The ATP binding cassette (ABC) transporter of mitochondria (Atm) family plays a vital (Leighton and Schatz, 1995), but enigmatic, role broadly related to transition metal homeostasis in eukaryotes (Lill et al., 2014). The best characterized member is Saccharomyces cerevisiae Atm1 (ScAtm1) present in the inner membrane of mitochondria (Leighton and Schatz, 1995) and required for the formation of cytosolic iron-sulfur cluster containing proteins (Kispal et al., 1999). Defects in ScAtm1 lead to an overaccumulation of iron in the mitochondria (Kispal et al., 1997). Atm1 is proposed to transport a sulfur containing intermediate (Kispal et al., 1999) that may also include iron (Pandey et al., 2019). It is also likely to transport a similar sulfur containing species from the mitochondria that is required for the cytoplasmic thiolation of tRNA (Pandey et al., 2018). While the precise substrate that is transported remains unknown, derivatives of glutathione have been implicated based on their ability to stimulate the ATPase activity of Atm1 (Kuhnke et al., 2006).

Structures for Atm family members are currently available for ScAtm1 (Srinivasan et al., 2014), the bacterial homolog NaAtm1 from Novosphingobium aromaticivorans (Lee et al., 2014) and human ABCB6 (Wang et al., 2020); the pairwise sequence identities between these homologous transporters range from 40% to 46%. These proteins occur as homodimers of half-transporters, where each half-transporter contains a transmembrane domain (TMD) followed by the canonical nucleotide binding domain (NBD) that defines the ABC transporter family. Each TMD consists of six transmembrane helices (TMs) that exhibit the exporter type I fold first observed for Sav1866 (Dawson and Locher, 2006); a recent re-classification now identifies this group as type IV ABC transporters (Thomas et al., 2020). The translocation of substrates across the membrane proceeds through an alternating access mechanism involving the ATP-dependent interconversion between inward- and outward-facing conformational states. Among the Atm1 family, these conformations have been most extensively characterized for NaAtm1 and include the occluded and closed states that provide a structural framework for the unidirectional transport cycle (Fan et al., 2020). Structures of ScAtm1 with reduced glutathione (GSH) (Srinivasan et al., 2014), and of NaAtm1 complexed with reduced (GSH), oxidized (GSSG), and metallated (GS-Hg-SG) (Lee et al., 2014), have defined the general substrate binding site in the TMD for the transport substrates.

Plants have been found to have large numbers of transporters (Hwang et al., 2016), including Arabidopsis with three Atm orthologues, AtAtm1, AtAtm2, and AtAtm3 (Chen et al., 2007). Of these, AtAtm3 (also known as ABCB25) rescues the ScAtm1 phenotype (Chen et al., 2007), and has been shown to be associated with maturation of cytosolic iron-sulfur proteins (Kushnir et al., 2001), confer resistance to heavy metals such as cadmium and lead (Kim et al., 2006), and participate in the formation of molybdenum-cofactor containing enzymes (Bernard et al., 2009; Teschner et al., 2010). Unlike yeast, defects in AtAtm3 are not associated with iron accumulation in mitochondria (Bernard et al., 2009). While the physiological substrate is unknown, AtAtm3 has been shown to transport GSSG and glutathione polysulfide, with the persulfidated species perhaps relevant to cytosolic iron-sulfur cluster assembly (Schaedler et al., 2014). The ability of AtAtm3 to export GSSG has been implicated in helping stabilize against excessive glutathione oxidation in the mitochondria and thereby serving to maintain a suitable reduction potential (Marty et al., 2019).

To help address the functional role(s) of Atm transporters, we have determined structures of AtAtm3 in multiple conformational states by single-particle cryo-electron microscopy (cryoEM). These structures not only provide a structural framework for defining the alternating access transport cycle, but they also illuminate an unappreciated feature of the glutathione binding site, namely the paucity of cysteine residues that could potentially form inhibitory mixed disulfides during the transport cycle. A survey of structurally characterized members of the type IV family of ABC transporters, including the Atm1 family, establishes that nucleotides are effectively required for the stabilization of the occluded, closed, and outward-facing conformations. In contrast to the nucleotide states, transport substrates and related inhibitors have only been observed associated with inward-facing conformational states. The absence of structures with dimerized NBDs containing both nucleotide and transport substrate suggests that this form of the ternary complex exists only transiently during the transport cycle.

AtAtm3 contains an N-terminal mitochondrial targeting sequence that directs the translated protein to the mitochondria, where it is cleaved following delivery to the inner membrane. Since this targeting sequence consists of ~80 residues and is anticipated to be poorly ordered, we generated three different N-terminal truncation mutants of AtAtm3 through deletion of 60, 70, or 80 residues to identify the best-behaved construct. Together with the wild-type construct, these three variants were heterologously overexpressed in Escherichia coli. The construct with the 80 amino acids deletion showed the highest expression level and proportionally less aggregation by size exclusion chromatography (Figure 1—figure supplement 1) and hence was used for further functional and structural studies.

ATPase activities

Using the 80-residue truncation construct, AtAtm3 was purified in the detergent n-dodecyl-β-D-maltoside (DDM) and reconstituted into nanodiscs formed from the membrane scaffolding protein (MSP) 1D1 and the lipid 1-palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC). The ATPase activity of this construct was measured as a function of MgATP concentration in the absence and presence of either 2.5 mM GSSG or 10 mM GSH, which approximate their physiological concentrations in E. coli (Bennett et al., 2009). The rate of ATP hydrolysis was determined by measuring phosphate release using a molybdate-based colorimetric ATPase activity assay (Chifflet et al., 1988). The basal ATPase activity, measured in the absence of glutathione derivatives, was significantly higher in detergent than in nanodiscs (104 vs. 7.7 nmol/min/mg, respectively; Figure 1ab), while the apparent K_ms for MgATP were within a factor of two (~0.16 and 0.08 mM, respectively). The ATPase activity of AtAtm3 is stimulated by both 2.5 mM GSSG and 10 mM GSH, but the extent of stimulation depends strongly on the solubilization conditions. In nanodiscs, the ATPase rates increase to 32 and 39 nmol/min/mg with 2.5 mM GSSG and 10 mM GSH, respectively, for an overall increase of 4–5× above the basal rate. The ATPase rates for AtAtm3 in DDM also increase with GSSG and GSH, to 117 and 154 nmol/min/mg, respectively. Because of the higher basal ATPase rate in detergent, however, the stimulation effect is significantly less pronounced, corresponding to only an ~50% increase for GSSG stimulation. Little change is observed for the K_ms of MgATP between the presence and absence of glutathione derivatives for either detergent solubilized or nanodisc reconstituted AtAtm3 (Figure 1).

Figure 1 with 1 supplement see all

Download asset Open asset

Figure 1—source data 1

Numerical data for the graphs depicted in Figure 1a and b.

https://cdn.elifesciences.org/articles/76140/elife-76140-fig1-data1-v2.xlsx

Download elife-76140-fig1-data1-v2.xlsx

Inward-facing, nucleotide-free conformational states

To map out the transport cycle, we attempted to capture AtAtm3 in distinct liganded conformational states using single-particle cryoEM. We first determined the structure of AtAtm3 reconstituted in nanodiscs at 3.4 Å resolution in the absence of either nucleotide or transport substrate (Figure 2a and Figure 2—figure supplement 1). This structure revealed an inward-facing conformation for AtAtm3 similar to those observed for the inward-facing conformations for ScAtm1 (PDB ID: 4myc) and NaAtm1 (PDB ID: 6vqu) with overall alignment root mean square deviations (rmsds) for the dimer of 2.6 Å (Figure 2—figure supplement 2a, b) and 2.1 Å (Figure 2), respectively, and half-transporter alignment rmsds of 2.3 and 2.0 Å (Figure 2—figure supplement 2c), respectively. The primary distinction between these structures is the presence of an approximately 20 amino acid loop between TM1 and TM2 of AtAtm3 that would be positioned in the intermembrane space and is absent from the structures of ABCB7 (Jumper et al., 2021; Varadi et al., 2022), ABCB6 (Song et al., 2021), ScAtm1 (Srinivasan et al., 2014), and NaAtm1 (Lee et al., 2014; Figure 2—figure supplement 3). While the functional relevance of this loop in AtAtm3 is not known, structural characterization of PglK, a lipid-linked oligosaccharide flippase, revealed a comparably positioned external helix between TM1 and TM2 that was implicated in substrate flipping (Perez et al., 2015), suggestive that the corresponding loop could also have a functional or structural role in AtAtm3.

Figure 2 with 8 supplements see all

Download asset Open asset

To further look at substrate binding, we determined a 3.6 Å resolution single-particle cryoEM structure of AtAtm3 purified in DDM with bound GSSG (Figure 2b and Figure 2—figure supplement 4). Although the overall resolution of the reconstruction was moderate (Figure 2—figure supplement 4d), we were able to model the GSSG molecule into the density map. In this structure, AtAtm3 adopts an inward-facing conformation, with an overall alignment rmsd to the ligand-free inward-facing structure of 2.9 Å (Figure 2—figure supplement 5a) and a corresponding half-transporter alignment rmsd of 1.6 Å (Figure 2—figure supplement 5b). The main difference between the two structures is the extent of NBD dimer separation (Figure 2—figure supplement 5a), where the GSSG-bound structure presents a more closed NBD dimer relative to the substrate-free structure. As previously noted with NaAtm1 (Fan et al., 2020), the TM6s in these inward-facing structures of AtAtm3 adopt a kinked conformation including residues 429–438 (Figure 2cd). This opens the backbone hydrogen bonding interactions to create the binding site for GSSG (Figure 2e) with binding pocket residues identified by PDBePISA (Krissinel and Henrick, 2007). The binding mode of GSSG in this AtAtm3 inward-facing conformation is similar to that observed in the inward-facing structure of the GSSG-bound NaAtm1 (Lee et al., 2014).

The plant mitochondrial Atm3 transporter has been implicated in a diverse set of functions associated with transition metal homeostasis that are reflective of the roles that have been described for the broader Atm1 transporter family. To provide a general framework for addressing the detailed function of this transporter in plants, we have structurally and functionally characterized Atm3 from Arabidopsis thaliana. We first identified a construct of AtAtm3 with the mitochondrial targeting sequence deleted that expressed well in E. coli (Figure 1—figure supplement 1). Following purification, the ATPase activities of AtAtm3 were measured in both detergent and MSP nanodiscs as a function of MgATP concentrations (Figure 1). Overall, the ATPase rate measured in detergent is about fivefold greater than that measured in nanodiscs, perhaps indicative of a more tightly coupled ATPase activity in a membrane-like environment. Both GSH and GSSG stimulate the ATPase activity by increasing V_max, with little change observed in the K_m for MgATP. The ability of GSSG to stimulate the ATPase activity of AtAtm3 agrees with previous reports (Schaedler et al., 2014), while the stimulation we observe with 10 mM GSH differs from the lack of stimulation noted in that work with 1.7 and 3.3 mM GSH. This discrepancy may reflect the higher GSH concentration utilized in the present studies, as well as differences in other experimental conditions including the use of a Lactococcus lactis expression system and ∆60 N-terminal truncation by Schaedler et al., 2014, compared to the E. coli expression system and the ∆80 N-terminal truncation employed in the present work.

ABC transporters are typically envisioned as utilizing an ‘alternating access’ mechanism, in which the substrate binding site transitions between inward- and outward-facing conformations coupled to the binding and hydrolysis of ATP. In an idealized two-state model, ABC transporters only adopt these two limiting conformations, but structural characterizations of ABC transporters in the presence of nucleotides and substrate analogs have identified a variety of intermediates, including occluded (with a ligand binding cavity exhibiting little or no access to either side of the membrane) and closed (no ligand binding cavity) conformations. The most extensive analysis of the conformational states of an Atm1-type exporter has been detailed for NaAtm1 and assigned to various states in the transport cycle (Fan et al., 2020; Lee et al., 2014). In the present work, we have determined four structures of AtAtm3 in three different conformational states by single-particle cryoEM: two inward-facing conformations (with and without bound GSSG) (Figure 2ab), together with closed and outward-facing states stabilized by MgADP-VO₄ (Figure 2fg). The parallels between the structurally characterized conformations of AtAtm3 and NaAtm1 support the idea that these conformational states are relevant to the transport cycle, and not simply an artifact of the specific conditions used to prepare each sample. The conformations observed for AtAtm3 and NaAtm1 do not completely correspond, however; most notably, the outward-facing conformation observed for AtAtm3 had not been previously observed with NaAtm1 (Fan et al., 2020; Lee et al., 2014), while the occluded conformations found with NaAtm1 were not observed for AtAtm3.

The closed conformation stabilized by MgADP-VO₄ has been observed for both NaAtm1 (Fan et al., 2020) and AtAtm3. In comparing the closed and the outward-facing conformations of AtAtm3, the arrangements of the NBDs are superimposable, with the major differences between the two conformational states involving the local conformation of TM6s. In the closed conformation, the TM6s adopt a kinked conformation at residues 438–441 that eliminates the substrate binding cavity, whereas the TM6s in the outward-facing conformation present straight TM6s (Figure 2hi). The TM6 helical kink in the closed conformation is adjacent to, but distinct from, the helical kink present at residues 429–438 in the inward-facing conformation. By eliminating the substrate binding cavity, the presence of the closed conformation in a post-ATP hydrolysis state enforces unidirectionality of the transport process by precluding the uptake of substrate from the outside. We note that MgADP-VO₄ was observed to stabilize two different conformational states for AtAtm3, the closed state in nanodiscs and the outward-facing conformation in detergent (Figure 2fg). The stabilization of multiple conformations of AtAtm3 with MgADP-VO₄ contrasts with our previous observations on NaAtm1, where only the closed conformation was observed (Fan et al., 2020). The underlying basis for these differences is not known but may reflect differences in the conformational stabilization of the membrane spanning regions between detergents and nanodiscs. Structural differences between detergents and nanodiscs have been previously reported for MsbA under MgADP-VO₄ stabilizing conditions (Mi et al., 2017; Ward et al., 2007), and in the functional analysis of other membrane proteins (Hänelt et al., 2013).

In contrast to differences in the structures of the TMDs between the closed and outward-facing conformations, the TMDs in the inward-facing conformation structures of AtAtm3 are similar. The primary differences between the two structures of inward-facing conformations of AtAtm3 are in the relative positioning of the NBDs which are more widely separated in the apo structure relative to the GSSG-bound structure. Similar substrate-induced NBD movements have been previously observed in MRP1 (Johnson and Chen, 2017) and ABCB1 (Barbieri et al., 2021).

The conformational changes in the TMDs underlying the transport cycle are associated with changes in the extent of kinking of TM6 and the positioning of TM4-TM5 relative to the core formed by the remaining four TMs. As noted for NaAtm1, we observed kinked TM6s in the inward-facing and closed state of AtAtm3 (Figure 2cdh), but not the outward-facing conformation (Figure 2i). These conformational changes lead to changes in the volume of the central cavity forming the glutathione binding site. Using the program CastP (Tian et al., 2018) with a probe radius of 2.5 Å, the cavity volumes of the inward-facing apo and GSSG-bound structures were measured to be ~6500 ų (Figure 3a) and ~4300 ų (Figure 3b), respectively, while the closed conformation exhibits a cavity volume of ~300 ų (Figure 3c), and the outward-facing conformation has a cavity volume of ~5700 ų (Figure 3d). We also measured the accessible solvent areas (ASA) of the key residues forming the binding site for GSSG in the different conformational states using Areaimol in CCP4 (Winn et al., 2011); the ASA of the inward-facing, inward-facing with GSSG bound, closed, and outward-facing structures are ~1500, ~1100, ~900, and ~1300 Ų, which are also highly correlated with the cavity volume calculations. Most of the binding pocket residues remain exposed in all conformations with a few having large relative changes than others (Figure 3—figure supplement 1). Further, the cavity volume measurements are comparable to those calculated for NaAtm1 (Fan et al., 2020). The similarities in conformational states between NaAtm1 and AtAtm3 indicate these transporters follow the same basic mechanism, in which straightening of TM6s in the transition from inward to outward conformation leads to the release of substrate to the opposite side of the membrane. Following substrate release, the transporter resets to the inward-facing conformation through the closed conformation adopted after ATP hydrolysis; the decreased size of the substrate binding cavity helps enforce substrate release and unidirectionality of substrate transport.

Figure 3 with 1 supplement see all

Download asset Open asset

The binding pocket for GSSG identified in this work primarily consists of residues from TM5 and TM6, with additional contributions from residues in TM3 and TM4 (Figure 4—figure supplement 1). The GSSG binding site for AtAtm3 largely overlaps with that identified previously for NaAtm1 (Lee et al., 2014) and for the binding of reduced GSH to ScAtm1 (Srinivasan et al., 2014). Inspection of a sequence alignment of Atm1 homologs (Figure 4—figure supplement 1) reveals that those residues forming the glutathione binding site are largely conserved, particularly if they are involved in polar interactions. A striking feature is the stretch of residues from P432 to R441 in the middle of TM6 (AtAtm3 sequence numbering) with sequence PLNFLGSVYR with a high degree of sequence conservation. P432 is associated with the TM6 kink in inward-facing conformations that permits formation of hydrogen bonds between exposed peptide groups with GSSG (Lee et al., 2014); as TM6 straightens in the occluded and outward-facing conformations, these peptide groups are no longer available to bind the transport substrate (Fan et al., 2020). A sequence alignment of the structurally characterized AtAtm3, NaAtm1, ScAtm1, and human ABCB7 and ABCB6 transporters establishes that residues in the binding pockets are conserved, including T317, R324, R328, N387, Q390, L433, G437, and R441 (Figure 4—figure supplement 1). The conservation of binding pocket residues as calculated by the program ConSurf (Landau et al., 2005) is illustrated in Figure 4 suggests that the substrates for these transporters may share common features, such as the glutathione backbone. Positions such where sequence variability is evident, such as residue 435 (Figure 4b and Figure 4—figure supplement 1), may reflect the binding of distinct GSSG derivatives by different eukaryotic and prokaryotic homologs.

Figure 4 with 1 supplement see all

Download asset Open asset

An important property of disulfide containing compounds such as GSSG is that they can undergo disulfide – thiol exchange with free -SH groups (Creighton, 1984; Nagy, 2013). This reactivity creates potential challenges for proteins such as AtAtm3 since reaction of a disulfide containing ligand such as GSSG with the thiol-containing side chain of cysteine could lead to formation of the mixed disulfide, thereby covalently connecting glutathione to the protein and releasing reduced GSH. Formation of the covalently linked mixed disulfide would be expected to restrict the access of exogenous ligands to the substrate binding cavity and hence would inhibit transport. For membrane proteins, cysteine residues are present in TMs with a frequency of about 1% (Baeza-Delgado et al., 2013). Although no cysteines residues are present in the AtAtm3 binding pocket, we analyzed additional Atm3 homologs from plants. For this analysis, we used the NCBI blastp server (Altschul et al., 1997) and selected 410 sequences with a sequence identity of 50–100% and query coverage of 80–100% with AtAtm3. Within the six TMs, the overall presence of cysteines was found to be ~0.4%. In this alignment, no cysteines were found in residues forming the glutathione binding cavity (Figure 5); more strikingly, no cysteine residues were found at any position of TM6 for these homologs (Figure 5f). Cysteines that are present in the TMs are either distant from the binding site, such as position 405 in TM5 of AtAtm3 (Figure 5eg) or if they are closer to the binding site, are positioned on the opposite side of the TMs, such as positions 149, 215, 290, and 307 (Figure 5abcd). The observed exclusion of cysteines from the glutathione binding site (Figure 5g) could consequently be the result of a selection against this residue to prevent the formation of inhibitory mixed disulfides during the transport cycle.

Figure 5

Download asset Open asset

As a general strategy to stabilize ABC transporters in distinct conformational states, different nucleotides or transport substrates are mixed with the transporter. The expectation is that a particular set of ligands will stabilize a specific conformational state, and so we were surprised to have captured with MgADP-VO₄ both a closed conformation in MSP nanodiscs and an outward-facing conformation in detergent. Given the similarities in the NBDs between these two structures, the distinctive arrangements of the TMDs between the closed and outward-facing conformations presumably arise from differences in the TMD environment provided by MSP nanodiscs and DDM, respectively. Furthermore, this phenomenon of obtaining different conformational states with MgADP-VO₄ is not unprecedented, however, since previously determined structures of MgADP-VO₄ stabilized ABC exporters include NaAtm1 in the closed conformation (Fan et al., 2020), Thermus thermophilus TmrAB in the occluded and outward-facing conformations (Hofmann et al., 2019), and E. coli MsbA in the closed conformation (Mi et al., 2017).

Despite extensive efforts, we were unable to prepare the ternary complex of AtAmt3 with both bound GSSG and MgATP. To assess more generally the relationship between the conformational states of ABC exporters and the presence or absence of nucleotide and transport substrate, we systematically compared the conformations of 80 half-transporters from the available structures of type IV ABC transporters (Supplementary file 2). To guide this analysis, the structures were mapped onto a one-dimensional conformational axis using principal component analysis (PCA; see Materials and methods). The PCA has the advantage for this purpose of separating all structures along a single axis such that transporters with similar structures will generally be positioned more closely together. Since only the dominant component is used, however, this analysis simplifies the conformational richness of ABC transporters. We note that the PCA does not provide a unique measure of the conformational state, and in particular, alternative metrics could be employed such as measuring distances between particular pairs of residues. A comparison is presented in Figure 6—figure supplement 1 comparing the dominant PCA component to alternative distance measures; these are generally anti-correlated, with the interesting exception of the distance between a pair of residues positioned in the outer facing (intermembrane space) that, not surprisingly, does a much better job of discriminating between outward conformations than the PCA, while less sensitive to differences between inward conformations. As the quantitative value of the principal component was utilized in this analysis only to separate structures along a single conformational analysis (Figure 6), we utilized the PCA for that purpose in this analysis.

Figure 6 with 1 supplement see all

Download asset Open asset

The principal component is dominated by the conformational state of the TMDs, which represents ~62% of total conformational change from the inward-facing to the outward-facing conformation. The distribution of component 1 of the PCA is shown in Figure 6 with the most extreme inward-facing conformations to the left and the outward-facing conformations to the right. To validate this approach, we color coded each structure according to their published conformational state (Figure 6a). It is evident that outward-facing conformations occur on the right-handed side of the figure with component 1 values between –25 and 0 Ų, while occluded, closed, and inward-facing conformations cluster around −25,–25, and –100 Ų, respectively. Hence, the magnitude of the principal component does capture the expected trends in conformational state, with increasing values corresponding to a progression from inward-facing to outward-facing conformations. The correspondence is not exact, however, since assigned conformational states overlap, which could reflect either limitations of the PCA and/or inconsistent assignments of conformational states between different structures. We further note that some conformational states have a very wide distribution, particularly the inward-facing conformations, with several structures exhibiting widely separated subunits (with values for component 1 below –150 Ų).

Using the PCA, we could relate protein conformation to binding of nucleotides (Figure 6b) and substrates/inhibitors (Figure 6c). This analysis establishes that occluded and outward-facing conformations are largely nucleotide bound with either ATP or an ATP analog, but there are exceptions, most notably the original ADP-bound structure of Sav1866 in the outward facing conformation (Dawson and Locher, 2006; a similar Sav1866 structure was subsequently solved with bound AMPPNP; Dawson and Locher, 2007). Inward-facing conformations are observed in both nucleotide-free and nucleotide-bound forms (Figure 6b). Only the MgADP-VO₄-bound structures are exclusively found to occupy a small conformational space in the closed/outward-facing region. In contrast to the binding of nucleotides to all conformational states of these transporters, a distinct pattern is observed for the binding of transport substrates where the transporter invariably adopts the inward-facing conformation (Figure 6c). Intriguingly, no structures with associated (dimerized) NBDs to date have been published that contain both nucleotide and transport substrate, suggesting that this ternary complex exists only transiently during the transport cycle. As this is a key intermediate for understanding how the binding of transport substrate stimulates the ATPase activity, characterization of the ternary complex represents an outstanding gap in the mechanistic characterization of ABC exporters.

The atomic coordinates for inward-facing, inward-facing with GSSG bound, closed and outward-facing conformations were separately deposited in the Protein Data Bank (PDB) and the Electron Microscopy Data Bank (EMDB) with accession codes: PDB 7N58, 7N59, 7N5A and 7N5B; EMDB EMD-24182, EMD-24183, EMD-24184 and EMD-24185. The plasmids encoding full-length AtAtm3 and the AtAtm3 with N-terminal 80 residue deletion were deposited in Addgene with Addgene ID 172321 and 173045, respectively. The raw data for ATPase assays presented in Figure 1 are provided in Source Data 1, while the essdyn.f Fortran source code used for the PCA analysis is provided as Source Code 1.

1. 1. Fan C
   2. Rees DC
   3. Drew D

   (2022) RCSB Protein Data Bank

   ID 7N58. Glutathione binding to the plant AtAtm3 transporter and implications for the conformational coupling of ABC transporters.

   https://www.rcsb.org/structure/7N58
2. 1. Fan C
   2. Rees DC
   3. Drew D

   (2022) RCSB Protein Data Bank

   ID 7N59. Glutathione binding to the plant AtAtm3 transporter and implications for the conformational coupling of ABC transporters.

   https://www.rcsb.org/structure/7N59
3. 1. Fan C
   2. Rees DC
   3. Drew D

   (2022) RCSB Protein Data Bank

   ID 7N5A. Glutathione binding to the plant AtAtm3 transporter and implications for the conformational coupling of ABC transporters.

   https://www.rcsb.org/structure/7N5A
4. 1. Fan C
   2. Rees DC
   3. Drew D

   (2022) RCSB Protein Data Bank

   ID 7N5B. Glutathione binding to the plant AtAtm3 transporter and implications for the conformational coupling of ABC transporters.

   https://www.rcsb.org/structure/7N5B
5. 1. Fan C
   2. Rees DC
   3. Drew D

   (2022) Electron Microscopy Data Bank

   ID EMD-24182. Glutathione binding to the plant AtAtm3 transporter and implications for the conformational coupling of ABC transporters.

   http://www.ebi.ac.uk/pdbe/entry/emdb/EMD-24182
6. 1. Fan C
   2. Rees DC
   3. Drew D

   (2022) Electron Microscopy Data Bank

   ID EMD-24183. Glutathione binding to the plant AtAtm3 transporter and implications for the conformational coupling of ABC transporters.

   http://www.ebi.ac.uk/pdbe/entry/emdb/EMD-24183
7. 1. Fan C
   2. Rees DC
   3. Drew D

   (2022) Electron Microscopy Data Bank

   ID EMD-24184. Glutathione binding to the plant AtAtm3 transporter and implications for the conformational coupling of ABC transporters.

   http://www.ebi.ac.uk/pdbe/entry/emdb/EMD-24184
8. 1. Fan C
   2. Rees DC
   3. Drew D

   (2022) Electron Microscopy Data Bank

   ID EMD-24185. Glutathione binding to the plant AtAtm3 transporter and implications for the conformational coupling of ABC transporters.

   http://www.ebi.ac.uk/pdbe/entry/emdb/EMD-24185

1. 1. Altschul SF
   2. Madden TL
   3. Schäffer AA
   4. Zhang J
   5. Zhang Z
   6. Miller W
   7. Lipman DJ

   (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs

   Nucleic Acids Research 25:3389–3402.

   https://doi.org/10.1093/nar/25.17.3389

   • PubMed
   • Google Scholar
2. 1. Amadei A
   2. Linssen AB
   3. Berendsen HJ

   (1993) Essential dynamics of proteins

   Proteins 17:412–425.

   https://doi.org/10.1002/prot.340170408

   • PubMed
   • Google Scholar
3. 1. Baeza-Delgado C
   2. Marti-Renom MA
   3. Mingarro I

   (2013) Structure-based statistical analysis of transmembrane helices

   European Biophysics Journal 42:199–207.

   https://doi.org/10.1007/s00249-012-0813-9

   • PubMed
   • Google Scholar
4. 1. Barbieri A
   2. Thonghin N
   3. Shafi T
   4. Prince SM
   5. Collins RF
   6. Ford RC

   (2021) Structure of ABCB1/P-Glycoprotein in the Presence of the CFTR Potentiator Ivacaftor

   Membranes 11:923.

   https://doi.org/10.3390/membranes11120923

   • PubMed
   • Google Scholar
5. 1. Bennett BD
   2. Kimball EH
   3. Gao M
   4. Osterhout R
   5. Van Dien SJ
   6. Rabinowitz JD

   (2009) Absolute metabolite concentrations and implied enzyme active site occupancy in Escherichia coli

   Nature Chemical Biology 5:593–599.

   https://doi.org/10.1038/nchembio.186

   • PubMed
   • Google Scholar
6. 1. Bernard DG
   2. Cheng Y
   3. Zhao Y
   4. Balk J

   (2009) An allelic mutant series of ATM3 reveals its key role in the biogenesis of cytosolic iron-sulfur proteins in Arabidopsis

   Plant Physiology 151:590–602.

   https://doi.org/10.1104/pp.109.143651

   • PubMed
   • Google Scholar
7. 1. Chen S
   2. Sánchez-Fernández R
   3. Lyver ER
   4. Dancis A
   5. Rea PA

   (2007) Functional characterization of AtATM1, AtATM2, and AtATM3, a subfamily of Arabidopsis half-molecule ATP-binding cassette transporters implicated in iron homeostasis

   The Journal of Biological Chemistry 282:21561–21571.

   https://doi.org/10.1074/jbc.M702383200

   • PubMed
   • Google Scholar
8. 1. Chifflet S
   2. Torriglia A
   3. Chiesa R
   4. Tolosa S

   (1988) A method for the determination of inorganic phosphate in the presence of labile organic phosphate and high concentrations of protein: application to lens ATPases

   Analytical Biochemistry 168:1–4.

   https://doi.org/10.1016/0003-2697(88)90002-4

   • PubMed
   • Google Scholar
9. 1. Crans DC
   2. Smee JJ
   3. Gaidamauskas E
   4. Yang L

   (2004) The chemistry and biochemistry of vanadium and the biological activities exerted by vanadium compounds

   Chemical Reviews 104:849–902.

   https://doi.org/10.1021/cr020607t

   • PubMed
   • Google Scholar
10. 1. Creighton TE

    (1984)

    Disulfide bond formation in proteins

    Methods Enzymology 107:305–329.

    • Google Scholar
11. 1. Davies DR
    2. Hol WGJ

    (2004) The power of vanadate in crystallographic investigations of phosphoryl transfer enzymes

    FEBS Letters 577:315–321.

    https://doi.org/10.1016/j.febslet.2004.10.022

    • PubMed
    • Google Scholar
12. 1. Dawson RJP
    2. Locher KP

    (2006) Structure of a bacterial multidrug ABC transporter

    Nature 443:180–185.

    https://doi.org/10.1038/nature05155

    • PubMed
    • Google Scholar
13. 1. Dawson RJP
    2. Locher KP

    (2007) Structure of the multidrug ABC transporter Sav1866 from Staphylococcus aureus in complex with AMP-PNP

    FEBS Letters 581:935–938.

    https://doi.org/10.1016/j.febslet.2007.01.073

    • PubMed
    • Google Scholar
14. 1. Emsley P
    2. Lohkamp B
    3. Scott WG
    4. Cowtan K

    (2010) Features and development of Coot

    Acta Crystallographica. Section D, Biological Crystallography 66:486–501.

    https://doi.org/10.1107/S0907444910007493

    • PubMed
    • Google Scholar
15. 1. Fan C
    2. Kaiser JT
    3. Rees DC

    (2020) A structural framework for unidirectional transport by A bacterial ABC exporter

    PNAS 117:19228–19236.

    https://doi.org/10.1073/pnas.2006526117

    • PubMed
    • Google Scholar
16. 1. Hänelt I
    2. Wunnicke D
    3. Bordignon E
    4. Steinhoff H-J
    5. Slotboom DJ

    (2013) Conformational heterogeneity of the aspartate transporter Glt(Ph

    Nature Structural & Molecular Biology 20:210–214.

    https://doi.org/10.1038/nsmb.2471

    • PubMed
    • Google Scholar
17. 1. Hofmann S
    2. Januliene D
    3. Mehdipour AR
    4. Thomas C
    5. Stefan E
    6. Brüchert S
    7. Kuhn BT
    8. Geertsma ER
    9. Hummer G
    10. Tampé R
    11. Moeller A

    (2019) Conformation space of a heterodimeric ABC exporter under turnover conditions

    Nature 571:580–583.

    https://doi.org/10.1038/s41586-019-1391-0

    • PubMed
    • Google Scholar
18. 1. Hwang J-U
    2. Song W-Y
    3. Hong D
    4. Ko D
    5. Yamaoka Y
    6. Jang S
    7. Yim S
    8. Lee E
    9. Khare D
    10. Kim K
    11. Palmgren M
    12. Yoon HS
    13. Martinoia E
    14. Lee Y

    (2016) Plant ABC Transporters Enable Many Unique Aspects of a Terrestrial Plant’s Lifestyle

    Molecular Plant 9:338–355.

    https://doi.org/10.1016/j.molp.2016.02.003

    • PubMed
    • Google Scholar
19. 1. Johnson ZL
    2. Chen J

    (2017) Structural Basis of Substrate Recognition by the Multidrug Resistance Protein MRP1

    Cell 168:1075–1085.

    https://doi.org/10.1016/j.cell.2017.01.041

    • PubMed
    • Google Scholar
20. 1. Jumper J
    2. Evans R
    3. Pritzel A
    4. Green T
    5. Figurnov M
    6. Ronneberger O
    7. Tunyasuvunakool K
    8. Bates R
    9. Žídek A
    10. Potapenko A
    11. Bridgland A
    12. Meyer C
    13. Kohl SAA
    14. Ballard AJ
    15. Cowie A
    16. Romera-Paredes B
    17. Nikolov S
    18. Jain R
    19. Adler J
    20. Back T
    21. Petersen S
    22. Reiman D
    23. Clancy E
    24. Zielinski M
    25. Steinegger M
    26. Pacholska M
    27. Berghammer T
    28. Bodenstein S
    29. Silver D
    30. Vinyals O
    31. Senior AW
    32. Kavukcuoglu K
    33. Kohli P
    34. Hassabis D

    (2021) Highly accurate protein structure prediction with AlphaFold

    Nature 596:583–589.

    https://doi.org/10.1038/s41586-021-03819-2

    • PubMed
    • Google Scholar
21. 1. Kim DY
    2. Bovet L
    3. Kushnir S
    4. Noh EW
    5. Martinoia E
    6. Lee Y

    (2006) AtATM3 is involved in heavy metal resistance in Arabidopsis

    Plant Physiology 140:922–932.

    https://doi.org/10.1104/pp.105.074146

    • PubMed
    • Google Scholar
22. 1. Kispal G
    2. Csere P
    3. Guiard B
    4. Lill R

    (1997) The ABC transporter Atm1p is required for mitochondrial iron homeostasis

    FEBS Letters 418:346–350.

    https://doi.org/10.1016/s0014-5793(97)01414-2

    • PubMed
    • Google Scholar
23. 1. Kispal G
    2. Csere P
    3. Prohl C
    4. Lill R

    (1999) The mitochondrial proteins Atm1p and Nfs1p are essential for biogenesis of cytosolic Fe/S proteins

    The EMBO Journal 18:3981–3989.

    https://doi.org/10.1093/emboj/18.14.3981

    • PubMed
    • Google Scholar
24. 1. Krissinel E
    2. Henrick K

    (2007) Inference of macromolecular assemblies from crystalline state

    Journal of Molecular Biology 372:774–797.

    https://doi.org/10.1016/j.jmb.2007.05.022

    • PubMed
    • Google Scholar
25. 1. Kuhnke G
    2. Neumann K
    3. Mühlenhoff U
    4. Lill R

    (2006) Stimulation of the ATPase activity of the yeast mitochondrial ABC transporter Atm1p by thiol compounds

    Molecular Membrane Biology 23:173–184.

    https://doi.org/10.1080/09687860500473630

    • PubMed
    • Google Scholar
26. 1. Kushnir S
    2. Babiychuk E
    3. Storozhenko S
    4. Davey MW
    5. Papenbrock J
    6. De Rycke R
    7. Engler G
    8. Stephan UW
    9. Lange H
    10. Kispal G
    11. Lill R
    12. Van Montagu M

    (2001) A mutation of the mitochondrial ABC transporter Sta1 leads to dwarfism and chlorosis in the Arabidopsis mutant starik

    The Plant Cell 13:89–100.

    https://doi.org/10.1105/tpc.13.1.89

    • PubMed
    • Google Scholar
27. 1. Landau M
    2. Mayrose I
    3. Rosenberg Y
    4. Glaser F
    5. Martz E
    6. Pupko T
    7. Ben-Tal N

    (2005) ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures

    Nucleic Acids Research 33:W299–W302.

    https://doi.org/10.1093/nar/gki370

    • PubMed
    • Google Scholar
28. 1. Lee JY
    2. Yang JG
    3. Zhitnitsky D
    4. Lewinson O
    5. Rees DC

    (2014) Structural basis for heavy metal detoxification by an Atm1-type ABC exporter

    Science (New York, N.Y.) 343:1133–1136.

    https://doi.org/10.1126/science.1246489

    • PubMed
    • Google Scholar
29. 1. Leighton J
    2. Schatz G

    (1995)

    An ABC transporter in the mitochondrial inner membrane is required for normal growth of yeast

    The EMBO Journal 14:188–195.

    • PubMed
    • Google Scholar
30. 1. Liebschner D
    2. Afonine PV
    3. Baker ML
    4. Bunkóczi G
    5. Chen VB
    6. Croll TI
    7. Hintze B
    8. Hung LW
    9. Jain S
    10. McCoy AJ
    11. Moriarty NW
    12. Oeffner RD
    13. Poon BK
    14. Prisant MG
    15. Read RJ
    16. Richardson JS
    17. Richardson DC
    18. Sammito MD
    19. Sobolev OV
    20. Stockwell DH
    21. Terwilliger TC
    22. Urzhumtsev AG
    23. Videau LL
    24. Williams CJ
    25. Adams PD

    (2019) Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix

    Acta Crystallographica. Section D, Structural Biology 75:861–877.

    https://doi.org/10.1107/S2059798319011471

    • PubMed
    • Google Scholar
31. 1. Lill R
    2. Srinivasan V
    3. Mühlenhoff U

    (2014) The role of mitochondria in cytosolic-nuclear iron–sulfur protein biogenesis and in cellular iron regulation

    Current Opinion in Microbiology 22:111–119.

    https://doi.org/10.1016/j.mib.2014.09.015

    • PubMed
    • Google Scholar
32. 1. Marty L
    2. Bausewein D
    3. Müller C
    4. Bangash SAK
    5. Moseler A
    6. Schwarzländer M
    7. Müller-Schüssele SJ
    8. Zechmann B
    9. Riondet C
    10. Balk J
    11. Wirtz M
    12. Hell R
    13. Reichheld J-P
    14. Meyer AJ

    (2019) Arabidopsis glutathione reductase 2 is indispensable in plastids, while mitochondrial glutathione is safeguarded by additional reduction and transport systems

    The New Phytologist 224:1569–1584.

    https://doi.org/10.1111/nph.16086

    • PubMed
    • Google Scholar
33. 1. Mi W
    2. Li Y
    3. Yoon SH
    4. Ernst RK
    5. Walz T
    6. Liao M

    (2017) Structural basis of MsbA-mediated lipopolysaccharide transport

    Nature 549:233–237.

    https://doi.org/10.1038/nature23649

    • PubMed
    • Google Scholar
34. 1. Nagy P

    (2013) Kinetics and mechanisms of thiol-disulfide exchange covering direct substitution and thiol oxidation-mediated pathways

    Antioxidants & Redox Signaling 18:1623–1641.

    https://doi.org/10.1089/ars.2012.4973

    • PubMed
    • Google Scholar
35. 1. Pandey A
    2. Pain J
    3. Dziuba N
    4. Pandey AK
    5. Dancis A
    6. Lindahl PA
    7. Pain D

    (2018) Mitochondria Export Sulfur Species Required for Cytosolic tRNA Thiolation

    Cell Chemical Biology 25:738–748.

    https://doi.org/10.1016/j.chembiol.2018.04.002

    • PubMed
    • Google Scholar
36. 1. Pandey AK
    2. Pain J
    3. Dancis A
    4. Pain D

    (2019) Mitochondria export iron-sulfur and sulfur intermediates to the cytoplasm for iron-sulfur cluster assembly and tRNA thiolation in yeast

    The Journal of Biological Chemistry 294:9489–9502.

    https://doi.org/10.1074/jbc.RA119.008600

    • PubMed
    • Google Scholar
37. 1. Perez C
    2. Gerber S
    3. Boilevin J
    4. Bucher M
    5. Darbre T
    6. Aebi M
    7. Reymond JL
    8. Locher KP

    (2015) Structure and mechanism of an active lipid-linked oligosaccharide flippase

    Nature 524:433–438.

    https://doi.org/10.1038/nature14953

    • PubMed
    • Google Scholar
38. 1. Pettersen EF
    2. Goddard TD
    3. Huang CC
    4. Couch GS
    5. Greenblatt DM
    6. Meng EC
    7. Ferrin TE

    (2004) UCSF Chimera--a visualization system for exploratory research and analysis

    Journal of Computational Chemistry 25:1605–1612.

    https://doi.org/10.1002/jcc.20084

    • PubMed
    • Google Scholar
39. 1. Punjani A
    2. Rubinstein JL
    3. Fleet DJ
    4. Brubaker MA

    (2017) cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination

    Nature Methods 14:290–296.

    https://doi.org/10.1038/nmeth.4169

    • PubMed
    • Google Scholar
40. 1. Schaedler TA
    2. Thornton JD
    3. Kruse I
    4. Schwarzländer M
    5. Meyer AJ
    6. van Veen HW
    7. Balk J

    (2014) A conserved mitochondrial ATP-binding cassette transporter exports glutathione polysulfide for cytosolic metal cofactor assembly

    The Journal of Biological Chemistry 289:23264–23274.

    https://doi.org/10.1074/jbc.M114.553438

    • PubMed
    • Google Scholar
41. 1. Song G
    2. Zhang S
    3. Tian M
    4. Zhang L
    5. Guo R
    6. Zhuo W
    7. Yang M

    (2021) Molecular insights into the human ABCB6 transporter

    Cell Discovery 7:55.

    https://doi.org/10.1038/s41421-021-00284-z

    • PubMed
    • Google Scholar
42. 1. Srinivasan V
    2. Pierik AJ
    3. Lill R

    (2014) Crystal structures of nucleotide-free and glutathione-bound mitochondrial ABC transporter Atm1

    Science (New York, N.Y.) 343:1137–1140.

    https://doi.org/10.1126/science.1246729

    • PubMed
    • Google Scholar
43. 1. Tan YZ
    2. Baldwin PR
    3. Davis JH
    4. Williamson JR
    5. Potter CS
    6. Carragher B
    7. Lyumkis D

    (2017) Addressing preferred specimen orientation in single-particle cryo-EM through tilting

    Nature Methods 14:793–796.

    https://doi.org/10.1038/nmeth.4347

    • PubMed
    • Google Scholar
44. 1. Teschner J
    2. Lachmann N
    3. Schulze J
    4. Geisler M
    5. Selbach K
    6. Santamaria-Araujo J
    7. Balk J
    8. Mendel RR
    9. Bittner F

    (2010) A novel role for Arabidopsis mitochondrial ABC transporter ATM3 in molybdenum cofactor biosynthesis

    The Plant Cell 22:468–480.

    https://doi.org/10.1105/tpc.109.068478

    • PubMed
    • Google Scholar
45. 1. Thomas C
    2. Aller SG
    3. Beis K
    4. Carpenter EP
    5. Chang G
    6. Chen L
    7. Dassa E
    8. Dean M
    9. Duong Van Hoa F
    10. Ekiert D
    11. Ford R
    12. Gaudet R
    13. Gong X
    14. Holland IB
    15. Huang Y
    16. Kahne DK
    17. Kato H
    18. Koronakis V
    19. Koth CM
    20. Lee Y
    21. Lewinson O
    22. Lill R
    23. Martinoia E
    24. Murakami S
    25. Pinkett HW
    26. Poolman B
    27. Rosenbaum D
    28. Sarkadi B
    29. Schmitt L
    30. Schneider E
    31. Shi Y
    32. Shyng SL
    33. Slotboom DJ
    34. Tajkhorshid E
    35. Tieleman DP
    36. Ueda K
    37. Váradi A
    38. Wen PC
    39. Yan N
    40. Zhang P
    41. Zheng H
    42. Zimmer J
    43. Tampé R

    (2020) Structural and functional diversity calls for a new classification of ABC transporters

    FEBS Letters 594:3767–3775.

    https://doi.org/10.1002/1873-3468.13935

    • PubMed
    • Google Scholar
46. 1. Tian W
    2. Chen C
    3. Lei X
    4. Zhao J
    5. Liang J

    (2018) CASTp 3.0: computed atlas of surface topography of proteins

    Nucleic Acids Research 46:W363–W367.

    https://doi.org/10.1093/nar/gky473

    • PubMed
    • Google Scholar
47. 1. Varadi M
    2. Anyango S
    3. Deshpande M
    4. Nair S
    5. Natassia C
    6. Yordanova G
    7. Yuan D
    8. Stroe O
    9. Wood G
    10. Laydon A
    11. Žídek A
    12. Green T
    13. Tunyasuvunakool K
    14. Petersen S
    15. Jumper J
    16. Clancy E
    17. Green R
    18. Vora A
    19. Lutfi M
    20. Figurnov M
    21. Cowie A
    22. Hobbs N
    23. Kohli P
    24. Kleywegt G
    25. Birney E
    26. Hassabis D
    27. Velankar S

    (2022) AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models

    Nucleic Acids Research 50:D439–D444.

    https://doi.org/10.1093/nar/gkab1061

    • PubMed
    • Google Scholar
48. 1. Wang C
    2. Cao C
    3. Wang N
    4. Wang X
    5. Wang X
    6. Zhang XC

    (2020) Cryo-electron microscopy structure of human ABCB6 transporter

    Protein Science 29:2363–2374.

    https://doi.org/10.1002/pro.3960

    • PubMed
    • Google Scholar
49. 1. Ward A
    2. Reyes CL
    3. Yu J
    4. Roth CB
    5. Chang G

    (2007) Flexibility in the ABC transporter MsbA: Alternating access with a twist

    PNAS 104:19005–19010.

    https://doi.org/10.1073/pnas.0709388104

    • PubMed
    • Google Scholar
50. 1. Winn MD
    2. Ballard CC
    3. Cowtan KD
    4. Dodson EJ
    5. Emsley P
    6. Evans PR
    7. Keegan RM
    8. Krissinel EB
    9. Leslie AGW
    10. McCoy A
    11. McNicholas SJ
    12. Murshudov GN
    13. Pannu NS
    14. Potterton EA
    15. Powell HR
    16. Read RJ
    17. Vagin A
    18. Wilson KS

    (2011) Overview of the CCP4 suite and current developments

    Acta Crystallographica. Section D, Biological Crystallography 67:235–242.

    https://doi.org/10.1107/S0907444910045749

    • PubMed
    • Google Scholar
51. 1. Zheng SQ
    2. Palovcak E
    3. Armache JP
    4. Verba KA
    5. Cheng Y
    6. Agard DA

    (2017) MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy

    Nature Methods 14:331–332.

    https://doi.org/10.1038/nmeth.4193

    • PubMed
    • Google Scholar

Decision letter after peer review:

Thank you for submitting your article "Glutathione binding to the plant AtAtm3 transporter and implications for the conformational coupling of ABC transporters" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, including David Drew as Reviewing Editor and Reviewer #1, and the evaluation has been overseen by Kenton Swartz as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Markus A. Seeger (Reviewer #2).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

1) We think that the closed conformation of Atm3 should be stressed further as it's only recently that a closed conformation was observed. This paper is another chance to announce/emphasize the existence of the closed conformation and the mechanistic conclusions that can be reached from it.

2) Please clarify the advantage of using principle component analysis (PC1) as compared to simply measuring distances between residues at strategic positions e.g. between coupling helices, at the extracellular gate and (to distinguish between closed and occluded) at the kinking prolines in TM6. One disadvantage with only PC1 is that one loses information on local changes described by PC2 etc. For example, we would assume that an inter-TM6 distance measurement is able to distinguish outward-occluded from closed; this being the most important advantage over the principal component analysis, but PC1 fails to distinguish outward-facing/outward-occluded and closed states here (as evident from Figure 6a). Some further discussion on this point is warranted and perhaps useful to clarify for those not familiar with using this statistical approach to clustering different conformational states.

3) The authors state in the abstract that "The absence of structures containing both nucleotide and transport substrate suggests that this ternary complex exists only transiently during the transport cycle." This provokes the question as to whether cryo-EM analyses were carried out in the presence of ATP and GSSG, i.e. under hydrolysis conditions and in the presence of the substrate? It is well established that outward-facing trapped ABC exporters (via vanadate trapping or EtoQ mutations at the Walker B motif) do not bind substrates (e.g. see MRP1 and ABCB1), hence it is not at all surprising that ATP-vanadate trapped AtAtm3 had no GSSG bound. However, the ternary complex may have been captured under hydrolysis conditions.

4) In Figure 1, the ATPase activities of AtAtm3 are quite different in detergent and membrane environment, with more than 10-fold decrease in basal activity from detergent to POPC containing nanodiscs. Please clarify if you think that the detergent ATPase signal is an over-estimate (e.g., due to increased dynamics without lipids present) or the nanodisc is an underestimate due to using POPC lipids rather than those found in the mitochondria membrane, e.g., requirement for cardiolipin? In one of the papers you cited, Schaedler et. al, 2014, ATM3's activity is not stimulated by GSH versus in your data, 10mM GSH strongly stimulated AtAtm3. How would you explain this discrepancy?

5) There are a couple of structural studies on Atm transporters available to date and the authors showed some structural comparisons in Figure S3 and S5. However, it is still not clear exactly what are new structural insights gained with these structures. It could be helpful to compare the substrate binding sites side-by-side or include a cartoon representation of different functional states in different systems. The authors brought readers attention to "a ~20 amino acid loop between TM1 and TM2 of AtAtm3 that would be positioned in the intermembrane space and is absent from the structures of ScAtm1 and NaAtm1". Does this loop have any proposed functional roles? Is it present in other Atm or ABC transporters?

6) The cysteine-exclusion hypothesis of Atm3 is interesting, but lacks experimental validation. Please provide some additional experimental data to support this (e.g., is a mixed disulfide formed between GSH and cysteine(s) introduced into binding site based on locations in structural homologues) and/or appropriately tone down the statement to more closely reflect the presented data.

Reviewer #1 (Recommendations for the authors):

The cysteine-exclusion of Atm3 would benefit from some additional experimental analysis, such as the introduction of cysteine residues as found in the closest structural homologues harbouring cysteines and their impact on oxidised glutathione export.

Reviewer #2 (Recommendations for the authors):

1) If this work is looked at in the larger context of type I ABC exporters (and in particular in the structure series of NaAtm1 published by the same authors in 2020 in PNAS), one burning question comes to my mind: How universal and generally relevant is what the authors call the closed conformation? In Fan et al., PNAS, 2020, the Rees lab for the first time (at least to the best of my knowledge) described a closed conformation for NaAtm1. The closed conformation has the NBDs fully dimerized/closed and the extracellular gate is closed, but in contrast to the more often observed outward-occluded conformation (first mentioning by Beis lab for McjD and ever since seen several times), there is no occluded cavity, but TM6-kinking results in a collapse of the internal cavity (thus called closed conformation). The major punch-line of the NaAtm1 story was that NaAtm1 assumes an outward-occluded conformation in the presence of non-hydrolysable AMP-PNP (pre-hydrolysis state), but a closed conformation in the presence of ATP-vanadate (post-hydrolysis state), suggesting a pumping/expelling mechanism that is dependent driven by ATP hydrolysis. Currently, the structural differentiation of AMP-PNP-bound versus ATP-vanadate bound state in light of the cavity collapse has only been described for NaAtm1 thus far, but not for any other type I ABC exporter. In line with the findings for NaAtm1, the authors describe in this paper the ATP-vanadate bound structure of AtAtm3 to assume again the closed conformation with kinked TMH6 and collapsed internal cavity (though, a small internal cavity still seems to exist according to Figure 3c). The obvious question (which does not appear to be mentioned in the paper) is whether the authors have attempted to obtain the AMP-PNP bound structure of AtAtm3 and if so, whether the outward-occluded conformation featuring an internal cavity was seen the way it was observed for NaAtm1?

Overall, I feel there is still limited appreciation in the ABC transporter field for the interesting and unique closed conformation, likely owing to the fact that the discovery of the first closed conformation is comparatively recent. This paper is another chance to announce/emphasize the existence of the closed conformation. Ideally, this is addressed by further structures (though that's only a reasonable request if it is straight-forward to achieve). Alternative possibilities are biochemical experiments (e.g. cross-linking, or DEER).

2) To stick to the "conformation business", I wonder whether the principal component analysis is superior to analyses simply measuring distances between residues at strategic positions (e.g. between coupling helices, at the extracellular gate and (to distinguish between closed and occluded) at the kinking prolines in TM6. Such distance measurements together with cavity determinations are in my view sufficient to distinguish between the relevant conformations and are more intuitive to grasp (see also PMID: 33098660).

In particular the distance between the coupling helices informs about NBD separation/completeness of NBD closure, hence allowing to differentiate between inward-facing/inward-occluded states (or the special case of the unlock return state of TmrAB) and outward-facing, outward-occluded and (as relevant here) the closed conformation. The closed conformation seems to correspond to an outward-occluded conformation when it comes the degree of intracellular/extracellular gate closure, but does not feature an enclosed cavity. I would assume that an inter-TM6 distance measurement is able to distinguish outward-occluded from closed; this being the most important advantage over the principal component analysis, which fails to distinguish outward-facing/outward-occluded and closed (as evident from Figure 6a). And finally, the degree of extracellular gate opening is of general interest when analysing different conformations with fully closed NBDs.

Unless I have overlooked something (which is well possible), I would consider such simple distance measurements as more informative than the principal component analysis.

3) The authors state in the abstract that "The absence of structures containing both nucleotide and transport substrate suggests that this ternary complex exists only transiently during the transport cycle." This provokes the question whether cryo-EM analyses were carried out in the presence of ATP and GSSG, i.e. under hydrolysis conditions and in the presence of the substrate. It is well established in the field that outward-facing trapped ABC exporters (via vanadate trapping or EtoQ mutations at the Walker B motif) do not bind substrates (e.g. see MRP1 and ABCB1), hence it is not at all surprising that ATP-vanadate trapped AtAtm3 had no GSSG bound. However, the ternary complex may have been captured under hydrolysis conditions.

Reviewer #3 (Recommendations for the authors):

1) In Figure 3, it could be helpful to label the functional states in panels in a, b, c, d, f, g, h. If the distances between 2 NBDs can be measured and labeled in panel a and b (apo and GSSG bound form), it would be more helpful for the non-ABC transporter audience to visualize the conformational change globally.

2) In Figure 6, the color codes in the legend do not match with those on the graph.

3) In Figure S4, there are 2 high-resolution reconstructions of AtAtm3 in an inward-facing conformation with GSSG bound (3.57A and 3.68A respectively), are they different structurally?

Essential revisions:

1) We think that the closed conformation of Atm3 should be stressed further as it's only recently that a closed conformation was observed. This paper is another chance to announce/emphasize the existence of the closed conformation and the mechanistic conclusions that can be reached from it.

Thank you for this comment. We have amplified our discussion of the closed conformation on page 10. We also included a statement in the abstract of the different conformational states we observed for AtAtm3, including the closed conformation.

2) Please clarify the advantage of using principle component analysis (PC1) as compared to simply measuring distances between residues at strategic positions e.g. between coupling helices, at the extracellular gate and (to distinguish between closed and occluded) at the kinking prolines in TM6. One disadvantage with only PC1 is that one loses information on local changes described by PC2 etc. For example, we would assume that an inter-TM6 distance measurement is able to distinguish outward-occluded from closed; this being the most important advantage over the principal component analysis, but PC1 fails to distinguish outward-facing/outward-occluded and closed states here (as evident from Figure 6a). Some further discussion on this point is warranted and perhaps useful to clarify for those not familiar with using this statistical approach to clustering different conformational states.

We utilized principal component analysis (PCA) to project the transporter structures onto a single conformational axis to distribute these structures so that similar structures will generally be positioned more closely together. Of course, since only the dominant component is used, this analysis simplifies the conformational richness of ABC transporters. We agree that the principal component is neither a completely accurate nor unique measure of conformational similarity, and in particular, alternative metrics could be employed such as measuring distances between particular pairs of residues. A comparison is presented in “Figure 6 supplement 1” comparing the dominant component to several alternative distance measures; these are generally anti-correlated, with the interesting exception of the distances between a pair of residues in the outer loops that, not surprisingly, do a much better job of discriminating between outward conformations than the PCA, while being less sensitive to differences between inward conformations. As the quantitative value of the principal component was only utilized in this analysis to separate structures along a single conformational analysis (Figure 6), we felt it was appropriate to use the PCA for that purpose in this analysis. This has been noted on page 14 of the text.

3) The authors state in the abstract that "The absence of structures containing both nucleotide and transport substrate suggests that this ternary complex exists only transiently during the transport cycle." This provokes the question as to whether cryo-EM analyses were carried out in the presence of ATP and GSSG, i.e. under hydrolysis conditions and in the presence of the substrate? It is well established that outward-facing trapped ABC exporters (via vanadate trapping or EtoQ mutations at the Walker B motif) do not bind substrates (e.g. see MRP1 and ABCB1), hence it is not at all surprising that ATP-vanadate trapped AtAtm3 had no GSSG bound. However, the ternary complex may have been captured under hydrolysis conditions.

We did try to capture this particular structure under turnover conditions in the presence of MgATP and the substrate GSSG, but unfortunately, we only observed the inward-facing conformation in these samples.

4) In Figure 1, the ATPase activities of AtAtm3 are quite different in detergent and membrane environment, with more than 10-fold decrease in basal activity from detergent to POPC containing nanodiscs. Please clarify if you think that the detergent ATPase signal is an over-estimate (e.g., due to increased dynamics without lipids present) or the nanodisc is an underestimate due to using POPC lipids rather than those found in the mitochondria membrane, e.g., requirement for cardiolipin? In one of the papers you cited, Schaedler et. al, 2014, ATM3's activity is not stimulated by GSH versus in your data, 10mM GSH strongly stimulated AtAtm3. How would you explain this discrepancy?

In our experience, the ATPase rates of ABC transporters are highly dependent on the precise solubilization conditions, particularly for detergents as some detergents/detergent combinations are stimulatory, while others are inhibitory. In addition, the POPC used in the nanodiscs might not be the best substitute of the natural lipids that are found in the mitochondrial membrane. Given the variability that has been observed in ATPase activities for different solubilization conditions, we do not know which condition more likely reflects the true physiological activity. (As an aside, we would have thought that basal ATPase rate should be zero in the absence of transported substrate to avoid uncoupled (and presumably wasteful) ATPase activity, but this has been rarely reported).

The Schaedler paper used GSH at two concentrations, 1.7 and 3.3 mM, to test the stimulation of ATPase activity. In our study, we used 10 mM GSSG for the ATPase activity assay to be consistent with our previous studies on NaAtm1. This discrepancy presumably reflects concentration differences utilized in these studies, and other experimental conditions including the nature of the protein construct. Different expression systems were used in the two studies, Schaedler et al., employed L. lactis, while we used E. coli; it is possible that different lipids co-purified with the proteins that could influence the ATPase activity. In addition, we used a truncation with an 80 residue deletion, while Schaedler et al., utilized a 60-residue truncation; our initial characterization of various constructs showed that the 80-residue deletion had a higher ATPase activity, although we did not characterize these differences in detail – certainly not for publication. While both studies used AtAtm3 solubilized in DDM, it is possible that there are still differences in the solubilization conditions reflecting the precise way in which detergents are introduced. We have added a sentence on page 9 noting this discrepancy in the GSH stimulation results, without ascribing a particular mechanism.

5) There are a couple of structural studies on Atm transporters available to date and the authors showed some structural comparisons in Figure S3 and S5. However, it is still not clear exactly what are new structural insights gained with these structures. It could be helpful to compare the substrate binding sites side-by-side or include a cartoon representation of different functional states in different systems. The authors brought readers attention to "a ~20 amino acid loop between TM1 and TM2 of AtAtm3 that would be positioned in the intermembrane space and is absent from the structures of ScAtm1 and NaAtm1". Does this loop have any proposed functional roles? Is it present in other Atm or ABC transporters?

We included a new figure, “Figure 2 supplement 3”, comparing different homologs, AtAtm3, human ABCB7 (an Alphafold model), human ABCB6, yeast Atm1 and prokaryotic Atm1. Among these five structures, only AtAtm3 presents a loop between TM1 and TM2. A quick protein sequence blast in PubMed showed the presence of this loop in many other plant atm transporters. This long loop has not been observed in other structure we know so far, but an external helix was observed for PglK, a lipid-linked oligosaccharide flippase, and this loop has been implicated in substrate interaction (Perez et al., Nature 524, 433 (2015); doi: 10.1038/nature14953)

6) The cysteine-exclusion hypothesis of Atm3 is interesting, but lacks experimental validation. Please provide some additional experimental data to support this (e.g., is a mixed disulfide formed between GSH and cysteine(s) introduced into binding site based on locations in structural homologues) and/or appropriately tone down the statement to more closely reflect the presented data.

We agree this would be an important test of our hypothesis. We did introduce cysteine mutations in NaAtm1 at position M317, M320 and the double mutations in TM6 to test this model. Unfortunately, there was no expression of any of these mutants. We recognize that a negative result like this can have many explanations, and that testing other residues would be appropriate (including in AtAtm3). At the time we were doing these studies, we were working under Covid restrictions and rather than testing additional residues, we focused on trapping additional conformational states.

Our apologies for the over-inflated claims about the mixed disulfides– we tried hard to appropriately describe our analysis, recognizing that this was based on inferences from sequence and structural comparisons (no cysteines in the glutathione binding site), but not from a direct test. We have reworded the abstract and page 13 to try to more accurately reflect the presented data.

Reviewer #1 (Recommendations for the authors):

The cysteine-exclusion of Atm3 would benefit from some additional experimental analysis, such as the introduction of cysteine residues as found in the closest structural homologues harbouring cysteines and their impact on oxidised glutathione export.

Please see our response to essential revision 6.

Reviewer #2 (Recommendations for the authors):

1) If this work is looked at in the larger context of type I ABC exporters (and in particular in the structure series of NaAtm1 published by the same authors in 2020 in PNAS), one burning question comes to my mind: How universal and generally relevant is what the authors call the closed conformation? In Fan et al., PNAS, 2020, the Rees lab for the first time (at least to the best of my knowledge) described a closed conformation for NaAtm1. The closed conformation has the NBDs fully dimerized/closed and the extracellular gate is closed, but in contrast to the more often observed outward-occluded conformation (first mentioning by Beis lab for McjD and ever since seen several times), there is no occluded cavity, but TM6-kinking results in a collapse of the internal cavity (thus called closed conformation). The major punch-line of the NaAtm1 story was that NaAtm1 assumes an outward-occluded conformation in the presence of non-hydrolysable AMP-PNP (pre-hydrolysis state), but a closed conformation in the presence of ATP-vanadate (post-hydrolysis state), suggesting a pumping/expelling mechanism that is dependent driven by ATP hydrolysis. Currently, the structural differentiation of AMP-PNP-bound versus ATP-vanadate bound state in light of the cavity collapse has only been described for NaAtm1 thus far, but not for any other type I ABC exporter. In line with the findings for NaAtm1, the authors describe in this paper the ATP-vanadate bound structure of AtAtm3 to assume again the closed conformation with kinked TMH6 and collapsed internal cavity (though, a small internal cavity still seems to exist according to Figure 3c). The obvious question (which does not appear to be mentioned in the paper) is whether the authors have attempted to obtain the AMP-PNP bound structure of AtAtm3 and if so, whether the outward-occluded conformation featuring an internal cavity was seen the way it was observed for NaAtm1?

Overall, I feel there is still limited appreciation in the ABC transporter field for the interesting and unique closed conformation, likely owing to the fact that the discovery of the first closed conformation is comparatively recent. This paper is another chance to announce/emphasize the existence of the closed conformation. Ideally, this is addressed by further structures (though that's only a reasonable request if it is straight-forward to achieve). Alternative possibilities are biochemical experiments (e.g. cross-linking, or DEER).

Thank you – we have discussed this situation in our response to essential revision 1.

2) To stick to the "conformation business", I wonder whether the principal component analysis is superior to analyses simply measuring distances between residues at strategic positions (e.g. between coupling helices, at the extracellular gate and (to distinguish between closed and occluded) at the kinking prolines in TM6. Such distance measurements together with cavity determinations are in my view sufficient to distinguish between the relevant conformations and are more intuitive to grasp (see also PMID: 33098660).

In particular the distance between the coupling helices informs about NBD separation/completeness of NBD closure, hence allowing to differentiate between inward-facing/inward-occluded states (or the special case of the unlock return state of TmrAB) and outward-facing, outward-occluded and (as relevant here) the closed conformation. The closed conformation seems to correspond to an outward-occluded conformation when it comes the degree of intracellular/extracellular gate closure, but does not feature an enclosed cavity. I would assume that an inter-TM6 distance measurement is able to distinguish outward-occluded from closed; this being the most important advantage over the principal component analysis, which fails to distinguish outward-facing/outward-occluded and closed (as evident from Figure 6a). And finally, the degree of extracellular gate opening is of general interest when analysing different conformations with fully closed NBDs.

Unless I have overlooked something (which is well possible), I would consider such simple distance measurements as more informative than the principal component analysis.

Our motivation was to provide a single parameter that could be used to project transporter structures onto a single conformational axis. The principal component of the PCA provides this descriptor, but we agree that this is not a unique description of the protein conformation. We have explored the use of several alternative metrics based on distances between pairs of residues – some are highly (anti-) correlated with the PCA analysis, while others (like the distance between residues in the outer loops) better discriminate between outward facing conformations, but at the expense of inward-facing conformations. For the purposes of our qualitative analysis, that did not seem to be a critical distinction. Please see additional comments in our response to essential revision 2.

3) The authors state in the abstract that "The absence of structures containing both nucleotide and transport substrate suggests that this ternary complex exists only transiently during the transport cycle." This provokes the question whether cryo-EM analyses were carried out in the presence of ATP and GSSG, i.e. under hydrolysis conditions and in the presence of the substrate. It is well established in the field that outward-facing trapped ABC exporters (via vanadate trapping or EtoQ mutations at the Walker B motif) do not bind substrates (e.g. see MRP1 and ABCB1), hence it is not at all surprising that ATP-vanadate trapped AtAtm3 had no GSSG bound. However, the ternary complex may have been captured under hydrolysis conditions.

Please see our response to essential revision 3.

Reviewer #3 (Recommendations for the authors):

1) In Figure 3, it could be helpful to label the functional states in panels in a, b, c, d, f, g, h. If the distances between 2 NBDs can be measured and labeled in panel a and b (apo and GSSG bound form), it would be more helpful for the non-ABC transporter audience to visualize the conformational change globally.

Thank you – please see our response to essential revision.

2) In Figure 6, the color codes in the legend do not match with those on the graph.

Thank you – please seen our response to essential revision.

3) In Figure S4, there are 2 high-resolution reconstructions of AtAtm3 in an inward-facing conformation with GSSG bound (3.57A and 3.68A respectively), are they different structurally?

Thank you – please seen our response to essential revision.

Author details

1. Chengcheng Fan

   Division of Chemistry and Chemical Engineering, Howard Hughes Medical Institute, California Institute of Technology, Pasadena, United States

   Present address

   Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, United States

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-4213-5758

2. Douglas C Rees

   Division of Chemistry and Chemical Engineering, Howard Hughes Medical Institute, California Institute of Technology, Pasadena, United States

   Contribution

   Conceptualization, Funding acquisition, Methodology, Project administration, Software, Supervision, Writing – review and editing

   For correspondence

   dcrees@caltech.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-4073-1185

• 1,213

  Page views

• 260

  Downloads

• 4

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.