Biochim. Biophys. Acta (1998) 1365, 37-45

Evolution of P-type ATPases


Michael G. Palmgren*, Kristian B. Axelsen



Department of Plant Biology, The Royal Veterinary and Agricultural University, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Copenhagen, Denmark

Key words: Superfamily; P-type ATPases; Pumps; Regulation; Evolution

*Corresponding author. Fax: +45 3528 3333. Email: palmgren@life.ku.dk

1. Introduction


Ion pumps have evolved independently several times during evolution. Their structure, mechanism of ion translocation and their energy source vary considerably [54]. Thus, various unrelated H+ pumps can be energized by either light, redox energy, pyrophosphate or ATP. H+-pumps driven by ATP may consist of one (P-type) and up to at least thirteen (FOF1 from mitochondria) different subunits.

P-type ATPases are ATP fuelled ion pumps with a single catalytic subunit, and have a phosphorylated reaction cycle intermediate [51]. The pumps become phosphorylated at an aspartate residue in the invariant sequence DKTGT [4]. At least two conformations exist, E1 and E2, with conformational changes being accompanied by ion translocation [37]. Phosphorylation forces the enzyme into the E2 conformation and following dephosphorylation the enzyme returns to the E1 state. In addition to the catalytic subunit, one or two additional subunits may be present.

P-type ATPases have been identified in almost all organisms studied thus far, the only current exception being the parasitic bacteria Borrelia burgdorferi (Table 1). P-type ATPases pump a variety of charged substrates such as K+, Na+, H+, Mg2+, Ca2+, Cu2+, Cd2+ and phospholipids. The number of P-type ATPases in any organism varies considerably (Table 1); from very few in parasitic bacteria to 7 - 9 in free living bacteria, 16 in the yeast Saccharomyces cerevisiae and probably more than 30 in plants and animals.

2. Evolution of conserved domains

P-type ATPases often show very low similarity to each other, making sequence alignments between distant members difficult or impossible. However, in each P-type ATPase, eight conserved segments comprising a total of 265 amino acids can be identified [4]. The conserved domains are situated in the small cytoplasmic loop between transmembrane segments 2 and 3 (4 and 5 in type IB ATPases; Fig. 1), the fourth transmembrane segment (6 in type IB), and the large cytoplasmic loop between transmembrane segments 4 and 5 (6 and 7 in type IB; Fig. 1). These parts of the polypeptide are believed to be involved in communication between ATP hydrolysis and conformational change (the energy transduction domain), in ion binding and in ATP binding, respectively, features that are common between all P-type ATPases [51]. In Na+/K+- and H+/K+-ATPases, the fourth transmembrane segments in addition contain specific targeting information [14].

When extracted from the full-length sequence, the core sequences can be easily aligned and compared. Evolutionary trees, generated from P-type ATPase core sequences present in some of the organisms listed in Table 1, are shown in Figs. 2 and 3. Evolutionary trees of P-type ATPases in a number of organisms (Fig. 2) and one composite tree (Fig. 3) are shown. A tree based on more sequences can be found in ref. 4.

Traditionally P-type ATPases have been divided into families according to a number of criteria such as ion specificity, bacterial or eucaryotic origin [17] or the number of transmembrane segments [44]. In evolutionary terms, a family can be defined as a set of related sequences that always form a monophyletic group in an evolutionary tree. Following this definition, the analysis of 159 P-type ATPase sequences present in the databases has revealed that there are five families of P-type ATPases, each of which may be divided into two ore more subfamilies [4]. These families are indicated in Table 1. Type V ATPases were recently identified as the result of genome sequencing projects [4,11] and seem to be abundant in eukaryotic cells. The function of these pumps is not known.

Any single P-type ATPase in an organism is most often more homologous to P-type ATPases of this class in distantly related organisms than to other ATPases in the same organism (Fig. 3). Closely related P-type ATPases appear to transport the same ions (Table 1). Therefore, it appears that at least some of the ion specificities evolved very early and before the division of bacteria, archaea and eukarya. Type IA ATPases are found only in bacteria and, based on their primitive structure, may be ancestral proteins. Type IB and type IIA ATPases are found in bacteria, archaea and eukarya, and must have evolved early in evolution. Type III ATPases are found in archaea, plants and fungi, but not in bacteria, and therefore must have evolved later. Type IV, type V, type IIB and type IID are only found in eukaryotes and probably evolved after the split between archaea and eukarya.

The P-type ATPase nomenclature described here differs from that found in the literature in the following aspects:

A) Type IA ATPases have been designated their own group by others [44]. However, in the phylogenetic tree these pumps always group together with type IB ATPases [4], and type IA and IB ATPases are, therefore, considered to constitute a monophyletic group.

B) Type II and type III ATPases have previously been placed in a single group [51]. In this analysis, type II and type III ATPases do not group together in the evolutionary tree and are, therefore, not considered a monophyletic group.

C) Type IIA (SERCA: sarco/endoplasmic reticulum Ca2+-ATPases) and type IIB pumps (PMCA; plasma membrane Ca2+-ATPases) have traditionally been named after their intracellular localization in animal cells. However, type IIA ATPases are also found in the Golgi apparatus [61] and in the plasma membrane of plant cells [20] and type IIB ATPases have been identified in an increasing number of intracellular membranes such as the chloroplast envelope [33], the vacuolar membrane [13,47,48] and the endoplasmic reticulum (J.F. Harper, personal communication). Therefore, a nomenclature based on subcellular distribution is inappropriate.

3. Evolution of non-conserved domains

The non-conserved domains of P-type ATPases have evolved into specialized functions and determine ion specificity, binding sites for regulatory proteins and intracellular targeting.

3.1 Membrane-spanning helices are involved in ion translocation
Mutational analyses have shown that residues important for occlusion and binding of ions in type IIA [46,65] and type IIC ATPases [38,42] are situated in membrane spanning helices 5 and 6, showing little homology between the various families, and in the more or less conserved transmembrane helix 4.

3.2 Several residues may be involved in determining ion specificity
When two closely related groups of pumps such as Na+/K+- and H+/K+-ATPases are compared, it is evident that several residues scattered throughout the polypeptide are conserved between all members of one group but not in the other. This suggests that a large number of residues in several domains are important for determining the three-dimensional structure that dictates ion specificity [4].

3.3 Role of terminal domains in regulation
A number of P-type ATPase have extended termini that serve a dual regulatory function: they contain sequences that inhibit pump activity and they contain binding motifs for regulatory proteins which release the inhibition imposed by the autoinhibitory sequences. The non-activated enzyme typically has low affinity for its ligands, ATP and the transported ion, whereas the activated enzyme is in a high-affinity state. It is possible that the low-affinity pumps represent a population of enzymes with reduced rates of dephosphorylation during the catalytic cycle. These would be predominantly in the E2 conformation, which is expected to have low affinity for its ligands.

C-terminal regulatory domains - In the late 1970's it was recognized that the plasma membrane Ca2+-ATPase of animal cells is activated by calmodulin [29,36]. Addition of calmodulin results in an enzyme with higher affinities for Ca2+ and ATP and with increased Vmax. The same kinetic changes are observed after removal of the C-terminus by proteolytic cleavage [68] or by removal of 120 amino acid residues by genetic means [64]. Calmodulin binds with high affinity to a sequence in the C-terminal end of the enzyme [35]. Adding a peptide representing a short sequence close to the calmodulin binding site reverts the C-terminally deleted ATPase to its low-affinity and low-activity state [15]. This autoinhibitory sequence may act by binding to a region close to the transduction domain of the enzyme [18,19]. Calmodulin binding to the Ca2+-ATPase apparently releases the constraint imposed by the autoinhibitory sequence.

Plasma membrane H+-ATPases are regulated by an autoinhibitory sequence in the C-terminal regulatory domain. When the last 12 C-terminal amino acids of the plasma membrane H+-ATPase of the yeast S. cerevisiae are removed at the gene level, the resulting enzyme has increased Vmax and a pH optimum shifted from an acidic pH optimum towards more neutral values [56]. The H+-ATPase becomes activated in vivo when glucose is added to starved cells [59] and the glucose-activated H+-ATPase has the same kinetic characteristics as the enzyme devoid of its C-terminus. After mutating a putative phosphorylation site in the C-terminus of the H+-ATPase, glucose activation in vivo is no longer possible [57]. This suggests a mechanism for glucose activation: glucose application in some way activates a protein kinase which phosphorylates the C-terminus of the H+-ATPase; this covalent modification directly or indirectly results in the displacement of the C-terminus from its intramolecular binding site.

Plant plasma membrane H+-ATPases have an extended C-terminus compared to yeast H+-ATPases. Removal of the C-terminal domain by proteases such as trypsin or chymotrypsin [55], or by truncating 51 amino acids at the gene level [58], results in an activated enzyme with increased Vmax, increased ATP affinity and a pH optimum shifted from acidic towards neutral pH. The plasma membrane H+-ATPase is activated in the same way in vivo by the fungal phytotoxin fusicoccin. Fusicoccin somehow stabilizes the association between the C-terminal domain of H+-ATPase and 14-3-3 proteins [34,53]. 14-3-3 proteins are abundant in eukaryotic cells where they function as mediators in signal transduction pathways and also act directly as modulators of enzyme activity [1]. Fusicoccin binds to a site created in a complex between the C-terminal domain of the plant H+-ATPase and 14-3-3 protein with activation of H+-ATPase as the result [5]. Since 14-3-3 proteins normally only bind to sequence motifs involving phosphorylated serine [50,67] it is possible that fusicoccin action mimicks the effect of protein kinase mediated phosphorylation of a serine residue in the C-terminus of the plant H+-ATPase.

Mutations scattered at several places around the polypeptide of yeast [16] and plant plasma membrane H+-ATPases [49] result in high affinity enzymes. The affected residues may directly or indirectly create an intramolecular receptor for the C-terminal regulatory domain. Alternatively, it is possible that they affect the E1-E2 conformational equilibrium of the enzymes.

N-terminal regulatory domains - A number of plant Ca2+-ATPases identified recently are closely related to animal plasma membrane Ca2+-ATPases and are also activated by calmodulin [30,47]. However, they differ in at least two important aspects. First, they are expressed in internal membranes (see above). Second, the calmodulin binding sites are situated in the long extended N-termini of these enzymes. When the N-terminus is deleted genetically, the resulting enzyme is more active than the wild-type enzyme. This suggests that an autoinhibitory sequence is also situated in the N-terminus. Whether the N- and C-terminal regulatory domains of plant and animal Ca2+-ATPases are evolutionary related is not known.

The heavy metal pumps in the P-type ATPase family (type IB ATPases) have very long extended N-termini which often contain several motifs directly involved in selective binding of heavy metals [45]. Although these have been suggested to be involved in coordinating the transported cation [45], it should be considered whether the N-terminal sequences of type IB ATPases represent regulatory domains that are modulated by binding of the transported ion.

Na+/K+-ATPase [6,21,23,43] and H+/K+-ATPase [62] are phosphorylated in their N-and C-terminal domains by various protein kinases but it has been difficult to demonstrate a link between phosphorylation and changes of activity of the enzymes. It has been hypothesized that the role of phosphorylation is to facilitate the binding of unknown regulatory proteins [22].

4. Evolution of additional subunits

The Kdp ATPase complexes (type IA ATPases) found in a number of bacteria consist of three subunits [2]. The catalytic subunit, KdpB, hydrolyzes ATP and is related to other P-type ATPases although it contains a reduced number of membrane helices. KdpA appears to be involved in binding of K+, the transported ion [10]. It has been hypothesized that Kdp represents an ancestral P-type ATPase and that in modern P-type ATPases the ATP hydrolyzing and the ion binding subunits have become fused [4]. The function of KdpC is unknown but involvement in stabilizing the complex between KdpA and KdpB has been suggested [10].

The Na+/K+- and H+-/K+-ATPases (type IIC ATPases) found in animals have a beta subunit in addition to the catalytic alpha subunit. The beta subunit seems to be important for maturation during transport through the secretory pathway [27], for intracellular sorting in polarized epithelial cells [12] and for determining ion binding properties of the alpha subunit. The evolutionary origin of the beta subunit is not known, but it has been suggested that it represents a remnant of the bacterial KdpC subunit [4]. In addition to the beta subunit, the Na+/K+-ATPase may be associated with a small proteolipid, the gamma subunit, which seems to be important for modulating K+ activation of the pump [7]. The plasma membrane H+-ATPase (type IIIA) of S. cerevisiae is likewise associated with a small proteolipid for which two genes exist, PMP1 and PMP2 [52]. The function of this component remains obscure.
Animal Ca2+-ATPases in internal membranes (type IIA ATPases) are distinct from those in the plasma membrane. They are not activated by calmodulin and have short N- and C-termini. Instead, activity is regulated by a small membrane-bound regulatory protein, phospholamban, which binds to and inhibits these enzymes [3]. There is functional analogy between the action of phospholamban and the C- and N-termini of animal and plant plasma membrane Ca2+-ATPases, respectively [66]. It is, therefore, possible that these autoinhibitory moieties have a common evolutionary origin, and that they, in some cases, fused to the protein they regulate whereas in other cases they remained distinct subunits.

Acknowledgwments

Work in the authors' laboratory is funded by the Danish Agricultural and Veterinary Research Council, the Danish Natural Science Research Council, NOVO Nordisk Fonden, and the European Communities' BIOTECH Programme.

References

[1] A. Aitken, Trends Cell Biol. 6 (1996) 341-347.
[2] K. Altendorf, A. Siebers, W. Epstein, Ann. N.Y. Acad. Sci. 671 (1992) 228-243.
[3] I.T. Arkin, P.D. Adams, A.T. Brunger, S.O. Smith, D.M. Engelman, Ann. Rev. Biophys. Biomol. Struct. 26 (1997) 157-179.
[4] K.B. Axelsen, M.G. Palmgren, J. Mol. Evol. 46 (1998) 84-101.
[5] L. Baunsgaard, A.T. Fuglsang, T. Jahn, H.A.A.J. Korthout, A.H. de Boer, M.G. Palmgren, Plant J. 13 (1998) 661-667.
[6] P. Beguin, M.C. Peitsch, K. Geering, Biochemistry 35 (1996) 14098-14108.
[7] P. Beguin, X.Y. Wang, D. Firsov, A. Puoti, D. Claeys, J.D. Horisberger, K. Geering, EMBO J. 16 (1997) 4250-4260.
[8] F.R. Blattner et al., Science 277 (1997) 1453-1474.
[9] C.J. Bult et al., Science 273 (1996) 1058-1073.
[10] E.T. Buurman, K.T. Kim, W. Epstein, J. Biol. Chem. 270 (1995) 6678-6685.
[11] P. Catty, A.D. D'Exaerde, A. Goffeau, FEBS Lett. 409 (1997) 325-332.
[12] N. Courtois-Coutry, D. Roush, V. Rajendran, J.B. McCarthy, J. Geibel, M. Kashgarian, M.J. Caplan, Cell 90 (1997) 501-510.
[13] K.W. Cunningham, G.R. Fink, J. Cell Biol. 124 (1994) 351-363.
[14] L.A. Dunbar, D.L. Roush, N. Courtois-Coutry, T.R. Muth, C.J. Gottardi, V. Rajendran, J. Geibel, M. Kashgarian, M.J. Caplan, Ann. N.Y. Acad. Sci. 834 (1997) 514-523.
[15] A. Enyedi, T. Vorherr, P. James, D.J. McCormick, A.G. Filoteo, E. Carafoli, J.T. Penniston, J. Biol. Chem. 264 (1989) 12313-12321.
[16] P. Eraso, F. Portillo, J. Biol. Chem. 269 (1994) 10393-10399.
[17] R. Falchetto, T. Vorherr, J. Brunner, E. Carafoli, J. Biol. Chem. 266 (1991) 2930-2936.
[18] M.J. Fagan, M.H. Saier Jr., J. Mol. Evol. 38 (1994) 57-99.
[19] R. Falchetto, T. Vorherr, E. Carafoli, Protein Sci. 1 (1992) 1613-1621.
[20] N. Ferrol, A.B. Bennett, Plant Cell 8 (1996) 1159-1169.
[21] M.S. Feschenko, K.J. Sweadner, J. Biol. Chem. 272 (1997) 17726-17733.
[22] M.S. Feschenko, R.K. Wetzel, K.J. Sweadner, Ann. N.Y. Acad. Sci. 834 (1997) 479-488.
[23] G. Fisone et al., J. Biol. Chem. 269 (1994) 9368-9373.
[24] R.D. Fleischmann et al., Science 269 (1995) 496-512.
[25] C.M. Fraser et al., Science 270 (1995) 397-403.
[26] C.M. Fraser et al., Nature 390 (1995) 580-586.
[27] K. Geering, A. Beggah, P. Good, S. Girardet, S. Roy, D. Schaer, P. Jaunin, J. Cell Biol. 133 (1996) 1193-1204.
[28] A. Goffeau et al., Nature 387(Suppl.) (1997) 5-105.
[29] R.M. Gopinath, F.F. Vincenci, Biochem. Biophys. Res. Commun. 77 (1977) 1203-1209.
[30] J.F. Harper, B. Hong, I. Hwang, H.Q. Guo, R. Stoddard, J.F. Huang, M.G. Palmgren, H. Sze, J. Biol. Chem. 273 (1998) 1099-1106.
[31] Hebsgaard, S.M., Korning, P.G., Tolstrup, N., Engelbrecht, J., Rouze, P. and Brunak, S. (1996) Nucleic Acids Res. 24, 3439-3452.
[32] R. Himmelreich, H. Hilbert, H. Plagens, E. Pirkl, B.C. Li, R. Herrmann, Nucleic Acids Res. 24 (1996) 4420-4449.
[33] L. Huang, T. Berkelman, A.E. Franklin, N.E. Hoffman, Proc. Natl. Acad. Sci. USA 90 (1993) 10066-10070.
[34] T. Jahn, A.T. Fuglsang, A. Olsson, I.M. Brüntrup, D.B. Collinge, D. Volkmann, M. Sommarin, M.G. Palmgren, C. Larsson, Plant Cell 9 (1997) 1805-1814.
[35] P. James, M. Maeda, R. Fischer, A.K. Verma, J. Krebs, J.T. Penniston, E. Carafoli, J. Biol. Chem. 263 (1988) 2905-2910.
[36] H.W. Jarrett, J.T. Penniston, Biochem. Biophys. Res. Commun. 77 (1977) 1210-1216.
[37] P.L. Jørgensen, J.P. Andersen, J. Membr. Biol. 103 (1988) 95-120.
[38] P.L. Jørgensen, J.H. Rasmussen, J.M. Nielsen, P.A. Pedersen, Ann. N.Y. Acad. Sci. 834 (1997) 161-174.
[39] T. Kaneko et al., DNA Res. 3 (1996) 109-136.
[40] H.-P. Klenk et al., Nature 390 (1997) 364-370.
[41] F. Kunst et al., Nature 390 (1997) 249-256.
[42] J.B. Lingrel, J.M. Argüello, J. van Huysse, T.A. Kuntzweiler, Ann. N.Y. Acad. Sci. 834 (1997) 194-206.
[43] N.S. Logvinenko, I. Dulubova, N. Fedosova, S.H. Larsson, A.C. Nairn, M. Esmann, P. Greengard, A. Aperia, Proc. Natl. Acad. Sci. U.S.A. 93 (1996) 9132-9137.
[44] S. Lutsenko, J.H. Kaplan, Biochemistry 34 (1995) 15607-15612.
[45] S. Lutsenko, K. Petrukhin, T.C. Gilliam, J.H. Kaplan, Ann. N.Y. Acad. Sci. 834 (1997) 155-157.
[46] D.H. MacLennan, W.J. Rice, A. Odermatt, Ann. N.Y. Acad. Sci. 834 (1997) 175-185.
[47] S. Malmström, P. Askerlund, M.G. Palmgren, FEBS Lett. 400 (1997) 324-328.
[48] J. Moniakis, M.B. Coukell, A. Forer, J. Biol. Chem. 270 (1995) 28276-28281.
[49] P. Morsomme, A. de Kerchove d'Exaerde, S. De Meester, D. Thinès, A. Goffeau, M. Boutry, EMBO J. 15 (1996) 5513-526.
[50] A.J. Muslin, J.W. Tanner, P.M. Allen, A.S. Shaw, Cell 84 (1996) 889-897.
[51] J.V. Møller, B. Juul, M. le Maire, Biochim. Biophys. Acta 1286 (1996) 1-51.
[52] C. Navarre, P. Catty, S. Leterme, F. Dietrich, A. Goffeau, J. Biol. Chem. 269 (1994) 21262-21268.
[53] C. Oecking, M. Piotrowski, J. Hagemeier, K. Hagemann, Plant J. 12 (1997) 441-453.
[54] M.G. Palmgren, Adv. Bot. Res. 28 (1998) 1-70.
[55] M.G. Palmgren, M. Sommarin, R. Serrano, C. Larsson, J. Biol. Chem. 266 (1991) 20470-20475.
[56] F. Portillo, I.F. de Larrinoa, R Serrano, FEBS Lett. 247 (1989) 381-385.
[57] F. Portillo, P. Eraso, R. Serrano, FEBS Lett. 287 (1991) 71-74.
[58] B. Regenberg, J.M. Villalba, F.C. Lanfermeier, M.G. Palmgren, Plant Cell 7 (1995) 1655-1666.
[59] R. Serrano, FEBS Lett. 156 (1983) 11-14.
[60] D.R. Smith et al., J. Bacteriol. 179 (1997) 7135-7155.
[61] A. Sorin, G. Rosas, R. Rao, J. Biol. Chem. 272 (1997) 9895-9901.
[62] K.T. Togawa, T. Ishiguro, K. Kaya, A. Shimada, T. Imagawa, K. Taniguchi, J. Biol. Chem. 270 (1995) 15475-15478.
[63] J.F. Tomb et al., Nature 388 (1997) 539-547.
[64] A.K. Verma, A. Enyedi, A.G. Filoteo, J.T. Penniston, J. Biol. Chem. 269 (1994) 1687-1691.
[65] B. Vilsen, D. Ramlov, J.P. Andersen, Ann. N.Y. Acad. Sci. 834 (1997) 297-309.
[66] T. Vorherr, M. Chiesi, R. Schwaller, E. Carafoli, Biochemistry 31 (1992) 371-376.
[67] M.B. Yaffe, K. Rittinger, S. Volinia, P.R. Caron, A. Aitken, H. Leffers, S.J. Gamblin, S.J. Smerdon, L.C. Cantley, Cell 91 (1997) 961-971.
[68] M. Zurini, J. Krebs, J.T. Penniston, E. Carafoli, J. Biol. Chem. 259 (1984) 618-627.


P-type ATPase superfamily: Tables

Table 1.

P-type ATPases in a number of organismsa
Accession numberb Typec Specificityd Name/Locus
Arabidopsis thaliana
D89981 IB
PAA1
Z99707 e IB

AC002392 f IB

AC002342 IB
T19K24.9
U93845 IIA Ca2+ ACA3
AF013294 IIA
A_TM018A10.4
L08468 IIB Ca2+ PEA1 / ACA1
AF025842 IIB
ACA2
AC002510 IIB
ACA4 / T32G6.8
P20649 IIIA H+ AHA1
P19456 IIIA H+ AHA2
P20431 IIIA H+ AHA3
X73676 IIIA H+ AHA9
S74033 IIIA H+ AHA10
AB005245 g IV
ALA1
AB005239 f IV

Caenorhabditis elegans
D83665 IB
cua-1
Z92807 h IIA

Z68221 IIB
W09C2.3
U18546 IIC Na+ / K+ eat-6
AF016433 IIC
C09H5.2
AF016446 IIC
C02E7.1
Z81457 h IIC
C01G12.?
Z92850 h IIC

U28940 i IV
T24H7.5
Z81078 IV
F36H2.1
Z93785 IV
W09D10.2
U41552 j V
K07E3.6
Z70271 V
W08D2.5
Z83217 V
C10C6.6
Homo sapiens
Q04656 IB Cu2+ ATP7A
P35670 IB Cu2+ ATP7B
U96781 IIA Ca2+ ATP2A1
P16614 IIA Ca2+ ATP2A2
Z69881 IIA Ca2+ ATP2A3
P20020 IIB Ca2+ ATP2B1
Q01814 IIB Ca2+ ATP2B2
Q16720 IIB Ca2+ ATP2B3
P23634 IIB Ca2+ ATP2B4
P05023 IIC Na+ / K+ ATP1A1
P50993 IIC Na+ / K+ ATP1A2
P13637 IIC Na+ / K+ ATP1A3
P54707 IIC H+ / K+ ATP1AL1
P20648 IIC H+ / K+ ATP4A
Methanobacterium thermoautotrophicum [60] k
AE000825 IB
MTH411
AE000854 IB
MTH755
AE000913 IB
MTH1535
AE000873 IIA
MTH1001
AE000912 IIA
MTH1516
Saccharomyces cerevisiae [28]
P38995 IB Cu2+ CCC2 / ydr270w
P38360 IB Cu2+ PCA1 / ybr295w
P13586 IIA Ca2+ PMR1 / ygl167c
P38929 IIB Ca2+ PMC1 / ygl006w
P13587 IID Na+ ENA1 / ydr040c
Q01896 IID Na+ ENA2 / ydr039c
Z74334 IID Na+ ENA5 / ydr038c
P05030 IIIA H+ PMA1 / ygl008c
P19657 IIIA H+ PMA2 / ypl036w
P39524 IV PL DRS2 / yal026c
P32660 IV
yer166w
P40527 IV
yil048w
Q12674 IV
ymr162c
Q12675 IV
ydr093w
P39986 V
yel031w
Q12697 V
yor291w
Synechocystis PCC6803 [39]
D90910 IA K+ kdpB / slr1729
D64005 IB Cd2+ cadA_1 / slr0797
D64005 IB Cd2+ cadA_2 / slr0798
D90904 IB
sll1920
D90915 IB
slr1950
P37367 IIA Ca2+ pma1 / sll1614
D64005 IIA
pacL / sll0672
D90911 IIA
pacL / slr0822
D90905 IIA
pacL / sll1076 l
Bacillus subtilis [41]
Z99111 IB
ykvW
Z99121 IB
yvgW
Z99121 IB
yvgX
Y13937 IIA
yloB
Escherichia coli [8]
P03960 IA K+ kdpB / f682
P37617 IB
yhhO / o732
Q59385 IB
f834
P39168 IIIB Mg2+ mgtA / o898
Haemophilus influenzae [24]
U32715 IB
HI290
Helicobacter pylori [63]
Q59465 IB
HP0791
P55989 IB
HP1072
AE000648 IB
HP1503
Methanococcus jannaschii [9]
U67563 IIIA
MJ1226
Mycobacterium tuberculosis
Z92539 IA
kdpB / MTCY10G2.19c
Q10866 IB
MTCY39.27
Z92771 IB
MTCY71.10
Z79700 IB
MTCY10D7.05c
Q10876 IB
ctpA / MTCY251.11
Q10877 IB
ctpB / MTCY251.22c
Q10860 IIA
MTCY39.21c
Q10900

MTCY22G10.22c
Z84724

MTCY21C12.02
Mycoplasma genitalium [25]
P47317

pacL / MG071
Mycoplasma pneumoniae [32]
P78036

mgtA / G07_872V
Archaeoglobus fulgidus [40]
AE001096 IB
AF0152
AE001071 IB
AF0473
Borrelia burgdorferi [26]
-m




Footnotes:
a The genomes of Arabidopsis thaliana, Caenorhabditis elegans and Homo sapiens are not fully sequenced. For each protein the accession number, protein name or locus name shown in bold are used in Figures 1 and 2. All data processing was performed using Biobase (The Danish Biotechnological Database).
b The accession numbers starting with a Q or P are from the SWISS-PROT database. All other accession numbers are from the EMBL database.
c The Type to which the P-Type ATPases belong are based on the phylogenetic tree shown in Fig. 2. The Types are defined in ref. 4.
d The ion specificity of the P-type ATPases is shown, when it is known. PL: Phospholipids.
e The proposed translation of the genomic sequence lacks several of the conserved regions of P-type ATPases [4]. The correct position of exons was determined by combining knowledge of amino acid composition from sequence alignments with other P-type ATPases and prediction of splice sites performed with NetPlantGene [31].
f No protein sequence has been provided with the genomic DNA data. Positions of splice sites has been elucidated by combining knowledge of amino acid composition from sequence alignments with other P-type ATPases and prediction of splice sites performed with NetPlantGene [31].
g No protein sequence data is provided with the DNA data. The amino acid sequence has kindly been provided by Dr. Eric Gomes.
h Only genomic sequence data is available and there is no proposed translation. The conserved segments have, therefore, been extracted directly from the genomic DNA data.
i The proposed translation of the genomic DNA sequence lacks the region around one of the universally conserved segments of P-type ATPases. The correct sequence has been elucidated by alignment with an expressed sequence tag (D68066) covering the concerned area of the protein.
j The proposed translation of the genomic DNA sequence lacks several of the conserved regions. The conserved segments have, therefore, been extracted directly from the genomic clone.
k Two fragments of a Type IIA P-type ATPase have been designated the locus names MTH481 and MTH482.
l The position of this sequence might change. In the investigation performed in ref. 4 the sequence was closer to type IIC ATPases.
m No P-type ATPases could be found in this genome.


Fig. 1.
Overview of the P-type ATPase superfamily.

Families are designated by roman numerals on the left followed by the name of the transported ion. Boxes indicate transmembrane segments; Filled circles, inhibitory sequences; Open circles, heavy metal binding sites. Abbreviations are: PL, phospholipids; NAS, no assigned specificity; plb, phospholamban. The ( subunit of type IIC ATPases so far has only been shown to be associated with Na+/K+-ATPase isozymes. It is uncertain whether association with a proteolipid is a general feature of type IIIA ATPases. This is indicated by a question mark.




Fig.2.
Phylogenetic trees of P-type ATPases present in six organisms, Synechocystis PCC6803, Methanobacterium thermoautotrophicum, Saccharomyces cerevisiae, Arabidopsis thaliana, Caenorhabditis elegans, and Homo sapiens.

Organisms from which the genome has been fully sequenced are indicated by an asterisk. Type IV and V ATPases are present in A. thaliana, C. elegans, and H. sapiens as evident from analysis of expressed sequence tags (ESTs) in the databases [4], but have only been included in the trees if full-length sequences have been reported. The tree was constructed using the neighbour joining method. Sequence alignments were based on conserved core sequences (265 amino acid residues in total [4]) extracted from the full-length sequences. The trees were bootstrapped 1000 times. The numbers at the nodepoints represent the number of times that particular node was present in the replicas.




Fig. 3.
Composite phylogenetic tree based on the P-type ATPases present in the organisms listed in the legend to Fig. 2.

When the substrate specificity of the ATPases present in each family is known, it corresponds in all cases to the name of the family. The name of branches corresponds either to the protein names, the locus names, or to the accession number of the DNA sequences in which the P-type ATPase sequence could be found. Table 1 lists all the P-type ATPases shown in the phylogenetic tree. The abbreviations are HM: heavy metals; NAS: no assigned specificity; PL: phospholipids.