Interactions between bacterially expressed cytoplasmic domains of the

Na, K-ATPase

Charles J. Costa^1,2, Kenneth Martin¹, Philip Weeks¹, and Craig Gatto^1
1Department of Biological Sciences, Illinois State University, Normal Illinois
²Department of Biological Sciences, Eastern Illinois University, Charleston, Illinois

Abstract

Information reported over the last several years suggests that the larger cytoplasmic domains of P-type ATPases interact with each other to form the intracellular active site of these enzymes. Further, these cytoplasmic domains probably change positions relative to each other during the pump cycle in order to transduce energy from ATP into ion translocation. The sodium pump α-subunit has 6 cytoplasmic domains: N- terminus, C-terminus, M2M3-loop, M4M5-loop, M6M7-loop, and M8M9-loop. Identification of the conformational changes that occur between the cytoplasmic domains during the pump cycle should help elucidate the mechanism that transfers energy from ATP into an ion motive force. We have overexpressed the N-term (Gly¹-Arg⁸⁸ ), M2M3 (Glu¹⁴³-Glu²⁹³), and M4M5 (Lys³⁵⁴-Lys⁷⁷⁴) domains in bacteria as His₆-tagged or GST-tagged constructs and used them to explore intracellular interactions that may occur during the pump cycle. In our experiments, GST-M4M5 (the ATP binding domain) was immobilized on glutathione-sepharose beads and then incubated with His6-N-term and His6-M2M3 individually, or together. We found that both M2M3 and N-term form stable interactions with M4M5. Further, these interactions are not dependent on the presence of MgATP as is required for the dimerization of the ATP binding domain (Costa et al., JBC 278, 9176-9184, 2003). Investigations are underway to identify the amino acids residues involved in the binding which hopefully provide insight into the coupling mechanism of this protein superfamily.

Introduction

The role of the sodium pump’s cytoplasmic loops has been of interest to those concerned with the changing structure-function relationships that occur during the pump cycle since the mid 1970’s (Jorgensen, 1975 and 1977). More recent functional studies have demonstrated dramatic effects on several parameters when specific regions of the pump’s N-terminal domain were removed (Segall, Lane, and Blostein, 2002). Further, while the crystal structure of the sodium pump has not been resolved, structural comparisons with the SERCA structure indicate the potential for interactions between the 3 largest cytoplasmic domains, N-term, M2M3, and M4M5 – the ATP binding domain (Sweadner and Donnet, 2001).

We are presenting preliminary data that suggest that N-term and M2M3 directly interact with the isolated ATP binding domain and that the observed interactions between N-term and the ABD depend on lysine residues in the N-terminus, consistent with observationgs by Segall, Lane and Blostein (2002).

Methods

Protein Expression and Purification. Escherichia coli was used to overexpress the three largest cytoplasmic domains of the α-1 subunit of rat Na,K-ATPase. The N-terminal cytoplasmic domain was produced as the His₆-fusion protein including residues Gly¹-Arg⁸⁸. The M2M3 loop, also expressed as the HIS-tagged fusion protein, included residues Glu¹⁴³-Glu²⁹³. The M4M5 loop or ATP-binding domain includes amino acid residues from Lys³⁵⁴-Lys⁷⁷⁴ of the rat α-1 subunit. The ATP-binding domain was labeled with glutathione S-transferase (GST) (see Kaplan et al., 1997). The presence of GST or His₆ allows both the isolation of the cytoplasmic domain of interest from other soluble proteins in the bacterial lysate using affinity chromatography and their detection via horseradish peroxidase (HRP) conjugated antibodies.

Binding Assay. Association assays were conducted by binding the GST-M4M5 onto glutathione sepharose resin then incubating the bound GST-ABD with its binding partner(s) in the presence of various ligands (see figures for details). Following a 2-hour incubation period, the proteins bound to the GSH-sepharose were washed thoroughly before eluting the bound proteins with Laemmli sample buffer. Bound proteins were separated by SDS-PAGE and identified on western blots with an anti-His₆ antibody.

Experiments

Interaction between GST-ABD and M2M3 does not depend on the presence of MgATP. Because we have shown that binding between two ATP binding domains depends on the presence of MgATP, we explored the possibility that the same requirement exists for interactions between the ABD and M2M3 loop.  GST-ABD bound to GSH-sepharose was incubated in the presence of MgATP, MgADP, MgP_i, ATP, ADP, and P_i all at 5mM concentration. Tris buffered saline and GSH-sepharose without GST-ABD served as controls. 

M2M3 and N-terminal domains appear to bind the ATP binding domain independently.
In order to begin to explore the interactions that occur between the 3 largest cytoplasmic domains, N-term and M2M3 were incubated with GST-ABD alone and together. The proteins were incubated together with 5 mM MgATP in the medium for 2 hours, then washed well and eluted from the GSH-sepharose. The eluted proteins along with positive controls were separated on a 15% SDS-PAGE and a subsequent western blot probed with anti-His₆ antibody. 

Chemical Modification. In order to begin to identify the amino acid residues responsible for binding between the cytoplasmic domains of the α-subunit, the GST-ABD, His₆-N-term, and His₆-M2M3 constructs were modified with the lysine reagent Sulfo-NHS acetate (Hydroxysulfosuccinimide Acetate Pierce #26777) or the cysteine reagent CPM (7-diethylamino-3-(4'-maleimidylphenyl)-4-methylcoumarin).

In the first experiment, (Figure 3) GST-ABD, His₆-N-term, and His₆-M2M3 were labeled with sulfo-NHS or CPM for 60 minutes at room temperature. The proteins were assayed pair-wise in all combinations. The chemical modification did not disrupt the GST-ABD/ His₆-M2M3 interaction (data not shown). Modification of N-term prevented the interaction regardless of the modification of the ATP binding domain.

In the second experiment, (Figure 4) the time course of lysine labeling was determined by treating His₆-N-term with 100 mM sulfo-NHS for 0, 5, 20, and 60 minutes prior to incubation with unlabeled GST-ABD. Labeling was quenched by diluting the solution 10 fold with tris incubation buffer.

Results

Conclusions

1.) Both the His₆ N-terminal domain and His₆ M2M3 loop bind to GST-M4M5 cytoplasmic loop.

2.) Binding is not dependent on the presence of Mg-ATP, unlike ABD-ABD interactions which is MgATP- dependent (Costa et al., 2003).

3.) Chemical modification of the peptide with sulfo-NHS abolishes binding of N-term with ABD.

4.) These preliminary observations extend the conclusions of Segall, Lane and Blostein (2002), and implicate a possible role for the lysine cluster in the beginning of the N-terminus for “A-domain” interaction with the ABD.

References

Costa, C.J., Gatto, C. and Kaplan, J.H. (2003) J. Biol. Chem, 278, 9176-9184
Jorgensen, P.L. (1975) Biochim. Biophys. Acta 401, 399-415
Jorgensen, P.L. (1977) Biochim. Biophys. Acta 466, 97-108
Segall, L., Lane, L.K., and Blostein, R. (2002) J. Biol. Chem, 277, 35202-35209
Sweadner, K.J. and Donnet, C. (2001) Biochem J. 356,685-704.

Acknowledgement

Support: NIH-GM61583 grant to C.G. & NIH-MARC supplement to K.L.M