Recently, new avenues for understanding the function of P-type ATPases have been opened by the availability of high resolution X-ray structures for a type II (P2A) enzyme, the sarcoplasmic reticulum Ca2+-ATPase (SERCA2). The original structure reported by Toyoshima et al.(2000) confirmed and illuminated previous biochemical and genetic studies, and in particular provided a clear indication of the residues that bind the transported ions. Moreover, it became possible in the overall structure to visualize particular amino acids, such as those conserved in the P-type ATPase family and/or implicated in transport function by mutation studies.
While a structure for a type IV P-type ATPase is not available, some insights into structure/function can be gained indirectly by taking advantage of the similarities between type II and type IV enzymes. Sequence analysis of the type IV subfamily not only indicated clear overall sequence similarity with type II transporters, but also suggested that the Ca2+ transporters and the amphipath transporters share a similar transmembrane domain organization, with two pairs of membrane-spanning helices on the amino terminal side and three pairs of helices on the carboxyl terminal side of the largest cytoplasmic domain. These findings suggest the possibility that type IV enzymes might be threaded onto the structure of the Ca2+-ATPase to generate a three dimensional structural model of the former, with the goal of visualizing interesting regions of the protein sequence.
The amino acid sequence of the Ca2+-ATPase (SERCA2) and the APLT--Amino Phospholipid Transporter (also referred to as "1a") were aligned using the CLUSTAL W program. Then, from viewing the structure of the Ca2+-ATPase, the alignment was refined manually. For example, substantial insertions in the 1a protein sequence that introduced amino acids into the middle of important structural domains were realigned to leave the basic structural elements intact. Similarly, deletions or insertions in transmembrane helices were adjusted to allow them to span the bilayer, as in the structural model, and in some cases were adjusted to align aromatic residues at the membrane interface. At later stages, the sequence was threaded onto the known Ca2+-ATPase structure (1EUL) and highly unfavorable sidechain clashes were identified using DeepView, the Swiss Protein Data Base (PDB) viewer, and the alignment further refined. Finally, prospective alignments were submitted to SWISS-MODEL for energy minimization. Alignments in which sidechain interactions could not be resolved were then further adjusted at the point in the sequence where the minimization failed to give an acceptable structure. The sequence alignment gave rise to the successful homology model shown here on the left side (A). All transmembrane helices are clustered together to form transmembrane domain M which is contained within the hydrophobic lipid bilayer. Above this domain in the figure are the three functional domains exposed in the aqueous compartment: the canonical DKTGTLT phosphorylation sequence of P-type ATPases is contained in the P domain, the N domain contains the ATP binding region, and the A domain is involved in activation.
Once a model is generated, specific amino acids of interest may be identified and visualized. For example, conserved amino acids are of great interest, since they represent functionally important sites under constant surveillance by natural selection. Such residues were identified first by aligning the sequences of 46 type IV enzymes using CLUSTAL W. Residues in the structure were then colored according to the degree to which they were conserved at each position , using the ConSurf server. For comparative purposes, a similar analysis was carried out on 46 homologs of the Ca2+-ATPase on the right side (B). In both cases, there is substantial conservation of the amino acids that comprise the helices in the interior of the transmembrane domain. The cytoplasmic domains are characterized by clusters of conserved sequences on the surfaces of the domains facing each other, with only a small degree of conservation of the residues on the faces of the domains facing outward. These conserved regions include the canonical phosphorylation sequence and nucleotide binding regions of P-type ATPases, and constitute the interacting faces of the domains in the closed, E2, conformation of the P-type ATPases (Toyoshima and Nomura, 2002) that alternates in the catalytic cycle with the more open, E1, conformation shown in these enzyme models (Toyoshima et al., 2000).
Among the amino acid sequences conserved in the type IV subfamily are many that are conserved across the entire family. On the other hand, there are also conserved sequences that are specific to the subfamily, or, in some cases, to specific classes within the subfamily. These subfamily-specific residues are scattered throughout the overall linear sequence, making it difficult to draw any conclusions about their functional implications. When these subfamily-specific sequences are visualized in the structure, however, an interesting pattern emerges. The first four of these conserved sequences (those designated A-D) include about 35 amino acids, and are dispersed over the first 400 residues of the protein (Halleck et al. 1998). As shown here, these residues are not dispersed in the structure, but rather are in immediate proximity and combine to form a single region in the protein. This region, which can be seen at the edges of the convex face on the right of the upper structure in partA of this figure, is not conserved in the corresponding location of the type II enzymes shown on the right side (B), upper structure, which instead contain only a few isolated conserved amino acids. The crevasse formed on the surface of the type IV enzyme (on the left, upper structure) extends from the A cytoplasmic domain down into the transmembrane region exposed to the lipid bilayer. The hydrophobicity of the residues in this crevasse is as expected - a mixture of polar and charged (largely cationic) residues in the cytoplasmic region, and hydrophobic side chains exposed to the bilayer in the transmembrane region. The function of this structural feature cannot be determined with confidence until functional transport assays for type IV enzymes can be developed. Its location, however, suggests that it may form part of a pathway permitting exit or entrance of substrates into the protein interior during the transport cycle, a function consistent with the assignment of a similar role for a conserved aspartate and glutamates in this location in the Ca2+-ATPase (Lee and East, 2001). If so, the larger size of the feature in the type IV enzymes is consistent with a role in the transport of more bulky substrates, such as phospholipids and other amphipaths, in comparison with the simple metal ions transported by the Ca2+-ATPase.
The mechanism of transport
in the P-type ATPases requires linkage between large movements of the three
great cytoplasmic domains and movements of the helices forming the transmembrane
domain that traverses the bilayer. While the cytoplasmic domains
must subsume similar functions of binding substrate and ATP across the
whole family, the membrane helices must include residues that bind the
tranported substrates, with the nature of these residues determining what
substrate is transported by each enzyme. In the sarcoplasmic reticulum
Ca2+-ATPase, two Ca2+ ions are transported in each round of the catalytic
cycle, a finding reflected in the structure of the enzyme by the presence
of two intramembrane Ca2+ binding sites, sites I and II shown
here in panel A. As predicted from earlier
studies, both of these sites include anionic carboxyl groups contributed
by aspartates and glutamates which neutralize the charge of the Ca2+ ion.
Interestingly, the plasma membrane Ca2+-ATPase (PMCA4), a quite different
P-type ATPase than its relative from the cell interior (SERCA2), transports
only a single Ca2+ ion at each turn of the transport cycle, and in this
case, two of the three anionic side chains at site I have been replaced
by uncharged residues (Table 1). The retention
of the third may be related to a role for this residue in transferring
the transported calcium ion into or out of the remaining binding site II.
The congregation of the anionic side chains results in the production of
highly charged binding sites in the interior of the transmembrane domain
(see panel A). The importance of
the corresponding residues in the type IV enzymes is indicated by their
high degree of conservation (Table 1), but as
seen here the nature of the site is quite
different. Unlike the highly charged binding sites in the Ca2+-ATPase,
the sites in the 1a protein, and thus in the subfamily in general, lack
charged residues, and are composed of a mixture of hydrophobic and polar
uncharged amino acids (see panel B).
Such substrate binding sites are consistent with the transport of more
bulky and complex amphipathic molecules, in contrast to the simple charged
ions of the Ca2+-ATPase and its relatives.
References
Toyoshima, C., Nakasako, M., Nomura, H., and Ogawa, H. (2000) Crystal structure of the calcium pump of sarcoplasmic reticulum at 2.6? resolution. Nature 405, 647-655.
Toyoshima, C., and Nomura, H. (2002) Structural changes in the calcium pump accompanying the dissociation of calcium, Nature 418, 605-611.
Lee, A.G.,
and East, M. (2001) What the structure of a calcium pump tells us about
its mechanism, Biochem. J. 356, 665-683.
Clarke, D.M., Loo, T.W., Inesi, G., and MacLennan, D.H. (1989) Location of high affinity Ca2+-binding sites within the predicted transmembrane of the sarcoplasmic reticulum Ca2+-ATPase, Nature 339, 476-478.
Andersen, J.P.(1995) Dissection of the functional domains of the sarcoplasmic reticulum Ca2+-ATPase by site-directed mutagenesis, Biosci. Rep. 15, 243-261.
Zhang, Z.S., Lewis, D., Strock, C., and Inesi, G. (2000) Detailed characterization
of the cooperative mechanism of Ca2+ binding and catalytic activation
in the Ca2+ transport (SERCA) ATPase, Biochemistry 39, 8758-8767.
Felsenstein, J. (1985) Confidence limits
on phylogenies: an approach using the bootstrap, Evolution 39, 783-791.
Sequence alignment of the sarcoplasmic reticulum
Ca2+-ATPase (SERCA2) and the 1a (Atp8a1) transporter. Following alignment
of the two sequences using CLUSTAL W, manual adjustments were made in the
alignment as described in the text. Subfamily-specific consensus
sequences A-J as well as those in P-type ATPase consensus sequences 1-
4 (white letters on black background) are boxed. Transmembrane domains
are underlined and the conserved aspartate characteristic of all P-type
ATPases is marked with an arrow.
Structural models showing conserved residues
in type IIA and type IV P-type ATPases. A CLUSTAL W alignment of
46 type IV P-type ATPases or 46 type IIA P-type ATPases was used with the
ConSurf server to shade either the space-filling structural model of the
1a protein generated by homology modeling using DeepView or (B) the known
space-filling structural model of the SERCA2 Ca2+ transporter (1EUL).
Two views are shown for each structure. Residues are colored according
to degree of conservation from most similar (dark red) to least similar
(dark blue-green). The three cytoplasmic domains, A (activation),
P (phosphorylation), and N (nucleotide binding) as well as the M (transmembrane)
domain are indicated.
Location of type IV-specific sequences in
the structural model of the 1a transporter. View of the space-filling structural
model of the 1a transporter with residues comprising type IV-specific consensus
regions A-D shaded in dark gray using Protein
Explorer. Domains identified above are labeled
with large, open letters.
Comparison of the Ca2+ binding sites of
the SERCA2 Ca2+-ATPase with analogous sites in the 1a transporter.
(A) Amino acids in Ca2+ binding site I, contributed by amino acids in TM(transmembrane
helix)5, TM6, and TM8, and binding site II, contributed by residues in
TM4 and TM6 of the Ca2+ transporter. (B) Amino acids occupying the same
positions in sites I and II in the 1a transporter. Hydrophobic (white),
polar (gray), or charged (black) residues in sites I and II are shown as
space-filling models. Traces of TM1 and TM10 are identified.
The phylogenetic relationship among
all known type IV subfamily sequences from diverse eukaryotic organisms,
and the (type IIB) plasma membrane Ca2+-ATPase, expressed as a Neighbor-Joining
tree, using p-distance, complete deletion and Felsenstein's
bootstrap test (1,000 replications).
Taken from: Schlegel,
R.A., Halleck, M.S., and Williamson, P. (2003) "Phospholipid Transporters
in the Brain" In: Phospholipids in Health and Disease (B.F. Szuhaj, ed.)
American Oil Chemists Society Press, Champaign, IL