Determination of proton flux and conductance at pH 6.8 through single FO sectors from Escherichia coli

Abstract

We have developed a mathematical model in concert with an assay that allows us to calculate proton (H+) flux and conductance through a single FO of the F1FO ATP synthase. Lipid vesicles reconstituted with just a few functional FO from Escherichia coli were loaded with 250 mM K+ and suspended in a low K+ solution. The pH of the weakly buffered external solution was recorded during sequential treatment with the potassium ionophore valinomycin, the protonophore carbonyl cyanide 3-chlorophenylhydrazone, and HCl. From these pH traces and separate determinations of vesicle size and lipid concentration we calculate the proton conductance through a single FO sector. This methodology is sensitive enough to detect small (15%) conductance changes. We find that wild-type FO has a proton flux of 3100 +/- 500 H+/s/FO at a transmembrane potential of 106 mV (25 degrees C and pH 6.8). This corresponds to a proton conductance of 4.4 fS.

FIGURE 1

The pH response of control and F_o-incorporated lipid vesicles. Before start of the trace, 50 μl of a solution high in K⁺ (250 mM) and containing vesicles with reconstituted F_o (50 μg) or protein-free (control) were added to 4 ml of a K⁺-free translocation buffer (see Materials and Methods). The left arrow shows where valinomycin, a K⁺ ionophore was added. This resulted in the efflux of K⁺ and the initial development of a membrane potential. This potential drives the influx of H⁺ and is manifested as the first rise in pH. The small H⁺ influx in control vesicles is due to a slow leak across the vesicle membrane, but in F_o-containing vesicles the influx is larger due to F_o. After the addition of a proton ionophore (CCCP, middle arrow), vesicles without F_o quickly exchange K⁺ for H⁺ causing the second rise in pH. Finally the solution was back-titrated with HCl (right arrow). Note that the total signal (sum of the first and second rises) is much larger for control samples. This has been described by Cao et al. (2001) and is due to small amounts of ion channel contaminants in the F_o preparation that make the vesicles leaky to both K⁺ and H⁺. However, the slope of the first pH jump increased monotonically with protein concentration as expected. Note that the pH scale bar is 0.025 pH units.

FIGURE 2

Key points from a typical proton flux assay used to define values for our model to calculate proton flux. Depicted is a typical time-lapse pH trace from our proton flux assay and its control. The rise between Points 1 and 3 is defined as the valinomycin signal. The difference between Points 2 and 3 is used as a correction factor (see Discussion). The rise between Points 3 and 4 is the CCCP signal. The drop between Points 4 and 5 is due to an HCl back-titration and is used to define ΔpH/ΔH⁺ in each assay. The total positive change between Points 1 and 4 is defined as the total signal. The line marked “10 Sec Slope” is used to calculate the initial valinomycin slope. All of these factors are used as input to our mathematical model calculating single F_o proton flux. Note that the slopes between the 10-Sec Slope line and Point 3 for both the control and 25-μg protein signals are nearly identical, indicating that the H⁺ leakage rate across the vesicle population is similar for both samples.

FIGURE 3

Time-lapse pH traces for the same sample 26 days apart. The repeatability of the time-lapse pH traces that provide the data for our proton flux calculation is evident by the similarity between these two traces. The pH trace stays relatively consistent for up to seven weeks between assays using the same sample. The sample is stored at room temperature in the dark in sealed culture tubes. The Day 1 graph was shifted by 14 mpH to overlay the two traces.

FIGURE 4

Temperature dependence of proton flux assay. The pH trace of a 100-μg F_o sample run at 10°C increments is representative of many temperature-dependent experiments that were executed. The valinomycin slope is significantly affected by the temperature at which the assay is run. Q10 varies from 1.6 to 2.8. Since the valinomycin slope is one of the most significant factors in our calculation, repeatability cannot be expected if the assays are not all run at the same temperature. The total signal is nearly identical for all temperatures, indicating the total vesicle population is consistent in all samples. The drop in the total signal for 35°C is possibly due to the breakdown of the vesicles at higher temperatures. This same effect was seen in all temperature-dependent experiments.