Structures of Rhodopseudomonas palustris RC-LH1 complexes with open or closed quinone channels

High-resolution structures of reaction-center light-harvesting 1 complexes provide new insights into quinone dynamics.

. Alignments of RC-LH1 core complex subunit amino acid sequences showing the resolved residues in the two RC-LH1 complexes. Residues that were not resolved in the structures are colored red in the sequences generated from the genome sequence (labelled genome). Structural sequences for LH1 α and β were taken from chains with the greatest level of coverage. Green boxes highlight α and β sequences observed by mass spectrometry with dashed lines showing alternate cleavage sites. The 'f' below the sequence indicates that Met1 is Nformylated. 16 at 880 nm. Spectral changes for RC-LH114-W (panels A, C) and RC-LH116 (panels B, D) at 895 (panels A, B; the Qy bleaching for the LH1 BChls) and 591 nm (panels C, D; the Qx bleaching for the LH1 BChls) following excitation with a 100 fs laser flash at 880 nm. Raw data (black circles) were fit to a multi-exponential model (red lines). The fitted values are shown in Table  S2.

Fig. S5. Transient absorption (TA) spectra and decay associated difference spectra (DADS)
for the RC-LH1 14 -W and RC-LH1 16 complexes. Panels A and B show TA spectra for RC-LH114-W and RC-LH116, respectively, in the visible and near-infrared (NIR) spectral regions. Panels C and D show DADS in the visible and NIR spectral regions for the RC-LH114-W and RC-LH116 core complexes, respectively. Global analysis was performed to deconvolute the timeresolved spectra in Carpetview. The visible and NIR spectra were combined (Fig. 2C-D) and modelled assuming parallel decay of temporally distinct non-reversible components. The resultant DADS are labelled with their associated process and time constants (inf being slower than the 7 ns measurement window time), assigned based upon known spectral signatures elucidated in previous studies (59,60). LH1* = singlet excited state of the LH1 BChls, RC = RC primary donor, EET = excitation energy transfer, g.s. = ground state, P + HA -= RC charge separated state, P R = RC primary donor triplet state. Panel E shows DADS of LH1 from Rba. sphaeroides lacking the RC (thus lacking the LH1RC EET and all RC processes), which was used as a reference for assignment of the LH1* annihilation and LH1*GS signals in the two Rps. palustris samples (preparation of the Rba. sphaeroides LH1 is detailed in ref (53))

Supplementary text
The DADS in Figure S5 show the competing processes/states present in the RC-LH1 samples following excitation. The process of particular interest to this work is LH1→RC energy transfer (EET, ~42 ps, red lines) and subsequent charge recombination following electron transfer (ET) within the RC (~5-6 ns, light blue lines). Here, LH1→RC EET (~42 ps, red lines) is the first step in a two-step process that involves EET from the excited LH1 (LH1*) to the RC to make the RC primary electron donor (P*), followed by rapid (3-4 ps) P*  P + HA -ET within the RC. The spectral signatures in the DADS are decay of LH1 bleaching at 896 nm (due to LH1*  P* EET) and the growing in of P + HAfeatures, namely the bleaching of the HAQx band at 542 nm and the formation of the HAanion band at 665 nm (due to P*  P + HA -ET). The subsequent decay of these P + HAfeatures (via P + HAcharge recombination) are clear in the ~6 ns spectra, as well as the infinity spectra (inf, orange lines) where some population of P + HAdecays to form the P triplet state (P R ) as well as ground state (g.s.) with a time constant greater than the timescale of the experiment. Note that it is known that P + HAcharge recombination is multi-exponential with at least three time constants spanning <1 ns to >10 ns; thus, the 6 ns observed here is an average. The other competing processes are as follows: (1) The ~2.5 ps spectra (green lines) indicate excited state LH1 (LH1*) intra-ring annihilation (resulting from two excited states within the same LH1 ring), as seen from the LH1 decay of BChl Qx bleaching (591 nm) and Qy bleaching (~895 nm).
(2) As described in the Materials and Methods Transient Absorption Spectroscopy section, a high concentration of ascorbate (~10,000x the RC concentration) is added to the RC-LH1 samples to ensure that P is reduced (i.e., not in the photooxidized state) prior to each excitation flash (preventing the RCs from becoming "closed" and thus unable to participate in LH1→RC EET). However, a small fraction of RC-LH1 complexes (~5-15%) exhibit excited state LH1*→g.s. decay (500-600 ps, dark blue lines), probably due to some RCs remaining oxidized when exposed to the excitation pulse or the RC being absent from a small fraction of the LH1 complexes (likely due to damage during sample preparation). The LH1* excited state lifetime in the absence of the RC (i.e. deactivation of LH1* to the g.s.) in Rba. sphaeroides is ~500 ps (shown in figure S5E), which is in close agreement with the ~500-600 ps component found in the Rps. palustris RC-LH1 samples.
(3) The P + HAstate should primarily decay by well-known charge recombination pathways to produce the g.s. or P R (59, 60) (~6 ns and inf spectra, dark blue and orange lines respectively); however, a small fraction of the RC-LH116 sample still has active QA and exhibits P + HA -→ P + QA -ET (e.g., 542 nm bleaching decay). This is despite the RC QB being replaced by terbutryn (which cannot participate in electron transfer) and QA being predominantly inactivated by pre-exposure of the sample to the excitation laser (see materials and methods). P + HA → P + QA -ET typically has a time constant of ~200 ps (59,60), but the presence of this process in such a small fraction (~10%) of the sample and close overlap in time with LH1-only decay (~500 ps time constant) prevents the two states/processes from being resolved in the DADS and therefore are combined in the ~500 ps spectrum (dark blue line).     Table S3. Error bars represent standard deviation from the mean of three biological replicates from three separate cultures (overlaid in black circles) and the p-value is derived from a paired, 2-tailed Student's t-test vs. the wild-type replicates with 'ns' indicating that the difference is not significant.                Fig. S4) and rates from global fitting (presented in Fig. S5). Avg 44+3 τ 1 = LH1 excited state (LH1*) singlet-singlet annihilation τ 2 = LH1 to RC energy transfer τ 3 = LH1* deactivation to ground state and/or RC P + H A - P + Q A -ET τ 4 = RC P + H Acharge recombination Mass spectral ion intensities for the RC-H, -M and -L subunits, LH 1-α and -β subunits and protein-W in wild-type, ΔpufW and PufW-His membranes. Proteins extracted from membranes isolated from the three strains (n = 3 biological replicates from three separate cultures) of Rps. palustris were digested with a combination of endoproteinase Lys-C and trypsin. The resultant peptides (500 ng) were analyzed by nano-flow reverse phase chromatography coupled to mass spectrometry. (A): Mass spectral ion intensities for peptides mapping to the RC-LH1-W subunits were extracted by MaxQuant. Only peptides validated by product ion spectra in all nine analyses were utilized. Intensities for peptides containing Met are shown as the sum of both native and sulfoxide forms. (B): Summed peptide ion intensities were used to generate Fig. S13 6.79E+08 6.43E+08 6.00E+08 8.05E+08 8.47E+08 9.08E+08 9.51E+08 1.01E+09 9.43E+08 VSV-AIR 4.72E+08 4.71E+08 5.17E+08 3.03E+08 3.24E+08 3.13E+08 3.10E+08 3.29E+08 3.20E+08 VSV-TSK 3.23E+08 3.32E+08 3.81E+08 4.39E+08 4.78E+08 4.82E+08 3.38E+08 3.43E+08 3.10E+08 YLE-VAK 7.11E+08 8.21E+08 9.12E+08 7.11E+08 6.46E+08 7.11E+08 6.19E+08 6.59E+08 6.31E+08