Powering the ABC multidrug exporter LmrA: How nucleotides embrace the ion-motive force

Researchers study how different forms of metabolic energy are coupled to drug extrusion by an ATP-binding cassette transporter.


This PDF file includes:
. Mascot search results for mass spectrometry data for purified LmrA-WT. Table S2. Mascot search results for mass spectrometry data for purified LmrA-ΔK388. Table S3. Speciation of HEPES at pH 6.5 as a function of the HEPES concentration. Data analysis S1. Determination of E rev values and ion stoichiometry. Data analysis S2. Comparisons of ion transport models.

Data analysis S1. Determination of E rev values and ion stoichiometry.
We consider a thermodynamic reaction cycle involving the coupled transport of n Na sodium ions from the outside to the inside of the phospholipid bilayer, and n Cl chloride ions, n H protons, and n D drug molecules from the inside to the outside, with z Na , z Cl , z H , and z D representing the charge of these ions. At equilibrium, the free energies of the coupled ions define a zero-flux equation relating the reversal potential (E rev ) to the transmembrane ion gradients: (C) Imposition of HEPES + gradients HEPES can exist in the form of monovalent cationic, zwitterionic, and monovalent anionic species, and has two protonatable moieties, a piperazin moiety with pKa 1 of 3, and a sulfonate moiety with pKa 2 of 7.5 (20 C) in the following equilibria: in which H 2 A + , HA and Arefer to the double protonated form (HEPES + , protonated on piperazin and sulfonate moiety), zwitterionic from (HEPES 0 , protonated on piperazin moiety only) and double deprotonated form of HEPES (HEPES -). Using the Henderson-Hasselbalch equation, the speciation at pH 6.5 was calculated: 2.9*10 -7 7.3*10 -7 2.9*10 -6 3.6*10 -6 [HEPES 0 ] (M) 9.1*10 -4 2.3*10 -3 9.1*10 -3 1.1*10 -2 [HEPES -] (M) 9.1*10 -3 2.3*10 -2 9.1*10 -2 1.1*10 -1 With In our previous measurements of the chemical proton gradient-dependent transport of radioactive 36 Clin LmrA-containing proteoliposomes, we directly established that LmrA mediates the symport of Cland proton with the stoichiometry of 1:1 (10).

Therefore, n H = n Cl (Eq. 7)
Our observations for symmetrical NaCl-containing buffer solutions that the ATP-induced ion conductance by LmrA-WT is directly proportional to the imposed membrane voltage (Fig. 2H) demonstrates the transbilayer movement of one positive charge per transport cycle, with the term: In the schematic below, in and out refer to the inside and outside of the membrane in lactococcal cells. The calculated E rev value of 69.5 mV is close to the experimental E rev value of 66.7 ± 6.1 mV (n = 3) (Fig. 2H). The calculated E rev value of 35.6 mV is close to the experimental E rev value of 37.6 ± 1.5 mV (n = 3) (Fig. 2H) Hence, the model correctly predicts E rev in response to simultaneous changes in ion gradients.

Data analysis S2. Comparisons of ion transport models.
(A) In previous work, Velamakanni and co-workers (10) focused on electrogenic proton-chloride symport by LmrA-MD and LmrA, and proposed that this reaction is based on apparent (1H + -1X + -1Cl -) + co-transport in which X + can be Na + or Drug + (Model 1). The transport of 100 µM 36 Cl -[added as NaCl] was measured in proteoliposomes containing transport proteins in an inside-out fashion. In the following calculations, the predicted direction of transport and accumulation of 36 Clfor Model 1 and 2Na + /(1H + -1Drug + -1Cl -) + exchange (Model 2) are compared with the experimental data.

A2. ∆ -dependent accumulation of 36 Cl -
To artificially impose a ∆ , the proteoliposomes were prepared in 20 mM (K)Pipes (pH 7.6) containing 100 mM K-acetate, and diluted 100-fold into 20 mM N-methyl-D-glucosamine (NMG) Pipes (pH 7.6) containing 100 mM (NMG) acetate in the presence of valinomycin to induce a potassium diffusion potential (∆ , interior negative) of -118.2 mV. The proteoliposomes accumulated 36 Clabout 17.4-fold above the equilibration level obtained for empty liposomes. Model 2 predicts a 100-fold accumulation of 36 Clin the lumen of the proteoliposomes.

Conclusions
With the imposed electrochemical ion gradients in the proteoliposomes, Model 2 and Model 1 predict Claccumulation in a similar range as the experimental values obtained in the study by Velamakanni and co-workers (10). This conclusion is further supported by the notion that the imposed ∆ (based on a potassium diffusion potential) will have been short-lived due to the export of K + from the lumen of the proteoliposomes via valinomycin (Model 1) and valinomycin and LmrA (Model 2).

(B) Further comparisons of ion transport models
In the electrophysiological experiments in Fig. 2H, the imposition of asymmetric solutions ([NaCl] in /[NaCl] out = 10 mM/150 mM) yielded a measured E rev = 66.7 ± 6.1 mV and calculated E rev = 70.5 mV based on Model 2 (data analyses S1 E1).
In the following section the E rev is calculated using Model 1.

Conclusion
Model 2 has a better predictive value of E rev than Model 1.