(Received for publication, August 28, 1996, and in revised form, November 13, 1996)
From the The extrinsic 33-kDa protein of
photosystem II (PSII) was modified with various reagents, and the
resulting proteins were checked for the ability to rebind to PSII and
to reactivate oxygen evolution. While modification of more than eight
carboxyl groups of aspartyl and glutamyl residues with glycine methyl
ester did not affect the rebinding and reactivating capabilities,
modification of amino groups of lysyl residues with either
N-succinimidyl propionate or 2,4,6-trinitrobenzene sulfonic
acid or modification of guanidino groups of arginyl residues with
2,3-butanedione resulted in a loss of rebinding and reactivating
capabilities of the 33-kDa protein. Moreover, the number of lysyl and
arginyl residues susceptible to modification was significantly
decreased when the protein was bound to PSII as compared with when it
was free in solution, whereas the number of carboxyl groups modified
was little affected. These results suggested that positive charges are
important for the electrostatic interaction between the extrinsic
33-kDa protein and PSII intrinsic proteins, whereas negative charges on
the protein do not contribute to such interaction. By a combination of
protease digestion and mass spectroscopic analysis, the domains of
lysyl residues accessible to N-succinimidyl propionate or
2,4,6-trinitrobenzene sulfonic acid modification only when the 33-kDa
protein is free in solution were determined to be Lys4,
Lys20, Lys66-Lys76,
Lys101, Lys105, Lys130,
Lys159, Lys186, and
Lys230-Lys236. These domains include those
previously reported accessible to N-hydroxysuccinimidobiotin only in solution (Frankel and
Bricker (1995) Biochemistry 34, 7492-7497), and may be
important for the interaction of the 33-kDa protein with PSII intrinsic
proteins.
The extrinsic 33-kDa protein of oxygen-evolving
PSII1 complex is associated with the
lumenal surface of thylakoid membranes and plays an important role in
maintaining functional binding of the manganese cluster that directly
catalyzes the H2O-splitting reaction (for reviews, see
Refs. 1-3). This protein can be removed from PSII membranes by washing
with high concentrations of divalent cations (4) or urea
plus NaCl (5), resulting in a substantial loss of the
oxygen-evolving activity, which can be restored by rebinding of the
protein (5-7). Removal of the protein also leads to a gradual
liberation of two of the four manganese per PSII (8). Cross-linking of
the protein with PSII intrinsic components by a water-soluble
carbodiimide, which directly links amino and carboxyl groups in
contact, prevents release of the protein upon CaCl2- or
urea/NaCl wash or heat treatment, thereby stabilizing binding of the
manganese cluster against these treatments (9-11). Based on these
studies, the association of the 33-kDa protein with PSII has been
proposed to involve electrostatic interactions between positive
and negative charges on both the extrinsic protein and some PSII
intrinsic subunits as well as hydrogen bonding (9, 12-14).
The 33-kDa protein from spinach is composed of 247 amino acid residues
including 14 Asp, 21 Glu, 23 Lys, and 6 Arg (15), with an isoelectric
point (pI) of 5.2 (16). Two Cys residues exist in the protein, which
form a disulfide bond that has been shown to be important for
maintaining the functional structure of the protein (17-19). Many
attempts have been made to identify possible PSII intrinsic components
with which the 33-kDa protein is associated. Trypsin digestion
experiments have suggested that CP43, an intrinsic chlorophyll-binding
43-kDa protein, is shielded by the 33-kDa protein against proteolytic
attack (20). A close association of the 33-kDa protein with CP47,
another intrinsic chlorophyll-binding protein, has been demonstrated by
cross-linking experiments with various bifunctional reagents (21-26).
Photoaffinity cross-linking studies suggested that the protein is
associated with D1 and D2, the two PSII reaction center proteins (27). The 33-kDa protein was also found to cross-link with the There is, however, only very limited information concerning the
residues on the 33-kDa protein responsible for its binding to PSII.
Proteolytic cleavage of the first 16-18 N-terminal residues of the
mature protein prevented its binding to PSII, suggesting the importance
of the N-terminal region for binding (29). Site-directed mutagenesis of
the protein in Synechocystis sp. PCC 6803 showed that
alteration of Asp159 to Asn decreased the binding affinity
of the protein to PSII (19), suggesting the possible involvement of
this residue in either the interaction with PSII or in maintaining a
functional structure of the protein required for binding. This is
consistent with the results observed for the 33-kDa protein from
spinach, where alteration of the corresponding Asp157
reduced the oxygen-evolving activity (3, 30). In addition, in the
spinach 33-kDa protein, alterations of Glu104 and
Asp109 were also shown to slightly decrease the oxygen
evolution, while changes of any one of the other conserved carboxylic
residues did not affect the activity significantly (3, 30, 31).
Chemical modification of charged amino acid residues can easily alter
surface charges on proteins and therefore is a useful technique to
identify residues possibly involved in the electrostatic interaction
between the 33-kDa protein and intrinsic PSII proteins. In an early
study, Frankel and Bricker (32) modified the lysyl residues of the
33-kDa protein with NHS-biotin either in solution or on PSII membranes
and showed that lysyl residues in domains Glu1-Lys4, Lys20,
Lys101-Lys105, and
Lys159-Lys186 were accessible to NHS-biotin
only when the 33-kDa protein was released from the PSII membranes. They
concluded, therefore, that these domains may be responsible for the
interaction of the 33-kDa protein with PSII intrinsic proteins. In the
present study, either negative charges of aspartyl and glutamyl
residues (and free C terminus) or positive charges of lysyl (and free N
terminus) and arginyl residues of the isolated extrinsic 33-kDa protein
were modified using chemical modification techniques, and the abilities of the resultant modified protein to rebind to PSII and to reactivate oxygen evolution were examined. Data are presented showing that positive charges on the 33-kDa protein are essential for its
electrostatic interaction with intrinsic PSII proteins, but elimination
of surface negative charges on the protein had essentially no effect on
its functional binding. Furthermore, the lysyl residues that are
modified either when the 33-kDa protein is free in solution or when it is associated with PSII membranes, were determined by a combination of
protease digestion and mass spectroscopic analysis. From these results,
the domains containing lysyl residues that are buried when the 33-kDa
protein is associated with PSII but become exposed to bulk solution
upon release of the protein from PSII membranes were determined. These
domains may be located in regions that interact with PSII intrinsic
components.
Oxygen-evolving PSII
membranes (BBY PSII membranes) were prepared from spinach chloroplasts
with Triton X-100 as in Ref. 33, with slight modifications as described
in Ref. 9. The isolated PSII membranes were suspended in medium A
containing 40 mM Mes (pH 6.5), 0.4 M sucrose,
10 mM NaCl, and 5 mM MgCl2 and
stored in liquid nitrogen until use. The extrinsic 33-kDa protein was extracted from the PSII membranes by 1 M CaCl2
treatment (4), incubated with 1 M CaCl2 for
3 h in the dark to suppress the activity of co-purified protease,
dialyzed against 5 mM Mes-NaOH (pH 6.5), and then purified
by column chromatography with a DEAE-Sepharose CL-6B column (Pharmacia
Biotech Inc.) (34). The concentration of the 33-kDa protein was
determined using an extinction coefficient of 16 mM Chemical modification of carboxyl groups on the purified 33-kDa protein
was carried out in 100 mM GME (pH 6.2) containing 37.75 µM of the 33-kDa protein and 2 mM EDC at
25 °C for 12 h in the dark. The reaction mixture was
concentrated and washed with 1 M NaCl and 20 mM
phosphate buffer (pH 6.5) by ultrafiltration to remove unreacted and
electrostatically attached reagents and then passed through a Sephadex
G-25 column equilibrated with 20 mM phosphate buffer (pH
6.5). For modification of the 33-kDa protein bound to PSII, the PSII
membranes were washed with 1 M NaCl to remove the 23- and
17-kDa proteins (35, 36) and then washed twice with and resuspended in
0.4 M sucrose, 100 mM GME (pH 6.2) to 1 mg of
Chl/ml. This solution was added by 4 mM EDC and incubated for 15 h at 25 °C. The reaction mixture was washed twice with the same GME buffer and then treated with 1 M
CaCl2, 25 mM Mes-NaOH (pH 6.5) to release the
modified 33-kDa protein. The released 33-kDa protein was incubated with
1 M CaCl2, dialyzed, and then concentrated by
ultrafiltration.
For modification of amino groups of lysyl residues and the free amino
terminus of the 33-kDa protein, the purified protein (48 µM) was incubated in a reaction mixture containing 20 mM phosphate buffer (pH 6.5) and 0.5-8 mM
N-succinimidyl propionate (NSP) at 25 °C for 90 min in
the dark. The reaction was stopped by passing the reaction mixture
through a Sephadex G-25 column equilibrated in 20 mM
phosphate buffer (pH 6.5), which removed unreacted NSP from the
reaction mixture. For modification of the 33-kDa protein associated
with PSII, the 23- and 17-kDa protein-depleted PSII membranes were
incubated in medium A containing 8 mM NSP at 1 mg of Chl/ml
for 90 min at 25 °C. After modification, the PSII membranes were
washed twice with medium A to remove the remaining NSP and treated with
1 M CaCl2 to release the modified 33-kDa protein. The released 33-kDa protein was concentrated as described above.
The lysyl residues and free N terminus of the 33-kDa protein were also
modified with 3 mM or 20 mM TNBS in medium A
for 15 h at 25 °C at a protein concentration of 4.4 µM. After modification, the reaction mixture was passed
through a Sephadex G-25 column, dialyzed against 20 mM
phosphate buffer (pH 6.5), and then concentrated with the
DEAE-Sepharose CL-6B column. For modification of the 33-kDa protein on
PSII membranes, the 23- and 17-kDa protein-depleted PSII membranes were
incubated with 4 mM TNBS at 1 mg of Chl/ml for 15 h at
25 °C in medium A. The reaction was stopped by centrifugation, and
the modified 33-kDa protein was released from the PSII membranes, dialyzed, and concentrated.
For modification of arginyl residues, the 33-kDa protein (33 µM) was incubated with 0.9-2.5 mM BD in 0.1 M KCl, 0.1 M H3BO3/NaOH (pH 8.4) at 25 °C for 70 min, and the reaction was stopped by passing the reaction mixture through the Sephadex G-25 column equilibrated in 20 mM phosphate (pH 6.5), 50 mM
H3BO3. For modification of the 33-kDa protein
bound to PSII, the 23- and 17-kDa protein-depleted PSII membranes were
washed twice with and resuspended in 0.4 M sucrose, 50 mM H3BO3, 20 mM NaCl
and 50 mM Mes-NaOH (pH 7.5) and then incubated with 10 mM BD at 1 mg of Chl/ml for 2 h at 25 °C (The pH of
the reaction mixture must be lowered to 7.5, since modification at pH
8.4 led to a release of the 33-kDa protein from PSII). After
incubation, the reaction mixture was washed twice with the same buffer
containing boric acid and then treated with 1 M
CaCl2, 50 mM H3BO3, 25 mM Mes-NaOH (pH 6.5) to release the 33-kDa protein. The
released 33-kDa protein was incubated with 1 M
CaCl2, dialyzed against 5 mM Mes-NaOH (pH 6.5),
50 mM H3BO3, and concentrated.
NSP was purchased from Wako Pure Chemicals (Tokyo, Japan); GME, EDC,
and TNBS were purchased from Nacalai Tesque Chemicals (Tokyo,
Japan).
For reconstitution,
native PSII membranes were washed with 2.6 M urea, 0.2 M NaCl to remove the three extrinsic proteins of 33, 23, and 17 kDa (5). The resultant PSII membranes were incubated with either
the native or the modified 33-kDa proteins at a protein:Chl ratio of
0.6 (w/w), in medium A at 0 °C for 30 min in the dark at a Chl
concentration of 0.5 mg/ml. The reconstituted PSII membranes were
collected by centrifugation at 35,000 × g for 10 min
and then washed once with and resuspended in medium A. SDS-polyacrylamide gel electrophoresis (SDS-PAGE) was carried out
according to Laemmli (37) with a slab gel of 11.5% acrylamide
containing 6 M urea. Samples were solubilized with 5%
lithium lauryl sulfate and 75 mM dithiothreitol. For
immunoblotting, proteins on the gel were electrotransferred onto a
nitrocellulose membrane, reacted with an antibody raised against the
extrinsic 33-kDa protein from spinach, and visualized with biotinylated
anti-rabbit IgG. Isoelectric focusing was carried out in a 5.5%
polyacrylamide-containing homogenous gel covering a pH range of 3.5-10
or 4.0-6.0. Proteins were stained with 0.048% CBB in 30% methanol
and 10% acetic acid. Oxygen evolution was measured at 25 °C with a
Clark-type oxygen electrode in medium A, to which 5 mM
CaCl2 and 0.4 mM
phenyl-p-benzoquinone were added. Chl concentration was
determined by the method of Porra et al. (38).
The 33-kDa protein (20-30 nmol)
modified with NSP either in solution or on PSII membranes was denatured
in 1 M Tris-HCl (pH 8.4), 8 M urea, and 1 mM mercaptoethanol at 37 °C for 4 h and then
treated with 2 mM iodoacetamide at 37 °C for 1 h to
block SH groups. The reaction mixture was added to a final
concentration of 10% of cold trichloroacetic acid and centrifuged, and
the resulting precipitate was washed twice with acetone. The final
precipitate was dried and then resolubilized in 0.1 ml of 0.1 M pyridine/acetic acid (pH 6.5). The denatured 33-kDa
protein was digested first with 15 µg of Staphylococcus V8
protease at 37 °C for 6 h and then with another 15-µg
protease at 37 °C overnight. The reaction was stopped by the
addition of 10 µl of 100% acetic acid and dried.
The TNBS-modified 33-kDa protein was similarly denatured with 8 M urea, 1 M Tris-HCl (pH 8.4), 1 mM
mercaptoethanol, precipitated, and then digested with the V8 protease.
The digested protein mixture was monitored by a reversed phase HPLC
every 8 h, and the digestion reaction was carried out for 2 days,
after which the elution pattern was not further changed.
The
protease-digested protein mixture of NSP-modified 33-kDa protein was
solubilized in water and directly applied to a MALDI-TOF mass
spectrometer (Reflex; Bruker), with a matrix of either
2,5-dihydroxybenzoic acid or The digestion mixture of TNBS-modified 33-kDa protein was separated
with a reversed phase column (Bondasphere 5 µ C4 300A, Waters Inc.)
in an HPLC set-up (LC-9A, Shimadzu Inc., Japan). The column was eluted
with a gradient of 0-75% acetonitrile in 0.1% trifluoroacetic acid
at a flow rate of 1 ml/min, and the elution pattern was monitored at
210 nm for all peptides and 345 nm for peptides modified by TNBS. Each
fraction was collected, dried, and resolubilized in 3 µl of 67%
acetic acid, out of which 1 µl was mixed with the same volume of the
matrix (a mixture of 1:1 (v/v) of glycerol and 3-nitrobenzyl alcohol)
and analyzed with a fast atom bombardment mass spectrometer (JEOL JMS
HX-110) at a voltage of 10 kV with xenon as the ionization gas. The
resulting mass spectra were analyzed with a DA5000 data system and
assigned to the known 33-kDa protein sequence.
The TNBS-modified peptides obtained from HPLC separation were also
analyzed for their N-terminal sequence by Edman degradation of the
peptides followed by sequence analysis with a protein sequencer (Applied Biosystems model 477A).
Amino acid composition of the native and BD-modified 33-kDa protein was
analyzed by an SII SAC4800 amino acid analyzer (Seiko Instrument Inc.,
Japan) following acidic hydrolysis of the proteins in the gaseous
phase. The relative abundance of each amino acid in the 33-kDa protein
was calculated with Leu as the standard.
Modification of Negative Charges on the 33-kDa Protein
The extrinsic 33-kDa protein has a pI of 5.2 (16) and is therefore
negatively charged at neutral pH. Thus, we first attempted to modify
the negatively charged carboxyl groups of aspartyl and glutamyl
residues (and free C terminus) on the 33-kDa protein using GME in the
presence of a water-soluble carbodiimide (EDC) (see Reaction R1)
(39).
In order to determine whether elimination of surface negative charges
affected binding of the 33-kDa protein to PSII, the ability of the
GME-modified protein to rebind to urea/NaCl-washed PSII membranes was
examined. Since the 33-kDa protein migrated much faster upon GME
modification, resulting in a comigration of the modified protein with
light-harvesting Chl apoproteins (data not shown), we used immunoblot
analysis to detect the protein. As Fig. 1B shows, the
GME-modified protein (lane 2) was able to rebind to the
protein-depleted PSII membranes as effectively as the unmodified
protein (lane 1), suggesting that elimination of the surface
negative charges has no significant effect on binding of the protein to
PSII. Table I shows that, upon rebinding of the
GME-modified 33-kDa protein to urea/NaCl-treated PSII membranes, the
oxygen evolution was restored to an extent comparable with that
achieved by rebinding of the unmodified protein, indicating that the
modified 33-kDa protein was fully functional upon rebinding. These
results clearly indicate that surface negative charges on the 33-kDa
protein do not participate in its functional binding to intrinsic PSII
proteins or are not important for maintaining the functional structure
of the protein.
Reactivation of oxygen evolution by reconstitution of the 33-kDa
protein modified with various reagents to urea/NaCl-washed PSII
membranes
Department of Biology, Faculty of Science,
Science University of Tokyo, Kagurazaka 1-3, Shinjuku-ku, Tokyo 162, Japan, the § Department of Applied Biological Science,
Faculty of Science and Technology, Science University of Tokyo,
Yamazaki 2641, Noda, Chiba 278, Japan, the ¶ Solar Energy Research
Group,
PRESTO,
Research Institute for Biosciences,
Science University of Tokyo, Yamazaki, Noda, Chiba 278, Japan
subunit of
cytochrome b559 and the psbI gene
product (28). These results suggested that the extrinsic 33-kDa protein
is associated with, or is in close proximity to, virtually all of the
intrinsic PSII components.
Preparation and Chemical Modifications
1 cm
1 at 276 nm (29).
-cyano-4-hydroxycinnamic acid. The mass
of each measured peptide fragment was assigned to the known 33-kDa
protein sequence.
In this reaction, a negatively charged carboxyl group is replaced
by an uncharged methyl ester group. As a result, the pI of the protein
is anticipated to shift toward higher values. Fig. 1A shows that the modification indeed
upshifted the pI value from 5.2 of unmodified protein (lane
1) to a value above 8.5 (lane 3). This change was
estimated to result from modification of more than eight negatively
charged carboxyl groups to uncharged groups, as calculated using a
computer pI/Mr tool (40). It should be noted
here that the band of the modified protein appeared much broader than
the native protein upon isoelectric focusing, implying that the
resulting protein products may be composed of proteins with different
numbers of carboxylic residues modified.
Fig. 1.
GME-modified 33-kDa protein and its
reconstitution with PSII. A, isoelectric focusing of the
native 33-kDa protein (lane 1), the 33-kDa protein modified
by 100 mM GME on PSII membranes (lane 2), and
the 33-kDa protein modified by 100 mM GME in solution (lane 3). The gel was stained with CBB. B,
reconstitution of the native or GME-modified 33-kDa protein with
urea/NaCl-washed PSII membranes. The native and reconstituted PSII
membranes were analyzed by SDS-PAGE, and the 33-kDa protein was
detected with an antibody raised against this protein from spinach.
Lane B, control BBY PSII-membranes; lane U,
urea/NaCl-washed PSII; lane 1, urea/NaCl-washed PSII
reconstituted with the native 33-kDa protein; lane 2,
urea/NaCl-washed PSII reconstituted with GME-modified 33-kDa protein.
For other details, see "Materials and Methods."
[View Larger Version of this Image (36K GIF file)]
Samples
pI
Oxygen evolution
Reactivation
µmol
O2/mg Chl/h
%
Control PSII
membranes
565
Urea/NaCl-washed PSII
71
PSII
reconstituted with
Unmodified 33-kDa protein
5.2
371
100
GME-modified 33-kDa protein
>8.5
373
101
0.5 mM NSP-modified 33-kDa
protein
4.9-5.1
347
92
2.0
mM NSP-modified 33-kDa protein
4.5-4.9
167
32
4.0 mM NSP-modified 33-kDa
protein
4.4-4.8
95
8
8.0 mM NSP-modified
33-kDa protein
4.4-4.7
72
0.2
3 mM
TNBS-modified 33-kDa protein
4.5-4.8
71
0
20
mM TNBS-modified 33-kDa protein
4.3-4.7
71
0
0.9 mM BD-modified 33-kDa
protein
185
38
1.5 mM BD-modified 33-kDa
protein
82
4
2.5 mM BD-modified 33-kDa
protein
61
0
Fig. 1A also shows that GME modification of the 33-kDa protein associated with PSII membranes resulted in a similar shift in the pI value (lane 2) as seen upon modification of the protein free in solution (lane 3). This indicates that a similar number of the carboxyl groups are modified even when the 33-kDa protein is bound to PSII and suggests that the carboxyl groups susceptible to GME modification when the 33-kDa protein is free in solution are also accessible to GME when the protein is bound to PSII membranes. Thus, these carboxylic amino acid residues are not located in regions interacting with PSII intrinsic components. This agrees with the above finding that modification of these residues did not affect the binding and reactivation of the 33-kDa protein.
Modification of Positive Charges on the 33-kDa Protein: Lysyl Residues
Modification with NSPThe -amino groups of lysyl residues
and
-amino group of free N terminus carry positive charges that can
be selectively modified with NSP according to Reaction R2
(41). Upon this modification, positively charged amino groups of lysyl
residues and the free N terminus are replaced by uncharged groups,
which results in a downshift of the pI value of the protein. Fig.
2A shows that the pI of NSP-modified 33-kDa
protein indeed shifted toward acidic pH with increasing NSP
concentration, although here again the modified proteins showed broader
bands due to mixing of proteins with slightly different pI values
resulting from inhomogeneous modification. In Fig. 2B, we
examined the rebinding capability of the NSP-modified proteins to
urea/NaCl-washed PSII membranes. The protein was able to rebind to PSII
after modification with NSP at a concentration of 0.5 mM,
which shifted the pI from 5.2 to 4.9-5.1. Increasing the NSP
concentration to above 2 mM, however, significantly
decreased the rebinding capability of the modified protein. This
decrease was paralleled by a decrease in the extent of restoration of
oxygen evolution upon reconstitution of the modified protein. As Table
I shows, a modified protein with pI of 4.4-4.7, which virtually cannot
rebind to PSII, completely lost the reactivating capability. Since
NSP-modified protein showed a weaker CBB-staining intensity and slower
migration in the SDS-PAGE gel than the unmodified protein, we
determined the amount of the modified proteins rebound by releasing
them with urea/NaCl wash and comparing the staining intensities of the
released proteins on the gels with the intensities of the known amounts
of equally modified proteins. The amount of the rebound protein
determined in this way showed a good correlation with oxygen evolution
restored (Fig. 3), irrespective of the degree of the NSP
modification. This indicates that loss of the reactivating capability
of the NSP-modified proteins was caused directly by loss of their
rebinding, which in turn suggests that the modified proteins, when
rebound, are fully functional and that there is apparently no
nonspecific binding of the modified protein.
Fig. R2.
In order to determine the exact pI value at which the protein loses its abilities to rebind to PSII and to reactivate oxygen evolution, the NSP-modified proteins with various pI values were separated by isoelectric focusing, and each band with different pI in the gel was cut out, extracted in 20 mM phosphate buffer (pH 6.5), and then reconstituted to urea/NaCl-washed PSII membranes. The results showed that the protein completely lost its rebinding and reactivating abilities at pI 4.7 (data not shown). A pI shift from 5.2 to 4.7 was calculated to correspond to elimination of about seven positive charges on the 33-kDa protein. As will be shown later, the number of domains of lysyl residues accessible to NSP modification in the 33-kDa protein bound to PSII is six less than in the protein free in solution. These results suggested that the lysyl residues modified by NSP when the 33-kDa protein is free in solution are largely located in regions interacting with PSII intrinsic components, and thus may be directly involved in binding to PSII.
Modification with TNBSSince chemical modification of a specific residue depends not only on the position of that residue in the protein but also in some cases on the chemical reagents used, we used another reagent, TNBS, to modify the amino groups of the lysyl residues and free N terminus of the 33-kDa protein. TNBS modifies the amino groups of lysyl residues and free N terminus according to Reaction R3 (42). Like NSP, TNBS also removed positive charges of the amino groups and thus lowered the pI value of the 33-kDa protein (Table I). Fig. 4 shows reconstitution of the TNBS-modified 33-kDa protein with urea/NaCl-washed PSII membranes. Upon modification of the 33-kDa protein with 3 mM TNBS, the 33-kDa protein completely lost its ability to rebind to the protein-depleted PSII membrane. This leads to a complete loss of the reactivation of oxygen evolution by the modified protein, as shown in Table I. Modification of the 33-kDa protein with TNBS at 3 mM resulted in a decrease of pI from 5.2 to 4.5-4.8, which corresponds to elimination of seven or eight positive charges of the amino groups on average. This is in agreement with the results of NSP modification and also roughly agrees with the number of domains that become inaccessible to TNBS when the 33-kDa protein is bound to PSII (see below). These results strongly suggest that lysyl residues (and/or the N terminus) of the 33-kDa protein are important for the electrostatic interaction between the protein and PSII intrinsic components.
[View Larger Version of this Image (6K GIF file)]Fig. R3.
Modification of Positive Charges on the 33-kDa Protein: Arginyl Residues
In addition to the amino group of lysyl residues, the guanidino group of arginyl residues also carries a positive charge. The positive charges of guanidino groups of the arginyl residues on the 33-kDa protein were converted to negative charges by modification of the 33-kDa protein with BD in the presence of boric acid, according to Reaction R4 (43). Fig. 5 shows rebinding of the BD-modified 33-kDa protein to urea/NaCl-washed PSII. While the 33-kDa protein was still able to rebind to PSII upon modification with BD at lower concentrations, its rebinding capability decreased upon modification with increasing BD concentration. When the 33-kDa protein was modified at BD concentrations higher than 1.5 mM, it was no longer able to rebind to PSII. Table I shows that the reactivation of oxygen evolution by the modified 33-kDa protein parallels its rebinding capability, implying that modification of positive charges of arginyl residues primarily inhibited the binding of the 33-kDa protein, which then resulted in the loss of reactivation of oxygen evolution.
[View Larger Version of this Image (7K GIF file)]Fig. R4.
Because the BD-modified arginyl residues are unstable in the absence of boric acid (43), it was impossible to estimate the number of modified arginyl residues by isoelectrophoretic analysis. Therefore, we determined the number of modified arginyl residues by analyzing the amino acid composition of the BD-modified protein. In the native 33-kDa protein, there are six arginyl residues, of which three were modified at 0.9 mM BD, whereas four were modified at 1.5 mM BD (Table II). In addition, we modified the 33-kDa protein bound to PSII membranes with BD and determined the arginyl residues that were modified. Table II shows that no arginyl residues were modified by BD when the 33-kDa protein was bound to PSII. These results suggest that the arginyl residues modified by BD in solution become shielded from contact with bulk solution upon binding of the protein to PSII, which in turn implies that these residues are located in regions of the protein in contact with PSII intrinsic proteins.
|
Identification of Lysyl Residues Possibly Involved in Binding of the 33-kDa Protein to PSII
The above results have shown that, while loss of negative charges did not affect binding and reactivating capabilities of the 33-kDa protein, loss of positive charges of either lysyl or arginyl residues led to a significant decrease in the rebinding capability and thereby reduced the reactivating capability of the 33-kDa protein. In order to identify the lysyl and arginyl residues possibly responsible for loss of the binding to PSII membranes of the 33-kDa protein, we modified the protein either in its free form (in solution) or in its bound form (on PSII membranes) and attempted to determine the location of lysyl and arginyl residues that are modified, by digesting the modified protein with Staphylococcus V8 protease followed by determination of the mass of the resultant peptide fragments with mass spectrometer. From the results obtained, the domains on the 33-kDa protein that contain lysyl residues modified only in solution but not on PSII membranes were determined; these domains may be located in regions interacting with PSII intrinsic components. Since the BD molecule is easily dissociated from arginyl residues due to the instability of the BD-modified arginyl residues in the absence of boric acid, we were unable to detect arginyl residues on the 33-kDa protein that were modified by BD in solution. (Note that there were no arginyl residues susceptible to BD modification when the 33-kDa protein was bound to PSII.) In the following experiments, only results for modification of amino groups of lysyl residues and free N terminus were presented.
Modification with NSPThe 33-kDa protein modified with 8 mM NSP either in its free form (in solution) or when it was
bound to NaCl-washed PSII membranes, was digested with
Staphylococcus V8 protease, and the resulting peptide
mixture was analyzed directly by a MALDI-TOF mass spectrometer. Since
whether a peptide fragment can be detected by the MALDI-TOF mass
spectrometer depends in some cases on the matrix employed, two
different matrices were used; one was 2,5-dihydroxybenzoic acid, and
the other was -cyano-4-hydroxycinnamic acid. This led to a more
complete identification of the peptide fragments resulting from the V8
protease digestion of the 33-kDa protein. Peptide fragments yielded
were assigned to the known amino acid sequences within a 0.2% mass
error. Modification of the amino group with each NSP molecule results
in an addition of an N-propionyl group, which corresponds to
an increase of 56.0 Da in the molecular weight. Table
III shows the results of peptide assignment obtained
from V8 protease digestion of the 33-kDa protein that was modified with
8 mM NSP in solution. In total, there were 41 peptides
identified ranging in mass from 517.9 to 4581.4 Da. Of these peptides,
eight were modified with one molecule of NSP
(Glu1-Asp9,
Pro54-Glu62,
Ser140-Asp168,
Ser140-Glu181,
Ser140-Glu182,
Ser140-Glu183,
Leu184-Glu187, and
Asn188-Glu209), whereas five peptides were
modified either with one or two molecules of NSP
(Gly33-Glu62,
Gly63-Glu87,
Lys105-Glu139,
Arg122-Glu139, and
Leu227-Glu238). These results indicate that
residues Lys66, Lys76, Lys130,
Lys137, Lys159, and Lys186 are
accessible to NSP when the 33-kDa protein is modified in solution.
Additionally, one amino group of the lysyl residues (or N terminus) in
domains Glu1-Lys4,
Lys56-Lys60, and
Lys190-Lys207 and two amino groups of the
lysyl residues in domains Lys41-Lys60,
Lys105-Lys137, and
Lys230-Lys236 are accessible to NSP
modification when the 33-kDa protein is free in solution.
|
Table IV presents results obtained for digestion of the 33-kDa protein that was modified with 8 mM NSP on the 23- and 17-kDa protein-depleted PSII membranes. In total, 32 peptides were identified, which ranged in mass from 518.1 to 4524.8 Da. Of these peptides, six were modified with one molecule of NSP (Gly33-Glu53, Pro54-Glu62, Gly63-Glu87, Arg122-Glu139, Asn188-Glu209, and Leu227-Glu238), one peptide was modified either with one or two molecules of NSP (Gly33-Glu62), and one peptide was modified with two molecules of NSP (Ser37-Glu62). These results indicate that one amino group of the lysyl residues in domains Lys41-Lys49, Lys56-Lys60, Lys66-Lys76, Lys130-Lys137, Lys190-Lys207, Lys230-Lys236, and two amino groups of the lysyl residues in domain Lys41-Lys60, were accessible to NSP when the 33-kDa protein was modified on PSII membranes. Among these domains, Lys137 was determined to be modified by NSP in domain Lys130-Lys137, by a postsource decay method of the MALDI-TOF mass analysis. Apparently, the numbers of domains modified when the 33-kDa protein is bound to PSII became much less in comparison with those when the protein is modified in solution; residues Lys130, Lys159, and Lys186 and one of the lysyl residues (and/or N terminus) in domains Glu1-Lys4, Lys66-Lys76, and Lys230-Lys236, were modified by NSP only when the 33-kDa protein was free in solution but not on PSII membranes. These results indicate that these domains become exposed to bulk solution when the 33-kDa protein is released from PSII membranes and therefore suggest that these domains are in regions interacting with PSII intrinsic proteins.
|
The lysyl residues (and free N
terminus) of the 33-kDa protein that are modified by TNBS in solution
and on PSII membranes were also determined by the combination of
protease digestion and mass spectroscopy. Since TNP-carrying lysyl
residues have an absorption peak at 345 nm (42), the TNBS-modified
peptides can be easily detected following protease digestion and
separation by HPLC. Thus, we digested the 33-kDa protein that had been
modified with TNBS either in solution or on PSII membranes with
Staphylococcus V8 protease, separated the peptide fragments
with a reversed phase HPLC, and determined either the mass or the
N-terminal sequence of the resulting peptide fragments. Fig.
6 shows the separation pattern by HPLC of the V8
protease-digested 33-kDa protein modified with TNBS either in solution
or on PSII membranes. Since the elution pattern in Fig. 6 was recorded
at 345 nm, only peptide fractions carrying TNP were detected. Nineteen
such fractions were obtained from the digested mixture of the 33-kDa
protein modified with TNBS in solution, whereas only one fraction was
obtained from the digested mixture of the 33-kDa protein modified with
TNBS on PSII membranes, implying that most of the lysyl residues
(and/or N terminus) that are accessible to TNBS in solution became
buried and not accessible to TNBS upon binding of the 33-kDa protein to
PSII membranes. The 19 fractions obtained, however, do not necessary
represent 19 peptides carrying different TNBS-modified lysyl residues
(see below).
Table V shows the results of mass spectroscopy combined
with N-terminal sequence analysis of the peptide fractions separated by
HPLC. The peptide in the only fraction obtained from V8 protease digestion of the 33-kDa protein modified with TNBS on PSII membranes had a measured mass of 1379.4 Da, which is consistent with the predicted mass of Glu1-Glu10 plus
one molecule of TNP (with an add-on mass of 211.0 Da). In the peptide
Glu1-Glu10, there are two amino groups, one is
in the N terminus and the other is in the Lys4. However,
N-terminal sequence analysis of fraction 1 revealed that the N terminus
was blocked. Thus, the N-terminal -amino group was concluded to be
modified by TNBS when the 33-kDa protein is bound to PSII membranes as
well as in solution.
|
Within the other 18 fractions obtained from V8 protease digestion of the 33-kDa protein modified with TNBS in solution, the mass of the peptides in fractions 9, 13, 17, 18, and 19 could not be determined by the fast atom bombardment mass spectrometer, presumably because these peptides had large masses and thus escaped from the detection of the mass spectrometer. We also tried to sequence the N termini of these peptides; only the N terminus of the peptide in fraction 13 could be sequenced, but the other fractions did not yield any significant phenylthiohydantoin signals. The sequence obtained from fraction 13 is VIGVFQSLQPSDTDLGAKVPXDV. This sequence corresponds to the sequence starting at Val213 in the 33-kDa protein. There are three lysyl residues in the peptide fragment from Val213 to the C terminus of the 33-kDa protein (Lys230, Lys233, and Lys236), among which Lys230 was detected by the N-terminal sequencing. Thus, the lysyl residues that were modified by TNBS in fraction 13 are most probably Lys233 and/or Lys236, although the possibility cannot be excluded that Lys230 was very weakly modified so that it could still be detected by the N-terminal sequencing. The mass of the peptides in the remaining 15 fractions were determined, among which fractions 3, 4, and 15 had masses that could not be assigned to a known sequence, probably due to nonspecific digestions of the V8 protease, which is often observed in such a prolonged digestion reaction as was conducted in the present study. From the measured masses of the remaining fractions, the residues or domains containing lysyl residues that were determined to be modified by TNBS in solution were Glu1, Lys4, Lys14, Lys20, Lys66-Lys76, Lys101, Lys105, Lys159, and Lys230-Lys236. Compared with the results of TNBS modification of the 33-kDa protein on PSII membranes, eight domains (residues) were identified to be modified with TNBS only in solution: Lys4, Lys14, Lys20, Lys66-Lys76, Lys101, Lys105, Lys159, and Lys230-Lys236. These domains, combined with the domains accessible to NSP only in solution, may be important for the interaction of the 33-kDa protein with PSII.
The present results demonstrated that, while elimination of a significant number of negative charges by modification of the carboxyl groups with GME did not affect the rebinding and reactivating capabilities of the 33-kDa protein, removal of positive charges by modification of the lysyl residues with NSP or TNBS or modification of the arginyl residues with BD significantly affected the rebinding and thus the reactivating capability of the 33-kDa protein. This is largely in agreement with the finding from site-directed mutagenesis that alteration of a single residue of most of the conserved Asp and Glu residues did not give rise to a significant loss of function of the 33-kDa protein, except for the residues Glu104, Asp109, and Asp157, alteration of which caused a slight decrease in the oxygen evolution (3, 19, 30, 31). These residues may be in an environment that is not accessible to GME even when the 33-kDa protein is free in solution. This in turn suggests that these residues are buried in the protein and thus are important for maintaining a functional structure of the protein rather than being involved directly in the interaction of the protein with PSII. However, since the effects of changing Glu104, Asp109, and Asp157 are only very marginal (30, 31), more work is needed before a decisive conclusion can be drawn.
A loss of the rebinding of the 33-kDa protein to PSII following chemical modification can, in principle, be caused by two different mechanisms. First, chemical modification may induce a conformational change of the protein, resulting in a protein structure that is no longer able to bind to PSII. Second, the residues that are modified may directly participate in the interaction of the protein with PSII; the modification of these residues may therefore inhibit binding of the protein to PSII. In the present study, however, we showed that the number of the carboxylic residues modified by GME when the 33-kDa protein is bound to PSII is almost the same as that when the protein is free in solution, indicating that these carboxylic residues are still exposed to bulk solution upon binding of the protein to PSII. Therefore, these carboxylic residues are neither in regions interacting with PSII components nor required for maintaining the functional conformation of the protein. This agrees with the finding that modification of the 33-kDa protein with GME in solution did not affect either the rebinding or reactivating capabilities of the protein. On the other hand, the number of lysyl and arginyl residues susceptible to modification decreased significantly upon binding of the 33-kDa protein to PSII. This may be caused by two possibilities. First, the lysyl and arginyl residues are located in regions interacting with PSII intrinsic components and thus became inaccessible to chemical modifiers upon binding. Second, binding of the 33-kDa protein to PSII may induce a conformational change of the protein, which leads to an inaccessibility of the lysyl and arginyl residues to the modifiers. There has been, however, no evidence reported so far for a conformational change (including homodimerization; see Ref. 23) of the 33-kDa protein upon its binding to PSII (see Ref. 32 for a more detailed discussion). Moreover, our present finding that the number of negatively charged carboxylic residues modified was almost the same between the 33-kDa protein free in solution and bound to PSII also suggests that no significant conformational changes occurred upon binding of the protein to PSII. We may conclude therefore that at least a large part of the lysyl and arginyl residues that were modified by NSP, TNBS, or BD when the 33-kDa protein was free in solution but not bound to PSII are located in regions directly interacting with PSII; modification of these residues then led to a loss of rebinding and reactivating capability of the 33-kDa protein.
The locations of lysyl residues that were accessible to modification by
either NSP or TNBS only when the 33-kDa protein was in solution were
determined and will be discussed later. None of the arginyl residues
were liable to modification by BD when the 33-kDa protein was bound to
PSII, whereas the location of the residues modified by BD in solution
could not be determined due to the instability of the BD-modified
arginyl residues. In the spinach 33-kDa protein, there are six arginyl
residues, of which three are completely conserved among 11 sequences
from cyanobacteria to high plants currently available from data bases
(Fig. 7): Arg80, Arg151, and
Arg161. These three conserved arginyl residues may thus be
located in regions interacting with PSII. The other three arginyl
residues, Arg5, Arg122, and Arg178,
are not strictly conserved and may not be important for interaction of
the protein with PSII.
It should be pointed out that modification of either one or two lysyl or one or two arginyl residues did not result in a significant loss of the binding of the 33-kDa protein to PSII. This implies that the interaction of the protein with PSII involves multiple sites and that this interaction did not change much upon destruction of one or two of them. This implies that alteration of only one lysyl or arginyl residue of the 33-kDa protein may not result in a distinct phenotype in site-directed mutagenesis studies on the protein; instead, multiple lysyl or arginyl residues should be mutagenized simultaneously in order to obtain a significantly altered phenotype of the mutant strain.
Domains of Lysyl Residues Possibly Responsible for Interaction of the 33-kDa Protein with PSIITable VI summarizes
the domains determined to be accessible to NSP or TNBS only when the
33-kDa protein is free in solution, together with the domains
accessible to NHS-biotin only in solution reported previously by
Frankel and Bricker (32). Four domains were accessible to both NSP and
TNBS when the 33-kDa protein was free in solution but were prevented
from modification when the protein was associated with PSII membranes:
Glu1-Lys4,
Lys66-Lys76, Lys159, and
Lys230-Lys236. Since the -amino group of
the N terminus was modified by TNBS even when the 33-kDa protein was
associated with PSII, we may conclude that Lys4 in the
domain Glu1-Lys4 is shielded from both NSP and
TNBS modification upon binding of the protein to PSII. Two residues,
Lys130 and Lys186, were modified by NSP only in
solution. These two residues, however, were not modified by TNBS either
in solution or on PSII membranes. On the other hand, residues
Lys14, Lys20, Lys101, and
Lys105 were modified by TNBS only in solution but were not
modified by NSP only in solution. Residues Lys14,
Lys20, and Lys101 were not modified by NSP
either in solution or on PSII membranes, whereas whether
Lys105 was modified by NSP only in solution could not be
determined unambiguously in the present study; two lysyl residues in
domain Lys105-Lys137 were modified by NSP in
solution, whereas only residue Lys130 was modified by NSP
when the 33-kDa protein was associated with PSII membranes. Apparently,
there is one lysyl residue in domain Lys105-Lys137 that is not accessible to NSP
when the 33-kDa protein is bound to PSII. This residue is probably
Lys130 as determined from another V8 protease-digested
fragment, but the possibility that it is Lys105 cannot be
excluded. In any case, the difference of modification observed between
NSP and TNBS may be caused by the difference between the two
modification reactions; although the molecular sizes of NSP and TNBS
are not much different, modification by NSP results in the addition of
a small carbon chain (N-propionyl, add-on mass 56.0 Da)
whereas modification with TNBS results in the addition of the TNP group
containing an aromatic ring, which is much larger in size (molecular
mass 211.0 Da). When the modification was carried out on the 33-kDa
protein free in solution, there was no significant difference between
the number of domains modified by NSP and those modified by TNBS; there
were 12 domains modified by NSP, whereas there were 9 modified by TNBS
(not taking into consideration some peptide fragments that were not
identified). When the modification was carried out on the 33-kDa
protein bound to PSII, on the other hand, the domains modified by TNBS
were much fewer than those modified by NSP; six domains were modified by NSP, but only one domain was modified by TNBS. This may be caused by
the steric hindrance of nearby residues of either the 33-kDa protein or
PSII membrane proteins against access of the large TNP group to the
lysyl residues that are otherwise liable to addition of the small
N-propionyl group when the modification reaction was carried
out with the 33-kDa protein bound to PSII.
|
Based on the above considerations, it is important to use reagents with different structural and reactive groups in the modifying reactions, and the results obtained should be combined in order to get a more complete understanding on the possibly important residues. We therefore consider that the domains modified either by NSP or TNBS only when the 33-kDa protein is in solution are all shielded from contact with bulk solution upon binding to PSII and may be important for the interaction of the protein with PSII. These are Lys4, Lys14, Lys20, Lys66-Lys76, Lys101, Lys105, Lys130, Lys159, Lys186, and Lys230-Lys236. These domains included five of the seven completely conserved lysyl residues present in 11 sequences of the 33-kDa protein from cyanobacteria to higher plants (Fig. 7): Lys66, Lys76, Lys130, Lys159, and Lys236. In addition, the residue Lys186 is completely conserved among eight organisms and is changed to Arg with a similar positive charge only in three cyanobacteria (Fig. 7). The present results thus strongly suggest that these conserved lysyl residues are important for binding of the 33-kDa protein to PSII membranes. The other two conserved lysyl residues that are not unambiguously identified as accessible to modification only in solution are Lys60 and Lys190; whether these two lysyl residues are important for interaction of the protein with PSII is not clear at present.
The domains determined above included the four domains that are accessible to NHS-biotin only when the 33-kDa protein is free in solution, as determined previously by Frankel and Bricker (32): Glu1-Lys4, Lys20, Lys101-Lys105, and Lys159-Lys186 (Table VI). Among these four domains, domains Glu1-Lys4 and Lys159-Lys186 were identified as being modified by both NSP and TNBS only when the 33-kDa protein is in solution. Lys20 and Lys101-Lys105 were modified by TNBS only in solution but not by NSP (the domain Lys101-Lys105 may be modified by NSP only in solution as mentioned above, but our present results did not allow us to draw an unambiguous conclusion). The other domains that were accessible to NSP or TNBS but not to NHS-biotin only in solution are Lys14K (TNBS), Lys130 (NSP), Lys66-Lys76, and Lys230-Lys237 (NSP and TNBS). Among these domains, only Lys14 was accessible to NHS-biotin both in solution and on PSII membranes (32), suggesting that this residue may not directly interact with PSII. The other domains were either completely inaccessible to NHS-biotin (Lys130) or less accessible to NHS-biotin than to the modifiers used in the present work (Lys66-Lys76, Lys230-Lys237). These differences may be attributable to the differences among, in addition to the modifying reactions as described above, the properties of the three modifying reagents used; NSP and TNBS are hydrophilic reagents and may be more efficient in modifying residues located in the hydrophilic part of the protein, whereas NHS-biotin is largely hydrophobic and therefore is expected to access residues in the hydrophobic part of the protein. In any events, our present studies confirmed and further extended the previous results of Frankel and Bricker as to the domains that are important for the interaction of the 33-kDa protein with PSII. These results should provide important clues to further site-directed mutagenesis studies on the 33-kDa protein.
In conclusion, the present results demonstrated that, while negative charges of carboxylic residues (and the C terminus) of the 33-kDa protein do not participate in the functional binding of the protein to PSII, positive charges of lysyl and arginyl residues are important for its binding to, and thus function in, PSII. The domains containing lysyl residues determined to be exposed to bulk solution only when the 33-kDa protein is free in solution are Lys4, Lys20, Lys66-Lys76, Lys101, Lys105, Lys130, Lys159, Lys186, and Lys230-Lys236. These domains may directly interact with PSII intrinsic components.