Augmenting light coverage for photosynthesis through YFP-enhanced charge separation at the Rhodobacter sphaeroides reaction centre

Abstract

Photosynthesis uses a limited range of the solar spectrum, so enhancing spectral coverage could improve the efficiency of light capture. Here, we show that a hybrid reaction centre (RC)/yellow fluorescent protein (YFP) complex accelerates photosynthetic growth in the bacterium Rhodobacter sphaeroides. The structure of the RC/YFP-light-harvesting 1 (LH1) complex shows the position of YFP attachment to the RC-H subunit, on the cytoplasmic side of the RC complex. Fluorescence lifetime microscopy of whole cells and ultrafast transient absorption spectroscopy of purified RC/YFP complexes show that the YFP-RC intermolecular distance and spectral overlap between the emission of YFP and the visible-region (Q_X) absorption bands of the RC allow energy transfer via a Förster mechanism, with an efficiency of 40±10%. This proof-of-principle study demonstrates the feasibility of increasing spectral coverage for harvesting light using non-native genetically-encoded light-absorbers, thereby augmenting energy transfer and trapping in photosynthesis.

Figure 1. Immunoblotting and spectroscopic analysis of ΔcrtB RC/YFP–LH1 ICMs.

(a) Immunoblotting with antibodies for the RC-H subunit and, separately, to YFP showing the synthesis of an RC-H/YFP polypeptide in Rba. sphaeroides. Lanes 1 and 2 indicate the ΔcrtB and ΔcrtB RC/YFP–LH1 strains, respectively. The asterisk indicates a non-specific signal from the RC-H antibodies. The numbers to the left of the anti-RC-H blot show the positions of protein standards, in kDa. (b) Room temperature and (c) 77 K absorption spectra of membranes purified from ΔcrtB (blue, dashed) and ΔcrtB RC/YFP–LH1 (red) normalized to 870 nm (b) or 880 nm (c). ΔcrtB RC/YFP–LH1 has a peak at 517 nm corresponding to YFP (indicated by arrow); this peak is shifted to 519 nm at 77 K. (d) Fluorescence excitation spectra of membranes with emission monitored at 910 nm show that YFP (peak indicated by arrow) contributes to emission at 910 nm.

Figure 2. Photosynthetic growth curve analysis of ΔcrtB RC/YFP–LH1.

(a) Light was provided using Megaman CFL bulbs at an intensity of 100 μmol photons s⁻¹ m⁻². Photosynthetic growth curves are shown for ΔcrtB RC/YFP–LH1 (red circles) and ΔcrtB (blue circles); a carotenoid-containing LH2-minus strain (green circles) is included as a positive control. (b) Superposition of the absorption of ΔcrtB, ΔcrtB RC/YFP–LH1 and purified YFP onto the emission output of the LEDs used for the photosynthetic growth rate experiment in c. ΔcrtB is shown in blue, ΔcrtB RC/YFP–LH1 in red and purified YFP in yellow; the spectra are normalized to their maximum values. The emission of the green LEDs (520 nm emission maximum, 35 nm FWHM) is indicated by the white region on the grey background. (c) Photosynthetic growth curves of ΔcrtB RC/YFP–LH1 (red circles) and ΔcrtB (blue circles) under green LEDs. A carotenoid-containing LH2-minus strain (green circles) is included as an additional control. Error bars indicate the s.d. from the mean; n=3. Each growth curve is representative of at least three independent experiments. The inset shows a quantification of YFP fluorescence, normalized for equal numbers of cells, showing that there was sufficient oxygen for maturation of the YFP chromophore.

Figure 3. Electron microscopy and 3D reconstruction of the RC/YFP–LH1 complex.

(a) Electron micrograph of purified, negatively stained RC/YFP–LH1 complexes; the two complexes in the central white box are enlarged in the inset, with white arrows indicating the density arising from YFP. Scale bar, 100 nm. (b) Gallery of 40 selected 2D averaged classes; the box size is 28 × 28 nm. Overall, 11,777 particles were used for reconstruction of the 3D model. (c–f) The reconstructed model of the RC/YFP–LH1 complex at 29 Å resolution; the diagram shows the electron density (grey mesh), and the following subunits: LH1β (blue), LH1α (yellow), PufX (red), RC-M (magenta), RC-H (cyan) and RC-L (orange). YFP is in gold. The RC special pair of BChls is shown in red. The structural model is viewed from (c) the cytoplasmic side of the membrane, (d) the plane of the membrane, (e) the periplasmic side, (f) the periplasmic side of the complex, showing only YFP and the LH1 and RC pigments. The dashed red lines show the distances between the YFP chromophore, the RC special pair BChls (red) and LH1 BChls (blue).

Figure 4. Spectral and lifetime imaging of YFP in ΔcrtB RC/YFP–LH1 and ΔcrtB pBBRBB–YFP whole cells.

(a–c) Fluorescence images of whole cells of ΔcrtB RC/YFP–LH1, ΔcrtB and ΔcrtB pBBRBB–YFP cells when excited at 495 nm. (d) Fluorescence emission spectra, each recorded on a single cell. (e) Fluorescence lifetime decay curves recorded at a central wavelength of 550 nm. The best fits, displayed as blue and red lines, were achieved using a double-exponential decay function. The measured instrument response function (IRF) of the system was ∼0.18 ns and taken into account during fitting. Scale bars, 5 μM.

Figure 5. Spectroscopic analysis of purified RC/YFP complexes.

(a) Room temperature absorption spectra of control RC complexes purified from the ΔcrtB strain and RC/YFP complexes purified from the ΔcrtB RC/YFP–LH1 strain. The arrow indicates the absorption maximum of YFP at 514 nm. (b) Fluorescence spectra of purified YFP (YFP-only; red line) and the RC/YFP complex (black line) normalized to the absorbance at λ_exc=486 nm. (c) Normalized fluorescence decay (at 550 nm using λ_exc=515 nm) and fits (using two exponentials plus a constant) of YFP-only (black) and RC/YFP (red) obtained via TCSPC.

Figure 6. Transient absorption spectra of P bleaching at ∼865 nm and P*-stimulated emission at ∼910 nm for RC-only and RC/YFP samples with 515 nm excitation into the YFP absorption band.

The dashed lines indicate absorbance changes due to P* (and decay to P⁺H_A⁻ and P⁺Q_A⁻) on direct excitation (at 515 nm) of the RC-only sample (a) and in RC/YFP, for which YFP is primarily excited by the flash (b). Note the difference in vertical scales for (a,b), such that the amplitude of the absorbance changes at early times (dashed) due to direct excitation of the RC are the same in (a,b) because the samples have approximately the same concentration. (c) Dynamics of energy flow from YFP* to the RC in RC/YFP followed by charge separation in the RC (in both RC/YFP and RC-only). Dashed and solid arrows correspond to the timescales listed in Figures 6a-6b.

Figure 7. Kinetic profiles and fits for evolution of RC P bleaching at 850 nm for RC/YFP and RC-only with 515 nm excitation of YFP.

The signals were normalized at the P absorbance maximum at 867 nm. RC/YFP data (red) was fit to three exponentials plus a constant (solid grey line) and RC-only data (blue, open triangles) to two exponentials plus a constant (cyan).

Figure 8. Modelling of the effect of LH1 removal on the position of YFP and the energy-transfer distances.

(a) The RC/YFP–LH1 complex shown is adapted from the data in Fig. 3. The subunits are: LH1β (blue), LH1α (yellow), PufX (red), RC-M (magenta), RC-H (cyan) and RC-L (orange). YFP is in gold, with its chromophore in orange. The RC special pair of BChls is shown in red. The 85 Å distance between the YFP chromophore and the RC BChl dimer in a is shown to shorten when the surrounding LH1 complex is removed during purification of the RC, down to a minimum of 45 Å in b.