determination of monoclonal antibody-induced alterations in na+/k+-atpase conformations using...

42 Biochimica et Biophysica Acta, 995 (1989) 42-53 Elsevier

BBA 33320

Determination of monoclonal antibody-induced alterations in N a+/K+-ATPase conformations using fluorescein-labeled enzyme

M a r k L. F r i e d m a n * and W i l l i a m J. Ball, Jr.

Department of Pharmacology and Cell Biophysics, University of Cincinnati College of Medicine, Cincinnati, OH (U.S.A.)

(Received 11 Jaly 1988)

Key words: ATPase, Na+/K+-; Enzyme conformation; Antibody binding site; Ion binding; Conformational transition

The flnomseein 5'-isothlecyanate (FITC).laheled lamb kidney Na+/K+.ATPase has been used to investigate enzyme ftmetioa and Ilgambindneed confornmtlenal changes. In these studies, we have determined the effects of two monodonal antibodies, wbieh inhibit Na+/K+-ATPase activity, on the conformational changes undergone by the FITC-labeled enzyme. Moaltming fluorescence intensity changes of FITC.laheled enzyme shows that antibody MI0-PS-CII, which inhibits E t - P interraediate formation (Ball, W.J. (1986) Biochemistry 25, 7155-7162), has little effect on the E t ~ E 2 transitions induced by Na +, K +, Mg2+P, or Mg 2+ • ouabaln. The MI0-P5-CII epitope, which appears to reside near the ATP-bind|ngo site, does not signifieantiy participate in these ligand interactions. In contrast, we find that antibody 9-AS (Seheult, D.B., Hubert, J.J. and Leffert, H .L (1984) J. Biol. Chem. 259, 14941-14951) inhibits both the Na+/K+-ATPase and p-nitmphenylphesphatase activity. Its binding produces a 'Na+like ' enhancement in FITC flnoroseenee, ~ the ability of K + to indnee the Et o E2 transition and converts E z • K + to an E t conformation. Mg 2+ binding to the enzyme alters both the conformation of this epitope region and its coupling of ligand interactions. In the presence of Mg 2+, 9-A$ binding stabilizes an E t • Mg 2+ conformation such that K+-, P,- and ouabain-induced E t ~ E a or E t --, E2.P , transitions are inhibited. Oualmin and P, added together overcome this stabilization. These studies indicate that the 9-AS epitope iwticiimtes in the E t o E 2 conformational transitions, linl~ Na+-K + interactions and mmlmin extraeellular binding site effects to both the phosphorylation site and the FITC-binding region. Antibody- binding studies and direct demonstration of 9-A5 inhibition of enzyme phosphorylafion by 132 PiP, confirm the results ebtaimal from the flmcesceaee studies. Antibody 9-A5 has also proven useful in demonstrating the independence of Mga+ATP and Mga+P, regulation of ouabaln binding, in addition, ISHlouabain and antibody.binding studies demoagrate that Frrc.laheling alL*rs the enzyme's resimnses to Mg 2+ as well as ATP regulation.

Introduction

Na+/K+-ATPase (ATP phosphohydrolase, EC 3.6.1.3) is a membrane protein composed of two sub- units, the a subunit (112 kDa) and the fl subunit (35 kDa). The primary function of the enzyme is the main-

* Present address: Miles Inc., P.O. Box 70, Elkhart, IN, U.S.A. Abbreviations: FITC, fluorescein 5'-isothiocyanate; SR, sarcoplasmic reticuham; pNPP, p-nitrophenyl phosphate; TCA, triehloroacetic acid; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophdre- sis; Pi, inorsanie phosphate.

Correspondence: Department of Pharmacology and Cell Biophysics, University of Cincinnati College of Medicine, Cincinnati, Ohio 452W- 0575, U.S.A.

tenance of intracellular K + levels and the excretion of Na + from cells. The energy for the transport of K + and Na + against their respective concentration gradients is derived from the hydrolysis of ATP. The a subunit contains the ATP- and cation-binding domains and appears to be the pharmacological receptor for cardiac glycosides, but none of these binding sites on the a protein have been clearly identified. No functional role has been ascribed to ft. The transport of Na + and K + ions involves the cycling of the enzyme between E 1 (Na + form) and E 2 (K + form) conformations with various phosphoenzyme intermediates, hence this enzyme belongs to a designated class of EIE 2 ATPase pumps (see reviews in Refs. 1-5). Although the order and number of conformational transitions is still under

0167-4838/'89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)

43

debate [6], the Albers-Post scheme represented below gives the widely accepted basic scheme [7,8].

Mg 2+ ATP ADP

E,Na+4 ~.~ / "

K +~--..~'~ i l ~ Na + Na + ......-~,1,

E2 K÷ ~ .--- E2-P

Pi K+

The existence of the E l, E 2 and the phosphorylated intermediate forms has been demonstrated by a number of researchers using a variety of techniques and chem- ical probes [9-12].

Fluorescein probes have been used extensively to study the E 1 ~ E 2 transitions of the enzyme [13-17]. One fluorescein derivative, fluorescein 5'-isothiocyanate (FITC), covalently labels the lamb kidney Na+/K +- ATPase a-subunit at Lys-501 [18,19]. This lysine is positioned intraeellularly as part of the peptide region which lies between the proposed H4 and H5 transmembrane stretches, and is approximately 132 amino acids towards the -COOH-terminal end from the phosphorylation site, Asp-369 [20,21]. Studies with sarcoplasmic reticulum (SR) Ca2+-ATPase [22] and gastric H + / K +- ATPase [23] indicate that this region of the polypeptide is highly conserved among the E~E 2 ATPase pumps. The presence of ATP prevents both labeling and in- activation of the enzyme by FITC. It is therefore gener- ally assumed that this region is part of the ATP-binding site, although some questions have been raised about this assumption [24,25]. The addition of various ligands to the FITC-labeled enzyme results in fluorescence intensity changes reflecting alterations in the probe's en- vironment which are coincident with accepted E1 ~ E 2 enzyme transitions [14]. This probe therefore is used to monitor enzyme conformational changes which occur during catalytic turnover.

Recently, a number of monoclonal antibodies have been raised to Na+/K+-ATPase for investigating b, Jth structural and functional aspects of the enzyme [26-30]. A limited ntlmber of these antibodies alter enzyme function. Monoclonal antibody M10-P5-Cll [27], raised to the lamb kidney enzyme, inhibits up to 90% of the Na+/K+-ATPase activity by inhibiting E ~ - P intermediate formation, but not dephosphorylation nor the enzyme's K+-dependent p-nitrophenylphosphatase activity. AntiLody 9-A5, raised to the rat kidney enzyme [29,31], inhibits ATPase activity apparently by reducing E-P formation without affecting the K m for ATP. Both antibodies bind to the a subunit on the intracellular side of the enzyme. 9-A5 binds to the 35 kDa chymotryptic fragment which contains the phosphorylation site [32,33] and M10-P5-Cll binds at an intracellular site that allows ATP binding but inhibits transfer of the ~,-phosphoryl group to the phosphorylation site [27].

The effects of these antibodies on some of the enzyme's partial reactions have been investigated, but their mechanisms of inhibition have not yet been determined. In the present study we have utilized the fluorescence changes occurring with FITC-labeled Na+/K+-ATPase to determine how these two antibodies affect ligand binding and ligand interactions, and how their epitopes participate in linking the regulatory interactions between ligands. Antibody-binding studies have demonstrated that antibody 9-A5 but not M10-P5- Cl l binding is sensitive to changes in enzyme conformation, and that its epitope is under regulatory control by Mg 2+. In addition, comparisons of 'native' and labeled enzyme relative to antibody binding, enzyme phosphorylation, and [3H]ouabain binding confirm that FITC fluorescence intensity changes do reflect protein conformation changes, and also demonstrate FITC- caused alterations in the enzyme.

Materials and Methods

Na + / K +.A TPase preparation. The membrane- bound enzyme was purified from the outer medulla of frozen lamb kidneys as described by Lane et al. [34]. The Na+/K+-ATPase activity of the enzyme preparation varied from 850 to 1000 #mol /mg per h. The enzyme was generously supplied by Dr. L.K. Lane.

Hybridoma cell lines and purification of antibodies. The cloned hybridoma cell line secreting antibody M10- P5-Cll was generated as described by Ball [27]. The 9-A5 antibody secreting hybridoma cell line was generously given to us by H.L. Leffert. The immunoglobulins were isolated from the ascites fluid of pristane-primed mice inoculated with the cloned hybridoma cells by methods already described [27].

Determination of antibody binding. M10-P5-C11 binding to the lamb kidney Na+/K+-ATPase was determined using an ELISA-type solid-surface adsorption assay [35]. Binding of 9-A5 could not be detected by this method, so a solution phase-binding assay was developed. In this procedure, enzyme (48/~g) and 9-A5 (2.4 #g) were combined at a ratio designed to ensure significant antibody binding in 1 ml buffer A (50 mM Tris-HCl/1 mM EGTA (pH 7.4)) at 4°C for 1 h. The Na+/K+-ATPase and any antibody-enzyme complex formed were removed from solution by centrifugation at 150000×g for 1 h. The unbound 9-A5 (in the supernatant) was adsorbed to microtiter plates (100 #l/well, Cooke flexible 96-well microtiter plates) for 2 h. Plate-bound 9-A5 was detected using a fl-galactosid- ase-sheep anti-mouse IgG F(ab')2 conjugate (Bethesda Research Laboratories). Solution phase binding of 9-A5 to the enzyme under various ligand combinations of Na + (100 mM), K +, Mg 2+, Pi, Tris-ATP (all at 5 mM), and ouabain (1-10 #M) was also examined. Controls consisting of samples of 9-A5 prepared similarily in the

44

absence of enzyme were used to determine control levels of antibody binding to the plate under each ligand condition.

Fluorescein 5'-isothiocyanate labeling of the Na +/ K +-,4 TPase. The labeling of the Na+/K+-ATPase with FITC (Molecular Probes) and determination of the extent of labeling was carried out as described by Kirley et al. [19]. Determination of the Na+/K+-ATPase and K+.dependent p-nitrophenylphosphatase activity of the labeled enzyme and SDS-PAGE of the labeled enzyme were performed as described below.

Enzyme assays. The Na+/K+-ATPase activity was determined by using a coupled enzyme spectrophoto- metric assay [36] in a medium containing 30 mM histidine, 5 mM Tris-ATP, 10 mM MgCl2, 100 mM NaCI, 10 mM KCI, 1 mM EGTA, 0.36 mM NADH, 2 mM phosphoenolpyruvate, and 15 FI of pyruvate kinase/lactate dehydrogenase (Sigma) (pH 7.2). The K+.dependent phosphatase activity was measured in a medium containing 30 mM histidine, 5 mM MgC! 2, 5 mM p-nitrophenylphosphate, 5 mM KCI and 1 mM EGTA (pH 7.4). To achieve maximal antibody effects, the enzyme was incubated with SDS at a 10.5 Fg SDS/7.5 F$ Na+/K ÷-ATpase ratio for 15 min at 4°C in approx. 40 FI of a 1 mM EGTA/Tris solution (pH 7.4). Various concentrations of the antibody (9-A5, MI0-PS-CII, or nonspecific IgG) were added to bring the final volume to 0.1 ml and the mixture was incubated for 45 min at 4°C. The enzyme/antibody mixture (0.75 Fg enzyme) was then used in the assay reaction mixture for activity determination. This concentration of SDS (0.010-0.017~) has previously been shown to increase the ATPase activity of this vesicular preparation about 20~ and antibody access to binding sites 20-50~ [27]. The binding affinity of MI0-PS-CII and 9-A5 was not altered. The SDS was omitted from all other experiments though, because it reduced enzyme retention on the millipore filters used in the ouabain-binding experiments [27] and this concentration was at the critical point where effects on vesicle light-scattering and FITC-fluorescence were observed (unpublished data).

[sH]Ouabain binding. [3H]Ouabain binding was carried out at 37°C essentially as described by Wallick and Schwartz [37]. The different binding conditions were (a) no iigands, (b) Mg 2+, (c) Mg2+ATP, (d) Na + Mg 2 +-ATP, (e) K + Na + Mg 2 + ATP, and (f) Mg 2 + Pi.

The ligand concentrations were 5 mM for MgCI2, Tris- ATP, KCI, Pi and 100 mM for NaCI. The enzyme, either native or labeled, (15 Fg) and the antibodies 9-A5, M10-PS-Cll, or sheep IgG Call at 60 Fg) were preincubated in 0.18 ml (buffer A) for 15 min at 4°C followed by the addition of iigands. Glycoside binding, under the various ligand conditions, was initiated by adding [3H]ouabain (1000 mCi/mmol) to obtain a reaction volume of 1 ml and a 0.1/tM concentration. The

binding experiments were carried out for various times depending upon the ligands present. All values were corrected for nonspecific binding as determined by the inclusion of 0.1 mM unlabeled ouabain prior to [ 3 H]ouabain adOition.

Enzyme [ S"P]P i phosphorylaffon and dephosphorylation. Phosphorylation of the enzyme by Pi was accomplished in a 0.30 ml buffer A solution. In this procedure, 20 Fg of enzyme (native or FITC-labeled) was incubated for 1 h at 4°C with approx. 80 Fg antibody (9-A5 or sheep IgG), followed by 1 h with Mg 2+ (5 mM) or Mg 2+ plus ouabain (1.0 FM). Following a 30 min incubation at 37°C, [32PIP i (9000 Ci/mmol) was added to achieve 75 F M and the phosphorylation reaction was stopped after 5 min with the addition of an equal volume of ice cold solution of 10~ trichloroacetic acid (TCA) and 400 mM Pi. The reaction mixtures were decanted onto 0.22 Fm MiUipore filters and washed with a 5% TCA/200 mM P~ solution. Nonspecific binding was determined using heat-denatured enzyme.

Fluorescence studies. The fluorescence of the FITC- labeled Na+/K+-ATPase was recorded on a Perkin- Elmer 650-10S spectrofluorometer with the sample tem- perature maintained at 25°C. Enzyme and antibody concentrations were 5 Fg/ml and 20-30 Fg/ml, respectively, in 2 ml, in either buffer A or 50 mM histidine, 1 mM EGTA, (pH 7.2). All solutions were allowed to come to thermal equilibrium for at least 5 min prior to initial measurements. A stirrer (suspended in the solution) was utilized to ensure rapid mixing upon addition of ligands or antibody and to prevent any settling of the enzyme preparation during the experiments. The fluorescence measurements were made with excitation at 495 nm and the intensity monitored at 520 nm. The emission spectrum was periodically monitored over the 500-600 nm range to check for wavelength shifts.

The addition of iigands was accomplished using Hamilton syringes. The final concentrations of ligands were: 100 mM NaCI, 2-5 mM KCI, 5 mM MgCI 2, 5 mM Pi, and 77 FM ouabain. Acetylphosphate (Sigma) was prepared by exchanging the Li + for K + via ion exchange on an AG50W-X12 column (Bio-B, ad) with buffer A. Total Pi concentrations present were determined by the method described by Chan et al. [38], while acetylphosphate concentrations were determined by the method of Lipmann and Turtle [39]. Fhe fluorescence results were corrected for the presence of hydrolysis products of acetylphosphate. The total volume changes for the experiments were less than 10% and corrections for dilution effects (unless stated otherwise) were incorporated into the final results.

Hill plots were obtained from the fluorescence intensity titration curves for the various ligands by utilizing the equation: log F ' = n log [l igand]- log k0. 5 [16]. The data are plotted in terms of the change in fluorescence ( F ' ) as a function of ligand concentration where

45

F ' = F/(Fma ~ - F ) . F represents the relative fluorescence at each ligand concentration and Fma ~ is the total change in fluorescence at infinite ligand concentration (deternfined from the y-intercept of the 1 / F vs. 1/[ligand] plot).

Other measurements and procedures. Protein concentrations were determined by the method of Lowry et al. [40] with bovine serum albumin as standard. The purity of the Na+/K+-ATPase, the antibodies, and the specificity of FITC labeling to the a subunit were checked by resolving the proteins with SDS-PAGE using 7.5% Laemmli gels [41]. The gels were stained with Coomassie brillant blue for detection of protein. The FITC-labeled a subunit was visualized in the gels using a $pectroline Model TS-365 transilluminator prior to gel staining.

Results

Antibody 9-A5 binding to and inhibition of the lamb kidney Na + / K +-.4 TPase

The effects of M10-PS-Cll on the activity and partial reactions of native enzyme as well as its ability to bind to the FITC-labded enzyme have been described previou,Ay [27]. Antibody 9-A5, raised to the rat kidney

140[ " 9-A~

~ 130 M10-P5-C11

Control

'7" 120

110 l ,

! I !

loo o 20 20 60 8'o

Na + (mM) Fig. 1. Na + titrations of FITC-labeled Na +/K +-ATPase fluorescence in the presence of M10-P5-CI1 or 9-A5. The enhancement in th,- fluorescence intensity of the FITC.labeled Na+/K+-ATPase [5 vg/m|) caused by increasing amounts of Na + was monitored in the absence of antibody (e) and in the presence of either M10-P5-CI 1 (A) or 9-A5 (S). The antibody concentration was in excess (20-7,0 ~g/ml). The experiments were performed in histidine buffer as aescribed in the Materials and Methods section. The initial fluon:scence in the absence of Na + is normalized to 10070 and each data point is the average of two experiments dor, e in duplicate. Tb,: Hill coefficients (nil) and ko.5 values were determined from plo~:s of log F ' vs. log [Na + ] (raM), which are not shown. The n H vair~,'es for the enzyme (no antibody), and enzyme with MI0-P5-Cll or 9-A5 present were all approx, n H = 1 and the ko.5 values were ~.2 mM, 3.7 mM and 4.1

mM, respectively.

Na+/K+-ATPase, has been shown to bind to Na' /K+-ATPase from several mammalian species [32], but its ability to recognize the lamb kidney enzyme has not been reported. We have found 9-A5 to inhibit 75% of the Na+/K+-ATPase activity of SDS-treated lamb kidney enzyme (> 40% of activity without SDS) with half-maximal inhibition achieved at slightly less than 1:1 ratio of antibody to enzyme. The K+-dependent hydrolysis of p -n i t rophenyl phosphate ( p - nitrophenylphosphatase activity) was similarly inhibited about 75%. Although the antibody and enzyme were usually incubated together prior to initiating the ATPase reaction, the preincubation period was not required and identical results were obtained when 9-A5 was added to the assay cuvette after ATP hydrolysis began.

When a solid-surface adsorption assay (ELISA) was used to indirectly detect antibody binding to the enzyme, 9-A5 did not bind to either our rat or lamb kidney Na+ /K +-ATPase preparations (although Schenk and Leffert [29] had used this technique for their 9-A5 binding studies). Since the state of the adsorbed enzyme is an unknown factor in ELISA assays, we implemented an 'in solution' assay to test antibody binding, These studies demonstrated that 9-A5 bound equally well to the native and FITC-labeled lamb kidney enzyme in the absence of added ligands. It also appeared that Na + and K + had no effect on binding. However, the results of these latter experiments were inconclusive because these ions reduced antibody adsorption to the microtiter plates.

Fluorescence studies with FITC-labeled Na + / K ÷ A TPase

Buffer and cation effects. The FITC-labeled enzyme was used to monitor possible antibody effects on enzyme conformational changes. The labeled enzyme had reduced Na+/K+-ATPase (85-95% inhibited) and K +- dependent p-nitrophenylphosphatase activities (40- 50%), and fluorescence excitation and emission maxima similar to those reported by Karlish [13] and Hegyvary and Jorgensen [14] (495 nm and 520 nm). In addition, Hegyvary and Jorgensen [14] have described the FITC- labeled Na+/K+-ATPase in Tris buffer as being in an E 1 state with Na + having little effect on the probe's fluorescence intensity and with K + quenching the fluorescence 20-25%. Skou and Esmann [15] suggest that in histidine buffer the enzyme exists in the E2 state. In agreement with these data, we found that: (1) the addition of Na + (100 mM) had little effect on the fluorescence of labeled enzyme in Tris buffer, but K ÷ (2-5 raM) caused a 20-25% decrease (Table I); (2) the level of fluorescence in lfistidine buffer was 20-25% lower than in Tris buffer; and (3) the addition of Na ÷ to the labeled enzyme in histidine buffer resulted in a fluorescence increase of approx. 25% (Fig. 1) while K + quenched the fluorescence less than 10% (data not

46

TABLE i

Effect of antibodies MIO.PS.CI 1 and 9.A5 on the fluorescence intensity of FITC.labeled Na +/K +.ATPase

Changes in the relative fluorescence intensity of FITC- Na ÷/K +-ATPase (5 Fg/ml in buffer A) upon the addition of various ligands, at the concentrations given in the Materials and Methods section, were determined in the absence or presence of antibodies MI0-P5-CII and 9-A5 (20.40 ~tg/ml). Controls also included determination of the effects of the nonspecific IgG secreted by cell line P3-X63-Ag8. The values represent the percent decrease in fluorescence relative to the initial fluorescence intensity, and they are the average of at least two experiments performed in duplicate. Standard deviations varied between I and 3%.

Ligand Relative percent decrease in fluorescence at 520 nm

ligand ligand addition addition with (controls) antibody present

M10-P5-Cll 9.A5

lit. value for ligands a

Na + 2 2 1 0

* 23 27 19 22 plus Na * 3 4 3 + 3

Mg 2* 3 i + i 0 plus K* 14 13 3 10 plus Na +

and K + 2 2 2 0

Mg 2+ and Pi 30 31 6 18 plus Na + 4 2 2 0

Mg 2+ 3 I I 0

plus ouabain 32 33 3 - plus P, and

ouabain 35 40 57 t, 44

Mg 2÷ and P, 30 35 8 18 plus ouabain 34 39 57 b 44

Ouabain i 2 1 - plus P, 2 5 0 - plus Mg'*

and P, 31 38 56 b 44

The reported literature values are those of Hegyvary and Jorgensen [14l,

h This level also reflects a reduction from the 25% enhanced level resulting from 9-A5 addition.

shown). The addi t ion of excess M10-PS-CI 1 or 9-A5 to the labeled enzyme caused an enhancemen t in f luorescence intensi ty of approx. 7 - 1 0 and 25~, respectively, regardless of the buffer composi t ion . This indica ted that

an t ibody b ind ing could be moni to red directly by following an increase in f luorescence intensity.

Antibody effects on N¢~ + responses. The effects of b o u n d an t ibody on the N a+- induced changes in fluorescein fluorescence in tens i ty were examined. T i t ra t ions were per fo rmed in h is t id ine buf fe r to obta in m a x i m u m N a + effects. The t i t rat ion plots (Fig. 1) showed that both an t ibodies enhanced the magni tude of the N a +-

caused increase in f luorescence intensi ty but not en-

zyme a f f in i ty (ko.5 = 4 m M ) for N a + nor its Hil l coeff i - cient (n H ---- 1, da ta not shown). T h e in teract ions of N a + with K + were also examined . W h e n N a + was a d d e d with 5 m M K + (Tris buf fer A) a l ready present , it induced the E2K + --* EmNa + t rans i t ion and restored the f luorescence level to near ly that seen in the absence of K + (Table I). The Hill plot of the N a + dependence for this effect showed that K + increased the n H coeff ic ient for N a + f rom 1 to 2 and the ko. 5 for N a + about 10-fold (4 m M to 50 raM, da ta not shown). M10-P5-CI1 a n d

9-A5 had li t t le effect on the K+- induced N a + cooper- at ivi ty (n H = 1.8), but M 1 0 - P 5 - C l l fur ther decreased

enzyme af f in i ty for Na + to a K0. 5 = 82 mM, and 9-A5 reversed m u c h of the K + effect and increased N a +

af f in i ty to K0. 5 - 11 mM. Ant ibody effects on K + Binding. A n t i b o d y effects on

the ac t ion of K + were also invest igated. The K + titra- t ions were pe r fo rmed in buf fe r A to give the m a x i m u m E~ ~ E 2 K * decreases in f luorescence intensity. T h e ad- d i t ion of K + ( 2 - 5 m M ) induced a 2 0 - 2 5 ~ decrease (Table I). Fig. 2 shows that MI0-P5-C11 had little effect on the m a g n i t u d e of the change , bu t there was a de-

crease in K + af f in i ty (k0. 5 increased about 3-fold f rom - 0.08 m M to 0.3 mM). 9-A5 reduced the m a g n i t u d e of f luorescence quench ing and decreased enzyme a f f in i ty

for K + over 5-fold to a K0. 5 -- 0.5 m M . The Hill coeffi-

100 0.s 2

0

• Log (K" IJM)

60 ~ 9-A5

M10-P5-C11

i i i i

K' (mM)

Fig. 2. K + titrations of FITC-labeled Na + /K +-ATPase fluorescence in the presence of M]0-P5-CI 1 or 9-A5. The relative percent decrease in the fluorescence intensity of the FITC-labeled Na+/K+-ATPase (5 /~g/ml) was monitored with increasing amounts of K + in the absence of antibody (e) and in the presence of M]0-P5-CI] (&) or 9-A5 (m). The antibody concentrations were 20-30 /Lg/ml. The experiments were performed in buffer A as described in the Materials and Meth- ods section. Each point is the average from two experiments done in duplicate, inset: Hill plots of K + titrations. The Hill coefficients (n H) and ko. 5 values were determined from the plots of log F ' vs. log [K + ] (pM) as described in the Materials and Methods Section. The n H and ko. s values in the absence of antibody were (®) 1.05 and 0.1 mM: with M10-P5-CII (t) 0.9 and 0.31 raM; with 9-A5 (B) 0.8 and

0.51 raM, respectively.

47

lOO

~ ~,~ ''~-..~ _ mM Na +

80 v o ~ 1~ "C°ntr°l

I I I I

0 1 2 3 4 b

K + (mM)

Fig. 3. Effect of N a + and 9-A5 on the K+-induced decrease in FITC-fluorescence intensity. The fluorescence intensity changes at varying K + concentrations are shown: in the absence of Na + (o); and in the presence of 5 mM Na + (,,,), and 10 mM Na + (n). In addition, the effect of antibody 9-A5 on this fluorescence decrease was determined in the absence of Na + (O) and the presence of 5 mM Na + (A) and 10 mM Na + (m). For all Na + concentrations, the fluorescence intensity levels achieved at 5 mM K + (control and plus 9-A5) were within 1-37o of the maximal effect achieved with 25 mM K +. These experiments were performed in buffer A as described in the Materials and Methods section. The Hill coefficients and k05 values for K + given in the text were calculated as described

previously.

cient of K + binding was not affected by either antibody (Fig. 2 insert, n H = 0.9).

As shown in Fig. 3, the presence of Na + decreased enzyme affinity for K +. With 5 and 10 mM Na + present, the K0. s for K + shifted from 0.1 mM (no Na +) to 0.25 and 0.8 mM, respectively. Na +, like K +, is a heterotropic allosteric regulator and the n H value for K + with 10 mM Na + present was 1.6. The effects of Na + and 9-A5 were cooperative in that together they caused additional reductions in K ÷ affinity and the extent of its effects. 9-A5 shifted the n H values for K + with 10 mM Na + present from 1.6 to 1.9 and its k0. 5 increased to 13 mM Na + enhanced 9-A5 inhibition of the fluorescence change. In conltrast, M10-P5-Cl l showed considerably less interaction with Na ÷ under the same conditions (data not shown).

Antibody effects on conformational changes with Mg 2+ present. Studies by Skou and Esmann [15] of eosin binding (a fluorescein analogue) to Na+ /K+-ATPase have identified Mg 2+ induced fluorescence changes. We found that Mg 2+ caused a 20% increase in FITC fluorescence intensity when the labeled enzyme was in histidine buffer (a pseudo Na+-like E2 ~ EiMg 2÷ transition) and a slight 2-3% decrease in Tris buffer (E 1 - , EIMg2+), which ;is distinguished from Na + effects (Ta- ble I). When antibody effects on the K+-dependent

E1Mg2++ K ÷--, E2Mg2+K ÷ transition (in buffer A) were analyzed, it was found that 9-A5 inhibited about 80% of the expected (14%) fluorescence decrease (Table I and Fig. 4A, curve C), and the addition of Na + had no effect, indicating that 9-A5 essentially prevented the K + transition from occurring. In the absence of Mg 2+, bound 9-A5 enhanced the Na + reversal of the K + titration (Fig. 4A curve B). In addition, with Mg 2+ present, 9-A5 also inhibited more than 80% of the Pi and ouabain, EIMga+---,E2-PiMg 2+ and E2Mg 2+ ouabain transitions, which produce 30 and 32% decreases in fluorescence intensity, respectively (Table I, F{gs. 4B curve B and 4C curve C). However, the addition of Pi subsequent to ouabain caused a 57% fluorescence quenching (Fig. 4C and curve C). The 57% quenching by Pi and ouabain with 9-A5 present is impressive, but if the initial 25% enhancement due to 9-A5 binding is subtracted from the total observed change, the decrease is similar to that normally ob-

"~+ served upon Pi .ouabain addition to EIMg- . These results indicated that 9-A5 stabilizes E1Mg 2÷ and inhibits the E iMg 2+- , E2Mg2+(X) transition (X = K +, Pi or ouabain) but that ouabain and Pi together were able to force the conformational change. Since the two ligands had the same effect regardless of their order of addition, 9-A5 did not appear to be blocking either ligand's access to enzyme. In contrast to 9-A5 antibody. M10-P5-Cl l had little effect on these ligand induced fluorescence changes (Table I).

Antibody effects on E I - P and Ee-P intermediate formation. While the effectiveness of 9-A5 in preventing the E lMg 2+ + X -~ E2Mg2+(X) transitions was demonstrated, antibody effects ota the N a ÷ M g 2+ ATP-dependent phosphoenzyme intermediate formation could not be monitored because the FITC-labeled enzyme does not accept ATP. As an alternative approach, the work of Beauge et al. [42], of Rephaeli et al. [43] and of Campos et al. [4,~,1 indicate that acetylphosphate (Ac - P) can be used :,n place of ATP to catalyze Na + and MgE+-dependent phosphoenzyme intermediate formation and that the E~ ~ E ~ - - P ~ E2-P conversion can be detected using FITC-labeled enzyme. Fig. 5 shows that the extent of the fluorescence decrease caused by Ac-- P v, as unaffected by antibody, but that both antibodies reduced enzyme affinity for A c - P. 9-A5 was the more potent inhibitor. In a reaction medium containing 5 mM Mg 2+ and 20 mM Na +, the apparent affinity for Ac - P was reduced from a k0. 5 = 0.07 mM to values of 0.31 and 0.80 mM by M10-PS-Cll and

9-A5, respectively. The effect of Na + / K +-A TPase confornlation on anti-

body binding. In addition to determining antibody effects on ligand binding, we investigated iigand effects on the antibody induced fluorescence change. By revers- ing the order of additions, i.e., adding antibody to the enzyme with the various ligands already present, the

48 9 - A 5 Mg K Titration

12o ~ J L ~

100

80

Ha Titration

Buffer d . . k ' - " ~ C - " -.--,,...,j B

. . , , j ( Mg K Titration L,~_ i • i t l

" " . - - - Na Titration .~ k,,---~ . . . . - -- , ,h-.--- A

, . . . . . , , . m R . m . " m m -

15 se¢

E1Mg 2+ - E2K+Mg 2+

B . 120

,J ~4

100 ..a U. 0 ,.~ ~ 8o 11

9-A5 l

t t ~ ' - t MO Pi

I ~ Na Titration Buffer

IS sec

" B

~ ~ . . ~ - ~ ' ~ " ~ " " " A

E1M92+ -- E2-PM92+

c

I" 9 - A 5 M 9 0 u a b

LMIOIb

1oo jLJMO/o, ; kBu,,.2

8 0

~e 15 sac ~ - - - A

E1Mo 2+ .~ E2M9 2+ * Ouab " * E2-PMg 2+ • Oul lb

Fig, 4. The effect of MS :+ and 9-A$ on the K +-, and Pi" and ouabaln-induccd decreases in fluorescence intensity. The fluorescence intensity changes with time of the FITC.labeled Na+/g+.ATPas¢ (5 Fg/ml) in buffer A upon the addition of (A) K +, (B) Pl or (C) ouabain as affected by the presence of Mg a+ and the prior addition of antibody an: shown as uncorrected data. The initial intensity changes seen following the addition of antibody in butter A and that caused by the addition of the same volume of buffer only are shown. Na + and K + were added in increments while the final concentrations achieved for the added figands were: $ mM for Mg 2+, K + and Pi; 90 mM for Na+; "/7/~M for ouabain. Antibody concentration was 30 Fg/ml. (A) The lower line (A) shows the intensity changes for the control sample of labeled enzyme upon addition of Mg 2+, then increasing K + concentrations followed by increasing Na + concentrations. Line B shows the K +-dependent fluorescence decrease occurring in the presence of 9-A$ and the subsequent ¢nlmnced ability of Na + to reverse this decrease. The upper line (C) shows that in the presence of Mg z+ and 9.A$, the K+-induuM decrease in fluorescence is prevented. The addition of Na + caused no increase in fluorescence. (B) Line A shows the hue __r~,_~__ace intensity decreases occurring upon sequential addition of Mg 2+, and Pi and then increasing concentrations of Ha +. Line B shows the effect of 9-AS in preventing the Mg2+Pi-dcpendent decrease, with no subsequent Na+-caused increase, (C) Line A shows the sequential fluorescence intensity changes upon addition of Mg 2+, ouabain and then Pi to labeled enzyme. Line B shows the effect of antibody M10-PS-C11 on the enzyme's response to these ligands, Line C shows the effect of 9-A5 in preventing the Mg2+-dependent ouabain-caused reduction in

fluorescence and the subsequent response upon P~ addition.

influence of the conformationai form of the enzyme upon antibody binding was determined. For MI0-PS- Cl l , the extent of fluorescence enhancement was reduced from that of the control (no ligands) in all cases except with Mg 2+ (Fig. 6A), but antibody concentration dependence or affinity was apparently not changed. Similar experiments (Fig. 6B) using 9-A5 showed that with Na + or Mg 2+ present the antibody-caused fluorescence increases appeared unaltered from that of the control. With K + or Mg2+Pi present there was a sub-

stantial additional increase in fluorescence and with Mg2+Pi.ouabain conditions there was essentially no response. These results suggested that 9-A5 binding initiated an E2--" E] reversal in enzyme conformation under K + and Mg2+P~ conditions, but was unable to bind in Mg2+P~ • ouabain conditions. The independence of the two antibody binding sites or epitopes was further demonstrated by the fact that bound M10-P5-Cll had no significant effect on the increase in fluorescence intensity caused by 9-A5 (Fig. 6B).

100

90

80

70

i

- - v •

Control M10-P5-Cl l

I I I I I

0 1 2 3 4 5

Acetyl Phosphate (rnM)

Fig. 5. Acetylphosphate thration of FITC-labeled Na+/K+-ATPase fluorescence in the presence of M10-P5-Cll or 9-A5. The relative percent decrease in FITC fluorescence under Na +/Mg 2 + conditions, upon the addition of varying Ac ..- P concentrations was determined in the absence (®) and presence of antibody M10-P5-CI 1 (A) or 9-A5 (m). The experiments were performed in buffer A as described in the Materials and Methods section. Each point is the average of two

experiments done in duplicate.

49

TABLE II

Effects of 9-A5 on [32P]P i phosphorylation of Na +/K + -A TPase and FITC-labeled Na +/K + -A TPase

Na+/K+-ATPase and FITC-Na+/K+_ATPase (20 /~g/ml) were phosphorylated for 5 min at 37 °C as described in the Materials and Methods section, either in the presence or absence of antibody (80/~g of 9-A5 or control nonspecific sheep IgG). Phosphorylation was done using Mg 2+ (5 mM) and Mg2+.ouabain (0.1 ~M) conditions with 75 taM [32P]P i added for 5 min. The levels of phosphorylation of unlabeled and labeled enzyme are given in parentheses, all other values are given as relative percentages for the enzymes under each ligand condition. Determinations were done at least twice in triplicate.

Relative percent phosphoenzyme

Ligand conditions: Mg 2 + M g 2 +. ouaba in

Na +/K +-ATPase 100 100 (0.47 nmol/mg) (1.5 nmol/mg)

+ Sheep lgG 76 90 + 9-A5 36 31

FITC-Na+/K +-ATPase 100 100 (0.24 nmol/mg) (0.98 nmol/mg)

+ Sheep lgG 62 92 + 9-A5 4 25

Phosphorylation of Na +/K +-d TPase in the presence o f M g 2+ and Pi

T h e 9-A5 s t ab i l i za t i on of the E ] M g 2+ c o n f o r m a t i o n

a n d i n d u c t i o n of the E2 K + - - , E~ c h a n g e s e e m e d

s t r a i g h t f o r w a r d . H o w e v e r , the f luorescence i n c r e a s e ob -

se rved u p o n a n t i b o d y a d d i t i o n u n d e r E 2 - P i M g 2+ c o n d i -

t i ons ra i sed ques t i ons as to w h e t h e r the e n z y m e u n d e r -

w e n t a n E2-P i ~ E n -- P t r a n s i t i o n wh ich was a n a l o g o u s

to that caused by high Na + concentrations in convertir~ E2-P i to the normal reaction E I -- P intermediate [45], or whether antibody trapping of the EIMg "-÷ form during the phosphorylation-dephosphorylation cycling of the enzyme accounted for the fluorescence change. An additional possibility was that 9-A5 binding could perturb FITC-fluorescence changes without actually affecting a

B

40 . . . . . . . ra . . . . . . . . . . . . . . . [] Mg" Pi . .

o

. . - °

O " K

P 30 d Control

Mg Na M 1 0 - P 5 - C l l .=

O ._= 20 U.

P

10

MgP, ' Ouab

1 2 3 4 5 6 7 8

O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . O i g

.,~,.~: " ° ' ' ' " • u Control

m ~ ~ Na

~ . . . . . . . . . . 4:1: . . . . . . . . . . . . . . . . . . . . . . . ~ . . . . . . . . . . . . . . . . . . . . . . Mg'Pi I O ~ "''C1 -- ~ K

--~, ... . . . . . o~ . . . . . . . . . . . . . . . . . . . . . . . ~- . . . . . . . . . . . . . . . . . . . . . . ,~ MgPi .Ouab

| i I i I I I I

1 2 3 4 5 6 7 8

M10-PS-C11 (pg/ml) 9 - A 5 (pg /ml )

Fig. 6. The effects of varying antibody concentrations on the fluorescence of FITC-iabeled Ha +/K +-ATPase in the presence of different iigands. (A) The enhancement ,of f e fluorescence intensity of FITC-iabeled Na+/K+-ATPase was followed upon addition of increasing MI0-P5-C11 concentrations. The cop,~,ions and concentrations of the ligands were as follows: e, no ligands: [ . 100 mM Na +; A. 5 mM K +: c~, 5 mM Mg-'" ' El, 5 mM Mg 2+ Pi; and ~. 5 mM Mg 2+ Pi plus 5 p,M ouabain. All titrations were performed in buffer A as described in the Materials and Methtxis section. (B) The enhancement of the fluorescence of the FITC-labeled Na +/K +-ATPase by 9-A5 titration. The concentration of the iigands were as follows: ®, no iigands; g, 100 mM Na+; A 5 mM K+; o, 5 mM Mg2+; El, 5 mM Mg 2+ Pi; A, 5 mM Mg 2+ Pi plus 5 #M ouabain: and x, 20/tg/ml

MI0-P5-CI 1. All titrations were performed in buffer A as described in the Materials and Methods section.

50

TABLE Iil

Ligand effects on 9.,45 'solution binding" to native and FITC-labeled Na +/K + -.4 TPase

The experimental conditions were as follows: Column A: Ligands were added to enzyme (48 pg/ml), then antibody 9-A5 (2.4 pg/ml) was added and allowed to bind 1 h at 4°C prior to centrifugation. Column B: Enzyme and antibody were incubated together for 1 h, then ligands were added and enzyme-bound antibody was removed by centrifugation 1 h later. Experiments were done in buffer A, while ligand concentrations were 5 mM for MgCI 2, Tris-ATP and Pi, and 1 pM for ouabain. Unbound 9-A5 was determined as described in the Materials and Methods section. The values are the averages of at least two experiments done in triplicate.

Ligands present Percent 9-A5 bound

Mg2 + ATP Pi ouabain Na +/K +-ATPase FITC-Na +/K +-ATPase

A B A B

. . . . 90%4-5 91~+4 + - - - 29_+_12 784-12 384-10 944- 1 - + - - 234- 8 68 4-10 99 4- 1 85 4- 1 + + - - 204-11 264- 8 994- 1 964- 1 + - + - 204- 3 734-11 524- 2 764-10 + - - + 124- 1 394-10 614-20 764- 8 + - + + 254- 2 304- 1 564-10 144-10

conformationai transition. Because of this, 9-A5 effects on E2-P i formation were determined directly. Table II shows that 9-A5 inhibited Pi phosphorylation of native (unlabeled) enzyme 60-70~ under Mg 2+ and Mg 2+. ouabain conditions. The extent of the decrease in E2-P i formation for unmodified enzyme was similar to the observed inhibition of the enzyme's pNPPase activity. 9-A5 also inhibited phosphorylation of the labeled enzyme about 95%, and 755 under Mg 2+ and Mg 2+- ouabain conditions, respectively, which was consistent with the spectroscopy results. At the concentration used

(0.1 pM), ouabain enhanced the level of E 2 - P i formed, which agrees with the recent work of Matsuda and lwata [46], but it did not abolish 9-A5 inhibition. Fur- ther, we found that the addition of antibody after steady-state phosphorylation was achieved caused a similar reduction in the level of phosphorylation. This ruled out the possibility of 9-A5 simply causing a E2-P ~ --) E ! --P transition reversal. Finally, dephosphorylation studies done under E 2 - P i M g 2 + . ouabain conditions showed that 9-A5 had no effect on the rate of E 2 - P i

hydrolysis (data not shown).

TABLE IV

Effect of antibody on [JH]ouabain binding to ?Ca +/K +-ATPase and FITC.labeled Na +/K +-ATPase

Maximal binding determined

Sample: Ha * / K +-ATPas¢

+ 9-A$

FITC-Na +/K +-ATPase, + 9-A$ + MI0-PS-CI I

Relative rates of binding: Na 4/K ÷-ATPase

+ 9-A$

FITC.Na +/K ÷-ATPase + 9-A5 + MIO-P$.CI I

Ligands present:

Relative percent [ 3Hlouabain bound

Mg 2 + Mg 2 +ATP Na ~ Mg2+ATP Mg 2+ Pl

100 100 100 100 50 113 105 77

100 (53) 100 (51) 100 (69) 100 (113) 12 34 61 66 7 87 47 102

100 100 100 100 89 94 100 56

100 (29) 100 (42) 100 (30) 100 (130) 8 38 7 23

87 56 45 107

Na+/K+-ATPase and FITC-Na+/K +-ATPase (15 pg) were preincubated with excess antibody (50 pg) and the various ligands for 15 rain at 4 ° C. [sH]Ouabain binding was conducted at 37 °C for varying times depending upon ligand conditions. The [ 3 H]ouabain binding times used to reach maximum bound levels were as follows: Mg 2+ and Mg2+ATP conditions, 4-6 h; Na + Mga+ATP, 2-3 hrs; and Mg 2+ Pi, 30-45 min. The binding limes used to determine rates of binding were: Mg 2+ and Mga+ATP conditions, 15 n'fin; Na + Mga+ATP, 10 min; and Mg 2+ Pi, 5 rain. The specific activity of [3HIouabain was 600-1000 pCi/mmol. Ouabain equilibrium binding was 1100 pmoi/mg protein and 840 pmol/mg protein, respectively, for Mga+pi and Mg 2+ conditions. The values presented are given as percent relative to each binding condition for both Na+/K+ATPase and FITC-Na+/K+ATPase and they represent the average of at least two experiments clone in triplicate. The values for the labeled enzyme given in parentheses are percentages relative to unlabeled enzyme.

The antibody clearly inhibited enzyme phosphorylation by Pi, while the FITC-labeled enzyme was more sensitive to 9-A5 inhibition and had reduced levels of phosphorylation.

Determination of antibody binding to holoenzyme and FITC-labeled enzyme

Additional 'in solution' experiments were then done in order to provide further corroboration of the fluorescence data, and comparison of the labeled and unlabeled enzyme. As mentioned previously, neither Na + nor K + had an apparent effect on either 9-A5 or MI0-PS-Cll binding. Table III, column A shows, though, that 9-A5 binding to native enzyme was reduced by the prior addition of all of the other regulatory ligands tested (Mg 2+, ATP, Mg2+Pi and Mg 2+. ouabain). If antibody binding preceded ligand addition, the effects of Mg 2+, ATP and Mg2+Pi were substantially reduced, while Mg2+ATP, and Mg2+.ouabain ( -+ Pi ) caused a displacement of bound antibody (Table III, Column B). These data indicated that 9-A5 binding was affected by the conformational form of the enzyme and that after 9-A5 was bound the enzyme could still undergo ligand-induced changes that reduced antibody affinity and promoted dissociation.

The control experiments with FITC-labeled enzyme (Table III) showed similar results in the absence of ligands, but the influence of Mg 2+ was reduced, while ATP and Mg2+ATP had little or no effect. If Mg 2+ was added in excess of ATP however, the expected Mg 2+- only effects were obtained (data not shown). FITC- labeling prevented the ATP effects on antibody binding. In addition the Mg2+pi, Mg2+.ouabain, and Mg2+Pi effects on 9-A5 binding were reduced by the bound probe.

Effects of antibody binding and FITC-labeling on [ 3H] ouabain binding

The fluorescence studies had indicated that 9-A5 prevented Mg 2+ but not Mg2+P~-dependent ouabain binding. Direct determination of [3H]ouabain binding to FITC-labeled enzyme, using lower ouabain concentrations (0.1 #M vs 77 #M), showed that excess 9-A5 did indeed alter ligand regulation of cardiac glycoside binding. Table IV shows that with only Mg 2+ present, 9-A5 reduced both the rate and level of ouabain binding to labeled enzyme about 90%. The addition of P~ signifi- cantly reversed this effect. Interestingly, despite the covalent-linked probe, ATP regulation of ouabain binding was not abolished and ATP reduced 9-A5 inhibition. Comparison of labeled enzyme to unlabeled e~- zyme showed that bound FITC reduced ouabain binding under all but Mg2+Pi conditions, where it actually enhanced the rate of binding. In addition, 9-A5 also inhibited Mg 2+ regulation of ouabain binding to native enzyme, but ATP reversed this effect. These results were

51

consistent with both the fluorescence and antibody binding studies. They also demonstrated that the covalently bound FITC moiety alters ,Mg 2÷ regulation of cardiac glycoside binding. Studies with antibody M10- P5-Cll showed that the labeled-enzyme was also more sensitive to this antibody under Mg 2+ and Na ÷, Mg2+ATP conditions than previously reported for unlabeled enzyme [27].

Discussion

While an enzyme's catalytic activity may depend only upon the three-dimensional arrangement of the amino acids comprising the active site, the way in which these residues arrive at or maintain this arrangement depends upon a much larger portion of the protein's sequence. Inhibition of enzyme activity by a monoclonal antibody bound at or near the active site could result from the steric blocking of substrate access to the site. Inhibition may also result from an allosteric alter- ation of the active site caused by antibody binding at a distant site, or by a 'freezing' of the enzyme into a single conformation that prevents the additional transitions needed to complete catalysis.

Our demonstration that stoichiometric concentrations of 9-A5 inhibit both Na + /K +-ATPase and K +-dependent p-nitrophenylphosphatase activities of the lamb kidney enzyme complements the localization studies by Farley et al. [33] which suggested that 9-A5 binds near the phosphorylation site on a. Antibody 9-A5 does not alter ATP binding [31] nor does it interfere with FITC- labeling of the enzyme (unpublished data). It's binding does perturb the enzyme, causing a Na+-like enhancement of FITC fluorescence. Like Na ÷, 9-A5 reduces the enzyme affinity for K + and converts E2K + to an Em conformation. Therefore it binds at a site which links Na + and K + interactions. Antibody binding augments or duplicates many of the Na+-induced conformational changes, except that its binding alone does not induce cooperativity between either the K + or the Na + sites. In addition, its enhancement of fluorescence occurs independently of the Na+-induced change.

The antibody's most impressive effect though is its stabilization of an EIMg 2+ conformation so that it inhibits the K ÷, Pi and ouabain induced E1 --, E 2 transitions and promotes the E 2 --, E z reversal. Interestingly, while Na + and K + do not appear to effect epitope conformation, the other regulatory ligands tested do, with Mg 2+ reducing antibody affinity but enhancing its Na+-like properties in preventing the K +, Pi and ouabain effects. Mg 2+ alters the conformation of the epitope and the manner in which its links the iigand-induced changes. Although 9-A5 may be binding near the phosphorylation site, it does not appear to be stericaily blocking ligand access because Pi and ouabain together can promote its dissociation and overcome the inhibi-

52

tion, and their order of addition does not matter. 9-A5 inhibition is achieved by prevention of ligand effects on enzyme conformation.

Therefore, 9-A5 has identified an intracellular region of the enzyme which is involved in most of the critical ligand interactions, but which does not appear to reside at a ligand binding site. As demonstrated by Ting-Beall, Beall, Hastings, Friedman and Ball (unpublished data), 9-A5 binding has been found to prevent enzyme crys- tallization. Interestingly, this epitope and its fundamen- tal conformation is highly conserved among Na+/K +- ATPases, since 9-A5 binds to the enzyme from different tissues and species [32]. This is in contrast to a number of other monoclonal antibodies which show either limited or no cross-reactivity with the enzyme from other tissues or species [28,47].

The mechanism of MI0-PS-C11 action is clearly dis- tinct from that of 9-A5. M10.P5.Cll inhibits Na+/K+-ATPase activity, with no effect on p- nitrophenylphosphatase activity. It has similar, but reduced, 'Na÷-like ' effects in that it causes an enhancement of FITC fluorescence and it decreases enzyme affinity for K +, but it does not reduce maximal K + effects. M10-P5-CI1 does not prevent any of the E ! E 2 ligand-induced changes monitored With the fluo- rescent probe. The M10-P5-Cll epitope clearly does not participate in these conformational changes. In addition, this epitope is an exposed accessible region that undergoes limited conformational changes. M10-P5-C11 binding is not altered by ligand binding nor does it alter the enzyme's ability to crystallize into dimeric (a/~)2 two-dimensional crystals (Ting-Beall etal., unpublished data). M10-P5-CI 1 binding is sensitive to major changes in conformation though because the antibody binds poorly to isolated a and not to Western blotted a. This epitope shows species variation, sittce MI0-P5-CI 1 does not bind to the rat kidney enzyme [47].

These studies confirm our earlier hypothesis that MI0-PS-Cll binds to a specific region adjacent to the ATP binding site without altering either ATP binding or the dephosphorylation steps. It does not 'freeze' the enzyme in a particular conformation or block substrate binding, rather it prevents a repositioning or rearrange- ment of bound ATP that is essential for cleavage of the ¥-phosphoryi group and enzyme phosphorylation. It also prevents an ATP-mediated transmembrane shift of a section of the polypeptide that stimulates ouabain binding. This ATP repositioning step may closely re- semble the ATP pseudorotation proposed by Strecken- bach etal . [4] in their studies with spin-labeled ATP. The M10-PS-Cll effect on enzyme phosphorylation by ATP -r :o highly specific in that Na+,Mg2+-dependent Ac -- P catalyzed E i --, E 1 ~ p ~ E2. p transitions are not blocked.

Comparison of 'native" and FITC-labeled enzyme has also proven informative. There are significant alter-

ations in the enzyme due to the covalently linked FITC. In addition to blocking ATP binding and inhibiting ATPase activity, the enzyme's p-nitrophenylphosphatase activity and affinity for K + are also reduced [49]. The inhibition of p-nitrophenyiphosphatase activity may result from a partial steric restriction of p-nitrophenylphosphate access to the phosphorylation site and/or from an altered Mg2+(ATP) binding site.

The [3H]ouabain binding studies show that the effectiveness of Mg 2+ in stimulating the rate of ouabain binding is substantially reduced by FITC labeling. In addition, 9-A5 caused inhibition of the Mg2+-dependent ouabain binding is enhanced by the FITC label, while the Mg2+-promoted dissociation of 9-A5 from the enzyme is reduced. This is in sharp contrast with the observation that ouabain binding regulation by Mg2+Pi is not inhibited, but is instead stimulated. Whether Mg '+ binds at two different sites or just differently depending upon whether ATP or P~ is present is not clear. Bonting et al. [50] have presented data that sug- gests that there are two binding sites for Mg 2+. The present experiments demonstrate that FITC-labeling alters some Mg 2+ effects, and that Mge+p~ binding at the phosphorylation site and Mg 2 + ATP binding at the ATP binding site(s) can independently, as well as coordi- nately, regulate ouabain binding to its extracellular receptor.

The usefulness of using both the antibodies and the FITC-label as probes of Na+/K+-ATPase conformational changes is evident, given the controversy as to whether different techniques such as the mapping of ligand-dependent changes in proteolytic cleavage sites or probe fluorescence changes are detecting aspects of the same or different structural changes. The ability of 9-A5 to differentiate between EiMg 2+ and EjNa + in inhibiting the K + effects complements Jorgensen and Peterson's [51] demonstration of differences in these conformations' sensitivity to tryptic digestion. The effects of ligands on 9-A5 binding to the enzyme demonstrate that its epitope undergoes conformational changes. In addition, the phosphorylation and antibody-binding studies confirm that the FITC fluorescence changes do reflect actual physical alterations in the Na+ /K 4- ATPase structure. The extent to which the various conformational forms of the enzyme differ at the secondary and tertiary structural levels is, however, not known. Gresalfi and Wallace [52] have presented data suggest- ing that the EmNa + and E2 K+ forms differ by a 7-10% change in the overall a helix and /~ sheet structures. Chetverin and Brazhnikov [53] and Hastings et al. [54] report little change in the secondary structure. Our work shows that significant changes occur at the 9-A5 binding site and that they can be transmitted to and link the Na +, K +, ouabain and FITC binding sites. The exact location of the ligand and antibody sites and the nature of these changes have, as yet, not been identified.

53

Acknowledgements

Supported by National Institutes of Health Grants T32-HL-07382 (M.L.F.), R01-HL-32214 and P01-HL- 22619. W.J.B. is an Established Investigator of the American Heart Association.

References

1 Wallick, E.T., Lane, L.K. and Schwartz, A. (1979) Ann. Rev. Physiol. 41,397-411.

2 Jorgensen, P.L. (1982) Biochim. Biophys. Acta 694, 27-68. 3 Glynn, I.M. (1985) in The Enzymes of Biological Membranes

(Martonosi, A.N., ed.), Vol. 3, pp. 35-114, Plenum Press, New York.

4 Racker, E. (1985) Reconstitutions of Transporters, Receptors and Pathological States, pp. 87-103, Academic Press, New York.

5 Repke, K.R.H. (1986) Bioehim. Biophys. Acta 864, 195-212. 6 Skou, J.C. (1988) Methods in Enzymology, Vol. 156 (Fleischer, S.

and Fleischer, B., eds.), pp 1-25, Academic Press, New York. 7 Albers, R.W. (1967) Annu. Rev. Biochem. 36, 727-756. 8 Post, R.L., Kume, S., Tobin, T., Orcutt, B. and Sen, A.K. (1969) J.

Gen. Physiol. 54, 306S-326S. 9 Karlish, S.J.D. and Yates, D.W. (1978) Biochim. Biophys. Acta

527, 115-130. 10 Moczydlowski, E.G. and Fortes, P.A.G. (1981) J. Biol. Chem. 256,

2357-2366. 11 Taniguchi, K., Suzuki, K., Sasaki, T., Shimokobe, H. and Iida, S.

(1986) J. Biol. Chem. 261, 3272-3281. 12 Jorgensen, P.L. (1983) in Current Topics in Membranes and

Transport tBronner, F. and Kleinzeller, A., eds.), Vol. 19, pp 377-401, Academic Press, New York.

13 Karlish, S.J.D. (1980) J. Bioenerg. Biomemb. 12, 111-136. 14 Hegyvary, C. and Jorgensen, P.L. (1981) J. Biol. Chem. 256,

62%-6303. 15 Skou, J.C. and Esw~,nn, M. (1983) Biochim. Biophys. Acta 746,

101-113. 1~, Kapakos, J.G. and Steinberg, M. (1986a) J. Biol. Chem. 261,

2084-2089. 17 Kapakos, J.G. and Steinberg, M. (1986b) J. Biol. Chem. 261,

2090-2095. 18 Farley, R.A., Tran, C.M., Carilli, C.T., Hawke, D. and Shively,

J.E. (1984) J. Biol. Chem. 259, 9532-9535. 19 Kirley, T.L., Wallick, E.T. and Lane, L.K. (1984) Biochem. Bio-

phys. Res. Commun. 125, 767-773. 20 Shull, G.E., Schwartz, A. and Lingrel, J.B. (1985) Nature 316,

691-695. 21 Ohta, T., Nagano, K. and Yoshida, M. (1986) Proc. Natl. Acad.

Sci. USA 83, 2071-2075. 22 MacLennan, D.H., Brandi, C.J., Korczak, B. and Green, N.M.

(1985) Nature 316, 696-700. 23 Shuli, G.E. and Lingrel, J.B. (1986) J. Biol. Chem. 261,

16788-16791. 24 Davis, R.L. and Robinson, J.D. (1988) Bioehim. Biophys. Acta

953, 26-36. 25 Bali, W.J. and Friedman, M.L. (1987) Biochem. Biophys. Res.

Commun. 148, 246-253.

26 Ball, W.J., Schwartz, A. and 1.essard, J.L. (1982) Biochim. Bio- phys. Acta 719, 413-423.

27 Bali, W.3. (1986) Biochemistry 25, 7155-7162. 28 Fambrough, D.M. and Bayne, E.K. (1983) J. Biol. Chem. 258,

3926-3935. 29 Schenk, D.B. and Leffert, H.L. (1983) Proc. Natl. Acad. Sci. USA

80, 5281-5285. 30 Urayama, O., Nagamune, H., Nakao, M., Hara, Y., Sugiyama, H.,

Sato, K. and Nakao, T. (1985) J. Biochem. 98, 209-217. 31 Schenk, D.B., Hubert, J.J. and Leffert, H.L. (1984) J. Biol. Chem.

259, 14941-14951. 32 Leffert, H.L., Schenk, D.B., Hubert, J.J., Skelly, H., Schumacher,

M., Ariyasu, R., Ellisman, M., Koch, K.S. and Keller, G.A. (1985) Hepatology 5, 501-507.

33 Farley, R.A., Ochoa, G.T. and Kudrow, A. (1986) Am. J. Physiol. 250, C896-C906.

34 Lane, L.K., Potter, J.D. and Collins, J.H. (1979) Prep. Biochem. 9, 157-170.

35 Ball, W.J. (1984) Biochemistry 23, 2275-2281. 36 Schwartz, A., Allen, J.C. and Harigaya, S. (1969) J. Pharmacol.

Exp. Ther. 168, 31-41. 37 Wallick, E.T. and Schwartz, A. (1974) J. Biol. Chem. 249,

5141-5147. 38 Chart, K.-M., Delfert, D. and Junger, K.D. (1986) Anal. Biochem.

157, 375-380. 39 Lipmann, F. and Tuttle, L.C. (1945) J. Biol. Chem. 159, 21-28. 40 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J.

(1951) J. Biol. Chem. 193, 265-275. 41 Laemmli, U.K. (1970) Nature 227, 680-685. 42 Beauge, L., Berberian, G., Campos, M. and Pedemonte, C. (1985)

in The Sodium Pump (Glynn, I.M. and Eliory, J.C., eds.), pp. 321-333. The Company of Biologists, Cambridge, U.K.

43 Rephaeli, A., Richards, D. and Karlish, S.J.D. (1986) J. Biol. Chem. 261, 6248-6254.

44 Campos, M., Berberian, G. and Beauge, L. (1988) Biochim. Bio- phys. Acta 938, 7-16.

45 Taniguchi, K. Post, R.L. (1975) J. Biol. Chem. 250, 3010-3018. 46 Matsuda, T. lwata, H. (1987) Arch. Biochem. Biophys. 258, 7-12. 47 Ball, W.J. and Lane, L.K. (1986) Biochim. Biophys. Acta 873,

79-87. 48 Streckenbach, B., Schwarz, D. and Repke, K.R.H. (1980) Biochim.

Biophys. Acta 601, 34-46. 49 Jorgensen, P.L. and Petersen, J. (1982) Biochim. Biophys. Acta

705, 38-47. 50 Bonting, S.L., Swartz, H.G.P., Peters, W.H.M., Steldaoven,

F.M.A.H.S. and DePont, J.J.H.H.M. (1983) in Current Topics in Membranes and Transport, (Bronner, F. and Kleinzeller, A., eds.), Vol. 19, pp. 403-423, Academic Press, New York.

51 Jorgensen, P.L. and Petersen, J. (1979) in Na+/K+-ATPase Struc- ture and Kinetics (Skou, J.C. and Norby, J.G., eds.), pp. 143-155, Academic Press, New York.

52 Gresalfi, T3. and Wallace, B.A. (1984) J. Biol. Chem. 259, 2622-2628.

53 Chetverin, A.B. and Brazhnikov, E.V. (1985) J. Biol. Chem. 260, 7817-7819.

54 Hastings, D.F., Reynolds, J.A. and Tanford, C. (1986) Biochim. Biophys. Acta 860, 566-569.

determination of monoclonal antibody-induced alterations in na+/k+-atpase conformations using...

Documents