Sizes of some protein complexes in Rb. sphaeroides photosynthetic chain

 Fig. showing model structures for protein complexes of the photosynthetic chain, taken from crystallographic data. LH1 and LH2 rings from Schulten's lab; bc1-complex dimer, from Ed Berry's coordinates for the chicken heart mitochondrial complex (but including only the three catalytic subunits).

Known protein complexes

Reaction center (RC) 65-72 Å diameter in membrane plane at maximal of cross-section = ~7 nm

LH2 a,b 20-22 Å at perpendicular intersect (the LH1 a,b pair likely has a similar dimension).

LH2 complex ((a,b)9 ring) 65 Å = 6.5 nm

bc1-complex dimer (catalytic subunits only, - cyt b, cyt c1, ISP) 90-110 Å = ~10 nm

RC + 2 LH1 (for RC inside LH1 palisade) = 70 + 40 = 110 A = ~11 nm

Supercomplex sizes expected

RC dimer = 70 + 70 = 140 = ~14 nm (with no LH1)

RC dimer with LH1 ~22 nm

"Standard" supercomplex:

2 RC + bc1-complex = ~14-19 nm

bc1-complex dimer + 4 RC (with no LH1, symmetrical RC) = 70 + 110 + 70 = 250 = ~25 nm

bc1-complex dimer + 4 RC (with LH1) = 110 x 3 = 330= ~33 nm

From freeze-etch images

The size (in nm) of the main particle populations in R. sphaeroides wild-type intracytoplasmic membranes (chromatophores) is 12.2 ±0.3 nm (1), or 8.25 - 9.75 (with the larger mean size for cells grown at low light) (2), or 7.7 ± 1.5 (with peaks at ~ 6.5, ~8 nm and ~10 nm) (3), or 7.8 ± 1.2 (for chromatophores), 10.4 ± 0.8 (for tubular intracytoplasmic membrane (ICM) in LH2- mutant), 10.5± 1.0 (for ordered cytoplasmic membrane) (4).

In histograms from ref. 2, no particles of diameter > 18 nm were reported. In the histograms from ref. 3, no particles > 12 nm were reported. None of the authors have noted any specific aggregation into dimeric or trimeric units containing three masses of similar size in chromatophores.
It has been suggested that the patterned array of particles seen in tubular ICM domains in some strains of Rb. sphaeroides (4, 5, 6) may represent supercomplexes (6). Very similar structures have been studied in a strain of Rb. sphaeroides lacking LH2 (4). In this latter work, it was noted that the 10.4 nm particles seen in the tubular arrays were distinct from the 7.8 nm particles seen in vesicular ICM domains (chromatophores), but were similar to particles seen in ordered domains of the cytoplasmic membrane. Recently, the structure of the particles in these tubular arrays has been resolved at higher resolution by electron microscopy and image averaging of negatively stained arrays (Vermeglio, A., personal communication). The pictures show that the arrays are composed of twins of reaction centers, each surrounded by an incomplete ring of LH1 light harvesting complexes. The arrays do not have room for a dimeric bc1 complex, but might accomodate a monomeric complex. If these structures represent supercomplexes, then they would require a dimer of reaction centers (the structures in the arrays) interacting with a monomeric bc1 complex, not resolve in the averaging process (Joliot, P. and Vermeglio, A., personal communication), and not co-isolated under mild detergent extraction.


If supercomplexes containing a dimeric bc1 complex exist, they are not identified in normal chromatophores as distinct particles by freeze-fracture electron microscopy. The main population of particles is of the size expected from LH2 rings. The discrete particles of ~11 nm diameter are the size expected from a reaction center surrounded by an LH1 ring, or a dimeric bc1 complex. The failure to find structures corresponding to the supercomplexes does not mean that these complexes can be excluded. However, if they exist, they do not give rise to obviously organized structures of the expected size, except perhaps in the tubular arrays. The strong curvature of the chromatophores, and their small size might make it dificult to distinguish organized structures from the discrete particles seen in electron micrographs.
  1. Lommen, M. A. and Takemoto, J. (1978) Comparison by freeze-fraction electronmicroscopy of chromatophores, spheroplast-derived membrane vesicles and whole cells of Rps. sphaeroides. J. Bacteriology 136, 730-741
  2. Yen, G. S. L., Cain, B. D. and Kaplan, S. (1984) Cell-cycle specific biosynthesis of the photosynthetic membrane of Rps. sphaeroides. Biochim. Biophys.Acta 777, 41-55
  3. Golecki, J. R. and Oelze, J. (1980) Differences in the architecture of the cytoplasmic and intracytoplasmic membranes of three chemotrophically and phototrophically grown species of Rhodospirillaceae. J. Bacteriology144, 781-788
  4. Golecki, J. R., Ventura, S. and Oelze, J. (1991) The architecture of unusual membranne tubes in the B800-850 light-harvesting bacteriochlorophyll-deficient mutant 19 of Rb. sphaeroides. FEMS Microbiology Lett. 77, 335-340.
  5. Kiley, P. J. and Kaplan, S. (1988) Micbrobiol. Rev. 52, 50-69
  6. Sabaty, M. Jappe, J. Olive, J. Vermeglio, A. (1994) Organization of electron transfer components in Rb. sphaeroides form a SP denitrificans whole cells. Biochim. Biophys. Acta 1187, 313-323