Info

(M

(M

(M

Ka for TNP-ATP

1.89 ± 0.72

10.36 ± 0.46

6.72 ± 1.45

Ka for ATP

268 ± 23

1137 ± 238

339 ± 80

Ka for ADP

85 ± 5

n.d.

n.d.

Ka for AMP

79 ± 18

52 ± 31

168 ± 37

is higher (18-23%), suggesting that the overall fold and some nucleotide-binding properties could be well preserved among all members of the P-type ATPase family.

We expressed the fragment k1010-K1325 of WNDP (Fig. 1) as a histidine tag (HT) fusion in E. coli, purified it from the soluble fraction, and demonstrated that it formed an independently folded domain (ATP-binding domain, or ATP-BD) (55). ATP-BD has both the nucleotide-binding and ATP-hydro-lyzing activities (55); the affinities of the purified ATP-BD for nucleotides are summarized in Table 1.

Analysis of the nucleotide-binding properties of ATP-BD yielded several interesting results. First, the ATP-binding domain of WNDP was found to bind ADP and AMP equally well and with significant affinity (see Table 1), in contrast to previously characterized domains of P2-type ATPases, which show fairly low affinity for these nucleotides and a large difference in the affinity for ADP and AMP (56,57). The lower selectivity of ATP-BD toward nucleotides resembles the property of the P-type ATPase from Methanococcus jannaschii (59), a soluble protein structurally equivalent to the isolated ATP-binding domain, and probably reflects the early evolutionary origin of copper-transporting ATPases.

Interestingly, both ATP-BD and the Methannococcus P-type ATPase have a low but measurable ATPase activity (55,59), a property that has not been observed in the isolated ATP-binding domains of the P2-ATPases. This interesting difference likely reflects a more compact folding of the ATP-binding domain of WNDP and the bacterial P-type ATPases; ATP-BD is 70-80 amino acid residues shorter than the corresponding domain of the P2-type ATPases and lacks several loops, which could be important for precise nucleotide selection in Ca2+-or Na+,K+-ATPase. The molecular modeling of the WNDP ATP-binding domain using the published crystal structure of Ca2+-ATPase further illustrates these points (Fig. 4).

As shown in Fig. 4, the ATP-binding domain of P-type ATPases consists of two distinct parts, the phosphorylation domain (P-domain), which includes the highly conserved residues DKTG, TGDN, and GDGxxD, and the N domain, which contains residues important for binding of the adenosine moiety of nucleotides (23). The P domains of WNDP and Ca2+-ATPase are structurally very similar (i.e., consistent with the common role these regions play in catalytic cycle of ATPases). In contrast, the N domains involved in the nucleotide binding are quite different. The differences in the number, length, and position of several loops in the N domain of P1-type and P2-type ATPases (see Fig. 4) are likely to be responsible for the differences in their nucleotide selectivity (see above).

Another novel and interesting property of the WNDP ATP-binding domain is its ability to bind ATP (the substrate of ATP hydrolysis) and ADP (the product of the reaction) simultaneously (55). Given the distinct subdomain organization of ATP-BD (Figs. 4 and 5), it is tempting to speculate that ATP binds in close proximity to the Asp residue in the DKTG motif (the residue, which in P-type ATPases accepts y-phosphate from ATP, forming phosphorylated intermediate) while ADP is bound in the adenosine-binding pocket of the N domain. During the catalytic cycle, two subdomains would

Fig. 4. Comparison of three-dimensional fold of the ATP-binding domains of WNDP and Ca-ATPase. The homology modeling was carried out using published coordinates for Ca2+-ATPase (accession N. 1EUL) and SwissPdbViewer software. The balls in the lower P-domain indicate the positions for invariant Asp in the DKTG motif, two Asp residues in the GDGxxD sequence, and the location of the TGDN motif. The chain of balls in the upper N domain marks the site of the TNP-AMP (the AMP analog) binding in the crystal structure of Ca2+-ATPase and the equivalent region in the structure of ATP-BD.

Fig. 4. Comparison of three-dimensional fold of the ATP-binding domains of WNDP and Ca-ATPase. The homology modeling was carried out using published coordinates for Ca2+-ATPase (accession N. 1EUL) and SwissPdbViewer software. The balls in the lower P-domain indicate the positions for invariant Asp in the DKTG motif, two Asp residues in the GDGxxD sequence, and the location of the TGDN motif. The chain of balls in the upper N domain marks the site of the TNP-AMP (the AMP analog) binding in the crystal structure of Ca2+-ATPase and the equivalent region in the structure of ATP-BD.

come together [as described for Ca2+-ATPase in (23)], forming a "closed state." The hydrolysis of ATP would then be accompanied by transfer of the adenosine moiety from the P domain to the N domain as shown in Fig. 5 with formation of an "open state" in which both ATP- and ADP-binding sites are ascessible.

It is significant that the nucleotide-binding properties of ATP-BD are modified in the presence of the N-terminal domain (see Table 1 and Section 5.). This change reflects the interaction between two functional domains of WNDP and suggests that domain-domain interactions play an important role in the functional activity of WNDP and homologous MNKP. The ability of ATP-BD to fold independently, to bind and hydrolyze ATP, and to interact with the N-WNDP specifically makes this isolated nucleotide-binding domain a convenient tool for analysis of numerous disease-causing mutations located in this region of WNDP (60,61) (see Fig. 1).

0 0

Post a comment