The Molar Mass of an Active Photosystem I Complex from the
Cyanobacterium Synechococcus PCC 7002
C. Tziatzios, P. Schuck, D. Schubert
In stitu t für Biophysik d er J. W. G o eth e-U n iv ersität, T h eo d o r-S tern -K ai 7, H aus 74,
D -60590 F ra n k fu rt am M ain, B undesrepublik D eutschland
and
G. Tsiotis
M ax -P lanck -In stitu t fü r B iophysik, H ein rich -H o ffm an n -S traß e 7,
D-60528 F ra n k fu rt am M ain, B undesrepublik D eutschland
Z. N aturfo rsch . 49c, 2 2 0 -2 2 2 (1994); received D ecem ber 8 , 1994
P hotosystem I, Synechococcus, P ro tein -D ete rg en t C om plexes, M o lar M ass D eterm ination,
A nalytical U ltracen trifu g atio n
T he m o lar m ass o f p h o to sy stem I (PS I) from the cyan o b acteriu m Synechococcus PC C 7002
was determ ined by sedim en tatio n equilibrium analysis in the analytical ultracentrifuge. F o r
th a t purpo se, the “trim eric” form o f PS I was isolated and studied in three different nonionic
detergents. A fter d eterm ining the p artial specific volum e o f the p rotein/pigm ent complex,
M r was obtained as 830,000 ± 60,000.
from the ultracentrifuge d ata, as (0.788 ± 0.010) ml g
Photosystem I (PS I) is a m ultiprotein complex,
which mediates light-driven electron transfer from
plastocyanin or cytochrome c 553 to ferredoxin. The
PS I complex is located in the thylakoid membrane
of higher plants, green algae and cyanobacteria
(Goldbeck and Bryant, 1991). Active “m onom er­
ic” and “trim eric” forms o f the cyanobacterial PS I
have been isolated by using nonionic detergents
for the solubilization o f the thylakoid membrane
(Rögner et al., 1990a,b; Tsiotis et al., 1993). The
isolated “m onom eric” PS I contains two protein
subunits o f about 83 kD a, to which 60-100 chlo­
rophyll a (Chi a) molecules are bound. In addition,
the PS I complex contains a series of other sub­
units with m olar masses below 20 kD a (Goldbeck
and Bryant, 1991). The m olar mass o f “m onom er­
ic” and “trim eric” PS I was estimated by electron
microscopy and measurements o f the apparent
Stokes radius by HPLC gel filtration and was
found tp be around 250-300 kD a for the “m ono­
meric” and 700-750 kD a for the “trim eric” form
(Rögner et al., 1990a, b; Tsiotis et al., 1993).
It is well-known that the m ethods used for the
m olar mass determ ination of the cyanobacterial
PS I complex are based on assum ptions which may
R eprint requests to Prof. D r. D. Schubert.
Verlag der Zeitschrift für Naturforschung,
D-72072 Tübingen
0939-5075/94/0300-0220 $03.00/0
not be fulfilled (Boonstra et al., 1993; Tanford and
Reynolds, 1976). We have, therefore, perform ed a
m olar mass determination by a rigorous therm o­
dynamic method, sedimentation equilibrium anal­
ysis in the analytical ultracentrifuge. The m ost crit­
ical param eter in the analysis, the partial specific
volume v of the protein/pigment complex, was de­
termined by a recently described adaptation of a
m ethod of Edelstein and Schachman (1967) to in­
trinsic membrane proteins in detergent solutions
(Schubert et al., 1994). In the procedure, which for
the first time allows a reliable experimental
v-determination of intrinsic membrane proteins,
data obtained in three different detergents were
combined, to yield simultaneously both M and v
of the complex. The system studied was the active
trimeric form of PS I from a phycobilisome-less
m utant of the cyanobacterium Synechococcus
PCC 7002 (Tsiotis et al., 1993).
Trimeric PS I was isolated as described by Tsio­
tis et al. (1993). The final purification step was
HPLC gel filtration on a TSK 4000 SW column
(Toso Haas) equilibrated with 20 mM MES, 10 mM
MgCl2, 10 mM CaCl2, 0.05% lauryl-ß-D-maltoside
(C 12M). Detergent exchange against 0.2% nonaethyleneglycol lauryl ether (C 12E9; Sigma) or 0.1%
of the hydrogenated form of Triton X-100 (TX;
Aldrich) was performed by an additional gel filtra­
tion step using the same column. The peak posi­
tion during elution of the protein was virtually un­
affected by the detergent exchange, which indi-
Unauthenticated
Download Date | 6/17/17 12:45 AM
C. Tziatzios et al. ■The Molar Mass of an Active Photosystem I Complex from the Cyanobacterium
cates that the protein complex was stable in the
detergents used. For the ultracentrifuge experi­
ments, the samples were brought to the required
concentrations of D 20 and D 2180 (Edelstein and
Schachman, 1967; Schubert et al., 1994). In addi­
tion, appropriate buffer stock solutions were add­
ed to restore the detergent and salt concentrations
given above plus 50 mM NaCl. Final protein con­
centration varied between 2.8 and 8.0 [ig/ml.
Sedimentation equilibrium experiments on the
purified PS I were performed in a Beckman O pti­
ma XL-A analytical ultracentrifuge, using
6-channel centerpieces. R otor speed was 8000 to
10000 rpm, and rotor tem perature 6 °C. The con­
centration distributions were recorded at 422 nm.
The samples contained either C 12M, C 12E9 or TX.
The param eter to be determined was the effective
m olar mass of the protein/pigment complex,
M efT= M(1 -vQo), at that solvent density q0 which
equals the density Qd of the corresponding deter­
gent (Schubert and Schuck, 1991; Schubert et al.,
1994; Tanford and Reynolds, 1976). In the case of
C 12M, extrapolation of the M efrvalues measured
at different values of qg to the density gd had to be
perform ed (Schubert et al., 1994). With the latter
two detergents, q0 could be adjusted to Qd, thus
avoiding the need for density variation series
(Schubert and Schuck, 1991; Schubert et al., 1994;
Tanford and Reynolds, 1976).
In m ost o f the experiments, the absorbance-versus-radius distributions obtained could be fitted
by a single exponential, which indicates homogeniety o f particle mass. However, with a few sam­
ples which had been stored frozen, the presence of
an additional com ponent o f much lower m olar
mass was apparent from the fits. This component,
with a typical Chi a-spectrum, obviously repre­
sented a degradation product. D ata obtained with
the latter samples were not considered in the evalu­
ations following.
In the sedimentation equilibrium experiments,
the effective m olar mass o f PS I in solutions of
C 12E9 und TX, at the density o f the respective de­
tergents (q(C 12E9)= 1.059 g m l-1 (Schubert et al.,
1994), q(TX) = 1.085 g m l-1*) was found to be
140,000 ± 2,000 (C]2E9; n = 6) and 122,000 ± 3,000
* T he density o f the hydrog en ated form o f TX , used in
this study, w as found to be distinctly low er th a n th a t
o f the u n su b stitu ted form (g = 1.115 g m l“ ').
221
(TX; n = 5). The data obtained in solutions of
C 12M, at different solvent densities q 0 , are shown
in Fig. 1; M eff extrapolated to the density of the de­
tergent (Qd = 1.229 g m l-1 (Schubert et al., 1994)) is
40,000 ± 5,000. The M efrvalues at the respective
detergent densities are plotted, as a function of qg,
in Fig. 2. The two figures yield approxim ate values
P [g/ml]
------
Fig. 1. PS I in solutions o f C ]2M: D ependency o f M efron
solvent density. ( • ) U n co rrected d ata; (O ) d a ta co rrect­
ed for H -D exchange. T he arro w indicates the b u o y an t
density o f C 12M.
P [g/ml]
Fig. 2. D ependency o f th e effective m o lar m ass o f PS I
on solvent density, u n d er the co n d itio n s o f density
m atching. ( • ) U n co rrected d ata ; (O ) d a ta corrected for
H -D exchange.
Unauthenticated
Download Date | 6/17/17 12:45 AM
222
C. Tziatzios et al. ■The Molar Mass of an Active Photosystem I Complex from the Cyanobacterium
(uncorrected for H-D exchange) for the m olar
mass of the protein/pigm ent and o f the protein/
pigm ent/C12M complex and thus also for the rela­
tive am ount of PS I-bound C 12M. Applying these
data, corrections for H-D exchange in the protein
and in the protein-bound C 12M were calculated,
assuming Akpmax = 0.008 (where (1 + A k pmax) is
the ratio of the protein’s m olar mass in 100% D 20
and H 20 ) (Schubert et al., 1994). The corrected
data are included in the figures. Those in Fig. 2
directly yield the final results for the protein/pig­
ment complex: v = (0.788 ± 0.010) ml g_1 and
Mr = 830,000 ± 60,000. Assuming Akp max = 0.0155
(Edelstein and Schachman, 1967), the corre­
sponding figures would be v = 0.797 ml/g and
M r = 860,000. As discussed earlier (Schubert et al.,
1994), we consider the former figures as the more
probable ones. - It follows that, in solutions o f
C 12M, the exact am ount of detergent bound to the
protein/pigm ent complex is (0.48 ± 0.06) g per g of
complex.
The figures obtained for the m olar mass o f the
protein/pigm ent complex are surprisingly close to
those estimated from electron microscopy and
from HPLC gel filtration, two techniques which
have to rely on quite disputable assumptions.
B oonstra A. F., Visschers R. W ., C alk o en F., van G ro n delle R., van Brüggen E. F. J. an d B oekem a E. J.
(1993), Structu ral ch aracterizatio n o f the B 8 0 0 -8 5 0
and B875 light-harvesting a n te n n a com plexes from
Rhodobacter sphaeroides by electron m icroscopy.
Biochim. Biophys. A cta 1142, 1 8 1 -1 8 8 .
Edelstein S. J. and S chachm an H. K. (1967), The sim ul­
taneous determ in atio n o f p artial specific volum es an d
m olecular w eights w ith m icrogram quan tities. J. Biol.
Chem . 242, 3 0 6 -3 1 1 .
G oldbeck J. and B ryant D. (1991), P ho to sy stem I. C u r­
rent topics in bioenergetics (Lee C. P. ed.) Vol. 16,
8 3 -1 7 7 .
R ögner M ., M ü h len h o ff U ., B oekem a E. J. an d W itt
H. T. (1990a), M ono-, di- an d trim eric PS I reaction
center com plexes isolated from the th erm ophilic
cyanobacterium Synechococcus sp. Size, shape an d
activity. Biochim. Biophys. A cta 1015, 4 1 5 -4 2 4 .
R ögner M ., N ixon P. J. and D iner B. A. (1990b), P u rifi­
cation and ch aracterizatio n o f p h otosystem I an d
photosystem II core com plexes from w ild-type an d
phycocyanin-deficient strains o f the cyanobacterium
Svnechocystis PC C 6803. J. Biol. Chem. 265, 6189 —
6196.
S chubert D. and Schuck P. (1991), A nalytical u ltracen ­
trifu g atio n as a tool for studying m em brane p ro tein s.
Progr. C olloid Polym . Sei. 8 6 , 1 2 -2 2 .
S chubert D ., T ziatzios C., van den B roek J. A., Schuck
P., G erm ero th L. and M ichel H. (1994), D ete rm in a­
tion o f the m o lar mass o f p igm ent-containing c o m ­
plexes o f intrinsic m em brane proteins: Problem s, so­
lutions an d application o f the light-harvesting co m ­
plex B 800/820 o f Rhodospirillum molischianum.
Progr. C olloid Polym . Sei. 94, 1 4 - 19.
T an fo rd C. and R eynolds J. A. (1976), C h aracterizatio n
o f m em brane proteins in d etergent solutions. B io­
chim . Biophys. A cta 457, 13 3 -1 7 0 .
T siotis G ., N itschke W ., H aase W . and M ichel H.
(1993), P urification and crystallization o f p h o to sy s­
tem I com plex from a phycobilisom e-less m u ta n t o f
the cyanobacterium Synechococcus PC C 7002. P h o to synth. Res. 35, 2 8 5 -2 9 7 .
Ackno wledgemen ts
We would like to thank Prof. Dr. H. Michel for
helpful discussions and Mrs. J. A. van den Broek
for skilful technical assistance. This work was sup­
ported by the Deutsche Forschungsgemeinschaft
(SFB 169).
Unauthenticated
Download Date | 6/17/17 12:45 AM