3Department of Biological Sciences, Eastern Illinois University, Charleston, IL

Abstract

Ionic gradients in most animal cells are established by the Na pump, a P₂-ATPase. The SERCA structure (Nature 405:647, 00) has provided a wealth of information towards understanding the molecular mechanism of P₂-ATPases, yet questions remain. In particular, what is the native oligomeric state of these proteins? Functional Na pump studies provide no compelling reason for invoking oligomerization (PNAS 97:3195, 00). Even so, there have been reports consistent with oligmerization. We present findings most easily explained by dimerization of two a-subunit domains. We measured interactions between kidney enzyme and 2 bacterially expressed ATP binding domains (His₆ABD or GST-ABD). 1) Both ABD constructs interact with C₁₂E₈-solublized enzyme in the presence of ATP, 2) Interactions are enhanced by Mg, 3) His₆ABD & GST-ABD also associate in the presence of MgATP, 4) Interactions occur when MgATP is bound to either of the protein partners, 5) Interactions do not involve phosphorylation. We propose that the Na pump can dimerize in vivo; dimerization may play a role in interactions between the pump and other proteins. Consistent with this is the finding that a GST-ABD interacted with foreign proteins in a MgATP-dependent manner (Caplan, M., Int. P-type Pump Meeting, Denmark, 2002). Supported by NIH grant HL30315 (JHK); NIH-GM61583 & AHA-SDG (CG)

Introduction

The Na,K-ATPase is an integral membrane protein that plays a central role in ionic homeostasis in animals by mediating the translocation of Na⁺ and K⁺ ions against their electrochemical gradients across the plasma membrane. The Na,K-ATPase functions as a heterodimer comprised of a 105 kDa a-subunit that spans the plasma membrane 10 times (Hu & Kaplan, 2000), and a ~55 kDa b-subunit that has a short cytoplasmic N-terminal domain, a single transmembrane domain, and a large extracellular domain. The b-subunit has no known transport function but has been implicated in targeting the enzyme complex to the plasma membrane (Gatto et al., 2001).

Whether or not the endogenous Na,K-ATPase exists as a single ab protomer or as a higher oligomer (e.g. a diprotomer) remains unresolved. Determining the quaternary structure of the Na pump is a central issue in furthering the understanding of the cell biological role mediated by this important transport system.

In this paper, we report ligand-induced interactions between nucleotide binding domains that provide a plausible mechanism for self-association between a-subunits.

Methods in brief

Binding Assay. Association assays were conducted by binding the HIS6- or GST-ABD onto an affinity resin then incubating the bound ABD with its binding partner in the presence of various ligands (see figures for details). Bound proteins were separated on SDS-PAGE and identified with an appropriate antibody.

Experiments

In order to evaluate the conditions necessary for interactions between GST-ABD and intact sodium pump, C₁₂E₈-solubilized enzyme purified from dog kidney was incubated with bacterially expressed GST-ABD in the presence of the following ligands in Tris buffer at pH 7.4: 50 mM Tris alone; 50 mM NaCl; 50 mM KCl; 3 mM MgAMP; 3 mM MgADP; 3 mM MgATP; 3 mM MgP_i. Panel A was probed with anti-KETYY antibody (specific for a-subunit). Panel B was probed with anti-b antibody. Dog kidney enzyme was assayed for ouabain sensitive ATPase activity before and after detergent treatment to determine the degree of denaturation.

Intact dog kidney ATPase binds most strongly to GST-ABD in the presence of MgATP (Fig 1A). Occasionally, binding was observed in the presence of MgADP as well, suggesting that MgADP may also promote an interacting conformation. Specifically, we observed a MgADP-facilitated interaction in three out of five experiments where the blot was probed with an anti-a antibody and one experiment probed with anti-b (Fig. 1B). GST-ABD/Na,K-ATPase interactions were never observed in the absence of substrates or in the presence of Na, K, or Mg, alone.

ATPase assays indicated that the detergent treated pump was not denatured: untreated and detergent solubilized enzyme specific activities were 17 and 16.5 mmol^.min^-1.mg^-1 respectively.

	Figure 1a
	Figure 1b

HIS₆-tagged ABD also Interacts with full-length Na,K-ATPase

It was possible that the interaction MgATP elicited between the C₁₂E₈-solubilized pump and GST-ABD was a function of the glutathione S-transferase tag and not interactions between the 2 nucleotide binding domains. To ensure this was not the case, detergent-treated pump was incubated with bacterially expressed HIS₆-ABD in the presence and absence of MgATP. The immunoblot in figure 2 was probed with anti-KETYY to identify a-subunit (gift from Dr. Jack Kyte, UCSD).

MgATP elicits the same strong interaction between the HIS₆-ABD and the intact Na,K-ATPase as was observed with the GST construct (center lane, Fig. 2). This is compelling evidence that the interactions observed are indeed mediated by interactions between between the ABD and Na,K-ATPase, not either fusion tag.

Determination of the MgATP concentration required for binding between GST-ABD and detergent solubilized sodium pump

The concentration of MgATP necessary to elicit binding between the GST-ABD and C₁₂E₈-solubilized pump was determined by performing binding assays in MgATP concentrations between 0 - 3.0 mM MgATP (Fig. 3).

Since no binding was observed at MgATP concentrations below 0.5 mM MgATP, it suggests that nucleotide binding by the full-length Na pump alone is not sufficient for the interaction.

The experiments described above provide strong evidence for a MgATP elicited interaction between the nucleotide binding domain and another site on the intact sodium pump, but offer little insight as to the exact region of contact. The nucleotide binding domain contains most of the cytoplasmic regions of the pump so for that reason alone is a strong candidate for the region of contact. We hypothesized that the MgATP elicited interaction takes place between 2 nucleotide binding domains.

We tested this hypothesis by measuring interactions between the two purified cytoplasmic loops themselves. The soluble His₆-ABD was incubated with GST-ABD bound to glutathione-sepharose resin in the presence of varying substrates (see Fig 4 for substrates). The resulting electroblots were probed with anti-penta-HIS.

We observed that the two constructs did associate and that this association depended upon the simultaneous presence of both Mg and ATP (Fig. 4A). That is, neither ATP alone nor Mg alone, were effective promoters of the interaction. MgADP and MgAMP were unable to facilitate interaction between the two nucleotide-binding domains (Fig. 4B). The inability of MgADP to promote interactions between the GST-ABD and the His₆-ABD, was somewhat surprising considering that it did facilitate interactions between the GST-ABD and intact Na,K-ATPase (Fig. 1A). Thus, it may suggest that MgADP binding to the full-length a-subunit elicits a conformation slightly different than when it is bound to the isolated M4M5 loop alone.

	Figure 4A
	Figure 4B

The fluorescent dye eosin has been shown to inhibit the Na, K-ATPase (Skou and Esmann, 1981). Its putative mechanism is competition with ATP for its binding site on the M4M5 loop. In order to begin to test whether the loop-loop interaction elicited by MgATP is due to ATP binding at its usual binding site, we assayed for domain-domain interaction in the presence of 10 mM eosin (Fig. 5).

The presence of 10mM eosin (lanes 1) greatly reduced the binding of His₆-tagged loop to GST-tagged loop in the presence of 1 mM MgATP (Fig. 5).

Both the Coomassie-blue stained gel (left panel) and the anti-penta-His stained western blot (right panel) indicate that a much smaller amount of the His₆-tagged loop bound to the GST-fusion protein in the presence of eosin.

Figure 5.

FITC is a fluorescent amine-reactive molecule that labels Lys⁵⁰¹ in the purified Na,K-ATPase as well as in both the His₆-ABD and the GST-ABD fusion proteins. FITC labeling is prevented by the simultaneous presence of ATP in all 3 proteins (Farley, et al., 1984; Gatto, et al., 1998; Kaplan, et al., 1997, respectively).

In order to determine if MgATP is required on one or both interaction partners, the binding assay was performed using FITC-labeled His₆-ABD and unlabeled GST-ABD (figure 6A-2) compared to both unlabeled proteins (figure 6A-1) and both proteins FITC labeled (figure 6B-1) compared to neither protein labeled (figure 6B-2).

(The FITC labeling protocol used for His₆-ABD modification was identical to those published previously Gatto, et al., 1998.)

When both the GST-ABD and His₆-ABD were labeled with FITC, domain-domain interactions are substantially reduced (Fig. 6B, right panel, lane 1). Figure 6B (middle panel, lane 1) shows FITC incorporation into the GST-ABD, whereas FITC-labeled His₆-ABD is not observed since it did not associate with the GST-ABD and thus was lost during the washing steps. Interestingly, FITC labeling of only one of the interacting fusion proteins did not inhibit the interaction; clearly, FITC-modification of the His₆-ABD (Fig. 6A, middle panel, lane 2) did not significantly reduce its ability to associate with the non-modified GST-ABD (Fig. 6A, right panel).

Figure 6.

Phosphorylation Cannot Explain Domain-Domain Interactions

Purified His₆-ABD, dog kidney Na,K-ATPase, and GST-ABD were incubated with [³²P]ATP on ice as described previously (Pedemonte & Kaplan, 1988). ³²Pi associated with protein denatured a priori with PCA was subtracted from non-denatured protein and the difference is shown in figure 7 as specific ³²Pi incorporation. There appears to be no covalent transfer of phosphate to either fusion protein, whereas native Na,K-ATPase was readily phosphorylated. Data represent means and bars the standard error from three experiments with triplicate determinations.

It is clear that MgATP, and not ATP alone, promotes the observed ABD interactions. An obvious possibility might be that the proteins are undergoing Mg-dependent phosphorylation. However, there was no difference between the K⁺-containing control with the intact Na,K-ATPase and either the His₆-ABD or the GST-ABD, in the presence of Na⁺ or K⁺ (Fig. 7). This is not surprising as the isolated ABDs are devoid of cation-binding sites. Nonetheless, the phosphorylation levels observed for the fusion proteins, compared to the level of ³²Pi captured on the filter with PCA-denatured protein (i.e. background), was minimal. Moreover, considering that the fusion proteins are less than half the molecular mass of the Na,K-ATPase and constitute a more pure protein preparation, the difference in phosphoprotein production between the intact enzyme and the fusion proteins is a very conservative estimate. Finally, this conclusion is supported by reports that the fusion proteins do not possess ATPase activity (Gatto et al., 1998).

Figure 7.

Significance of a-a interactions

The functional significance of the interactions between ABDs is an interesting question. It appears that single pumps can transport Na and K at the expense of ATP so higher-order associations are not necessary to explain primary function.

One plausible explanation is based on Caplan and coworker’s observations that several proteins seemingly unrelated to pump function bind to a-subunit (polycystin, SNAPAP, and protein phosphatase 2a). The association of such cellular regulatory proteins may depend on the pump first self-associating and then forming a binding site (i.e. membrane anchor) for the regulatory partner.

Another explanation is that our system is mimicking intramolecular interactions which normally occur between the N and P subdomains of the M4M5 loop.

Conclusions

Both His₆- and GST-ABD constructs interact with C₁₂E₈-solublized enzyme in the presence of ATP and the interactions are significantly enhanced by the presence of Mg²⁺.  His₆ABD and GST-ABD also associate with each other in the presence of MgATP.  Interactions occur when MgATP is bound to either of the protein partners but are prevented if the ATP binding site on both partners is blocked with eosin or FITC.  Interactions do not involve phosphorylation of the proteins.  We propose that the Na pump is able to dimerize in vivo; dimerization may play a role in interactions between the pump and other proteins. This conclusion is consistent with Caplan’s (2002) finding that a GST-ABD interacted with foreign proteins in a MgATP-dependent manner.

References

Caplan, M. (2002), Ann. N.Y. Acad Sci. 986.

Farley, R.A., Tran, M.C., Carilli, C.T., Hawke, D., and Shively, J.E. J. Biol. Chem. 259, 9532-9535

Gatto, C., Wang, A.X., and Kaplan, J.H. (1998) J. Biol. Chem. 273, 10578-10585.

Gatto, C., McLoud, S.M., and Kaplan, J.H. (2001) Am. J. Phys. Cell Phys. 281, C982-C992.

Hu, Y-K., and Kaplan, J.H. (2000) J. Biol. Chem. 275, 19185-19191

Kaplan, J.H., Lutsenko, S., Gatto, C., Daoud, S., and Kenney, L.J. (1997) Ann. N.Y. Acad. Sci.

Pedemonte & Kaplan, Biochem. 27, 7966, 1988.

Skou, J.C. and Esmann, M. (1981) Biochim. Biophys. Acta. 647, 232-240.