Research
a) Thylakoid membranes from oxygenic phototrophs such as pea and chromatophore membranes from anoxygenic phototrophs such as Rba. sphaeroides have complementary absorbance spectra due to differences in the electronic structures of the macrocycle π electron systems of chlorophyll and bacteriochlorophyll.
b) The major plant light-harvesting complex LHCII harvests solar energy in regions where absorbance by Rba. sphaeroides RCs is weak, notably around 650nm, and its emission spectrum overlaps the absorbance spectrum of the RC between 640 and 800nm.
c) The red-enhanced emission spectrum of heterodimeric plant LHCI has a stronger overlap with the absorbance spectrum of the Rba. sphaeroides RC, particularly the coincident absorbance bands of the bacteriopheophytins (HA/HB).
d) Architecture of the RC cofactors and the route of four-step charge separation which oxidises P870 and reduces QB. The bacteriochlorophylls (orange carbons) and bacteriopheophytins (yellow carbons) give rise to the absorbance bands labelled in c. Further descriptions of pigment-protein structures and their sources are given in Supplementary Fig. 1.
LHCII emission (excitation at 475nm) and LHCI emission (excitation at 500nm) in the absence and presence of WT RCs. The latter spectra are offset for clarity.
Data and fits for photobleaching and dark recovery of P870 absorbance for the WTRCin the absence and presence of LHCII (using variant LHCII-T).
Photobleaching and dark recovery of P870 absorbance in WT RCs in the absence and presence of LHCI (using variant Td-LHCI-Td).
Sketmatic
Schematic of photocurrent generation on a nanostructured silver electrode; black arrows show the route of electron transfer and red arrows show energy flow.
Solution absorbance and EQE spectra for WT RCs compared with those for mixtures of WT RCs with either LHCII-T or Td-LHCI-Td. The absorbance spectra were normalised at 804nm, while each EQE spectrum was normalised to the corresponding absorbance spectrum at the maximum of the long-wavelength P870 band.
Construct designs for adaptation of the RC. For purification the WT RC was modified with a Histag on PufM.
Construct designs for adaptation of LHCII. The control LHCII was truncated at its N-terminus (dLHCII—see text) and was His-tagged at its C-terminus.
Construct designs for adaptation of LHCI which is an Lhca1/Lhca4 heterodimer. For b and c protein sequences are given in Supplementary Fig. 5a.
Sucrose density gradient fractionation of RCs (red bands) and LHCIIs (green bands). RC-LHCII chimeras migrate to a lower position in gradients than either RC or LHCII monomers, with no dissociation into components.
Blue NativePAGE showing the formation of high molecular weight products by mixing LHCI-Td or Td-LHCI-Td with RCC (see Supplementary Fig. 7a for the full gel with more combinations). The multiple bands seen for the high molecular weight products are likely to be due to conformational heterogeneity.
Sucrose density gradient fractionation of RCs (red bands) and LHCIs (green bands). LHCI#RC chimeras and larger RC#LHCI#RC chimeras migrate to lower positions than either RCs or LHCI.
TEM images of an equimolar mixture of the WT RC and dLHCII (top/left), the LHCII#RC chimera (top/right), the LHCI#RC chimera (bottom/left) and the RC#LHCI#RC chimera (bottom/right). Additional images shown in Supplementary Fig. 17. Scale bar represents 50nm.
Molecular model of the LHCII#RC chimera. The RC (maroon) N-terminally adapted with SpyCatcherΔ (blue) is covalently linked to LHCII (green) C-terminally adapted with SpyTag (yellow). Cofactor colours are as described in Supplementary Fig. 1.
Molecular models of the LHCI#RC and RC#LHCI#RC chimeras. Colours as for panel