Interdomain-linkers control conformational transitions in the SLC23 elevator transporter UraA

Abstract

Uptake of nucleobases and ascorbate is an essential process in all living organisms mediated by SLC23 transport proteins. These transmembrane carriers operate via the elevator alternating-access mechanism, and are composed of two rigid domains whose relative motion drives transport. The lack of large conformational changes within these domains suggests that the interdomain-linkers act as flexible tethers. Here, we show that interdomain-linkers are not mere tethers, but have a key regulatory role in dictating the conformational space of the transporter and defining the rotation axis of the mobile transport domain. By resolving a wide inward-open conformation of the SLC23 elevator transporter UraA and combining biochemical studies using a synthetic nanobody as conformational probe with hydrogen-deuterium exchange mass spectrometry, we demonstrate that interdomain-linkers control the function of transport proteins by influencing substrate affinity and transport rate. These findings open the possibility to allosterically modulate the activity of elevator proteins by targeting their linkers.

Fig. 1. Functional analysis of inter-domain linker mutants in UraA.

A Dimeric UraA structure (PDB:5XLS) with scaffold and transport domain in green or purple, respectively, and the interdomain-linkers in blue. B Topology plot of UraA (center) with color code as in (A). Upper and lower panels depict the sequence conservation of the extracellular and cytoplasmic interdomain-linkers in the UraA-subfamily, respectively, shown as sequence logo. C Absolute differences in dihedral angles between interdomain-linker residues in UraA_IO (PDB:3QE7) and UraA_OCC (PDB:5XLS). Δφ and Δψ are colored dark and light gray, respectively, and glycine and proline residues flanking the spacer helices are highlighted in red and orange, respectively. D In vivo uptake rates of [³H]-uracil by wild type UraA and interdomain-linker mutants upon expression in E. coli BW25113(ΔuraA) with technical replicates shown as scatter plot and derived mean values ± SEM as bars (control: n = 6, WT: n = 4, G112P, P121G, G320P, and P330G: n = 3). A UraA variant with three alanine substitutions in the substrate binding site (E241A, H245A, and E290A) served as negative control. E Representative size-exclusion chromatograms of decylmaltoside-solubilized UraA variants in the absence of substrate. F Melting temperature of interdomain-linker mutants as determined by differential scanning fluorimetry in absence and presence of 1 mM uracil with technical replicates shown as scatter plot and derived mean values ± SEM as bars (n = 3 for all samples).

Fig. 2. Structural characterization of the Sy45 binding site.

A Structure of UraA(G320P)-Sy45 complex viewed from within the membrane. Transport and scaffold domain are purple and green, respectively. Interdomain-linkers are blue. Sy45 is colored gray and CDR1, CDR2, and CDR3 are yellow, orange, and red, respectively. B View of the UraA(G320P)-Sy45 complex in the plane of the membrane on the scaffold domain showing the access of cytoplasmic water to the substrate binding site. C View of UraA(G320P) from the cytoplasmic side. Sy45 has been left out for clarity. Regions of UraA within 4 Å of CDR1, CDR2, or CDR3 are colored yellow, orange, and red, respectively.

Fig. 3. Conformational transitions in UraA are mediated by interdomain linkers.

A Cartoon representation of single UraA_WIO and UraA_OCC protomers with scaffold and transport domain in green or purple, respectively, and the spacer helices in blue. B Surface representation of single UraA_WIO and UraA_OCC protomers clipped through the substrate binding sites. The location of the substrate is indicated by a red asterisk. C Conformational changes in TM4 to TM5 and the cytoplasmic interdomain-linker (left panel) and TM11 to TM12 and the external interdomain-linker (right panel). View is from within the plane of the membrane (top panels) or on the external side of the membrane (bottom panels). For superimposition of the structures TM5 and TM12 were used.

Fig. 4. Conformational equilibrium in UraA is affected by interdomain-linker mutants with an impact on substrate binding.

A The reporter cysteine mutants L34C and V248C are surface-exposed in an outward-open model or wide inward-open structure of UraA, respectively. B Degree of cysteine alkylation of L34C (red, outward-open reporter) and V248C (blue, wide inward-open reporter) in the presence and absence of Sy45. UraA variants were alkylated with mPEG5k for one hour and alkylation was quantified by mobility shift in SDS-PAGE and densitometry analysis. Data was normalized to the maximal alkylation obtained in the presence of 1% SDS. Shown are three technical replicates as scatter plot and derived mean values ± SEM as bars. A two-tailed, unpaired t-test was performed to test for statistical significance, The respective p-values are indicated. C Thermal stability of inter-domain linker mutants analyzed by differential scanning fluorimetry in absence (gray, n = 3) and presence (blue, n = 4) of Sy45 with technical replicates shown as scatter plot and derived mean values ± SEM as bars. A two-tailed, unpaired t-test was performed to test for statistical significance (p  = 0.0153). D Scintillation proximity assay of UraA wild type, G320P and P330G in absence and presence of Sy45 as indicated with three technical replicates shown as gray scatter. Mean values and derived SEM shown as black error bars. Scintillation data was fitted in Origin with a binding curve for homologous competition to calculate the dissociation constants.

Fig. 5. Hydrogen-deuterium exchange mass spectrometry analysis reveals changes in the conformational dynamics of UraA.

A HDX of wild type UraA in the presence and absence of 100 µM uracil. Transport and scaffold domain are oriented as if the protein was folded open as indicated in the central panel. The interdomain-linkers are shown in both domains as a reference point. The circle on the scaffold domain represents the position of the substrate in the opposing transport domain assuming the UraA_OCC conformation. B Substrate-dependent HDX for UraA(G320P) and UraA(P330G). C Differential HDX resulting from the G320P (top panel) and P330G (lower panel) mutations in the absence of ligand compared to wild type UraA. The color range representing the relative deuterium uptake, ranging from red (decreased uptake) over white (no change) to blue (increased deuterium uptake), applies to all panels. Gray stretches indicate regions of not identified amino acids.

Fig. 6. Role of the interdomain-linker in the transport mechanics.

A Theoretical profile of the mutation-induced changes in the occupancy of the conformational space relative to wild type UraA. WIO, OCC, and OO refer to the wide inward-open, occluded, and outward-open conformations, respectively. Positive or negative deviations indicate increased or reduced occupancy of the respective conformation. B Toy model of the external interdomain-linker demonstrating how the conversion from UraA_WIO to UraA_OCC results in a change in the vertical position of the substrate. View is in the plane of the membrane. Substrate is represented by a red asterisk. C Model of TM4-5 and TM11-12 highlighting the approximate location of the rotation axis for the transport domain and its consequences on the trajectory of the substrate with respect to the barrier helices TM 5 and 12. View is in the membrane on the scaffold domain.