Interconvertibility of lipid- and translocon-bound forms of the bacterial Tat precursor pre-SufI

Abstract

Signal peptides target protein cargos for secretion from the bacterial cytoplasm. These signal peptides contain a tri-partite structure consisting of a central hydrophobic domain (h-domain), and two flanking polar domains. Using a recently developed in vitro transport assay, we report here that a central h-domain position (C17) of the twin arginine translocation (Tat) substrate pre-SufI is especially sensitive to amino acid hydrophobicity. The C17I mutant is transported more efficiently than wild type, whereas charged substitutions completely block transport. Transport efficiency is well-correlated with Tat translocon binding efficiency. The precursor protein also binds to non-Tat components of the membrane, presumably to the lipids. This lipid-bound precursor can be chased through the Tat translocons under conditions of high proton motive force. Thus, the non-Tat bound form of the precursor is a functional intermediate in the transport cycle. This intermediate appears to directly equilibrate with the translocon-bound form of the precursor.

Figure 1. Quantification of precursor-lipid and precursor-translocon interactions

All gels in this figure are anti-SufI immunoblots. (A) Membrane binding assay. The indicated amounts of pre-SufI were incubated with pTat or pTat* IMVs at pH 8.0 and then sedimented. The gel shows the amount of precursor protein recovered in the washed pellet fraction. Lanes 1–3 are quantification standards. (B) Competition between Sec and Tat precursors. Shown is the amount of membrane-bound (top and middle) and transported (bottom) pre-SufI in the presence of various molar equivalents of proOmpA-HisC. The proOmpA-HisC protein was preincubated with IMVs for 10 min prior to pre-SufI addition (lanes 4–9), or simultaneously with pre-SufI (lanes 10–15). Transport reactions were initiated with 4 mM NADH. Each lane contains the Sec chaperone SecB (25 μM). (C) Concentration dependence of membrane binding and transport. The binding of pre-SufI to ΔTat (top) and pTat (middle) IMVs was assayed as in A (pH 8.0). Transport into pTat* IMVs (bottom) was initiated by the addition of 4 mM NADH. Control lanes are devoid of pre-SufI for the binding assays, or with 8.3 pmol pre-SufI but no NADH for the transport assay. The plot shows the quantification of the pre-SufI bound to non-translocon components of the membrane (ΔTat IMVs, red), the pre-SufI bound to both Tat translocons and non-Tat membrane component (pTat IMVs, blue), and the pre-SufI transported into pTat* IMVs (black) (n = 3). The amount of pre-SufI bound to TatABC (green) was estimated by subtracting the red curve from the blue curve. (D) Binding to trypsin-treated IMVs. Membrane binding interactions were assayed as in A.

Figure 2. Effects of chemical agents and the RR → KK mutation on membrane binding and transport of pre-SufI

All gels in this figure are anti-SufI immunoblots. (A) Effect of pH on binding and transport efficiency. Shown is the effect of pH on pre-SufI binding (top), and on transport in the presence of 4 mM ATP with an ATP regenerating system (middle) or 4 mM NADH (bottom). The data were quantified as in Fig. 1C: (blue) precursor bound to pTat IMVs, (red) lipid-bound precursor, (green) translocon-bound precursor, (black, solid) precursor transported with NADH, and (black, open circles) precursor transported with ATP (n = 3). (B) Effect of KCl concentration on binding and transport efficiency. Key as in A (n = 3). (C) Effect of KK substitution for the RR motif (pre-SufI-KK-CCC) on pre-SufI binding and transport efficiency (Tr) (n = 3). (D) Effect of urea concentration on binding and transport efficiency. Key as in A (n = 3).

Figure 3. Effect of chemical agents on the membrane binding and transport efficiency of GFP precursors

The gels in this figure are anti-SufI (B, top) or anti-GFP (all others) immunoblots. (A) Concentration dependence of the binding efficiency of spTorA-GFP and spSufI-GFP. The data were quantified as in Fig. 1C: (dashed) precursor bound to pTat IMVs, (dotted) lipid-bound precursor, (solid) translocon-bound precursor, (blue) spTorA-GFP, and (red) spSufI-GFP (n = 3). The upper band on the anti-GFP immunoblot results from immuno-crossreactivity to a non-GFP protein and was only occasionally observed. (B) Effect of 2 M urea (U), 1 M KCl (K) and 2 M urea + 1 M KCl (U + K) on the membrane binding efficiency of pre-SufI (red), spTorA-GFP (green) and spTorA12-GFP (blue). The plot shows averaged data (n = 3) for ΔTat (hatched) and pTat (solid) IMVs. The precursor bound to pTat IMVs is considered the control (100%). (C) Effect of the magnitude of the PMF on the transport efficiency of spTorA-GFP and spTorA12-GFP. In B and C, 8.3 pmol of spTorA-GFP and spTorA12-GFP were added to the reactions; in B, 3.1 pmol of pre-SufI was added.

Figure 4. Effect of the hydrophobicity of the amino acid at position 17 in the pre-SufI signal peptide on membrane binding and transport efficiency

(A) Domain structure of pre-SufI. C17 and C295 are endogenous cysteines and C497 was added for fluorescent dye labeling, yielding pre-SufI-CCC. (B) Membrane binding and transport efficiency of pre-SufI-CCC. These anti-SufI immunoblots show that addition of C497 has no effect on the membrane binding and transport efficiency of pre-SufI. (C) Membrane binding and transport efficiency of C17 mutants (n = 3). (D) Pairwise correlation between the ΔΔG (according to cyclohexane-to-water partition coefficients (Radzicka et al., 1988), the translocon binding efficiency and the transport efficiency for C17 mutants. For the left and middle plots, the data points for asparatic acid (D) and glycine (G) were not included in the best-fit line determination because they appear to be outliers. Once the transport efficiency reaches zero, a more negative ΔΔG is expected to have no effect, which explains the D point. (E) Kyte and Doolittle (Kyte & Doolittle, 1982) hydropathy plot for the signal peptides of pre-SufI (black), pre-SufI-ICC (red), pre-SufI-TCC (violet) and pre-TorA (green) using a scanning window of 9 residues.

Figure 5. Membrane binding and transport efficiency of pre-SufI-ICC

All gels in this figure are anti-SufI immunoblots. (A) Concentration dependence of membrane binding efficiency. The amount of pre-SufI-CCC (black) and pre-SufI-ICC (red) bound to ΔTat (solid) and pTat IMVs (dashed) was quantified (n = 3). (B) Correlation between transport efficiency and the amount of translocon-bound precursor. The gel shows the concentration dependent transport efficiency of pre-SufI-CCC and pre-SufI-ICC using 4 mM NADH. The plot shows the amount of transported pre-SufI-CCC (black) and pre-SufI-ICC (red) from three independent experiments (solid lines). The amount of translocon-bound protein was calculated from the data in A as described in Fig. 1C (dashed lines).

Figure 6. Transport kinetics of Atto565-labeled pre-SufI

(A) Transport of pre-SufI, pre-SufI-CCC, pre-SufI-ICC and pre-SufI-IAC (n = 3). 3.1 pmol (90 nM) precursor protein was added to each reaction. (B) Membrane binding and transport efficiency of pre-SufI-IAC^atto. The membrane binding efficiency of unlabeled (top) and Atto565-labeled (middle) pre-SufI-IAC to ΔTat and pTat IMVs was determined for various precursor concentrations. The transport yield (bottom) of both of these precursors was determined with 4 mM NADH and pTat* IMVs. The plot shows the quantified results: (black) pre-SufI-IAC, (red) pre-SufI-IAC^atto, (dotted) lipid-bound precursor, (dashed) lipid + translocon-bound precursor, and (solid) transported precursor (n = 3). (C) Transport kinetics of pre-SufI-IAC^atto. The top and middle gels are anti-SufI immunoblots of pre-SufI-IAC (black) and pre-SufI-IAC^atto (red), respectively. In the bottom gel, the Atto565-labeled protein was detected by fluorescence emission (blue) (n = 3). All gels in A and B are anti-SufI immunoblots.

Figure 7. Properties of the pre-SufI-IAC^biotin/avidin complex

(A) Effect of avidin on the transport of pre-SufI-IAC^biotin. Proteins were detected by Western blotting using avidin-HRP. When present, biotin (16 μM) was in large excess over neutravidin (1.6 μM), and neutravidin was in large excess over the precursor protein (40 nM, 1.4 pmol). Thus, the binding interactions were saturated. [IMV] = 5 (A₂₈₀) (B) Competition between pre-SufI-IAC^biotin and pre-SufI-IAC^atto in the presence of avidin. The transport efficiency of pre-SufI-IAC^atto (40 nM, 1.4 pmol) was determined in the absence (black) and presence (red) of 25 μM neutravidin and the indicated molar equivalents of pre-SufI-IAC^biotin (n = 3). Gels were visualized by the Atto565 fluorescence emission. [NADH] = 4 mM; [IMV] = 5 (A₂₈₀). (C) Membrane binding efficiency of pre-SufI-IAC^biotin in the presence of avidin. This anti-SufI immunoblot shows the binding efficiency of pre-SufI-IAC and pre-SufI-IAC^biotin to ΔTat and pTat IMVs in the presence and absence of 25 μM neutravidin, as indicated. In lanes 10–15, the boxes indicate which reagent was added first. In lanes 10–12, neutravidin was preincubated with pre-SufI-IAC^biotin for 10 min prior to addition of IMVs (A₂₈₀ = 2). In lanes 13–15, IMVs were preincubated with pre-SufI-IAC^biotin for 10 min prior to addition of neutravidin. The lower plot corresponds to integrated values from the lanes below in the gel (averaged over 3 identical experiments). The upper plot corresponds to the amount of translocon-bound precursor, calculated for each set of experiments by subtracting the ΔTat IMV values from the pTat IMV values. (D) Effect of precursor addition order on the transport efficiency of pre-SufI-IAC^atto. For lanes 5–9, the reactions contained pre-SufI-IAC^atto, 25 μM neutravidin, and pTat* IMVs (A₂₈₀ = 5). For lanes 7–9, the reactions contained pre-SufI-IAC^biotin. The boxes indicate which precursor (90 nM, 3.1 pmol each) was added first to the reaction solution. In lane 7, both precursors were added simultaneously, and 4 mM NADH was added 5 min later. In lane 8, pre-SufI-IAC^biotin was added 5 min prior to the simultaneous addition of pre-SufI-IAC^atto and NADH. Lane 9 was obtained similarly to lane 8, but the precursors were reversed. This gel was visualized by the Atto565 fluorescence emission.

Figure 8. Transport of lipid-bound precursor

Pre-SufI (90 nM) was incubated with pTat* IMVs for 10 min at 37°C. IMVs were recovered by centrifugation through a 0.7 M sucrose cushion to remove the unbound precursor protein (Experimental Procedures). Lane 4 of this anti-SufI immunoblot shows the total (T) amount of pre-SufI bound to the IMVs. To probe the reversibility of the membrane binding reaction, IMVs with bound pre-SufI were incubated at 37°C for 10 min and re-isolated by centrifugation. Most of the pre-SufI was recovered with IMVs in the pellet (P) fraction (lanes 5–6). In contrast, addition of 2 M urea resulted in the release of a significant fraction of the membrane-bound pre-SufI into the supernatant (S) fraction (lanes 7–8). In the presence of ATP, more pre-SufI was transported than was released from the membrane by urea (compare lanes 11 and 7), indicating that some of the lipid-bound precursor must have transported under these conditions. For this figure, pTat* IMVs were used for both the membrane binding experiments and the transport experiments (n = 3). [IMVs] = 2 (A₂₈₀) for each lane.

Figure 9. Model of E. coli Tat transport

TatB (orange) and TatC (yellow) form a receptor complex, and TatA (blue) forms ring-like oligomers in resting membranes (A). The precursor protein (red) binds to the membrane surface (B) in addition to its binding site on the TatBC complex (C). The cargo/TatBC complex associates with an oligomerized TatA complex (D). Cargo is transported across the membrane in the presence of a transmembrane electric field gradient (Δψ). Signal peptide (magneta) cleavage occurs after transport of the precursor protein (E). We reported earlier that there are two ψ-dependent steps (Bageshwar & Musser, 2007). The first Δψ-dependent step can occur in the absence of precursor protein (not shown). The second Δψ-dependent step occurs after the precursor protein binds to the membrane, suggested here as the D → E step. For clarity, the likely possibility that the TatBC complexes form higher order oligomers is not shown. The fate of the signal peptide after transport is not known.