeloma cell line according to the procedures of Galfre et al. [20] and Kohler and Milstein [21]. The monoclonal antibodies were obtained by Protein A column chromatography of the immunoglobulin fraction of the ascites fluid from hybridoma cell inoculated mice. The procedures for the production, isolation and characterization of the antibodies from the isolated hybridoma cell lines have been described previously [22,231. Determination of antibody binding. Antibody binding was measured using an indirect solid-surface adsorption binding assay similar to that developed by Engvall [24] and which has been extensively characterized [18,19,22]. Microtiter plate wells (Cooke flexible 96-well plates) were treated for 30 min with 100/~l/well of test proteins at 0.05-0.5 mg/ml. The antigen-containing solution was removed and the wells were washed several times with buffer A (5 m g / m l bovine serum albumin), 10 mM Tris and 150 mM NaC1, pH 7.4), exposed to the antisera or antibody diluted in buffer A for 2 h. Rabbit immunoglobulins bound to the antigens were washed with buffer A and detected using 1251-labeled Protein A (from Staphylococcus aureus). Mouse immunoglobulins bound to the antigen were detected by adding a/3-galactosidase sheep anti-mouse IgG F(ab')2 conjugate (Bethesda Research Laboratories) for 90 min. When using mouse immunoglobulins the titer plates were rinsed with buffer B (5 m g / m l bovine serum albumin, 1.5 mM MgC12, 2 mM 2-mercaptoethanol, 0.05% Triton X-100 and 0.05% sodium azide). The bound second antibody was detected using o-nitrophenyl-/3-galactoside as substrate for the fl-galactosidase conjugate. After a 10 min incubation, the reaction was terminated with 0.5 M Na2CO 3 and each sample was transferred to a fresh microtiter plate well and the absorbance of

the reaction solutions was determined at 404 nm. The a and /3 subunits of lamb and rat kidney (Na ~ + K+)-ATPase were resolved on 7.5% SDSpolyacrylamide Laemmli gels [25], and 2-mm gel slices were placed into separate wells of the plastic Cooke plates. The gel proteins were then extracted into 100-~1 volumes of 0.5 M Na2CO 3 for 12 h. The gel slices were removed, and the plates were washed with buffer A. The wells were then exposed to a substantial excess of antisera or monoclonal antibody diluted in buffer A for 90-120 min, and the rabbit antibody-antigen complexes or the mouse antibody-antigen complexes were determined by using 125I-labeled protein A or the /3-galactosidase-anti-mouse IgG F(ab') 2 conjugate, respectively. Immunoblots (Western blots) were made by electrophoretic transfer of the slab gel (7.5% Laemmli gels) resolved proteins to nitrocellulose sheets as described by Towbin et al. [26]. Blots were then exposed to rabbit antisera or monoclonal antibody and then either horseradish peroxidase-conjugated goat anti-rabbit IgG or peroxidase goat anti-mouse IgG conjugate. 4'Chloro-l-naphthol was used as the color reagent to develop the blots. Results

Antigenic similarities as detected by holoenzyme-directed antibodies An indirect solid-surface adsorption binding assay was used to quantitate the immunological cross-reactivities of the rabbit polyclonal antibody populations raised to the lamb and rat kidney holoenzymes. The extent of the cross-reactivity and the relative affinities of these antibodies for both the immunizing antigen and the other enzyme preparation are shown in Table I. Antibodies raised to the lamb holoenzyme exhibited approx. 30-40% maximal cross-reactivity with the rat holoenzyme with a substantially reduced affinity (Fig. la and Table I). With early bleed samples (7-10 weeks immunization), rat holoenzyme-directed antibodies cross-reacted approx. 100% with lamb enzyme. As the immunization protocol was continued, this cross-reactivity declined and stabilized at approx. 50% cross-reactivity with the holoenzyme, similar to what was observed with the "late bleed" anti-lamb enzyme sera.

82 TABLE l DETERMINATION LAMB A N D RAT BODIES

Titer values are given as the dilution values for half-maximal binding. Maximal bindings are given as percentages relative to that for the immunizing antigen. The binding is quantitated using 12SI-protein A. Antiserum numbers are given as designation numbers for particular rabbits, while the subscript number indicates bleed number. Antiserum

80 60 40 o

o J~ k~

Lamb (Na+ + K + ) ATPase

Rat (Na+ + K + ) ATPase

c

Titer

Titer

Q..

Lamb directed SGP 94 1 : 300 SGP 810 1 : 500 SGP 1011 1 : 1000

Percentage binding

Percentage binding

a

100

OF B I N D I N G PROPERTIES OF HOLOENZYME-DIRECTED ANTI-

20 0

i

.04

.01

4

0

b

8O 6O

100 100 100

1 : 50 1 : 120 1 : 50

30 37 37

4O

20 A 0 .// .2 .1

1 : 240 1:110 1:120 1 : 300 1 : 75

.02

100(

I

Rat directed RGP 42 RGP52 RGP48 RGP 5~ RG P 15

.03

105 97 30 66 34

1 : 1000 1:800 1:800 1 : 2500 1 : 1200

100 100 100 100 100

'l'he effect of these antibody populations on inhibition of ( N a + + K+)-ATPase activity of the two enzymes was also determined. Concentrations of the lamb enzyme-directed antisera which inhibited the lamb enzyme by 50% had no inhibitory effect on the rat enzyme. The early bleed samples of the anti-rat enzyme sera, however, inhibited both the rat and lamb enzyme ( N a + + K + ) ATPase activity similarly (53% vs. 42%). By the seventh bleed (approx. 5 months immunization) the anti-rat enzyme sera had little inhibitory effect on the rat enzyme and none on the lamb enzyme. These results indicate that the inhibition of enzyme activity is a poor indicator of immunological cross-reactivity in comparison to the adsorption binding assay. To determine the immunological cross-reactivities of the holoenzyme-directed antibodies for the component c~ and /3 subunits of the lamb and rat enzymes, the a and ,8 subunits of the two enzymes were separated by SDS-gel electrophoresis, allowed to bind to the plastic microtiter plates, and

_

.08

.06

.04

.02

Dilution Ratio

Fig. 1. Determination of antibody binding to lamb and rat kidney (Na + + K + )-ATPase as a function of antiserum concentration. (a) shows the binding of lamb holoenzyme-directed antibodies to the lamb enzyme; C), SGP81o; zx, SGP1011: and to the rat enzyme; • and A, respectively. (b) shows the binding of anti-rat enzyme antibodies to the rat enzyme (©, RGP42) and binding to the lamb enzyme • , RGP42 and *, RGP4s). The titer curve of RGP48 binding to the rat enzyme is similar to RGP42 and not shown.

the extent of antibody cross-reactivity for the individual subunits was determined as described in Experimental procedures. By comparing the ratios of the cpm of antibody-bound t25I-Protein A to the relative amount of Coomassie blue staining in each of the gel protein peaks, it was possible to estimate the extent of cross-reactivity. Fig. 2 illustrates the immunological reactivity of the lamb holoenzyme-directed antibodies to the lamb and rat subunits. Identical experiments were conducted with the rat holoenzyme-directed antibodies. As shown in Table II, two different preparations of lamb holoenzyme-directed antibodies exhibited about 60% cross-reactivity for the c~ and/3 subunits of the rat kidney enzyme, while the rat enzyme-directed antibodies exhibited about 40% cross-reactivity for the a and /~ subunits of the lamb enzyme.

83 12[

]r

o

I ~2

i

J

TABLE I1

1o T

< c

8~

2

6 I

o_

7 13

0s ! 06

l

F

o Lo

t g 0

0

20

40

60

80

20

40

60

80

0

Gel Length (mm)

Fig. 2. Determination of lamb holoenzyme-directed antibody binding to the c~ and /~ subunits of the lamb and rat (Na + + K + )-ATPase. (a) shows the immunoglobulin (SGP8]0) binding to the SDS-PAGE resolved lamb enzyme subunits. (b) shows immunoglobulin cross-reactivity with the gel-resolved rat enzyme subunits. The open circles (©) show quantitation of antibody binding to the subunit using ]251-protein A. The dotted line represents the 560 nm absorbance profile of approx. 5/~g of the gel-resolved Coomassie blue-stained protein.

Antigenic similarities detected by subunit-specific antibodies Since it is logical to assume that SDS-denatured (Na + + K + ) - A T P a s e s u b u n i t s would present more antigenic sites than the holoenzyme, rabbit polyclonal antibodies were raised against ( N a + + K + ) - A T P a s e t~ a n d /~ s u b u n i t s which had been purified by SDS-gel chromatography. The extent of a n t i b o d y b i n d i n g of two sets of lamb enzyme subunit-directed antisera to the electrophoretically resolved s u b u n i t s of the lamb a n d rat enzymes was d e t e r m i n e d as described previously. The extent of cross-reactivity was f o u n d to be approx. 60% a n d 45% for the rat a a n d /~ subunits, respectively. The above results suggest that there is a 40-60% similarity in the antigenic regions of the rat and lamb proteins, whether they are in a ' n a t i v e ' or d e n a t u r e d form. However, since these results were o b t a i n e d using heterogeneous p o p u l a t i o n s of rabbit serum antibodies, it is not possible to estimate either the n u m b e r of distinct or overlapping det e r m i n a n t sites to which the antibodies have been raised or the n u m b e r of d e t e r m i n a n t sites represented by a particular percentage decrease in total a n t i b o d y binding.

DETERMINATION OF HOLOENZYME-DIRECTED ANTIBODY BINDING TO THE SDS-POLYACRYLAM1DE GEL SEPARATED (Na + + K + )-ATPase SUBUNITS The extent of cross-reactivity of holoenzyme-directedantibodies for the subunits of the (Na++K+)-ATPase was determined by quantitating the amount of antibody binding to the titer plate-adsorbed catalytic and glycoprotein subunits relative to amount of protein as indicated by Coomassie blue staining of the 7.5% polyacrylamide gel resolved proteins. Values are given as percentage binding relative to that of the immunizing protein. Similar studies using a-directed antisera designated ~-114 and c~-4a6gave 55% and 75% cross-reactivity with rat a. ,8-directed antisera, ,8-17 and /~-36, gave 42% and 44% cross-reactivitywith rat /~. Antiserum

Antigen Lamb kidney subunits

Rat kidney subunits

Catalytic

Catalytic Glycoprotein

Glycoprotein

Lamb enzyme specific SGP 810 100% SGP 1011 100%

100% 100%

60% 64%

57% 75%

Rat enzyme specific RGP 42 41% RGP 52 36%

38% 37%

100% 100%

100% 100%

Studies with monoclonal antibodies To examine specific antigenic regions of the ( N a + + K + ) - A T P a s e , a collection of m o n o c l o n a l antibodies which are directed against the a subu n i t of the l a m b kidney enzyme were isolated. The titer values, or a p p a r e n t affinity values, for the b i n d i n g of six different m o n o c l o n a l antibodies to the lamb holoenzyme were d e t e r m i n e d using an e n z y m e - l i n k e d i m m u n o s o r b e n t assay as shown in T a b l e III. In addition, five of the six antibodies were f o u n d to exhibit a higher a p p a r e n t affinity towards the d e n a t u r e d lamb c~ than to the lamb holoenzyme. The sixth antibody, M10-P5-C11, b i n d s almost exclusively to the ' n a t i v e ' lamb holoenzyme. All of these antibodies make a clear immunological distinction between the rat and lamb holoenzymes. They show no cross-reactivity with the rat kidney holoenzyme. The ability of these antibodies to b i n d to the electrophoretically resolved ~ s u b u n i t s of the l a m b

84 TABLE 111 ANTI-LAMB K I D N E Y (Na + + K ÷ )-ATPase ANTIBODY P R O D U C I N G HYBRIDOMA CELLS Antibody binding to plate-adsorbed antigen (0.2-0.5 m g / m l ) was detected using the EL1SA assay. Titer values are defined as concentration of purified antibody given 50% maximal binding. The extent of cross-reactivity was determined by comparing the anaount of detected antibody binding to the titer plate-adsorbed catalytic subunit relative to the absorbance of Coomassie blue staining of the protein resolved on 7.5% polyacrylamide gel. n.d., not determined. Cell line:

Binding characteristics

Binding titers or specificity

designation of clones

lgG isotype

lamb kidney (Na + + K + )-ATPase: subunits

rat kidney (Na + + K + )-ATPase

catalytic (nM)

glycoprotein

holoenzyme

% Binding SDS-gelresolved catalytic subunit

3 3 1.5 1,500 0.7 1.4

0 0 0 0 0 0

0 0 0 0 0 0

8% 47% 45% 0% 40% n.d.

M7-PB-E9 (E5) M8-PI-A3 MI0-P6-B7 MI0-P5-Cll M 12-P4-E8 MI 2-P7-F11

holoenzyme (nM)

IgG 1 IgG 1 IgG 1 IgG 1 IgG I IgG ~

5.2 7.5 4.6 7.0 7.0 1.5

05

1.0

b 0.8

~4

0.6

3.3

0.4

II

0.2

II I'

:).2

l D.I

'I I, % .,

O

0

o

1.0

0.5

8c e~

d

C 0.8

0.4

0.6

0.3 ,i

0.4

0.2

t;

0.2

0

t 0

2O

40

60

80

0

. ii 20

Gel Length (mm)

Ji 40

60

80

o.I

O e~ v

Fig. 3. Determination of monoclonal antibody binding to the lamb and rat a subunits. Panel a shows antibody M10-P6-B7 binding to the gel-resolved a subunit of the lamb enzyme. Panel b shows antibody M10-P6-B7 cross-reactivity with the rat enzyme ~ subunit. Panel c shows antibody M7-PB-E9 binding to the gel-resolved c~ subunit of the lamb enzyme. Panel d shows antibody binding to the rat a subunit. The open circles ( O ) show quantitation of antibody binding to cc using the /~-galactosidase sheep anti-mouse IgG F(ab') 2 conjugate. The dotted line represents the absorbance profile of the stained proteins at 560 nm.

85

R !.

m l

M-B7 R I.

M+E9 Fig. 4. Immunoblot analysis of the specificity of lamb kidney (Na + + K + )-ATPase-directed antibodies. Samples of purified ( N a + + K + )-ATPase were loaded onto 14.2-cm-wide slab gels of 7.5% polyacrylamide, electrophoresed and then electroblotted onto nitrocellulose sheets. Individual nitrocellulose strips were then stained by amido black to detect the transferred protein or treated individually with rabbit anti-(Na + + K + )-ATPase sera or the monoclonal antibodies. Bound antibody was detected with the peroxidase-conjugated second antibody. The first two strips (upper left) show amido black staining of the blotted rat (strip R, 12 /zg) and lamb (strip L, 12 /zg) kidney (Na + + K + )-ATPase. The subsequent pairs of strips designated R and L demonstrate the binding of rabbit antiserum (AS), and monoclonal antibodies, M8-P1-A3 (M-A3), M10-P6-B7 (M-B7), M12-P4-E8 (M-E8), M12-P7-Fll (M-Fll), M10-P5-Cll ( M - C l l ) and M7-PB-E9 (M-E9), to rat and lamb enzymes, respectively. The rabbit antiserum (SGP-108) was diluted 1 : 200 and the monoclonal antibodies were used at 100 nM.

and rat enzymes was then determined (Fig. 3 and Table III). As expected, antibody M10-P5-Cll showed little, if any, detectable binding to either denatured lamb or rat c~. Antibodies M8-P1-A3, M10-P6-B7 (Fig. 3a) and M12-P4-E8 all bound to the lamb a and exhibited 47%, 45%, and 40% cross-reactivity with rat ~, despite their lack of binding to the rat holoenzyme. Antibody M7-PBE9 (Fig. 3b) was found to bind to denatured lamb but it exhibited only an 8% cross-reactivity with the rat c~.

Immunoblot cross-reactivity studies Because antibody binding had been found to be sensitive to the conformational form of the enzyme proteins, electrophoretic immunoblots (i.e., Western blots) of the purified lamb and rat enzymes were also used to assess antibody specificity or cross-reactivity. As shown in Fig. 4, the polyclonal lamb kidney holoenzyme-directed antibodies cross-reacted with the a and the/~ subunits of the rat (Na + + K+) ATPase. In addition, monoclonal antibodies M8P1-A3, M10-P6-B7, M12-P4-E8 and M12-P7-F11 also reacted with both the lamb and rat ( N a + + K+)-ATPase a subunits (M10-P5-Cll binds only the native lamb enzyme). Finally antibody M7PB-E9 again showed only trace cross-reactivity with the rat a. The observed antigenic difference between the rat and lamb (Na + + K+)-ATPases at the M7-PB-E9 epitope therefore appears to reside in the primary structure of the c~ subunit. Discussion Previous studies have demonstrated an immunological cross-reactivity between the ( N a + + K+)-ATPase of different tissues in the same species [6] and between the ( N a + + K+)-ATPases of various vertebrate species [6-8]. However, other than comparisons of the extent to which a particular antiserum can inhibit the ATPase activity of various preparations of the enzyme, no quantitation of the immunological cross-reactivity between ( N a + + K+)-ATPases from different species has been made. Since the antibodies which inhibit ( N a + + K+)-ATPase activity probably represent only a small fraction of the total population of antibodies produced against this complex antigen,

86 such comparisons would be expected to reflect immunological differences of only limited regions of the enzymes. These studies demonstrate that antibody inhibition of enzyme activity is, in fact, not a reliable indicator of immunological cross-reactivity. In the present study, using 'late bleed' antisera samples, we have shown that rabbit polyclonal antibodies raised against the lamb kidney (Na + + K+)-ATPase holoenzyme cross-react about 40% with rat kidney holoenzyme, but with a substantially lowered affinity. By using both 'early' and 'late' bleed samples of antisera raised against the rat kidney holoenzyme, we found a changing pattern of binding specificity, with a similar 40-50% cross-reactivity of rat enzyme-directed antibodies to the lamb enzyme being observed after several months of immunization. In addition, these rat and lamb holoenzyme-directed antibodies show substantial binding to the denatured ( N a + + K+)ATPase proteins. Polyclonal antibodies monospecific for the SDS-denatured a and fl subunits of the lamb enzyme also show about a 50% cross-reactivity towards the same subunit of the rat enzyme. Thus~ there appear to be both significant antigenic similarities and differences between the denatured rat and lamb subunits. Since these studies were carried out with denatured, lipid-depleted proteins, some of the observed differences in antigenicity must reflect differences in the primary structures of these proteins. It should, of course, be pointed out that while the results obtained for each of these sets of antisera appear to be consistent with each other, the variability between animals as to their immunological response to a complex antigen may be substantial. Varying, and sometimes contradictory, effects of antibodies on (Na + + K +)-ATPase functions have been observed in previous studies [27,28-31]. It is difficult to know whether these discrepancies can be explained by differences in the experimental procedures used or by differing antibody populations. In addition, there are limitations to any antibody binding assay. The use of the solid-phase binding assay as a quantitative method has its limitations and the assay conditions must be very carefully standardized. It is possible that the adsorption of proteins onto the

plastic microtiter plates could produce a heterogeneous array of adsorbed antigens and that the differing quantities of bound antibody could reflect this variation in the proteins. However, our previous competition binding studies [18,19] carried out using the rabbit polyclonal antibodies have shown that the same antibody population binds to both the 'in solution' active enzyme and its subunits and to the plate-adsorbed proteins, Competition binding using the monoclonal antibodies have given similar results (unpublished results). Clearly, the results obtained using the six monoclonal antibodies demonstrate the utility of using specific antibodies as probes of both conformational and structural aspects of proteins. These six monoclonal antibodies bind to different, apparently non-overlapping determinant sites. Five of the monoclonal antibodies bind to the isolated lamb c~ subunit with a higher "affinity" than they do to the holoenzyme, while one antibody shows a greatly reduced binding to isolated c~. From these results, it is apparent that each of these antibodies is sensitive to some conformational differences between the isolated lamb c~ and c~ as complexed with fi and lipids in the lamb holoenzyme. Since none of these antibodies bind to the rat holoenzyme, it is reasonable to assume that substantial differences in either the conformation a n d / o r the primary structure exist at these six determinant sites of the rat and lamb c~ subunits. Because antibodies M8-P1-A3, M10-P6-B7, M12-P4-E8 and M I 2 - P 7 - F l l show substantial cross-reactivity with the denatured rat c~ the primary structures of these four determinant sites must be quite similar, while the tertiary structures are dissimilar. The monoclonal antibodies have demonstrated that there are substantial tertiary structure differences between the lamb and rat c~. The observation that antibody M7-PB-E9 binds to the denatured lamb o~ but not to the denatured rat ~ demonstrates that the primary sequences of the lamb and rat ~ differ at this antigenic site. Previous studies from this laboratory [22] have shown that antibody M7-PB-E9 binding to the lamb (Na + + K+)-ATPase acts competitively with ATP to inhibit ( N a + + K+)-ATPase activity. In the presence of Mg 2+, the antibody also acts in an ATP-like manner to increase the affinity of the

87

enzyme for ouabain. This antigenic site appears to be involved in the regulation of both catalytic activity and ouabain binding. These results suggest that antibody M7-PB-E9 has identified a functionally important sequence difference between these two enzymes. These studies also demonstrate that the inability of a monoclonal antibody to cross-react with an homologous 'native' enzyme is not substantial proof that this determinant site is absent in a putative isoenzyme or homologous enzyme. Fambrough and Bayne [32] reported the isolation of a monoclonal antibody raised to the ( N a ÷ + K+)-ATPase of embryonic chicken muscle cells which does not bind to all chicken ( N a + + K÷) ATPases. On the basis of this variable cross-reactivity to holoenzymes from different tissues, they concluded that there are multiple tissue-specific forms of the chicken ( N a ÷ + K+)-ATPase. The present study demonstrates that the variable affinity of an antibody for closely related antigens is a function of both the conformation and the primary structure of the proteins; and, further, that an antibody's sensitivity to these aspects can be highly variable.

Acknowledgments The authors thank Chuck Loftice, Kelly Crawford, Kathryn Kalvin-Hartmann and Purabi Dey for their excellent technical assistance. This work was supported by a grant from the American Heart Association (W.J.B.) and grants from the National Institutes of Health (R01-HL32214 (W.J.B.), R01-HL25545 (L.K.L.) and P01HL22619). W.J.B. is an Established Investigator of the American Heart Association. References 1 Glynn, I.M. and Karlish, S.J.D. (1975) Annu. Rev. Physiol. 37, 13-55 2 Wallick, E.T., Lane, L.K. and Schwartz, A. (1979) Annu. Rev. Physiol. 41,397-419 3 Robinson, J.D. and Flashner, M.S. (1979) Biochim. Biophys. Acta 549, 145-176

4 Jorgensen, P.L. (1982) Biochim. Biophys. Acta 694, 27-68 5 Carilli, C.J., Farley, R.A., Perlman, D.M. and Cantley, L.C, (1980) Curr. To. Bioenerg. 11,201-237 6 McCanns, J.L., Lindenmayer, G.E., Pitts, B.J.R., Ray, M.V., Raynor, B.D., Butler, V.P., Jr. and Schwartz, A. (1975) J, Biol. Chem. 250, 7257-7265 7 Schellenberg, G.D., Pech, I.V. and Stahl, W.L. (1981) Biochim. Biophys. Acta 649, 691-700 8 McDonough, A.A., Hiatt, A. and Edelman, I.S. (1982) J. Membrane Biol. 69, 13-22 9 Sweadner, K.J. (1979) J. Biol. Chem. 254, 6060-6067 10 Sweadner, K.J. and Gildeson, R.C. (1985) Biol. Chem. 250, 9016-9022 11 Wallick, E.T., Pitts, B.J.R., Lane, L.K. and Schwartz, A. (1980) Arch. Bicohem. Biophys. 202, 442-449 12 Periyasamy, S.M., Lane, L.K. and Askari, A. (1979) Biochem. Biophys. Res. Commun. 86, 742-747 13 Periyasamy, S.M., Huang, W.H. and Askari, A. (1983) Comp. Biochem. Physiol. 76B (3), 449-454 14 Akera, T., Yamamoto, S., Chibb, J., McNish, R. and Brody, T.M. (1979) Naunyn-Schmiedeberg's Arch. Pharmacol. 208, 31-88 15 Lane, L.K., Potter, J.D. and Collins, J.H. (1979) Prep. Biochem. 9, 157-170 16 Reeves, A.S., Collins, J.H. and Schwartz, A. (1980) Bicohem. Biophys. Res. Commun. 95, 1591-1598 17 Schwartz, A., Allen, J.C. and Harigaya, S. (1969) J. Pharmacol. Exper. Therap. 168, 31-40 18 Ball, W.J., Jr. and Schwartz, A. (1982) Arch. Bicohem. Biophys. 217, 110-119 19 Ball, W.J., Jr., Collins, J.H., Lane, L.K. and Schwartz, A. (1983) Arch. Biochem. Biophys. 221,371-380 20 Galfre, G., Howe, S., Milstein, C., Butcher, G. and Howard, J. (1977) Nature 266, 550-552 21 Kohler, G. and Milstein, C. (1975) Nature 256, 495-497 22 Ball, W.J., Jr., Schwartz, A. and Lessard, J.L. (1982) Biochim. Biophys. Acta 719, 413-423 23 Ball, W.J.~ Jr. (1984) Biochemistry, 2275-2281 24 Engvall, E. (1978) Scand. J. Immunol. 8 (Suppl. 7), 25-31 25 Laemmli, U.K. (1970) Nature 227, 680-685 26 Towbin, A., Shaekelin, T. and Gordon (1979) J. Proc. Natl. Acad. Sci. USA 76, 4340 27 Askari, A. (1974) Ann. N.Y. Acad. Sci. 242, 372-388 28 Rhee, H.M. and Hokin, L.E. (1975) Biochem. Biophys. Res. Commun. 63, 1139-1145 29 Kyte, J. (1974) J. Biol. Chem. 249, 3652-3660 30 McCanns, J.L., Lane, L.K., Lindenmayer, G.E., Butler, V.P., Jr. and Schwartz, A. (1974) Proc. Natl. Acad. Sci. 71, 2449-2452 31 Jean, D.H. and Albers, R.W. (1977) J. Biol. Chem. 252, 2450-2451 32 Fambrough, D. and Bayne, E. (1983) J. Biol. Chem. 258, 3926-3935