STRUCTURAL BIOLOGY Cryo-EM structures capture the ...STRUCTURAL BIOLOGY Cryo-EM structures capture the transportcycle of the P4-ATPase flippase Masahiro Hiraizumi1,2, Keitaro Yamashita1,3,

STRUCTURAL BIOLOGY

Cryo-EM structures capturethe transport cycle of theP4-ATPase flippaseMasahiro Hiraizumi1,2, Keitaro Yamashita1,3, Tomohiro Nishizawa1*, Osamu Nureki1*

In eukaryotic membranes, type IV P-type adenosine triphosphatases (P4-ATPases)mediate the translocation of phospholipids from the outer to the inner leaflet andmaintain lipid asymmetry, which is critical for membrane trafficking and signalingpathways. Here, we report the cryo–electron microscopy structures of sixdistinct intermediates of the human ATP8A1-CDC50a heterocomplex at resolutions of2.6 to 3.3 angstroms, elucidating the lipid translocation cycle of this P4-ATPase.ATP-dependent phosphorylation induces a large rotational movement of the actuatordomain around the phosphorylation site in the phosphorylation domain, accompaniedby lateral shifts of the first and second transmembrane helices, thereby allowingphosphatidylserine binding. The phospholipid head group passes through thehydrophilic cleft, while the acyl chain is exposed toward the lipid environment.These findings advance our understanding of the flippase mechanism and thedisease-associated mutants of P4-ATPases.

In eukaryotic cells, the phospholipid com-positions differ between the outer and in-ner leaflets of the plasma and organellarmembranes: phosphatidylcholine (PC) andsphingomyelin are enriched in the outer

leaflet, whereas phosphatidylserine (PS) andphosphatidylethanolamine (PE) are confined tothe inner leaflet (1). The maintenance and dis-ruption of the asymmetric composition affectprocesses, such asmembrane biogenesis, mem-brane trafficking, signaling, and apoptosis. Threetypes of transporters—scramblase, floppase, andflippase—have been reported to function as phos-pholipid translocators (2–4). Scramblases catalyzebidirectional phospholipid translocations thatdissipate the membrane asymmetry withoutenergy consumption. In contrast, flippase andfloppase, which use the energy of adenosinetriphosphate (ATP) hydrolysis to mediate specificphospholipid translocations against their con-centration gradients, maintain the asymmetricphospholipid composition. ATP-binding cassettetransporters function as floppases that drive theinner-to-outer translocation of lipids, whereas typeIV P-type ATPases (P4-ATPase) are flippases thatdrive the outer-to-inner translocation of lipids.The transport by P-type ATPases occurs essen-

tially according to the Post-Albers mechanism(5, 6), wherein ATP hydrolysis–coupled phospho-rylation and dephosphorylation within the cyto-plasmic ATPase domain mediates the transition

between two intermediate states, E1 and E2,which have different affinities for the substrates,enabling the substrate transport across themem-brane (fig. S1A). Among the P-type ATPase familymembers, the P1- to P3-ATPases are ion trans-porters, with the most representative being theP2-ATPase family, which includes the sarcoplas-mic reticulumCa2+pump(SERCA),Na+/K+-ATPase,and H+/K+-ATPase. The P4-ATPase is the onlymember that functions as a lipid transporter (7).The reaction of the P4-ATPases also follows thePost-Albers mechanism, but it simply translocatesphospholipids from exoplasmic leaflet to cyto-plasmic leaflet and does not require any counter-transported substrate (fig. S1B) (8). The humangenome encodes 14 P4-ATPase subclasses, whichdiffer in their lipid selectivities and tissue expres-sion (9). For most P4-ATPases, heterodimeriza-tion with a CDC50 family protein is essential forproper expression and flippase activity (10, 11).The first P4-ATPasemember identified, ATP8A1,

was found in bovine erythrocytes and chromaffingranules as an aminophospholipid translocase(12, 13). P4-ATPases, includingATP8A1, are presentin plasma and organellarmembranes and seques-ter PS lipids from the outer to the inner leafletin resting cells. In apoptotic cells, P4-ATPasesare cleaved and inactivated by proteases such ascaspases and calpains. The PS that is subsequentlyexposed on the cell surface acts as an “eat me”signal to induce phagocytosis (14, 15). Further-more, the ATP8A1-catalyzed flipping of PS inthe organellar membrane is necessary for thetransport of recycling endosomes, membranefission, and cell migration (16, 17). Several diseasesare associated with P4-ATPases. For example,ATP8B1 mutations cause the liver diseases knownas benign recurrent intrahepatic cholestasis 1 andprogressive familial intrahepatic cholestasis 1,ATP10A is associated with type 2 diabetes and

insulin resistance, and ATP11A is associated withcancer (18). Furthermore, ATP8A1 and ATP8A2have been identified as causative genes for neuro-logical disorders. ATP8A1 knockout mice showhippocampus-dependent learning deficits associ-ated with the exposure of PS on the outer surfaceof the plasma membrane in hippocampal neu-rons (19–21).As comparedwith the canonical ion-transporting

P-type ATPases, P4-ATPase has a large transportsubstrate and thus is expected to use a differentmechanism for substrate recognition and trans-location (22–24). However, despite substantialefforts, themolecular mechanism underlying thelipid flippase activity by the P4-ATPases has re-mained elusive. Here, we report the cryo–electronmicroscopy (cryo-EM) structures of the humanATP8A1-CDC50a heterodimer complex in its sixdistinct intermediates: an apo state (E1), the non-hydrolyzable ATP analog b,g-methyleneadenosine5′-triphosphate (AMPPCP)–bound state (E1-ATP),the adenosine diphosphate–inorganic phosphate(ADP-Pi) analog AlF4

−-ADP–bound state (E1P-ADP), the phosphate-analog AlF4

−-bound tran-sient phosphorylated state (E1P), the BeF3

−- boundphosphoenzyme ground state (E2P), and theAlF4

−-bound dephosphorylation state with the sub-strate phospholipid (E2Pi-PL), revealing the trans-port cycle along the lipid flipping reaction.

Overall structure

We performed a cryo-EM analysis of the P4-ATPase lipid translocator family to elucidate thelipid translocation mechanism (Fig. 1). We ex-pressed full-length human ATP8A1 and humanCDC50a together in mammalian human embry-onic kidney–293F cells and purified the complexin glycol-diosgenin (GDN) micelles (fig. S1C). SDS–polyacrylamidegel electrophoresis analysis showedhigher molecular weight bands of CDC50a, prob-ably derived frommultiple glycosylations (fig. S1D)(25, 26). The purified ATP8A1-CDC50a complexshowed PS-dependent ATPase activity, with aMichaelis constantKmof 111 ± 26.4 mMand amaxi-mum rate of reaction Vmax of 99.7 ± 9.50 nmolmin−1 mg−1, as well as weak PE-dependentATPase activity (Fig. 1C), consistent with previousreports (13, 17). The ATPase activity was inhibitedby general inhibitors of P-type ATPases, such asberyllium fluoride (BeF3

−) and aluminum fluoride(AlF4

−) (fig. S1F). The purified ATP8A1-CDC50acomplex was subjected to cryo-EM single-particleanalyses under several different conditions; name-ly, without any inhibitors and in the presence ofAMPPCP, ALF4

−-ADP, BeF3−, and ALF4

− (fig. S1B).The acquired movies were motion-corrected andprocessed in RELION 3.0 (27), which providedcryo-EMmaps at overall resolutions of 2.6 to 3.3 Å,according to the gold-standard Fourier shell cor-relation 0.143 criterion (figs. S4 to S8). The flexiblecytoplasmic ATPase domain is most stabilizedin the AlF4

−-ADP and BeF3−-bound states, allow-

ing the de novo modeling of almost the entireATP8A1-CDC50a complex, except for someminordisordered regions (Fig. 1, A and B, and fig. S9).The overall structure shows the typical P-typeATPase fold, composed of three large cytoplasmic

RESEARCH

Hiraizumi et al., Science 365, 1149–1155 (2019) 13 September 2019 1 of 7

1Department of Biological Sciences, Graduate School ofScience, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,Tokyo, 113-0033, Japan. 2Discovery Technology Laboratories,Innovative Research Division, Mitsubishi Tanabe PharmaCorporation, 1000 Kamoshida, Aoba-ku, Yokohama, 227-0033, Japan. 3RIKEN SPring-8 Center, 1-1-1 Kouto, Sayo-cho,Sayo-gun, Hyogo 679-5148, Japan.*Corresponding author. Email: [email protected] (T.N.);[email protected] (O.N.)

on June 5, 2021

http://science.sciencemag.org/

Dow

nloaded from


domains (A, actuator; N, nucleotide binding; P,phosphorylation) and ten membrane-spanninghelices (M1 to M10). CDC50a has two transmem-brane helices (TM1 and TM2) at the N and Ctermini, an ectodomain consisting of an antipar-allel b-sandwich (b1 to b8), and extensions of ~60and 70 amino acids with less secondary structurein the b3-b4 and b5-b6 loops, respectively, whichare stabilized by two intrachain disulfide bonds(fig. S3D). The three N-linked glycosylation sitesof CDC50a, which are important for the properfolding as well as the membrane trafficking ofthe P4-ATPases, are clearly visible in the cryo-EMmap (fig. S3, D and E) (11, 28).

Interaction between ATP8A1 and CDC50a

CDC50a is an essential component for P4-ATPasesand is required for the proper expression andfolding of ATP8A1 (fig. S1E) (10, 11). CDC50aand ATP8A1 interact extensively through the

extracellular TM and intracellular regions (fig. S3).In the extracellular region, the CDC50a ectodo-main covers all of the extracellular loops of ATP8A1,except for the M1-M2 loop, interacting in an com-plementary electrostaticmanner: the extracellularloops of ATP8A1 bear negative charges, whereasCDC50a bears positive charges (fig. S3, A and B).In particular, Asp961 andGlu1026 of ATP8A1 form asalt bridgewith Arg262 of CDC50a. In addition, theM3-M4 loop of ATP8A1 extends toward CDC50a,and the two bulky residues at the tip of the loop,Trp328 andTyr329, formhydrophobic interactionsinvolving Phe127, Tyr299, Pro300, Val301, and theN-glycan attached to Asn180 of CDC50a (fig. S3E).In the TM region, several bulky residues, such asTrp942, Ala947 (M9), Met1038, Phe1042, and Leu1049

(M10) of ATP8A1, and Phe54, Ile57, Phe61 (TM1),Phe324, Leu325, Ala328, Tyr329, and Val332 (TM2) ofCDC50, are engaged in the complex interaction(fig. S3G). Furthermore, we observed a strong

planar density at the interface between M7 andM10 of ATP8A1 and TM2 of CDC50a (fig. S3C),which could be assigned to the cholesteryl hemi-succinate added during solubilization. Therefore,cholesterol may bind to the same site and facil-itate the heterodimeric interaction of ATP8A1and CDC50a. In the cytoplasmic region, the N-terminal tail of CDC50a adopts an unstructuredloop conformation that extends parallel to theplasmamembrane and interacts with theM6-M7and M8-M9 loops and the short segment con-necting M4 and the P domain (fig. S3F). Overall,CDC50a envelops the bulk of the TM segmentsand forms extensive interactions with ATP8A1,which explains the chaperone activity of CDC50afor the P4-type ATPases.

Entire transport cycle of P4-ATPase

The cryo-EM structures revealed the clear den-sities of the inhibitors in their respective maps,


Fig. 1. Biochemical and cryo-EM studies of the ATP8A1-CDC50acomplex. (A) Topology diagram of ATP8A1-CDC50a. Conserved domainsand TM helices are schematically illustrated. In the cytoplasmic regions,the A, N, and P domains and the C-terminal regulatory domain are coloredyellow, red, blue, and green, respectively. M1-M2 and M3 to M10 ofATP8A1 are purple and orange, respectively, and CDC50a is pink. TheN-glycosylation sites are shown as sticks. cyto, cytoplasmic side; exo,exoplasmic side. (B) Overall structure of ATP8A1-CDC50a complex.

Cryo-EM maps (top) and ribbon models (bottom). The same color schemeis used throughout the manuscript. (C) Phospholipid-dependent ATPaseactivity of ATP8A1. Data points represent the mean ± SEM of three to sixmeasurements at 37°C. By nonlinear regression of the Michaelis-Mentenequation, ATP8A1-CDC50a in GDN micelles has a Km of 111.0 ± 26.4 mMfor 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (POPS) and amaximal ATPase activity of 99.7 ± 9.5 nmol min−1 mg−1. POPE, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylethanolamine.

RESEARCH | RESEARCH ARTICLEon June 5, 2021


Dow

nloaded from


bound at the catalytic site of ATP8A1 and sta-bilizing the ATPase domain in different conforma-tions (Fig. 2), whereas CDC50a adopted almostthe same conformation in all of these states [rootmean square deviation (RMSD) (Å) = 0.24 to 0.66].

Most notably, the TM region of ATP8A1 remainsstructurally rigid throughout the transport cycle,probably because of the tight association withCDC50a, in contrast to other P-type ATPases, suchas SERCA (figs. S10 and S11 and movie S1).

The structures obtained under three condi-tions, namely without inhibitors, with AMPPCP,and with AlF4

−-ADP, describe the conformationalchanges upon ATP binding and autophospho-rylation, which correspond to the E1, E1-ATP, andE1P-ADP conformations, respectively, in the Post-Albers scheme (Fig. 2A). The densities of the Nand A domains are only weakly visible in the E1state, indicating the highly flexible motion of thesedomains without any ligand (E1 in Fig. 2 and fig.S4). The particles were then classified according tothe densities of the N and P domains, and thesedomains were modeled into the class with thestrongest densities, which probably representsthemost likely arrangement in theE1 state (class 2in fig. S4). The particles of the AMPPCP-boundstate can be classified into three similar confor-mations, wherein the N and P domains adoptslightly different orientations (fig. S5). The com-parison of these classes indicated that ATP bindingat the N domain induces the mutual approach oftheNandPdomains. The density for theAMPPCPis most clearly visible within the class where thesedomains are proximal and bridged by AMPPCP(E1-ATP in Fig. 2B and fig. S5D): The adenine ringinteractswith Phe534 of theNdomain,whereas thephosphate group interacts with Asp409 and Thr411

(the DKTGmotif), Asn789, and Asp790 at the phos-phorylation site of the P domain, in cooperationwith a Mg2+ ion. The AlF4

−-ADP–bound state issimilar to the E1-AMPPCP conformation, but theNand P domains are more tightly bridged by ADPand AlF4

− (E1P-ADP in Fig. 2 and fig. S6), capturedin the phosphoryl transfer intermediate (E1P-ADP).Overall, ATP binding and the subsequent

phosphoryl transfer reaction induce the proximalarrangement of the N and P domains, which isaccompanied by a slight outward shift of the Adomain by ~6.5 Å (E1, E1-ATP, and E1P-ADP inFig. 2A, fig. S10, and movie S1). The phosphoryl-ation reaction is mediated by the motions of theATPase domain and does not require any changesin the TM region. The TM segments of ATP8A1adopt almost the same conformation throughoutthe transition (fig. S11), which is consistent withthe substrate-independent autophosphorylationof P4-type ATPases (8).The two phosphate analogs, BeF3

− and AlF4−,

occupy the phosphorylation site in a similar man-ner, but their coordination geometries are slightlydifferent (E1P, E2P, andE2Pi-PL in Fig. 2). BeF3

− iscovalently attached to the carboxylate side chainof Asp409, in coordination with a Mg2+ ion, andcaptures the phosphoenzyme ground state (E2Pin Fig. 2B). The A domain is tightly fixed to thephosphorylation site (E2P in Fig. 2A and fig. S7)through the backbone carbonyls of Asp189 andGly190 in the conservedDGETmotif (residues 189to 192) (E2P in Fig. 2B). The N domain is pushedapart from the P domain and no longer has accessto the phosphorylation site, thus representing theADP-insensitive E2P state (8). The particles of theAlF4

−-bound state could be separated into twodifferent classes (fig. S8), and both showed clearAlF4

− density at the phosphorylation site. In thefirst class, the bound AlF4

− does not mediate anyinterdomain interactions, and the catalytic domains


Fig. 2. Entire transport cycle of ATP8A1-CDC50a. (A) The six different intermediates ofATP8A1-CDC50a during the phospholipid translocation cycle are shown, arranged clockwise asin the Post-Albers reaction cycle: E1, E1-ATP, E1P-ADP, E1P, E2P, and E2Pi-PL.The bound inhibitors areshown in space-filling model representations. (B) Comparison of the phosphorylation sites in eachintermediate. AMPPCP and ADP are shown as sticks, and AlF4

− and BeF3− are shown as spheres.

Densities are shown as green mesh, contoured at 3.5s. Single-letter abbreviations for the aminoacid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu;M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.



Dow

nloaded from


adopt a conformation similar to the AlF4−- ADP

bound state (E1P-ADP and E1P in Fig. 2B), likelyrepresenting the transient phosphorylated state(E1P) immediately after the ADP release. In thesecond class, AlF4

− mediates the interaction be-tween the N and A domains through the DGETmotif in a similar manner to the BeF3

−-boundstate, but the A domain is rotated by ~22° aroundthe phosphorylation site, as compared with theBeF3

−-bound state (Fig. 3A). This allows the re-positioning of the carboxyl side chain of Glu191 intheDGETmotif to provide a catalytic base for thedephosphorylation reaction (E2PandE2Pi-PL inFig.2B) (8), therebymimicking the dephosphorylationtransition–like intermediate (29). The rearrange-ment of the A domain accompanies the swing-outmotion of the TM1-TM2 segment, which is di-rectly connected to the A domain, consequentlycreating a large cleft between the M1-M2 andM4-M5 segments, in which the clear density of aglycerophospholipid is observed (Fig. 3, B to D).Therefore, this second AlF4

− class structure repre-sents the substrate-bound E2Pi state (E2Pi-PL).The rearrangement of the A domain is likely tobe coupled to the binding of the substrate lipid,as it occupies the cleft and pushes out theM1-M2segment (figs. S10 and S11), which explains thesubstrate-dependent dephosphorylation of P4-ATPases (Fig. 1C) (8).

Phospholipid recognition

Given that the substrate lipids (such as PS or PE)were not added during purification, it is likely

that endogenous phospholipid contained in theGDN micelles is specifically bound to ATP8A1in the AlF4

−-bound state (Fig. 3, A to D, and fig.S12A). ATP8A1 shows PG-, PE-, and PS-dependentATPase activity, with the highest preference for PS(13, 17), and the size and shape of the head groupdensity are in good agreement with those of theserine moiety (fig. S1H). Therefore, we modeledPS into the density. PS is recognized within theopen cleft, in which the phosphate group is co-ordinated by the backbone amide groups of Ile357

and Ser358 in the conserved PISL motif at the un-wound kink of M4 and further stabilized by theGln88, Asn353, andAsn882 side chains (Fig. 3, C andE), whereas the attached acyl chains are exposedto the bulk lipid environment and partly accom-modated in the hydrophobic pocket formed bythe conserved residues in TM2 and TM4, such asVal103, Pro104, Phe107 (M2), Val361, Val365 (M4),Val883, and Leu891 (M6) (Fig. 3C). In the currentcryo-EMmap, the acyl chains aremost visible nearthe attached glycerol moiety, and PS moleculeswith shorter acyl chains showed weaker ATPaseactivity (fig. S1G), indicating that the acyl chains,as well as the hydrophilic head group, are specif-ically recognized in the substrate binding pocketof ATP8A1.The head group of PS is situatedwithin a small

cavity on the extracellular half of the cleft andis surrounded by hydrophilic residues, such asGln88, Asn352, Asn353, and Asn882 (Fig. 3C), withwhich the serine moiety forms hydrogen bondinginteractions. Consistently, mutational studies have

shown the importance of the uncharged polarresidues Gln88, Gln89, Asn352, and Asn353 for PSselectivity (22–24). Such interactions explain thehead group preferences of ATP8A1, which has weakselectivities for PE and PG, with head groups thatcan form similar hydrogen bonding interactions,and no selectivity for PC, with headmethyl groupsthat cannot form such hydrogen bonding inter-actions. The PC selective P4-ATPases have nonpolarresidues, such as Ala and Gly, at the correspond-ing positions (fig. S2), also supporting the pro-posal that the residues constituting this exoplasmiccavity primarily define the head group selectivity.

Lipid translocation pathway of ATP8A1

In the P2-ATPases, conformational changes inthe cytoplasmic ATPase domains are coupled torearrangement of the core TM helices that con-stitute the cation binding sites (Fig. 4A) (30, 31).Especially in SERCA,Glu309 of the conserved PEGLmotif, located in theunwoundM4kink, constitutespart of the ion binding sites, enabling coupling be-tween ion binding and release and rearrangementof theATPasedomain. InATP8A1, theM4 segmentis similarly kinked at the PISL motif, but the ionbinding sites are lost by the substitution with hy-drophobic residues Ile357, Leu854, and Val977 (Fig.4B). Although the PS binding site partially over-laps the Ca2+ binding site in SERCA (site II), thearrangement of the surrounding residues remainsalmost unchanged throughout the transport cycle.It has been suggested that the P4-ATPases use adifferent translocation pathway for a large lipid


Fig. 3. Phospholipid recognition. (A) Structural comparison of the E2P and E2Pi-PL states, showing the large rearrangement of M1-M2 and the N and Adomains upon phospholipid binding. (B and C) Phospholipid binding site, viewed (B) parallel to the membrane plane and (C) as a close-up of the headgroup. Residues within 4 Å of the bound phospholipid are shown as sticks. Hydrogen bond interactions are shown as black dashed lines. (D) Cryo-EMdensity showing the bound endogenous phospholipid (green mesh, 2.5s). (E) Residues constituting the hydrophobic gate are shown. The putativetranslocation pathway is indicated by an orange arrow.



Dow

nloaded from


substrate (9), and according to previous muta-tion studies on the yeast and bovine P4-ATPases,the residues associated with the head group selec-tivity are mapped along the hydrophilic cleft be-tween theM1-M2andM3-M4 segments (fig. S12B).In the PS-bound structure, the head group entersfrom the exoplasmic leaflet and stays occludedin the middle of the pathway by the side chain ofIle357 in the PISL motif. The mutation of Ile357 tobulky residues, such as Met and Phe, drasticallyreduced the lipid transport activity, whereas themutations to smaller residues only moderatelyaffected the transport activity (22), suggesting thatIle357 constitutes a central hydrophobic gate forthe lipid translocation, together with other resi-dues on the M1-M2 segment, such as Phe81 andIle108 (Fig. 3E). Because the exoplasmic end isalso closed by the M1-M2 loop (fig. S12B), thecurrent structure probably represents a partiallyoccluded state. Although PS binding induces aslight reorientation of the Ile357 side chain towardthe M1-M2 segments (Fig. 4B), the translocationof the hydrophilic head group requires furtherrearrangement of the central gate residues, whichis probably coupled with the phosphate releasefrom the P domain. By analogy to the E2P-to-E2transition inSERCA, thephosphate release “unlocks”the A domain and allows a further outward shiftof the M1-M2 segment, thus inducing the open-ing of the central hydrophobic gate (32). Previousstructures of P2-ATPases revealed several lipidbinding sites; for example, the E2 structure ofSERCA stabilized by thapsigargin and an inhib-

itor, 2,5-di-tert-butyl-1,4-dihydroxybenzene (BHQ)(33), showed PE binding between theM2 andM4segments at the intracellular leaflet, correspond-ing to the putative exit of the lipid translocationpathway in ATP8A1 (fig. S13A). Furthermore,phospholipids are anchored to the positivelycharged residues at the protein-lipid interfaceand interplay with the protein conformationalchanges during the transport cycle in SERCA(34). The ATP8A1-CDC50A complex has clustersof positively charged residues at both the en-trance and exit of the translocation pathway,which may play important roles in lipid trans-location (fig. S12C).

C-terminal autoregulatory domain

In the BeF3−-stabilized E2P state, we observed an

extra density extending through the cytoplasmiccatalytic domains (Fig. 5, A and B), which we as-signed as the C-terminal autoregulatory domain(residues 1117 to 1140) (35, 36), consisting of theconserved GYAFS motif (residues 1119 to 1123)and a short helical domain (residues 1131 to 1137),although the ~50–amino acid linker connectedto the M10 helix was disordered. The regulatorydomain interacts with the N domain, and theGYAFS motif is specifically recognized by a shortloop region of the N domain (residues 533 to539) (Fig. 5B). Notably, Phe1122 occupies the ATPbinding site and stacks with Phe534. The densitiesof the C-terminal residues are only visible in theBeF3

−-stabilized E2P conformation and are com-pletely disordered in the other conformations,

including the ligand-unbound E1 state. This sug-gests that the regulatory domain specifically sta-bilizes ATP8A1 in the E2P conformation, in whichthe N domain is somewhat farther apart (Fig. 5Cand fig. S10).The C-terminal regulatory domain has differ-

ent effects between the yeast and mammalianP4-ATPases. In the yeast Drs2p flippase, it exertsan autoinhibitory effect on ATPase activity (37).However, in the mammalian ATP8A2 flippase, itmediates a rather complicated regulation mode.The partial truncation of the GYAFS motif andthe short helical domain results in decreasedATPase activity, whereas the complete loss of theC-terminal residues, including the disordered loopregion, restores the ATPase activity to the samelevel as the wild-type enzyme (38), indicating thatthe GYAFS motif and the short helical domainobserved in the current cryo-EM map positivelymodulate the enzymatic reaction. We hypoth-esize that the regulatory domain keeps the Ndomain apart from the A domain in the E2P stateand thus facilitates the rotational rearrangementof the A domain that is required for PS binding.The conformation of the N domain in the E1 andE1-ATP states would sterically prevent this rota-tional motion (Fig. 5D).

Mechanism of the P4-type ATPase

The current cryo-EM structures revealed sixdifferent intermediates of ATP8A1, namely, E1,E1-ATP, E1P-ADP, E1P, E2P, and E2Pi-PL, dem-onstrating the transport cycle of the lipid flippase


Fig. 4. Comparison of the phospholipidbinding sites. (A) Ca2+ binding site of SERCAin the Ca2+ binding state (right: PDB ID 1T5S) andthe H+ binding state (left: PDB ID 3B9R),viewed from the cytoplasmic side. Residuesinvolved in Ca2+ and H+ transport are shownas ball-stick representations. Hydrogen bondsare shown as black dashed lines and the boundCa2+ are pale blue spheres. (B) Phospholipidbinding site of ATP8A1 in the unbound state(right: E1-ATP) and the phospholipid-bound state(left: E2Pi-PL) from the same viewpoint as in(A). Residues involved in phospholipid translocationand other residues corresponding to thosecoordinating H+ and Ca2+ in SERCA are shown asball-and-stick representations. Hydrogenbonds are shown as black dashed lines. Cryo-EMdensity showing the side chain of Ile357

in unwound M4 kink (green mesh, 3.0s).



Dow

nloaded from


reaction (Fig. 6). Although ATP8A1 shares a sim-ilar ATP hydrolysis–dependent catalytic reactionwith the ion-transporting P2-type ATPases, such asSERCA (30, 31), Na+/K+-ATPase (39), and H+/K+-ATPase (40), there are notable differences in theirtransport mechanisms and substrate transportingpathways (fig. S14). These canonical ion-transportingP-type ATPases undergo extensive rearrangementsof the TM region, especially in the M1 to M6 seg-ments that constitute the ion translocating path-

way (fig. S14A). In contrast, ATP8A1maintains thestructural rigidity of the TM region throughoutthe transport cycle, as the core TM segments (M3to M10) could be superimposed well in all of theintermediates [RMSD (Å) = 0.30 to 0.69]. Conse-quently, ATP8A1 has a distinct pathway for thelipid head group, between the M1-M2 and M3-M4 segments, and the lipid translocation is es-sentially accomplished by the mobile segmentsof M1-M2 (fig. S14B).

The critical rearrangement in SERCA occursduring the E1P-to-E2P transition, in which theA-domain rearrangement toward the phospho-rylation site induces the opening of the “luminalgate” composed of the M1 to M4 segments andalters the affinity for the substrate ions (fig. S15A)(31). Although theAdomain ofATP8A1 undergoesa similar rearrangement during this transition,the conformational change is limited to the M1and M2 segments in the region proximal to the A


Fig. 5. ATP8A1 autoregulation by the C-terminal domain. (A) Inthe BeF3

−-stabilized E2P state, an extra density is observed around thecytoplasmic catalytic domains, corresponding to the C-terminal auto-regulatory domain. The density is shown as a green mesh, contouredat 3s. (B) Close-up view of the interaction between a short loopregion (residues 533 to 539) in the N domain and the regulatorydomain. An atomic model of the GYAFS motif and a short helicalregion in the regulatory domain are modeled into the density.

(C and D) Arrangements of the N and A domains and the regulatorydomain, shown for the E2P (C) and E1-ATP (D) states, viewed fromthe cytoplasmic side. The N and A domains of the E2Pi-PL state aresuperimposed in a transparency representation. The regulatorydomain keeps the N domain apart from the A domain and thus facilitatesthe rotational movement of the A domain around the phosphorylationsite in the E2P state (C), whereas the similar rearrangement ishindered by the N domain in the E1-ATP state (D).

Fig. 6. Proposed mechanism of phospholipid trans-location. Schematic model of the phospholipidtranslocation cycle by ATP8A1-CDC50a, accordingto the Post-Albers mechanism. The model is depicted withthe same colors as in Fig. 1A. ATP binding induces theproximal arrangement of the N and P domains, by bridgingthese domains and slightly forcing out the A domain.After the phosphoryl transfer reaction, ADP is releasedfrom the N domain, and the A domain approaches theN domain and interacts with it, through the DGETmotif, to form the E2P state. The C-terminal regulatorydomain penetrates between the P and N domains andstabilizes the E2P state. The rearrangement of the Adomain induces flexibility in the M1-M2 segments, thusallowing phospholipid binding at the interface between theM1-M2 and bulk TM segments. Phospholipid bindinginduces further rearrangement of the A domain, therebyfacilitating the dephosphorylation reaction (E2Pi-PL).Ile357 constitutes a hydrophobic gate that occludes themiddle of the translocation pathway. Phospholipidtranslocation to the cytoplasmic leaflet is probably coupledto the phosphate release at the P domain, allowing thefurther outward shift of the M1-M2 segment (E2-PL).The translocated phospholipid laterally diffuses to thecytoplasmic leaflet, and the enzyme adopts the E1 confor-mation, ready to initiate another reaction cycle.



Dow

nloaded from


domain, and the luminal side remains unchanged(fig. S15B). This rigidity is probably achieved bythe tight association with CDC50a, which holdsthe M3 to M10 segments of ATP8A1 on both theluminal and cytoplasmic sides. Most notably, theloop connectingM3-M4 and the cytoplasmic endof M4 is constrained by the interaction withCDC50a, which probably hinders theM3 andM4rearrangement (figs. S3, E and F, and S15B). Thedeletion of the CDC50a N-terminal tail, whichinteracts with the cytoplasmic end of the M4segment, decreased the flippase activities of P4-ATPases (25). Therefore, the rigidity of the TMsegments is important for the transport activityof P4-ATPases. Although the phosphorylation-induced A-domain rearrangement in ATP8A1causes only minor changes on the luminal side,the density of the M1-M2 segment near the Adomain ismore disordered in the E2P conforma-tion (fig. S16), suggesting higher flexibility in thelinker region. This flexibility may facilitate thesubsequent binding of the phospholipid betweentheM1-M2 andM3-M4 segments by allowing theswing-out motion of the M1-M2 segment, as ob-served in the AlF4

−-stabilized dephosphorylationtransition-like state. Overall, the P4-ATPases haveevolved a distinct mechanism for the lipid trans-location, while sharing the similar rearrangementof the cytoplasmic domains with the canonicalion-transporting P-type ATPases.

REFERENCES AND NOTES

1. H. M. Hankins, R. D. Baldridge, P. Xu, T. R. Graham, Traffic 16,35–47 (2015).

2. H.-W. Shin, H. Takatsu, FASEB J. 33, 3087–3096(2019).

3. C. Montigny, J. Lyons, P. Champeil, P. Nissen, G. Lenoir,Biochim. Biophys. Acta 1861, 767–783 (2016).

4. Y. Yang, M. Lee, G. D. Fairn, J. Biol. Chem. 293, 6230–6240(2018).

5. S. Lutsenko, J. H. Kaplan, Trends Biochem. Sci. 21, 467(1996).

6. P. L. Pedersen, E. Carafoli, Trends Biochem. Sci. 12, 146–150(1987).

7. M. Palmgren, J. T. Østerberg, S. J. Nintemann, L. R. Poulsen,R. L. López-Marqués, Biochim. Biophys. Acta Biomembr. 1861,1135–1151 (2019).

8. J. A. Coleman, A. L. Vestergaard, R. S. Molday, B. Vilsen,J. P. Andersen, Proc. Natl. Acad. Sci. U.S.A. 109, 1449–1454(2012).

9. J. P. Andersen et al., Front. Physiol. 7, 275 (2016).10. S. Bryde et al., J. Biol. Chem. 285, 40562–40572

(2010).11. J. A. Coleman, R. S. Molday, J. Biol. Chem. 286, 17205–17216

(2011).12. X. Tang, M. S. Halleck, R. A. Schlegel, P. Williamson, Science

272, 1495–1497 (1996).13. J. K. Paterson et al., Biochemistry 45, 5367–5376

(2006).14. K. Segawa et al., Science 344, 1164–1168 (2014).15. W. Jing et al., Blood Adv. 3, 219–229 (2019).16. U. Kato et al., J. Biol. Chem. 288, 4922–4934 (2013).17. S. Lee et al., EMBO J. 34, 669–688 (2015).18. V. A. van der Mark, R. P. Elferink, C. C. Paulusma, Int. J. Mol. Sci.

14, 7897–7922 (2013).19. K. Levano et al., J. Neurochem. 120, 302–313 (2012).20. O. E. Onat et al., Eur. J. Hum. Genet. 21, 281–285

(2013).21. S. Alsahli, M. T. Alrifai, S. Al Tala, F. A. Al Mutairi,

M. Alfadhel, J. Cent. Nerv. Syst. Dis. 10, 1179573518759682(2018).

22. A. L. Vestergaard et al., Proc. Natl. Acad. Sci. U.S.A. 111,E1334–E1343 (2014).

23. R. D. Baldridge, T. R. Graham, Proc. Natl. Acad. Sci. U.S.A. 109,E290–E298 (2012).

24. R. D. Baldridge, T. R. Graham, Proc. Natl. Acad. Sci. U.S.A. 110,E358–E367 (2013).

25. J. A. Coleman, R. S. Molday, J. Biol. Chem. 286, 17205–17216(2011).

26. K. Segawa, S. Kurata, S. Nagata, J. Biol. Chem. 293, 2172–2182(2018).

27. J. Zivanov et al., eLife 7, e42166 (2018).28. J. A. Coleman, M. C. M. Kwok, R. S. Molday, J. Biol. Chem. 284,

32670–32679 (2009).29. C. Olesen, T. L. M. Sørensen, R. C. Nielsen, J. V. Møller,

P. Nissen, Science 306, 2251–2255 (2004).30. T. L. M. Sørensen, J. V. Møller, P. Nissen, Science 304,

1672–1675 (2004).31. C. Olesen et al., Nature 450, 1036–1042 (2007).32. M. Dyla et al., Nature 551, 346–351 (2017).33. K. Obara et al., Proc. Natl. Acad. Sci. U.S.A. 102, 14489–14496

(2005).34. Y. Norimatsu, K. Hasegawa, N. Shimizu, C. Toyoshima, Nature

545, 193–198 (2017).35. P. Natarajan et al., Nat. Cell Biol. 11, 1421–1426 (2009).36. X. Zhou, T. T. Sebastian, T. R. Graham, J. Biol. Chem. 288,

31807–31815 (2013).

37. A. Jacquot et al., J. Biol. Chem. 287, 13249–13261(2012).

38. M. Chalat, K. Moleschi, R. S. Molday, Mol. Biol. Cell 28,452–462 (2017).

39. J. P. Morth et al., Nature 450, 1043–1049 (2007).40. K. Abe, K. Irie, H. Nakanishi, H. Suzuki, Y. Fujiyoshi, Nature

556, 214–218 (2018).

ACKNOWLEDGMENTS

We thank H. Hirano for assistance in generating the movie,H. Nishimasu for fruitful discussions, T. Nakane for assistance withthe single-particle analysis, and the structural biophysics team atMitsubishi Tanabe Pharma Corporation, especially H. Kishida, fortechnical advice about model building. We also thank the staffscientists at the University of Tokyo’s cryo-EM facility, especiallyK. Kobayashi, T. Kusakizako, H. Yanagisawa, A. Tsutsumi,M. Kikkawa, and R. Danev. Funding: This work was supportedby a MEXT Grant-in-Aid for Specially Promoted Research (grant16H06294) to O.N. Author contributions: M.H. prepared thecryo-EM samples and performed the functional analyses. M.H.and T.N. collected and processed the cryo-EM data and built thestructures. K.Y. assisted data processing and structure refinement.M.H., T.N., and O.N. wrote the manuscript. T.N. and O.N.supervised the research. Competing interests: M.H. is agraduate student at Mitsubishi Tanabe Pharma Corporation andis supported by the company with nonresearch funds. Thecompany has no financial or other interest in this research.Data and materials availability: Cryo-EM density maps havebeen deposited in the Electron Microscopy Data Bank under theaccession codes EMD-9931 (E1 class1), EMD-9932 (E1 class2),EMD-9933 (E1 class3), EMD-9935 (E1-ATP class1), EMD-9934(E1-ATP class2), EMD-9936 (E1-ATP class3), EMD-9937(E1P-ADP), EMD-9938 (E2P class1), EMD-9939 (E2P class2),EMD-9940 (E2P class3), EMD-9941 (E2Pi-PL), and EMD-9942(E1P). Atomic coordinates have been deposited in the ProteinData Bank under IDs 6K7G (E1 class1), 6K7H (E1 class2), 6K7J(E1-ATP class1), 6K7I (E1-ATP class2), 6K7K (E1P-ADP), 6K7L(E2P-class2), 6K7M (E2Pi-PL), and 6K7N (E1P). The raw imageshave been deposited in the Electron Microscopy Public ImageArchive, under accession code EMPIAR-10303.

SUPPLEMENTARY MATERIALS

science.sciencemag.org/content/365/6458/1149/suppl/DC1Materials and MethodsFigs. S1 to S16Table S1References (41–58)Movie S1

10 June 2019; accepted 6 August 2019Published online 15 August 201910.1126/science.aay3353




Dow

nloaded from

https://science.sciencemag.org/content/365/6458/1149/suppl/DC1http://science.sciencemag.org/

Cryo-EM structures capture the transport cycle of the P4-ATPase flippaseMasahiro Hiraizumi, Keitaro Yamashita, Tomohiro Nishizawa and Osamu Nureki

originally published online August 15, 2019DOI: 10.1126/science.aay3353 (6458), 1149-1155.365Science

, this issue p. 1149Sciencelipids bind differently, powering translocation.requires for function, CDC50. ATP binding and autophosphorylation of ATP8A1 drive a cycle of conformations in which

electron microscopy structure of six intermediates of the human flippase ATP8A1 bound to the partner protein it−the cryo report et al.ATPases that are important in processes such as membrane trafficking, signaling, and apoptosis. Hiraizumi

against a concentration gradient from the outer to inner or inner to outer leaflets, respectively. Flippases are P4-typeknown as flippases and floppases use the energy from adenosine triphosphate (ATP) hydrolysis to translocate lipids

The membranes of eukaryotic cells have different lipid compositions in their inner and outer leaflets. EnzymesFlipping a lipid

ARTICLE TOOLS http://science.sciencemag.org/content/365/6458/1149

MATERIALSSUPPLEMENTARY http://science.sciencemag.org/content/suppl/2019/08/14/science.aay3353.DC1

REFERENCES

http://science.sciencemag.org/content/365/6458/1149#BIBLThis article cites 58 articles, 23 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Science. No claim to original U.S. Government WorksCopyright © 2019 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of

on June 5, 2021


Dow

nloaded from
http://science.sciencemag.org/content/365/6458/1149http://science.sciencemag.org/content/suppl/2019/08/14/science.aay3353.DC1http://science.sciencemag.org/content/365/6458/1149#BIBLhttp://www.sciencemag.org/help/reprints-and-permissionshttp://www.sciencemag.org/about/terms-servicehttp://science.sciencemag.org/

Documents

STRUCTURAL BIOLOGY Cryo-EM structures capture the ...STRUCTURAL BIOLOGY Cryo-EM structures capture the transportcycle of the P4-ATPase flippase Masahiro Hiraizumi1,2, Keitaro Yamashita1,3,