Probing the affinity of SecA for signal peptide in different environments

Abstract

SecA, the peripheral subunit of the Escherichia coli preprotein translocase, interacts with a number of ligands during export, including signal peptides, membrane phospholipids, and nucleotides. Using fluorescence resonance energy transfer (FRET), we studied the interactions of wild-type (WT) and mutant SecAs with IAEDANS-labeled signal peptide, and how these interactions are modified in the presence of other transport ligands. We find that residues on the third alpha-helix in the preprotein cross-linking domain (PPXD) are important for the interaction of SecA and signal peptide. For SecA in aqueous solution, saturation binding data using FRET analysis fit a single-site binding model and yielded a Kd of 2.4 microM. FRET is inhibited for SecA in lipid vesicles relative to that in aqueous solution at a low signal peptide concentration. The sigmoidal nature of the binding curve suggests that SecA in lipids has two conformational states; our results do not support different oligomeric states of SecA. Using native gel electrophoresis, we establish signal peptide-induced SecA monomerization in both aqueous solution and lipid vesicles. Whereas the affinity of SecA for signal peptide in an aqueous environment is unaffected by temperature or the presence of nucleotides, in lipids the affinity decreases in the presence of ADP or AMP-PCP but increases at higher temperature. The latter finding is consistent with SecA existing in an elongated form while inserting the signal peptide into membranes.

Figure 1

Design of the fluorescence resonance energy transfer assay. (A) HPLC spectra of the WT signal peptide labeled with IAEDANS (SP–P). Signal peptide absorbs at 220 nm (⎯), and the probe absorbs at 336 nm (- - -). In the inset, nonlabeled peptide absorbs at 220 nm but not at 336 nm. The peptide was eluted from a C4 analytical column with 61% acetonitrile in 0.1% trifluoroacetic acid. (B) Spectral overlap of SecA tryptophan emission fluorescence and SP–P: absorption spectrum of SecA (- - -), emission spectrum of SecA (⎯), absorption spectrum of SP–P (– – –), and emission spectrum of SP–P (– - - –). (C) SecA tryptophan fluorescence quenching and IAEDANS fluorescence enhancement: SecA alone (– – –), SP–P alone (⎯), and SecA with 1 (- - -) and 4 μM (– - - –) SP–P. Fluorescence emission maxima are observed at 340 and 480 nm for SecA and IAEDANS, respectively.

Figure 2

Equilibrium binding of IAEDANS-labeled signal peptide (SP–P) and SecA monitored via fluorescence resonance energy transfer. (A) Saturation binding of SP–P to SecA. The fluorescence of SecA with each addition of SP–P was normalized to the SecA fluorescence without SP–P. The curve was fit to a one-site binding model. F_o is the fluorescence of SecA alone, and F is the fluorescence of SecA after addition of SP–P at 345 nm. (B) Competition binding of functional (WT) and nonfunctional (3K2L) signal peptides (SP) to SecA. SecA was incubated with 3 μM SP–P prior to addition of nonlabeled wild-type (■) or 3K2L (▽) signal peptide (shown in the inset). (C) Contribution to FRET of specific tryptophans (indicated below the bar) in the region of the proposed peptide-binding site. Numbers in bold, above the bars, indicate the fraction of FRET remaining for each mutant relative to WT SecA. Each data point represents an average of at least three separate experiments performed in duplicate.

Figure 3

Screening for the signal peptide binding site on SecA using FRET and mutants of SecA. (A) Effect of SecA residue substitutions on the efficiency of FRET. F_o is the fluorescence of SecA alone, and F is the fluorescence of SecA after addition of 6 μM SP–P at 345 nm. Each data point represents an average of at least two separate experiments performed in duplicate. (B) CD spectrum of WT SecA (⎯) and a representative mutant SecA (- - -) spectrum. (C and D) Structures of B. subtilis SecA translocation domain in closed [PDB entry 1M6N (14)] and open [PDB entry 1TF5 (20)] conformations, respectively, shown at different angles. Positions of mutants examined are marked in bold, and the third α-helix (residues 292–319) of the PPXD domain is highlighted in the box. Space-filled amino acids are tryptophans analyzed in Figure 2C and are marked by the residue number corresponding to the E. coli sequence.

Figure 4

Comparison of binding of the signal peptide to SecA in aqueous solution and lipid vesicles. (A) Saturation binding of the IAEDANS-labeled signal peptide (SP–P) to SecA in an aqueous environment (■) and in E. coli phospholipids (□). The inset shows a magnification of the binding curve at low signal peptide concentrations for SecA in lipid. F_o is the fluorescence of SecA alone, and F is the fluorescence of SecA after addition of SP–P at 345 nm. (B) Stern–Volmer plots of acrylamide quenching of SP–P. The fluorescence of the labeled peptide without the quencher (F_o) and fluorescence of the sample after the addition of acrylamide (F) at 500 nm were monitored in the absence of SecA in aqueous solution (■) and in the presence of lipid vesicles (□). Parallel measurements were taken for SP–P with 50 nM SecA in aqueous solution (▲) and in lipid vesicles (△). Each data point represents an average of at least three separate experiments performed in duplicate.

Figure 5

Effect of the signal peptide on the oligomeric state of SecA. SecA (500 nM) was incubated in the presence or absence of vesicles composed of E. coli phospholipids and WT alkaline phosphatase signal peptide as indicated (see Experimental Procedures). Molecular weight standards are as follows: catalase (232 kDa), BSA dimer (132 kDa), and BSA monomer (66 kDa). The location of SecA in its monomeric, dimeric, and tetrameric forms is indicated.

Figure 6

Factors affecting the binding of the signal peptide to SecA. (A) Effect of nucleotide on signal peptide–SecA binding in an aqueous environment: without nucleotide (■), with 1 mM ADP (●), and with 1 mM AMP-PCP (▲). F_o is the fluorescence of SecA alone, and F is the fluorescence of SecA after addition of the IAEDANS-labeled signal peptide (SP–P) at 345 nm. (B) Effect of nucleotide on SecA–signal peptide binding in lipid vesicles: no nucleotide (□) and with 1 mM ADP (○). The curve generated in the presence of 1 mM AMP-PCP (△) is shown in the inset in comparison with that of ADP. (C) Effect of temperature on the concentration dependence of binding of the signal peptide to SecA in aqueous solution at 22 °C (■), lipid vesicles at 22 °C (□), aqueous solution at 37 °C (▲), and lipid vesicles at 37 °C (△). Each data point represents an average of at least two separate experiments performed in duplicate.