Structural Insights into the Broad Substrate Recognition Mechanism of Human ABCD3

Structural Insights into the Broad Substrate Recognition Mechanism of Human ABCD3 | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Structural Insights into the Broad Substrate Recognition Mechanism of Human ABCD3 Osamu Nureki, Shinichiro Aiba, Hiroyuki Okamoto, Atsuhiro Tomita, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7304961/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Human peroxisomal ABC subfamily D transporters, ABCD1, ABCD2, and ABCD3, transport acyl-CoA molecules and bile acid precursors, and thereby participate in fatty acid metabolism and cholesterol homeostasis. ABCD3 can recognize a wide range of substrates, including polyunsaturated fatty acyl-CoAs, dicarboxylic acyl-CoAs, and ATP molecules. However, the structural mechanism of this broad substrate recognition by ABCD3 remains largely unknown. Here we report five cryo-electron microscopy structures of ABCD3: the apo state, the eicosapentaenoyl coenzyme A (C20:5-CoA)-bound state, the cholesteryl hemisuccinate (CHS)-bound state, the ATP-bound inward-facing state, and the ATP-bound outward-facing state. In the C20:5-CoA-bound state, the CoA moiety is recognized by ABCD3 in a similar manner to ABCD1, but the long, unsaturated bent acyl chain moiety is accommodated in a large cavity between protomers, distinct from the saturated straight acyl chain moiety-binding site observed in ABCD1. Furthermore, we found that the structurally and chemically different substrates, ATP and CHS, bind to distinct hydrophilic and hydrophobic sites, respectively, in the transmembrane domain of ABCD3. Combined with molecular dynamics simulations, these structures, encompassing both inward- and outward-facing conformations, reveal the molecular mechanism by which ABCD3 recognizes and transports a broad range of substrates. Main text Biological sciences/Structural biology/Electron microscopy/Cryoelectron microscopy Biological sciences/Cell biology/Organelles/Peroxisomes Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Fatty acids are used in diverse biological processes, such as lipid membrane formation, energy storage, and post-translational modifications 1 , 2 . Intracellular organelles, including mitochondria, the endoplasmic reticulum, and peroxisomes, play important roles in fatty acid metabolism 3 . In mitochondria and peroxisomes, fatty acids are mainly processed by β-oxidation 4 . Mitochondrial β-oxidation is involved in energy production. In contrast, peroxisomes are not directly involved in energy production, and instead degrade a wide range of fatty acids, including branched fatty acids and very long-chain fatty acids 5 , since they cannot be broken down in mitochondria. As fatty acids have low membrane permeability, ABC subfamily D transporters (ABCD1, ABCD2, and ABCD3) are responsible for their uptake into peroxisomes 6 . Fatty acid transport is accomplished in the form of an acyl moiety bound to CoA, and ABCD1–3 transport different types of acyl-CoAs 6 . ABCD1–3 transport long-chain fatty acyl-CoAs such as C16:0-CoA, while very long-chain fatty acyl-CoAs such as C22:0-, C24:0-, and C26:0-CoA are only transported by ABCD1 and ABCD2 6 . In contrast, unsaturated fatty acyl-CoAs are transported by ABCD2 and ABCD3, and not by ABCD1 6 . Dicarboxylic acyl-CoAs, branched fatty acyl-CoAs, bile acid precursors, and acylcarnitines are exclusively transported by ABCD3 6,7 . Thus, ABCD1 and ABCD2 are responsible for very long, straight-chain fatty acid transport, whereas ABCD3 transports a wide range of acyl-CoA molecules that are shorter but bulkier and more hydrophilic. Consequently, the phenotypes that emerge from each dysfunction are different: ABCD1 is associated with adrenoleukodystrophy 8 , while ABCD3 dysfunction is associated with hepatosplenomegaly 7 , 9 . Despite the functional differences of ABCD1–3, structural information is only available for ABCD1 10–14 and the apo state of ABCD3 15 . Previous studies reported that the acyl chains are recognized by hydrophobic pockets in ABCD1 10 , which shares 36.7% sequence identity with ABCD3 (Supplementary Fig. 1). However, the mechanism by which ABCD3 recognizes and transports bulkier, unsaturated long-chain acyl-CoA molecules, which are not accommodated by ABCD1, remains elusive. In addition to acyl-CoA molecules, ABCD3 also transports bile acid precursors 7 and ATP molecules 16 into peroxisomes. ATP molecules in peroxisomes are essential for the β-oxidation of fatty acids 4 , and peroxisomes also biosynthesize bile acids, playing a crucial role in cholesterol homeostasis 3 , 9 . However, the structural basis for the recognition and transport of such structurally distinct substrates, such as acyl-CoAs, bile acid precursors, and ATP, by ABCD transporters remains poorly understood. In this study, we present cryo-electron microscopy structures of human ABCD3 in distinct transporting states: the apo, unsaturated long-chain acyl-CoA (C20:5 CoA)-bound, CHS-bound, and substrate ATP-bound inward-facing, and ATP-bound outward-facing states. The C20:5 CoA-bound structure reveals details about the acyl-CoA substrate preference of the ABCD3 transporter. Moreover, the bile acid precursor-like CHS-bound and substrate ATP-bound structures exhibit different substrate-binding modes. These ATP-bound, inward-facing and outward-facing structures represent the conformational changes in ABCD3 during the substrate transport cycle. Collectively, our study provides structural insights into the substrate preferences and transport mechanisms of ABCD3. Results Unsaturated long-chain acyl-CoA recognition mechanism To elucidate the structural basis for the unsaturated long-chain acyl-CoA recognition by ABCD3, we determined the cryo-EM structure of C20:5-CoA-bound ABCD3 with C2 symmetry at an overall resolution of 3.3 Å (Fig. 1 a,b, and Supplementary Figs. 2 and 3a). The structure reveals the typical ABC half-transporter architecture, with each subunit comprising a transmembrane domain (TMD) and a nucleotide-binding domain (NBD) (Fig. 1 a,b). The TMD consists of six TM helices, and the inter-subunit internal cavity is largely exposed to the intracellular and lateral sides (the inner leaflet of the membrane) with 41.9 Å between the two corresponding Y282 residues in the TMDs, similar to the C22:0-CoA-bound ABCD1 structure 10 (PDB ID: 7VZB) (Fig. 1 a,b, and Supplementary Figs. 4 and 5). The structure revealed that the C20:5-CoA molecule is located within the cavity between TM bundles, with the CoA moiety of C20:5-CoA forming polar interactions with hydrophilic residues (Fig. 1 c), in a similar manner to the C22:0-CoA-bound ABCD1 structure 10 (Fig. 1 d). Specifically, the adenine ring forms hydrogen bonds with S205 and K209, and a stacking interaction with R373 (Fig. 1 c). The 5′-diphosphate and 3′-phosphate groups of the CoA moiety interact the with conserved hydrophilic residues R96, N140, K322’ (apostrophe indicates residue from the opposite protomer) and Y323’ and with K144, respectively (this site is referred to as substrate-binding site 1) (Fig. 1 c). Both K322’ and Y323’ are located on the TM5’ π-helix, which is observed only in ABCD1 and ABCD3, but not in ABCD4 17 , a member of the same subfamily. This unique structural feature allows K322’ and Y323’ to come into the proximity of the acyl-CoA substrate (Fig. 1 c,d). The acyl moiety of C20:5-CoA is accommodated in the large central cavity between TM bundles (this region is referred to as substrate-binding site 2), which comprises hydrophobic residues such as Y330’, V333’, and V365’ (Fig. 1 e). The unsaturated, long acyl chain adopts a bent conformation, with the terminal region poorly resolved in the density map, suggesting its flexibility. This recognition mode is significantly distinct from that of ABCD1 bound to C22:0-CoA (PDB ID: 7VZB), in which the acyl moiety of C22:0-CoA is recognized by the hydrophobic pocket formed by TMs 3’, 4’, 5’, and 6’ 10 (Fig. 1 f,g). In our C20:5-CoA-bound and apo inward-facing ABCD3 structure, the hydrophobic pocket observed in ABCD1 collapses due to the proximity of TMs 3’ and 4’ (Fig. 1 h,i), which block substrate-binding 10 . In addition, the luminal loop 2 (LL2) between TMs 3’ and 4’ is shorter in ABCD3 than in ABCD1 (Supplementary Fig. 1), whose LL2 protrudes further toward the luminal side compared with that of ABCD3 (Supplementary Fig. 5a,b). This structural difference induces closer packing of TM3’ and TM4’ in ABCD3, thus eliminating the hydrophobic pocket. In particular, the M233’ residue on TM4’ in ABCD3, which is not conserved in ABCD1 or ABCD2 (Supplementary Fig. 1), is located at the hydrophobic pocket and blocks substrate entry (Fig. 1 i). On the surface at the peroxisomal end of substrate-binding site 2, R362 in ABCD3 forms a salt bridge with E111 (Fig. 1 e), which is not conserved among the ABCD family members 10 , 12 (Supplementary Fig. 1), and adopts a distinct conformation from the corresponding R389 residue in ABCD1, which forms the gate towards the peroxisomal side 12 (Fig. 1 j). A superimposition of the acyl-CoA-bound inward-facing ABCD3 (C20:5) and ABCD1 (C22:0) structures reveals that the acyl moiety of C20:5-CoA bound to ABCD3 sterically clashes with R389 in ABCD1, suggesting that the distinct conformation of R362 in ABCD3 allows the accommodation of the acyl moiety (Fig. 1 j). Together, these findings reveal the unique mechanism of unsaturated long-chain acyl-CoA recognition by ABCD3, which differs substantially from that of saturated long-chain acyl-CoA recognition by ABCD1. Bile acid precursor recognition mechanism In addition to unsaturated long-chain acyl-CoAs, ABCD3 transports the bile acid precursors 3α,7α-dihydroxycholestanoic acid (DHCA) and 3α,7α,12α-trihydroxycholestanoic acid (THCA) 7 , 9 , which share structural similarity with cholesteryl hemisuccinate (CHS) (Supplementary Fig. 6), to biosynthesize bile acids in peroxisomes 18 . To understand how ABCD3 recognizes and transports bile acid precursors, which are structurally quite different from acyl-CoA molecules, we determined the cryo-EM structure of ABCD3 in the CHS-bound state at 3.8 Å resolution (Fig. 2 a,b, and Supplementary Figs. 3b and 6). The EM density map clearly shows that the CHS molecule binds to substrate-binding site 2 (Fig. 2 c), where the thioester bond and the acyl chain moiety are accommodated in the C20:5-CoA-bound state. The cholesteryl moiety of CHS is recognized by the hydrophobic surface of substrate-binding site 2, while its hemisuccinate moiety interacts with Y330 and R362 at the peroxisomal end of the site (Fig. 2 c). M233 on TM4, which constitutes the hydrophobic surface of substrate-binding site 2, protrudes into the lateral gate formed by TMs 3–6, thereby narrowing it and forming the CHS-binding pocket (Fig. 2 c,d and Supplementary Fig. 5a). By contrast, the CHS-unbound ABCD1 structure exhibits a wider lateral gate compared with ABCD3, with TM4 shifted away, thereby failing to form the CHS-binding pocket observed in the present CHS-bound ABCD3 10,11,13,14 . These structural observations suggest that the binding mode of CHS at substrate-binding site 2 in ABCD3 mimics that of physiological substrate bile acid precursors, highlighting the key role of ABCD3 in bile acid precursor transport 9 . To further investigate the bile acid precursor recognition mechanism, we performed MD simulations of CHS-bound and THCA-bound ABCD3, in which the THCA molecule was modeled at substrate-binding site 2, based on our CHS-bound inward-facing ABCD3 structure. Our MD simulation results showed that the cholesteryl backbones of both CHS and THCA remained stable at substrate-binding site 2 throughout the 100-ns simulations (Fig. 2 e,f and Supplementary Fig. 7), supporting the notion that bile acid precursors such as THCA bind to this site similarly to CHS. By contrast, the terminus of the hemisuccinate moiety of CHS and the side chains of both CHS and THCA exhibited higher flexibility (Supplementary Fig. 7g,h). Together, these results suggest that ABCD3 recognizes substrates promiscuously by a hydrophobic surface at substrate-binding site 2, which interacts with both acyl chains and steroid moieties. ATP recognition in the substrate-binding site To clarify the transport mechanism by ABCD3, we purified the catalytically inactive mutant, in which the catalytic E596 residue in the NBD is substituted by a glutamine residue (ABCD3 EQ ), to capture the ATP-bound structures. In the presence of ATP-Mg 2+ , we determined two distinct ATP-bound ABCD3 EQ structures in the presence (Fig. 3 a,b and Supplementary Fig. 8) or absence (Supplementary Figs. 9 and 10a,b) of cholic acid, which shares structural similarity with transport substrates. The ATP-bound structure in the presence of cholic acid adopts the inward-facing conformation similar to the CHS-bound state, with the distance between NBDs almost unchanged among the structures (Fig. 3 a,b and Supplementary Fig. 4), whereas the ATP-bound structure in the absence of cholic acid adopts the outward-facing conformation (described in the next section). We determined the structure of ABCD3 EQ in the ATP-bound inward-facing state, in the presence of cholic acid, with C2 symmetry at 2.9 Å resolution (Fig. 3 a,b, and Supplementary Figs. 8 and 11a). Although we observed undefined density at the lateral gate, we could not determine whether this originates from cholic acid due to the poor resolution. The structure revealed that ATP binds to the Walker motif in the NBDs, in which the distance between the two corresponding Y282 residues is 59.1 Å (Supplementary Fig. 4), indicating that this structure represents a conformational state just before NBD dimerization upon ATP binding. Notably, in addition to the ATPs bound to the NBDs, an ATP molecule is located at substrate-binding site 1 (Fig. 3 a,c,d), suggesting ATP transport by ABCD3, consistent with the previous study 16 . The adenosine and triphosphate moieties of ATP are recognized by Q369 and R373, and by the positively charged residues R96, K209, R373, and K322’, which also interact with the CoA moiety of acyl-CoA in the C20:5-CoA bound state, respectively (Fig. 3 c). This structural observation indicates that the binding mode of ATP in substrate-binding site 1 differs significantly from that of the ADP moiety of C20:5-CoA (Figs. 1 c and 3 c). The 3D classification with C1 symmetry revealed that the two substrate ATP molecules bind asymmetrically and form a hydrogen bond with each other (Fig. 3 e), likely representing an intermediate state during ATP transport. In both the C2 and C1 maps, the triphosphate densities at substrate-binding site 1 are well resolved, whereas the adenosine densities are poorly defined, suggesting their comparatively weaker recognition and greater flexibility. Together, the substrate ATP-bound state structure provides the first view of this new substrate recognition mode by the ABC family transporters. Conformational change of ABCD3 during the transport cycle We sought to obtain the cryo-EM map of the branched acyl-CoA-bound ABCD3, and acquired cryo-EM images of ABCD3 under phytanoyl-CoA-containing conditions (Supplementary Fig. 12). However, the reconstituted 3D map at 3.3 Å resolution lacked clear density corresponding to the phytanoyl-CoA molecule (Supplementary Figs. 10e and 12). Thus, we refer to this structure as an apo state. The apo state structure adopts the inward-facing conformation with a 59.1 Å distance between the two corresponding Y282 residues in the TMDs (Supplementary Fig. 4), thereby highlighting the conformational change upon acyl-CoA binding (Fig. 4 a, and Supplementary Fig. 4). Previous studies 10 , 12 , 14 suggested that the interactions between the two C-terminal helices may facilitate the dimerization of NBDs in apo ABCD1. Interestingly, the apo inward-facing conformation of ABCD3 lacking the corresponding C-terminal helices is similar to the ABCD1 apo inward-facing conformation (PDB ID: 7VWC) (Supplementary Fig. 5a), indicating that the distance between the NBDs in the ABCD3 apo inward-facing conformation is stable without the C-terminal helices. Although this apo inward-conformation of ABCD3 resembles the apo inward-conformation of ABCD1, the lateral gate in ABCD3 is narrower than that in ABCD1 10 (Supplementary Fig. 5a). The cryo-EM map of the apo ABCD3 state shows several lipid-like densities on the protein surface. In both apo structures of ABCD1 and ABCD3, no density corresponding to transport substrates was observed, but lipid densities were present. In particular, we observed an acyl-CoA or phospholipid-like density at the narrow lateral gate in the inner leaflet (Supplementary Fig. 10e,f), as previously observed in the apo state structure of ABCD1 10,14 , although we could not model the specific molecule due to its low resolution. This density interacts with L237, V238, L242, L371, and V375 on the membrane side and with T245, R248, and R379 on the membrane surface, implying that the transport of amphiphilic molecules such as acyl-CoAs may be nonspecifically mediated by the lateral gate. Superimposition of one side of the TM bundle (TM1, TM2, TM3, TM6, and TM4′/TM5′) between the apo and C20:5-CoA-bound states shows good alignment, with an RMSD of 0.89 Å over 261 Cα atoms (Fig. 4 a). Among the residues involved in the CoA moiety recognition, R96 and K209 adopt distinct conformations between the apo and C20:5-CoA-bound states (Fig. 4 a,c). Upon CoA binding, R96, which interacts with D213 in the apo state, undergoes a large structural rearrangement and newly interacts with the 5′-diphosphate group of C20:5-CoA. This structural rearrangement of R96 creates a space that is occupied by K209, which then interacts with the adenine ring of C20:5-CoA, allowing it to bind to substrate-binding site 1. By contrast, a structural comparison of the opposite TM bundles between the apo and C20:5-CoA-bound states reveals a conformational change, particularly attributed to electrostatic interactions between K144 and the 3′-phosphate group of C20:5-CoA (Figs. 1 c and 4 a,d). These structural observations indicate the rigid body movement between TM bundles upon C20:5-CoA binding, thereby explaining the different NBD distances between the apo and C20:5-CoA inward-facing conformations. We determined the cryo-EM structure of ABCD3 EQ in the ATP-bound outward-facing state at 3.2 Å resolution, in the absence of cholic acid (Supplementary Figs. 9a, 10a,b and 11b). The structure revealed that ATP binds to the Walker motif in the NBDs, in which the distance between Y282 is 30.6 Å (Supplementary Fig. 4). In the outward-facing structure, TM6 adopts a straight conformation, and the luminal loop 3 (LL3) between TMs 5 and 6 is ordered (Supplementary Figs. 10a and 11b). This arrangement reduces the exposure of the central cavity to the peroxisomal side, as compared to that of the ABCD1 outward-facing state (PDB ID: 7RR9), in which TM6 is bent with LL3 disordered (Supplementary Fig. 5c), thereby facilitating the release of substrates from the central cavity into the peroxisomes. A structural comparison between the inward- and outward-facing states reveals that the residues constituting substrate-binding sites 1 and 2 adopt distinct conformations (Fig. 4 b,e), which probably reduce the substrate affinity in the outward-facing state. Together, our ABCD3 structures in the apo, CHS-, substrate ATP-, and C20:5-CoA-bound inward-facing, and ATP-bound outward-facing states reveal the conformational transitions of ABCD3 during the transport cycle (Fig. 5 ), suggesting the substrate transport mechanism of ABCD3. Discussion In this study, we determined different substrate-bound and multiple conformations of ABCD3 during the transport cycle (Fig. 5 ). In the inward-facing states, the transport substrates are mediated by the lateral gate and/or from the intracellular side, and hydrophilic and subsequently hydrophobic moieties of substrates are recognized at the substrate-binding sites 1 and 2, respectively. The conformational transition from the inward-facing state to the outward-facing state upon ATP-binding in the NBDs induces the substrate release into the peroxisomes. The C20:5-CoA-bound structure provides insights into the substrate specificity of ABCD3 for acyl-CoA. ABCD1 recognizes saturated long-chain acyl-CoA molecules that fit into the hydrophobic cavity formed by TMs 3’–6’, but does not transport dicarboxylic and bulky acyl-CoA molecules. By contrast, ABCD3 transports bulky and unsaturated acyl-CoA molecules via the large cavity within the TMD. In addition, ABCD3 is capable of transporting dicarboxylic acyl-CoA, possibly owing to the hydrophilic surface at the peroxisomal end of substrate-binding site 2, unlike the hydrophobic pocket in ABCD1. ABCD3 lacks the hydrophobic pocket observed in ABCD1, which accommodates extended saturated acyl chains, thereby explaining why ABCD1 preferentially transports such substrates over ABCD3 19 . Although the 3′-phosphate ADP moiety of CoA is structurally similar to ATP, their binding modes at substrate-binding 1 are completely different. This difference likely stems from the chemical distinction at the 3′ position of the ribose: CoA contains a phosphate group, whereas ATP possesses a hydroxyl group. The 3′-phosphate group of CoA is recognized by K144, and this interaction induces a rigid-body shift of the TMD. By contrast, the 3′-hydroxyl group in ATP is insufficient to induce such a conformational change, resulting in the distance between the NBDs remaining unchanged from the apo state, and the substrate ATP failing to adopt the binding mode observed for the 3′-phosphate ADP moiety of CoA. The CHS-bound structure suggests that bile acid precursor molecules such as DHCA and THCA, substrates of ABCD3 7,9 , probably bind to substrate-binding site 2, similar to the CHS-binding mode. Our structural comparison between the CHS- and C20:5-CoA-bound states of ABCD3 revealed that the binding site of CHS overlaps with that of the thioester bond and acyl chain moiety of acyl-CoA at substrate-binding site 2. Given that ABCD3 transports DHCA and THCA in their CoA-conjugated forms 20 , these findings suggest that the binding modes of DHCA-CoA and THCA-CoA may differ slightly from those of typical acyl-CoAs. We determined the substrate ATP-bound ABCD3 structure under conditions with cholic acid and an excess of ATP and Mg 2+ (Fig. 3 a,b and Supplementary Fig. 8). We first expected that the cholic acid molecules would bind to the outward-facing ABCD3, as indeed the ATP-bound ABCD3 structure determined under the excess ATP and Mg 2+ -containing conditions adopts the outward-facing conformation (Supplementary Fig. 10a,b). However, the substrate ATP-bound ABCD3 structure in the presence of cholic acid adopts the inward-facing conformation, suggesting a potential preference of cholic acid binding to the inward-facing rather than the outward-facing state of ABCD3, although we could not confidently assign the cholic acid model to the density observed at the lateral gate. In addition, the structure revealed that the adenine ring is located near substrate-binding site 2, implying that ATP binding prevents cholic acid from occupying this site, in a similar manner to that observed in the CHS-bound state. Alternatively, the different ATP-incubation conditions may have affected the captured conformations. We incubated ATP with ABCD3 EQ in glyco-diosgenin (GDN) for 1 h at 4°C for the inward-facing conformation, whereas we incubated ATP with ABCD3 EQ in nanodiscs for 20 min at room temperature, followed by concentration, for the outward-facing conformation (see Methods). Collectively, this work provides structural insights into the acyl-CoA substrate specificity and the distinct mechanisms of substrate recognition and transport involving the lateral gate and two spatially separate substrate-binding sites of ABCD3, and paves the way for a deeper understanding of disorders associated with ABCD3. Methods Protein preparation ABCD3 WT and ABCD3 EQ were expressed using the Bac-to-Bac baculovirus expression system (Thermo Fischer Scientific), according to the method reported previously 21 . Spodoptera frugiperda Sf9 insect cells (Thermo Fischer Scientific) were infected at a density of 3-4 ×10 6 cells ml -1 . The infected Sf9 cells were incubated in Sf900II medium (Gibco) at 27℃ for 48 h, and then collected by centrifugation at 5,000 x g for 10 min. The cell pellets were resuspended in lysis buffer, containing 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, and 10% glycerol (v/v), and homogenized. For the apo state, the membrane proteins were extracted from the cells with 1% (w/v) LMNG and protease inhibitors, and rotated gently at 4°C for 1.5 h. After centrifugation at 40,000 rpm for 30 min, the supernatant was mixed with CNBr-activated Sepharose resin (GE) coupled with the GFP-binding nanobody ‘enhancer’ 22 and stirred at 4°C for 1 h. The resin was washed with 20 column volumes of wash buffer 2, containing 50 mM Tris (pH 8.0), 500 mM NaCl, 0.03% (w/v) GDN, and 10% glycerol. HRV3C protease was added, and the resin was rotated gently at 4°C for 3 h. The eluate was then mixed with Ni-NTA resin and stirred at 4°C for 30 min. The eluate was collected with wash buffer 2 and concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma). The solution was centrifuged at 40,000 rpm for 15 min. The supernatant was purified by SEC on a Superose 6 Increase 10/300 GL column, equilibrated with SEC buffer 2 (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.03% GDN). The fractions containing protein peaks were collected. The protein was concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma) and centrifuged at 40,000 rpm for 15 min. The supernatant was used for grid preparation. For the acyl-CoA- and CHS-bound states, protein expression and cell collection were performed using the same procedure as for the apo state. The membrane proteins were extracted from the cells with a solution containing 1% (w/v) LMNG, 0.05% (w/v) DDM, 0.02% (w/v) cholesteryl hemisuccinate (CHS) (Sigma), 5.2 mg ml -1 aprotinin, 2.0 mg ml -1 leupeptin, 1.4 mg ml -1 pepstatin A, and 100 mM phenylmethylsulfonyl fluoride, by gentle rotation at 4°C for 2 h. After solubilization purification with a GFP-nanobody, SEC was performed, using the same procedure as for the apo state. For the ATP-bound inward-facing state, protein expression, cell collection, solubilization, and GFP-nanobody purification were using the same procedure as for the apo state. The eluate from the GFP-nanobody purification was purified by SEC on a Superose 6 Increase 10/300 GL column, equilibrated with SEC buffer 3 (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.03% GDN, 0.01% cholic acid). The fractions containing protein peaks were collected. The protein was concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma) and centrifuged at 40,000 rpm for 15 min. The supernatant was used for grid preparation. For the ATP-bound outward-facing state ABCD3 EQ , protein expression and cell collection were performed using the same procedure as for the apo state. The membrane proteins were extracted from the cells with a solution containing 1% (w/v) n-dodecyl-beta-D-maltopyranoside (DDM, Merck), 5.2 mg ml -1 aprotinin (Calbiochem), 2.0 mg ml -1 leupeptin (Calbiochem), 1.4 mg ml -1 pepstatin A (Sigma), and 100 mM phenylmethylsulfonyl fluoride (PMSF; Roche), by gentle rotation at 4°C for 1.5 h. After centrifugation at 40,000 rpm for 30 min (Beckman, Type 70 Ti), the supernatant was mixed with TALON resin (Takara Bio Company) and stirred at 4°C for 1 h. The resin was washed with 20 column volumes of buffer, containing 50 mM Tris (pH 8.0), 500 mM NaCl, 0.05% (w/v) DDM, 0.01% (w/v) CHS, and 15 mM imidazole. The protein was eluted with 4 column volumes of buffer, containing 50 mM Tris (pH 8.0), 500 mM NaCl, 0.05% (w/v) DDM, 0.01% (w/v) CHS, 10% (v/v) glycerol, and 200 mM imidazole. The eluate was concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma) before the nanodisc reconstitution. Nanodisc reconstitution of ATP-bound ABCD3EQ Before reconstitution, 10 mg Brain Total Lipid Extract (BL, Avanti), dissolved in chloroform, was dried using a nitrogen stream and the residual chloroform was further removed by vacuum desiccation (O/N). Lipids were then rehydrated in buffer, containing 1% DDM, 50 mM Tris-HCl (pH 8.0), and 150 mM NaCl. The purified protein was mixed with BPL (please define-brain phospholipids?) and MSP1D1 (membrane scaffold protein 1D1) at a molar ratio of 1:8:100 (ABCD3:MSP1D1:BPL). Nanodisc reconstitution proceeded for 1 h at 4℃. For detergent removal, Bio-Beads SM2 (Bio-Rad) pre-equilibrated with SEC buffer 1 (50 mM Tris-HCl, pH 8.0, 150 mM NaCl) were added and the reconstituted protein mixture was gently rotated for 2 h at 4℃. Afterward, the Bio-Beads were exchanged with fresh Bio-Beads and the mixture was gently rotated overnight at 4℃. The solution, excluding the Bio-Beads, was centrifuged at 40,000 rpm for 15 min. The supernatant was purified by size exclusion chromatography (SEC) on a Superose 6 Increase 10/300 GL column (GE Healthcare), equilibrated with SEC buffer 1. The fractions containing protein peaks reconstituted into nanodiscs were collected and incubated with 5 mM ATP and 5 mM MgCl 2 for 20 min at room temperature (RT). The mixture was concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma) and centrifuged at 40,000 rpm for 15 min. The supernatant was used for grid preparation. Grid preparation and Cryo-EM data collection For the apo state, phytanoyl CoA was added to the purified ABCD3 and incubated at 4℃ for 1 h. The solution was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4°C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained with a 300 kV Titan Krios G3i microscope (Thermo Fisher Scientific), equipped with a GIF Quantum energy filter (Gatan) and a K3 Summit direct electron detector (Gatan). In total, 8,525 movies were acquired at a nominal magnification of 105,000×, with a calibrated pixel size of 0.83 Å pix -1 and a defocus range of -0.8 to -1.6 μm, using the EPU software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 50 e - Å -2 at the grid. For the acyl-CoA-bound state, 1 mM 5Z,8Z,11Z,14Z,17Z-eicosapentaenoyl Coenzyme A (Avanti) was added to the purified ABCD3 and incubated at 4℃ for 1 h. The solution was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4°C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained as described above. In total, 9,829 movies were acquired at a nominal magnification of 105,000×, with a calibrated pixel size of 0.83 Å pix -1 and a defocus range of -0.8 to -1.6 μm, using the EPU software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 48 e - Å -2 at the grid. For the CHS-bound ABCD3, the purified ABCD3 was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4°C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained as described above. In total, 5,119 movies were acquired at a nominal magnification of 105,000× with a calibrated pixel size of 0.83 Å pix -1 and a defocus range of -0.8 to -1.6 μm, using the SerialEM software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 49.027 e - Å -2 at the grid. For the ATP-bound inward-facing ABCD3 EQ , 5 mM ATP and 5 mM MgCl 2 were added to the purified ABCD3 solution and incubated at 4℃ for 1 h. The solution was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4°C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained as described above. In total, 4,532 movies were acquired at a nominal magnification of 105,000×, with a calibrated pixel size of 0.83 Å pix -1 and a defocus range of -0.8 to -1.6 μm, using the EPU software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 50 e - Å -2 at the grid. For the ATP-bound outward-facing ABCD3 EQ in nanodiscs, the nanodisc-reconstituted and ATP/Mg 2+ -incubated ABCD3 EQ was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4°C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained as described above. In total, 3,942 movies were acquired at a nominal magnification of 105,000×, with a calibrated pixel size of 0.83 Å pix -1 and with a defocus range of -0.8 to -1.6 μm, using the SerialEM software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 51.362 e - Å -2 at the grid. Image processing For the acyl-CoA-bound state, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, using RELION 3.1 23 . The CTF parameter was calculated in CryoSPARC 24 . The following steps were carried out with CryoSPARC. Blob picking was performed on all images, and 1,198,153 particles were extracted at 3.32 Å/pixel. These particles were subjected to 2D classification. After 2D classification, 32,216 particles were used for Topaz 25 picking, and 3,647,328 were extracted at 3.32 Å/pixel. Please note: Kindly check these numbers. These particles were subjected to 2D classification, Ab-Initio Reconstruction and Heterogeneous Refinement. Selected particles were re-extracted at 1.66 Å/pixel. The re-extracted particles were subjected to Ab-Initio Reconstruction and Heterogeneous Refinement with C2 symmetry. Selected particles were re-extracted at 1.107 Å/pixel, and then selected by the same procedure. Afterward, 204,250 particles were symmetry expanded and subjected to 3D classification without alignment in RELION 4.0 26 . Because the substrate density was located near the axis of symmetry, a mask was placed around the substrate in the 3D classification to obtain a substrate-bound density on only one side. Finally, 119,378 particles were selected, and a density map with a global resolution of 3.33 was obtained by Local-refinement in CryoSPARC without symmetry. With C2 symmetry, 204,250 particles were 3D classified by RELION 4.0 and a density map with a global resolution of 3.29 Å was obtained by NU-refinement. The C2 map was used for modeling. For the CHS-bound state, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, using RELION 3.1. The CTF parameter was calculated in CryoSPARC. The following steps were carried out with CryoSPARC. Blob picking was performed on all images, and 1,267,885 particles were extracted at 3.32 Å/pixel. These particles were subjected to 2D classification, and 1,447,563 particles were extracted at 3.32 Å/pixel by Topaz picking. Please note: Kindly check these numbers. These particles were subjected to 2D classification, Ab-Initio Reconstruction, and Heterogeneous Refinement. The resulting particles were re-extracted at 1.107 Å/pixel. Re-extracted particles were selected by Ab-Initio Reconstruction and Heterogeneous Refinement with C2 symmetry. Selected particles were subjected 3D classification without alignment in RELION 4.0. Finally, 44,696 particles were selected and a density map with a global resolution of 3.81 Å was obtained by NU refinement in CryoSPARC. This map was used for modeling. For the ATP-bound inward-facing state, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, using RELION 3.1. The CTF parameter was calculated in CryoSPARC. The following steps were carried out within CryoSPARC. Blob picking was performed on all images, and 1,327,401 particles were extracted at 1.66 Å/pixel. These particles were subjected to 2D classification, and then 50,694 particles were used for Topaz picking, resulting in 1,378,789 particles extracted at 3.32 Å/pixel. Please note: Kindly check these numbers. These particles were subjected to 2D classification, Ab-Initio Reconstruction, and Heterogeneous Refinement. Selected particles were re-extracted at 1.107 Å/pixel and the density map was reconstructed by NU-Refinement with C2. Aligned particles were subjected 3D classification without alignment in RELION 4.0. Finally, 65,336 particles were selected and a density map with a global resolution of 2.94 Å was obtained by NU-refinement in CryoSPARC. This map was used for modeling. For the apo state, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, using RELION 3.1. The CTF parameter was calculated in CryoSPARC. The following steps were carried out with CryoSPARC. Blob picking was performed on all images, and 2,479,005 particles were extracted at 3.32 Å/pixel. These particles were subjected to 2D classification, which yielded 19,690 particles for Topaz picking. After 2,914,648 particles were extracted at 3.32 Å/pixel by Topaz picking, they were subjected to 2D classification, Ab-Initio Reconstruction, and Heterogeneous Refinement. Please note: Kindly check these numbers. The resulting particles were re-extracted at 1.66 Å/pixel. Re-extracted particles were selected by Ab-Initio Reconstruction and Heterogeneous Refinement with C2 symmetry. Selected particles were re-extracted at 1.107 Å/pixel and subjected 3D classification without alignment in RELION 4.0. Finally, 73,450 particles were selected, and a density map with a global resolution of 3.28 Å was obtained by NU refinement in CryoSPARC. This map was used for modeling. For the ATP-bound outward-facing state ABCD3 EQ in nanodiscs, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, and the contrast transfer function (CTF) parameter was calculated using RELION 3.1. Laplacian-of-Gaussian picking was performed on 1,000 of these images, and 26,7297 particles were extracted at 3.403 Å/pixel. These particles were subjected to 2D classification. The 8,564 particles thus generated were used for template picking, and 3,875,206 were extracted at 3.403 Å/pixel. After 2D classification, 20,179 particles were used for Topaz picking. Please note: Kindly check these numbers. The data processing steps after Topaz picking were carried out with RELION 4.0. These particles were subjected to 2D and 3D classification and 3D refinement. The resulting particles were subjected to CTF refinement and postprocessing. Finally, 20,043 particles extracted at 1.51244 Å/pixel were selected to produce a map with a global resolution of 3.3 Å. This map was used for modeling. Model building and refinement The density map exhibited sufficient quality to build a model in COOT 27 . The structure predicted by AlphaFold 2 28 (AF2) was fitted by MOLREP. Next, we manually readjusted ABCD3 and modeled ATP and Mg 2+ molecules. The models were refined using phenix.real_space_refine 29 and Servalcat 30 , and validated using Molprobity 31 . The F o − F c omit map was calculated using Servalcat. MD simulations The CHS-bound ABCD3 model was constructed based on the cryo-EM structure, and the THCA-bound ABCD3 model was generated using the CHS-bound model as the template. The unresolved loop regions were modeled as flexible segments. All molecular dynamics (MD) simulations were performed using GROMACS 2020.3 32 . The protein structures were embedded in a 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) lipid bilayer, using the CHARMM-GUI 33 web server. The system was neutralized by adding Cl⁻ ions and consisted of approximately 360 lipids, 52,000 water molecules, 240 chloride ions, and 210 sodium ions. The ionic concentration was adjusted to 150 mM NaCl, and TIP3P water molecules were used. Ligand topologies were generated using the CGenFF 34 web server, and all other topologies and force field parameters were taken from the CHARMM36m 35 force field. A time step of 2 fs was used for both equilibration and production runs. The systems were energy-minimized until the maximum force was below 1,000 kJ mol⁻¹ nm⁻², with positional restraints applied to all non-hydrogen atoms. Equilibration was performed in the NVT and NPT ensembles for 0.1 ns and 10 ns, respectively, using positional restraints of 1,000 kJ mol⁻¹ nm⁻². Production runs were carried out three times for 100 ns each, without any restraints, under the NPT ensemble with semi-isotropic pressure coupling. The temperature was maintained at 310 K using the Nosé–Hoover thermostat, and the pressure was maintained at 1 bar using the Parrinello–Rahman barostat. Long-range electrostatic interactions were calculated using the particle mesh Ewald (PME) method. Simulation results were analyzed using the PyMOL program (http://www.pymol.org/pymol), MDAnalysis 36,37 , and MDTraj 38 . Declarations Data availability Cryo-EM density maps have been deposited in the Electron Microscopy Data Bank, under the accession codes EMD-65674, EMD-65675, EMD-65676, EMD-65677 and EMD-65678 for the apo state, the C20:5-CoA-bound state, the CHS-bound state, the ATP-bound inward-facing state, and the ATP-bound outward-facing state structures, respectively. Atomic coordinates have been deposited in the Protein Data Bank, under the IDs 9W62, 9W63, 9W64, 9W65 and 9W66 for the apo state, the C20:5-CoA-bound state, the CHS-bound state, the ATP-bound inward-facing state, and the ATP-bound outward-facing state structures, respectively. The raw images have been deposited in the Electron Microscopy Public Image Archive under the accession code EMPIAR-XXXXX. Acknowledgements We thank K. Ogomori, C. Harada, T. Tanaka, and T. A. Kobayashi at The University of Tokyo for their technical and editorial assistance in structural analysis, and K. Kawaguchi, M. Morita, and T. So at Toyama University for their help with plasmid preparation. This work was supported by JSPS KAKENHI Grant Numbers JP16H06294 (O.N.), JP20K15754 (T.K.), JP22K15072 (T.K.), JP24K01961 (T.K.), JP25KJ1114 (S.A.), JST PRESTO Grant Number JPMJPR22E4 (T.K.), JST CREST Grant Number JPMJCR20E2 (O.N.) and Research Support Project for Life Science and Drug Discovery (Basis for Supporting Innovative Drug Discovery and Life Science Research (BINDS)) from AMED under Grant Numbers JP21am0101115 (support number 3272), and JP24ama121002 (support number 3272). Author contributions S.A. prepared the cryo-EM samples, and H.H.O. and A.T. assisted with ABCD3 construction, expression, and purification. S.A. and F.K.S. collected the cryo-EM data. S.A., T.K., A.T. and F.K.S. processed the cryo-EM data. S.A. and T.K. build the structures. S.A. and F.K.S. performed MD simulations and analyzed the results. S.A., T.K. and O.N. wrote the manuscript. H.H.O., T.K. and O.N. supervised the research. Competing interest declaration The authors declare no competing interests. References Maxfield, F.R., Tabas, I.: Role of cholesterol and lipid organization in disease. Nature. 438 , 612–621 (2005) Wang, M., Casey, P.J.: Protein prenylation: unique fats make their mark on biology. Nat. Rev. Mol. Cell. Biol. 17 , 110–122 (2016) Nguyen, P., et al.: Liver lipid metabolism. J. Anim. Physiol. Anim. Nutr. 92 , 272–283 (2008) Grevengoed, T.J., Klett, E.L., Coleman, R.A.: Acyl-CoA metabolism and partitioning. Annu. Rev. Nutr. 34 , 1–30 (2014) Watkins, P.A., Ellis, J.M.: Peroxisomal acyl-CoA synthetases. Biochim. Biophys. Acta - Mol. Basis Dis. 1822 , 1411–1420 (2012) Kawaguchi, K., Morita, M.A.B.C., Transporter Subfamily, D.: Distinct Differences in Behavior between ABCD1-3 and ABCD4 in Subcellular Localization, Function, and Human Disease. BioMed Res. Int. (2016). (2016) Ranea-Robles, P., et al.: The peroxisomal transporter ABCD3 plays a major role in hepatic dicarboxylic fatty acid metabolism and lipid homeostasis. J. Inherit. Metab. Dis. 44 , 1419–1433 (2021) Wiesinger, C., Kunze, M., Regelsberger, G., Forss-Petter, S., Berger, J.: Impaired very long-chain acyl-CoA β-oxidation in human X-linked adrenoleukodystrophy fibroblasts is a direct consequence of ABCD1 transporter dysfunction. J. Biol. Chem. 288 , 19269–19279 (2013) Ferdinandusse, S., et al.: A novel bile acid biosynthesis defect due to a deficiency of peroxisomal ABCD3. Hum. Mol. Genet. 24 , 361–370 (2015) Chen, Z.P., et al.: Structural basis of substrate recognition and translocation by human very long-chain fatty acid transporter ABCD1. Nat. Commun. 13 , (2022) Wang, R., Qin, Y., Li, X.: Structural basis of acyl-CoA transport across the peroxisomal membrane by human ABCD1. Cell. Res. (2021). 10.1038/s41422-021-00585-8 Le, L.T.M., Thompson, J.R., Dang, P.X., Bhandari, J., Alam, A.: Structures of the human peroxisomal fatty acid transporter ABCD1 in a lipid environment. Commun. Biol. 5 , (2022) Jia, Y., et al.: Structural and functional insights of the human peroxisomal ABC transporter ALDP. eLife 11, 1–11 (2022) Xiong, C., et al.: Structural insights into substrate recognition and translocation of human peroxisomal ABC transporter ALDP. Signal. Transduct. Target. Ther. 8 , (2023) Li, Y., et al.: Structural insights into human ABCD3-mediated peroxisomal acyl-CoA translocation. Cell. Discov 10 , (2024) van Roermund, C.W.T., IJlst, L., Linka, N., Wanders, R.J.A., Waterham, H.R.: Peroxisomal ATP Uptake Is Provided by Two Adenine Nucleotide Transporters and the ABCD Transporters. Front. Cell. Dev. Biol. 9 , 1–12 (2022) Xu, D., et al.: Cryo-EM structure of human lysosomal cobalamin exporter ABCD4. Cell. Res. 29 , 1039–1041 (2019) Ferdinandusse, S., Denis, S., Faust, P.L., Wanders, R.J.: A. Bile acids: the role of peroxisomes. J. Lipid Res. 50 , 2139–2147 (2009) van Roermund, C.W.T., IJlst, L., Wagemans, T., Wanders, R.J.A., Waterham, H.: R. A role for the human peroxisomal half-transporter ABCD3 in the oxidation of dicarboxylic acids. Biochim. Biophys. Acta BBA - Mol. Cell. Biol. Lipids. 1841 , 563–568 (2014) Prydz, K., Kase, B.F., Björkhem, I., Pedersen, J.I.: Subcellular localization of 3 alpha, 7 alpha-dihydroxy- and 3 alpha,7 alpha,12 alpha-trihydroxy-5 beta-cholestanoyl-coenzyme A ligase(s) in rat liver. J. Lipid Res. 29 , 997–1004 (1988) Dror, R.O., et al.: Structural basis for nucleotide exchange in heterotrimeric G proteins. Science. 348 , 1361–1365 (2015) Kirchhofer, A., et al.: Modulation of protein properties in living cells using nanobodies. Nat. Struct. Mol. Biol. 17 , 133–138 (2010) Zivanov, J., et al.: New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife 7, e42166 (2018) Punjani, A., Rubinstein, J.L., Fleet, D.J., Brubaker, M.A.: cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods. 14 , 290–296 (2017) Bepler, T., et al.: Positive-unlabeled convolutional neural networks for particle picking in cryo-electron micrographs. Nat. Methods. 16 , 1153–1160 (2019) Kimanius, D., Dong, L., Sharov, G., Nakane, T., Scheres, S.H.: W. New tools for automated cryo-EM single-particle analysis in RELION-4.0. Biochem. J. 478 , 4169–4185 (2021) Emsley, P., Lohkamp, B., Scott, W.G., Cowtan, K.: Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 66 , 486–501 (2010) Jumper, J., et al.: Highly accurate protein structure prediction with AlphaFold. Nature. 596 , 583–589 (2021) Liebschner, D., et al.: Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr. Sect. Struct. Biol. 75 , 861–877 (2019) Yamashita, K., Palmer, C.M., Burnley, T., Murshudov, G.N.: Cryo-EM single-particle structure refinement and map calculation using Servalcat. Acta Crystallogr. Sect. Struct. Biol. 77 , 1282–1291 (2021) Davis, I.W., et al.: MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res. 35 , W375–W383 (2007) GROMACS 4: 5: a high-throughput and highly parallel open source molecular simulation toolkit | Bioinformatics | Oxford Academic. https://academic.oup.com/bioinformatics/article/29/7/845/253065 Jo, S., et al.: CHARMM-GUI 10 years for biomolecular modeling and simulation. J. Comput. Chem. 38 , 1114–1124 (2017) Vanommeslaeghe, K., et al.: CHARMM general force field: A force field for drug-like molecules compatible with the CHARMM all-atom additive biological force fields. J. Comput. Chem. 31 , 671–690 (2010) Huang, J., et al.: CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nat. Methods. 14 , 71–73 (2017) Michaud-Agrawal, N., Denning, E.J., Woolf, T.B., Beckstein, O.: MDAnalysis: A toolkit for the analysis of molecular dynamics simulations. J. Comput. Chem. 32 , 2319–2327 (2011) Gowers, R.J., et al.: MDAnalysis: A Python Package for the Rapid Analysis of Molecular Dynamics Simulations. scipy (2016). 10.25080/Majora-629e541a-00e McGibbon, R.T., et al.: MDTraj: A Modern Open Library for the Analysis of Molecular Dynamics Trajectories. Biophys. J. 109 , 1528–1532 (2015) Liu, Y., Schmidt, B., Maskell, D.L.: MSAProbs: multiple sequence alignment based on pair hidden Markov models and partition function posterior probabilities. Bioinformatics. 26 , 1958–1964 (2010) Robert, X., Gouet, P.: Deciphering key features in protein structures with the new ENDscript server. Nucleic Acids Res. 42 , W320–W324 (2014) Tables Table 1 | Cryo-EM data collection, refinement and validation statistics. ABCD3 (EMD-65674) (PDB 9W62) ABCD3-20:5 CoA (EMD-65675) (PDB 9W63) ABCD3-CHS (EMD-65676) (PDB 9W64) ABCD3- ATP(IF) (EMD-65677) (PDB 9W65) ABCD3-ATP(OF) (EMD-65678) (PDB 9W66) Data collection and processing Magnification ×105,000 ×105,000 ×105,000 ×105,000 ×105,000 Voltage (kV) 300 300 300 300 300 Electron exposure (e–/Å 2 ) 50 48 49 50 51 Defocus range (μm) −1.6 to −0.8 −1.6 to −0.8 −1.6 to −0.8 −1.6 to −0.8 −1.6 to −0.8 Pixel size (Å) 0.83 0.83 0.83 0.83 0.83 Symmetry imposed C2 C2 C2 C2 C2 Initial particle images (no.) 2,914,648 3,647,328 119,378 1,378,789 2,805,407 Final particle images (no.) 73,450 204,250 44,696 65,336 20,043 Map resolution (Å) FSC threshold 3.28 0.143 3.29 0.143 3.81 0.143 2.94 0.143 3.20 0.143 Map resolution range (Å) 2.79-37.93 2.73-43.47 3.22-43.27 2.59-44.22 2.14-4.92 Refinement Model resolution (Å) FSC threshold 3.28 0.5 3.29 0.5 3.81 0.5 2.94 0.5 3.20 0.5 Map sharpening B factor (Å 2 ) -100.2 -115.9 -111.5 -97.0 -69.4 Model composition Non-hydrogen atoms Protein residues Ligands 9548 1148 6 9736 1146 12 9418 1150 2 9588 1150 12 9722 1162 12 B factors (Å 2 ; mean) Protein Ligand 119.76 133.23 198.42 174,89 187.34 209.85 124.64 151.02 104.37 136.12 R.m.s. deviations Bond lengths (Å) Bond angles (°) 0.0110 2.07 0.0103 2.00 0.0108 2.01 0.0111 2.02 0.0089 1.54 Validation MolProbity score Clashscore Poor rotamers (%) 1.70 4.33 1.38 1.82 5.49 1.77 1.78 2.42 0.00 1.96 3.31 3.92 2.25 10.86 4.66 Ramachandran plot Favored (%) Allowed (%) Disallowed (%) 94.39 5.61 0.00 95.25 4.75 0.00 96.50 3.50 0.00 94.40 5.60 0.00 96.88 3.12 0.00 Additional Declarations There is NO Competing Interest. Supplementary Files 0804sup.pdf Supplementary Fig. 1 | Sequence alignment of human ABC transporters belonging to subfamily D. Multiple sequence alignment for human ABC transporters belonging to subfamily D. The alignment was performed by MSAProbs v0.9.7 39 and the figure was prepared with ESPrint3 40 . The secondary structures based on the cryo-EM structure of human ABCD3 are shown above the sequences. Supplementary Fig. 2 | Cryo-EM analysis of ABCD3 in the C20:5-CoA-bound inward-facing state. a, Flowchart of cryo-EM data processing for apo-state ABCD3, using cryoSPARC v4.2 and RELION 4.0. b, SEC profile of purified ABCD3. c, SDS-PAGE analysis of the eluted peak fraction from SEC. d, Representative 2D class averages. e, Representative cryo-EM micrograph. f, FSC curve of the final map, reconstructed using cryoSPARC v4.6. g, Structural formula of C20:5-CoA. Supplementary Fig. 3 | Density maps of ABCD3 in the C20:5-CoA-bound inward-facing state and the CHS-bound inward-facing state. a, Cryo-EM densities of TMs, loops, NBDs, and CoA molecules with the corresponding structure of C20:5-CoA-bound ABCD3. Contour levels are set at 2.1σ. b, Cryo-EM densities of TMs, Loops, NBDs, and CHS molecules with the corresponding structure of CHS-bound ABCD3. Contour levels are set at 2.2σ. Supplementary Fig. 4 | Distances between NBDs. Distances between NBDs are compared based on the distances between Cα atoms of Y282. Apo state, green; 20:5-CoA bound state, light blue; CHS-bound state, brown; ATP-bound-IF-state, yellow; and ATP-bound-OF-state, red. Supplementary Fig. 5 | Structural comparisons between ABCD1 and ABCD3. a-c, Structural superimpositions of ABCD1 and ABCD3, using secondary structure matching by CueMol. a, Apo state (ABCD1 (PDB ID: 7VWC) yellow-green, ABCD3 green). b, Acyl-CoA-bound state (ABCD1 (PDB ID: 7VZB) purple, ABCD3 light blue). c, ATP-bound-OF-state (ABCD1 (PDB ID: 7RR9) brown, ABCD3 red). Supplementary Fig. 6 | Cryo-EM analysis of ABCD3 in the CHS-bound inward-facing state. a, Flowchart of cryo-EM data processing for apo-state ABCD3, using cryoSPARC v3.1 and RELION 4.0. b, SEC profile of purified ABCD3. c, SDS-PAGE analysis of the eluted peak fraction from SEC. d, Representative 2D class averages. e, Representative cryo-EM micrograph. f, FSC curve of the final map, reconstructed using cryoSPARC v3.3. g, Chemical formulas of cholesteryl hemisuccinate (CHS) and 3α,7α,12α-trihydroxycholestanoic acid (THCA). Supplementary Fig. 7 | MD simulations. The results of the three independent runs are shown. All plots show the moving average as a solid line and the raw data as a semi-transparent line. a, All-atom RMSD plot from the MD simulation of the CHS-bound state. b, All-atom RMSD plot from the MD simulation of the THCA-bound state. c, Plot of the distance between the centers of mass of CHS and M233. d, Plot of the distance between the centers of mass of CHS and L365. e, Plot of the distance between the centers of mass of THCA and M233. f, Plot of the distance between the centers of mass of THCA and L365. g, Visualization of CHS atoms, colored according to RMSF values calculated from MD simulations. h, Visualization of THCA atoms, colored according to RMSF values calculated from MD simulations. Supplementary Fig. 8 | Cryo-EM analysis of ABCD3 in the ATP-bound inward-facing state. a, Flowchart of cryo-EM data processing for apo-state ABCD3, using cryoSPARC v3.3 and RELION 4.0. b, SEC profile of purified ABCD3. c, SDS-PAGE analysis of eluted fractions from SEC. d, Representative 2D class averages. e, Representative cryo-EM micrograph. f,g, FSC curves of the final C2 (f) and C1 (g) maps, reconstructed using cryoSPARC v4.4. h, Structural formulas of cholic acid and ATP. Supplementary Fig. 9 | Cryo-EM analysis of ABCD3 in the ATP-bound outward-facing state. a, Flowchart of cryo-EM data processing for apo-state ABCD3 using RELION 3.1 and RELION 4.0. b, SEC profile of purified ABCD3. c, SDS-PAGE analysis of the eluted peak fraction from SEC. d, Representative 2D class averages. e, Representative cryo-EM micrograph. f, FSC curve of the final map, reconstructed using RELION 4.0. Supplementary Fig. 10 | ATP-bound inward-facing state and apo inward-facing state of ABCD3. a,b, Overall structure of ATP-bound outward-facing state ABCD3, viewed from the membrane (a) and luminal (b) sides. Each protomer is colored red or gray. ATP and Mg 2+ molecules are shown by ball and stick models. c,d, Overall structure of apo state ABCD3, viewed from the membrane (c) and luminal (d) sides. Each protomer is colored green or gray. The peroxisome membrane is shown in gray. The TMDs, NBDs, and lateral gate are labeled. e, Side view of the density map, displayed using UCSF ChimeraX 1.6.1. The F o − F c map and the F o map are depicted by blue and green surfaces, respectively. The F o − F c map was calculated by Servalcat. f, The F o − F c map of the lateral gate, shown as a blue mesh. Residues around the density are indicated by stick models. Supplementary Fig. 11 | Density maps of ABCD3 in the ATP-bound outward-facing state, ATP-bound inward-facing state, and apo inward-facing state. a, Cryo-EM densities of TMs, loops, NBDs, ATP, and Mg 2+ molecules in the ATP-bound inward-facing state ABCD3. Contour levels are set at 1.9σ for TMs, loops, and NBDs, and at 3.0σ for ATP and Mg 2+ molecules. b, Cryo-EM densities of TMs, loops, NBDs, ATP, and Mg 2+ molecules in the ATP-bound outward-facing state ABCD3. Contour levels are set at 0.9σ for TMs, loops, and NBDs, and at 1.4σ for ATP and Mg 2+ molecules. c, Cryo-EM densities of TMs, loops, and NBDs in the apo-state ABCD3. Supplementary Fig. 12 | Cryo-EM analysis of ABCD3 in the apo inward-facing state. a, Flowchart of cryo-EM data processing for apo-state ABCD3, using cryoSPARC v3.3 and RELION 4.0. b, Size exclusion chromatography (SEC) profile of purified ABCD3. c, SDS-PAGE analysis of the eluted peak fraction from SEC. d, Representative 2D class averages. e, Representative cryo-EM micrograph. f, Gold-standard Fourier shell correlation (FSC) curve of the final map, reconstructed using cryoSPARC v3.3. Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7304961","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":502111604,"identity":"bd1bcd9b-76f3-4179-90d3-c6376477ccb2","order_by":0,"name":"Osamu Nureki","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBElEQVRIiWNgGAWjYJACxgYDmwQ4TwJKMxPQkgbTYkCsFobDmFpwAnOJ3IcPZxScz5NvYH/AzMPwR05yRvIBhh81DOzmOLRYzkg3NtxgcLvY4ACPAVCLgbG0RFoCY88xBmbLBuxaDG6ksUk+MLiduIGBh4GZ959B4jyJHAMG3gYGZoMDOLWw/3xgcC5xPsRhEC2Mf/FrYWPcYHAgseEAA9hhibOBWpjx2nLmGbPkDIPkYoPDPAYH5zAYG0v2PEs4LHNMArdfjqcxfuz5Y5cn397+8MEbBjk5iePJBx++qbFJxhViCACMOohLBBJADIlkA4Ja4IAfotOOBC2jYBSMglEwvAEAunlRe60Pqs0AAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0003-1813-7008","institution":"University of Tokyo","correspondingAuthor":true,"prefix":"","firstName":"Osamu","middleName":"","lastName":"Nureki","suffix":""},{"id":502111605,"identity":"d497e7c2-9675-4507-8fe5-7f35a2337611","order_by":1,"name":"Shinichiro Aiba","email":"","orcid":"","institution":"University of Tokyo","correspondingAuthor":false,"prefix":"","firstName":"Shinichiro","middleName":"","lastName":"Aiba","suffix":""},{"id":502111606,"identity":"19c7a2e7-6bdb-47e2-b691-c8e11785c97e","order_by":2,"name":"Hiroyuki Okamoto","email":"","orcid":"","institution":"University of Tokyo","correspondingAuthor":false,"prefix":"","firstName":"Hiroyuki","middleName":"","lastName":"Okamoto","suffix":""},{"id":502111607,"identity":"67bc63a9-e6e6-477b-96b4-7eafee4bb650","order_by":3,"name":"Atsuhiro Tomita","email":"","orcid":"https://orcid.org/0000-0003-3887-1429","institution":"Preferred Networks, Inc.","correspondingAuthor":false,"prefix":"","firstName":"Atsuhiro","middleName":"","lastName":"Tomita","suffix":""},{"id":502111608,"identity":"12f51543-d1eb-4916-ac97-cf8a0e03d2cf","order_by":4,"name":"Fumiya Sano","email":"","orcid":"https://orcid.org/0000-0002-8965-788X","institution":"The University of Tokyo","correspondingAuthor":false,"prefix":"","firstName":"Fumiya","middleName":"","lastName":"Sano","suffix":""},{"id":502111609,"identity":"3fb4bc42-7bb6-42a2-9c74-6ed7778f2955","order_by":5,"name":"Tsukasa Kusakizako","email":"","orcid":"https://orcid.org/0000-0002-6186-6647","institution":"The University of Tokyo","correspondingAuthor":false,"prefix":"","firstName":"Tsukasa","middleName":"","lastName":"Kusakizako","suffix":""}],"badges":[],"createdAt":"2025-08-06 02:50:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7304961/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7304961/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":95525695,"identity":"ea3443b4-a891-447f-9bf4-74e3534203c8","added_by":"auto","created_at":"2025-11-10 10:05:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6969412,"visible":true,"origin":"","legend":"","description":"","filename":"0804main.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/829d617ae2be954a3565b3c8.pdf"},{"id":95525717,"identity":"0f180ca7-8cd1-4fee-a134-7d910c332f90","added_by":"auto","created_at":"2025-11-10 10:05:36","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":113690,"visible":true,"origin":"","legend":"","description":"","filename":"ABCD3MS250805.docx","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/620c476363a620cc0cb08dad.docx"},{"id":95526302,"identity":"20ca3b9f-ab0d-4334-9af3-6bed854f8ae9","added_by":"auto","created_at":"2025-11-10 10:06:45","extension":"json","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7338,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS2561519.json","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/cc9b5e4f07621f7b3179054a.json"},{"id":95377154,"identity":"d5eba62d-f810-4799-ba5a-c1b4e12bd894","added_by":"auto","created_at":"2025-11-07 11:02:24","extension":"pdf","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6969412,"visible":true,"origin":"","legend":"","description":"","filename":"0804main.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/1c0c029e99c31023054682fe.pdf"},{"id":95377155,"identity":"836575db-4a45-4d24-9080-ba3e074887e0","added_by":"auto","created_at":"2025-11-07 11:02:24","extension":"xml","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":132842,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25615190structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/a84a5537e170606fe3fce8e1.xml"},{"id":95377146,"identity":"c082f810-ccf5-4763-92c8-bc869850c851","added_by":"auto","created_at":"2025-11-07 11:02:24","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1697712,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eC20:5-CoA recognition mechanism of ABCD3.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e,\u003cstrong\u003eb\u003c/strong\u003e, Overall structure of the C20:5-CoA-bound state of ABCD3, viewed from the membrane (\u003cstrong\u003ea\u003c/strong\u003e) and luminal (\u003cstrong\u003eb\u003c/strong\u003e) sides. Each protomer is colored light blue or gray. C20:5-CoA molecules are shown by stick models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e,\u003cstrong\u003ed\u003c/strong\u003e, Recognition of CoA moiety of acyl-CoA by ABCD3 (\u003cstrong\u003ec\u003c/strong\u003e) and ABCD1 (\u003cstrong\u003ed\u003c/strong\u003e). The acyl-CoA molecules and residues involved in acyl-CoA recognition are shown by stick models. Interactions between acyl-CoA and ABCD3 within 3.5 Å are shown by dotted lines.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Recognition of the acyl-chain moiety of C20:5-CoA by ABCD3. The acyl-CoA molecules and residues involved in acyl-CoA recognition are shown by stick models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, Cutaway view of the molecular surface of C22:0-CoA-bound ABCD1 (PDB ID: 7VZB), sliced along the hydrophobic pocket that accommodates the acyl chain. The surface is colored based on electrostatic potential, with negatively charged regions shown in red and positively charged regions in blue.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg\u003c/strong\u003e, Saturated acyl chain binding pockets in ABCD1, viewed from the luminal side. The residues forming the binding pockets are shown by stick and semi-transparent CPK models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh\u003c/strong\u003e, Cutaway view of the molecular surface of C20:5-CoA-bound ABCD3, superimposed with the C22:0-CoA-bound ABCD1 and sliced along the same plane as in panel (\u003cstrong\u003ef\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ei\u003c/strong\u003e, View of ABCD3 superimposed with ABCD1, shown from the same viewpoint as in panel (\u003cstrong\u003eg\u003c/strong\u003e). Residues in ABCD3 that occupy the space where the hydrophobic pocket is observed in ABCD1 are shown by stick and semi-transparent CPK models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ej\u003c/strong\u003e, Comparison of the acyl chain binding sites between ABCD3 (light blue) and ABCD1 (magenta). R362 (R389 is the corresponding residue in ABCD1) and E111 in ABCD3 are important for the structural changes.\u003c/p\u003e","description":"","filename":"0804main1.png","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/b8104485139db58d0ae9cf9c.png"},{"id":95526687,"identity":"42ddc9c9-44bf-4d47-ab29-696e5f7f4113","added_by":"auto","created_at":"2025-11-10 10:07:35","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1073552,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStructure of CHS-bound ABCD3.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e,\u003cstrong\u003eb\u003c/strong\u003e, Overall structure of CHS-bound ABCD3, viewed from the membrane (\u003cstrong\u003ea\u003c/strong\u003e) and luminal (\u003cstrong\u003eb\u003c/strong\u003e) sides. Each protomer is colored orange or gray. CHS molecules are shown by stick models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, Close-up view of the CHS recognition site. The CHS molecule and surrounding residues are shown by stick models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Apo ABCD1 (PDB ID: 7VWC), superimposed onto CHS-bound ABCD3, shown from the same viewpoint as the CHS-binding site. The lipid and its surrounding residues are displayed as sticks.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e,\u003cstrong\u003ef\u003c/strong\u003e,\u003cstrong\u003e \u003c/strong\u003eRMSD plots of CHS (\u003cstrong\u003ee\u003c/strong\u003e) and THCA (\u003cstrong\u003ef\u003c/strong\u003e) molecules from MD simulations. RMSD values were calculated for each frame after structural superimposition of the protein moiety.\u003c/p\u003e","description":"","filename":"0804main2.png","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/dfa24c095d83266c981ed26b.png"},{"id":95377148,"identity":"510fe63e-5815-499b-8f6e-15696882cd0f","added_by":"auto","created_at":"2025-11-07 11:02:24","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1108619,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eATP-bound inward-facing state of ABCD3.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e,\u003cstrong\u003eb\u003c/strong\u003e, Overall structure of ATP-bound inward-facing ABCD3, viewed from the membrane (\u003cstrong\u003ea\u003c/strong\u003e) and luminal (\u003cstrong\u003eb\u003c/strong\u003e) sides. Each protomer is colored yellow or gray. ATP and Mg\u003csup\u003e2+\u003c/sup\u003e molecules are shown by ball and stick models. Interactions between ATP and ABCD3 within 3.5 Å are shown by dotted lines.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, Substrate ATP binding site in the TMD. The ATP molecule and surrounding residues are shown as stick models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Overlay of the \u003cem\u003eF\u003c/em\u003e\u003csub\u003eo\u003c/sub\u003e−\u003cem\u003eF\u003c/em\u003e\u003csub\u003ec\u003c/sub\u003e omit map and ATP molecules. The \u003cem\u003eF\u003c/em\u003e\u003csub\u003eo\u003c/sub\u003e−\u003cem\u003eF\u003c/em\u003e\u003csub\u003ec\u003c/sub\u003e omit map calculated by Servalcat without ATP molecules is shown as a green mesh, contoured at 2.8σ. The ATP molecules surrounding the residues are displayed as sticks.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Overlay of the \u003cem\u003eF\u003c/em\u003e\u003csub\u003eo\u003c/sub\u003e−\u003cem\u003eF\u003c/em\u003e\u003csub\u003ec\u003c/sub\u003e omit map and ATP molecules. The ATP molecules were modeled based on the density map, refined using particles obtained from 3D classification with C1 symmetry. The \u003cem\u003eF\u003c/em\u003e\u003csub\u003eo\u003c/sub\u003e−\u003cem\u003eF\u003c/em\u003e\u003csub\u003ec\u003c/sub\u003e omit map calculated by Servalcat without ATP molecules is shown as a green mesh, contoured at 1.8σ.\u003c/p\u003e","description":"","filename":"0804main3.png","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/829efac150693747c9932952.png"},{"id":95377152,"identity":"e56b98d2-0f02-457e-acfc-36797ae8e3a3","added_by":"auto","created_at":"2025-11-07 11:02:24","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1815507,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConformational changes of ABCD3 during the transport cycle.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Structural superimpositions of the apo and 20:5-CoA-bound states performed by least squares fitting, using one TM bundle (TM1, TM2, TM3, TM6, and TM4’, TM5’) as the fitting reference. The structures are shown from the membrane and luminal sides.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, Structural superimpositions of the 20:5-CoA-bound and ATP-bound OF states performed by least squares fitting, using one TM bundle (TM1, TM2, TM3, TM6, and TM4’, TM5’) as the fitting reference. The structures are shown from the membrane and luminal sides.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, Structural rearrangements of R96 and K209 upon CoA binding.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Conformational changes of the TMD upon CoA binding. The CoA molecule and surrounding residues are shown as sticks.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Conformational changes of the TMD upon NBD dimerization. The CoA molecule and residues that undergo notable conformational changes are shown as sticks.\u003c/p\u003e","description":"","filename":"0804main4.png","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/540205d9779bb979ad41c652.png"},{"id":95526335,"identity":"98809843-3e59-4b0d-9555-2e4c38e9945a","added_by":"auto","created_at":"2025-11-10 10:06:48","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":68081,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSubstrate recognition and transport mechanism of ABCD3.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eModel of the substrate recognition mechanism by ABCD3. For acyl-CoA, the CoA moiety is rigidly recognized at substrate-binding site 1, while the acyl chain is loosely accommodated at substrate-binding site 2, enabling ABCD3 to transport a variety of acyl-CoAs. For bile acid precursors, the cholesteryl backbone is recognized by the hydrophobic substrate-binding site 2. Substrate ATP molecules are recognized through interactions between their triphosphate groups and hydrophilic residues at substrate-binding site 1.\u003c/p\u003e","description":"","filename":"0804main5.png","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/3228e3d711b183004538e1d2.png"},{"id":95531051,"identity":"896a4cac-c3c2-4245-9ca2-abbf8b226aa1","added_by":"auto","created_at":"2025-11-10 10:22:32","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6792253,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/7b4c5489-dd46-432e-8300-862c3fc433fc.pdf"},{"id":95377157,"identity":"b591cbdf-c193-455a-83c7-2c70516f3c78","added_by":"auto","created_at":"2025-11-07 11:02:24","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":23509283,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Fig. 1 | Sequence alignment of human ABC transporters belonging to subfamily D.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMultiple sequence alignment for human ABC transporters belonging to subfamily D. The alignment was performed by MSAProbs v0.9.7\u003csup\u003e39\u003c/sup\u003e and the figure was prepared with ESPrint3\u003csup\u003e40\u003c/sup\u003e. The secondary structures based on the cryo-EM structure of human ABCD3 are shown above the sequences.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 2 | Cryo-EM analysis of ABCD3 in the C20:5-CoA-bound inward-facing state.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Flowchart of cryo-EM data processing for apo-state ABCD3, using cryoSPARC v4.2 and RELION 4.0.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, SEC profile of purified ABCD3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, SDS-PAGE analysis of the eluted peak fraction from SEC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Representative 2D class averages.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Representative cryo-EM micrograph.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, FSC curve of the final map, reconstructed using cryoSPARC v4.6.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg\u003c/strong\u003e, Structural formula of C20:5-CoA.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 3 | Density maps of ABCD3 in the C20:5-CoA-bound inward-facing state and the CHS-bound inward-facing state.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Cryo-EM densities of TMs, loops, NBDs, and CoA molecules with the corresponding structure of C20:5-CoA-bound ABCD3. Contour levels are set at 2.1σ.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, Cryo-EM densities of TMs, Loops, NBDs, and CHS molecules with the corresponding structure of CHS-bound ABCD3. Contour levels are set at 2.2σ.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 4 | Distances between NBDs.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDistances between NBDs are compared based on the distances between Cα atoms of Y282. Apo state, green; 20:5-CoA bound state, light blue; CHS-bound state, brown; ATP-bound-IF-state, yellow; and ATP-bound-OF-state, red.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 5 | Structural comparisons between ABCD1 and ABCD3.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea-c\u003c/strong\u003e, Structural superimpositions of ABCD1 and ABCD3, using secondary structure matching by CueMol.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Apo state (ABCD1 (PDB ID: 7VWC) yellow-green, ABCD3 green).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, Acyl-CoA-bound state (ABCD1 (PDB ID: 7VZB) purple, ABCD3 light blue).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, ATP-bound-OF-state (ABCD1 (PDB ID: 7RR9) brown, ABCD3 red).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 6 | Cryo-EM analysis of ABCD3 in the CHS-bound inward-facing state.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Flowchart of cryo-EM data processing for apo-state ABCD3, using cryoSPARC v3.1 and RELION 4.0.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, SEC profile of purified ABCD3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, SDS-PAGE analysis of the eluted peak fraction from SEC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Representative 2D class averages.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Representative cryo-EM micrograph.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, FSC curve of the final map, reconstructed using cryoSPARC v3.3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg\u003c/strong\u003e, Chemical formulas of cholesteryl hemisuccinate (CHS) and 3α,7α,12α-trihydroxycholestanoic acid (THCA).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 7 | MD simulations.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe results of the three independent runs are shown. All plots show the moving average as a solid line and the raw data as a semi-transparent line.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, All-atom RMSD plot from the MD simulation of the CHS-bound state.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, All-atom RMSD plot from the MD simulation of the THCA-bound state.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, Plot of the distance between the centers of mass of CHS and M233.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Plot of the distance between the centers of mass of CHS and L365.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Plot of the distance between the centers of mass of THCA and M233.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, Plot of the distance between the centers of mass of THCA and L365.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eg\u003c/strong\u003e, Visualization of CHS atoms, colored according to RMSF values calculated from MD simulations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh\u003c/strong\u003e, Visualization of THCA atoms, colored according to RMSF values calculated from MD simulations.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 8 | Cryo-EM analysis of ABCD3 in the ATP-bound inward-facing state.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Flowchart of cryo-EM data processing for apo-state ABCD3, using cryoSPARC v3.3 and RELION 4.0.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, SEC profile of purified ABCD3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, SDS-PAGE analysis of eluted fractions from SEC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Representative 2D class averages.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Representative cryo-EM micrograph.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e,\u003cstrong\u003eg\u003c/strong\u003e, FSC curves of the final C2 (\u003cstrong\u003ef\u003c/strong\u003e) and C1 (\u003cstrong\u003eg\u003c/strong\u003e) maps, reconstructed using cryoSPARC v4.4.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eh\u003c/strong\u003e, Structural formulas of cholic acid and ATP.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 9 | Cryo-EM analysis of ABCD3 in the ATP-bound outward-facing state.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Flowchart of cryo-EM data processing for apo-state ABCD3 using RELION 3.1 and RELION 4.0.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, SEC profile of purified ABCD3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, SDS-PAGE analysis of the eluted peak fraction from SEC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Representative 2D class averages.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Representative cryo-EM micrograph.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, FSC curve of the final map, reconstructed using RELION 4.0.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 10 | ATP-bound inward-facing state and apo inward-facing state of ABCD3.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e,\u003cstrong\u003eb\u003c/strong\u003e, Overall structure of ATP-bound outward-facing state ABCD3, viewed from the membrane (\u003cstrong\u003ea\u003c/strong\u003e) and luminal (\u003cstrong\u003eb\u003c/strong\u003e) sides. Each protomer is colored red or gray. ATP and Mg\u003csup\u003e2+\u003c/sup\u003e molecules are shown by ball and stick models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e,\u003cstrong\u003ed\u003c/strong\u003e, Overall structure of apo state ABCD3, viewed from the membrane (\u003cstrong\u003ec\u003c/strong\u003e) and luminal (\u003cstrong\u003ed\u003c/strong\u003e) sides. Each protomer is colored green or gray. The peroxisome membrane is shown in gray. The TMDs, NBDs, and lateral gate are labeled.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e,\u003cstrong\u003e \u003c/strong\u003eSide view of the density map, displayed using UCSF ChimeraX 1.6.1. The \u003cem\u003eF\u003c/em\u003e\u003csub\u003eo\u003c/sub\u003e−\u003cem\u003eF\u003c/em\u003e\u003csub\u003ec\u003c/sub\u003e map and the \u003cem\u003eF\u003c/em\u003e\u003csub\u003eo\u003c/sub\u003e map are depicted by blue and green surfaces, respectively. The \u003cem\u003eF\u003c/em\u003e\u003csub\u003eo\u003c/sub\u003e−\u003cem\u003eF\u003c/em\u003e\u003csub\u003ec\u003c/sub\u003e map was calculated by Servalcat.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, The \u003cem\u003eF\u003c/em\u003e\u003csub\u003eo\u003c/sub\u003e−\u003cem\u003eF\u003c/em\u003e\u003csub\u003ec\u003c/sub\u003e map of the lateral gate, shown as a blue mesh. Residues around the density are indicated by stick models.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 11 | Density maps of ABCD3 in the ATP-bound outward-facing state, ATP-bound inward-facing state, and apo inward-facing state.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Cryo-EM densities of TMs, loops, NBDs, ATP, and Mg\u003csup\u003e2+\u003c/sup\u003e molecules in the ATP-bound inward-facing state ABCD3. Contour levels are set at 1.9σ for TMs, loops, and NBDs, and at 3.0σ for ATP and Mg\u003csup\u003e2+\u003c/sup\u003e molecules.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, Cryo-EM densities of TMs, loops, NBDs, ATP, and Mg\u003csup\u003e2+\u003c/sup\u003e molecules in the ATP-bound outward-facing state ABCD3. Contour levels are set at 0.9σ for TMs, loops, and NBDs, and at 1.4σ for ATP and Mg\u003csup\u003e2+\u003c/sup\u003e molecules.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, Cryo-EM densities of TMs, loops, and NBDs in the apo-state ABCD3.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eSupplementary Fig. 12 | Cryo-EM analysis of ABCD3 in the apo inward-facing state.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e, Flowchart of cryo-EM data processing for apo-state ABCD3, using cryoSPARC v3.3 and RELION 4.0.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eb\u003c/strong\u003e, Size exclusion chromatography (SEC) profile of purified ABCD3.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ec\u003c/strong\u003e, SDS-PAGE analysis of the eluted peak fraction from SEC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ed\u003c/strong\u003e, Representative 2D class averages.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ee\u003c/strong\u003e, Representative cryo-EM micrograph.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ef\u003c/strong\u003e, Gold-standard Fourier shell correlation (FSC) curve of the final map, reconstructed using cryoSPARC v3.3.\u003c/p\u003e","description":"","filename":"0804sup.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7304961/v1/c2ed39845c69208f76d42f0a.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Structural Insights into the Broad Substrate Recognition Mechanism of Human ABCD3","fulltext":[{"header":"Introduction","content":"\u003cp\u003eFatty acids are used in diverse biological processes, such as lipid membrane formation, energy storage, and post-translational modifications\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Intracellular organelles, including mitochondria, the endoplasmic reticulum, and peroxisomes, play important roles in fatty acid metabolism\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. In mitochondria and peroxisomes, fatty acids are mainly processed by β-oxidation\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Mitochondrial β-oxidation is involved in energy production. In contrast, peroxisomes are not directly involved in energy production, and instead degrade a wide range of fatty acids, including branched fatty acids and very long-chain fatty acids\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, since they cannot be broken down in mitochondria.\u003c/p\u003e\u003cp\u003eAs fatty acids have low membrane permeability, ABC subfamily D transporters (ABCD1, ABCD2, and ABCD3) are responsible for their uptake into peroxisomes\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Fatty acid transport is accomplished in the form of an acyl moiety bound to CoA, and ABCD1–3 transport different types of acyl-CoAs\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. ABCD1–3 transport long-chain fatty acyl-CoAs such as C16:0-CoA, while very long-chain fatty acyl-CoAs such as C22:0-, C24:0-, and C26:0-CoA are only transported by ABCD1 and ABCD2\u003csup\u003e6\u003c/sup\u003e. In contrast, unsaturated fatty acyl-CoAs are transported by ABCD2 and ABCD3, and not by ABCD1\u003csup\u003e6\u003c/sup\u003e. Dicarboxylic acyl-CoAs, branched fatty acyl-CoAs, bile acid precursors, and acylcarnitines are exclusively transported by ABCD3\u003csup\u003e6,7\u003c/sup\u003e. Thus, ABCD1 and ABCD2 are responsible for very long, straight-chain fatty acid transport, whereas ABCD3 transports a wide range of acyl-CoA molecules that are shorter but bulkier and more hydrophilic. Consequently, the phenotypes that emerge from each dysfunction are different: ABCD1 is associated with adrenoleukodystrophy\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, while ABCD3 dysfunction is associated with hepatosplenomegaly\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Despite the functional differences of ABCD1–3, structural information is only available for ABCD1\u003csup\u003e10–14\u003c/sup\u003e and the apo state of ABCD3\u003csup\u003e15\u003c/sup\u003e. Previous studies reported that the acyl chains are recognized by hydrophobic pockets in ABCD1\u003csup\u003e10\u003c/sup\u003e, which shares 36.7% sequence identity with ABCD3 (Supplementary Fig.\u0026nbsp;1). However, the mechanism by which ABCD3 recognizes and transports bulkier, unsaturated long-chain acyl-CoA molecules, which are not accommodated by ABCD1, remains elusive.\u003c/p\u003e\u003cp\u003eIn addition to acyl-CoA molecules, ABCD3 also transports bile acid precursors\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e and ATP molecules\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e into peroxisomes. ATP molecules in peroxisomes are essential for the β-oxidation of fatty acids\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, and peroxisomes also biosynthesize bile acids, playing a crucial role in cholesterol homeostasis\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. However, the structural basis for the recognition and transport of such structurally distinct substrates, such as acyl-CoAs, bile acid precursors, and ATP, by ABCD transporters remains poorly understood.\u003c/p\u003e\u003cp\u003eIn this study, we present cryo-electron microscopy structures of human ABCD3 in distinct transporting states: the apo, unsaturated long-chain acyl-CoA (C20:5 CoA)-bound, CHS-bound, and substrate ATP-bound inward-facing, and ATP-bound outward-facing states. The C20:5 CoA-bound structure reveals details about the acyl-CoA substrate preference of the ABCD3 transporter. Moreover, the bile acid precursor-like CHS-bound and substrate ATP-bound structures exhibit different substrate-binding modes. These ATP-bound, inward-facing and outward-facing structures represent the conformational changes in ABCD3 during the substrate transport cycle. Collectively, our study provides structural insights into the substrate preferences and transport mechanisms of ABCD3.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eUnsaturated long-chain acyl-CoA recognition mechanism\u003c/h2\u003e\u003cp\u003eTo elucidate the structural basis for the unsaturated long-chain acyl-CoA recognition by ABCD3, we determined the cryo-EM structure of C20:5-CoA-bound ABCD3 with C2 symmetry at an overall resolution of 3.3 \u0026Aring; (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea,b, and Supplementary Figs.\u0026nbsp;2 and 3a). The structure reveals the typical ABC half-transporter architecture, with each subunit comprising a transmembrane domain (TMD) and a nucleotide-binding domain (NBD) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea,b). The TMD consists of six TM helices, and the inter-subunit internal cavity is largely exposed to the intracellular and lateral sides (the inner leaflet of the membrane) with 41.9 \u0026Aring; between the two corresponding Y282 residues in the TMDs, similar to the C22:0-CoA-bound ABCD1 structure\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e (PDB ID: 7VZB) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea,b, and Supplementary Figs.\u0026nbsp;4 and 5). The structure revealed that the C20:5-CoA molecule is located within the cavity between TM bundles, with the CoA moiety of C20:5-CoA forming polar interactions with hydrophilic residues (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec), in a similar manner to the C22:0-CoA-bound ABCD1 structure\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). Specifically, the adenine ring forms hydrogen bonds with S205 and K209, and a stacking interaction with R373 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). The 5\u0026prime;-diphosphate and 3\u0026prime;-phosphate groups of the CoA moiety interact the with conserved hydrophilic residues R96, N140, K322\u0026rsquo; (apostrophe indicates residue from the opposite protomer) and Y323\u0026rsquo; and with K144, respectively (this site is referred to as substrate-binding site 1) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec). Both K322\u0026rsquo; and Y323\u0026rsquo; are located on the TM5\u0026rsquo; π-helix, which is observed only in ABCD1 and ABCD3, but not in ABCD4\u003csup\u003e17\u003c/sup\u003e, a member of the same subfamily. This unique structural feature allows K322\u0026rsquo; and Y323\u0026rsquo; to come into the proximity of the acyl-CoA substrate (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec,d).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe acyl moiety of C20:5-CoA is accommodated in the large central cavity between TM bundles (this region is referred to as substrate-binding site 2), which comprises hydrophobic residues such as Y330\u0026rsquo;, V333\u0026rsquo;, and V365\u0026rsquo; (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee). The unsaturated, long acyl chain adopts a bent conformation, with the terminal region poorly resolved in the density map, suggesting its flexibility. This recognition mode is significantly distinct from that of ABCD1 bound to C22:0-CoA (PDB ID: 7VZB), in which the acyl moiety of C22:0-CoA is recognized by the hydrophobic pocket formed by TMs 3\u0026rsquo;, 4\u0026rsquo;, 5\u0026rsquo;, and 6\u0026rsquo;\u003csup\u003e10\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ef,g). In our C20:5-CoA-bound and apo inward-facing ABCD3 structure, the hydrophobic pocket observed in ABCD1 collapses due to the proximity of TMs 3\u0026rsquo; and 4\u0026rsquo; (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eh,i), which block substrate-binding\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. In addition, the luminal loop 2 (LL2) between TMs 3\u0026rsquo; and 4\u0026rsquo; is shorter in ABCD3 than in ABCD1 (Supplementary Fig.\u0026nbsp;1), whose LL2 protrudes further toward the luminal side compared with that of ABCD3 (Supplementary Fig.\u0026nbsp;5a,b). This structural difference induces closer packing of TM3\u0026rsquo; and TM4\u0026rsquo; in ABCD3, thus eliminating the hydrophobic pocket. In particular, the M233\u0026rsquo; residue on TM4\u0026rsquo; in ABCD3, which is not conserved in ABCD1 or ABCD2 (Supplementary Fig.\u0026nbsp;1), is located at the hydrophobic pocket and blocks substrate entry (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ei).\u003c/p\u003e\u003cp\u003eOn the surface at the peroxisomal end of substrate-binding site 2, R362 in ABCD3 forms a salt bridge with E111 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ee), which is not conserved among the ABCD family members\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e (Supplementary Fig.\u0026nbsp;1), and adopts a distinct conformation from the corresponding R389 residue in ABCD1, which forms the gate towards the peroxisomal side\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ej). A superimposition of the acyl-CoA-bound inward-facing ABCD3 (C20:5) and ABCD1 (C22:0) structures reveals that the acyl moiety of C20:5-CoA bound to ABCD3 sterically clashes with R389 in ABCD1, suggesting that the distinct conformation of R362 in ABCD3 allows the accommodation of the acyl moiety (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ej). Together, these findings reveal the unique mechanism of unsaturated long-chain acyl-CoA recognition by ABCD3, which differs substantially from that of saturated long-chain acyl-CoA recognition by ABCD1.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eBile acid precursor recognition mechanism\u003c/h3\u003e\n\u003cp\u003eIn addition to unsaturated long-chain acyl-CoAs, ABCD3 transports the bile acid precursors 3α,7α-dihydroxycholestanoic acid (DHCA) and 3α,7α,12α-trihydroxycholestanoic acid (THCA)\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e, which share structural similarity with cholesteryl hemisuccinate (CHS) (Supplementary Fig.\u0026nbsp;6), to biosynthesize bile acids in peroxisomes\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. To understand how ABCD3 recognizes and transports bile acid precursors, which are structurally quite different from acyl-CoA molecules, we determined the cryo-EM structure of ABCD3 in the CHS-bound state at 3.8 \u0026Aring; resolution (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea,b, and Supplementary Figs.\u0026nbsp;3b and 6). The EM density map clearly shows that the CHS molecule binds to substrate-binding site 2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec), where the thioester bond and the acyl chain moiety are accommodated in the C20:5-CoA-bound state. The cholesteryl moiety of CHS is recognized by the hydrophobic surface of substrate-binding site 2, while its hemisuccinate moiety interacts with Y330 and R362 at the peroxisomal end of the site (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). M233 on TM4, which constitutes the hydrophobic surface of substrate-binding site 2, protrudes into the lateral gate formed by TMs 3\u0026ndash;6, thereby narrowing it and forming the CHS-binding pocket (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec,d and Supplementary Fig.\u0026nbsp;5a). By contrast, the CHS-unbound ABCD1 structure exhibits a wider lateral gate compared with ABCD3, with TM4 shifted away, thereby failing to form the CHS-binding pocket observed in the present CHS-bound ABCD3\u003csup\u003e10,11,13,14\u003c/sup\u003e. These structural observations suggest that the binding mode of CHS at substrate-binding site 2 in ABCD3 mimics that of physiological substrate bile acid precursors, highlighting the key role of ABCD3 in bile acid precursor transport\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo further investigate the bile acid precursor recognition mechanism, we performed MD simulations of CHS-bound and THCA-bound ABCD3, in which the THCA molecule was modeled at substrate-binding site 2, based on our CHS-bound inward-facing ABCD3 structure. Our MD simulation results showed that the cholesteryl backbones of both CHS and THCA remained stable at substrate-binding site 2 throughout the 100-ns simulations (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ee,f and Supplementary Fig.\u0026nbsp;7), supporting the notion that bile acid precursors such as THCA bind to this site similarly to CHS. By contrast, the terminus of the hemisuccinate moiety of CHS and the side chains of both CHS and THCA exhibited higher flexibility (Supplementary Fig.\u0026nbsp;7g,h). Together, these results suggest that ABCD3 recognizes substrates promiscuously by a hydrophobic surface at substrate-binding site 2, which interacts with both acyl chains and steroid moieties.\u003c/p\u003e\n\u003ch3\u003eATP recognition in the substrate-binding site\u003c/h3\u003e\n\u003cp\u003eTo clarify the transport mechanism by ABCD3, we purified the catalytically inactive mutant, in which the catalytic E596 residue in the NBD is substituted by a glutamine residue (ABCD3\u003csub\u003eEQ\u003c/sub\u003e), to capture the ATP-bound structures. In the presence of ATP-Mg\u003csup\u003e2+\u003c/sup\u003e, we determined two distinct ATP-bound ABCD3\u003csub\u003eEQ\u003c/sub\u003e structures in the presence (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea,b and Supplementary Fig.\u0026nbsp;8) or absence (Supplementary Figs.\u0026nbsp;9 and 10a,b) of cholic acid, which shares structural similarity with transport substrates. The ATP-bound structure in the presence of cholic acid adopts the inward-facing conformation similar to the CHS-bound state, with the distance between NBDs almost unchanged among the structures (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea,b and Supplementary Fig.\u0026nbsp;4), whereas the ATP-bound structure in the absence of cholic acid adopts the outward-facing conformation (described in the next section).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe determined the structure of ABCD3\u003csub\u003eEQ\u003c/sub\u003e in the ATP-bound inward-facing state, in the presence of cholic acid, with C2 symmetry at 2.9 \u0026Aring; resolution (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea,b, and Supplementary Figs.\u0026nbsp;8 and 11a). Although we observed undefined density at the lateral gate, we could not determine whether this originates from cholic acid due to the poor resolution. The structure revealed that ATP binds to the Walker motif in the NBDs, in which the distance between the two corresponding Y282 residues is 59.1 \u0026Aring; (Supplementary Fig.\u0026nbsp;4), indicating that this structure represents a conformational state just before NBD dimerization upon ATP binding. Notably, in addition to the ATPs bound to the NBDs, an ATP molecule is located at substrate-binding site 1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea,c,d), suggesting ATP transport by ABCD3, consistent with the previous study\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. The adenosine and triphosphate moieties of ATP are recognized by Q369 and R373, and by the positively charged residues R96, K209, R373, and K322\u0026rsquo;, which also interact with the CoA moiety of acyl-CoA in the C20:5-CoA bound state, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). This structural observation indicates that the binding mode of ATP in substrate-binding site 1 differs significantly from that of the ADP moiety of C20:5-CoA (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). The 3D classification with C1 symmetry revealed that the two substrate ATP molecules bind asymmetrically and form a hydrogen bond with each other (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ee), likely representing an intermediate state during ATP transport. In both the C2 and C1 maps, the triphosphate densities at substrate-binding site 1 are well resolved, whereas the adenosine densities are poorly defined, suggesting their comparatively weaker recognition and greater flexibility. Together, the substrate ATP-bound state structure provides the first view of this new substrate recognition mode by the ABC family transporters.\u003c/p\u003e\n\u003ch3\u003eConformational change of ABCD3 during the transport cycle\u003c/h3\u003e\n\u003cp\u003eWe sought to obtain the cryo-EM map of the branched acyl-CoA-bound ABCD3, and acquired cryo-EM images of ABCD3 under phytanoyl-CoA-containing conditions (Supplementary Fig.\u0026nbsp;12). However, the reconstituted 3D map at 3.3 \u0026Aring; resolution lacked clear density corresponding to the phytanoyl-CoA molecule (Supplementary Figs.\u0026nbsp;10e and 12). Thus, we refer to this structure as an apo state.\u003c/p\u003e\u003cp\u003eThe apo state structure adopts the inward-facing conformation with a 59.1 \u0026Aring; distance between the two corresponding Y282 residues in the TMDs (Supplementary Fig.\u0026nbsp;4), thereby highlighting the conformational change upon acyl-CoA binding (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea, and Supplementary Fig.\u0026nbsp;4). Previous studies\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e suggested that the interactions between the two C-terminal helices may facilitate the dimerization of NBDs in apo ABCD1. Interestingly, the apo inward-facing conformation of ABCD3 lacking the corresponding C-terminal helices is similar to the ABCD1 apo inward-facing conformation (PDB ID: 7VWC) (Supplementary Fig.\u0026nbsp;5a), indicating that the distance between the NBDs in the ABCD3 apo inward-facing conformation is stable without the C-terminal helices. Although this apo inward-conformation of ABCD3 resembles the apo inward-conformation of ABCD1, the lateral gate in ABCD3 is narrower than that in ABCD1\u003csup\u003e10\u003c/sup\u003e (Supplementary Fig.\u0026nbsp;5a). The cryo-EM map of the apo ABCD3 state shows several lipid-like densities on the protein surface. In both apo structures of ABCD1 and ABCD3, no density corresponding to transport substrates was observed, but lipid densities were present. In particular, we observed an acyl-CoA or phospholipid-like density at the narrow lateral gate in the inner leaflet (Supplementary Fig.\u0026nbsp;10e,f), as previously observed in the apo state structure of ABCD1\u003csup\u003e10,14\u003c/sup\u003e, although we could not model the specific molecule due to its low resolution. This density interacts with L237, V238, L242, L371, and V375 on the membrane side and with T245, R248, and R379 on the membrane surface, implying that the transport of amphiphilic molecules such as acyl-CoAs may be nonspecifically mediated by the lateral gate.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSuperimposition of one side of the TM bundle (TM1, TM2, TM3, TM6, and TM4\u0026prime;/TM5\u0026prime;) between the apo and C20:5-CoA-bound states shows good alignment, with an RMSD of 0.89 \u0026Aring; over 261 Cα atoms (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea). Among the residues involved in the CoA moiety recognition, R96 and K209 adopt distinct conformations between the apo and C20:5-CoA-bound states (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea,c). Upon CoA binding, R96, which interacts with D213 in the apo state, undergoes a large structural rearrangement and newly interacts with the 5\u0026prime;-diphosphate group of C20:5-CoA. This structural rearrangement of R96 creates a space that is occupied by K209, which then interacts with the adenine ring of C20:5-CoA, allowing it to bind to substrate-binding site 1. By contrast, a structural comparison of the opposite TM bundles between the apo and C20:5-CoA-bound states reveals a conformational change, particularly attributed to electrostatic interactions between K144 and the 3\u0026prime;-phosphate group of C20:5-CoA (Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea,d). These structural observations indicate the rigid body movement between TM bundles upon C20:5-CoA binding, thereby explaining the different NBD distances between the apo and C20:5-CoA inward-facing conformations.\u003c/p\u003e\u003cp\u003eWe determined the cryo-EM structure of ABCD3\u003csub\u003eEQ\u003c/sub\u003e in the ATP-bound outward-facing state at 3.2 \u0026Aring; resolution, in the absence of cholic acid (Supplementary Figs.\u0026nbsp;9a, 10a,b and 11b). The structure revealed that ATP binds to the Walker motif in the NBDs, in which the distance between Y282 is 30.6 \u0026Aring; (Supplementary Fig.\u0026nbsp;4). In the outward-facing structure, TM6 adopts a straight conformation, and the luminal loop 3 (LL3) between TMs 5 and 6 is ordered (Supplementary Figs.\u0026nbsp;10a and 11b). This arrangement reduces the exposure of the central cavity to the peroxisomal side, as compared to that of the ABCD1 outward-facing state (PDB ID: 7RR9), in which TM6 is bent with LL3 disordered (Supplementary Fig.\u0026nbsp;5c), thereby facilitating the release of substrates from the central cavity into the peroxisomes. A structural comparison between the inward- and outward-facing states reveals that the residues constituting substrate-binding sites 1 and 2 adopt distinct conformations (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb,e), which probably reduce the substrate affinity in the outward-facing state. Together, our ABCD3 structures in the apo, CHS-, substrate ATP-, and C20:5-CoA-bound inward-facing, and ATP-bound outward-facing states reveal the conformational transitions of ABCD3 during the transport cycle (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), suggesting the substrate transport mechanism of ABCD3.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we determined different substrate-bound and multiple conformations of ABCD3 during the transport cycle (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). In the inward-facing states, the transport substrates are mediated by the lateral gate and/or from the intracellular side, and hydrophilic and subsequently hydrophobic moieties of substrates are recognized at the substrate-binding sites 1 and 2, respectively. The conformational transition from the inward-facing state to the outward-facing state upon ATP-binding in the NBDs induces the substrate release into the peroxisomes.\u003c/p\u003e\u003cp\u003eThe C20:5-CoA-bound structure provides insights into the substrate specificity of ABCD3 for acyl-CoA. ABCD1 recognizes saturated long-chain acyl-CoA molecules that fit into the hydrophobic cavity formed by TMs 3\u0026rsquo;\u0026ndash;6\u0026rsquo;, but does not transport dicarboxylic and bulky acyl-CoA molecules. By contrast, ABCD3 transports bulky and unsaturated acyl-CoA molecules via the large cavity within the TMD. In addition, ABCD3 is capable of transporting dicarboxylic acyl-CoA, possibly owing to the hydrophilic surface at the peroxisomal end of substrate-binding site 2, unlike the hydrophobic pocket in ABCD1. ABCD3 lacks the hydrophobic pocket observed in ABCD1, which accommodates extended saturated acyl chains, thereby explaining why ABCD1 preferentially transports such substrates over ABCD3\u003csup\u003e19\u003c/sup\u003e. Although the 3\u0026prime;-phosphate ADP moiety of CoA is structurally similar to ATP, their binding modes at substrate-binding 1 are completely different. This difference likely stems from the chemical distinction at the 3\u0026prime; position of the ribose: CoA contains a phosphate group, whereas ATP possesses a hydroxyl group. The 3\u0026prime;-phosphate group of CoA is recognized by K144, and this interaction induces a rigid-body shift of the TMD. By contrast, the 3\u0026prime;-hydroxyl group in ATP is insufficient to induce such a conformational change, resulting in the distance between the NBDs remaining unchanged from the apo state, and the substrate ATP failing to adopt the binding mode observed for the 3\u0026prime;-phosphate ADP moiety of CoA.\u003c/p\u003e\u003cp\u003eThe CHS-bound structure suggests that bile acid precursor molecules such as DHCA and THCA, substrates of ABCD3\u003csup\u003e7,9\u003c/sup\u003e, probably bind to substrate-binding site 2, similar to the CHS-binding mode. Our structural comparison between the CHS- and C20:5-CoA-bound states of ABCD3 revealed that the binding site of CHS overlaps with that of the thioester bond and acyl chain moiety of acyl-CoA at substrate-binding site 2. Given that ABCD3 transports DHCA and THCA in their CoA-conjugated forms\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, these findings suggest that the binding modes of DHCA-CoA and THCA-CoA may differ slightly from those of typical acyl-CoAs.\u003c/p\u003e\u003cp\u003eWe determined the substrate ATP-bound ABCD3 structure under conditions with cholic acid and an excess of ATP and Mg\u003csup\u003e2+\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea,b and Supplementary Fig.\u0026nbsp;8). We first expected that the cholic acid molecules would bind to the outward-facing ABCD3, as indeed the ATP-bound ABCD3 structure determined under the excess ATP and Mg\u003csup\u003e2+\u003c/sup\u003e-containing conditions adopts the outward-facing conformation (Supplementary Fig.\u0026nbsp;10a,b). However, the substrate ATP-bound ABCD3 structure in the presence of cholic acid adopts the inward-facing conformation, suggesting a potential preference of cholic acid binding to the inward-facing rather than the outward-facing state of ABCD3, although we could not confidently assign the cholic acid model to the density observed at the lateral gate. In addition, the structure revealed that the adenine ring is located near substrate-binding site 2, implying that ATP binding prevents cholic acid from occupying this site, in a similar manner to that observed in the CHS-bound state. Alternatively, the different ATP-incubation conditions may have affected the captured conformations. We incubated ATP with ABCD3\u003csub\u003eEQ\u003c/sub\u003e in glyco-diosgenin (GDN) for 1 h at 4\u0026deg;C for the inward-facing conformation, whereas we incubated ATP with ABCD3\u003csub\u003eEQ\u003c/sub\u003e in nanodiscs for 20 min at room temperature, followed by concentration, for the outward-facing conformation (see Methods).\u003c/p\u003e\u003cp\u003eCollectively, this work provides structural insights into the acyl-CoA substrate specificity and the distinct mechanisms of substrate recognition and transport involving the lateral gate and two spatially separate substrate-binding sites of ABCD3, and paves the way for a deeper understanding of disorders associated with ABCD3.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eProtein preparation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eABCD3 WT and ABCD3\u003csub\u003eEQ\u003c/sub\u003e were expressed using the Bac-to-Bac baculovirus expression system (Thermo Fischer Scientific), according to the method reported previously\u003csup\u003e21\u003c/sup\u003e. \u003cem\u003eSpodoptera frugiperda\u003c/em\u003e Sf9 insect cells (Thermo Fischer Scientific) were infected at a density of 3-4 \u0026times;10\u003csup\u003e6\u003c/sup\u003e cells ml\u003csup\u003e-1\u003c/sup\u003e. The infected Sf9 cells were incubated in Sf900II medium (Gibco) at 27℃ for 48 h, and then collected by centrifugation at 5,000 x \u003cem\u003eg\u003c/em\u003e for 10 min. The cell pellets were resuspended in lysis buffer, containing 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, and 10% glycerol (v/v), and homogenized.\u003c/p\u003e\n\u003cp\u003eFor the apo state, the membrane proteins were extracted from the cells with 1% (w/v) LMNG and protease inhibitors, and rotated gently at 4\u0026deg;C for 1.5 h. After centrifugation at 40,000 rpm for 30 min, the supernatant was mixed with CNBr-activated Sepharose resin (GE) coupled with the GFP-binding nanobody \u0026lsquo;enhancer\u0026rsquo;\u003csup\u003e22\u003c/sup\u003e and stirred at 4\u0026deg;C for 1 h. The resin was washed with 20 column volumes of wash buffer 2, containing 50 mM Tris (pH 8.0), 500 mM NaCl, 0.03% (w/v) GDN, and 10% glycerol. HRV3C protease was added, and the resin was rotated gently at 4\u0026deg;C for 3 h. The eluate was then mixed with Ni-NTA resin and stirred at 4\u0026deg;C for 30 min. The eluate was collected with wash buffer 2 and concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma). The solution was centrifuged at 40,000 rpm for 15 min. The supernatant was purified by SEC on a Superose 6 Increase 10/300 GL column, equilibrated with SEC buffer 2 (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.03% GDN). The fractions containing protein peaks were collected. The protein was concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma) and centrifuged at 40,000 rpm for 15 min. The supernatant was used for grid preparation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFor the acyl-CoA- and CHS-bound states, protein expression and cell collection were performed using the same procedure as for the apo state. The membrane proteins were extracted from the cells with a solution containing 1% (w/v) LMNG, 0.05% (w/v) DDM, 0.02% (w/v) cholesteryl hemisuccinate (CHS) (Sigma), 5.2\u0026nbsp;mg ml\u003csup\u003e-1\u003c/sup\u003e aprotinin, 2.0 mg ml\u003csup\u003e-1\u003c/sup\u003e leupeptin, 1.4 mg ml\u003csup\u003e-1\u003c/sup\u003e pepstatin A, and 100 mM phenylmethylsulfonyl fluoride, by gentle rotation at 4\u0026deg;C for 2 h. After solubilization purification with a GFP-nanobody, SEC was performed, using the same procedure as for the apo state.\u003c/p\u003e\n\u003cp\u003eFor the ATP-bound inward-facing state, protein expression, cell collection, solubilization, and GFP-nanobody purification were using the same procedure as for the apo state. The eluate from the GFP-nanobody purification was purified by SEC on a Superose 6 Increase 10/300 GL column, equilibrated with SEC buffer 3 (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.03% GDN, 0.01% cholic acid). The fractions containing protein peaks were collected. The protein was concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma) and centrifuged at 40,000 rpm for 15 min. The supernatant was used for grid preparation.\u003c/p\u003e\n\u003cp\u003eFor the ATP-bound outward-facing state ABCD3\u003csub\u003eEQ\u003c/sub\u003e, protein expression and cell collection were performed using the same procedure as for the apo state. The membrane proteins were extracted from the cells with a solution containing 1% (w/v) n-dodecyl-beta-D-maltopyranoside (DDM, Merck), 5.2\u0026nbsp;mg ml\u003csup\u003e-1\u003c/sup\u003e aprotinin (Calbiochem), 2.0 mg ml\u003csup\u003e-1\u003c/sup\u003e leupeptin (Calbiochem), 1.4 mg ml\u003csup\u003e-1\u003c/sup\u003e pepstatin A (Sigma), and 100 mM phenylmethylsulfonyl fluoride (PMSF; Roche), by gentle rotation at 4\u0026deg;C for 1.5 h. After centrifugation at 40,000 rpm for 30 min (Beckman, Type 70 Ti), the supernatant was mixed with TALON resin (Takara Bio Company) and stirred at 4\u0026deg;C for 1 h. The resin was washed with 20 column volumes of buffer, containing 50 mM Tris (pH 8.0), 500 mM NaCl, 0.05% (w/v) DDM, 0.01% (w/v) CHS, and 15 mM imidazole. The protein was eluted with 4 column volumes of buffer, containing 50 mM Tris (pH 8.0), 500 mM NaCl, 0.05% (w/v) DDM, 0.01% (w/v) CHS, 10% (v/v) glycerol, and 200 mM imidazole. The eluate was concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma) before the nanodisc reconstitution.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNanodisc reconstitution of ATP-bound ABCD3EQ\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBefore reconstitution, 10\u0026thinsp;mg Brain Total Lipid Extract (BL, Avanti), dissolved in chloroform, was dried using a nitrogen stream and the residual chloroform was further removed by vacuum desiccation (O/N). Lipids were then rehydrated in buffer, containing 1% DDM, 50 mM Tris-HCl (pH 8.0), and 150 mM NaCl.\u0026nbsp;The purified protein was mixed with BPL (please define-brain phospholipids?) and MSP1D1 (membrane scaffold protein 1D1) at a molar ratio of 1:8:100 (ABCD3:MSP1D1:BPL). Nanodisc reconstitution proceeded for 1 h at 4℃. For detergent removal, Bio-Beads SM2 (Bio-Rad) pre-equilibrated with SEC buffer 1 (50 mM Tris-HCl, pH 8.0, 150 mM NaCl) were added and the reconstituted protein mixture was gently rotated for 2 h at 4℃. Afterward, the Bio-Beads were exchanged with fresh Bio-Beads and the mixture was gently rotated overnight at 4℃. The solution, excluding the Bio-Beads, was centrifuged at 40,000 rpm for 15 min. The supernatant was purified by size exclusion chromatography (SEC) on a Superose 6 Increase 10/300 GL column (GE Healthcare), equilibrated with SEC buffer 1. The fractions containing protein peaks reconstituted into nanodiscs were collected and incubated with 5 mM ATP and 5 mM MgCl\u003csub\u003e2\u003c/sub\u003e for 20 min at room temperature (RT). The mixture was concentrated using 100 kDa cut-off Amicon Ultra filters (Millipore-Sigma) and centrifuged at 40,000 rpm for 15 min. The supernatant was used for grid preparation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGrid preparation and Cryo-EM data collection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor the apo state, phytanoyl CoA was added to the purified ABCD3 and incubated at 4℃ for 1 h. The solution was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4\u0026deg;C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained with a 300 kV Titan Krios G3i microscope (Thermo Fisher Scientific), equipped with a GIF Quantum energy filter (Gatan) and a K3 Summit direct electron detector (Gatan). In total, 8,525 movies were acquired at a nominal magnification of 105,000\u0026times;, with a calibrated pixel size of 0.83 \u0026Aring; pix\u003csup\u003e-1\u003c/sup\u003e and a defocus range of -0.8 to -1.6 \u0026mu;m, using the EPU software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 50 e\u003csup\u003e-\u003c/sup\u003e \u0026Aring;\u003csup\u003e-2\u003c/sup\u003e at the grid.\u003c/p\u003e\n\u003cp\u003eFor the acyl-CoA-bound state, 1 mM 5Z,8Z,11Z,14Z,17Z-eicosapentaenoyl Coenzyme A (Avanti) was added to the purified ABCD3 and incubated at 4℃ for 1 h. The solution was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4\u0026deg;C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained as described above. In total, 9,829 movies were acquired at a nominal magnification of 105,000\u0026times;, with a calibrated pixel size of 0.83 \u0026Aring; pix\u003csup\u003e-1\u003c/sup\u003e and a defocus range of -0.8 to -1.6 \u0026mu;m, using the EPU software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 48 e\u003csup\u003e-\u003c/sup\u003e \u0026Aring;\u003csup\u003e-2\u003c/sup\u003e at the grid.\u003c/p\u003e\n\u003cp\u003eFor the CHS-bound ABCD3, the purified ABCD3 was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4\u0026deg;C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained as described above. In total, 5,119 movies were acquired at a nominal magnification of 105,000\u0026times; with a calibrated pixel size of 0.83 \u0026Aring; pix\u003csup\u003e-1\u003c/sup\u003e and a defocus range of -0.8 to -1.6 \u0026mu;m, using the SerialEM software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 49.027 e\u003csup\u003e-\u003c/sup\u003e \u0026Aring;\u003csup\u003e-2\u003c/sup\u003e at the grid.\u003c/p\u003e\n\u003cp\u003eFor the ATP-bound inward-facing ABCD3\u003csub\u003eEQ\u003c/sub\u003e, 5 mM ATP and 5 mM MgCl\u003csub\u003e2\u003c/sub\u003e were added to the purified ABCD3 solution and incubated at 4℃ for 1 h. The solution was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4\u0026deg;C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained as described above. In total, 4,532 movies were acquired at a nominal magnification of 105,000\u0026times;, with a calibrated pixel size of 0.83 \u0026Aring; pix\u003csup\u003e-1\u003c/sup\u003e and a defocus range of -0.8 to -1.6 \u0026mu;m, using the EPU software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 50 e\u003csup\u003e-\u003c/sup\u003e \u0026Aring;\u003csup\u003e-2\u003c/sup\u003e at the grid.\u003c/p\u003e\n\u003cp\u003eFor the ATP-bound outward-facing ABCD3\u003csub\u003eEQ\u003c/sub\u003e in nanodiscs, the nanodisc-reconstituted and ATP/Mg\u003csup\u003e2+\u003c/sup\u003e-incubated ABCD3\u003csub\u003eEQ\u003c/sub\u003e was applied to a glow-discharged Quantifoil holey carbon grid (R1.2/1.3, Au, 300 mesh) and blotted at 4\u0026deg;C in 100% humidity. The grid was then plunge-frozen in liquid ethane by using a Vitrobot Mark IV (FEI). Movies were obtained as described above. In total, 3,942 movies were acquired at a nominal magnification of 105,000\u0026times;, with a calibrated pixel size of 0.83 \u0026Aring; pix\u003csup\u003e-1\u003c/sup\u003e and with a defocus range of -0.8 to -1.6 \u0026mu;m, using the SerialEM software. Each movie was split into 48 frames, resulting in an accumulated exposure of about 51.362 e\u003csup\u003e-\u003c/sup\u003e \u0026Aring;\u003csup\u003e-2\u003c/sup\u003e at the grid.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eImage processing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor the acyl-CoA-bound state, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, using RELION 3.1\u003csup\u003e23\u003c/sup\u003e. The CTF parameter was calculated in CryoSPARC\u003csup\u003e24\u003c/sup\u003e. The following steps were carried out with CryoSPARC. Blob picking was performed on all images, and 1,198,153 particles were extracted at 3.32 \u0026Aring;/pixel. These particles were subjected to 2D classification. After 2D classification, 32,216 particles were used for Topaz\u003csup\u003e25\u003c/sup\u003e picking, and 3,647,328 were extracted at 3.32 \u0026Aring;/pixel. Please note: Kindly check these numbers. These particles were subjected to 2D classification, Ab-Initio Reconstruction and Heterogeneous Refinement. Selected particles were re-extracted at 1.66 \u0026Aring;/pixel. The re-extracted particles were subjected to Ab-Initio Reconstruction and Heterogeneous Refinement with C2 symmetry. Selected particles were re-extracted at 1.107 \u0026Aring;/pixel, and then selected by the same procedure. Afterward, 204,250 particles were symmetry expanded and subjected to 3D classification without alignment in RELION 4.0\u003csup\u003e26\u003c/sup\u003e. Because the substrate density was located near the axis of symmetry, a mask was placed around the substrate in the 3D classification to obtain a substrate-bound density on only one side. Finally, 119,378 particles were selected, and a density map with a global resolution of 3.33 was obtained by Local-refinement in CryoSPARC without symmetry. With C2 symmetry, 204,250 particles were 3D classified by RELION 4.0 and a density map with a global resolution of 3.29 \u0026Aring; was obtained by NU-refinement. The C2 map was used for modeling.\u003c/p\u003e\n\u003cp\u003eFor the CHS-bound state, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, using RELION 3.1. The CTF parameter was calculated in CryoSPARC. The following steps were carried out with CryoSPARC. Blob picking was performed on all images, and 1,267,885 particles were extracted at 3.32 \u0026Aring;/pixel. These particles were subjected to 2D classification, and 1,447,563 particles were extracted at 3.32 \u0026Aring;/pixel by Topaz picking. Please note: Kindly check these numbers. These particles were subjected to 2D classification, Ab-Initio Reconstruction, and Heterogeneous Refinement. The resulting particles were re-extracted at 1.107 \u0026Aring;/pixel. Re-extracted particles were selected by Ab-Initio Reconstruction and Heterogeneous Refinement with C2 symmetry. Selected particles were subjected 3D classification without alignment in RELION 4.0. Finally, 44,696 particles were selected and a density map with a global resolution of 3.81 \u0026Aring; was obtained by NU refinement in CryoSPARC. This map was used for modeling.\u003c/p\u003e\n\u003cp\u003eFor the ATP-bound inward-facing state, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, using RELION 3.1. The CTF parameter was calculated in CryoSPARC. The following steps were carried out within CryoSPARC. Blob picking was performed on all images, and 1,327,401 particles were extracted at 1.66 \u0026Aring;/pixel. These particles were subjected to 2D classification, and then 50,694 particles were used for Topaz picking, resulting in 1,378,789 particles extracted at 3.32 \u0026Aring;/pixel. Please note: Kindly check these numbers. These particles were subjected to 2D classification, Ab-Initio Reconstruction, and Heterogeneous Refinement. Selected particles were re-extracted at 1.107 \u0026Aring;/pixel and the density map was reconstructed by NU-Refinement with C2. Aligned particles were subjected 3D classification without alignment in RELION 4.0. Finally, 65,336 particles were selected and a density map with a global resolution of 2.94 \u0026Aring; was obtained by NU-refinement in CryoSPARC. This map was used for modeling.\u003c/p\u003e\n\u003cp\u003eFor the apo state, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, using RELION 3.1. The CTF parameter was calculated in CryoSPARC. The following steps were carried out with CryoSPARC. Blob picking was performed on all images, and 2,479,005 particles were extracted at 3.32 \u0026Aring;/pixel. These particles were subjected to 2D classification, which yielded 19,690 particles for Topaz picking. After 2,914,648 particles were extracted at 3.32 \u0026Aring;/pixel by Topaz picking, they were subjected to 2D classification, Ab-Initio Reconstruction, and Heterogeneous Refinement. Please note: Kindly check these numbers. The resulting particles were re-extracted at 1.66 \u0026Aring;/pixel. Re-extracted particles were selected by Ab-Initio Reconstruction and Heterogeneous Refinement with C2 symmetry. Selected particles were re-extracted at 1.107 \u0026Aring;/pixel and subjected 3D classification without alignment in RELION 4.0. Finally, 73,450 particles were selected, and a density map with a global resolution of 3.28 \u0026Aring; was obtained by NU refinement in CryoSPARC. This map was used for modeling.\u003c/p\u003e\n\u003cp\u003eFor the ATP-bound outward-facing state ABCD3\u003csub\u003eEQ\u003c/sub\u003e in nanodiscs, all acquired movies were dose-fractionated and subjected to beam-induced motion correction, and the contrast transfer function (CTF) parameter was calculated using RELION 3.1. Laplacian-of-Gaussian picking was performed on 1,000 of these images, and 26,7297 particles were extracted at 3.403 \u0026Aring;/pixel. These particles were subjected to 2D classification. The 8,564 particles thus generated were used for template picking, and 3,875,206 were extracted at 3.403 \u0026Aring;/pixel. After 2D classification, 20,179 particles were used for Topaz picking. Please note: Kindly check these numbers. The data processing steps after Topaz picking were carried out with RELION 4.0. These particles were subjected to 2D and 3D classification and 3D refinement. The resulting particles were subjected to CTF refinement and postprocessing. Finally, 20,043 particles extracted at 1.51244 \u0026Aring;/pixel were selected to produce a map with a global resolution of 3.3 \u0026Aring;. This map was used for modeling.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel building and refinement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe density map exhibited sufficient quality to build a model in COOT\u003csup\u003e27\u003c/sup\u003e. The structure predicted by AlphaFold 2\u003csup\u003e28\u003c/sup\u003e (AF2) was fitted by MOLREP. Next, we manually readjusted ABCD3 and modeled ATP and Mg\u003csup\u003e2+\u003c/sup\u003e molecules. The models were refined using phenix.real_space_refine\u003csup\u003e29\u003c/sup\u003e and Servalcat\u003csup\u003e30\u003c/sup\u003e, and validated using Molprobity\u003csup\u003e31\u003c/sup\u003e. The \u003cem\u003eF\u003c/em\u003e\u003csub\u003eo\u003c/sub\u003e\u0026minus;\u003cem\u003eF\u003c/em\u003e\u003csub\u003ec\u003c/sub\u003e omit map was calculated using Servalcat.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMD simulations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe CHS-bound ABCD3 model was constructed based on the cryo-EM structure, and the THCA-bound ABCD3 model was generated using the CHS-bound model as the template. The unresolved loop regions were modeled as flexible segments. All molecular dynamics (MD) simulations were performed using GROMACS 2020.3\u003csup\u003e32\u003c/sup\u003e. The protein structures were embedded in a 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) lipid bilayer, using the CHARMM-GUI\u003csup\u003e33\u003c/sup\u003e web server. The system was neutralized by adding Cl⁻ ions and consisted of approximately 360 lipids, 52,000 water molecules, 240 chloride ions, and 210 sodium ions. The ionic concentration was adjusted to 150 mM NaCl, and TIP3P water molecules were used. Ligand topologies were generated using the CGenFF\u003csup\u003e34\u003c/sup\u003e web server, and all other topologies and force field parameters were taken from the CHARMM36m\u003csup\u003e35\u003c/sup\u003e force field. A time step of 2\u0026thinsp;fs was used for both equilibration and production runs. The systems were energy-minimized until the maximum force was below 1,000 kJ mol⁻\u0026sup1; nm⁻\u0026sup2;, with positional restraints applied to all non-hydrogen atoms. Equilibration was performed in the NVT and NPT ensembles for 0.1 ns and 10 ns, respectively, using positional restraints of 1,000 kJ mol⁻\u0026sup1; nm⁻\u0026sup2;. Production runs were carried out three times for 100 ns each, without any restraints, under the NPT ensemble with semi-isotropic pressure coupling. The temperature was maintained at 310 K using the Nos\u0026eacute;\u0026ndash;Hoover thermostat, and the pressure was maintained at 1 bar using the Parrinello\u0026ndash;Rahman barostat. Long-range electrostatic interactions were calculated using the particle mesh Ewald (PME) method. Simulation results were analyzed using the PyMOL program\u0026nbsp;(http://www.pymol.org/pymol), MDAnalysis\u003csup\u003e36,37\u003c/sup\u003e, and MDTraj\u003csup\u003e38\u003c/sup\u003e.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCryo-EM density maps have been deposited in the Electron Microscopy Data Bank, under the accession codes EMD-65674, EMD-65675, EMD-65676, EMD-65677 and EMD-65678 for the \u0026nbsp;apo state, the C20:5-CoA-bound state, the CHS-bound state, the ATP-bound inward-facing state, and the ATP-bound outward-facing state structures, respectively. Atomic coordinates have been deposited in the Protein Data Bank, under the IDs 9W62, 9W63, 9W64, 9W65 and 9W66 for the \u0026nbsp;apo state, the C20:5-CoA-bound state, the CHS-bound state, the ATP-bound inward-facing state, and the ATP-bound outward-facing state structures, respectively. The raw images have been deposited in the Electron Microscopy Public Image Archive under the accession code EMPIAR-XXXXX.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank K. Ogomori, C. Harada, T. Tanaka, and T. A. Kobayashi at The University of Tokyo for their technical and editorial assistance in structural analysis, and K. Kawaguchi, M. Morita, and T. So at Toyama University for their help with plasmid preparation. This work was supported by JSPS KAKENHI Grant Numbers JP16H06294 (O.N.), JP20K15754 (T.K.), JP22K15072 (T.K.), JP24K01961 (T.K.), JP25KJ1114 (S.A.), JST PRESTO Grant Number JPMJPR22E4 (T.K.), JST CREST Grant Number JPMJCR20E2 (O.N.) and Research Support Project for Life Science and Drug Discovery (Basis for Supporting Innovative Drug Discovery and Life Science Research (BINDS)) from AMED under Grant Numbers JP21am0101115 (support number 3272), and JP24ama121002 (support number 3272).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eS.A. prepared the cryo-EM samples, and H.H.O. and A.T. assisted with ABCD3 construction, expression, and purification. S.A. and F.K.S. collected the cryo-EM data. S.A., T.K., A.T. and F.K.S. processed the cryo-EM data. S.A. and T.K. build the structures. S.A. and F.K.S. performed MD simulations and analyzed the results. S.A., T.K. and O.N. wrote the manuscript. H.H.O., T.K. and O.N. supervised the research.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interest declaration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eMaxfield, F.R., Tabas, I.: Role of cholesterol and lipid organization in disease. Nature. \u003cb\u003e438\u003c/b\u003e, 612\u0026ndash;621 (2005)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, M., Casey, P.J.: Protein prenylation: unique fats make their mark on biology. Nat. Rev. Mol. Cell. Biol. \u003cb\u003e17\u003c/b\u003e, 110\u0026ndash;122 (2016)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNguyen, P., et al.: Liver lipid metabolism. J. Anim. Physiol. Anim. Nutr. \u003cb\u003e92\u003c/b\u003e, 272\u0026ndash;283 (2008)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGrevengoed, T.J., Klett, E.L., Coleman, R.A.: Acyl-CoA metabolism and partitioning. Annu. Rev. Nutr. \u003cb\u003e34\u003c/b\u003e, 1\u0026ndash;30 (2014)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWatkins, P.A., Ellis, J.M.: Peroxisomal acyl-CoA synthetases. Biochim. Biophys. Acta - Mol. Basis Dis. \u003cb\u003e1822\u003c/b\u003e, 1411\u0026ndash;1420 (2012)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKawaguchi, K., Morita, M.A.B.C., Transporter Subfamily, D.: Distinct Differences in Behavior between ABCD1-3 and ABCD4 in Subcellular Localization, Function, and Human Disease. \u003cem\u003eBioMed Res. Int.\u003c/em\u003e (2016). (2016)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRanea-Robles, P., et al.: The peroxisomal transporter ABCD3 plays a major role in hepatic dicarboxylic fatty acid metabolism and lipid homeostasis. J. Inherit. Metab. Dis. \u003cb\u003e44\u003c/b\u003e, 1419\u0026ndash;1433 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWiesinger, C., Kunze, M., Regelsberger, G., Forss-Petter, S., Berger, J.: Impaired very long-chain acyl-CoA β-oxidation in human X-linked adrenoleukodystrophy fibroblasts is a direct consequence of ABCD1 transporter dysfunction. J. Biol. Chem. \u003cb\u003e288\u003c/b\u003e, 19269\u0026ndash;19279 (2013)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFerdinandusse, S., et al.: A novel bile acid biosynthesis defect due to a deficiency of peroxisomal ABCD3. Hum. Mol. Genet. \u003cb\u003e24\u003c/b\u003e, 361\u0026ndash;370 (2015)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen, Z.P., et al.: Structural basis of substrate recognition and translocation by human very long-chain fatty acid transporter ABCD1. Nat. Commun. \u003cb\u003e13\u003c/b\u003e, (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, R., Qin, Y., Li, X.: Structural basis of acyl-CoA transport across the peroxisomal membrane by human ABCD1. Cell. Res. (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41422-021-00585-8\u003c/span\u003e\u003cspan address=\"10.1038/s41422-021-00585-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLe, L.T.M., Thompson, J.R., Dang, P.X., Bhandari, J., Alam, A.: Structures of the human peroxisomal fatty acid transporter ABCD1 in a lipid environment. Commun. Biol. \u003cb\u003e5\u003c/b\u003e, (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJia, Y., et al.: Structural and functional insights of the human peroxisomal ABC transporter ALDP. \u003cem\u003eeLife\u003c/em\u003e 11, 1\u0026ndash;11 (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXiong, C., et al.: Structural insights into substrate recognition and translocation of human peroxisomal ABC transporter ALDP. Signal. Transduct. Target. Ther. \u003cb\u003e8\u003c/b\u003e, (2023)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, Y., et al.: Structural insights into human ABCD3-mediated peroxisomal acyl-CoA translocation. Cell. Discov \u003cb\u003e10\u003c/b\u003e, (2024)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003evan Roermund, C.W.T., IJlst, L., Linka, N., Wanders, R.J.A., Waterham, H.R.: Peroxisomal ATP Uptake Is Provided by Two Adenine Nucleotide Transporters and the ABCD Transporters. Front. Cell. Dev. Biol. \u003cb\u003e9\u003c/b\u003e, 1\u0026ndash;12 (2022)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXu, D., et al.: Cryo-EM structure of human lysosomal cobalamin exporter ABCD4. Cell. Res. \u003cb\u003e29\u003c/b\u003e, 1039\u0026ndash;1041 (2019)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFerdinandusse, S., Denis, S., Faust, P.L., Wanders, R.J.: A. Bile acids: the role of peroxisomes. J. Lipid Res. \u003cb\u003e50\u003c/b\u003e, 2139\u0026ndash;2147 (2009)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003evan Roermund, C.W.T., IJlst, L., Wagemans, T., Wanders, R.J.A., Waterham, H.: R. A role for the human peroxisomal half-transporter ABCD3 in the oxidation of dicarboxylic acids. Biochim. Biophys. Acta BBA - Mol. Cell. Biol. Lipids. \u003cb\u003e1841\u003c/b\u003e, 563\u0026ndash;568 (2014)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePrydz, K., Kase, B.F., Bj\u0026ouml;rkhem, I., Pedersen, J.I.: Subcellular localization of 3 alpha, 7 alpha-dihydroxy- and 3 alpha,7 alpha,12 alpha-trihydroxy-5 beta-cholestanoyl-coenzyme A ligase(s) in rat liver. J. Lipid Res. \u003cb\u003e29\u003c/b\u003e, 997\u0026ndash;1004 (1988)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDror, R.O., et al.: Structural basis for nucleotide exchange in heterotrimeric G proteins. Science. \u003cb\u003e348\u003c/b\u003e, 1361\u0026ndash;1365 (2015)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKirchhofer, A., et al.: Modulation of protein properties in living cells using nanobodies. Nat. Struct. Mol. Biol. \u003cb\u003e17\u003c/b\u003e, 133\u0026ndash;138 (2010)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZivanov, J., et al.: New tools for automated high-resolution cryo-EM structure determination in RELION-3. \u003cem\u003eeLife\u003c/em\u003e 7, e42166 (2018)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePunjani, A., Rubinstein, J.L., Fleet, D.J., Brubaker, M.A.: cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods. \u003cb\u003e14\u003c/b\u003e, 290\u0026ndash;296 (2017)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBepler, T., et al.: Positive-unlabeled convolutional neural networks for particle picking in cryo-electron micrographs. Nat. Methods. \u003cb\u003e16\u003c/b\u003e, 1153\u0026ndash;1160 (2019)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKimanius, D., Dong, L., Sharov, G., Nakane, T., Scheres, S.H.: W. New tools for automated cryo-EM single-particle analysis in RELION-4.0. Biochem. J. \u003cb\u003e478\u003c/b\u003e, 4169\u0026ndash;4185 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEmsley, P., Lohkamp, B., Scott, W.G., Cowtan, K.: Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. \u003cb\u003e66\u003c/b\u003e, 486\u0026ndash;501 (2010)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJumper, J., et al.: Highly accurate protein structure prediction with AlphaFold. Nature. \u003cb\u003e596\u003c/b\u003e, 583\u0026ndash;589 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiebschner, D., et al.: Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr. Sect. Struct. Biol. \u003cb\u003e75\u003c/b\u003e, 861\u0026ndash;877 (2019)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYamashita, K., Palmer, C.M., Burnley, T., Murshudov, G.N.: Cryo-EM single-particle structure refinement and map calculation using Servalcat. Acta Crystallogr. Sect. Struct. Biol. \u003cb\u003e77\u003c/b\u003e, 1282\u0026ndash;1291 (2021)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDavis, I.W., et al.: MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res. \u003cb\u003e35\u003c/b\u003e, W375\u0026ndash;W383 (2007)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGROMACS 4: 5: a high-throughput and highly parallel open source molecular simulation toolkit | Bioinformatics | Oxford Academic. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://academic.oup.com/bioinformatics/article/29/7/845/253065\u003c/span\u003e\u003cspan address=\"https://academic.oup.com/bioinformatics/article/29/7/845/253065\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJo, S., et al.: CHARMM-GUI 10 years for biomolecular modeling and simulation. J. Comput. Chem. \u003cb\u003e38\u003c/b\u003e, 1114\u0026ndash;1124 (2017)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVanommeslaeghe, K., et al.: CHARMM general force field: A force field for drug-like molecules compatible with the CHARMM all-atom additive biological force fields. J. Comput. Chem. \u003cb\u003e31\u003c/b\u003e, 671\u0026ndash;690 (2010)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuang, J., et al.: CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nat. Methods. \u003cb\u003e14\u003c/b\u003e, 71\u0026ndash;73 (2017)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMichaud-Agrawal, N., Denning, E.J., Woolf, T.B., Beckstein, O.: MDAnalysis: A toolkit for the analysis of molecular dynamics simulations. J. Comput. Chem. \u003cb\u003e32\u003c/b\u003e, 2319\u0026ndash;2327 (2011)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGowers, R.J., et al.: MDAnalysis: A Python Package for the Rapid Analysis of Molecular Dynamics Simulations. \u003cem\u003escipy\u003c/em\u003e (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.25080/Majora-629e541a-00e\u003c/span\u003e\u003cspan address=\"10.25080/Majora-629e541a-00e\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcGibbon, R.T., et al.: MDTraj: A Modern Open Library for the Analysis of Molecular Dynamics Trajectories. Biophys. J. \u003cb\u003e109\u003c/b\u003e, 1528\u0026ndash;1532 (2015)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu, Y., Schmidt, B., Maskell, D.L.: MSAProbs: multiple sequence alignment based on pair hidden Markov models and partition function posterior probabilities. Bioinformatics. \u003cb\u003e26\u003c/b\u003e, 1958\u0026ndash;1964 (2010)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRobert, X., Gouet, P.: Deciphering key features in protein structures with the new ENDscript server. Nucleic Acids Res. \u003cb\u003e42\u003c/b\u003e, W320\u0026ndash;W324 (2014)\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1 | Cryo-EM data collection, refinement and validation statistics.\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"567\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003eABCD3\u003c/p\u003e\n \u003cp\u003e(EMD-65674)\u003c/p\u003e\n \u003cp\u003e(PDB 9W62)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003eABCD3-20:5 CoA\u003c/p\u003e\n \u003cp\u003e(EMD-65675)\u003c/p\u003e\n \u003cp\u003e(PDB 9W63)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5752%;\"\u003e\n \u003cp\u003eABCD3-CHS\u003c/p\u003e\n \u003cp\u003e(EMD-65676)\u003c/p\u003e\n \u003cp\u003e(PDB 9W64)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.7434%;\"\u003e\n \u003cp\u003eABCD3- ATP(IF)\u003c/p\u003e\n \u003cp\u003e(EMD-65677)\u003c/p\u003e\n \u003cp\u003e(PDB 9W65)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.7434%;\"\u003e\n \u003cp\u003eABCD3-ATP(OF)\u003c/p\u003e\n \u003cp\u003e(EMD-65678)\u003c/p\u003e\n \u003cp\u003e(PDB 9W66)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eData collection and processing\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eMagnification\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026times;105,000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026times;105,000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e\u0026times;105,000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026times;105,000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026times;105,000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eVoltage (kV)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e300\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eElectron exposure (e\u0026ndash;/\u0026Aring;\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e51\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eDefocus range (\u0026mu;m)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026minus;1.6 to \u0026minus;0.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026minus;1.6 to \u0026minus;0.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e\u0026minus;1.6 to \u0026minus;0.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026minus;1.6 to \u0026minus;0.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026minus;1.6 to \u0026minus;0.8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003ePixel size (\u0026Aring;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e0.83\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eSymmetry imposed\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003eC2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003eC2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003eC2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003eC2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003eC2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eInitial particle images (no.)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e2,914,648\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e3,647,328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e119,378\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e1,378,789\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e2,805,407\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eFinal particle images (no.)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e73,450\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e204,250\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e44,696\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e65,336\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e20,043\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eMap resolution (\u0026Aring;)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; FSC threshold\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e3.28\u003c/p\u003e\n \u003cp\u003e0.143\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e3.29\u003c/p\u003e\n \u003cp\u003e0.143\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e3.81\u003c/p\u003e\n \u003cp\u003e0.143\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e2.94\u003c/p\u003e\n \u003cp\u003e0.143\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e3.20\u003c/p\u003e\n \u003cp\u003e0.143\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eMap resolution range (\u0026Aring;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e2.79-37.93\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e2.73-43.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e3.22-43.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e2.59-44.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e2.14-4.92\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRefinement\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eModel resolution (\u0026Aring;)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; FSC threshold\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e3.28\u003c/p\u003e\n \u003cp\u003e0.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e3.29\u003c/p\u003e\n \u003cp\u003e0.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e3.81\u003c/p\u003e\n \u003cp\u003e0.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e2.94\u003c/p\u003e\n \u003cp\u003e0.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e3.20\u003c/p\u003e\n \u003cp\u003e0.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eMap sharpening \u003cem\u003eB\u003c/em\u003e factor (\u0026Aring;\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e-100.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e-115.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e-111.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e-97.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e-69.4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eModel composition\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Non-hydrogen atoms\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Protein residues\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Ligands\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e9548\u003c/p\u003e\n \u003cp\u003e1148\u003c/p\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e9736\u003c/p\u003e\n \u003cp\u003e1146\u003c/p\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e9418\u003c/p\u003e\n \u003cp\u003e1150\u003c/p\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e9588\u003c/p\u003e\n \u003cp\u003e1150\u003c/p\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e9722\u003c/p\u003e\n \u003cp\u003e1162\u003c/p\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003e\u003cem\u003eB\u003c/em\u003e factors (\u0026Aring;\u003csup\u003e2\u003c/sup\u003e; mean)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Protein\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Ligand\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e119.76\u003c/p\u003e\n \u003cp\u003e133.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e198.42\u003c/p\u003e\n \u003cp\u003e174,89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e187.34\u003c/p\u003e\n \u003cp\u003e209.85\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e124.64\u003c/p\u003e\n \u003cp\u003e151.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e104.37\u003c/p\u003e\n \u003cp\u003e136.12\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003eR.m.s. deviations\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Bond lengths (\u0026Aring;)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Bond angles (\u0026deg;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e0.0110\u003c/p\u003e\n \u003cp\u003e2.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e0.0103\u003c/p\u003e\n \u003cp\u003e2.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e0.0108\u003c/p\u003e\n \u003cp\u003e2.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e0.0111\u003c/p\u003e\n \u003cp\u003e2.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e0.0089\u003c/p\u003e\n \u003cp\u003e1.54\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003e\u0026nbsp;Validation\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; MolProbity score\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Clashscore\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Poor rotamers (%) \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e1.70\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e4.33\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e1.38\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e1.82\u003c/p\u003e\n \u003cp\u003e5.49\u003c/p\u003e\n \u003cp\u003e1.77\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e1.78\u003c/p\u003e\n \u003cp\u003e2.42\u003c/p\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e1.96\u003c/p\u003e\n \u003cp\u003e3.31\u003c/p\u003e\n \u003cp\u003e3.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e2.25\u003c/p\u003e\n \u003cp\u003e10.86\u003c/p\u003e\n \u003cp\u003e4.66\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 27.0796%;\"\u003e\n \u003cp\u003e\u0026nbsp;Ramachandran plot\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Favored (%)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Allowed (%)\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; Disallowed (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e94.39\u003c/p\u003e\n \u003cp\u003e5.61\u003c/p\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.9292%;\"\u003e\n \u003cp\u003e95.25\u003c/p\u003e\n \u003cp\u003e4.75\u003c/p\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.5752%;\"\u003e\n \u003cp\u003e96.50\u003c/p\u003e\n \u003cp\u003e3.50\u003c/p\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e94.40\u003c/p\u003e\n \u003cp\u003e5.60\u003c/p\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 12.7434%;\"\u003e\n \u003cp\u003e96.88\u003c/p\u003e\n \u003cp\u003e3.12\u003c/p\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7304961/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7304961/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eHuman peroxisomal ABC subfamily D transporters, ABCD1, ABCD2, and ABCD3, transport acyl-CoA molecules and bile acid precursors, and thereby participate in fatty acid metabolism and cholesterol homeostasis. ABCD3 can recognize a wide range of substrates, including polyunsaturated fatty acyl-CoAs, dicarboxylic acyl-CoAs, and ATP molecules. However, the structural mechanism of this broad substrate recognition by ABCD3 remains largely unknown. Here we report five cryo-electron microscopy structures of ABCD3: the apo state, the eicosapentaenoyl coenzyme A (C20:5-CoA)-bound state, the cholesteryl hemisuccinate (CHS)-bound state, the ATP-bound inward-facing state, and the ATP-bound outward-facing state. In the C20:5-CoA-bound state, the CoA moiety is recognized by ABCD3 in a similar manner to ABCD1, but the long, unsaturated bent acyl chain moiety is accommodated in a large cavity between protomers, distinct from the saturated straight acyl chain moiety-binding site observed in ABCD1. Furthermore, we found that the structurally and chemically different substrates, ATP and CHS, bind to distinct hydrophilic and hydrophobic sites, respectively, in the transmembrane domain of ABCD3. Combined with molecular dynamics simulations, these structures, encompassing both inward- and outward-facing conformations, reveal the molecular mechanism by which ABCD3 recognizes and transports a broad range of substrates.\u003c/p\u003e\u003cp\u003eMain text\u003c/p\u003e","manuscriptTitle":"Structural Insights into the Broad Substrate Recognition Mechanism of Human ABCD3","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-07 11:02:19","doi":"10.21203/rs.3.rs-7304961/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"communications-biology","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"commsbio","sideBox":"Learn more about [Communications Biology](http://www.nature.com/commsbio/)","snPcode":"","submissionUrl":"","title":"Communications Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Communications Series","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"5bedf02b-3edb-4238-a31d-48d00c4d5d42","owner":[],"postedDate":"November 7th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":53332397,"name":"Biological sciences/Structural biology/Electron microscopy/Cryoelectron microscopy"},{"id":53332398,"name":"Biological sciences/Cell biology/Organelles/Peroxisomes"}],"tags":[],"updatedAt":"2025-11-07T11:02:19+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-07 11:02:19","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7304961","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7304961","identity":"rs-7304961","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}