- Email: [email protected]
oxygenase from Alcaligenes eutrophus H16 cells, in the active and CABP-inhibited forms
J. Mol. Biol. (1989) 207, 621-623
Comparative Studies of Ribulose-1,5biphosphate Carboxylase/Oxygenase from Alcaligenes eutrophus H16 cells, in the Active and CABP-inhibited Forms RuRisCO (D-ribulose-1,5-biphosphate carboxylase/oxygenase; EC 4.1.1.39) has been isolated from the autotrophic hydrogen-oxidizing bacterium Alcaligenes eutrophus H16. Combining photon correlation and sedimentation analysis transport parameters of the enzyme were investigated in the active, (E * CO,.Mg’+) as a ternary complex, and inactive 2+.CABP) as a quaternary complex, where RuBisCO is complexed with state. (E.CO,.Mg the transition state analogue CABP (2-C-carboxy-o-arabinitol-1,5-biphosphate). Within experimental error, no difference has been detected between the diffusion and sedimentation coefficients (I$!,,, = 2.72( +007) x 1Om~7cm2 s-l, s;~,,, = 17.Q + 05)s) of active and CARPcomplexed enzyme thus leading to the conclusion t’hat the molecule, at least in solution. does not assume a different conformation when complexed with CABP.
crystallographic S-fold axis parallel to the crystallographic a axis. It has been proposed t’hat this sliding is associated with the highly specific binding of CABP. In order to answer the question whether active and CABP-inhibited enzyme indeed assume different structural conformations, we have undertaken solution measurements on both enzyme forms. We employed the technique of photon correlation spectroscopy (Berne & Pecora, 1976) combined with sedimentation analysis for deriving reliable estimates of the diffusion and sedimentation coefficients of the en?yme. Conformational changes, resulting from a 36 A lateral sliding of the subunits, should be detectable from changes in the transport properties of the two enzyme forms. ,411 materials used were of analvtical grade. Before each type of measurement, active RuBisCO solutions were dialysed against 20 mM-Tris . HCl (pH 8.0), 10 mM-MgC1,.6 H,O, 50 mM-,l’aHCO,, 1 mM-DTE, 1 mM-EDTA. CABP was synthesized as described (Pierce et al., 1980) and its structure verified by nuclear magnetic resonance spectroscopy. The isolation of active RuBisCO was accomplished according t’o the scheme suggested by Kowien et al. (1976) with minor modifications. All experiments were conducted at 20*0( kO.l)Y’. For photon correlation experiments, active RuBisCO solutions were repeatedly passed through 025 ,um pore size Millex filters and injected into round light-scattering cells. Similarly filtered portions of CABP were added later when measurements on the active enzyme were completed. The enzyme to CARP stoichiometry was 1 : 40. Two optical arrangements manufactured by ALV (Langen, FRG) were used. The first apparatus was equipped with an Ar+ ion laser operated at a wavelength of 457.9 nm. Data were collected with a 512channel ALV correlator at scattering angles of 30”, 40”, 60”, 90”, 120” and, in a few cases, 150”. For
RuBisCOt is one of the most important bifunctional enzymes known as it plays a key role in both the photorespiration and photosynthetic cycles. Several laboratories have made noteworthy progress in elucidating the crystal structure of various RuBisCO types (for references, see Chapman et al., 1987; Suh et aZ., 1987). Despite its usually prohibitively large size and consequently vast amounts of data required for attaining atomic-level resolution, a significant amount of knowledge concerning the crystal structure of this enzyme is available. Several solution studies Including neutron scattering (Donnely et al., 1984), small-angle X-ray scattering (Meisenberger et al., 1984), and analytical ultracentrifugation studies (Bowien & Gottschalk, 1982; Suh et al., 1987) provide a rather contradictory pict)ure, describing changes in the conformation of the active (in the presence of Mg2+ and HCO,), C’ABP-inhibited and inactive (in the absence of Mg2 + and HCO;) forms of the enzyme from various sources. We have investigated the transport properties of the active and CARP-inhibited RuBisCO, when complexed with the transition state analogue CABP (Pierce et al., 1980). The inhibitor binds to the large subunits with a very high affinity, exhibiting a dissociat,ion constant of the order of 10-l’ M. We have reported (Holzenburg et aZ., 1987) an electron microscopy study of the active enzyme f;om Alcaligems eutrophus. as well as a 5 A (1 A = 0.1 nm) resolution X-ray crystal structure of CABP-inhibited RuBisCO. The crystal structure of the enzyme shows two local 4-fold axes for the “top” and “bottoy” L,S, halves which, however, are shifted by 36 .4 relative to each other and related by a t Abbreviations used: RuBisCO, n-ribulose-l&biphosphate c*arboxylase/oxygenase (EC 4.1.1.39); CABP, Z-c’carboxy-D-arabinitol-1,5-biphosphate; DTE, dithiorrythritol: PMSF. Y-phenylmethan-1-sulphonylfluoride.
621 0022-2836/89/l
10621+3 $03.00/0
0 1989 Academic Press Limited
622
H.-W.
higher concentrations of RuBisCO, data collection was performed with the second instrument, which was equipped with a He-Ne laser operated at 632.8 nm and a B12030 12%channel correlator (Brookhaven Instruments, U.S.A.) using the same scattering angles. A comparative semi-logarithmic representation of autocorrelation functions for both enzymes forms versus delay time is shown displayed in Figure I. Spectra evaluation was carried out with the program CONTTN (Provencher, 1982a,b) in the linewidth space using a logarithmic 31-point inversion grid as described (Georgalis et al., 1987). The obtained particle distributions were narrow and monomodal for both RuBisCO forms, with variances between 0.1 and 0.12. The free-particle diffusion coefficient determined (Fig. 2(a)) was 2.72( kOO7) x lop7 cm2 s-l for both the active and CABP-inhibited RuBisCO forms, correspondjng to an effective hydrodynamic radius of 74( + 3) A. The value of the frictional ratio, flfmi” = 1.37, suggests an elongated and highly similar shape for both enzyme types. A Heckman model E analytical ultracentrifuge equipped with schlieren and ultraviolet scanner optics was employed for sedimentation velocity experiments. A capillary-type synthetic boundary cell and an An-H rotor was used throughout, at speeds between 1.2 x lo4 and 1.4 x lo4 revs/min. Photographs and scans were recorded every eight and within detectable limits. minutes. Single, symmetric peaks were observed and sedimentation coefficients were determined from the rate of movement> of the peak maximum. The sedimentation coefficient was 17+3( fO5) S for both RuBisCO forms (Fig. 2(b)). Combining t’he corrected sedimentation and diffusion data into the Svedberg relation (Tanford, 1961) and assuming a partial specific volume of 0.73 ml/g, we obtained a molecular weight of 5.75 x lo5 g/mol for RuBisCO. This value is within 5:& of the value
2.0 -I-bl
--km
I.0 -
9.0 @P se2 40 co*rJ l 20 l .0. l
0.0 -
.“.o
o
l .2 000
l ‘.;.“o
l “.O0 l ..“o.
-0.1 0.0
15.0
30-O T (ps)
I 45.0
l. I 60.0
Figure 1. Typical plots of the logarithm of t,he field autocorrelation function g’(t) ~ers’us delay time T for the (a) active. and (0) CARP-inhibited RuBisCO at an enzyme concentration of 263 mg/ml. Data were collected at a srattering angle of 90”. For clarity. only every 2nd data point is plotted.
Choe et al.
o,o
2.0
4.0
6.0
C (mg/ml)
Figure 2. (a) Diffusion coefficient and (b) sedimentation c-oefficient as a function of concentration for (u) active and (0) CARP-inhibited RuBisCO. Open symbols denote data obtained using the ultraviolet light optics at 285 rim. The best fit straight lines to each dataset are indicated.
deduced by sequencing the DNA of the ATCC17707 A. eutrophus RuBisCO (Andersen & Caton, 1987). Theories for determining the transport properties of multisubunit systems from the properties of the constituent protomers are valid only for spatially equivalent promoters (Andrews & Jeffrey, 1976, 1980; Jeffrey & Andrews, 1980; Garcia De La Torre & Bloomfield, 1981). If no lateral shift is present between the two L,&, subunits, the overall structure of RuBisCO can be approximated as a prolate ellipsoid of revolution with semiaxes 60 8, x 50 A. Assuming that the shape of an LS complex is close to spherical, we can attempt a Kirkwood-type calculation for a multisubunit structure like RuBisCO. Despite its approximate nature. this calculation is sufficient for working out conformational changes of large magnitude. We estimated that a 36 .& lateral shift between the L,S4 subunits will yield at least a 16% increment of the frictional coefficient of the CABP-inhibited enzyme. Since our experiments were conducted with less than 5?& error, we conclude that within the accuracy of the methods employed, RuBisCO in the active and CABPinhibited form has the same overall shape. This is in line with the work of Chapman et al. (1987), who suggested that a sliding layer-type conformational change should be limited by the interdigitating quaternary structure of RuBisCO from Nicotiana tnhncum. We have no explanation for the discrepancy with our earlier results (Holzenberg et al.: 1987). We hope t,hat. the picture will become clearer when a higher-resolution crystal structure of il. eutrophus RuBisCO is available. We thank Professor B. Bowien for supplying t’he dlcaligenes cells and Drs Cr. Meier and M. Schmtdt of t,he MaxPlanck-Institut fiir Polymerforschung, Mainz. for making available to us their photon c-or-relation fac4ities. The
Letters to the Editor
present work was financially supported through Sonderforschungsbereich 9, by the Fonds der Chemischen Industrie and by a grant of the European Community to Y.G. (ST2J-0034-l/F).
Hui-Woog Choe Yam& Georgalist Wolfram Saenger Institut fur Kristallographie Freie Universitat Berlin Takust. 6,lOOO Berlin 33, F.R.G. Received 9 May 1988, and in revised form 27 January 1989 t Author to whom all correspondence
should be sent,
623
Bowien, B., Mayer, F., Codd, G. A. & Schlegel, H. Q. (1976). Arch. Microbial. 110. 157-166. Chapman, M. S., Suh, 6. W., Cascio. D.. Smith, W. W. & Eisenberg, D. (1987). Nature (London), 379, 3,54356. Donnely, M. I., Hartman, F. C. & Ramakrishnan, V. (1984). J. Biol. Chem. 259, 406-411. Garcia De La Torre, J. & Bloomfield. V. A. (1981). Quart. Rev. Biophys. 14, 81-139. Georgalis, Y., Ruf, H. & Grell, E. (1987). In Topics in Molecular Pharmacology (Burgen, A. S. V.. Roberts. G. C. K. & Anner, B. M., eds), pp. l-20, Elsevier, Amsterdam. Holzenburg. A.. Mayer, F.. Harauz. G.. van Heel. M.. Tokuoka: R., Ishida, T.. Harata, K., Pal. G. 1’. & 325, 730-732. Saenger, W. (1987). Nature (Lon.don), Jeffrey, P. D. & Andrews, P. R. (1980). Hiophys. (‘hem.
11, 61-70. References Andersen, K. & Caton, J. (1987). J. Bacterial. 169, 45474558. Andrews, P. R. & Jeffrey, P. D. (1976). Biophys. Chem. 4, 933102. Andrews, P. R. & Jeffrey, P. D. (1980). Biophys. Chem.
11, 49-59.
Meisenberger. 0.. Pilz. I.. Bowien, B., Pal, G. 1’ & Saenger, W. (1984). J. Biol. Chem. 259. 4463-446.5. Pierce, J., Tolbert, N‘. E. & Barker. R. (I 980). Hiochemiatry, 19, 934-942. Provencher, S. 1%‘. (1982a). Camp. Ph,ys. (‘ommun. 27. 213-227. Provenrher. S. W. (1982b). Camp. Phys. (Jommun. 27, 229-242.
Berne, B. J. & Pecora, R. (1976). Dynamic Light tering, Wiley, New York. Bowien. B. & Gottschalk, E.-M. (1982). J. Bid. 257. 1184511847.
ScatChem.
Sub. W. 8.. Cascio, D.. Chapman. M. S. & Eisenberg, I). (1987). J. Mol. Biol. 197, 363-365. Tanford. C. (1961). In Physical Chemistry of Macromolecules, Wiley, New York.
Edited by B. W. Matthews
Comparative Studies of Ribulose-1,5biphosphate Carboxylase/Oxygenase from Alcaligenes eutrophus H16 cells, in the Active and CABP-inhibited Forms RuRisCO (D-ribulose-1,5-biphosphate carboxylase/oxygenase; EC 4.1.1.39) has been isolated from the autotrophic hydrogen-oxidizing bacterium Alcaligenes eutrophus H16. Combining photon correlation and sedimentation analysis transport parameters of the enzyme were investigated in the active, (E * CO,.Mg’+) as a ternary complex, and inactive 2+.CABP) as a quaternary complex, where RuBisCO is complexed with state. (E.CO,.Mg the transition state analogue CABP (2-C-carboxy-o-arabinitol-1,5-biphosphate). Within experimental error, no difference has been detected between the diffusion and sedimentation coefficients (I$!,,, = 2.72( +007) x 1Om~7cm2 s-l, s;~,,, = 17.Q + 05)s) of active and CARPcomplexed enzyme thus leading to the conclusion t’hat the molecule, at least in solution. does not assume a different conformation when complexed with CABP.
crystallographic S-fold axis parallel to the crystallographic a axis. It has been proposed t’hat this sliding is associated with the highly specific binding of CABP. In order to answer the question whether active and CABP-inhibited enzyme indeed assume different structural conformations, we have undertaken solution measurements on both enzyme forms. We employed the technique of photon correlation spectroscopy (Berne & Pecora, 1976) combined with sedimentation analysis for deriving reliable estimates of the diffusion and sedimentation coefficients of the en?yme. Conformational changes, resulting from a 36 A lateral sliding of the subunits, should be detectable from changes in the transport properties of the two enzyme forms. ,411 materials used were of analvtical grade. Before each type of measurement, active RuBisCO solutions were dialysed against 20 mM-Tris . HCl (pH 8.0), 10 mM-MgC1,.6 H,O, 50 mM-,l’aHCO,, 1 mM-DTE, 1 mM-EDTA. CABP was synthesized as described (Pierce et al., 1980) and its structure verified by nuclear magnetic resonance spectroscopy. The isolation of active RuBisCO was accomplished according t’o the scheme suggested by Kowien et al. (1976) with minor modifications. All experiments were conducted at 20*0( kO.l)Y’. For photon correlation experiments, active RuBisCO solutions were repeatedly passed through 025 ,um pore size Millex filters and injected into round light-scattering cells. Similarly filtered portions of CABP were added later when measurements on the active enzyme were completed. The enzyme to CARP stoichiometry was 1 : 40. Two optical arrangements manufactured by ALV (Langen, FRG) were used. The first apparatus was equipped with an Ar+ ion laser operated at a wavelength of 457.9 nm. Data were collected with a 512channel ALV correlator at scattering angles of 30”, 40”, 60”, 90”, 120” and, in a few cases, 150”. For
RuBisCOt is one of the most important bifunctional enzymes known as it plays a key role in both the photorespiration and photosynthetic cycles. Several laboratories have made noteworthy progress in elucidating the crystal structure of various RuBisCO types (for references, see Chapman et al., 1987; Suh et aZ., 1987). Despite its usually prohibitively large size and consequently vast amounts of data required for attaining atomic-level resolution, a significant amount of knowledge concerning the crystal structure of this enzyme is available. Several solution studies Including neutron scattering (Donnely et al., 1984), small-angle X-ray scattering (Meisenberger et al., 1984), and analytical ultracentrifugation studies (Bowien & Gottschalk, 1982; Suh et al., 1987) provide a rather contradictory pict)ure, describing changes in the conformation of the active (in the presence of Mg2+ and HCO,), C’ABP-inhibited and inactive (in the absence of Mg2 + and HCO;) forms of the enzyme from various sources. We have investigated the transport properties of the active and CARP-inhibited RuBisCO, when complexed with the transition state analogue CABP (Pierce et al., 1980). The inhibitor binds to the large subunits with a very high affinity, exhibiting a dissociat,ion constant of the order of 10-l’ M. We have reported (Holzenburg et aZ., 1987) an electron microscopy study of the active enzyme f;om Alcaligems eutrophus. as well as a 5 A (1 A = 0.1 nm) resolution X-ray crystal structure of CABP-inhibited RuBisCO. The crystal structure of the enzyme shows two local 4-fold axes for the “top” and “bottoy” L,S, halves which, however, are shifted by 36 .4 relative to each other and related by a t Abbreviations used: RuBisCO, n-ribulose-l&biphosphate c*arboxylase/oxygenase (EC 4.1.1.39); CABP, Z-c’carboxy-D-arabinitol-1,5-biphosphate; DTE, dithiorrythritol: PMSF. Y-phenylmethan-1-sulphonylfluoride.
621 0022-2836/89/l
10621+3 $03.00/0
0 1989 Academic Press Limited
622
H.-W.
higher concentrations of RuBisCO, data collection was performed with the second instrument, which was equipped with a He-Ne laser operated at 632.8 nm and a B12030 12%channel correlator (Brookhaven Instruments, U.S.A.) using the same scattering angles. A comparative semi-logarithmic representation of autocorrelation functions for both enzymes forms versus delay time is shown displayed in Figure I. Spectra evaluation was carried out with the program CONTTN (Provencher, 1982a,b) in the linewidth space using a logarithmic 31-point inversion grid as described (Georgalis et al., 1987). The obtained particle distributions were narrow and monomodal for both RuBisCO forms, with variances between 0.1 and 0.12. The free-particle diffusion coefficient determined (Fig. 2(a)) was 2.72( kOO7) x lop7 cm2 s-l for both the active and CABP-inhibited RuBisCO forms, correspondjng to an effective hydrodynamic radius of 74( + 3) A. The value of the frictional ratio, flfmi” = 1.37, suggests an elongated and highly similar shape for both enzyme types. A Heckman model E analytical ultracentrifuge equipped with schlieren and ultraviolet scanner optics was employed for sedimentation velocity experiments. A capillary-type synthetic boundary cell and an An-H rotor was used throughout, at speeds between 1.2 x lo4 and 1.4 x lo4 revs/min. Photographs and scans were recorded every eight and within detectable limits. minutes. Single, symmetric peaks were observed and sedimentation coefficients were determined from the rate of movement> of the peak maximum. The sedimentation coefficient was 17+3( fO5) S for both RuBisCO forms (Fig. 2(b)). Combining t’he corrected sedimentation and diffusion data into the Svedberg relation (Tanford, 1961) and assuming a partial specific volume of 0.73 ml/g, we obtained a molecular weight of 5.75 x lo5 g/mol for RuBisCO. This value is within 5:& of the value
2.0 -I-bl
--km
I.0 -
9.0 @P se2 40 co*rJ l 20 l .0. l
0.0 -
.“.o
o
l .2 000
l ‘.;.“o
l “.O0 l ..“o.
-0.1 0.0
15.0
30-O T (ps)
I 45.0
l. I 60.0
Figure 1. Typical plots of the logarithm of t,he field autocorrelation function g’(t) ~ers’us delay time T for the (a) active. and (0) CARP-inhibited RuBisCO at an enzyme concentration of 263 mg/ml. Data were collected at a srattering angle of 90”. For clarity. only every 2nd data point is plotted.
Choe et al.
o,o
2.0
4.0
6.0
C (mg/ml)
Figure 2. (a) Diffusion coefficient and (b) sedimentation c-oefficient as a function of concentration for (u) active and (0) CARP-inhibited RuBisCO. Open symbols denote data obtained using the ultraviolet light optics at 285 rim. The best fit straight lines to each dataset are indicated.
deduced by sequencing the DNA of the ATCC17707 A. eutrophus RuBisCO (Andersen & Caton, 1987). Theories for determining the transport properties of multisubunit systems from the properties of the constituent protomers are valid only for spatially equivalent promoters (Andrews & Jeffrey, 1976, 1980; Jeffrey & Andrews, 1980; Garcia De La Torre & Bloomfield, 1981). If no lateral shift is present between the two L,&, subunits, the overall structure of RuBisCO can be approximated as a prolate ellipsoid of revolution with semiaxes 60 8, x 50 A. Assuming that the shape of an LS complex is close to spherical, we can attempt a Kirkwood-type calculation for a multisubunit structure like RuBisCO. Despite its approximate nature. this calculation is sufficient for working out conformational changes of large magnitude. We estimated that a 36 .& lateral shift between the L,S4 subunits will yield at least a 16% increment of the frictional coefficient of the CABP-inhibited enzyme. Since our experiments were conducted with less than 5?& error, we conclude that within the accuracy of the methods employed, RuBisCO in the active and CABPinhibited form has the same overall shape. This is in line with the work of Chapman et al. (1987), who suggested that a sliding layer-type conformational change should be limited by the interdigitating quaternary structure of RuBisCO from Nicotiana tnhncum. We have no explanation for the discrepancy with our earlier results (Holzenberg et al.: 1987). We hope t,hat. the picture will become clearer when a higher-resolution crystal structure of il. eutrophus RuBisCO is available. We thank Professor B. Bowien for supplying t’he dlcaligenes cells and Drs Cr. Meier and M. Schmtdt of t,he MaxPlanck-Institut fiir Polymerforschung, Mainz. for making available to us their photon c-or-relation fac4ities. The
Letters to the Editor
present work was financially supported through Sonderforschungsbereich 9, by the Fonds der Chemischen Industrie and by a grant of the European Community to Y.G. (ST2J-0034-l/F).
Hui-Woog Choe Yam& Georgalist Wolfram Saenger Institut fur Kristallographie Freie Universitat Berlin Takust. 6,lOOO Berlin 33, F.R.G. Received 9 May 1988, and in revised form 27 January 1989 t Author to whom all correspondence
should be sent,
623
Bowien, B., Mayer, F., Codd, G. A. & Schlegel, H. Q. (1976). Arch. Microbial. 110. 157-166. Chapman, M. S., Suh, 6. W., Cascio. D.. Smith, W. W. & Eisenberg, D. (1987). Nature (London), 379, 3,54356. Donnely, M. I., Hartman, F. C. & Ramakrishnan, V. (1984). J. Biol. Chem. 259, 406-411. Garcia De La Torre, J. & Bloomfield. V. A. (1981). Quart. Rev. Biophys. 14, 81-139. Georgalis, Y., Ruf, H. & Grell, E. (1987). In Topics in Molecular Pharmacology (Burgen, A. S. V.. Roberts. G. C. K. & Anner, B. M., eds), pp. l-20, Elsevier, Amsterdam. Holzenburg. A.. Mayer, F.. Harauz. G.. van Heel. M.. Tokuoka: R., Ishida, T.. Harata, K., Pal. G. 1’. & 325, 730-732. Saenger, W. (1987). Nature (Lon.don), Jeffrey, P. D. & Andrews, P. R. (1980). Hiophys. (‘hem.
11, 61-70. References Andersen, K. & Caton, J. (1987). J. Bacterial. 169, 45474558. Andrews, P. R. & Jeffrey, P. D. (1976). Biophys. Chem. 4, 933102. Andrews, P. R. & Jeffrey, P. D. (1980). Biophys. Chem.
11, 49-59.
Meisenberger. 0.. Pilz. I.. Bowien, B., Pal, G. 1’ & Saenger, W. (1984). J. Biol. Chem. 259. 4463-446.5. Pierce, J., Tolbert, N‘. E. & Barker. R. (I 980). Hiochemiatry, 19, 934-942. Provencher, S. 1%‘. (1982a). Camp. Ph,ys. (‘ommun. 27. 213-227. Provenrher. S. W. (1982b). Camp. Phys. (Jommun. 27, 229-242.
Berne, B. J. & Pecora, R. (1976). Dynamic Light tering, Wiley, New York. Bowien. B. & Gottschalk, E.-M. (1982). J. Bid. 257. 1184511847.
ScatChem.
Sub. W. 8.. Cascio, D.. Chapman. M. S. & Eisenberg, I). (1987). J. Mol. Biol. 197, 363-365. Tanford. C. (1961). In Physical Chemistry of Macromolecules, Wiley, New York.
Edited by B. W. Matthews