Phycobilisome Diffusion Is Required for Light-State Transitions in Cyanobacteria1
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
This article is published in Plant Physiology Online, Plant Physiology Preview Section, which publishes manuscripts accepted for
publication after they have been edited and the authors have corrected proofs, but before the final, complete issue is published
online. Early posting of articles reduces normal time to publication by several weeks.
Phycobilisome Diffusion Is Required for Light-State
Transitions in Cyanobacteria1
Sarah Joshua and Conrad W. Mullineaux*
Department of Biology, University College London, London WC1E 6BT, United Kingdom
Phycobilisomes are the major accessory light-harvesting complexes of cyanobacteria and red algae. Studies using fluorescence
recovery after photobleaching on cyanobacteria in vivo have shown that the phycobilisomes are mobile complexes that rapidly
diffuse on the thylakoid membrane surface. By contrast, the PSII core complexes are completely immobile. This indicates that
the association of phycobilisomes with reaction centers must be transient and unstable. Here, we show that when cells of the
cyanobacterium Synechococcus sp. PCC7942 are immersed in buffers of high osmotic strength, the diffusion coefficient for the
phycobilisomes is greatly decreased. This suggests that the interaction between phycobilisomes and reaction centers becomes
much less transient under these conditions. We discuss the possible reasons for this. State transitions are a rapid physiological
adaptation mechanism that regulates the way in which absorbed light energy is distributed between PSI and PSII. Immersing
cells in high osmotic strength buffers inhibits state transitions by locking cells into whichever state they were in prior to
addition of the buffer. The effect on state transitions is induced at the same buffer concentrations as the effect on phycobilisome
diffusion. This implies that phycobilisome diffusion is required for state transitions. The main physiological role for
phycobilisome mobility may be to allow such flexibility in light harvesting.
Phycobilisomes are the main accessory light- coefficients of membrane components (Sarcina et al.,
harvesting complexes of cyanobacteria and red algae. 2001; Mullineaux and Sarcina, 2002). FRAP measure-
They are large, highly structured aggregates of phy- ments show that the phycobilisomes are highly mo-
cobiliproteins associated with the cytoplasmic or stro- bile, with diffusion coefficients typically around 4 3
mal surface of the thylakoid membrane (for review, see 10210 cm2 s21 (Sarcina et al., 2001; Aspinwall et al.,
Grossman et al., 1993; MacColl, 1998). In structural 2004). By contrast, PSII appears completely immobile
terms, little is known of the interaction between the (Mullineaux et al., 1997; Sarcina et al., 2001). This
phycobilisomes and the reaction center core com- indicates that there is no stable association between
plexes. However, energy transfer and mutational stud- phycobilisomes and PSII, and it is likely that the
ies indicate that phycobilisomes can couple directly association of phycobilisomes with PSI is also tran-
to PSI as well as to PSII (Glazer et al., 1994; Mullineaux, sient. Studies with mutant strains of Synechococcus
1994; Ashby and Mullineaux, 1999; Rakhimberdieva 7942 have shown that the phycobilisome diffusion
et al., 2001). coefficient is influenced by a number of factors in-
Fluorescence recovery after photobleaching (FRAP) cluding phycobilisome size and membrane lipid com-
can be used to measure the diffusion of thylakoid position (Sarcina et al., 2001) and the oligomerization
membrane components in cyanobacteria. The tech- of PSI (Aspinwall et al., 2004). The physiological role of
nique involves the use of a highly focused confocal phycobilisome mobility has remained unclear.
laser spot to selectively bleach the fluorophores in State 1-state 2 transitions (state transitions) are
a small region of the cell. The diffusion of the fluo- a rapid physiological adaptation of the photosynthetic
rophores can then be monitored by observing the light-harvesting apparatus, resulting in changes in the
spread and recovery of the bleach (Mullineaux and distribution of excitation energy between PSI and PSII.
Sarcina, 2002). The cyanobacterium Synechococcus sp. State transitions in green plants involve the redistri-
PCC7942 (Synechococcus 7942) has relatively large cells bution of a proportion of the light-harvesting chloro-
and a very regular thylakoid membrane organization, phyll a/b-binding protein of PSII (LHCII). This is
allowing quantitative measurements of the diffusion triggered by LHCII phosphorylation. In State 1, most
of the LHCII is associated with PSII complexes in the
thylakoid grana. On transition to State 2, a proportion
1
of the LHCII decouples from PSII and reassociates
This work was supported by Biotechnology and Biological with PSI in the stroma lamellae. Thus, state transitions
Sciences Research Council (BBSRC) and by The Wellcome Trust
involve a relatively long-range migration of the LHCII
(grants to C.W.M.). S.J. is supported by a BBSRC research student-
ship.
complexes (for review, see Allen and Forsberg, 2001).
* Corresponding author; e-mail c.mullineaux@ucl.ac.uk; fax 44– Although cyanobacteria lack LHCII, they perform
20–7679–7096. state transitions that are analogous to those of green
Article, publication date, and citation information can be found at plants (Fork and Satoh, 1983). State transitions in cyano-
www.plantphysiol.org/cgi/doi/10.1104/pp.104.046110. bacteria involve the redistribution of phycobilisome-
Plant Physiology, August 2004, Vol. 135, pp. 1–8, www.plantphysiol.org Ó 2004 American Society of Plant Biologists 1 of 8
Downloaded on May 8, 2021. - Published by https://plantphysiol.org
Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.Joshua and Mullineaux
absorbed energy between PSII and PSI (van Thor et al.,
1998; McConnell et al., 2002), and one effect of the
transition to State 2 is the functional decoupling of
phycobilisomes from PSII and their reassociation with
PSI (Mullineaux, 1992). The biochemical mechanism is
not known. However, it is clear that state transitions
can be triggered by changes in the redox state of an
intersystem electron carrier (Mullineaux and Allen,
1990), and one gene specifically required for state tran-
sitions in cyanobacteria has been identified (Emlyn-
Jones et al., 1999).
This study establishes a direct connection between
phycobilisome mobility and state transitions. It has
previously been shown that state transitions in cyano-
bacteria are inhibited when cells are immersed in
buffers containing high concentrations of phosphate.
Interestingly, treatment with the buffer locks the cells Figure 1. The 77 K fluorescence emission spectra for cells of Synecho-
into the state to which they were adapted prior to coccus 7942 adapted to State 1 or to State 2. Cells were adapted to State
addition of the buffer, so cells can be locked in either 1 (black line) by incubation in red light or to State 2 (gray line) by
State 1 or in State 2, as judged from fluorescence incubation in the dark before freezing in liquid nitrogen. Fluorescence
spectroscopy (Mullineaux, 1993). Here, we use FRAP spectra were recorded with excitation at 600 nm and normalized to the
to show that treatment of cells of Synechococcus 7942 phycocyanin fluorescence peak (654 nm).
with high-phosphate buffers drastically decreases the
mobility of phycobilisomes. We further show that this
effect is induced at the same phosphate concentrations to red light (Fig. 2A) or dark (Fig. 2B) before freezing
that are required to lock state transitions. We propose samples in liquid nitrogen for 77 K fluorescence
that the buffers lock state transitions because they spectroscopy. The spectra in Figure 2A are essentially
prevent phycobilisomes from decoupling from reac- identical to those in Figure 2B. Therefore, the fluores-
tion centers, and we discuss the reasons this may cence emission spectrum reflects the acclimation of the
occur. cells before addition of the buffer: subsequent accli-
mation to red light or dark has no effect on the
spectrum (Fig. 2). The effect is not species specific as
RESULTS a similar result was obtained with the cyanobacterium
Synechocystis sp. PCC6803 (data not shown).
Adaptation to State 1 or to State 2 can be monitored Exposure to 0.5 M phosphate buffer changes the
by recording fluorescence emission spectra on frozen shape of the fluorescence emission spectrum, with the
samples at 77 K. In State 1, there is greater fluorescence peak at about 685 nm becoming more prominent
emission from PSII relative to the phycobilins and PSI (compare Figs. 1 and 2). This effect was not observed
than there is when cells are adapted to State 2. This when PSII was directly excited at 435 nm (data not
reflects the higher proportion of excitation energy that shown), indicating that the intrinsic fluorescence emis-
is transferred to PSII when cells are adapted to State 1 sion from PSII is not changed by the buffer. PSII has
(Murata, 1969). Figure 1 shows 77 K fluorescence emis- emission peaks at 685 and 695 nm, but with phycobilin
sion spectra for cells of Synechococcus 7942 adapted excitation the 685-nm peak comes partly from the
to State 1 or State 2 prior to freezing the cells. The State terminal emitter pigments of the phycobilisome core
1 spectra show greater relative emission from (Ashby and Mullineaux, 1999). Therefore, the increase
PSII (685–695 nm) as compared to phycocyanin in the 685-nm peak relative to the other peaks, in-
(654 nm) and PSI (720 nm; Fig. 1). cluding the shoulder at 695 nm (Figs. 1 and 2),
It was shown previously that state transitions in the indicates that energy transfer from phycobilisomes to
cyanobacterium Synechococcus sp. PCC6301 could be reaction centers becomes slightly less efficient. The
inhibited by immersing cells in K2HPO4/KH2PO4 effect would be consistent with a structural change
buffers with phosphate concentrations of about 0.2 M leading to slightly slower energy transfer.
or greater. Furthermore, exposure to the buffer had the We have used 77 K fluorescence emission spectra to
effect of locking the cells in either State 1 or State 2, quantify the extent of fixation of light-state. The ratio
depending on the state to which the cells were adapted of fluorescence at 685 nm to fluorescence at 654 nm
prior to addition of the buffer (Mullineaux, 1993). (F685/F654) gives an indication of light state, with
Here, we show a similar result with Synechococcus a higher F685/F654 indicating adaptation to State 1
7942. Cells were adapted to State 1 or to State 2 by (Fig. 1). The extent of fixation in State 1 may be quan-
incubation in respectively red light or dark for 5 min. tified as (LD2DD)/(LL2DD), where LD is the F685/
Phosphate buffer (0.5 M K2HPO4/KH2PO4, pH 6.8) was F654 for cells adapted to red light and then readapted
then added, and the suspension was again acclimated to dark after addition of phosphate buffer, DD is
2 of 8 Plant Physiol. Vol. 135, 2004
Downloaded on May 8, 2021. - Published by https://plantphysiol.org
Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.Phycobilisome Diffusion and State Transitions
Figure 2. The 77 K fluorescence emission spectra for
cells of Synechococcus 7942 in 0.5 M phosphate
buffer. Fluorescence spectra recorded with excitation
at 600 nm and normalized to the phycocyanin
fluorescence peak (654 nm). Cells were adapted to
red light (black line) or to dark (gray line) before
addition of phosphate buffer. A, Cells readapted to
red light after addition of phosphate buffer. B, Cells
readapted to dark after addition of phosphate buffer.
F685/F654 for cells adapted to dark both before and after There is no fixation of light state at 0.1 M phosphate:
addition of phosphate buffer, and LL is F685/F654 for the negative values for fixation at this concentration
cells adapted to red light both before and after addi- indicate that state transitions are slightly enhanced
tion of phosphate buffer. If the cells were completely (Fig. 3). State transitions are partially inhibited at 0.2 to
fixed in State 1, then dark adaptation after addition of 0.3 M phosphate, and cells can be completely locked in
phosphate would have no effect on the spectrum. Then either state at 0.4 M phosphate or above. As observed
LD would be equal to LL and (LD2DD)/(LL2DD) previously in Synechococcus 6301 (Mullineaux, 1993),
would be 1. Conversely, if there was no fixation of light the concentration of phosphate required to fix cells in
state, then LD would be equal to DD and (LD2DD)/ State 1 appears slightly lower than that required to fix
(LL2DD) would be 0. The extent of fixation in State 2 cells in State 2 (Fig. 3). Figure 4 shows 77 K fluores-
can be quantified in a similar way as (DL2DD)/ cence emission spectra for cells treated with 0.2 M
(LL2DD), where DL is the F685/F654 for cells adapted phosphate. At this concentration, there is significant
to dark and then readapted to red light after addition fixation of cells in State 1 but not in State 2. Therefore,
of phosphate buffer. Figure 3 shows the extent of when cells are readapted to dark after addition of
fixation in State 1 and State 2 in Synechococcus 7942 phosphate buffer, cells initially adapted to State 1
as a function of phosphate concentration. Fixation remain close to State 1 (Fig. 4B). However, when cells
increases with increasing phosphate concentration. are readapted to red light after addition of buffer, cells
initially in State 2 revert to State 1 (Fig. 4A).
We have also examined the effect of phosphate
buffers on the kinetics of state transitions, monitored
by fluorescence time courses at room temperature.
Cells were adapted to State 2 in the dark. Phosphate
buffer was then added, and 3-(3,4-dichlorophenyl)-
1,1-dimethylurea (DCMU) was added to block elec-
tron transport at the acceptor side of PSII. Cells were
then illuminated with bright light, and red fluores-
cence was monitored over a time course of a few
minutes. The transition to State 1 results in a fluores-
cence rise on a timescale of a few seconds to a few
minutes (Schluchter et al., 1996; Aspinwall et al., 2004).
The bright illumination ensures that the kinetics of the
state transition are not limited by the kinetics of
electron transport, and in the presence of DCMU, PSII
centers remain closed, so that the fluorescence kinetics
are not complicated by changes in PSII trap closure
(Schluchter et al., 1996; Aspinwall et al., 2004). The
Figure 3. Fixation of cells in State 1 and State 2 as a function of results are shown in Figure 5. The State 1 transition is
phosphate concentration. Cells were adapted to State 1 (black) or to not inhibited at 0.1 M phosphate, and in fact the fluores-
State 2 (gray) before addition of phosphate buffer. Cells were then cence rise is more rapid in 0.1 M phosphate than in
adapted to the opposite light regime before being frozen for 77 K growth medium. At 0.2 M phosphate, the fluorescence
fluorescence spectroscopy. Fixation of light state is calculated from rise becomes much smaller and slower. At 0.3 M
F685/F654 from 77 K spectra. Fixation in State 1 is defined as (LD2DD)/
phosphate and above, there is apparently complete
(LL2DD) and fixation in State 2 is defined as (DL2DD)/(LL2DD),
where LD is F685/F654 for cells adapted to red light and then readapted to
inhibition of the State 1 transition, and fluorescence
dark after addition of phosphate buffer, DL is F685/F654 for cells adapted falls slightly during illumination (Fig. 5).
to dark and then readapted to red light after addition of buffer, and so We found that fixation is not influenced by small
on. Mean data from five samples is presented. SDs are shown by the changes in pH; similar results were obtained with
error bars. phosphate buffers at pH 6.0 and pH 8.0 (data not
Plant Physiol. Vol. 135, 2004 3 of 8
Downloaded on May 8, 2021. - Published by https://plantphysiol.org
Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.Joshua and Mullineaux
Figure 4. The 77 K fluorescence emission spectra for
cells of Synechococcus 7942 in 0.2 M phosphate
buffer. Fluorescence spectra recorded with excitation
at 600 nm and normalized to the phycocyanin
fluorescence peak (654 nm). Cells were adapted to
red light (black line) or to dark (gray line) before
addition of phosphate buffer. A, Cells readapted to
red light after addition of phosphate buffer. B, Cells
readapted to dark after addition of phosphate buffer.
shown). We also found cells could be fixed in either fixation of cells in State 1 is 0.97 6 0.16, but the fixation
state with sucrose (Suc) solutions or with high con- of cells in State 2 is only 0.17 6 0.09. The critical con-
centrations of potassium chloride (data not shown). As centration of potassium chloride was also about 0.8 M,
with phosphate, low concentrations of Suc and potas- again higher than for potassium phosphate. In this
sium chloride enhanced state transitions rather than case both the anion and the cation are small, so the
inhibiting them (data not shown). Thus, the effects are solute is likely to leak across the plasma membrane,
not specific to phosphate and are presumably related reducing the osmotic effect. Comparison of the critical
to osmotic strength rather than to any more specific concentrations of potassium phosphate and potassium
chemical properties of the buffer. The minimum con- chloride confirms that this is not an electrostatic effect
centrations of Suc and potassium chloride required to dependent on ionic strength in the cytoplasm. If this
inhibit the State 1 transition were estimated using were the case, then potassium chloride would be more
kinetic experiments of the type shown in Figure 5. The effective than potassium phosphate because potas-
minimum concentration of Suc required for complete sium chloride will penetrate the plasma membrane
inhibition of the State 1 transition was about 0.8 M. This more easily.
compares with a critical concentration of 0.3 M potas- In view of the effect of high osmotic strength buffers
sium phosphate (Fig. 5). The higher critical concentra- on state transitions, it is of interest to see if these
tion of Suc is consistent with an osmotic effect: buffers affect the diffusion of phycobilisomes. We
a potassium phosphate solution will exert a greater therefore carried out FRAP measurements using a
osmotic effect than an equimolar solution of Suc laser-scanning confocal microscope as described pre-
because potassium phosphate dissociates into anions viously (Sarcina et al., 2001; Aspinwall et al., 2004). A
and cations in solution. At the intermediate Suc 633-nm red He-Ne laser was used to selectively excite
concentration of 0.5 M, 77 K fluorescence measure- phycocyanin, and fluorescence emission was moni-
ments of the type shown in Figures 2 and 4 show that tored at .665 nm. At room temperature, most fluo-
rescence at these wavelengths comes from the
phycobilisome cores, so these settings can be used
to monitor the diffusion of intact phycobilisomes
(Sarcina et al., 2001). Prior to the measurement, cells
were grown in the presence of 0.5% dimethylsulfox-
ide; this generates elongated cells that are more suit-
able for quantitative FRAP measurements (Aspinwall
et al., 2004). The elongated cells do not exhibit any
changes in thylakoid membrane structure as judged
by electron microscopy and fluorescence imaging, and
their photosynthetic properties as judged by fluores-
cence spectroscopy and oxygen-electrode measure-
ments are normal (Mullineaux and Sarcina, 2002).
Where appropriate, cells were dark adapted and then
treated with phosphate buffer. Cells were then immo-
bilized by adsorption onto agar containing growth
medium or phosphate buffer. The confocal spot was
scanned across the center of a cell for 2 s to bleach the
Figure 5. Effect of phosphate buffers on the kinetics of state transitions.
Cells were dark adapted (State 2) prior to addition of phosphate buffer.
phycobilins. The laser power was then reduced by a
Cells were then illuminated in the presence of DCMU as described in factor of 8 to prevent further bleaching, and the cell
‘‘Materials and Methods.’’ The faster phase of the fluorescence rise was repeatedly imaged. Phycobilisome diffusion cau-
(appearing immediate on this timescale) has been subtracted. Fluores- ses the bleached line to spread and fill in, and the
cence is expressed relative to this initial fluorescence for cells with no phycobilisome diffusion coefficient can be calculated
added phosphate. from the rate of recovery of fluorescence at the center
4 of 8 Plant Physiol. Vol. 135, 2004
Downloaded on May 8, 2021. - Published by https://plantphysiol.org
Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.Phycobilisome Diffusion and State Transitions
of the bleach (Mullineaux et al., 1997; Mullineaux and
Sarcina, 2002). Figure 6 shows a typical result for a cell
in growth medium. Under these conditions the fluo-
rescence in the center of the bleach recovers on
a timescale of seconds, and we measured a mean
phycobilisome diffusion coefficient of (4.3 6 1.7) 3
10210 cm2 s21, which is comparable with that pre-
viously obtained (Sarcina et al., 2001; Aspinwall et al.,
2004). By contrast, when cells are treated with 0.5 M
phosphate buffer, no fluorescence recovery is seen on
these short timescales, and only partial recovery is
seen after 20 min (Fig. 7). Under these conditions, we
estimated the mean phycobilisome diffusion coeffi-
cient to be (9.8 6 0.6) 3 10212 cm2 s21. Thus, exposure
to the buffer decreases the rate of diffusion of the
phycobilisomes by a factor of about 40 to 50. Figure 8
shows the dependence of the mean phycobilisome
Figure 7. FRAP image sequence for a cell of Synechococcus 7942 in
diffusion coefficient on phosphate concentration. At
0.5 M phosphate buffer Confocal fluorescence images taken from
0.1 M phosphate, phycobilisomes diffuse slightly faster a typical FRAP sequence. Scale bar 5 5 mm. Fluorescence from the
on average than in growth medium, but we cannot be phycobilisomes is imaged. After recording the prebleach image, a line
sure that this is significant due to the substantial was bleached across the cell by increasing the laser power and
variation in diffusion coefficient from cell to cell. At scanning the laser for 2 s in the X-direction. The laser power was then
0.2 M phosphate, there is a significant decrease in the decreased again and a series of images were recorded. For each image,
diffusion coefficient, and a further significant decrease the time after recording the first post-bleach image is shown.
at 0.3 M phosphate.
We found that the effect of phosphate bufferon phyco-
bilisome mobility is reversible. Cells were treated
with 0.5 M phosphate buffer and then adsorbed onto cm2 s21 with no significant difference from untreated
agar containing growth medium rather than 0.5 M cells (t test P 5 0.3).
phosphate. Under these conditions, the phosphate The FRAP data presented above (Figs. 7 and 8) were
will be diluted by the growth medium in the agar. obtained from cells that were dark adapted prior to
The rapid diffusion of phycobilisomes is then restored. addition of phosphate buffer. At high phosphate con-
The mean diffusion coefficient was (6.0 6 3.3) 3 10210 centrations, the cells will therefore be fixed in State 2
(Figs. 2 and 3). We obtained similar results with cells
fixed in State 1 with 0.5 M phosphate buffer (data not
shown), and we could not detect any significant
differences between State 1- and State 2-adapted cells
at lower phosphate concentrations (data not shown).
However, Suc solutions also affected the phycobili-
some diffusion coefficient. When we carried out FRAP
measurements on cells in 0.5 M Suc, a concentration at
which cells are efficiently fixed in State 1 but not in
State 2 (see above), we found a significant difference
depending on adaptation prior to addition of Suc.
For cells adapted to State 1, the mean phycobili-
some diffusion coefficient was (6.5 6 0.3) 3 10212 cm2
s21, whereas for cells adapted to State 2 it was (4.7 6
0.3) 3 10211 cm2 s21. The difference is significant
(t test, P 5 0.019). Thus, at this Suc concentration, the
rate of phycobilisome diffusion is much more strongly
affected when cells are adapted to State 1.
Figure 6. FRAP image sequence for a cell of Synechococcus 7942 in
growth medium. Confocal fluorescence images taken from a typical DISCUSSION
FRAP sequence. Scale bar 5 5 mm. Fluorescence from the phycobi-
lisomes is imaged. After recording the prebleach image, a line was Previous FRAP studies have shown that the phyco-
bleached across the cell by increasing the laser power and scanning the bilisomes are highly mobile complexes, diffusing rap-
laser for 2 s in the X-direction. The laser power was then decreased idly on the thylakoid membrane (Mullineaux et al.,
again and a series of images were recorded. For each image, the time 1997; Sarcina et al., 2001). Figure 6 shows the rapid
after recording the first post-bleach image is shown. diffusion of phycobilisomes under physiological con-
Plant Physiol. Vol. 135, 2004 5 of 8
Downloaded on May 8, 2021. - Published by https://plantphysiol.org
Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.Joshua and Mullineaux
in high-phosphate buffers. Furthermore, cells are
locked in the state to which they were adapted prior
to addition of the buffer (Mullineaux, 1993). Our
current results confirm this effect for Synechococcus
7942 (Fig. 2). We found similar effects with Suc and
potassium chloride solutions, and glycine (Gly) beta-
ine also has this effect (Li et al., 2004). Thus, the effect
on state transitions also appears to depend on the
osmotic strength of the buffer. We tested the effect of
increasing concentrations of phosphate (Figs. 3–5). The
0.1 M phosphate does not inhibit state transitions (Fig.
3), which actually occur more rapidly in 0.1 M phos-
phate than in growth medium (Fig. 5). At 0.2 M
phosphate, the transition to State 2 is partially inhibi-
ted (Fig. 3). There is little effect on the transition to
Figure 8. Phycobilisome diffusion coefficients in Synechococcus 7942 State 1 as judged from 77 K fluorescence spectra
cells in different concentrations of phosphate buffer. Diffusion coef- recorded on cells frozen after 5 min of adaptation
ficients are means from measurements on six cells. SDs are shown by the (Figs. 3 and 4). However, the kinetics of the State 1
error bars. transition as judged from room-temperature fluores-
cence time courses are strongly affected (Fig. 5). At 0.3
M phosphate and above, both the State 1 and the State 2
ditions in vivo. However, we show here that when transitions are very strongly inhibited (Figs. 3 and 5).
cells of Synechococcus 7942 are immersed in concen- The phosphate concentration dependence of the state
trated phosphate buffers, the rate of diffusion of transition effect correlates with the phosphate concen-
phycobilisomes is drastically decreased (Figs. 7 and tration dependence of the phycobilisome diffusion
8). This appears to be a rather nonspecific osmotic coefficient (Fig. 8). The 0.1 M phosphate slightly
effect, since it is pH independent and we obtained increases the mean phycobilisome diffusion coeffi-
similar results with Suc solutions. The most plausible cient, although the variation from cell to cell means
cause of such an extreme decrease in the diffusion that this may not be significant (Fig. 8). Phycobilisomes
coefficient is a strong interaction with another, immo- diffuse significantly slower at 0.2 M phosphate, and the
bile component in the membrane (Zhang et al., 1993). diffusion coefficient becomes minimal at 0.3 M phos-
FRAP studies have shown that PSII is completely phate and above (Fig. 8). Therefore, we propose that
immobile under normal conditions (Mullineaux et al., the two effects are linked.
1997; Sarcina et al., 2001). We have no direct informa- A number of models for the mechanism of state
tion on the diffusion of PSI, but its diffusion is likely to transitions in cyanobacteria have been proposed
be very slow in the crowded environment of a cyano- (Allen and Holmes, 1986; Bruce et al., 1989; Meunier
bacterial thylakoid membrane (Mullineaux, 1999). et al., 1997). Recent work has led to a consensus that
Therefore, we suggest that high osmotic strength state transitions involve changes in the relative energy
buffers greatly increase the stability of phycobili- transfer from phycobilisomes to the PSI and PSII
some-reaction center complexes and that this restricts reaction centers (Mullineaux, 1994; Schluchter et al.,
the diffusion of the phycobilisomes. It is notable that 1996; McConnell et al., 2002). This is usually accom-
high osmotic strength buffers are essential for the panied by a redistribution of chlorophyll-absorbed
isolation of membranes and reaction centers with energy, although the two effects can be separated
functionally coupled phycobilisomes. The buffer de- (Emlyn-Jones et al., 1999; McConnell et al., 2002). The
veloped for this purpose by Gantt and co-workers structural basis of state transitions is still controversial.
contains 0.5 M phosphate and 0.5 M Suc (Katoh and The mobility of phycobilisomes (Mullineaux et al.,
Gantt, 1979; Gantt et al., 1988). It is possible that high 1997) suggests that state transitions may involve the
osmotic strength buffers stabilize phycobilisome-- physical decoupling of phycobilisomes from one type
reaction center complexes by decreasing the avail- of reaction center and their reassociation with the
ability of water molecules around the thylakoid other, as originally proposed by Allen and co-workers
membranes. Consistent with this, we found the effect (Allen and Holmes, 1986). The transition to State 1
on the phycobilisome diffusion coefficient to be fully occurs more rapidly in mutants lacking PsaL, the
reversible and to be pH independent. The 0.5 M subunit required for trimerization of PSI (Schluchter
phosphate buffer changes the shape of the 77 K et al., 1996; Aspinwall et al., 2004). Phycobilisomes also
fluorescence spectrum, with the 685-nm peak becom- diffuse more rapidly in PsaL mutants, which is con-
ing more prominent (Figs. 1 and 2). This is consistent sistent with the idea that state transitions involve
with an altered structural interaction between phyco- phycobilisome mobility (Aspinwall et al., 2004). How-
bilisomes and reaction centers. ever, other models that do not involve the movement
It was previously shown that state transitions in of phycobilisomes have been proposed (Schluchter
cyanobacteria are inhibited when cells are immersed et al., 1996; McConnell et al., 2002).
6 of 8 Plant Physiol. Vol. 135, 2004
Downloaded on May 8, 2021. - Published by https://plantphysiol.org
Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.Phycobilisome Diffusion and State Transitions
Our current results indicate that phycobilisome Adaptation of Cells to State 1 or to State 2
mobility is critical for state transitions: the same Cells were harvested by centrifugation and resuspended in growth
conditions that immobilize phycobilisomes also lock medium to a final chlorophyll concentration of 5 mM. Chlorophyll concen-
cells into the light state to which they were adapted. trations were determined by methanol extraction (Porra et al., 1989). Aliquots
Under physiological conditions, the association be- of the cell suspension were injected into quartz capillary tubes (2.5 mm
internal diameter). For State 2, the sample was dark adapted for 5 min, and for
tween phycobilisomes is unstable and transient, and
State 1 the sample was adapted for 5 min to a red light defined by a Schott
a phycobilisome will frequently detach from a reaction RG665 glass filter (transmitting wavelengths longer than about 665 nm). The
center, diffuse, and reassociate with another reaction light intensity was 20 mE m22 s21.
center (Mullineaux et al., 1997; Sarcina et al., 2001). We
propose that high osmotic strength buffers prevent Fluorescence Spectroscopy
phycobilisomes from decoupling from reaction centers
and that this prevents any redistribution of phycobi- Cell suspensions at 5 mM chlorophyll in growth medium or buffer in quartz
lisomes between PSII and PSI. capillary tubes were frozen by dropping the tube into liquid nitrogen.
Fluorescence emission spectra were recorded at 77 K in a Perkin-Elmer
With 0.1 M phosphate, state transitions are enhanced (Foster City, CA) LS50 luminescence spectrometer equipped with a liquid-
(Figs. 3 and 5) and the mean phycobilisome diffusion nitrogen sample housing and a red-sensitive photomultiplier. The excitation
coefficient increases, although this may not be signif- wavelength was 600 nm, and emission was scanned from 620 to 750 nm.
icant (Fig. 8). It is therefore possible that under these Excitation and emission slitwidths were 5 nm. Because the absolute ampli-
tudes of low-temperature spectra are unreliable, spectra were routinely
conditions, phycobilisome-reaction center interactions normalized to the phycocyanin fluorescence emission peak at 654 nm.
actually become more labile than in growth medium. Fluorescence ratios were averaged from spectra obtained from five samples.
At intermediate osmotic strengths (0.2 M phosphate or
0.5 M Suc), cells are fixed in State 1 much more
Treatment with High Osmotic Strength Buffer
efficiently than they are fixed in State 2 (Figs. 3 and
4). The difference is particularly pronounced in 0.5 M The phosphate buffers used were K2HPO4/KH2PO4 solutions at phosphate
Suc, and at this Suc concentration, phycobilisome concentrations of 0.1 to 0.5 M. The pH was 6.8 unless otherwise specified. An
aliquot of cell suspension at a chlorophyll concentration of 50 mm was placed
mobility is reduced significantly more when cells are
in a stirred beaker and preadapted to State 1 or to State 2 using the
adapted to State 1 prior to addition of Suc. This illumination conditions described above. Nine volumes of buffer were then
suggests that the phycobilisome-PSII complex is sta- added, and the illumination conditions were maintained for a further 5 min.
bilized at slightly lower osmotic strength than the For fluorescence spectroscopy, an aliquot of the cell suspension was then
phycobilisome-PSI complex. injected into a quartz capillary tube and adapted for a further 5 min to either
red light or dark before freezing and recording fluorescence spectra as
described above.
CONCLUSIONS
Kinetics of State Transitions
Phycobilisome mobility is critical for state transi- Fluorescence transients were recorded at room temperature using a
tions in cyanobacteria. Our data support a model in laboratory-built fluorimeter (Peter Rich, University College London, UK).
which excitation energy distribution from phycobili- Cells were resuspended in growth medium at a chlorophyll concentration of
somes to reaction centers is governed by a dynamic 30 mM and dark adapted for 5 min. Nine volumes of phosphate buffer were
added, and the suspension was kept in the dark for a further 5 min. DCMU
equilibrium in which PSII and PSI reaction centers was then added to a final concentration of 50 mM. The cells were then illumi-
compete to bind phycobilisomes. State transitions nated with phycobilin-absorbed light defined by a combination of Schott
change the position of the equilibrium by changing RG610 and Ealing 660-nm short-pass filters. The illumination was controlled
the binding constant of phycobilisomes with one or by an electronic shutter opening in about 1 ms and was at an intensity of
100 mE m22 s21. Fluorescence was detected by a photomultiplier screened by
both of the reaction centers, although the biochemical a Schott RG695 red glass filter.
mechanism is not known. High osmotic strength
buffers stabilize phycobilisome-reaction center associ-
ation, and this has the effect of drastically slowing the FRAP Measurements
diffusion of phycobilisomes and preventing any re- Cells grown in the presence of 0.5% dimethylsulfoxide, as described above,
distribution of the phycobilisomes between PSI and were used. Where specified, cells were pretreated with phosphate buffer as
PSII. The major physiological role of phycobilisome described above. In this case, cells were preadapted to State 2 by dark
incubation unless otherwise specified. Cell suspensions were spotted onto
mobility may be to allow flexibility in light harvesting.
1.5% agar plates (Difco Bacto-Agar). The agar was made up either with growth
medium (for untreated cells) or phosphate buffer at the appropriate concen-
tration. When the cell suspension was adsorbed onto the agar, a small block of
MATERIALS AND METHODS agar was cut out and placed in a laboratory-built water-jacketed sample
holder with a 0.2-mm glass coverslip pressed onto the agar surface. During
Strains and Culture Conditions measurements, the temperature was maintained at 30°C by connecting the
sample holder to a circulating water bath.
Wild-type Synechococcus sp. PCC7942 was obtained from the Pasteur FRAP measurements were carried out essentially as described previously
Culture Collection. Cells were grown in liquid culture in BG11 medium (Sarcina et al., 2001; Aspinwall et al., 2004) using a Nikon (Tokyo) PCM2000
(Castenholz, 1988) supplemented with 10 mM NaHCO3. Cells were grown in laser-scanning confocal microscope equipped with 20 mW red He-Ne laser
batch culture in an illuminated orbital incubator at 30°C under white light at (633 nm). A 60 3 oil-immersion objective lens was used, with a 50-mm
9 mE m22 s21. For FRAP studies, the growth medium was supplemented with confocal pinhole. Fluorescence emission was defined by a Schott RG665 red
dimethylsulfoxide at 0.5% (v/v) for 3 d prior to the experiment, as this induces glass filter, transmitting wavelengths above about 665 nm. When recording
cell elongation (Aspinwall et al., 2004). images, the laser intensity was reduced by a factor of 8 with a neutral density
Plant Physiol. Vol. 135, 2004 7 of 8
Downloaded on May 8, 2021. - Published by https://plantphysiol.org
Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.Joshua and Mullineaux
filter and scanned over a sample region of 20.5 3 20.5 mm. Cells aligned in the Grossman AR, Schaefer MR, Chiang GG, Collier JL (1993) The phycobi-
Y-direction were used for FRAP measurements. After recording a prebleach lisome, a light-harvesting complex responsive to environmental con-
image by scanning the laser in the XY-plane, the microscope was switched to ditions. Microbiol Rev 57: 725–749
X-scanning mode, the laser intensity was increased 8-fold by removing the Katoh T, Gantt E (1979) Photosynthetic vesicles with bound phycobili-
neutral density filter, and the confocal spot was scanned across the cell for 2 s somes from Anabaena variabilis. Biochim Biophys Acta 546: 383–393
to bleach a line across the cell. The laser intensity was then decreased again Li D, Xie J, Zhao J, Xia A, Li D, Gong Y (2004) Light-induced excitation
and images were recorded every 3 s for 30 s. Where appropriate, further energy redistribution in Spirulina platensis cells: ‘‘spillover’’ or ‘‘mobile
images were recorded every 5 min up to 30 min after the bleach. PBSs’’? Biochim Biophys Acta 1608: 114–121
MacColl R (1998) Cyanobacterial phycobilisomes. J Struct Biol 124:
311–334
FRAP Data Analysis McConnell MD, Koop R, Vasil’ev S, Bruce D (2002) Regulation of the
distribution of chlorophyll and phycobilin-absorbed excitation energy
Diffusion coefficients were obtained from the image series as described
in cyanobacteria. A structure-based model for the light state transition.
previously (Mullineaux et al., 1997; Sarcina et al., 2001; Mullineaux and
Plant Physiol 130: 1201–1212
Sarcina, 2002; Aspinwall et al., 2004). Optimas 5.2 image analysis software was
Meunier PC, Colón-López MS, Sherman LA (1997) Temporal changes in
used to obtain a one-dimensional fluorescence profile along the long axis of
state transitions and photosystem organization in the unicellular
the cell, summing pixel values across the cell in the X-direction. Fluorescence
diazotrophic cyanobacterium Cyanothece sp. ATCC51142. Plant Physiol
profiles for bleached cells were corrected by subtracting the prebleach profile
115: 991–1000
and then fitted to Gaussian curves using SigmaPlot software (Jandel Scientific,
Mullineaux CW (1992) Excitation energy transfer from phycobilisomes
San Rafael, CA). Diffusion coefficients were obtained by plotting bleach depth
to photosystem I in a cyanobacterium. Biochim Biophys Acta 1100:
versus time according to a one-dimensional diffusion equation (Mullineaux
285–292
et al., 1997; Aspinwall et al., 2004; Mullineaux, 2004). Values shown are means
Mullineaux CW (1993) Inhibition by phosphate of light-state transitions in
from six cells, with SDs.
cyanobacterial cells. Photosynth Res 38: 135–140
Mullineaux CW (1994) Excitation energy transfer from phycobilisomes to
photosystem I in a cyanobacterial mutant lacking photosystem II.
ACKNOWLEDGMENTS Biochim Biophys Acta 1184: 71–77
Mullineaux CW (1999) The thylakoid membranes of cyanobacteria: struc-
We thank Santiago Garcia for his excellent technical support and con- ture, dynamics and function. Aust J Plant Physiol 26: 671–677
struction of the microscope sample holder, Peter Rich for help with the kinetic Mullineaux CW (2004) FRAP analysis of photosynthetic membranes. J Exp
measurements, and Mary Sarcina for helping to develop the FRAP technique. Bot 55: 1207–1211
Mullineaux CW, Allen JF (1990) State 1-state 2 transitions in the cyano-
Received May 10, 2004; returned for revision May 28, 2004; accepted May 28,
bacterium Synechococcus 6301 are controlled by the redox state of
2004.
electron carriers between photosystems I and II. Photosynth Res 22:
157–166
LITERATURE CITED Mullineaux CW, Sarcina M (2002) Probing the dynamics of photosynthetic
membranes with fluorescence recovery after photobleaching. Trends
Allen JF, Forsberg J (2001) Molecular recognition in thylakoid structure Plant Sci 7: 237–240
and function. Trends Plant Sci 6: 317–326 Mullineaux CW, Tobin MJ, Jones GR (1997) Mobility of photosynthetic
Allen JF, Holmes NG (1986) A general model for regulation of photosyn- complexes in thylakoid membranes. Nature 390: 421–424
thetic unit function by protein phosphorylation. FEBS Lett 202: 175–181 Murata N (1969) Control of excitation transfer in photosynthesis. 1. Light-
Ashby MK, Mullineaux CW (1999) The role of ApcD and ApcF in energy induced change of chlorophyll a fluorescence in Porphyridium cruentum.
transfer from phycobilisomes to PSI and PSII in a cyanobacterium. Biochim Biophys Acta 172: 242–251
Photosynth Res 61: 169–179 Porra RJ, Thompson WA, Kriedeman PE (1989) Determination of accurate
Aspinwall CL, Sarcina M, Mullineaux CW (2004) Phycobilisome mobility extinction coefficients and simultaneous equations for assaying chlo-
in the cyanobacterium Synechococcus sp. PCC7942 is influenced by the rophylls a and b extracted with four different solvents: verification of the
trimerisation of Photosystem I. Photosynth Res 79: 179–187 concentration of chlorophyll standards by atomic absorption spectros-
Bruce D, Brimble S, Bryant DA (1989) State transitions in a phycobilisome- copy. Biochim Biophys Acta 975: 384–394
less mutant of the cyanobacterium Synechococcus sp. PCC7002. Biochim Rakhimberdieva MG, Boichenko VA, Karapetyan N, Stadnichuk IN
Biophys Acta 974: 66–73 (2001) Interaction of phycobilisomes with photosystem II dimers and
Castenholz RW (1988) Culturing methods for cyanobacteria. In L Packer, photosystem I monomers and trimers in the cyanobacterium Spirulina
AN Glazer, eds, Methods in Enzymology, Vol 167. Academic Press, San platensis. Biochemistry 40: 15780–15788
Diego, pp 68–93 Sarcina M, Tobin MJ, Mullineaux CW (2001) Diffusion of phycobilisomes
Emlyn-Jones D, Ashby MK, Mullineaux CW (1999) A gene required for on the thylakoid membranes of the cyanobacterium Synechococcus 7942.
the regulation of photosynthetic light-harvesting in the cyanobacterium Effects of phycobilisome size, temperature and membrane lipid com-
Synechocystis 6803. Mol Microbiol 33: 1050–1058 position. J Biol Chem 276: 46830–46834
Fork DC, Satoh K (1983) State I-state II transitions in the thermophilic blue- Schluchter WM, Shen G, Zhao J, Bryant DA (1996) Characterization of
green alga (cyanobacterium) Synechococcus lividus. Photochem Photobiol psaI and psaL mutants of Synechococcus sp. strain PCC7002: a new
37: 421–427 model for state transitions in cyanobacteria. Photochem Photobiol 64:
Gantt E, Clement-Metral JD, Chereskin BM (1988) Photosystem II-phyco- 53–66
bilisome complex preparations. In L Packer L, AN Glazer, eds, Methods van Thor JJ, Mullineaux CW, Matthijs HCP, Hellingwerf KJ (1998) Light-
in Enzymology, Vol 167. Academic Press, San Diego, pp 286–290 harvesting and state transitions in cyanobacteria. Bot Acta 111:
Glazer AN, Gindt YM, Chan CF, Sauer K (1994) Selective disruption of 430–443
energy flow from phycobilisomes to photosystem I. Photosynth Res 40: Zhang F, Lee GM, Jacobson K (1993) Protein lateral mobility as a reflection
167–173 of membrane microstructure. Bioessays 15: 579–588
8 of 8 Plant Physiol. Vol. 135, 2004
Downloaded on May 8, 2021. - Published by https://plantphysiol.org
Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.You can also read
