File S1 - Structural Basis for TatA Oligomerization: An NMR Study of <i>Escherichia coli</i> TatA Dimeric Structure

Supporting results and discussion, Supporting Table S1, and Supporting Figures S1-S12. Figure S1, 2D 1H-15N HSQC spectrum of TatA. 2D 1H-15N HSQC spectrum of TatA in DPC micelles annotated with the backbone assignments. The spectrum was collected on a Bruker Avance 800 MHz spectrometer (with a cryo-probe) at 35°C. The assignments are labeled with the one-letter amino acid code and residue number. The side chain NH2 peaks of Asn and Gln are connected by horizontal lines. The asterisk indicates residues from the His-tag. Figure S2, Secondary structures of E. coli TatA. Secondary chemical shifts for 13Cα, 13Cβ, 13C′ and 1Hα resonances versus residue numbers and the consensus chemical shift index (CSI) plot of TatA in DPC micelles. The values of CSI for β-strand, α-helix and random coil are 1, -1 and 0, respectively. Figure S3, DPR-dependent spectral changes of full-length TatA. (A) The protein sequence of TatA is shown at the top to show the region displaying DPR-dependent chemical shift perturbations (underlined) and residues that have multiple sets of peaks (colored in red). (B) 1H-15N HSQC spectrum of full-length TatA at DPR 300 (black) and 20 (red) with dashed lines indicating the signals from the same residues. The population ratios of different conformers were calculated based on peak volumes (V1∶V2∶V3). (C) The composite chemical shift differences between the two peak sets calculated using the formula . The red bars represent the residues with multiple sets of signals and the CSD was calculated between the 1st and 2nd peak sets. The black bars represent residues that show only chemical shift perturbations. Grey background represents residues that could not be analyzed. Figure S4, Characterization of TatA at very low DPR. 1H-15N HSQC spectrum of full-length TatA at DPR of 300 (black) and an extremely low DPR (<20, red). Representative residues are labeled with black and red lines. Figure S5, Characterization of TatA1-55 at DPR of 30. (A) 1H-15N HSQC spectrum of TatA1-55 at DPR of 30 with dashed lines indicating the signals from the same labeled residues. The population ratios of different conformers were calculated based on peak volumes. (B) Chemical shift perturbations between the 1st and 3rd peak sets of residues in the hinge region. Open and filled bars correspond to the chemical shift changes of backbone 15N and 1HN atoms, respectively. Figure S6, Estimation of the TatA-DPC complexes sizes based on backbone dynamics. The upper panel shows the 2D illustrations of back-calculated sizes of TatA-DPC complexes based on the 1st (upper left) and 2nd sets of peaks (upper right). The data of the 1st set corresponds well to a monomeric model (upper left), whereas the 2nd set exhibits a 3 Å increase in radius compared to the 1st set. The lower panel shows schematic models of a dimeric TatA-DPC complex and a tetrameric TatA-DPC complex. The averaged radius contributed by THM is estimated to be 8.3 Å or 12 Å based on the dimeric or tetrameric models, and would result in an apparent radius increase of 2.8 Å and 6.6 Å respectively. The diameter of a single helix is assumed to be 11 Å in all calculations. The experimental data of the 2nd set fits better to the dimeric model. Figure S7, Backbone dynamics of d-MCG-TatA1-55. The backbone 15N relaxation parameter R2/R1 values of d-MCG-TatA1–55 at DPR of 100 (filled circle) in comparison with full-length TatA (open circle). The N-terminal MCG-tripeptide extension is numbered as residue -3, -2 and -1 respectively. Residues in the C-terminus of full-length TatA are not shown. Figure S8, Illustration of inter-subunit PRE effects of different TatA oligomerization models. Calculation of the population distribution of different oligomeric species and inter-subunit PRE effects using dimeric, trimeric and tetrameric models for paramagnetic labeling on the inter-subunit interface (panel A, TatA1-55-I12C-MTSL) or the opposite side of APH (B, TatA1-55-I11C-MTSL). The 15N-labeled and spin-labeled samples are mixed at 1∶1 molar ratio, and the 15N-labeled sample is the only source generating observable NMR signals. Small yellow star designates the position of the MTSL label, and red and blue crosses indicate expected complete or partial signal broadening by the PRE effect. The expected signal reduction ratio is calculated for each model. The experimental results obtained at DPR ∼40 for MTSL labeling at either position 12 or 11 best correlate with the dimeric model. Figure S9, Formation of disulfide-linked d-MCG-TatA dimer. (A) SDS-PAGE spectra of freshly eluted d-MCG-TatA from Ni-NTA and DTT reduced MCG-TatA. Positions of dimer and monomer are labeled. (B) HSQC spectra of d-MCG-TatA and MCG-TatA showing the different positions of the residue cysteine (C-2) in the oxidized and reduced states. Figure S10, Spectral comparisons of MCG-TatA, d-MCG-TatA and wt-TatA. 1H-15N HSQC spectra of full-length TatA at DPR of 300 (black) and 20 (red) in comparison with MCG-TatA (DPR 110, reduced state) and d-MCG-TatA (DPR 85, oxidized state). The red and blue lines indicate peaks of d-MCG-TatA and the 2nd peak set in TatA (DPR 20), respectively. Representative peaks from the same residue are grouped together and labeled. Apart from residues 1–8 that are most significantly affected by the N-terminal extension, the other residues show chemical shift differences less than 0.04 ppm for protons and 0.5 ppm for nitrogens between d-MCG-TatA and the 2nd peak set in TatA. The primary sequences of wt-TatA and MCG-TatA are shown at the bottom. Figure S11, Verification of inter-subunit NOEs. Representative strips from the 3D 13C/15N-filtered (ω1), 13C-edited (ω3) NOESY-HSQC spectra (mixing time 300 ms) of uniformly labeled dimer (left panels) and mixed dimer (right panels). Assignments for inter-subunit NOE cross peaks are labeled in the right panels, whereas no signals are observable in the corresponding positions in the control sample (left panels). The strong peaks present in the control experiment (left panels) originate from detergent and water signals. Figure S12, A schematic model of the TatA dimerization. Table S1, Protein samples and conditions used in this study. (PDF)