Copyright © Physiologia Plantarum 1999PHYSIOLOGIA PLANTARUM 105: 746–755. 1999Printed in Ireland—all rights reser6ed ISSN 0031-9317

Oxygen-dependent electron flow influences photosystem II function andpsbA gene expression in the cyanobacterium Synechococcus sp. PCC7942

Douglas Campbella,*, Adrian K. Clarkeb, Petter Gustafssonb and Gunnar O8 quistb

aDepartment of Biology, Mount Allison Uni6ersity, Sack6ille, NB, E4L 1G7, CanadabDepartment of Plant Physiology, Uni6ersity of Umea, S-90187 Umea, Sweden*Corresponding author, e-mail: dcampbell@mta.ca

Received 11 June 1998; revised 16 November 1998

During acclimated growth in Synechococcus sp. PCC 7942 a message, encoding the D1:1 protein of PSII, and induction ofpsbAII/AIII encoding the alternate D1:2 protein. The changessubstantial proportion of the electrons extracted from water

by photosystem II ultimately flow back to oxygen. This flow in the mRNA pool are not, however, reflected at the proteinincreases rapidly under high light, which allows Synechococ- level, and D1:1 remains in the thylakoid membranes. There iscus to maintain photosystem II centers largely open, even no accumulation of D1:2, despite some continued synthesis of

other proteins. PSII closure, therefore, results in repression ofunder excessive excitation. The electron flow to oxygen withpsbAI and induction psbAII/AIII expression, but D1:1/D1:2increasing light accounts for the progressive discrepancy be-

tween the light response curve of measured oxygen evolution, exchange is blocked by anoxia, downstream from transcrip-tion. D1:1 protein and PSII activity are quite stable underand the light response curve of photosystem II activity esti-anoxia and moderate illumination. Nevertheless, upon recov-mated from fluorescence measures. In cells under anoxia thisery under oxygenic conditions, the existing D1:1 is lost fromflexible electron sink is lost and photosystem II centers sufferthe membranes, resulting in a transient drop in PSII activity.partial closure at the growth light intensity, with closureThis suggests that under normal conditions the cells usebecoming more severe under excess light. As predicted from

earlier work this PSII closure results in rapid loss of psbAI oxygen to facilitate preemptive turnover of D1 proteins.

Introduction

Under illumination, cyanobacteria and other photobiontsmust continually balance the capture of light energy withits subsequent conversion to chemical energy. The activityof the photosystem II (PSII) reaction center is key tomaintaining this balance between photon capture and elec-tron flow. Furthermore, in cyanobacteria, electrons origi-nating from photosystem II centers enter a complex andflexible web of electron carriers. Electron fluxes elsewherein the transport system strongly influence the activity andregulation of photosystem II (Mullineaux and Allen 1986,Dominy and Williams 1987, Miller et al. 1991, Herbert etal. 1995, Mano et al. 1995, Mi et al. 1995, Mir et al. 1995,

Schreiber et al. 1995, Schubert et al. 1995, Li and Canvin1997a,b,c, Tanaka et al. 1997).

One such flux is the removal of electrons from the thy-lakoid-based transport system through donation of elec-trons to oxygen. Several paths for electron donation tooxygen exist in cyanobacteria (Peschek et al. 1982, Alpes etal. 1984, Binder et al. 1984, Scherer et al. 1988, Badgerand Schreiber, 1993, Shyam et al. 1993, Geerts et al. 1994,Schubert et al. 1995, Li and Canvin, 1997a,b), includingflow from the lumenal electron carriers to oxygen, medi-ated by thylakoidal cytochrome oxidase (Manna and Ver-maas 1997). Earlier work has shown that changes in

Abbre6iations – Chl: Chlorophyll; DCMU: 3-(3,4-dichlorophenyl)-1,1-dimethylurea; Fm: maximal fluorescence; Fm% : maximal fluorescencemeasured under a light-acclimated state; Fo: minimal fluorescence; Fo% : minimal fluorescence measured under a light-acclimated state; Fs:steady-state fluorescence measured under actinic light; Fv=Fm−Fo; Fv%=Fm% −Fo% ; MOPS: 3-(n-morpholino) propanesulfonic acid; PAM:Pulse amplitude modulated Chl fluorometer; PC: phycocyanin; fPSII= (Fm% −Fs/Fm% −PSII: photosystem II; qp= (Fm% −Fs)/(Fm% −Fo% ).

Physiol. Plant. 105, 1999746

respiratory (Shyam et al. 1993) and cyclic electron flow(Herbert et al. 1995) are important factors in cyanobacte-rial tolerance and acclima tion to excess light. Withdrawalof electrons from the transport system can prevent over-reduction of redox pools, thereby avoiding feed-back inhi-bition of photosystem II.

In cyanobacteria, the physiological regulations of PSIIactivity through changes in electron flux and excitationdistribution (Mullineaux and Allen, 1986, Dominy andWilliams, 1987) are complemented by changes in gene ex-pression, which can rapidly alter the protein compositionand activity of PSII. Cyanobacteria, unlike plants, containsmall psbA gene families encoding the D1 protein of PSII(Curtis and Haselkorn 1984, Mulligan et al. 1984, Goldenet al. 1986, Jansson et al. 1987). In several species, themember psbA genes are differentially regulated accordingto environmental conditions. The strain Synechococcus sp.PCC 7942 contains three psbA genes (Golden et al. 1986).The psbAI gene is expressed during acclimated growth andencodes the constitutive form of D1 protein, D1:1,whilst psbAII and psbAIII encode an alternate form ofthe D1 protein, D1:2. Expression from psbAII/III isinduced when cells experience excitation stress andD1:2 transiently replaces D1:1 within the PSII reactioncenters (Schaefer and Golden, 1989, Bustos et al. 1990,Clarke et al. 1993a, Kulkarni and Golden, 1994, Clarkeet al. 1995; reviewed in Golden, 1995). The two formsof D1 protein confer functional differences on PSII(Krupa et al. 1991, Clarke et al. 1993b, Campbell etal. 1996), and the exchange between the alternate forms isan essential phase of acclimation to excitation stress inSynechococcus (Clarke et al. 1993a, 1995).

Our previous papers have demonstrated that the switchfrom psbAI to psbAII/III expression occurs under condi-tions which drive fractional PSII reduction or closure.This predictive correlation holds whether PSII closure isinduced by increased light (Clarke et al. 1995), de-creased temperature, or specific inhibitors of electron flow(Campbell et al. 1995). Furthermore, electron transportstatus is important for the regulation of various genes inother organisms (Allen, 1993, Campbell et al. 1993,Danon and Mayfield 1994, Allen et al. 1995, Maxwell etal. 1995).

Chlorophyll (Chl) fluorescence analysis allows non-invasive assessment of the activity and redox statusof PSII (reviewed in Campbell et al. 1998). We thereforesought to explore the contribution of oxygen-depen-dent electron consumption to maintaining PSII oxidizedand active in the model cyanobacterium Synechococcussp. PCC 7942. To inhibit all the potential paths for ele-ctron donation to oxygen, we removed oxygen fromthe culture media (Miller et al. 1991). We then moni-tored the effects of anoxia on PSII redox and activity,psbA transcripts and D1 content. Based on our ear-lier results we predicted that if the loss of oxygen asan electron sink drove photosystem II closure, anoxiawould trigger changes in psbA expression and D1 compo-sition.

Materials and methods

Culture material and growth conditions

Synechococcus sp. PCC 7942 was grown in BG-11 inorganicmedium (Rippka et al. 1979), supplementally buffered with10 mM 3-(n-morpholino) propanesulfonic acid (MOPS), pH7.5. Cultures of 300 ml were grown in flat flasks, bubbledwith 5% CO2 in air (ca 1 ml s−1) at 37°C with continuous,even illumination of 50 mmol photons m−2 s−1 from incan-descent Philips PAR 38 Economy 120 W lamps. Chl andphycocyanin (PC) content were determined using whole cellspectra according to Myers et al. (1980). Cultures wereinoculated with liquid pre-culture to a concentration ofabout 0.5 mg Chl ml−1 and used for anoxia treatments whenthey reached about 2 mg Chl ml−1 after about 24 h, inexponential growth phase.

Anoxia treatments

Sodium bicarbonate solution (5 mM final concentration)was added to the culture flasks to provide a source of CO2.Gas bubbling was stopped 30 min later and 1 mg of glucoseoxidase (Boehringer), 1 mg of catalase (Sigma) and glucose(Boehringer, 10 mM final concentration, 0.18% w/v) wereadded to the flask (Miller et al. 1991). Before the flask wassealed, a sample was taken for the time 0 control photosyn-thetic measurements. The action of glucose oxidase con-sumes oxygen and the catalase destroys the resultinghydrogen peroxide. Successful removal of oxygen was ver-ified using the oxygen electrode; oxygen concentrations fellto undetectable levels within 5 min, remaining there for theduration of the experiment (4 h). Five replicate culturesgrown at 37°C were shifted to anoxia with the light heldconstant. Samples for photosynthetic analysis, pigment de-termination, RNA isolation and protein extraction weretaken 0, 0.5, 1, 2, and 4 h after the transfer. For theanoxia/recovery experiments, cultures were made anoxic byvigorous and continual bubbling with 5% CO2 in nitrogen(ca 1 l min−1). Anoxia was thereby maintained for 45 min,and then bubbling was switched to 5% CO2 in air, forrecovery under oxygenic conditions.

Photosynthetic measurements

Chl a fluorescence parameters and oxygen evolution weremeasured simultaneously using a system of cuvette, mag-netic stirrer, oxygen electrode, and actinic lamp (Hansatech,King’s Lynn, UK) (Walker, 1987), compatible with a Pulseamplitude modulated Chl fluorometer (PAM 101, Walz,Effeltrich, Germany), using techniques described in detail in(Clarke et al. 1995, Campbell et al. 1996, reviewed inCampbell et al. 1998).

Cells were harvested from the treatment flask and trans-ferred to the measurement cuvette, for an initial dark-incu-bation period of 3–5 min. The weak modulated measuringbeam of the PAM was then activated (intensity ca 0.1 mmolphotons m−2 s−1) and Fo fluorescence of the dark-accli-mated cells was recorded. A 1-s saturating white-light flashwas applied to measure maximal fluorescence measured

Physiol. Plant. 105, 1999 747

under a light-acclimated state (Fm% ) of the dark-acclimatedcyanobacterial cells, which typically show an Fm% consider-ably below maximal fluorescence (Fm) (Campbell et al.1998). Then a series of actinic light levels were applied to thesample using an incandescent source similar in quality to thegrowth light. At each actinic light level steady-state FS

fluorescence was measured, a saturating flash was applied tomeasure Fm% under each light level, and the actinic light wasbriefly interrupted to measure Fo% . After the measurementswere performed at the highest light level used, 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU) was injectedinto the cuvette (0.5 mmol l−1 final concentration) to closePSII centers and block electron transport. This drivesfluorescence to the maximal Fm level. Cells were not CO2

limited during measurements. Photochemical (qP) quenchingof PSII variable fluorescence was calculated according tovan Kooten and Snel (1990), qP= (Fm% −Fs)/(Fm% −Fo% ). Pho-tochemical (qP) quenching reflects the proportion of PSIIcenters open and available to perform photochemistry, on ascale from qP=0 when all centers are closed and Fs=Fm% , toqP=1 when all centers are open and Fs=Fo% . We alsocalculated a relative measure of the efficiency of excitationenergy capture by PSII reaction centers (Fv/Fm= (Fm−Fo)/Fm). Net oxygen evolution was measured for the oxygeniccontrol sample, simultaneously with the fluorescence mea-surements. Electron flux through PSII was estimated andexpressed on the basis of oxygen evolution using the fluores-cence parameter fPSII= (Fm% −Fs)/Fm% ) (Genty et al. 1989),which is a measure of the effective quantum yield of PSII inthe light, which is lowered by the combination of reactioncenter closure (measured by qP) and drops in energy captureefficiency (measured by Fv% /Fm% ). fPSII was multiplied by theincident light intensity and an empirical conversion factor of12 incident photons per O2 (Sundberg et al. 1996, Campbellet al. 1998). This estimator correlates well with net oxygenevolution at the growth light intensity, and provided ameans to follow approximate PSII electron flow underanoxic conditions. The applications of these fluorescenceparameters to cyanobacteria are discussed in detail in Camp-bell et al. (1998). In the plots of fluorescence parametersthere are two values at time 0, measured before and after theshift to anoxia.

Detection of psbA messages

Total RNA was isolated from Synechococcus sp. PCC 7942using a procedure developed from those of Glatron andRapoport (1972) and Mohamed and Jansson (1989), asdescribed previously (Campbell et al. 1995). RNA (3 mg) wasdenatured by glyoxylation (Sambrook et al. 1989), elec-trophoretically fractionated in 1% agarose gels in 10 mMNaPO4 (pH 7.0) and then transferred to Hybond-N mem-brane (Amersham, Buckinghamshire, UK). Transcripts ofall three psbA genes were then detected using DNA probesspecific for the unique 5%-untranslated regions of each gene(100 bp). The amount of hybridization to each psbA messagewas measured using a GS 250 Molecular Imager (Bio-Rad,Hercules, CA, USA) and accompanying PhosphoAnalystsoftware, and quantified relative to 16S RNA content asdescribed by Kulkarni et al. 1992.

D1 protein detection by immunoblotting

Total cellular proteins were extracted as described previ-ously (Clarke et al. 1993a). Protein samples containingequivalent amounts of Chl were loaded into lanes of linear15% lithium dodecyl sulphate-polyacrylamide gels (0.3–0.6mg per lane depending on the gel) and separated, thentransferred electrophoretically to Immobilon-P (0.2 mm poresize, Millipore, USA) followed by immunoblot analysis ac-cording to Clarke and Critchley (1992) with detection usingthe ECL chemiluminescent kit (Amersham, UK). The poly-clonal antibodies against D1:1 and D1:2 are completelyform-specific and show no detectable cross-reactivity, asshown by Clarke et al. (1993a). The specificities of thepolyclonal antibodies against total D1 polypeptide havebeen characterized by Tyystjarvi et al. (1994). Density scan-ning of films from the immunoblot exposures was carriedout using the ImageMaster Desktop Scanner and software(Pharmacia, LKB, Sweden) following the manufacturer’srecommendations.

Protein synthesis

To measure protein synthesis cells were grown as described,harvested by centrifugation and resuspended in BG-11 lack-ing added sulfate. Anoxia and control treatments wereinitiated as described except that small culture flasks wereused, and after 10 min, 5.92 MBq of carrier free 35SO4

(Amersham, UK) was added to each flask, containing 30 mlof culture. Samples for protein analysis were taken 30 and 60min after the addition of radioactivity (40 and 70 min afteranoxia was initiated). Protein samples containing 0.75 mgChl were separated by electrophoresis, as described. The gelwas fixed and soaked in Amplify solution (Amersham, UK)according to the manufacturer’s directions, dried under vac-uum, and exposed to autoradiography film. Radioactivity inthe gel was quantified using a GS 250 Molecular Imager(Bio-Rad, Hercules, CA, USA) with a high-sensitivityscreen, and the accompanying PhosphoAnalyst software.

Results

PSII suffers closure and slow inactivation under anoxia

Under control conditions (light intensity of 50 mmol photonsm−2 s−1, 37°C, bubbling with 5% CO2 in air) the replicatebatch cultures used in all experiments grew with an exponen-tial doubling time of 7.3 h90.4 h (Fig. 1), based onmeasures of Chl content calculated from whole cell spectra(Myers et al. 1980). Growth declined and halted after 1 h ofanoxia, and remained depressed throughout the 4 htreatment.

Fig. 2 presents key photosynthetic characteristics of thecultures under control conditions and immediately after theshift to anoxia, estimated from Chl fluorescence parameters.Fig. 2A shows simple light response curves of photochemicalquenching (qP), which is approximately equivalent to theproportion of open PSII centers. Under oxygenic conditions,the PSII centers remain largely open even under excessiveexcitation four times higher than the growth light

Physiol. Plant. 105, 1999748

intensity. Under anoxia, however, the cultures suffered sig-nificant PSII closure even at the growth light intensity.This indicates that a proportion of electrons from PSIIare normally passed back to oxygen, and if this sink isremoved, a feedback of electrons results in partial closureof the PSII population. The gap between oxygenic andanoxic PSII closure became much wider under higher irra-diance, with the anoxic cells suffering severe PSII closure.This shows the flux of electrons to oxygen is highly flex-ible and responds almost instantaneously to increasinglight, since the light response curves were measuredrapidly by increasing the irradiance of a culture sample.

Fig. 2B shows the estimated electron flow through PSIIexpressed on an oxygen basis (equivalent to gross oxygenevolution). The dashed line shows the actual measured netoxygen evolution under oxygenic conditions. Under oxy-genic conditions the estimated PSII electron flow progres-sively exceeded the measured net oxygen evolution as thelight intensity increased; at four times the growth lightintensity, estimated PSII electron flow is double the netoxygen evolution. In marked contrast, estimated PSII elec-tron flow under anoxia closely corresponds to the net oxy-gen evolution measured under oxygenic conditions, afterallowing for a downward offset of oxygen evolutioncaused by dark respiration. Thus the discrepancy betweenmeasured oxygen evolution and estimated PSII electronflow under oxygenic conditions is dependent on the pres-ence of oxygen, and disappears if oxygen is removed.

Fig. 3 follows PSII status over time after the shift toanoxia, with the measurements performed under the treat-ment conditions of 50 mmol photons m−2 s−1 and 37°C.As shown in Fig. 3A, upon the shift to anoxia the pro-portion of open PSII centers immediately dropped, andthen continued to fall for 1 h after the shift. Fv/Fm

Fig. 2. Light Response Curves of (A) the proportion of open PSIIcenters (qP) and (B) predicted gross oxygen evolution under control(�) and anoxic conditions (), predicted from the Chl fluorescenceparameter fPSII as described in Materials and methods. Dashedline shows actual measured net oxygen evolution under controlconditions. n=59SE.

Fig. 1. Growth of a representative Synechococcus sp. PCC 7942culture under control and anoxic conditions. Cultures were trans-ferred to anoxia upon reaching 2 mg Chl ml−1, in mid-exponentialphase.

reflects the relative efficiency of excitation energy captureby PSII reaction centers. This index of PSII activity wasstable for 1 h after the shift to anoxia, but then declinedto a very low level by 4 h after the shift, largely becauseof declining Fm. The increase in qP from 2 h onwardsmay reflect re-opening of the residual PSII centers, but theqP measure may not be reliable when Fv/Fm reaches lowlevels. Fig. 3B plots the estimated gross oxygen evolutionduring the anoxia treatment, predicted from the fluores-cence parameter fPSII, as described in the materials andmethods. fPSII combines the proportion of PSII centersopen (qP, Fig. 3A) with the efficiency of quantum excita-tion capture by these open centers (Fv% /Fm% ) (Genty et al.1989). The estimated gross oxygen evolution thereforeshows an initial drop up to 1 h reflecting PSII closure,followed by further decline reflecting the later loss of PSIIactivity.

Rapid changes in psbA gene expression under anoxia

Under our conditions of acclimated growth, psbAI mRNAwas the predominant psbA species, consistent with previ-ous observations (Golden et al. 1986, Campbell et al.

Physiol. Plant. 105, 1999 749

1995). Within 30 min of anoxia, psbAI message dropped totrace levels (Fig. 4A), where it remained for the duration ofthe treatment (4 h). In contrast, anoxia induced a largeaccumulation of psbAII and AIII message within 15 min(Fig. 4B,C). Although some of the induced psbAII and AIIImessage was detected as the 1.3 kb full-length transcripts,most was present as a shorter 0.22 kb 5%-fragment. Weearlier interpreted this fragment as a cleavage product gener-ated from the full-length psbAII/II transcripts (Campbell etal. 1995), as has been shown to occur for psbAI (Golden etal. 1986). More recently, Soitamo et al. (1998) characterizedthis small fragment as resulting from specific pre-maturetermination of transcription, which occurs particularly whentranslation is slow. A somewhat larger 0.32 kb 5%-fragmentdoes result from cleavage of the psbAI transcript and isvisible in the 15 min sample in Fig. 4A. This 0.32 kbcleavage fragment was not detected from the psbAII orpsbAIII transcripts (Fig. 4B,C Golden et al. 1986, Soitamoet al. 1998). After 15 min of anoxia, the level of maturepsbAII/III transcripts gradually declined whereas that of thesmaller termination product remained consistently high(Fig. 4).

Fig. 4. psbA transcripts under control (�) and anoxia treatment.RNA-DNA hybridizations are representative of two separate exper-iments. (A) psbAI ; (B) psbAII ; and (C) psbAIII.

Fig. 3. Photosynthetic status during anoxia treatment. (A) Propor-tion of open PSII centers (qP, �) and the relative efficiency ofphotochemical excitation capture of PSII (Fv/Fm �) under treat-ment conditions. The qP value is high at the 4 h point, but thecalculation may not be valid at the very low Fv/Fm value at thispoint. (B) Gross oxygen evolution predicted from fPSII as de-scribed in Materials and methods. n=49SE.

D1:1 protein is initially stable under anoxia and D1:2 doesnot accumulate

Fig. 5 shows the amount and composition of D1 proteinunder control and anoxia treatments. For the first 60 min ofanoxia there is no detectable change in D1:1 content, butsubsequently D1:1 is lost (Fig. 5A). This pattern closelyparallels the initial stability and subsequent drop in Fv/Fm

under anoxia (Fig. 3). Under oxygenic conditions the half-life of D1:1 under equivalent light and temperature (50 mmolphotons m−2 s−1, 37°C) is only 40 min (Clarke et al.1993b). In contrast, under anoxia little or no synthesis ofD1:1 occurs because the psbAI message pool is depletedwithin 30 min (Fig. 4A). The existing D1:1, therefore, isstable for at least 60 min but is then lost with a half-life ofapproximately 25 min. The Fv/Fm measures show that notonly is the D1:1 protein maintained stable in the cellsbeyond the control half-life period, but that it also remainsfunctional until the delayed degradation occurs.

Physiol. Plant. 105, 1999750

Although expression of the psbAII/AIII messages is in-duced, no accumulation of D1:2 occurs under anoxia, withthe trace levels present in the control sample declining in apattern similar to that of D1:1 (Fig. 5B). The total pool ofD1 protein, detected with an antibody recognizing bothD1:1 and D1:2 (Fig. 5C), closely reflects the pool of D1:1protein alone.

Protein synthesis continues but at a reduced rate underanoxia

Fig. 6 presents the incorporation of radioactive 35S intoprotein, supplied to cells as 35Sulfate. Under control condi-tions a large number of proteins are labeled (Fig. 6A) andthe total sum of radioactive incorporation into all proteinbands is nearly linear over a 1 h labeling period (Fig. 6B). Incells in which labeling commenced 10 min after the initiationof anoxia, some protein synthesis continued over the course

Fig. 6. Synthesis of protein under control and anoxia treatment.Newly synthesized proteins were radiolabeled by exposing cells to35SO4. (A) Fluorograph of radiolabeled proteins separated by poly-acrylamide gel electrophoresis. For clarity of presentation theanoxia lanes are a longer exposure than the control lanes. (B)Accumulation of radioactivity in proteins synthesized under control(�) and anoxia treatments (). The total radioactivity from allprotein bands for each lane was determined using a Bio-Rad GS250 Molecular Imager.

Fig. 5. D1 protein content under control (0) and anoxia treatment.Immunoblots are representative of two separate experiments. (A)D1:1; (B) D1:2; and (C) total D1 (D1:1+D1:2).

of 1 h, but the specific pattern of labeling is significantlydifferent from the control (Fig. 6A). Labeling of manybands is greatly decreased, while a few novel bands appearrelatively strongly. Furthermore, the rate of total incorpora-tion of label under anoxia is only about 9% of the control(Fig. 6B).

D1:1 and PSII activity are transiently lost upon aerobicrecovery from anoxia

Fig. 7 presents PSII activity and D1:1 protein during ashorter 45 min anoxic incubation (−45 to 0 min), followedby a 60 min recovery period (0 to 60 min) under aerobicconditions. Anoxia was maintained for only 45 min to avoid

Physiol. Plant. 105, 1999 751

the loss of D1 protein and drop in Fv/Fm previously ob-served under more prolonged anoxia (Fig. 5). As expected,PSII activity measured as Fv/Fm was stable during the shortanoxia treatment (Fig. 7), as was the level of D1:1 protein(cf Fig. 5). Upon a return to oxygenic conditions, however,PSII activity transiently dropped within the first 15 min, inparallel with a transient loss of D1:1 protein (Fig. 7).Afterwards, D1:1 content recovered rapidly, as did PSIIactivity. No accumulation of D1:2 occurred during thisrecovery period (data not shown), in spite of the transientloss of PSII function.

Discussion

Electron flow from PSII to oxygen in Synechococcus

PSII centers suffer significant closure upon removal of oxy-gen from a Synechococcus culture, while light, temperatureand other conditions are held constant. This reflects a dropin electron consumption once electron transfer to oxygen isblocked, and demonstrates a significant flow of electronsfrom PSII to oxygen under steady-state growth. Under thegrowth light intensity of 50 mmol photons m−2 s−1, re-moval of oxygen instantaneously drives the approximatefraction of open PSII centers from 0.89 down to 0.77,indicating that under acclimated growth, approximately13% of electrons leaving PSII are ultimately transferred tooxygen. For the same cells measured under excess light, theproportion of electrons flowing from PSII to oxygen ismuch higher, reaching about 47% for cells measured under200 mmol photons m−2 s−1. Electron consumption byoxygen is therefore an important sink for removal of elec-

trons from PSII, and can largely account for the surprisingcapacity of many cyanobacteria to maintain PSII centersopen in the face of excess excitation (Miller et al. 1991,Clarke et al. 1993b, Luttge et al. 1995, Campbell, 1996,Campbell et al. 1996). Importantly, the extent of this fluxresponds quickly to the level of excitation pressure on PSII,increasing progressively as light exceeds the acclimated level(Shyam et al. 1993). Our data provide no indication of theprecise paths connecting PSII and oxygen, but work byothers demonstrates several potential contributing mecha-nisms such as cytochrome oxidase activity or the Mehler-Ascorbate Peroxidase cycle reactions (Peschek et al. 1982,Alpes et al. 1984, Binder et al. 1984, Scherer et al. 1988,Miller et al. 1991, Badger and Schreiber 1993, Shyam et al.1993, Geerts et al. 1994, Schreiber et al. 1995, Schubert et al.1995, Manna and Vermaas, 1997).

In rye plants, a 50% inhibition of mitochondrial respira-tion caused some PSII closure, but the effect was fairlyconstant at light intensities above the growth level, reaching13–20% under saturating light (Hurry et al. 1995). Thesedata suggest that although respiratory electron flow con-tributes to excitation tolerance in plants, the response hasless instantaneous flexibility than the rapid changes in respi-ration found in cyanobacteria.

fPSII is a fluorescence measure equivalent to the productof the quantum excitation capture efficiency (Fv% /Fm% ) and theproportion of open PSII centers (qP). fPSII thus measuresthe effective quantum yield of electron transport throughPSII, and can be multiplied by the incident light intensity toestimate electron flow through PSII, expressed on the basisof gross oxygen evolution using an empirical conversionfactor. For cyanobacterial cells under aerobic conditions,this estimated PSII electron flow accords well with directmeasures of oxygen evolution under irradiances up to theacclimated growth level (Sundberg et al. 1996, Campbell etal. 1998). Under higher light the estimated electron fluxthrough PSII progressively exceeds the measured oxygenevolution. Our data indicate that the discrepancy reflectsflow of electrons away from PSII and ultimately to oxygen,which results in intracellular consumption of oxygen. There-fore under aerobic conditions the measured oxygen evolu-tion underestimates the actual production of oxygen byPSII, and electron flux through PSII exceeds the detectedevolution of oxygen. When oxygen is removed from themedia, the estimated electron flux through PSII slows, andmatches well with the net oxygen evolution measured incontrol cells.

Anoxia triggers rapid changes in psbA mRNA composition.

We have demonstrated using other treatments that condi-tions causing fractional closure of PSII centers result in arapid change from psbAI to psbAII/AIII expression (Camp-bell et al., 1995, Clarke et al., 1995). These earlier experi-ments shared the common element of driving PSII closure,but also caused other changes in PSII status, either throughphotoinhibitory inactivation under excess light or low tem-perature, or direct chemical inhibition of PSII activity. Incontrast, removal of oxygen under moderate light has no

Fig. 7. PSII activity and D1:1 protein are transiently lost duringaerobic recovery from anoxia. The plot presents Fv/Fm as an indexof total PSII activity versus time, under anoxia (−45 to 0) andduring recovery under aerobic conditions (0 to 60 min). The insetshows an immunoblot detection of D1:1 protein with equivalentloads (mg Chl basis) in each lane after 45 min of anoxia, followed byprotein prepared from cells after 15, 30, 45, and 60 min of aerobicrecovery. Data is representative of two separate experiments.

Physiol. Plant. 105, 1999752

immediate influence on PSII activity, as indicated by theabsence of photoinhibition, and the maintenance of D1:1protein in the thylakoid membranes. Nevertheless, the lossof oxygen as an electron sink does cause fractional closureof PSII, which as predicted results in immediate suppressionof psbAI expression and activation of psbAII/AIII.

These rapid changes in expression do not result in a largeaccumulation of mature psbAII/III transcripts, as occursunder excess light (Kulkarni and Golden, 1994) or decreasedtemperature (Campbell et al. 1995). Some full-length tran-scripts do accumulate in the initial period of anoxia, butthese appear to be degraded as the treatment progresses.

There is a large and sustained accumulation of short 5%fragments of all three gene transcripts. We earlier inter-preted these fragments as cleavage products derived fromthe mature messages (Campbell et al. 1995). More recently,Soitamo et al. (1998) have shown that 5% 0.22 kb fragmentsresult from premature termination of transcription of allthree psbA genes. They found this premature termination tooccur primarily under conditions when translation is slow,as is the case here under anoxia. Soitamo et al. (1998) alsoobserved a 0.32 kb fragment, probably derived from cleav-age of the psbAI transcript (Golden et al. 1986) but not fromthe psbAII and psbAIII transcripts. Consistent with theirresults, in Fig. 4A probed with 5% region of psbAI, a weakband of about 0.3 kb is present. This band is not detectedon the psbAII and psbAIII blots (Fig. 4B,C).

Others have observed fast rates of psbAII/III turnoverduring shifts to high light (Kulkarni and Golden 1994),indicating these transcripts are unstable upon induction, andrequire sustained gene expression to accumulate. Therefore,although fractional PSII closure is sufficient to repress psbAIand initiate psbAII/AIII gene expression, the transcriptionalactivation is not sufficient to maintain full-length psbAII/AIII messages during anoxia. This supports the hypothesisthat regulation of psbA mRNA stability and prematuretranscriptional termination is linked to translation and D1turnover (Aro et al. 1993, Soitamo et al. 1998), both ofwhich processes are inhibited under anoxia.

PSII closure alone is not sufficient to drive exchange ofD1:1 and D1:2.

The complete switch in expression from psbAI to psbAII/IIIunder anoxia and moderate light does not result in anexchange of D1:1 and D1:2 proteins in PSII centers. Instead,the existing D1:1 proteins remain active and stable in thePSII centers for well beyond their normal half-life of 40 min,as measured under comparable control conditions (Clarke etal. 1993a). In isolated spinach thylakoids incubated understrong light, anoxia caused a similar stabilization of D1protein, although under those conditions PSII electrontransport was inhibited (Hundal et al. 1990, Vass et al.1992). Ultimately, after 1 h of anoxia treatment of Syne-chococcus, PSII activity starts to decline, and the D1:1protein is progressively lost from the thylakoids. There is nodetectable accumulation of D1:2, even though some transla-tional activity is maintained under anoxia, and some full-length psbAII/III transcripts are present. These data further

indicate that accumulation of D1:2 protein and psbA mes-sage stability respond to factors separate from the transcrip-tional regulation triggered in response to PSII closure. Onelikely regulatory signal could be the loss of D1:1 from PSIIcenters. In general, PSII closure is followed by photoin-hibitory inactivation and loss of D1:1 protein, so that thetranscriptional activation remains coupled to D1:1/D1:2exchange. Under the special case of anoxia, however, PSIIcenters close but inactivation and clearance of D1:1 proteinsis slowed, so that the transcriptional changes are not suffi-cient in themselves to result in exchange of the proteins inthe reaction centers.

D1:1 damaged during anoxia is cleared using oxygen

Although D1:1 is somewhat stabilized under anoxia, it isultimately lost from the PSII centers. After cells are exposedto anoxia for a short period, a return to aerobic conditionsresults in loss of D1:1 protein from the membranes and atransient inhibition of PSII activity, even though the PSIIcenters rapidly re-open upon addition of oxygen. AlthoughPSII activity suffers inhibition and D1:1 protein contentdrops, no accumulation of D1:2 occurs in these cells. Thisexperiment demonstrates the complement of the above sec-tion; that is, loss of PSII activity is not alone sufficient todrive exchange between D1 forms, if the remaining PSIIcenters are maintained in an open state and psbAII/AIIItranscription is not triggered.

Under normal growth conditions, oxygen is used in apreemptive process to clear D1:1 proteins from thylakoidmembranes, before they suffer terminal inactivation. Underanoxia, the D1:1 proteins are stabilized (Aro et al. 1993) andmaintain activity beyond their normal half-life, but continueto accumulate some form of damage (Hundal et al. 1990,Vass et al. 1992). Upon the return of oxygen, these D1:1proteins are rapidly cleared from the membranes, even atthe cost of a transient drop in PSII activity.

In cyanobacteria, the presence of oxygen is a key factorfor coping with the changing balance between excitation andenergy consumption. For instantaneous response to excessexcitation, oxygen serves as an important and highly flexiblesink for electrons originating from PSII. This intersection ofrespiratory and photosynthetic electron flow largely ac-counts for the strong capacity of cyanobacteria to maintainPSII centers open under excess excitation. If the electrontransport systems prove insufficient to maintain PSII centersopen, transcriptional regulation is triggered to stop produc-tion of psbAI message and induce the alternate psbAII/AIIImessages. As well as acting as an electron sink, oxygen isalso essential for the orderly turnover and replacement ofD1 proteins (Aro et al. 1993). In cyanobacteria exchangingthe alternate forms of D1 is a key means of maintaining andadjusting PSII activity under changing conditions (O8 quist etal. 1995).

Acknowledgements – This work was supported by operating grantsto G. O8 quist and a post-doctoral fellowship to D. Campbell fromthe Swedish Natural Science Research Council, and by an operatinggrant to D. Campbell from the Natural Sciences and EngineeringResearch Council of Canada. We thank our colleagues Dr AmandaCockshutt, Prof Anthony Miller, Dr Henrik Schubert, Dr Kristin

Physiol. Plant. 105, 1999 753

Palmqvist and Dr Bodil Sundberg for valuable discussions andsuggestions.

ReferencesAllen CA, Hakansson G, Allen JF (1995) Redox conditions specify

the proteins synthesised by isolated chloroplasts and mitochon-dria. Redox Rep 1: 119–123

Allen JF (1993) Redox control of transcription: sensors, responseregulators, activators and repressors. FEBS Lett 323: 203–207

Alpes I, Sturzl E, Scherer S, Boger P (1984) Interaction of photo-synthetic and respiratory electron transport: Effect of a cy-tochrome c-553 specific antibody. Z Naturforsch 360: 623–627

Aro EM, Virgin I, Andersson B (1993) Photoinhibition of photosys-tem II: Inactivation, protein damage and turnover. BiochimBiophys Acta 1143: 113–134

Badger MR, Schreiber U (1993) Effects of inorganic carbon accu-mulation on photosynthetic oxygen reduction and cyclic elec-tron flow in the cyanobacterium Synechococcus PCC 7942.Photosynth Res 37: 177–191

Binder A, Hauser R, Krogmann D (1984) Respiration in energytransducing membranes of the thermophilic cyanobacteriumMastigocladus laminosus. I: Relation of the respiratory andphotosynthetic electron transport. Biochim Biophys Acta 765:241–246

Bustos SA, Schaefer MR, Golden SS (1990) Different and rapidresponses of four cyanobacterial psbA transcripts to changes inlight intensity. J Bacteriol 172: 1998–2004

Campbell D (1996) Complementary chromatic adaptation altersphotosynthetic strategies in the cyanobacterium Calothrix. Mi-crobiol 142: 1255–1263

Campbell D, Bruce D, Carpenter C, Gustafsson P, O8 quist G (1996)Two forms of the photosystem II D1 protein alter energydissipation, state transitions in the cyanobacterium Synechococ-cus sp. PCC 7942. Photosynth Res 47: 131–144

Campbell DA, Houmard J, Tandeau de Marsac N (1993) Electrontransport regulates cellular differentiation in the filamentouscyanobacteria Calothrix. Plant Cell 5: 451–463

Campbell D, Hurry VM, Clarke AK, Gustafsson P, O8 quist G(1998) Chlorophyll fluorescence analysis of cyanobacterial pho-tosynthesis and acclimation. Microbiol Molec Biol Rev 62:667–683

Campbell D, Zhou G, Gustafsson P, O8 quist G, Clarke AK (1995)Electron transport regulates exchange of two forms of Photosys-tem II D1 protein in the cyanobacterium Synechococcus. EMBOJ 14: 5457–5466

Clarke AK, Critchley C (1992) The identification of a heat-shockprotein complex in chloroplasts of barley leaves. Plant Physiol100: 2081–2089

Clarke AK, Campbell D, Gustafsson P, O8 quist G (1995) Dynamicresponses of photosystem II and phycobilisomes to changinglight in the cyanobacterium Synechococcus sp. PCC 7942. Planta197: 553–562

Clarke AK, Soitamo A, Gustafsson P, O8 quist G (1993a) Rapidinterchange between two distinct forms of cyanobacterial photo-system II reaction-center protein D1 in response to photoinhibi-tion. Proc Natl Acad Sci USA 90: 9973–9977

Clarke AK, Hurry VM, Gustafsson P, O8 quist G (1993b) Twofunctionally distinct forms of the photosystem II reaction-centerprotein D1 in the cyanobacterium. Proc Natl Acad Sci USA 90:11985–11989

Curtis SE, Haselkorn R (1984) Isolation, sequence and expressionof two members of the 32-kD thylakoid membrane protein genefamily from the cyanobacterium Anabaena 7120. Plant Mol Biol3: 249–255

Danon A, Mayfield SP (1994) Light-regulated translation of chloro-plast messenger RNAs through redox control. Science 226:1717–1719

Dominy PJ, Williams WP (1987) The role of respiratory electronflow in the control of excitation energy distribution in blue-green algae. Biochim Biophys Acta 892: 264–274

Geerts D, Schubert H, de Vrieze G, Borrias M, Matthijs HCP,Weisbeek PJ (1994) Expression of Anabaena PCC 7937 plasto-cyanin in Synechococcus PCC 7942 enhances photosyntheticelectron transfer and alters the electron distribution between

photosystem I and cytochrome-c oxidase. J Biol Chem 269:28068–28075

Genty B, Briantais JM, Baker NR (1989) The relationship be-tween the quantum yield of photosynthetic electron transportand quenching of chlorophyll fluorescence. Biochim BiophysActa 990: 87–92

Glatron MF, Rapoport G (1972) Biosynthesis of the parasporalinclusion of Bacillus thuringiensis : Half-life of its correspondingmessenger RNA. Biochimie 54: 1291–1301

Golden SS (1995) Light-responsive gene expression in cyanobacte-ria. J Bacteriol 177: 1651–1654

Golden SS, Brusslan J, Haselkorn R (1986) Expression of a familyof psbA genes encoding a photosystem II polypeptide in thecyanobacterium Anacystis nidulans R2. EMBO J 5: 2789–2798

Herbert SK, Martin RE, Fork DC (1995) Light adaptation ofcyclic electron-transport through Photosystem-I in thecyanobacterium Synechococcus sp. PCC 7942. Photosynth Res46: 277–285

Hundal T, Aro EM, Carlberg I, Andersson B (1990) Restorationof light induced photosystem II inhibition without de novoprotein synthesis. FEBS Lett 267: 203–206

Hurry V, Tobiæson M, Kromer S, Gardestrom P, O8 quist G(1995) Mitochondria contribute to increased photosynthetic ca-pacity of leaves of winter rye (Secale secale L.) following cold-hardening. Plant Cell Env 18: 69–76

Jansson C, Debus RJ, Osiewacz HD, Gurevitz M, McIntosh L(1987) Construction of an obligate photoheterotrophic mutantof the cyanobacterium Synechocystis 6803. Plant Physiol 85:1021–1025

Krupa Z, O8 quist G, Gustafsson P (1991) Photoinhibition andrecovery of photosynthesis in psbA gene-inactivated strains ofcyanobacterium Anacystis nidulans. Physiol Plant 82: 1–8

Kulkarni RD, Golden SS (1994) Adaptation to high light intensityin Synechococcus sp. Strain PCC 7942: Regulation of threepsbA genes and two forms of the D1 protein. J Bacteriol 176:959–965

Kulkarni RD, Schaefer MR, Golden SS (1992) Transcriptionaland posttranscriptional components of psbA response to highlight intensity in Synechococcus sp. strain PCC 7942. J Bacte-riol 174: 3775–3781

Li QL, Canvin DT (1997a) Effect of the intracellular inorganicpool on chlorophyll-a fluorescence quenching and O2 photore-duction in air-grown cells of the cyanobacterium Synechococ-cus UTEX-625. Can J Bot 75: 946–954

Li QL, Canvin DT (1997b) Oxygen photoreduction and its effecton CO2 accumulation and assimilation in air-grown cells ofSynechococcus UTEX-625. Can J Bot 75: 274–283

Li QL, Canvin DT (1997c) Inorganic carbon accumulation stimu-lates linear electron flow to artificial electron-acceptors of Pho-tosystem-I in air-grown cells of the cyanobacteriumSynechococcus UTEX-625. Plant Physiol 114: 1273–1281

Luttge U, Budel B, Ball E, Strube F, Weber P (1995) Photosyn-thesis of terrestrial cyanobacteria under light and desiccationstress as expressed by chlorophyll fluorescence and gas ex-change. J Exp Bot 46: 309–319

Manna P, Vermaas W (1997) Lumenal proteins involved in res-piratory electron-transport in the cyanobacterium Synechocys-tis sp. PCC 6803. Plant Mol Biol 35: 407–416

Mano J, Miyake C, Schreiber U, Asada K (1995) Photoactivationof the electron flow from NADPH to plastoquinone inspinach-chloroplasts. Plant Cell Physiol 36: 1589–1598

Maxwell DP, Falk S, Huner NPA (1995) Photosystem II excita-tion and development of resistance to photoinhibition. PlantPhysiol 107: 687–694

Mi HI, Endo T, Ogawa T, Asada K (1995) Thylakoid membrane-bound, NADPH-specific pyridine-nucleotide dehydrogenasecomplex mediates cyclic electron-transport in the cyanobac-terium Synechocystis sp pcc-68038. Plant Cell Physiol 36: 661–668

Miller AG, Espie GS, Canvin DT (1991) The effects of inorganiccarbon and oxygen on fluorescence in the cyanobacteriumSynechococcus UTEX 625. Can J Bot 69: 1151–1160

Mir NA, Salon C, Canvin DT (1995) Photosynthetic nitrite reduc-tion as influenced by the internal inorganic carbon pool inair-grown cells of Synechococcus UTEX-625. Plant Physiol108: 313–318

Mohamed A, Jansson C (1989) Influence of light on accumula-tion of photosynthesis-specific transcripts in the cyanobac-terium Synechocystis 6803. Plant Mol Biol 13: 693–700

Physiol. Plant. 105, 1999754

Mulligan B, Schultes N, Chen L, Bogorad L (1984) Nucleotidesequence of a multiple-copy gene for the B protein of photosys-tem II of a cyanobacterium. Proc Natl Acad Sci USA 81:2693–2697

Mullineaux CW, Allen JF (1986) The state 2 transition in thecyanobacterium Synechococcus 6301 can be driven by respira-tory electron flow into the plastoquinone pool. FEBS Lett 205:155–160

Myers J, Graham JR, Wang RT (1980) Light harvesting in Anacys-tis nidulans studied in pigment mutants. Plant Physiol 66: 1144–1149

O8 quist G, Campbell D, Clarke AK, Gustafsson P (1995) Thecyanobacterium Synechococcus modulates photosystem II func-tion in response to excitation stress through D1 exchange.Photosynth Res 46: 151–158

Peschek GA, Schmetterer G (1982) Evidence for plastoquinal-cy-tochrome f/b-563 reductase as a common electron donor toP-700 and cytochrome oxidase in cyanobacteria. Biochem Bio-phys Res Commun 108: 1188–1195

Rippka R, Deruelles J, Waterbury JB, Herdman M, Stanier RY(1979) Generic assignments, strain histories and properties ofpure cultures of cyanobacteria. J Gen Microbiol 111: 1–61

Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: Alaboratory manual. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, NY

Schaefer MR, Golden SS (1989) Light availability influences theratio of two forms of D1 in cyanobacterial thylakoids. J BiolChem 264: 7412–7417

Scherer S, Almon H, Boger P (1988) Interaction of photosynthesis,respiration and nitrogen fixation in cyanobacteria. PhotosynthRes 15: 95–114

Schreiber U, Endo T, Mi H, Asada K (1995) Quenching analysis ofchlorophyll fluorescence by the saturation pulse method: Partic-ular aspects relating to the study of eukaryotic algae andcyanobacteria. Plant Cell Physiol 36: 873–882

Schubert H, Matthijs H, Mur L (1995) In vivo assay of P700 redoxchanges in the cyanobacterium Fremyella diplosiphon and therole of cytochrome-c-oxidase in regulation of photosyntheticelectron transfer. Photosynthetica 31: 517–527

Shyam R, Raghavendra AS, Sane PV (1993) Role of dark respira-tion in photoinhibition of photosynthesis and its reactivation inthe cyanobacterium Anacystis nidulans. Physiol Plant 88: 446–452

Soitamo AJ, Sippola K, Aro EM (1998) Expression of psbA genesproduces prominent 5% psbA mRNA fragments in Synechococcussp. PCC 7942. Plant Mol Biol 37: 1023–1033

Sundberg B, Campbell D, Palmqvist K (1996) Predicting CO2 gainand photosynthetic light acclimation from measurements offluorescence yield and quenching in cyanolichens. Planta 201:13–14

Tanaka Y, Katada S, Ishikawa H, Ogawa T, Takabe T (1997)Electron flow from NAD(P)H dehydrogenase to Photosystem-Iis required for adaptation to salt shock in the cyanobacterium-Synechocystis sp. PCC 6803. Plant Cell Physiol 38: 1311–1318

Tyystjarvi E, Kettunen R, Aro EM (1994) The rate constant ofphotoinhibition in vitro is independent of the antenna size ofphotosystem II but depends on temperature. Biochim BiophysActa 1186: 177–185

van Kooten O, Snel JFH (1990) The use of chlorophyll fluorescencenomenclature in plant stress physiology. Photosynth Res 25:147–150

Vass I, Styring S, Hundal T, Koivuniemi A, Aro EM, Andersson B(1992) Reversible and irreversible intermediates during photoin-hibition of photosystem II: Stable reduced QA species promotechlorophyll triplet formation. Proc Natl Acad Sci USA 89:1408–1412

Walker D (1987) The use of the oxygen electrode and fluorescenceprobes in simple measurements of photosynthesis. HansatechInstruments Limited, Paxman Road, Hardwick Industrial Es-tate, Kings Lynn, UK

Edited by B. Andersson

Physiol. Plant. 105, 1999 755