- Email: [email protected]
Photosynthesis in Rhodobacter sphaeroides
COMMENT
compatible solutes, spore formation and/or germination, mechanosensation, bacterial virulence and transport of toxic metalloids is being investigated, and other roles might yet be discovered. By contrast, unknown physiological mechanisms could be compensating for the lack of canonical MIP channels in some microorganisms1. Although microbial MIP proteins have potential medical, pharmaceutical, biotechnological and agricultural applications, use of the E. coli AqpZ (Ref. 9) and GlpF (Ref. 10) channels as model systems is already providing information useful for elucidating the biophysical, structural and biochemical features of the MIP family. However, considerable effort will be needed to identify additional MIP family members and to fully understand the existing ones.
Giuseppe Calamita Dipartimento di Fisiologia Generale e Ambientale, Università degli Studi di Bari, Via Amendola 165/A, 70126 Bari, Italy References 1 Hohmann, S. et al. (2000) Microbial MIP channels. Trends Microbiol. 8, 33–38 2 Park, J.H. and Saier, M.H., Jr (1996) Phylogenetic characterization of the MIP family of transmembrane channel proteins. J. Membr. Biol. 153, 171–180 3 Agre, P. et al. (1998) The aquaporins, blueprints for cellular plumbing systems. J. Biol. Chem. 273, 14659–14662 4 Kashiwagi, S. et al. (1995) A Synechococcus gene encoding a putative pore-forming intrinsic membrane protein. Biochim. Biophys. Acta 1237, 189–192
Photosynthesis in Rhodobacter sphaeroides
I
n their recent review, Verméglio and Joliot1 emphasized their view that the components of the Rhodobacter sphaeroides photosynthetic chain are organized as supercomplexes. This model was developed to explain the low equilibrium constant between cytochrome (cyt) c2 and the oxidized reaction center (RC) observed in kinetic experiments2. We have suggested an alternative view: electron transfer requires diffusion of ubiquinone, and cyt c2 couples between separate complexes3–6. We have been able to explain their original observations as the effects arising from the statistical distribution of components between small chromatophore vesicles, which would produce the same apparent change in equilibrium constant as they observed, without the need for supercomplexes6. It is an essential part of the supercomplex hypothesis that, on a time scale that is long compared with the ~1 ms turnover, cyt c2 is not able to diffuse from the supercomplex2. We have shown that in chromatophores, contrary to this expectation, QH2 or cyt c2 can rapidly visit ~9 bc1 complexes (approximately the number expected in a chromatophore vesicle)4. We have also demonstrated that in mutant strains in which the components are expressed in
TRENDS
IN
stoichiometric ratios that differ from those required for a supercomplex, all bc1 complexes, cyt c2 and RCs that interact rapidly following excitation by a short (5 ms) actinic flash, show the behavior expected from the diffusional model5,6 (S.J. Hong and A.R. Crofts, unpublished). Verméglio and Joliot1 discuss the structures of the ordered arrays of C-shaped particles observed in tubular membranes in R. sphaeroides strains deficient in light-harvesting complex (LH) 2 (Ref. 7), and claim that these support the supercomplex hypothesis. The structures probably represent the dimeric LH1–RC complex of Francia et al.8, but in that work, no bc1 complex was found in the dimer fraction. Although Jungas et al.7 reported the presence of the bc1 complex in tubular membranes at the stoichiometry expected, they found no structural evidence for an association between the bc1 and LH1–RC complexes. Verméglio and Joliot suggest ‘the electron density localized outside the C-shaped structures is tentatively attributed to the bc1 complex’. However, in the analysis by Jungas et al.7, the electron density was almost completely accounted for by RC and LH1, and any unaccounted density would have to include the one PufX protein molecule per RC found by Francia et al.8 The mitochondrial bc1 complex is dimeric, and requires dimeric
MICROBIOLOGY
105
VOL. 8
NO. 3
5 Yasui, M. et al. (1999) Rapid gating and anion permeability of an intracellular aquaporin. Nature 402, 184–187 6 Calamita, G. et al. (1998) Regulation of the Escherichia coli water channel gene aqpZ. Proc. Natl. Acad. Sci. U. S. A. 95, 3627–3631 7 Tamas, M.J. et al. (1999) Fps1p controls the accumulation and release of the compatible solute glycerol in yeast osmoregulation. Mol. Microbiol. 31, 1087–1104 8 Flick, K.M. et al. (1997) The wacA gene of Dictyostelium discoideum is a developmentally regulated member of the MIP family. Gene 195, 127–130 9 Scheuring, S. et al. (1999) Highresolution AFM topographs of the Escherichia coli water channel aquaporin Z. EMBO J. 18, 4981–4987 10 Lagrée, V. et al. (1998) Oligomerization state of water channels and glycerol facilitators. Involvement of loop E. J. Biol. Chem. 273, 33949–33953
association to bind the iron sulfur subunit9; the bc1 complex from R. sphaeroides also appears to be dimeric10. Such a dimeric complex would have roughly the same volume as that of an LH1–RC monomer, but there is no room for such a volume in the array, either within or outside the LH1–RC dimer. Indeed, there is no room for even a bc1 complex monomer between the C-shaped structures. No immediate external location seems likely, as the bc1 complex (one per LH1–RC dimer) would impose an asymmetry in the array (whether dimer or monomer); no such symmetry is apparent. Furthermore, any external location would require a substantial diffusional pathway between binding sites on the RC and bc1 complex, because of the thickness of the LH1 ring. This would be difficult to reconcile with the tight binding the supercomplex mechanism requires2.
Antony R. Crofts Dept of Biochemistry, University of Illinois at UrbanaChampaign, IL 61801, USA References 1 Verméglio, A.T. and Joliot, P. (1999) The photosynthetic apparatus of Rhodobacter sphaeroides. Trends Microbiol. 7, 435–440
MARCH 2000
COMMENT
2 Joliot, P. et al. (1989) Evidence for supercomplexes between reaction centers, cytochrome c2 and cytochrome bc1 complex. Biochim. Biophys. Acta 975, 336–345 3 Crofts, A.R. (1986) Reaction center and UQH2:cyt c2 oxidoreductase act as independent enzymes in Rps. sphaeroides. J. Bioenerg. Biomembr. 18, 437–446 4 Fernàndez-Velasco, J. and Crofts, A.R. (1991) Complexes or supercomplexes: inhibitor titrations show that electron transfer in chromatophores from Rhodobacter sphaeroides involves a
dimeric UQH2: cytochrome c2 oxidoreductase, and is delocalized. Biochem. Soc. Trans. 19, 588–593 5 Witthuhn, V.C. et al. (1996) The reactions of isocytochrome c2 in the photosynthetic chain of Rhodobacter sphaeroides. Biochemistry 36, 903–911 6 Crofts, A.R. et al. (1998) Chromatophore heterogeneity explains effects previously attributed to supercomplexes. Photosynth. Res. 55, 357–362 7 Jungas, C. et al. (1999) Supramolecular organization of the photosynthetic
Response from Verméglio and Joliot
C
rofts presents three main arguments against the supramolecular organization that we propose for the photosynthetic apparatus of Rhodobacter sphaeroides. Firstly, he suggests that the anomalous low apparent equilibrium constant between cytochrome (cyt) c2 and the photooxidized reaction center (RC) is a result of the statistical distribution of components between small chromatophore vesicles1. We agree that the relative amounts of RC and cyt c2 can vary, depending on the isolated chromatophores considered. However, why should such a statistical distribution be present in intact cells where all chromatophores are connected, especially if the cyt c2 can freely diffuse? Concerning the second argument based on inhibitor titrations, Crofts has admitted that ‘in contrast to the chromatophores results2, when inhibitor titration experiments were performed with whole cells, the electron transfer linked to cyt c2 is restricted to a local domain of 1–2 units, ... as would be expected from the supercomplex model’ (http://ahab.life.uiuc. edu/hetchrm.html). Crofts envisages two possibilities for this paradoxical behavior: (1) ‘supercomplexes are present in whole cells, but lost on the preparations of chromatophores’ and (2) ‘the restricted diffusion seen in cells is not due to supercomplexes, but to some structure in the periplasm, for example the murein sacculus...’. This second hypothesis is hardly tenable as it is well known that most of the photosynthetic apparatus is localized in the invaginated part of the membrane (i.e. at a significant distance from the
TRENDS
IN
bacterial wall). In contradiction with the Crofts’ proposal, cyt c2 can freely diffuse in the periplasmic space associated with the smooth part of the membrane where a small fraction of the photosynthetic apparatus interacts with the respiratory chains3. This diffusion process is significantly blocked by the addition of glycerol or magnesium ions, although these compounds do not affect the kinetics of photo-induced electron transfer for the photosynthetic chains organized in supercomplexes3–5. We conclude that both Crofts’ results and ours support the supercomplex hypothesis for whole cells of R. sphaeroides. In the related species Rhodospirillum rubrum, the cyt c2 can freely diffuse to interact with the bc1 complex and this reaction is blocked below 2108C (Ref. 6; A. Verméglio and P. Joliot, unpublished). By contrast, the complete photo-induced cyclic electron transfer occurs at temperatures down to 2208C in frozen medium for R. sphaeroides cells7. How can this result be explained if cyt c2 must diffuse to shuttle between the RC and bc1 complexes? Regarding the third argument, we agree with Crofts that there is no room for a dimeric bc1 complex in the ordered arrays observed in tubular membranes8. However, we believe that a monomeric bc1 complex can be located in the basic unit of this array. The statement that the functional bc1 complex has to be dimeric is by no means obvious. For example, a supercomplex associating one bc1 complex and one cytochrome oxidase complex has been purified from the thermophilic bacterium Bacillus PS3 (Ref. 9). Therefore, the bc1 complex can be found in monomeric form
MICROBIOLOGY
106
VOL. 8
NO. 3
apparatus of Rhodobacter sphaeroides. EMBO J. 18, 534–542 8 Francia, F. et al. (1999) Biochemistry 38, 6834–6845 9 Crofts, A.R. and Berry, E.A. (1998) Structure and function of the cytochrome bc1 complex of mitochondria and photosynthetic bacteria. Curr. Opin. Struct. Biol. 8, 501–509 10 Guergova-Kuras, M. et al. (1998) Expression and one-step purification of a fully active poly-histidine tagged bc1 complex from Rhodobacter sphaeroides. Protein Expres. Purif. 15, 370–380
if associated with other membrane proteins – which is our proposal for the photosynthetic apparatus of R. sphaeroides.
André Verméglio CEA-Cadarache-DSV, 13108 Saint Paul-lez-Durance Cedex, France Pierre Joliot Institut de Biologie PhysicoChimique, CNRS UPR 9072, 13 rue Pierre et Marie Curie, 75005 Paris, France References 1 Crofts, A.R. et al. (1998) Chromatophore heterogeneity explains phenomena seen in Rhodobacter sphaeroides previously attributed to supercomplexes. Photosynth. Res. 55, 357–362 2 Fernàndez-Velasco, J. and Crofts, A.R. (1991) Complex or supercomplex: inhibitor titration shows that electron transfer in chromatophores from Rhodobacter sphaeroides involves a dimeric UQH2: cytochrome c2 oxidase, and is delocalized. Biochem. Soc. Trans. 19, 588–593 3 Verméglio, A. et al. (1993) The rate of cytochrome c2 photooxidation reflects the subcellular distribution of reaction centers in Rhodobacter sphaeroides Ga cells. Biochim. Biophys. Acta 1183, 352–360 4 Sabaty, M. et al. (1994) Organization of electron transfer components in Rhodobacter sphaeroides forma sp. denitrificans whole cells. Biochim. Biophys. Acta 1187, 313–323 5 Matsuura, K. et al. (1988) Heterogeneous pools of cytochrome c2 in photodenitrifying cells of Rhodobacter sphaeriodes forma sp. denitrificans. J. Biochem. 104, 1016–1020
MARCH 2000
compatible solutes, spore formation and/or germination, mechanosensation, bacterial virulence and transport of toxic metalloids is being investigated, and other roles might yet be discovered. By contrast, unknown physiological mechanisms could be compensating for the lack of canonical MIP channels in some microorganisms1. Although microbial MIP proteins have potential medical, pharmaceutical, biotechnological and agricultural applications, use of the E. coli AqpZ (Ref. 9) and GlpF (Ref. 10) channels as model systems is already providing information useful for elucidating the biophysical, structural and biochemical features of the MIP family. However, considerable effort will be needed to identify additional MIP family members and to fully understand the existing ones.
Giuseppe Calamita Dipartimento di Fisiologia Generale e Ambientale, Università degli Studi di Bari, Via Amendola 165/A, 70126 Bari, Italy References 1 Hohmann, S. et al. (2000) Microbial MIP channels. Trends Microbiol. 8, 33–38 2 Park, J.H. and Saier, M.H., Jr (1996) Phylogenetic characterization of the MIP family of transmembrane channel proteins. J. Membr. Biol. 153, 171–180 3 Agre, P. et al. (1998) The aquaporins, blueprints for cellular plumbing systems. J. Biol. Chem. 273, 14659–14662 4 Kashiwagi, S. et al. (1995) A Synechococcus gene encoding a putative pore-forming intrinsic membrane protein. Biochim. Biophys. Acta 1237, 189–192
Photosynthesis in Rhodobacter sphaeroides
I
n their recent review, Verméglio and Joliot1 emphasized their view that the components of the Rhodobacter sphaeroides photosynthetic chain are organized as supercomplexes. This model was developed to explain the low equilibrium constant between cytochrome (cyt) c2 and the oxidized reaction center (RC) observed in kinetic experiments2. We have suggested an alternative view: electron transfer requires diffusion of ubiquinone, and cyt c2 couples between separate complexes3–6. We have been able to explain their original observations as the effects arising from the statistical distribution of components between small chromatophore vesicles, which would produce the same apparent change in equilibrium constant as they observed, without the need for supercomplexes6. It is an essential part of the supercomplex hypothesis that, on a time scale that is long compared with the ~1 ms turnover, cyt c2 is not able to diffuse from the supercomplex2. We have shown that in chromatophores, contrary to this expectation, QH2 or cyt c2 can rapidly visit ~9 bc1 complexes (approximately the number expected in a chromatophore vesicle)4. We have also demonstrated that in mutant strains in which the components are expressed in
TRENDS
IN
stoichiometric ratios that differ from those required for a supercomplex, all bc1 complexes, cyt c2 and RCs that interact rapidly following excitation by a short (5 ms) actinic flash, show the behavior expected from the diffusional model5,6 (S.J. Hong and A.R. Crofts, unpublished). Verméglio and Joliot1 discuss the structures of the ordered arrays of C-shaped particles observed in tubular membranes in R. sphaeroides strains deficient in light-harvesting complex (LH) 2 (Ref. 7), and claim that these support the supercomplex hypothesis. The structures probably represent the dimeric LH1–RC complex of Francia et al.8, but in that work, no bc1 complex was found in the dimer fraction. Although Jungas et al.7 reported the presence of the bc1 complex in tubular membranes at the stoichiometry expected, they found no structural evidence for an association between the bc1 and LH1–RC complexes. Verméglio and Joliot suggest ‘the electron density localized outside the C-shaped structures is tentatively attributed to the bc1 complex’. However, in the analysis by Jungas et al.7, the electron density was almost completely accounted for by RC and LH1, and any unaccounted density would have to include the one PufX protein molecule per RC found by Francia et al.8 The mitochondrial bc1 complex is dimeric, and requires dimeric
MICROBIOLOGY
105
VOL. 8
NO. 3
5 Yasui, M. et al. (1999) Rapid gating and anion permeability of an intracellular aquaporin. Nature 402, 184–187 6 Calamita, G. et al. (1998) Regulation of the Escherichia coli water channel gene aqpZ. Proc. Natl. Acad. Sci. U. S. A. 95, 3627–3631 7 Tamas, M.J. et al. (1999) Fps1p controls the accumulation and release of the compatible solute glycerol in yeast osmoregulation. Mol. Microbiol. 31, 1087–1104 8 Flick, K.M. et al. (1997) The wacA gene of Dictyostelium discoideum is a developmentally regulated member of the MIP family. Gene 195, 127–130 9 Scheuring, S. et al. (1999) Highresolution AFM topographs of the Escherichia coli water channel aquaporin Z. EMBO J. 18, 4981–4987 10 Lagrée, V. et al. (1998) Oligomerization state of water channels and glycerol facilitators. Involvement of loop E. J. Biol. Chem. 273, 33949–33953
association to bind the iron sulfur subunit9; the bc1 complex from R. sphaeroides also appears to be dimeric10. Such a dimeric complex would have roughly the same volume as that of an LH1–RC monomer, but there is no room for such a volume in the array, either within or outside the LH1–RC dimer. Indeed, there is no room for even a bc1 complex monomer between the C-shaped structures. No immediate external location seems likely, as the bc1 complex (one per LH1–RC dimer) would impose an asymmetry in the array (whether dimer or monomer); no such symmetry is apparent. Furthermore, any external location would require a substantial diffusional pathway between binding sites on the RC and bc1 complex, because of the thickness of the LH1 ring. This would be difficult to reconcile with the tight binding the supercomplex mechanism requires2.
Antony R. Crofts Dept of Biochemistry, University of Illinois at UrbanaChampaign, IL 61801, USA References 1 Verméglio, A.T. and Joliot, P. (1999) The photosynthetic apparatus of Rhodobacter sphaeroides. Trends Microbiol. 7, 435–440
MARCH 2000
COMMENT
2 Joliot, P. et al. (1989) Evidence for supercomplexes between reaction centers, cytochrome c2 and cytochrome bc1 complex. Biochim. Biophys. Acta 975, 336–345 3 Crofts, A.R. (1986) Reaction center and UQH2:cyt c2 oxidoreductase act as independent enzymes in Rps. sphaeroides. J. Bioenerg. Biomembr. 18, 437–446 4 Fernàndez-Velasco, J. and Crofts, A.R. (1991) Complexes or supercomplexes: inhibitor titrations show that electron transfer in chromatophores from Rhodobacter sphaeroides involves a
dimeric UQH2: cytochrome c2 oxidoreductase, and is delocalized. Biochem. Soc. Trans. 19, 588–593 5 Witthuhn, V.C. et al. (1996) The reactions of isocytochrome c2 in the photosynthetic chain of Rhodobacter sphaeroides. Biochemistry 36, 903–911 6 Crofts, A.R. et al. (1998) Chromatophore heterogeneity explains effects previously attributed to supercomplexes. Photosynth. Res. 55, 357–362 7 Jungas, C. et al. (1999) Supramolecular organization of the photosynthetic
Response from Verméglio and Joliot
C
rofts presents three main arguments against the supramolecular organization that we propose for the photosynthetic apparatus of Rhodobacter sphaeroides. Firstly, he suggests that the anomalous low apparent equilibrium constant between cytochrome (cyt) c2 and the photooxidized reaction center (RC) is a result of the statistical distribution of components between small chromatophore vesicles1. We agree that the relative amounts of RC and cyt c2 can vary, depending on the isolated chromatophores considered. However, why should such a statistical distribution be present in intact cells where all chromatophores are connected, especially if the cyt c2 can freely diffuse? Concerning the second argument based on inhibitor titrations, Crofts has admitted that ‘in contrast to the chromatophores results2, when inhibitor titration experiments were performed with whole cells, the electron transfer linked to cyt c2 is restricted to a local domain of 1–2 units, ... as would be expected from the supercomplex model’ (http://ahab.life.uiuc. edu/hetchrm.html). Crofts envisages two possibilities for this paradoxical behavior: (1) ‘supercomplexes are present in whole cells, but lost on the preparations of chromatophores’ and (2) ‘the restricted diffusion seen in cells is not due to supercomplexes, but to some structure in the periplasm, for example the murein sacculus...’. This second hypothesis is hardly tenable as it is well known that most of the photosynthetic apparatus is localized in the invaginated part of the membrane (i.e. at a significant distance from the
TRENDS
IN
bacterial wall). In contradiction with the Crofts’ proposal, cyt c2 can freely diffuse in the periplasmic space associated with the smooth part of the membrane where a small fraction of the photosynthetic apparatus interacts with the respiratory chains3. This diffusion process is significantly blocked by the addition of glycerol or magnesium ions, although these compounds do not affect the kinetics of photo-induced electron transfer for the photosynthetic chains organized in supercomplexes3–5. We conclude that both Crofts’ results and ours support the supercomplex hypothesis for whole cells of R. sphaeroides. In the related species Rhodospirillum rubrum, the cyt c2 can freely diffuse to interact with the bc1 complex and this reaction is blocked below 2108C (Ref. 6; A. Verméglio and P. Joliot, unpublished). By contrast, the complete photo-induced cyclic electron transfer occurs at temperatures down to 2208C in frozen medium for R. sphaeroides cells7. How can this result be explained if cyt c2 must diffuse to shuttle between the RC and bc1 complexes? Regarding the third argument, we agree with Crofts that there is no room for a dimeric bc1 complex in the ordered arrays observed in tubular membranes8. However, we believe that a monomeric bc1 complex can be located in the basic unit of this array. The statement that the functional bc1 complex has to be dimeric is by no means obvious. For example, a supercomplex associating one bc1 complex and one cytochrome oxidase complex has been purified from the thermophilic bacterium Bacillus PS3 (Ref. 9). Therefore, the bc1 complex can be found in monomeric form
MICROBIOLOGY
106
VOL. 8
NO. 3
apparatus of Rhodobacter sphaeroides. EMBO J. 18, 534–542 8 Francia, F. et al. (1999) Biochemistry 38, 6834–6845 9 Crofts, A.R. and Berry, E.A. (1998) Structure and function of the cytochrome bc1 complex of mitochondria and photosynthetic bacteria. Curr. Opin. Struct. Biol. 8, 501–509 10 Guergova-Kuras, M. et al. (1998) Expression and one-step purification of a fully active poly-histidine tagged bc1 complex from Rhodobacter sphaeroides. Protein Expres. Purif. 15, 370–380
if associated with other membrane proteins – which is our proposal for the photosynthetic apparatus of R. sphaeroides.
André Verméglio CEA-Cadarache-DSV, 13108 Saint Paul-lez-Durance Cedex, France Pierre Joliot Institut de Biologie PhysicoChimique, CNRS UPR 9072, 13 rue Pierre et Marie Curie, 75005 Paris, France References 1 Crofts, A.R. et al. (1998) Chromatophore heterogeneity explains phenomena seen in Rhodobacter sphaeroides previously attributed to supercomplexes. Photosynth. Res. 55, 357–362 2 Fernàndez-Velasco, J. and Crofts, A.R. (1991) Complex or supercomplex: inhibitor titration shows that electron transfer in chromatophores from Rhodobacter sphaeroides involves a dimeric UQH2: cytochrome c2 oxidase, and is delocalized. Biochem. Soc. Trans. 19, 588–593 3 Verméglio, A. et al. (1993) The rate of cytochrome c2 photooxidation reflects the subcellular distribution of reaction centers in Rhodobacter sphaeroides Ga cells. Biochim. Biophys. Acta 1183, 352–360 4 Sabaty, M. et al. (1994) Organization of electron transfer components in Rhodobacter sphaeroides forma sp. denitrificans whole cells. Biochim. Biophys. Acta 1187, 313–323 5 Matsuura, K. et al. (1988) Heterogeneous pools of cytochrome c2 in photodenitrifying cells of Rhodobacter sphaeriodes forma sp. denitrificans. J. Biochem. 104, 1016–1020
MARCH 2000