Capturing the interplay of membrane lipids and structural transitions in human ABCA7

Phospholipid extrusion by ABC subfamily A (ABCA) exporters is central to cellular physiology, although the specifics of the underlying substrate interactions and transport mechanisms remain poorly resolved at the molecular level. Here we report cryo-EM structures of lipid-embedded human ABCA7 in an open state and a nucleotide-bound, closed state at resolutions between 3.6-4.0 Å. The former reveals an ordered patch of bilayer lipids traversing the transmembrane domain (TMD), while the latter reveals a lipid-free, closed TMD with a small extracellular opening. These structures offer a structural framework for both substrate entry and exit from the ABCA7 TMD and highlight conserved rigid-body motions that underlie the associated conformational transitions. Combined with functional analysis and molecular dynamics (MD) simulations, our data also shed light on lipid partitioning into the ABCA7 TMD and localized membrane perturbations that underlie ABCA7 function and have broader implications for other ABCA family transporters.


transitions. Combined with functional analysis and molecular dynamics (MD) simulations, our
Introduction 28 ABCA family exporters mediate efflux of phospholipids and sterols from cells, 29 contributing to membrane homeostasis, bilayer structure and asymmetry, and the formation of 30 serum lipoproteins, among other key physiological processes 1 . Their dysfunction therefore 31 underlies several human diseases 2-4 . The molecular details governing the ABCA exporter substrate 32 transport cycle are not fully resolved. To fill this knowledge gap, here we present the structural 33 and functional analysis of human ABCA7, whose dysfunction has been strongly linked to 34 Alzheimer's Disease (AD) 5-12 , in a lipid environment. Deficient ABCA7 activity leads to 35 alterations in both brain lipid profiles and fatty acid and phospholipid biosynthetic pathways 13 , 36 impaired memory, and reduced immune responses 14,15 . Both in vitro lipid flipping 16 and lipid 37 extrusion to apolipoproteins by cells over-expressing ABCA7 17 have been demonstrated, although 38 the correlation between the two processes, if any, remains unclear. To date, no direct structural 39 information exists for ABCA7. Understanding the molecular details of the ABCA7 transport cycle 40 and how its dysfunction alters inflammatory and immune responses, lipid homeostasis, and 41 phagocytosis, which all contribute to AD progression 18-22 , may therefore pave the way for novel 42 therapeutics for AD. 43 ABCA7 encodes a 2146 amino acid membrane transporter found in many tissues and 44 blood, hippocampal neurons, macrophages, and microglia 23,24 . Like the phospholipid and sterol 45 exporter ABCA1 and retinal importer ABCA4, with which it shares 54% and 59% sequence 46 similarity, respectively, ABCA7 comprises two halves assembled as a full transporter. Each half 47 consists of a TMD, with the first two transmembrane helices (TMs) of each separated by a large 48 extracellular domain (ECD), and a nucleotide binding domain (NBD) attached to a cytoplasmic 49 regulatory domain (RD). To visualize its ATP-dependent conformational cycle in a lipid 50 environment, we resolved the structures of human ABCA7 in multiple conformations in lipid and 51 detergent environments using cryo-EM and probed its lipid interactions using ATPase assays and 52 MD simulations. Our data allow us to directly visualize lipid partitioning into the TMDs and the 53 associated conformational changes in ABCA7 that provide insights into its mechanisms of 54 substrate entry and export that likely hold true for other members of the ABCA family. 55

Results 56
Dependence of ABCA7 ATPase activity on its lipid environment 57 Human ABCA7 expressed in a tetracycline inducible stable HEK293 cell line was 58 reconstituted in liposomes and nanodiscs comprising 80% brain polar lipids (BPL) and 20% 59 cholesterol (Chol) and its ATPase activity was measured ( Figure 1A, Figure S1A). Although ATP 60 hydrolysis was slowest in nanodiscs, it followed Michaelis-Menten kinetics similar to ABCA7 in 61 detergent or liposomes comprising the same lipid/cholesterol composition and Michaelis constant 62 (KM) values for all three were in the 0.5-0.8 mM range. ATPase rates for a hydrolysis-deficient 63 mutant carrying E965Q and E1951Q substitutions (ABCA7EQ) were drastically reduced compared 64 to wildtype in both nanodiscs and detergents, demonstrating that the observed activity was specific 65  To gain additional molecular insight into the TMD-mediated lipid partitioning, membrane  178   perturbation, and lipid extrusion from ABCA7, we performed multi-microsecond MD simulations  179   of the open conformation of ABCA7PE after embedding into two distinct lipid bilayers, one  180 containing PE/Chol (4:1) and the other PC/Chol (4:1) ( Figure 5, Figure S8). Each simulation was 181 performed for 2 µs using a system including four copies of the protein ( Figure 5A apolipoproteins in a physiologically relevant manner and, more broadly, the structural basis for 225 apolipoprotein interactions with ABCA7 or ABCA1 is unknown. Second, while our structures 226 provide a potential basis for lipid entry into and extrusion from the ABCA7 TMD, the mechanism 227 whereby lipids get flipped, remains unresolved. Third, while our data point to a potential 228 preference for PE over PC for TMD partitioning, the exact mechanism whereby ABCA1 or 229 ABCA7 achieve lipid specificity are unknown. The observed enhancement in conformational 230 homogeneity, quality of lipid density, and ATPase activity of our ABCA7PE sample may be, in 231 part, due to the smaller PE headgroup, which has been shown to aid in folding and stabilization of 232 membrane proteins 43 . Finally, it is unknown how the opposite direction of substrate transport for 233 ABCA4 is achieved considering similarities in ATP bound closed and open state structures of 234 ABCA4 and ABCA7. Overall, our data will help devise better in vitro and in silico models to 235 answer these questions, which will further aid in dissecting the unique roles these proteins play in 236 cellular physiology. 237

Protein Purification 240
We utilized the Flp-In TREX system (Thermo Fisher Scientific) for tetracycline inducible 241 expression of human ABCA7. In short, a codon optimized synthetic gene construct 242 (PMSF) and 20 µg ml -1 soybean trypsin inhibitor (both Sigma), and mechanically cracked using 258 a dounce homogenizer before addition of a 0.5%/0.1% w:v mixture of dodecyl maltoside (DDM) 259 and cholesteryl hemisuccinate (CHS) (both Anatrace). Protein extraction was allowed to proceed 260 for 90 minutes at 4°C with gentle agitation, after which, the suspension was centrifuged at 48,000 261 r.c.f for 30 minutes and the supernatant applied to rho-1D4 antibody (University of British 262 Columbia) coupled Sepharose resin (Cytiva). Binding was allowed to proceed for 3 hours before 263 the unbound fraction was discarded and beads rinsed with 4 x 10 bed volumes (BVs) of wash 264 buffer (25 mM Hepes pH 7.5, 150 mM NaCl, 20% glycerol, 0.02%/0.004% w:v DDM/CHS). 265 Protein was eluted by incubation with 3 BVs elution buffer (wash buffer supplemented with either 266 3C protease (1:10 w:w 3C:ABCA7) or 0.5 mg ml -1 1D4 peptide (GenScript)) for 2-18 hours. 267 The EQ variant of ABCA7 contained two site mutations, E965Q and E1951Q. The E965Q 268 site was created within the ABCA7 construct using site directed mutagenesis by PCR with the 269 primers: A7eq1for 5'-GGTCATCCTGGATCAACCTACAGCAGGCGTGG-3' and A7eq1rev 270 5'-GCCTGCTGTAGGTTGATCCAGGATGACCACC-3'. E1951Q was generated using a 271 synthesized dsDNA block and the enzymes NheI and BsiWI. ABCA7EQ with C-terminal eYFP-272 Rho1D4 tag was then transferred to a pCAG vector using KpnI and NotI restriction sites. The England Biolabs). One-liter cultures of Terrific Broth (TB) supplemented with 50 ug ml -1 288 kanamycin were grown from 10 ml overnight cultures from single colonies grown in LB. Cells 289 were grown to an OD600 of 0.8 in a shaking incubator at 37 °C and induced with 1 mM isopropyl 290 β-d-1-thiogalactopyranoside (IPTG). Protein expression was allowed to proceed at 20°C for 12 291 hours. Cells were centrifuged at 12,000 r.c.f, and pellets were flash frozen in liquid nitrogen and 292 stored at -80°C until required. Frozen pellets were resuspended in 8 ml/ gram cell pellet 293 resuspension buffer comprising 25 mM Hepes pH 7.5, 150 mM NaCl and 1 mM 294 phenylmethylsulfonyl fluoride (PMSF) and sonicated. The suspension was spun down at 16,000 295 r.c.f at 4 °C for 30 min and the supernatant was applied 5 ml Ni-NTA resin (Qiagen)/ L culture 296 medium. After discarding the flowthrough, the resin was washed with 25 mM Hepes pH 7.5, 297 150 mM NaCl, 1 mM PMSF and 20 mM imidazole until a pre-established baseline A280 reading 298 was achieved. ApoA1 was eluted in 4 BVs of 25 mM Hepes pH 7.5, 150 mM NaCl, 1 mM PMSF 299 and 200 mM imidazole, concentrated using a 10 kDa molecular weight cutoff (MWCO) Amicon 300 filter (Millipore-Sigma) and desalted using a PD10 column (Cytiva) into 25 mM Hepes pH 7.5, 301 150 mM NaCl. The concentration of apoA1 was adjusted to 1 mg ml -1 for flash freezing in liquid 302 nitrogen and storage at -80°C. 303

ABCA7 and ABCA7EQ nanodisc and proteoliposome preparation 304
For nanodisc reconstitution, peptide eluted or 3C cleaved ABCA7 was mixed with 305 MSP1D1 and a mixture of BPL (brain polar lipid extract from Avanti) and cholesterol (80:20 w:w) 306 with 0.5%/0.1% DDM:CHS using a 1:10:350 (ABCA7:MSPD1:lipid mix) molar ratio in nanodisc 307 buffer (25 mM Hepes pH 7.5, 150 mM NaCl) that contained up to 4% glycerol for 30 minutes at 308 room temperature (RT). Nanodisc reconstitution was induced by removing detergent with 0.8 mg 309 ml -1 pre-washed Biobeads SM-2 (Bio-Rad) for 2 hours with gentle agitation at RT. For different 310 phospholipid compositions, BPL was replaced with brain PE, PS, or PS (all from Avanti Polar 311 Lipids). For structural studies, nanodisc-reconstituted ABCA7 bearing the eYFP-Rho1D4 tag was 312 bound to rho-1D4 resin for an additional 2 hours, washed with 4 BV of nanodisc buffer, and eluted 313 with 3C protease for 2 hours at 4°C. The eluted ABCA7 nanodiscs were concentrated using a 314 100,000 MWCO kDa Amicon filter and further purified by size exclusion chromatography using 315 a G4000swxl column (TOSOH biosciences) equilibrated with nanodisc buffer at 4°C. Figure S1A  316 shows a SEC chromatogram for pure ABCA7 placed into BPL/Ch nanodiscs, while a SEC 317 chromatogram for ABCA7 in PE/Ch nanodiscs is in Figure S2A. Generally, three fractions were 318 pooled from the main resultant peak. 319 ABCA7EQ was reconstituted in nanodiscs for cryo-EM preparation using the same 320 approach of ABCA7; except that an additional 2 mM ATP and 10 mM MgCl2 were present until 321 the end of the purification procedure prior to grid preparation. A SEC chromatogram for ABCA7EQ 322 in BPL/Ch nanodiscs is shown in Figure S5A, where the trace is affected by the additional ATP 323 added during the run. 324 ABCA7 proteoliposomes were generated by mixing detergent purified ABCA7 with 325 liposomes at a protein: liposome ratio of 1:10 w:w. Liposomes were prepared by extruding a 20 326 mg ml -1 , 80:20 w:w BPL/Ch lipid mixture 11 times using a previously described protocol 45 . 327 Briefly, detergent purified ABCA7 and liposomes were added to 0.14% and 0.3% Triton X100 328 (Sigma), respectively, then incubated for 30 minutes at RT. These two samples were mixed and 329 incubated for 60 minutes. Detergent was removed by adding 40 mg fresh Biobeads SM2 (Bio-330 Rad) per ml reaction mixture during five successive incubation steps, 30 minutes at RT, 60 minutes 331 at 4°C, overnight at 4°C, and two periods of 60 minutes at 4°C with gentle agitation. The 332 suspension was centrifuged at 80,000 rpm for 20 minutes in an ultracentrifuge. The supernatant 333 was removed, and the liposomal pellet was washed once with reconstitution buffer containing 150 334 mM NaCl, 25 mM Hepes pH 7.5. The ABCA7-liposome suspension was then centrifuged to 335 remove the supernatant, and the proteoliposomes were resuspended at a final concentration of 336 0.5 -1 mg ml -1 for ATPase assays. 337

ABCA7DIGITONIN preparation 338
For the digitonin solubilized ABCA7 purification, ABCA7 was extracted from the cell in 339 a lysis buffer (Buffer L) using the same approach above, and the supernatant was applied to rho-340 1D4 resin for a 3-hour binding period. Then, the resin was rinsed with the 4 x 10 BVs of wash 341 buffer containing 25 mM Hepes pH 7.5, 150 mM NaCl, 20% glycerol (v/v), and 0.06% digitonin 342 (w/v). Protein was eluted by incubation with 3 BVs elution buffer, which was wash buffer 343 supplemented with 3C protease (1:10 w:w 3C:ABCA7). Interestingly we obtained better particle 344 distribution and ice quality with addition of a 1:2.5 molar excess of apoA1, prepared in house, to 345 3C cleaved ABCA7 prior to grid preparation. The mixture was concentrated by a 100,000 MWCO 346 kDa Amicon filter (Millipore) and further purified by size exclusion chromatography using a 347 G4000swxl column (TOSHOH biosciences) equilibrated with a buffer containing 25 mM Hepes 348 pH 7.5, 150 mM NaCl, 0.035% digitonin (w/v), as shown in Figure S4A. Peak fractions were 349 pooled and concentrated for cryo-EM grid preparation. ATP and ABCA7 in digitonin, two sample droplets were applied to glow discharged grids to obtain 376 more particles per hole. 377

Cryo-electron microscopy data collection and processing 378
Grids were clipped as per manufacturer guidelines and cryo-EM data was collected using 379 a Titan Krios electron microscope operating at 300kV and equipped with a Falcon 3EC direct 380 electron detector (Thermo Fisher Scientific.). Automated data collection was carried out using EPU 2.8.0.1256REL software package (Thermo Fisher Scientific) over multiple sessions in 382 counting mode at a nominal magnification of 96,000x, corresponding to a calibrated pixel size of 383 0.895 Å for nanodisc reconstituted ABCA7BPL. Image stacks comprising 60 frames were collected 384 at a defocus range of -0.6 to -2.6 µm and estimated dose rate of 1 electron/Å 2 /frame and further 385 processed in Relion-3.1 (beta). Motion correction was done using Motioncor2 (Relion 386 implementation) 47 and contrast transfer function (CTF) correction was performed using Gctf 1.06 387 48 . A summary of the overall data processing scheme and the quality was presented in Figure S1C-388 E. In brief, 11802 micrographs were used for template free picking of 6725108 particles, followed 389 by particle extraction at a 3x binned pixel size of 2.685 Å/pix. The dataset was processed in two 390 batches. After 2-3 rounds of 2D classification 1259324 particles from Set 1 and 1088487 particles 391 from Set 2 were selected for independent 3D classification steps (number of classes (K)=8 for 392 both). The structure of human ABCA1 (EMDB6724) was used as a 3D reference for an initial 3D 393 classification of a subset of the total data to yield an initial sub-nanometer resolution map of 394 ABCA7 that was used as a 3D reference for the full datasets. After 1 round of 3D classification, 395 both sets of data yielded a similar ensemble of classes. A total of 113291 particles from similar 396 looking classes (black boxes) were subjected to an additional round of classification (K=3), ~80% 397 of which fell into a high-resolution class that yielded a 3.6 Å map after refinement and particle 398 polishing steps. Similarly, 124114 particles from a second set of two similar classes (red boxes in 399 Figure S1D) were selected for subsequent refinement, particle polishing, and post processing to of 3×10 -4 bar and 5 ps, respectively. The systems were energy minimized for 1,000 steps, followed 492 by short equilibration runs of 18 ns, while restraints were applied to lipid bilayer headgroups and 493 protein backbones. During this time the restraints on bilayer headgroups were reduced gradually 494 from k = 200 kJ.mol -1 .nm -2 to zero, whereas the protein backbones' restraints (k = 1,000 kJ.mol -495 to the protein backbones, resulting in an aggregate sampling of 8 μs (4 copies × 2 μs). All the 497 systems were simulated following the same MD protocol. 498 All the molecular images were generated using VMD 56 . The membrane deformation 499 induced by ABCA7 was quantified by calculating the z distance of the lipid phosphate moieties 500 (PO4 bead type in MARTINI) with respect to the bilayer midplane, over the last 1 μs of each 501 trajectory. The generated histogram (binned in 2×2 Å 2 bins) in each leaflet illustrates the spatial 502 distribution of the height of the lipid head groups within each leaflet. We quantified the differential 503 movement of POPE and POPC within the protein lumen by calculating the number of 504 phospholipids located within the TMDs. If the PO4 bead of a phospholipid was within 22.5 Å and 505 12.5 Å in x and y, respectively, with respect to a protein's center in the x-y plane (membrane plane), 506 then the phospholipid was considered to be within the TMD lumen ( Figure S9C). 507 Acknowledgments 508 We would like to thank Dr. Kaspar Locher at ETH, Zurich, Switzerland, for providing the 509 synthetic gene construct of ABCA7. We would also like to thank the cryo-EM and shared 510 instruments core facilities at the Hormel Institute for help with experimental setup, and Dr.