Structural characterization of the N terminus of IpaC from Shigella flexneri

Abstract

The primary effector for Shigella invasion of epithelial cells is IpaC, which is secreted via a type III secretion system. We recently reported that the IpaC N terminus is required for type III secretion and possibly other functions. In this study, mutagenesis was used to identify an N-terminal secretion signal and to determine the functional importance of the rest of the IpaC N terminus. The 15 N-terminal amino acids target IpaC for secretion by Shigella flexneri, and placing additional amino acids at the N terminus does not interfere with IpaC secretion. Furthermore, amino acid sequences with no relationship to the native IpaC secretion signal can also direct its secretion. Deletions introduced beyond amino acid 20 have no effect on secretion and do not adversely affect IpaC function in vivo until they extend beyond residue 50, at which point invasion function is completely eliminated. Deletions introduced at amino acid 100 and extending toward the N terminus reduce IpaC's invasion function but do not eliminate it until they extend to the N-terminal side of residue 80, indicating that a region from amino acid 50 to 80 is critical for IpaC invasion function. To explore this further, the ability of an IpaC N-terminal peptide to associate in vitro with its translocon partner IpaB and its chaperone IpgC was studied. The N-terminal peptide binds tightly to IpaB, but the IpaC central hydrophobic region also appears to participate in this binding. The N-terminal peptide also associates with the chaperone IpgC and IpaB is competitive for this interaction. Based on additional biophysical data, we propose that a region between amino acids 50 and 80 is required for chaperone binding, and that the IpaB binding domain is located downstream from, and possibly overlapping, this region. From these data, we propose that the secretion signal, chaperone binding region, and IpaB binding domain are located at the IpaC N terminus and are essential for presentation of IpaC to host cells during bacterial entry; however, IpaC effector activity may be located elsewhere.

FIG. 1.

IpgC associates with a region near the N terminus of IpaC in vitro. Fluorescence polarization was used to monitor the interaction between fluorescein-labeled IpaC derviatives and nonfluorescent IpgC. The increase in polarization caused by the addition of increasing concentrations of IpgC is shown for FM-labeled IpaC^−19-363 (•), FITC-labeled region I' (IpaC^−19-100) (○), FM-labeled IpaC^Cys (▴), and FITC-labeled IpaC^Δ1-62 (▵). IpaC^Cys is composed of IpaC^50-363 with a MetCys pair at the immediate N terminus. The fluorescent proteins were used at approximately 20 nM.

FIG. 2.

IpgC binds to both IpaC and IpaB in vitro. (A) IpaC and proteins derived from IpaC were purified and labeled with FITC for fluorescence polarization measurements. The increase in polarization caused by the addition of increasing concentrations of IpaB is shown for FITC-labeled IpaC^−19-363 (○), region I' (IpaC^−19-100) (•), IpaC^Δ21-100 (▴), and IpaC^171-363 (▵). The values shown are an average of three measurements of a single sample and are representative of multiple (three or more) independent experiments. (B) The increase in the polarization of FITC-labeled IpgC (20 nM) was monitored following the addition of increasing concentrations of nonfluorescent IpaB.

FIG. 3.

Association of IpaB and IpgC to FM-IpaC^Cys results in protection of the probe from solute quenching. (A) The addition of increasing concentrations of acrylamide results in quenching of FM-IpaC^−19-100 fluorescence (•). The degree of acrylamide quenching is not greatly affected by the addition of IpaB (○) or IpgC (▴). The fluorescent protein was used at 30 nM in each case, and the results are from a single representative experiment. Acrylamide quenching of FM-IpaC^Cys alone is shown by the closed circles, and quenching with IpgC is shown by the closed triangles. (B) The same experiment was carried out with FM-IpaC^Cys at approximately 30 nM. The data shown are from a single representative experiment. The average K_sv value in each case (n ≥ 3) is given in the text.

FIG. 4.

Region I' (IpaC^−19-100) associates with IpgC and IpaB in a competitive manner. (A) The emission spectrum of 230 nM CPI-labeled region I' is shown in the presence of an equimolar amount of IpgC (dashed line) or FITC-labeled IpgC (solid line). The decrease in emission with FITC-IpgC is due to FRET of CPI excitation energy from the coumarin to the fluorescein moiety. (B) The same experiment was carried out with CPI-labeled region I' and IpgC (dashed line) and FITC-labeled IpgC (solid line), except that 250 nM IpaB is also present in the sample. CPI fluorescence was measured by using an excitation wavelength of 385 nM. The data are from a single representative experiment, which was performed three times.