Background
Members of the organic anion transporter (OAT) family, including
OAT1, are expressed on the epithelial membrane of the kidney,
liver, brain, intestine, and placenta. OAT1 regulates the transport
of organic anion drugs from the blood into kidney epithelial
cells by utilizing the α-ketoglutarate (α-KG) gradient across the
membrane established by the tricarboxylic acid (TCA) cycle.The organic anion transporter 1 (OAT1) also plays a key role in excreting waste from organic drug metabolism and
contributes significantly to drug-drug interactions and drug disposition. However, the structural basis of specific
substrate and inhibitor transport by human OAT1 (hOAT1) has remained elusive. Here are four
cryo-electron microscopy (cryo-EM) structures of hOAT1 in its inward-facing conformation: the apo
form, the substrate (olmesartan)-bound form with different anions, and the inhibitor (probenecid)-bound
form.
Cryo-EM structure of hOAT1
Click the green links below to change the molecular scene. Drag to rotate. Zoom the molecule with your mouse wheel, or Shift-Drag up/down.
|
Human
OAT1 adopts an inward-facing conformation in a membrane. OAT1 consist of structural features
including intracellular helices domain (ICD),
extracellular domain (ECD), N-lobe helices
(TM1-6), and C-lobe helices (TM7-12). (right) The
border of the binding cavity (described in solvent
exclude-surface) is formed by residues N35,
Y230, Y353, Y354 (upper), and M207 and F442
(lower).
Key Structural Characteristics:
- Adopts the classic Major Facilitator Superfamily (MFS) fold.
- Comprises 12 transmembrane helices (TMs 1-12).
- Exhibits pseudo-two-fold symmetry, divided into an N-lobe (TMs 1-6) and a C-lobe (TMs 7-12).
- The cavity is located between the N-lobe (formed by TM1, TM2, TM4, TM5) and the C-lobe (formed by TM7, TM8, TM10, TM11).
- It possesses a positively charged electrostatic environment, which explains its strong preference for transporting anionic substrates.
- The cavity is lined by 29 residues, forming a hydrophobic and aromatic-rich environment.
- Cavity Borders and Cytosolic Gate:
- The top border (extracellular side) of the cavity is formed by residues including N35, Y230, Y353, and Y354.
- The bottom border (cytosolic side) features a narrow "thin bottom gate" formed by residues M207 and F442. The interaction between these two residues splits the cytosolic entrance into two distinct pathways:
- Path A: Located between TM2 and TM11.
- Path B: Located between TM5 and TM8.
- In the apo state, the transporter is in a relaxed, inward-open conformation, providing access for substrates from the cytoplasm.
- The structure serves as a baseline for understanding the conformational changes that occur upon substrate or inhibitor binding.
Olmesartan recognition by hOAT1
The structural and functional analysis of hOAT1 in complex with the high-affinity antihypertensive drug olmesartan provides a detailed blueprint for substrate specificity and binding.
1. Binding Location and Pose
- Olmesartan binds within the central cavity of hOAT1 in an inward-facing conformation.
- It occupies Site 3 of the binding pocket, which is the primary polyspecific site for anionic substrates.
- The drug adopts a diagonal orientation relative to the membrane plane, a pose that requires more space than the smaller inhibitor probenecid. This orientation is similar to its conformation when bound to the angiotensin receptor.
2. Key Interacting Residues
Olmesartan is surrounded by residues from multiple transmembrane helices (TM1, TM4, TM5, TM7, TM10, TM11) within a 5 Å distance. The critical interactions involve:
- Aromatic and Hydrophobic Cage:
- The biphenyl group of olmesartan is nestled near residue F438.
- The tetrazole ring is positioned between the bottom-gate residues M207 and F442.
- The imidazole moiety is located close to Y354.
- Upon olmesartan binding, the side chain of Y230 undergoes a vertical rotation to accommodate and interact with the substrate.
- Mutagenesis studies confirm its importance: the Y230F mutation increased the IC₅₀ for olmesartan inhibition from 845.3 nM (Wild Type) to 2.36 µM, indicating a reduction in binding affinity.
- The Bottom Gate Residues (M207 and F442):
- These residues are crucial for high-affinity olmesartan binding.
- The M207A mutant caused a 4-fold reduction in affinity (IC₅₀ = 3.78 µM).
- The F442A mutant caused a dramatic 12-fold reduction in affinity (IC₅₀ = 10.32 µM).
- This suggests these residues not only form a gate but also directly interact with large, transportable substrates like olmesartan.
3. Chloride Ion Coordination is Essential
A key finding is the role of a chloride ion in stabilizing the olmesartan-bound state.
- The Chloride-Binding Site: A chloride ion (or bromide, used for confirmation) is observed coordinated between residues S203, Y230, and R466.
- Indirect Role of S203: While S203 does not directly contact olmesartan, it is critical for chloride coordination. This is a major species-specific difference, as rat OAT1 has an alanine at this position.
- Functional Evidence of Chloride Dependence:
- The IC₅₀ of olmesartan is 2.01 µM in chloride-rich conditions but improves to 0.91 µM in chloride-depleted conditions, suggesting a more complex relationship where chloride may facilitate transport.
- The S203A mutant shows a severe ~5-fold reduction in olmesartan binding affinity specifically in the presence of chloride (IC₅₀: WT = 2.47 µM; S203A = 29.52 µM).
- The S203A-Y230F double mutant has an even more profound effect, increasing the IC₅₀ to 93.30 µM in chloride conditions, highlighting their synergistic role in chloride-dependent substrate binding.
The OmcS monomer has .
Alpha Helices,
310 Helices,
Beta Strands ,
Loops .
The structure assigned by the authors is 77% loops; Jmol objectively assigns 82% loops. The authors assigned 10% alpha helices, 7% 310 helices, and 6% beta strands.
The OmcS structure determined by Filman et al. [2]was very similar, with 80% loops assigned by the authors (86% by Jmol), having only 3% beta strand but otherwise very similar. We compared OmcS with three other c-type multi-heme cytochrome crystal structures: 1ofw, 3ucp, and 3ov0 had 45%, 49%, and 60% loops respectively.
Mechanism of OAT1 inhibition by probenecid
The cryo-EM structure of hOAT1 bound to the classic inhibitor probenecid reveals a dual-mechanism of action that goes beyond simple competition, effectively arresting the transporter in a restricted state.
1. Binding Mode and Direct Competition
- Probenecid binds at the top of the central cavity, parallel to the membrane plane.
- Its binding site overlaps with both Site 1 (partially) and Site 3.
- It engages in specific, high-affinity interactions with key residues:
- K382 on TM8 forms a hydrogen bond with the carboxylate group of probenecid.
- Y354 on TM7 forms a hydrogen bond with its sulfonyl group.
- Crucially, K382 is also the residue that interacts with the counter-substrate α-ketoglutarate (α-KG), establishing a direct competitive inhibition mechanism by blocking α-KG binding.
2. Conformational Arrest and Cytoplasmic Path Blockage
The primary inhibitory mechanism is a probenecid-induced conformational change that physically blocks substrate access and exit.
- Constriction of the Binding Pocket: Compared to the apo state, the cytoplasmic opening of the binding pocket narrows from ~15 Å to ~12 Å in the probenecid-bound state.
- Dual-Pathway Blockade: The cytosolic entrance is split into two paths. Probenecid binding critically affects both:
- Path A (between TM2 and TM11) is narrowed from ~5 Å to ~4 Å.
- Path B (between TM5 and TM8) is completely blocked.
This structural rearrangement is caused by a slight inward movement of the cytoplasmic ends of TM5, TM8, TM10, and TM11 toward the binding pocket.
3. Locked Conformation
By constricting the cytoplasmic access routes, probenecid does not just compete for the substrate-binding site; it stabilizes the transporter in an apo-like, inward-facing conformation that is inaccessible to cytosolic substrates. This prevents the entry of new substrates and likely traps the transporter in this non-functional state, effectively "locking" it and preventing the conformational changes necessary for the transport cycle.
Each OmcS monomer :
C
O
N
Fe.
The hemes are arranged in parallel-displaced pairs. Each pair is orthogonal to the next pair.
The , which likely contributes to the stability of the filament. More importantly, this produces a . This continuous chain of hemes is believed to be the basis of the electrical conductivity.
Cysteine Anchors
Each heme is , which form thioether bonds with the heme vinyl groups (opposite the heme carboxyls):
C
O
N
S
Fe.
12 CxxCH motifs in the OmcS sequence anchor the 6 hemes within each OmcS chain.
Histidine to Iron
Each heme , in addition to the four heme nitrogens. The iron of heme 5 (the next to last heme at the carboxy end of the chain) is bound to His 332 from its own chain (Chain A), and His 16 in the N-terminal "bulge" of the next protein chain (Chain B) in the filament. This inter-chain histidine-iron bond is undoubtedly important in strengthening the monomer-monomer interfaces in the filament. The histidines bound to hemes 1, 2, 3, 4, and 6 are all in the same protein chain that contains those hemes.
Salt Bridges
Using a 4.0 Å cutoff, 6ef8 has 7 salt bridges between amino acid sidechains (not shown). One of these, (Chain A, Chain B,
O,
N),
is between protein chains, further strengthening the interfaces between monomers in the filament. (These opposing charges are 4.9 Å apart in 6nef.)
The amino-terminal NH3+ on Phe 1 forms a salt bridge with one carboxy of heme 2 (HEC503; 3.65 Å; not shown).
Each heme has close to zero net charge, since the two carboxyls are compenated by Fe++. About half of the heme carboxyls are on the surface, exposed to water (not shown). Several of the heme carboxyls form salt bridges with sidechains of arginine or lysine (not shown).
Buried Cations
The . None have anions within 5 Å (not shown). The sidechain nitrogens of Arg333 and Arg344 touch each other (3.0 Å). These characteristics are confirmed in 6nef. The presence of these cations deep within OmcS is plausible, since proteins of this size have, on average, several buried charges[3][4]. Moreover, on average from many proteins, more than half of all arginine guanidiniums are buried[3]. Burying charge seems to be an important factor in how evolution regulates protein stability[3][4].
The buried contact between two usually-cationic sidechains of Arg333 and Arg344 is also plausible because, when buried, the positive charge of the guanidinium group can be greatly diminished due to dehydration and nearby positive charges[3]. Although hydrated guanidinium retains more than half of its charge when the pH is below ~12 (its intrinsic pKa[3]), dehydration due to burial decreases the pKa. Furthermore, the samples for cryo-electron microscopy were prepared at pH 10.5[1] (despite the pH being incorrectly stated as 7.0 in REMARK 245 of the PDB file).
Other Findings & Conclusions
References for the assertions below are cited in the journal publication[1].
Seamless micrometer-long polymerization of hundreds of cytochromes is without precedent, to the knowledge of the authors. The filaments whose structure was determined here were obtained from electrode-grown cells. However, fumarate-grown cells produced filaments with similar sinusoidal morphology. The purified OmcS filaments have morphology and power spectra similar to cell-attached filaments previously thought to be type IV pili. Direct current electrical conductivity of individual wild type ~4 nm OmcS filaments was confirmed, and was comparable to previously reported filament conductivity values.
Cells with the omcS gene deleted (ΔomcS) produced thinner (~1.7 nm) filaments that were smooth (not sinusoidal) and had electrical conductivity >100-fold lower than the OmcS filaments. ΔomcS cells can produce electrically conductive biofilms, but that conductivity might well depend on filaments of OmcZ, whose expression is known to increase in ΔomcS cells.
Previous studies showed that PilA is required for export of OmcS. However, PilA was not found in the structure of the OmcS nanowires studied here. Thus, PilA appears to be required for production of OmcS nanowires, but not to be a structural component of those nanowires.