Phowsynthesis Research 48: 147-162, 1996. (~) 1996 Kluwer Academic Publishers. Printed in the Netherlands.

Regular paper

Chimaeric CP47 mutants of the cyanobacterium Synechocystis sp. PCC 6803 carrying spinach sequences: Construction and function Wim F.J. Vermaas l, Gaozhong Shen 1'3 & Itzhak Ohad2 IDepartment of Botany, and Center for the Study of Early Events in Photosynthesis, Arizona State University, Tempe AZ 85287-1601, USA; 2Department of Biological Chemistry, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel; 3Present address: Department of Molecular and Cell Biology, Pennsylvania State University, University Park PA 16802, USA Received 24 October 1995; accepted in revised form 6 December 1995

Key words: gene replacement, Photosystem II, photosynthesis, thylakoid membranes, transformation

Abstract

Chimaeric mutants of the cyanobacterium Synechocystis sp. PCC 6803 have been generated carrying part or all of the spinach psbB gene, encoding CP47 (one of the chlorophyll-binding core antenna proteins in Photosystem II). The mutant in which the entire psbB gene had been replaced by the homologous gene from spinach was an obligate photoheterotroph and lacked Photosystem II complexes in its thylakoid membranes. However, this strain could be transformed with plasmids carrying selected regions of SynechocystispsbB to give rise to photoautotrophs with a chimaeric spinach/cyanobacterial CP47 protein. This process involved heterologous recombination in the cyanobacterium between psbB sequences from spinach and Synechocystis 6803; which was found to be reasonably effective in Synechocystis. Also other approaches were used that can produce a broad spectrum of chimaeric mutants in a single experiment. Functional characterization of the chimaeric photoautotrophic mutants indicated that if a decrease in the photoautotrophic growth rates was observed, this was correlated with a decrease in the number of Photosystem II reaction centers (on a chlorophyll basis) in the thylakoid membrane and with a decrease in oxygen evolution rates. Remaining Photosystem II reaction centers in these chimaeric mutants appeared to function rather normally, but thermoluminescence and chlorophyll a fluorescence measurements provided evidence for a destabilization of Q~. This illustrates the sensitivity of the functional properties of the PS II reaction center to mild perturbations in a neighboring protein.

Abbreviations: diuron- 3-(3,4-dichlorophenyl)- 1,1-dimethylurea;

F v - variable chlorophyll a fluorescence; HEPES-N-(2-hydroxyethyl)piperazine-N'-(2-ethanesulfonic acid); (k)bp-(kilo)base pairs; PS II-Photosystem II; QA-primary electron-accepting plastoquinone in Photosystem II; QB- secondary electron-accepting plastoquinone in Photosystem II; SDS - sodium dodecyl sulfate Introduction

Photosystem II (PS II) is the pigment-protein complex in thylakoid membranes that catalyzes the lightinduced reduction of plastoquinone by water. Light is absorbed by antenna pigments (such as chlorophylls a and b in plants, and chlorophyll a and phycobilins in cyanobacteria), and light energy is transferred to P680, the primary donor in the PS II reaction center (reviewed

by Nixon et al. 1992; Barber 1992; Vermaas 1993). Two chlorophyll-binding antenna proteins, CP43 and CP47 (approximately 43 000 and 47 000 Mr, respectively), are associated directly with the D1 and D2 proteins that form the PS II reaction center (reviewed by Bricker 1990). Of these two chlorophyll-binding antenna proteins, CP47 appears to be closer to the reaction center complex as PS II reaction centers retaining CP47 but lacking CP43 can be isolated (Ghanotakis et

148 al. 1989; Dekker et al. 1989). The relatively peripheral arrangement of CP43 in PS II complexes is in agreement with ultrastructural observations (Boekema et al. 1994, 1995). CP47 is also important for stable assembly of the PS II complex: In mutants in which the psbB gene (encoding CP47) has been interrupted or deleted, D1 and D2 are virtually absent from the thylakoid membrane (Vermaas et al. 1988). Structurally, CP47 is quite conserved throughout evolution: The amino acid sequence of CP47 from Synechocystis sp. PCC 6803 is 76% identical compared to CP47 from spinach (Vermaas et al. 1987). However, functionally the constraints on the chlorophyll-binding proteins in the PS II core complex may be different in cyanobacteria as compared to that in plants. In the first place, the peripheral antenna complex is different (phycobilisomes in cyanobacteria and the integral chlorophyll-binding light-harvesting complex LHC II in plants; see Gantt (1988) and Thornber et al. (1988)), and thus recognition between the core (CP47/CP43) and peripheral antenna may be different in the two groups of organisms. Secondly, CP47 is involved in binding extrinsic proteins associated with the water-splitting apparatus, particularly the '33 kDa' manganese-stabilizing protein (Bricker 1990; Vermaas et al. 1993). The manganese-stabilizing protein occurs both in plants and cyanobacteria, but its sequence is rather divergent between the two phyla. Also, two proteins in contact with the manganese-stabilizing protein, the 16 and 23 kDa proteins, are present in plants but absent in cyanobacteria. In cyanobacteria, these proteins may have been functionally replaced by cytochrome c-550 (Shen and Inoue 1993). Thus, an evolutionary divergence between cyanobacteria and higher plants may have occurred with respect to CP47 to optimize, for example, the energy transfer efficiency and the interaction with extrinsic components at the donor side of PS II. To test the exchangeability of CP47 between Synechocystis and higher plants, we set out to determine to which extent CP47 from spinach can functionally replace cyanobacterial CP47 in Synechocystis through replacing part or all of the psbB gene from Synechocystis by that from spinach. One of the methodologies to create such chimaeric Synechocystis/spinach psbB genes involves recombination between the non-identical spinach and cyanobacterial psbB sequences. Homologous recombination by means of double-crossover events occurring at sequences that are identical between genome and plasmid is known to occur in Synechocystis 6803 at a high frequency (Shes-

takov and Reaston 1987; Williams 1988). The length of sequence identity required for efficient recombination generally is found to be > 100 nucleotides at each recombination site. However, in this study double recombination in Synechocystis was found to occur also if the sequences in the domains where crossover takes place are similar, but non-identical. This corroborates and extends observations made when creating chimaeric CP43 mutants (Carpenter et al. 1993). Functional characterization of the chimaeric CP47 mutants indicates that at least parts of the Synechocystis CP47 protein are needed for stability of the PS II reaction center of Synechocystis 6803 and for optimal function of the secondary quinone-type electron acceptor in PS II, QB.

Materials and methods

Growth conditions and transformation protocols of

Synechocystis sp. PCC 6803 have been described (Vermaas et al. 1987). Cyanobacterial growth rates in liquid culture were determined by measurements of the optical density (light scattering) at 730 nm as a function of time in a Shimadzu UV-160 spectrophotometer. Linearity between optical density and cell number was demonstrated up to an optical density of 0.4 at 730 nm (not shown). At optical densities exceeding 0.4, cell cultures were diluted in the cuvette before measurements were made. After restriction digestion of genomic DNA from cyanobacteria, Southern blotting to GeneScreen Plus (NEN-Du Pont) was performed, and blots were hybridized with a nick-translated cyanobacterial psbB probe according to the manufacturer's recommendations. The hybridized blots were washed at reduced stringency (two 10-min washes in 0.45 M NaCl/45 mM Na-citrate (pH 7.0), followed by two 30-min washes in 0.3 M NaCI/30 mM Na-citrate (pH 7.0) and 0.5% (w/v) sodium dodecyl sulfate (SDS), and two 30-min washes in 15 mM NaC1/1.5 mM Na-citrate (pH 7.0); all washes were done at room temperature) to retain some hybridization between the cyanobacterial psbB probe and the genomic spinach psbB present in some mutants. Oxygen evolution measurements were performed on a Gilson oxygraph (model KM) using intact cells suspended in 25 mM N-(2-hydroxyethyl)piperazineN'-(2-ethanesulfonic acid) (HEPES)/NaOH, pH 7.0. The chlorophyll concentration was 10/~g/ml, and the suspension was kept at 30 °C during the measurement.

149 Illumination was provided by a 150 W Xenon lamp, and light was focused on the sample after passing through a water filter and a yellow cut-off filter, transmitting wavelengths above 580 nm. The light intensity was 7000/~E m -2 s - l . As electron acceptor, 0.25 mM 2,6-dimethyl-p-benzoquinone was added. In addition, 0.5 mM K3Fe(CN)6 was present to reoxidize the quinone. Fluorescence excitation and emission spectra were determined at room temperaure and at 77 K using a SPEX Fluorolog 2 instrument. For fluorescence emission spectra, the excitation wavelength was 440 nm (chlorophyll excitation) or 590 nm (mostly exciting phycobilisomes). The bandwidth of the monochromator for the excitation light was 12 nm; the bandwidth of the emission monochromator was 2.4 nm. For fluorescence excitation spectra, emission was monitored at 660 nm (mostly from phycobilisome pigments) or at 690 nm (fluorescence being emitted mostly by chlorophyll with a contribution from phycobilisome components close to the thylakoid). The excitation and emission monochromator bandwidths were 2.4 and 6 nm, respectively. Kinetic fluorescence measurements were made using a PAM fluorometer (Walz, Germany). For these measurements, cyanobacterial cultures in logarithmic growth phase were harvested and resuspended in 25 mM HEPES/NaOH pH 7.0 to a chlorophyll concentration of 10 #g/ml. Where indicated, 25 #M diuron was added 5 min before the measurement. Appropriate frequencies and intensities of modulated measuring light were chosen to avoid closing of significant amounts of centers as a result of the measuring light. Herbicide binding experiments using whole ceils and procedures for thylakoid preparation were done as described by Vermaas et al. (1990). Methods used for SDS-polyacrylamide gel electrophoresis, Western blotting, and immunoanalysis have been described by Vermaas et al. (1988). Thermoluminescence measurements were carried out on a home-made instrument described in Carpenter et al. (1993). Wild-type and mutant cells to be used for thermoluminescence measurements were grown in BG-11 medium (Rippka et al. 1979) supplemented with 5 mM glucose, and were harvested during the logarithmic phase of growth. Cells were washed once, and resuspended in 25 mM HEPES/NaOH (pH 7.0) in 15% (v/v) glycerol at 75 #g/ml chlorophyll. Samples were kept in darkness or very dim light, and after transfer to the thermoluminescence stage the samples were dark-adapted further for three min at 20-30 °C.

After this additional dark adaptation, cells were cooled to - 5 °C on the thermoluminescence stage, flashed with the indicated number of saturating #s flashes, and cooled to - 4 0 °C before the heating process was started. Other experimental details have been described in Carpenter et al. (1993).

Results

Introduction of the spinach psbB gene into the cyanobacterial genome. The first step in our studies was to determine whether spinach psbB in its entirety could functionally replace the wild-type psbB gene in Synechocystis 6803 and yield a functional PS II complex. To introduce the spinach psbB gene into the Synechocystis genome, first wild-type psbB was deleted from the cyanobacterial genome. To this purpose, site-directed mutagenesis was applied to create a Nco I site at the translational start site of the gene. The region between this Nco I site and another Nco I site 0.5 kbp downstream of psbB was replaced by a Sm (spectinomycin)-resistance marker derived from pHP45f~ (Prentki and Krisch 1984). After introducing this construct into Synechocystis and after genomic segregation, the cyanobacterial psbB deletion mutant psbB-D1 (D denoting 'deletion') was obtained. A summary of the genetic makeup of the psbB region in this and other psbB mutants generated and analyzed in this study is provided in Figure 1. To introduce spinach psbB, the psbB-D1 mutant was transformed with a construct containing spinach psbB, flanked on both ends by cyanobacterial DNA from up- and downstream ofpsbB. This construct was made using the naturally occurring Nco I site at the translational start site of spinachpsbB, and using a Dra I site downstream of the gene in comparable locations in spinach and Synechocystis. In this way, the entire spinach psbB coding region could be introduced into Synechocystis while retaining the native Synechocystis upstream region, including the ribosome-binding site. The resulting Synechocystis mutant carrying spinach psbB is referred to as psbB-S 1 (S denoting the spinach origin of the gene). The psbB-S 1 mutant is an obligate photoheterotroph (see below), indicating that the spinach CP47 protein cannot functionally replace the cyanobacterial protein in the Synechocystis 6803 PS II complex. The sequence differences between the CP47 proteins from the two species are summarized in Table 1. The CP47 proteins are different at 120 residues, and in many cases the substitutions are con-

150 b-~rain

photoautol~ophlc

growth

H

wild type

B

t

++

t

P

K

S

L L

DNNK

L ~ , - I J J LJ

H

t

H

t

1,4

1,4

K • I I%\\\\\\\N

IN.~\\\\\\\~

psbB.Dl

I

14 I

H S I I H

N B

/

/

IN\\\\\\\\,"

~-$I

Hl ~ r" ~ 'j.,.,~ -"--

/

/

..........

~

psbB I

I

psbB-Cl

+

I

I

~'-

L

I

k \ \ \ \ \ \ \ \ N . ~ . " ..".".."."~.', . ' i ; ; ;

l

I

psbB-C2

++

psbB-C6

+

B

P

H

B

P K

I

I I

NN S I

~

~

~r--r ~,'~.v

I

H

t I. t . ~ I" I N N \ X \ \ \ N , \ " ~ " -". -"."-". -"-"."-"." ~ \ \ \ \ \ \ \ \ \ \ ~ I i i psbB S NN 8

i

DNNK

H

~,.'~'~.,~.,~..'~..'.'~..\\\\\\\\\\\\~I psbB H S I

I k\\\\\\\\xl....

psbB-C7

I~

I

~,,,.1~' r~,,,,*'~

I

H

/

~"~'JJl%\\\\\\\\\\l H

I

H

I~

C -~-iJ'~K N ~ \ \ \ \ \ \ \ \ I NN H S

~=...;;

I

~

I

i;=.. I; -. ~'~IN~\NNNNN\N\'NJ

psbB

I I

NN 1 kb

Synechocysti$sequence spinach sequence

~

spectinomyctn-resJstance c a ~ l d g e

~

kanamydn-resistance cartridge

Figure 1. Schematic description of the genetic makeup of the psbB region of the genome of wild-type and mutant strains analyzed. The capacity of strains to grow photoantotrophically has been listed (++: normal photoautotrophic growth; + : decreased rates of photoantotrophic growth; - : obligate photoheterotroph), Restriction sites have been indicated (B: BamH I; D: Dra I; H: Hind III; K: Kpn I; N: Nco I; P: Pvu II; S: Sma I). See text for a description of the generation and properties of the mutant strains.

servative. Most of the sequence differences are located in loop regions. A total of 19 differences are observed in the major hydrophobic domains that are assumed to span the membrane; these major hydrophobic regions make up about 25% of the total CP47 protein. To determine the region(s) ofpsbB that need to be of cyanobacterial origin to support photoautotrophic competence, three different approaches were utilized that are presented in the following sections.

Chimaeric spinach/cyanobacterial psbB mutants. In the first approach, the obligate photoheterotrophic

psbB-S1 mutant was transformed with various plasmids carrying different parts of Synechocystis psbB, and we checked for the occurrence of photoautotrophic transformants (Figure 2). In this way CP47 domains that need to be of cyanobacterial origin to have photoautotrophic competence may be identified. For a transformation with part of Synechocystis psbB to be successful, one must assume that recombination can occur between similar regions of homologous genes. As is indicated in Figure 2, only transformation with a relatively large Synechocystis psbB fragment (from a BamH I site about 250 nucleotides from the 5' end

151

Table 1. Amino acid differences in CP47 from Synechocystis 6803 and spinach, and summary of the spinach residues present in the various chirnaeric mutants. The location of the residues has been determined from Vermaas et al (1987), Bricker (1990) and Vermaas (1993) Residue number

Residue in

Residue in spinach

Location

Synechocystis6803

Chimaeric mutants with spinach residue

24 44

Leu lie

lie Val

helix I loop I-II

47 50 53 63 67 71 73 76 80 83 85 86 89 90 93

Ser Ala Asn Leu Ala Val Ser Ash Val Glu Gly Leu Gly Phe Phe

Pro Pro Asp lie Thr lie Asn Giy lie Gly Thr Thr Ser lie Tyr

loop loop loop loop loop loop loop loop loop loop loop loop loop loop loop

I-II I-II I-II I-II I-II I-II I-II I-II I-II I-II I-II I-II I-II I-II I-II

98 102 103 106 112

Ala Val Leu Leu Val

Gly Met Phe Cys lie

helix helix helix helix helix

II II II II II

117 122 124 126 130 131 132 138 148 149 158 161 162 166 177 182 183 184 187 188 191 194 196 198 199 200 207 208

Phe Leu Val Pro Glu Ser Ala Met Leu Leu Leu Val Trp Met His Ala Pro Glu Pro Ala Asn Ash Gly Val Val Ala lie Val

Tyr lie Ser Glu Lys Pro Ser lie Val Ala Val Leu Tyr lie Lys Cys Ser Ala Val Glu Asp Val Arg Ile Ala Ser Thr Leu

helix II loop II-III loop II-III loop II-III loop II-III loop II-III loop II-III loop II-III helix III helix III helix III loop Ill-IV loop Ill-IV loop Ill-IV loop III-IV loop Ill-IV loop Ill-IV loop Ill-IV loop Ill-IV loop Ill-IV loop Ill-IV loop Ill-IV loop Ill-IV helix IV helix IV helix IV helix IV helix IV

9sbB-C6 9sbB-C6 9sbB-C6 9sbB-C6 ~sbB-C6 ~sbB-C6 ~sbB-C6 ~sbB-C6 ~sbB-C6 ~sbB-C6 ~sbB-C6 ~sbB-C6 psbB-C6 psbB-C6 psbB-C6, psbB-C7 psbB-C6, psbB-C7 psbB-C6, psbB-C7 psbB-C7 psbB-C7 psbB-C'l psbB-C7 psbB-C'l psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7

152

Table 1. Continued Residue number

Residue in spinach

Location

Synechocystis6803

Residue in

Chimaeric mutants with spinach residue

211 218 221 223 228 260 277 282 285 289 290 291 295 296 298 304 315 318 327 329 331 332 338 339 341 347 355 361 369 370 371 373 376 381 386 390 395 400 405 407 409 410 411 413 415 416 420 423 429 431 432 434

Ile Thr Pro Glu Ala Asn Lys Glu Gin Asp Ser Gln Gly Ala Leu Thr Val Ser Thr Ala Asn Ser Gin Glu lie Lys Glu Asn lie Met Thr Ala Val lie Ser Phe Thr Ser Ala Asp Gin Thr Phe Asn Ser Asp Phe Lys Gly Asp Phe Thr

Leu Ser Ser Gin Gly Ser Gin Gin Tyr Ser Ala Gly Asn Gln Phe Lys lie Asn Ala Ser Asp Asn Val Gly Leu Arg Phe Thr Val Leu lie Gly Ile Val Ala Tyr Val Glu Glu Asn Val Ser Tyr Asp Ala Thr Tyr Arg lie Glu Leu Arg

helix IV helix IV loop IV-V loop IV-V loop IV-V helix V loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI loop V-VI

psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-C7 psbB-Cl,psbB-C7 psbB-C1,psbB-C7 psbB-Cl,psbB-C7 psbB-Cl,psbB-C7 psbB-Cl,psbB-C7 psbB-Cl,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1, psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1, psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1, psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-Cl,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-Cl,psbB-C7 psbB-Cl,psbB-C7 psbB-C1, psbB-C7 psbB-C1,psbB-C7 psbB-Cl,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1, psbB-C7 psbB-Cl,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-C1,psbB-C7 psbB-Cl,psbB-C7 psbB-C1, psbB-C7 psbB-C1, psbB-C7

153

Table 1. Continued Residue number 435 437 438 445 457 482 485 487 488 494 496 498 501 502 505 506 507 508

Residuein

Residue in

Location

Chimaericmutants with spinach residue

loop V-VI loop V-VI loop V-VI loop V-VI helix VI C-terminus C-terrmnus C-terminus C-terminus C-terminus C-terrmnus C-terminus C-terminus C-terminus C-terminus C-terminus C-terminus C-terminus

psbB-C1, psbB-C7 psbB-C1, psbB-C7 psbB-C1, psbB-C7 psbB-C1, psbB-C7 psbB-C1, psbB-C7 psbB-C1,psbB-C2,psbB-C7 psbB-Cl, psbB-C2, ~sbB-C7 psbB-Cl, psbB-C2, 7sbB-C7 psbB-Cl, psbB-C2, 9sbB-C7 psbB-Cl, psbB-C2, 7sbB-C7 psbB-Ci, psbB-C2, 9sbB-C7 psbB-C1,psbB-C2, ~sbB-C7 psbB-Cl, psbB-C2, ~sbB-C7 psbB-Cl, psbB-C2, 7sbB-C7 psbB-Cl, psbB-C2, 7sbB-C7 psbB-Cl, psbB-C2, 7sbB-C7 psbB-Cl, psbB-C2, ~sbB-C7 psbB-Cl, psbB-C2, ~sbB-C7

Synechocystis6803 spinach Glu Phe Asn Thr Val Val Gly Glu Giu Val Ala Val Leu Ser Lys Glu Ala

Ala Leu Lys Ser Ser lie Asp Asp Val Ala Gln Ile Pro Thr Arg Gln Gly Val

of the gene up to the downstream region) yielded photoautotrophic transformants. This indicates that recombination between similar (but non-identical) DNAs can occur, and that a major part of psbB needs to be of cyanobacterial origin for restoration of photoautotrophic growth. One photoautotrophic transformant isolated by this method (psbB-C6; C indicating the chimaeric nature ofpsbB in this mutant) was sequenced, and its site of recombination in psbB was found to be in a region 20 nucleotides downstream from the Barn HI site; in this region, a sequence of five nucleotides is identical between Synechocystis 6803 and spinach. This mutant was characterized further (see below).

Transformation of wild-type Synechocystis with spinach psbB. A second approach to introduce part of spinach psbB into the cyanobacterial genome is to transform wild-type Synechocystis with a plasmid construct containing spinach psbB, linked at its 3 I end to a kanamycin (Km)-resistance cartridge; the 31 end of this cartridge was connected to a region of cyanobacterial D N A originating from downstream of psbB. Kanamycin-resistant transformants of Synechocystis will be generated only upon recombination between plasmid and chromosome at each side of the Km-resistance marker, i.e. downstream of psbB, and between spinach psbB on the plasmid and cyanobacte-

rial psbB in the chromosome (Figure 3). Km-resistant transformants were obtained at a reasonable frequency (about 1 transformant per 3 x 105 cells), indicating that recombination indeed can occur between similar regions of highly homologous, but non-identical sequences. Several transformants obtained by this approach were selected and purified. A number of these did not grow photoautotrophically, some showed slow photoautotrophic growth, whereas others resembled wild type in their growth rates. For further characterization, one photoautotrophic transformant with a somewhat decreased growth rate (psbB-C1) and one exhibiting normal growth (psbB-C2) under photoautotrophic conditions were selected. To identify the precise site of the crossover between plasmid and chromosome in psbB in the psbB-C1 and psbB-C2 strains, psbB was cloned out of these mutants and the region containing the crossover was sequenced. Figure 4 shows selected regions of the Synechocystis and spinach psbB sequences, and the sequence of the psbB-C 1 and psbB-C2 mutants in this region. For psbBC 1, crossover between the cyanobacterial genome and the spinach psbB-containing plasmid occurred in the region between base pairs 730 and 734 of psbB, in a region of the gene that, at the amino acid level, is extremely well-conserved between Synechocystis

154

p~B

I

~

I regionof psbB-S1 genome

T m n ~ m u d l ~ ~ ~b_~S1 strain ~

~a~mids ~ntaln~g:

K

a.

Pbotmautotrophlc ~'ansformants:

K

none

L\\\\\\\\\\\\I B

b.

yes (e.g., psbB-C6)

L\\\\~\\\\\\\\\\\\\N B K

none K

d.

S

none

~

1 kb Syn~.hocy=~ soquonce P77"~ spinachsequence I

I kanamydn-resistance cartridge

Figure 2. Transformation of the Synechocystis mutant carrying spinach psbB (mutant psbB-Sl) with plasmids carrying different regions of Synechocystis psbB and flanking regions (a~l). Photoautotrophic transformants were selected; only transformation with construct (b), which carries most of the cyanobacterial psbB gene, yielded photoautotrophic transformants. For generation of the plasmid inserts indicated in (a~t) the following restriction enzymes cutting in SynechocystispsbB have been used (see Figure 1): Barn HI (B), which has a restriction site 258 bp downstream from the psbB translation start site; Kpn I (K), which cuts approximately in the middle of the psbB gene (762 bp from the start site); Sma I (S), which recognizes a site inpsbB 1342 bp from the translation start site, and 171 bp upstream from the end of the gene.

[~\\\\1-"

psbe -" -" C .,"C.,'.," -."-" -'-" ~ , " ~ \ \ \ \ ~ 1

Syncchocys~ 6803 wgd type

+

psbe ~ K r n r ~

"'~

plasmidvathspLnachpsbe

~'-'-'//~';;Z/;;--%'/J~\\\~J

and Km-mslstancemarker

~ b'ansformation and selection for Km resistance

I

IN\\\\"

~

I

k",,"d d d d d -." d d ..~d ~-'J~\\\\~'l psbB-C2mutant

Synechocys#ssequence

Ikb

spinach sequence I

I kanamycin-resistance car~'Idge

Figure 3. Flow scheme for transformation of wild-type Synechocystis sp. PCC 6803 with a plasmid with a spinach DNA segment coveting psbB and a 370 bp sequence downstream of this gene. This downstream region was flanked at its 3 r end by a kanamycin-resistance cartridge. A region from downstream of Synechocystis psbB followed the antibiotic-resistance cartridge. Kanamycin-resistant transformants (psbB-C1 and psbB-C2) are created by a crossover in the homologous region downstream of the cartridge coupled to one in the heterologous psbB region. This results in transformants carrying a chimaeric cyanobacterial/spinach psbB gene, with a 5 r region from cyanobacterial origin, and a complementing 31 region from spinach.

155 6803

psbB-Cl spi

720 730 .... CAGTAGTATTGCCGCTG

740 TATTTTTTGCC

.... C A GT A GT A TTGCCC~"rG

T ~-F~ u u " Y r G C A G . . . .

.... TAGTAGTATCGCTGCTGT~uu'F~"FFGCAG

G ....

....

6803

1340 1350 1360 .... CCAGTCCCCGGGGTTGGTTTACCTTTGGCC

....

psbB-C2

.... CCAGTCCCCGGGGTT~GTTTACTTTGGGAC

....

spi

.... GTAGTCCGAGGGGTTGGTTTACTTTGGGAC

....

Figure4. Sequenceof regions of the psbBgenes fromspinach (spi) and Synectu~cystissp. PCC 6803 (6803), and from the chimaeric mutantspsbB-Cl andpsbB-C2.Nucleotidenumberingstarted at the A of the translation initiation codonof the gene. Regionsof crossover in the chimaeric mutants have been underlined.

and spinach (a stretch of 31 amino acids are identical in this region). This domain is located in the fifth major hydrophobic region of the CP47 protein, before the long hydrophilic, lumenally exposed loop between transmembrane regions V and VI. However, the DNA sequence is not as much conserved in this domain (77% sequence identity on the nucleotide level). In psbB-C2, crossover has occurred in a very homologous stretch of DNA, just beyond the cyanobacterial Sma I site that is located near the Y end of the gene, in the region coding for the last major hydrophobic region of CP47. To investigate whether the sites of recombination are linked to regions of maximal homology between spinach and Synechocystis psbB, the percentage of sequence identity within a 10-nucleotide window in the regions of recombination in psbB-C 1 and psbB-C2 were compared to the percentage of sequence identity between spinach and Synechocystis in other areas of the gene. Recombination events have occurred in regions of rather high sequence identity in these two chimaeric mutants, but regions with similarly high sequence identity are scattered throughout the psbB gene (data not shown). Thus, we expect that transformation of wild type with the plasmid containing the spinachpsbB gene can give rise to a myriad of different mutants, suggesting that this method is useful to generate a broad spectrum of psbB chimaeras.

Chimaeric spinach / Synechocystis psbB gene constructs. A third avenue to create chimaeric psbB mutants is to make use of restriction sites that are conserved between spinach and Synechocystis to construct hybrid psbB plasmids that can be used to transform psbB-D1 (the deletion mutant). A convenient conserved restriction site in psbB is one for BamH I, located about 260 bp downstream from the translation start site of the gene. A psbB plasmid construct was made

carrying the spinach sequence from the BamH I site to a Dra I site beyond the end of the gene, and flanked by appropriate Synechocystis sequence. A Km-resistance cartridge was inserted in the sequence downstream of psbB. Transformation of the psbB-D1 mutant (lacking psbB) with this plasmid construct yielded obligatorily photoheterotrophic transformants, indicating that the 5' 260 bp ofpsbB being of cyanobacterial origin is not sufficient to restore a functional, stable PS II complex. One of these mutants, psbB-C7, was used for further characterization. In this strain, the first 260 bp of the translated region of psbB is from Synechocystis, the remainder is from spinach.

Southern blotting. To confirm that upon transformation the recombination between plasmid and genome has occurred by a proper double-crossover event, a Southern blot of DNA isolated from the Synechocystis wild type and mutants and cut with different restriction enzymes was prepared and probed with a cyanobacterial Kpn I/Kpn I DNA fragment containing the Y half of psbB. The results are shown in Figure 5. Note that hybridization of the cyanobacterial probe to spinach psbB is rather weak, and that relatively low stringency was required to obtain sufficient hybridization. From the results shown in the Southern blot, we conclude that recombination indeed has occurred by doublecrossover events in the mutants.

Growth rates. Growth of wild type and mutants in liquid medium under photoautotrophic and photoheterotrophic conditions is shown in Figure 6. Under photoautotrophic conditions, psbB-D1, psbB-S1 and psbB-C7 do not propagate, whereas the rate of growth of the other mutants is somewhat lower than that of wild type. In the presence of glucose (when PS II is not required for growth),psbB-D 1, psbB-S 1 and psbBC7 grow at a rate reasonably similar to that of wild type, indicating that in these mutants PS II is impaired. Interestingly, the photoautotrophic growth rate of psbB-C1 is found to be about a factor of two lower than that of wild type. The psbB sequence of this mutant indicates that the first five hydrophobic domains (potentially membrane-spanning regions) and the short hydrophilic regions between these domains of CP47 are of cyanobacterial origin, whereas the long hydrophilic loop and the last hydrophobic domain are from spinach. Further extension of the spinach region towards the N-terminal end of CP47 to include four more hydrophobic domains from spinach (as is the case in psbB-C7) eliminates the photoautotrophic capacity

156

Figure 5. Southern blot of genomic DNA from wild type (lane 1), psbB-Dl (lane 2), psbB-S1 (lane 3), psbB-C7 (lane 4), psbB-C1 (lane 5), psbB-C6 (lane 6), and psbB-C2 (lane 7) cut with Hind III, Kpn I, Sma I, or Pvu II. The blot was probed with a 1.4 kb cyanobacterial Kpn l/Kpn I fragment carrying the 3t half ofpsbB and a 0.7 kb downstream region. Restriction sites in psbB and its surroundings from spinach and Synechocystis sp. PCC 6803 are indicated in Figure 1. Table 2. Chlorophyll/PS II ratio, [14C]-diuron dissociation constant, and oxygen evolution rates in wild type and psbB mutants Strain

Chlorophyll/PS II

KD, nM

Oxygen evolution (% of wild typea)

Wild type 7sbB-C1

650 1600 850 1450 N.D.b N.D. N.D.

16 18 16 17 N.D. N.D. N.D.

100 39 74 42 0 0 0

9sbB-C2 9sbB-C6 ~sbB-C7 9sbB-Dl 9sbB-Sl

The error in these measurements is approximately 20%. Data shown here are the average of three or more independent experiments. aThe oxygen evolution rate of wild type was 270 /zmol 02 (mg chlorophyll)- l h - I. bN.D.: No measurable [14C]-diuronbinding could be detected in these strains. Therefore, the chlorophyll/PS II ratio and the dissociation constant of the herbicide could not be determined.

of the mutant. The absorption spectrum of cells of any o f the m u t a n t s is indistinguishable from that of wild type (not shown), indicating that p h y c o b i l i s o m e s a c c u m u l a t e in all m u t a n t s to a level similar to that in wild type.

Quantitation of PS II. In order to m o n i t o r the effect of the psbB mutations on the a c c u m u l a t i o n of PS II in thylakoids, specific b i n d i n g of a radiolabeled PS II-directed herbicide, [laC]diuron, to intact cells was measured at different d i u r o n concentrations. F r o m the resulting data, the d i u r o n affinity and the n u m b e r of herbicide b i n d i n g sites on a chlorophyll basis was calculated (Tischer and S t r o t m a n n 1977). The results are s u m m a r i z e d in Table 2. Wild type was f o u n d to contain one diuron b i n d i n g site per 650 chlorophyll molecules, whereas forpsbB-C2 the n u m b e r of PS II centers on a chlorophyll basis appeared to have decreased a little. The chimaeric mutants psbB-C1 and psbB-C6, which contain a larger part of CP47 from spinach, had a m o r e p r o n o u n c e d loss of PS II centers on a chlorophyll basis. Mutants c o n t a i n i n g an even larger portion o f CP47 from spinach origin, however, did not b i n d diuron, suggesting that they did not a c c u m u l a t e a significant a m o u n t of intact PS II complexes in their thylakoids. Oxygen evolution. The capability to perform PS II electron transport was m o n i t o r e d through o x y g e n evolution m e a s u r e m e n t s at saturating light intensity. A n artificial electron acceptor (0.25 m M 2,6-dimethyl-pb e n z o q u i n o n e ) was added, and 0.5 m M K3Fe(CN)6 was present to keep the q u i n o n e in oxidized state. The

157 (a) 0.4 -]"

l

-0.0 N

-o.4

o 0

-0.8

$ -1.2 -1.6 0

20

40

60

80

100

TIME ( hours ) (b) 0.5

E

~ 0

~ 1

0.0

-0.5 -1.0 -1.5 0

20

40 60 TIME ( hours )

80

100

Figure 6. Growth curves of wild type and selected psbB mutants under photoautotrophic (a) and photoheterotrophic (b) conditions in BG-l I medium (without and with the addition of 5 mM glucose, respectively), o: wild type; ~ : psbB-Dl; O: psbB-Sl; A: psbB-C7; D: psbB-Cl; +: psbB-C6; &: psbB-C2.

Figure 7. Western blot of thylakoid proteins from wild type (lanes 1), psbB-D1 (lanes 2), psbB-S1 (lanes 3), psbB-C7 (lanes 4), psbB-C1 (lanes 5), psbB-C6 (lanes 6), and psbB-C2 (lanes 7) challenged with antisera raised against spinach D2 and DI (A) and CP47 and CP43 (B).

results are summarized in Table 2. No oxygen evolution could be observed in strains lacking herbicide binding and lacking photoautotrophic capacity. The amount of oxygen evolution on a chlorophyll basis measured in the other strains is decreased to approximately the same degree as the PS II / chlorophyll ratio in these strains. In all strains that showed oxygen evolution, the rate of oxygen evolution was stable for at least several minutes at the high light intensity (7 000 #E m -2 s -1) used for the measurement. This stability resembled that of the wild type. Therefore, remaining PS II centers in the photoautotrophic chimaeric strains appear to be rather normal in a functional sense.

PS Hproteins. To determine the PS II protein composition in thylakoids from the various mutants, thylakoid proteins were size-separated by SDS-polyacrylamide gel electrophoresis, blotted onto nitrocellulose, and reacted with antisera raised against D1, D2, CP43 and CP47. The results are shown in Figure 7. As expect-

ed, in the psbB deletion mutant (psbB-D1) no CP47 could be detected in the thylakoid. Also replacement ofpsbB by the homologous gene from spinach (mutant psbB-S 1) did not support stable incorporation of CP47 in the thylakoid. CP47 was missing also in the obligate photoheterotrophic mutant psbB-C7. As expected, all photoautotrophic mutants contained CP47; however, the mobility of hybrid proteins was altered in some cases. This is particularly evident in psbB-C1. All mutants contained a relatively normal amount of CP43 in their thylakoids. D1 and D2 were found to be present in apparently normal amounts in the photoautotrophic psbB mutants. The mutants lacking a significant accumulation of CP47 in the thylakoids contained reduced (but significant) amounts of D1 and D2. However, these proteins may not have been functionally incorporated in the thylakoid, because no herbicide binding could be detected (Table 2).

158 100

n

-,-I

75 -.-.I -0 ,.-I Q) I,..I

50

.,u -,,-t r~ a) .,u

25

-.-I [-I

I -40

I

i

-20

I 0

I

I 20

I

I 40

I

I 60

temperature, Oc Figure 8.

Thermoluminescence

glow curves of wild type (--) and

the chimaeric strains psbB-C1 (- -) and psbB-C6 (- - -). Diuron (20/zM), which shifts the thermoluminescence maximumto lower temperatures (5-7 °C), has been added to samples that showpeaks near the '+'. The otherthree curvespresented here were recordedin the absenceof diuron(-). Thermoluminescencewas measuredafter 2 flashes and subsequent freezing (see 'Materials and methods'). A full vertical scale (of 100)correspondsto 14400cps detected by the photomultiplier.

Fluorescence.

The absorption spectra of wild type and all mutants are indistinguishable, indicating that changes in PS II levels in the thylakoid do not affect accumulation of phycobilisomes in the cell (not shown). To investigate energy transfer between chlorophylls, and between phycobilisomes and chlorophylls in wild type and mutants, fluorescence emission and excitation spectra were recorded at room temperature and at 77 K (data not shown). The results indicated normal energy transfer between phycobilisomes and remaining PS II centers in the photoautotrophic mutants (psbB-C1 and psbB-C6). Thus, energy transfer between phycobilisomes and chlorophyll a, discovered by W. Arnold (see Arnold and Oppenheimer 1950), does not appear to have been affected considerably by introduction of spinach sequences in the CP47 protein.

Thermoluminescence. Thermoluminescence in photosynthetic organisms was discovered by W. Arnold (see, for example, Arnold 1966) and is a sensitive monitor of thermodynamic properties of redox components involved in PS II electron transfer at both the electron donor and electron acceptor sides (reviewed in Vass and Inoue 1992; Vass and Govindjee, this issue). To determine whether the introduction of chimeric CP47

has affected the midpoint redox potential of electron transport components, thermoluminescence was measured in intact cells of wild type and various mutants (Figure 8). The glow peak of thermoluminescence originating from charge recombination between Qff and the donor side (the B band) consistently was shifted to lower temperatures by 6--7 °C in both psbB-C1 and psbB-C6 as compared to wild type (Figure 8). This shift was found regardless of flash number (data not shown). No significant thermoluminescence was observed in mutants lacking a functional PS II complex, such aspsbB-S 1. In the presence of diuron, which blocks electron transfer between QA and QB, the glow peak (the Q band) in wild type was at 6-7 °C and at 4-6 °C in the psbB-C1 and psbB-C6 mutants; this difference is not considered to be significant. The glow peak observed in the presence of diuron represents a back reaction between QA and the donor side. As the glow temperature of thermoluminescence involving QB was shifted in thepsbB-C1 andpsbB-C6 strains as compared to wild type while the QA glow peaks did not show significant differences between wild type and mutants, we interpret this to indicate that the mutations in CP47 affect mostly the properties of QB. It is interesting to note that the glow temperature in these two chimaeric mutants shifted by a similar degree, even though the nature of the mutants is different: psbB-C1 carries a spinach C-terminal half of CP47, whereas psbB-C6 has an N-terminal CP47 region from spinach.

QA oxidation. As presence of a chimaeric CP47 protein in PS II appeared to affect the properties of Qff as determined by the thermoluminescence peak temperature, it was of interest to monitor the kinetics of Q~ formation and the location of the QA " QB ~- QA • Qa equilibrium. To monitor whether the kinetics of the forward reaction QA "QB ~ QA" QB have been changed, changes in variable chlorophyll a fluorescence yield after a single turnover flash were measured. The results are presented in Table 3. The rate of QA oxidation (as determined from the decay of variable fluorescence) is decreased somewhat in psbB-Cl: the half-time of the decay is about 40% longer in this mutant than in the psbB-C2 and psbB-C6 strains. Fluorescence decay kinetics in the latter two strains were similar to those of wild type. An indication of the value of the apparent QA " Q B ~-- QA • QB equilibrium constant is provided by the value to which the variable fluorescence decays after a flash; the presence of a significant equilibrium concentration of QA (i.e. a decreased equilibrium constant) will be visible as an increased variable

159

Table3. Kinetics and extent of QA oxidation in wild-type and mutant strains in the absence and presence of 25 #M diuron as measured by variable fluorescence Strain

tl/2 (ms)a (-diuron)

~,r*l tl/2 (ms)~ Fb,rgl (25 ms after flash) (+diuron) (5 s after 5 s ill.) ( - diuron) (+ dinron)

Wild type

0.66 0.97 0.66 0.71

0.09 0.15 0.08 0.14

psbB-C1 psbB-C2 psbB-C6

600 480 560 560

0.03 0.03 0.03 0.03

Data shown here are the average of three or more independent experiments. The error in the determined halftimes is approximately 10%; the determined relative Fv values are reproducible within 0.02-0.03. a Values reflect the time after a single-turnover Xenon flash where the variable (chlorophyll a) fluorescence yield has decayed to 50% of its initial value. Fluorescence decay is multi-phasic in these systems; however, the relative amplitude of the various phases is similar for the different strains. b Values reflect the fraction of variable (chlorophyll) fluorescence remaining at 40 ms after the single turnover Xenon flash (without diuron) or at 5 s after the end of a 5-s illumination in the presence of diuron.

fluorescence after an electron distribution equilibrium between QA and QB has been established. The results shown in Table 3 indicate an increased variable fluorescence yield at 40 ms after a flash in the psbBC1 and psbB-C6 mutants, consistent with a shift in the semiquinone equilibrium constant predicted from the thermoluminescence data. However, the apparent equilibrium constant of the semiquinone equilibrium remains significantly larger than one (favoring QA • QB) as the amount o f the 'stable' variable fluorescence yield after a flash is relatively small and as the thermoluminescence temperature involving Q~ is still much higher than that in the presence of diuron (involving QA only). As indicated in the introduction, the large hydrophilic region in the C-terminal half of CP47 may interact with components at the PS II donor side, and therefore it was important to test whether introduction of spinach sequences might affect properties of the donor side. For this purpose, the kinetics of QA decay in the presence of diuron (mostly reflecting QA oxidation by donor-side components) were monitored. As shown in Table 3, these kinetics were essentially unaltered. In addition, the accessibility of the donor side to exogenous reductants was monitored by measuring the amount o f stable variable fluorescence after a 5 s-illumination o f cells that had been dark-adapted for 5 min in the presence of diuron (Chu et al. 1994). Five s after the end of the illumination, essentially no variable fluorescence remained, indicating a full oxidation of Q~ mostly by oxidizing equivalents that had

been formed at the donor side. This implies that in these mutants the donor side does not easily lose its oxidizing equivalents and that thus the accessibility to exogenous reductants has not been altered significantly in any of the PS II-containing mutants covered in this study.

Discussion

Recombination mechanisms. Generally, homologous recombination in Synechocystis is utilized to target gene constructs to desired sites in the cyanobacterial genome (Shestakov and Reaston 1987; Williams 1988). However, this study on psbB and a previous study on psbC by Carpenter et al. (1993) show that a long stretch o f sequence identity between plasmid and chromosome at at least one of the sites of double-crossover is not an absolute requirement for efficient transformation. Upon double-crossover, an overall sequence identity of 7 5 - 8 0 % sequence identity on one side appears to be sufficient to obtain frequent recombination. However, it should be kept in mind that there are limits to the efficiency with which recombination between homologous genes from different species occurs in Synechocystis. For example, transformation of wild type Synechocystis with an apcE gene (coding for the phycobilisome anchor protein) from Synechococcus sp. PCC 7002 interrupted by a kanamycinresistance marker did not yield Synechocystis transformants with interrupted apcE, in spite of the fact that the

160 overall percentage of sequence identity between apcE from the two organisms is about 65% (G. Shen and W. Vermaas, unpub, observations). Nonetheless, the property of Synechocystis 6803 to undergo recombination even if perfect sequence identity does not exist expands the suitability of Synechocystis6803 as a system for targeted mutagenesis, in that gene interruption may be possible even if the Synechocystis 6803 gene is not available and a homologous gene from another organism is used.

Functional consequences of chimaeric CP47. As evidenced by the obligate photoheterotrophic nature of psbB-S 1, spinach CP47 does not functionally replace the cyanobacterial protein. However, since the psbBC 1 mutant is photoautotrophic, the 3 ~half ofpsbB can be of spinach origin without dramatic effects on translation, function, or stability of CP47. This suggests that the large lumen-exposed loop between transmembrane regions V and VI can be of spinach origin without having a significant effect on the overall donor-side properties. The lack of a significant donor-side effect in the psbB-C1 mutant is corroborated by the observation that Q~, oxidation in the presence of diuron is normal, both after a single flash and after a 5-s illumination period. The psbB-C1 mutant appears to be normal in its functional accommodation of the Synechocystis 6803 33-kDa manganese-stabilizing protein (PS II-O), which interacts with the long hydrophilic loop of CP47 (reviewed in Bricker (1990) and Vermaas et al. (1993)). The lack of functional PS II-O leads to large changes in the thermoluminescence glow curve, both in the presence and absence of diuron (Burnap et al. 1992), whereas in psbB-C1 such changes are not observed as compared to the control. Another indication that PS II-O is bound functionally to the PS II complex in the psbB-C1 mutant is the stability of oxygen evolution at saturating light intensity: mutants lacking PS II-O are easily photoinactivated (Mayes et al. 1991; Philbrick et al. 1991). The implication that cyanobacterial PS II-O can be accommodated in systems carrying a CP47 protein with the large hydrophilic loop originating from spinach is in line with the observation that the manganese-stabilizing protein from spinach can partially restore oxygen evolution in PS II complexes from a cyanobacterium (Koike and Inoue 1985). However, in the psbB-C1 mutant the properties of the acceptor side are somewhat modified. The Q~ oxidation rate by QB has decreased and the semiquinone equilibrium has shifted somewhat to the left. The lat-

ter property appears to be shared with the psbB-C6 strain, in which the 5 ~ 20% of the psbB gene is of spinach origin. This suggests that both a region near the N-terminus and one near the C-terminus may interact with the acceptor side of PS II. The somewhat altered characteristics of QB do not appear to affect the binding properties of PS II-directed herbicides to the site: the diuron affinity remains unchanged in these mutants (Table 2). This is not very surprising as the precise binding sites of the quinones and herbicides are close, but not overlapping (see Bowyer et al. 1991 for a review). The effect of CP47 on the properties of QB presumably are indirect, but the fact that other parts of PS II appear unaffected in the psbB-C1 and psbB-C6 mutants implies that the influence of CP47 on QB may be rather specific. On one hand, cyanobacterial and spinach CP47 are very similar: their primary sequence is about 80% identical, and regions of the cyanobacterial CP47 can be replaced by homologous regions from spinach without dramatic effects on its function in PS II. However, on the other hand, it is obvious that regions of CP47 need to be of cyanobacterial origin in order to maintain stability in a cyanobacterial PS II complex. This bears some resemblance to what was observed for CP43 (Carpenter et al. 1993), but is different from the situation for DI: the psbA genes from Synechocystiscan be replaced by that from Poa annua, and a functional cyanobacterial PS II complex can be obtained (Nixon et al. 1991). Based on the observation that both the N-terminal end and the C-terminal half of CP47 can be of spinach origin, but not the region in between, it appears that a domain within the area between residues 97 and 229 needs to be of cyanobacterial origin to allow stable assembly of CP47 within the PS II complex of Synechocystis. In this region, 38 out of 131 amino acid residues are different between the spinach and Synechocystis CP47 sequences (see Table 1). It is impractical to test the importance of each of the non-conserved residues in terms of their necessity for stable assembly of the CP47 protein. However, it is possible that one or more of the non-conserved residues that potentially can ligate chlorophyll (Asn, Gin and His) in loop regions of the protein serve as chlorophyll-binding residues in Synechocystis 6803 but are absent in the homologous location in the spinach sequence. Chlorophyll binding to CP47 has been shown to be a major factor in stable assembly of the PS II complex (Shen et al. 1993; Shen and Vermaas 1994).

161 In psbB mutants lacking CP47 in their thylakoids, also D1 and D2 were significantly depleted. This most likely is due to destabilization of these components; pleiotropic effects o f mutations on other PS II components in Synechocystis 6803 generally have been found to occur on the post-translational level. Little effect on transcript levels o f genes coding for other major PS II proteins has been detected in a Synechocystis psbBinterruption mutant (Yu and Vermaas 1990). We have not yet been successful in unequivocally determining by pulse labeling whether CP47 is synthesized in psbB-S 1: the apparent instability of spinach CP47 in the cyanobacterial PS II complex precludes accumulation o f sufficiently labeled CP47 in the thylakoid to be unambiguously detectable (not shown). However, it is probable that spinach psbB is properly transcribed and translated in Synechocystis: the fully photoautotrophic capacity of psbB-C2 indicates that the changes introduced downstream o f psbB (spinach up to the Dra I restriction site, and a kanamycinresistance marker introduced at the border between the spinach and cyanobacterial sequences downstream of psbB) do not affect transcription or transcript stability to a physiologically relevant extent. Also, all codons in spinach psbB are used in SynechocystispsbB or in one of the other major PS II genes from this cyanobacteria (not shown). Thus, it is highly unlikely that spinach CP47 cannot be translated in Synechocystis. Introduction o f spinach CP47 regions into the cyanobacterial PS II complex does not have marked effects on the efficiency with which light energy absorbed by phycobilisomes can be utilized by PS II. In various chimaeric mutants having a relatively stable PS II complex only small changes in the fluorescence properties can be observed, indicating that energy transfer between phycobilisomes and PS II remains efficient. Therefore, functional accommodation of a light-harvesting complex in the membrane versus a phycobilisome outside the membrane does not appear to require modifications in either the N-terminal 25% or C-terminal 50% o f the CP47 sequence. The results presented here emphasize that the properties o f the PS II reaction center complex are susceptible towards rather minor disturbances in the structure of the individual components. This illustrates the highly complex and evolved nature of the PS II reaction center complex, and protein engineering techniques coupled to mutant analysis provide a suitable approach to elucidate the role o f individual components and domains in PS II structure and function. The observation that heterologous recombination does occur in

Synechocystis 6803 with reasonable frequency paves the way toward gene replacement approaches aimed at investigating regions of particular functional relevance.

Acknowledgement This research is supported by the US Department of Energy (Grant #DE-FG03-95ER20180 to WV).

References Arnold W (1966) Light reaction in green plant photosynthesis: A method of study. Science 154:1046--1050 Arnold W and Oppenheimer R (1950) Internal conversion in the photosynthetic mechanism of blue-green algae. J Gen Physiol 33:423-435 Barber J (1992) The Photosystems: Structure, Function and Molecular Biology. Elsevier, Amsterdam Boekema EJ, Boonstra AF, Dekker JP and R6gner M (1994) Electron microscopic structural analysis of Photosystem I, Photosystern 1I and the cytochrome b61fcomplex from green plants and cyanobacteria. J Bioenerg Biomembr 26:17-29 Boekema EJ, Hankamer B, Bald D, Kruip J, Nield J, Boonstra AF, Barber J and R6gner M (1995) Supramolecular structure of the Photosystem II complex from green plants and cyanobacteria. Proc Natl Acad Sci USA 92:175-179 Bricker TM (1990) The structure and function of CPa-1 and CPa-2 in Photosystem II. Photosynth Res 24:1-13 Bowyer JR, Camilleri P and Vermaas WFJ (1991) Photosystem II and its interaction with herbicides. In: Baker NR and Percival MP (eds) Herbicides, pp 27-85. Elsevier, Amsterdam Burnap RL, Shen JR, Jursinic PA, Inoue Y and Sherman LA (1992) Oxygen yield and thermoluminescence characteristics of a cyanobacterium lacking the manganese-stabilizing protein of Photosystem II. Biochemistry 31:7404-7410 Carpenter SD, Ohad I and VermaasWFJ (1993) Analysis of chimeric spinach/cyanobacterialCP43 mutants of Synechocystissp. PCC 6803: The chlorophyll-protein CP43 affects the water-spliUing system of PhotosystemII. Biochim Biophys Acta 1144:204-212 Chu H-A, Nguyen AP and Debus RJ (1994) Site-directed Photosystem II mutants with pertu~,at oxygen-evolving properties. 1. Instability or inefficient assembly of the manganese cluster in vivo. Biochemistry 33:6137~149 Dekker JP, Bowlby NR and Yocum CF (1989) Chlorophyll and cytochrome b-559 content of the photochemical reaction center of Photosystem II. FEBS Lett 254:150-154 Gantt E (1988) Phycobilisomes: assessment of the core structure and thylakoid interaction. In: Stevens Jr. SE and Bryant DA (eds) Light-EnergyTransductionin Photosynthesis: Higher Plant and Bacterial Models, pp 91-101. American Society of Plant Physiologists,Rockville, MD GhanotakisDF, de PaulaJC, DemetriouDM, Bowlby NR, PetersenJ, Babcock GT and YocumCF (1989) Isolation and characterization of the 47 kDa protein and the D1-D2-cytochromeb-559 complex. Biochim Biophys Acta 974:44-53 Koike H and Inoue Y (1985) Properties of a peripheral 34 kDa protein in SynechococcusvulcanusPhotosystem II particles. Its

162 exchangeability with spinach 33 kDa protein in reconstitution of 02 evolution. Biochim Biophys Acta 807:64-73 Mayes SR, Cook KM, Self S J, Zhang ZH and Barber J (1991) Deletion of the gene encoding the Photosystem II 33 kDa protein from Synechocystis sp. PCC 6803 does not inactivate water-splitting but increases vulnerability to photoinhibition. Biochim Biophys Acta 1060:1-12 Nixon PJ, ROgner M and Diner BA (1991) Expression of a higher plant psbA gene in Synechocystis 6803 yields a functional hybrid Photosystem II reaction center complex. Plant Cell 3:383-395 Nixon P J, Chisholm DA and Diner BA (1992) Isolation and functional analysis of random and site-directed mutants of Photosystem II. In: Shewry P and Gutteridge S (eds) Plant Protein Engineering, pp 93-141. Cambridge University Press, Cambridge Philbrick JB, Diner BA and Zilinskas BA (1991) Construction and characterization of cyanobacterial mutants lacking the manganese-stabilizing potypeptide of Photosystem II. J Biol Chem 266:13370-13376 Prentki P and Krisch HM (1984) In vitro insertional mutagenesis with a selectable DNA fragment. Gene 29:303-313 Rippka R, Deruelles J, Waterbury JB, Herdman M and Stanier RY (1979) Generic assignments, strain histories and properties of pure cultures of cyanobacteria. J Gen Microbiol 111: 1-61 Shen G and Vermaas WFJ (1994) Mutation of chiorophyll ligands in the chlorophyll-binding CP47 protein as studied in a Synechocystis sp. PCC 6803 Photosystem I-less background. Biochemistry 33:7379-7388 Shen G, Eaton-Rye JJ and Vermaas WFJ (1993) Mutation ofhistidine residues in CP47 leads to destabilization of the Photosystem II complex and to impairment of light energy transfer. Biochemistry 32:5109-5115 Shen JR and Inoue Y (1993) Binding and functional properties of two new extrinsic components, cytochrome c-550 and a 12-kDa protein, in cyanobacterial Photosystem II. Biochemistry 32: 18251832 Shestakov SV and Reaston J (1987) Gene transfer and host-vector systems of cyanobacteria. Oxford Surv Plant Moi Cell Biol 4: 136-166

Thomber JP, Peter GF, Chitnis PR, Nechushtai R and Vainstein A (1988) The light-harvesting complex of Photosystem II of higher plants. In: Stevens Jr SE and Bryant DA (eds) Light Energy Transduction in Photosynthesis: Higher Plant and Bacterial Models, pp 137-154. American Society of Plant Physiologists, Rockville, MD Tischer W and Strotmann H (1977) Relationship between inhibitor binding by chloroplasts and inhibition of photosynthetic electron transport. Biochim Biophys Acta 460:113-125 Vass I and Inoue Y (1992) Thermoluminescence in the study of Photosystem II. In: Barber J (ed) The Photosystems: Structure, Function and Molecular Biology, pp 259-294. Elsevier, Amsterdam Vermaas W (1993) Molecular-biological apparoaches to analyze Photosystem II structure and function. Annu Rev Plant Physiol Plant Mol Biol 44:457-481 Vermaas WFJ, Williams JGK and Arntzen CJ (1987) Sequencing and modification of psbB, the gene encoding the CP-47 protein of Photosystem II, in the cyanobacterium Synechocystis 6803. Plant Mol Biol 8:317-326 Vermaas WFJ, Ikeuchi M and Inoue Y (1988) Protein composition of the Photosystem II core complex in genetically engineered mutants of the cyanobacterium Synechocystis sp. PCC 6803. Photosynth Res 17:97-113 Vermaas W, Charit6 J and Shen G (1990) Glu-69 of the D2 protein in Photosystem II is a potential ligand to Mn involved in photosynthetic oxygen evolution. Biochemistry 29:5325-5332 Vermaas WFJ, Styring S, Schr0der W and Andersson B (1993) Photosynthetic water oxidation: The protein framework. Photosynth Res 38:249-263 Williams JGK (1988) Construction of specific mutations in the Photosystem II photosynthetic reaction center by genetic engineering methods in the cyanobacterium Synechocystis 6803. Meth Enzymol 167:766-778 Yu J and Vermaas XVFJ (1990) Transcript levels and synthesis of Photosystem II components in cyanobacterial mutants with inactivated Photosystem II genes. Plant Cell 2:315-322